From 9a08c352d05305ca7651540c3b107da1e4e1f40b Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Wed, 17 Feb 2016 16:49:38 -0800 Subject: fs: add filp_clone_open API I need an API that allows me to obtain a clone of the current file pointer to pass in to an exec handler. I've labelled this as an internal API because I can't see how it would be useful outside of the fs subsystem. The use case will be a persistent binfmt_misc handler. Signed-off-by: James Bottomley Acked-by: Serge Hallyn Acked-by: Jan Kara --- fs/internal.h | 1 + fs/open.c | 20 ++++++++++++++++++++ 2 files changed, 21 insertions(+) diff --git a/fs/internal.h b/fs/internal.h index b71deeecea17..c8ca0c957743 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -108,6 +108,7 @@ extern long do_handle_open(int mountdirfd, struct file_handle __user *ufh, int open_flag); extern int open_check_o_direct(struct file *f); extern int vfs_open(const struct path *, struct file *, const struct cred *); +extern struct file *filp_clone_open(struct file *); /* * inode.c diff --git a/fs/open.c b/fs/open.c index 17cb6b1dab75..bfe6f2b8345f 100644 --- a/fs/open.c +++ b/fs/open.c @@ -1002,6 +1002,26 @@ struct file *file_open_root(struct dentry *dentry, struct vfsmount *mnt, } EXPORT_SYMBOL(file_open_root); +struct file *filp_clone_open(struct file *oldfile) +{ + struct file *file; + int retval; + + file = get_empty_filp(); + if (IS_ERR(file)) + return file; + + file->f_flags = oldfile->f_flags; + retval = vfs_open(&oldfile->f_path, file, oldfile->f_cred); + if (retval) { + put_filp(file); + return ERR_PTR(retval); + } + + return file; +} +EXPORT_SYMBOL(filp_clone_open); + long do_sys_open(int dfd, const char __user *filename, int flags, umode_t mode) { struct open_flags op; -- cgit v1.2.3 From 948b701a607f123df92ed29084413e5dd8cda2ed Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Wed, 17 Feb 2016 16:51:16 -0800 Subject: binfmt_misc: add persistent opened binary handler for containers This patch adds a new flag 'F' to the binfmt handlers. If you pass in 'F' the binary that runs the emulation will be opened immediately and in future, will be cloned from the open file. The net effect is that the handler survives both changeroots and mount namespace changes, making it easy to work with foreign architecture containers without contaminating the container image with the emulator. Signed-off-by: James Bottomley Acked-by: Serge Hallyn --- fs/binfmt_misc.c | 41 +++++++++++++++++++++++++++++++++++++++-- 1 file changed, 39 insertions(+), 2 deletions(-) diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index 3a3ced779fc7..8a108c435bc6 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -26,6 +26,8 @@ #include #include +#include "internal.h" + #ifdef DEBUG # define USE_DEBUG 1 #else @@ -43,6 +45,7 @@ enum {Enabled, Magic}; #define MISC_FMT_PRESERVE_ARGV0 (1 << 31) #define MISC_FMT_OPEN_BINARY (1 << 30) #define MISC_FMT_CREDENTIALS (1 << 29) +#define MISC_FMT_OPEN_FILE (1 << 28) typedef struct { struct list_head list; @@ -54,6 +57,7 @@ typedef struct { char *interpreter; /* filename of interpreter */ char *name; struct dentry *dentry; + struct file *interp_file; } Node; static DEFINE_RWLOCK(entries_lock); @@ -201,7 +205,13 @@ static int load_misc_binary(struct linux_binprm *bprm) if (retval < 0) goto error; - interp_file = open_exec(iname); + if (fmt->flags & MISC_FMT_OPEN_FILE && fmt->interp_file) { + interp_file = filp_clone_open(fmt->interp_file); + if (!IS_ERR(interp_file)) + deny_write_access(interp_file); + } else { + interp_file = open_exec(iname); + } retval = PTR_ERR(interp_file); if (IS_ERR(interp_file)) goto error; @@ -285,6 +295,11 @@ static char *check_special_flags(char *sfs, Node *e) e->flags |= (MISC_FMT_CREDENTIALS | MISC_FMT_OPEN_BINARY); break; + case 'F': + pr_debug("register: flag: F: open interpreter file now\n"); + p++; + e->flags |= MISC_FMT_OPEN_FILE; + break; default: cont = 0; } @@ -543,6 +558,8 @@ static void entry_status(Node *e, char *page) *dp++ = 'O'; if (e->flags & MISC_FMT_CREDENTIALS) *dp++ = 'C'; + if (e->flags & MISC_FMT_OPEN_FILE) + *dp++ = 'F'; *dp++ = '\n'; if (!test_bit(Magic, &e->flags)) { @@ -590,6 +607,11 @@ static void kill_node(Node *e) } write_unlock(&entries_lock); + if ((e->flags & MISC_FMT_OPEN_FILE) && e->interp_file) { + filp_close(e->interp_file, NULL); + e->interp_file = NULL; + } + if (dentry) { drop_nlink(d_inode(dentry)); d_drop(dentry); @@ -698,6 +720,21 @@ static ssize_t bm_register_write(struct file *file, const char __user *buffer, goto out2; } + if (e->flags & MISC_FMT_OPEN_FILE) { + struct file *f; + + f = open_exec(e->interpreter); + if (IS_ERR(f)) { + err = PTR_ERR(f); + pr_notice("register: failed to install interpreter file %s\n", e->interpreter); + simple_release_fs(&bm_mnt, &entry_count); + iput(inode); + inode = NULL; + goto out2; + } + e->interp_file = f; + } + e->dentry = dget(dentry); inode->i_private = e; inode->i_fop = &bm_entry_operations; @@ -716,7 +753,7 @@ out: if (err) { kfree(e); - return -EINVAL; + return err; } return count; } -- cgit v1.2.3 From 4af75df6a410ce76d9f60f27b07e5645ecc2c5ed Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 25 Feb 2016 08:32:51 -0800 Subject: binfmt_misc: add F option description to documentation Signed-off-by: James Bottomley --- Documentation/binfmt_misc.txt | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/Documentation/binfmt_misc.txt b/Documentation/binfmt_misc.txt index 6b1de7058371..ec83bbce547a 100644 --- a/Documentation/binfmt_misc.txt +++ b/Documentation/binfmt_misc.txt @@ -66,6 +66,13 @@ Here is what the fields mean: This feature should be used with care as the interpreter will run with root permissions when a setuid binary owned by root is run with binfmt_misc. + 'F' - fix binary. The usual behaviour of binfmt_misc is to spawn the + binary lazily when the misc format file is invoked. However, + this doesn't work very well in the face of mount namespaces and + changeroots, so the F mode opens the binary as soon as the + emulation is installed and uses the opened image to spawn the + emulator, meaning it is always available once installed, + regardless of how the environment changes. There are some restrictions: -- cgit v1.2.3 From d68778b80dd78ad6ed998397af182163d0ef16b0 Mon Sep 17 00:00:00 2001 From: Uwe Kleine-König Date: Wed, 11 May 2016 09:11:23 +0200 Subject: rtc: initialize output parameter for read alarm to "uninitialized" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit rtc drivers are supposed to set values they don't support to -1. To simplify this for drivers and also make it harder for them to get it wrong initialize the values to -1. Signed-off-by: Uwe Kleine-König Signed-off-by: Alexandre Belloni --- drivers/rtc/interface.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/drivers/rtc/interface.c b/drivers/rtc/interface.c index 9ef5f6f89f98..99475908e556 100644 --- a/drivers/rtc/interface.c +++ b/drivers/rtc/interface.c @@ -104,7 +104,17 @@ static int rtc_read_alarm_internal(struct rtc_device *rtc, struct rtc_wkalrm *al else if (!rtc->ops->read_alarm) err = -EINVAL; else { - memset(alarm, 0, sizeof(struct rtc_wkalrm)); + alarm->enabled = 0; + alarm->pending = 0; + alarm->time.tm_sec = -1; + alarm->time.tm_min = -1; + alarm->time.tm_hour = -1; + alarm->time.tm_mday = -1; + alarm->time.tm_mon = -1; + alarm->time.tm_year = -1; + alarm->time.tm_wday = -1; + alarm->time.tm_yday = -1; + alarm->time.tm_isdst = -1; err = rtc->ops->read_alarm(rtc->dev.parent, alarm); } -- cgit v1.2.3 From 7af6a2e1c2565396f6b941ff2a989b989b24a97b Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Sun, 22 May 2016 11:06:16 +0200 Subject: MAINTAINERS: Add file patterns for rtc device tree bindings Submitters of device tree binding documentation may forget to CC the subsystem maintainer if this is missing. Signed-off-by: Geert Uytterhoeven Cc: Alessandro Zummo Cc: Alexandre Belloni Cc: rtc-linux@googlegroups.com Signed-off-by: Alexandre Belloni --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 7304d2e37a98..72c28f4520d5 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9558,6 +9558,7 @@ L: rtc-linux@googlegroups.com Q: http://patchwork.ozlabs.org/project/rtc-linux/list/ T: git git://git.kernel.org/pub/scm/linux/kernel/git/abelloni/linux.git S: Maintained +F: Documentation/devicetree/bindings/rtc/ F: Documentation/rtc.txt F: drivers/rtc/ F: include/linux/rtc.h -- cgit v1.2.3 From f09c5142ee42efce956ba93f5a75ddaf4835c9e8 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:49 +0200 Subject: rtc: cmos: remove empty asm/mc146818rtc.h files Nothing on these architectures ever includes the asm/mc146818rtc.h file, the drivers that used to do this have been fixed long ago, and the remaining users are all PC-specific. This removes the files for good. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/frv/include/asm/mc146818rtc.h | 16 ---------------- arch/h8300/include/asm/mc146818rtc.h | 9 --------- arch/ia64/include/asm/mc146818rtc.h | 10 ---------- arch/parisc/include/asm/mc146818rtc.h | 9 --------- arch/sh/include/asm/mc146818rtc.h | 7 ------- 5 files changed, 51 deletions(-) delete mode 100644 arch/frv/include/asm/mc146818rtc.h delete mode 100644 arch/h8300/include/asm/mc146818rtc.h delete mode 100644 arch/ia64/include/asm/mc146818rtc.h delete mode 100644 arch/parisc/include/asm/mc146818rtc.h delete mode 100644 arch/sh/include/asm/mc146818rtc.h diff --git a/arch/frv/include/asm/mc146818rtc.h b/arch/frv/include/asm/mc146818rtc.h deleted file mode 100644 index 90dfb7a633d1..000000000000 --- a/arch/frv/include/asm/mc146818rtc.h +++ /dev/null @@ -1,16 +0,0 @@ -/* mc146818rtc.h: RTC defs - * - * Copyright (C) 2005 Red Hat, Inc. All Rights Reserved. - * Written by David Howells (dhowells@redhat.com) - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - */ - -#ifndef _ASM_MC146818RTC_H -#define _ASM_MC146818RTC_H - - -#endif /* _ASM_MC146818RTC_H */ diff --git a/arch/h8300/include/asm/mc146818rtc.h b/arch/h8300/include/asm/mc146818rtc.h deleted file mode 100644 index ab9d9646d241..000000000000 --- a/arch/h8300/include/asm/mc146818rtc.h +++ /dev/null @@ -1,9 +0,0 @@ -/* - * Machine dependent access functions for RTC registers. - */ -#ifndef _H8300_MC146818RTC_H -#define _H8300_MC146818RTC_H - -/* empty include file to satisfy the include in genrtc.c/ide-geometry.c */ - -#endif /* _H8300_MC146818RTC_H */ diff --git a/arch/ia64/include/asm/mc146818rtc.h b/arch/ia64/include/asm/mc146818rtc.h deleted file mode 100644 index 407787a237ba..000000000000 --- a/arch/ia64/include/asm/mc146818rtc.h +++ /dev/null @@ -1,10 +0,0 @@ -#ifndef _ASM_IA64_MC146818RTC_H -#define _ASM_IA64_MC146818RTC_H - -/* - * Machine dependent access functions for RTC registers. - */ - -/* empty include file to satisfy the include in genrtc.c */ - -#endif /* _ASM_IA64_MC146818RTC_H */ diff --git a/arch/parisc/include/asm/mc146818rtc.h b/arch/parisc/include/asm/mc146818rtc.h deleted file mode 100644 index adf41631449f..000000000000 --- a/arch/parisc/include/asm/mc146818rtc.h +++ /dev/null @@ -1,9 +0,0 @@ -/* - * Machine dependent access functions for RTC registers. - */ -#ifndef _ASM_MC146818RTC_H -#define _ASM_MC146818RTC_H - -/* empty include file to satisfy the include in genrtc.c */ - -#endif /* _ASM_MC146818RTC_H */ diff --git a/arch/sh/include/asm/mc146818rtc.h b/arch/sh/include/asm/mc146818rtc.h deleted file mode 100644 index 0aee96a97330..000000000000 --- a/arch/sh/include/asm/mc146818rtc.h +++ /dev/null @@ -1,7 +0,0 @@ -/* - * Machine dependent access functions for RTC registers. - */ -#ifndef _ASM_MC146818RTC_H -#define _ASM_MC146818RTC_H - -#endif /* _ASM_MC146818RTC_H */ -- cgit v1.2.3 From 5ab788d7383289bfc141ab357767bc6c11bbf77f Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:50 +0200 Subject: rtc: cmos: move mc146818rtc code out of asm-generic/rtc.h Drivers should not really include stuff from asm-generic directly, and the PC-style cmos rtc driver does this in order to reuse the mc146818 implementation of get_rtc_time/set_rtc_time rather than the architecture specific one for the architecture it gets built for. To make it more obvious what is going on, this moves and renames the two functions into include/linux/mc146818rtc.h, which holds the other mc146818 specific code. Ideally it would be in a .c file, but that would require extra infrastructure as the functions are called by multiple drivers with conflicting dependencies. With this change, the asm-generic/rtc.h header also becomes much more generic, so it can be reused more easily across any architecture that still relies on the genrtc driver. The only caller of the internal __get_rtc_time/__set_rtc_time functions is in arch/alpha/kernel/rtc.c, and we just change those over to the new naming. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/alpha/kernel/rtc.c | 6 +- arch/x86/include/asm/mc146818rtc.h | 1 - arch/x86/platform/efi/efi_64.c | 1 + arch/x86/platform/intel-mid/intel_mid_vrtc.c | 1 + drivers/rtc/rtc-cmos.c | 12 +- include/asm-generic/rtc.h | 206 +-------------------------- include/linux/mc146818rtc.h | 194 +++++++++++++++++++++++++ 7 files changed, 209 insertions(+), 212 deletions(-) diff --git a/arch/alpha/kernel/rtc.c b/arch/alpha/kernel/rtc.c index f535a3fd0f60..ceed68c7500b 100644 --- a/arch/alpha/kernel/rtc.c +++ b/arch/alpha/kernel/rtc.c @@ -15,8 +15,6 @@ #include #include -#include - #include "proto.h" @@ -81,7 +79,7 @@ init_rtc_epoch(void) static int alpha_rtc_read_time(struct device *dev, struct rtc_time *tm) { - __get_rtc_time(tm); + mc146818_get_time(tm); /* Adjust for non-default epochs. It's easier to depend on the generic __get_rtc_time and adjust the epoch here than create @@ -112,7 +110,7 @@ alpha_rtc_set_time(struct device *dev, struct rtc_time *tm) tm = &xtm; } - return __set_rtc_time(tm); + return mc146818_set_time(tm); } static int diff --git a/arch/x86/include/asm/mc146818rtc.h b/arch/x86/include/asm/mc146818rtc.h index 0f555cc31984..24acd9ba7837 100644 --- a/arch/x86/include/asm/mc146818rtc.h +++ b/arch/x86/include/asm/mc146818rtc.h @@ -6,7 +6,6 @@ #include #include -#include #ifndef RTC_PORT #define RTC_PORT(x) (0x70 + (x)) diff --git a/arch/x86/platform/efi/efi_64.c b/arch/x86/platform/efi/efi_64.c index 6e7242be1c87..663c282801f3 100644 --- a/arch/x86/platform/efi/efi_64.c +++ b/arch/x86/platform/efi/efi_64.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include #include diff --git a/arch/x86/platform/intel-mid/intel_mid_vrtc.c b/arch/x86/platform/intel-mid/intel_mid_vrtc.c index ee40fcb6e54d..58024862a7eb 100644 --- a/arch/x86/platform/intel-mid/intel_mid_vrtc.c +++ b/arch/x86/platform/intel-mid/intel_mid_vrtc.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include diff --git a/drivers/rtc/rtc-cmos.c b/drivers/rtc/rtc-cmos.c index fbe9c72438e1..cf8eb98382ce 100644 --- a/drivers/rtc/rtc-cmos.c +++ b/drivers/rtc/rtc-cmos.c @@ -43,7 +43,7 @@ #include /* this is for "generic access to PC-style RTC" using CMOS_READ/CMOS_WRITE */ -#include +#include struct cmos_rtc { struct rtc_device *rtc; @@ -190,10 +190,10 @@ static inline void cmos_write_bank2(unsigned char val, unsigned char addr) static int cmos_read_time(struct device *dev, struct rtc_time *t) { /* REVISIT: if the clock has a "century" register, use - * that instead of the heuristic in get_rtc_time(). + * that instead of the heuristic in mc146818_get_time(). * That'll make Y3K compatility (year > 2070) easy! */ - get_rtc_time(t); + mc146818_get_time(t); return 0; } @@ -205,7 +205,7 @@ static int cmos_set_time(struct device *dev, struct rtc_time *t) * takes effect exactly 500ms after we write the register. * (Also queueing and other delays before we get this far.) */ - return set_rtc_time(t); + return mc146818_set_time(t); } static int cmos_read_alarm(struct device *dev, struct rtc_wkalrm *t) @@ -1142,14 +1142,14 @@ static __init void cmos_of_init(struct platform_device *pdev) if (val) CMOS_WRITE(be32_to_cpup(val), RTC_FREQ_SELECT); - get_rtc_time(&time); + cmos_read_time(&pdev->dev, &time); ret = rtc_valid_tm(&time); if (ret) { struct rtc_time def_time = { .tm_year = 1, .tm_mday = 1, }; - set_rtc_time(&def_time); + cmos_set_time(&pdev->dev, &def_time); } } #else diff --git a/include/asm-generic/rtc.h b/include/asm-generic/rtc.h index 4e3b6558331e..4fcff22cd707 100644 --- a/include/asm-generic/rtc.h +++ b/include/asm-generic/rtc.h @@ -12,12 +12,12 @@ #ifndef __ASM_RTC_H__ #define __ASM_RTC_H__ -#include #include -#include -#include -#ifdef CONFIG_ACPI -#include + +#ifndef get_rtc_time +#include +#define get_rtc_time mc146818_get_time +#define set_rtc_time mc146818_set_time #endif #define RTC_PIE 0x40 /* periodic interrupt enable */ @@ -31,202 +31,6 @@ #define RTC_24H 0x02 /* 24 hour mode - else hours bit 7 means pm */ #define RTC_DST_EN 0x01 /* auto switch DST - works f. USA only */ -/* - * Returns true if a clock update is in progress - */ -static inline unsigned char rtc_is_updating(void) -{ - unsigned char uip; - unsigned long flags; - - spin_lock_irqsave(&rtc_lock, flags); - uip = (CMOS_READ(RTC_FREQ_SELECT) & RTC_UIP); - spin_unlock_irqrestore(&rtc_lock, flags); - return uip; -} - -static inline unsigned int __get_rtc_time(struct rtc_time *time) -{ - unsigned char ctrl; - unsigned long flags; - unsigned char century = 0; - -#ifdef CONFIG_MACH_DECSTATION - unsigned int real_year; -#endif - - /* - * read RTC once any update in progress is done. The update - * can take just over 2ms. We wait 20ms. There is no need to - * to poll-wait (up to 1s - eeccch) for the falling edge of RTC_UIP. - * If you need to know *exactly* when a second has started, enable - * periodic update complete interrupts, (via ioctl) and then - * immediately read /dev/rtc which will block until you get the IRQ. - * Once the read clears, read the RTC time (again via ioctl). Easy. - */ - if (rtc_is_updating()) - mdelay(20); - - /* - * Only the values that we read from the RTC are set. We leave - * tm_wday, tm_yday and tm_isdst untouched. Even though the - * RTC has RTC_DAY_OF_WEEK, we ignore it, as it is only updated - * by the RTC when initially set to a non-zero value. - */ - spin_lock_irqsave(&rtc_lock, flags); - time->tm_sec = CMOS_READ(RTC_SECONDS); - time->tm_min = CMOS_READ(RTC_MINUTES); - time->tm_hour = CMOS_READ(RTC_HOURS); - time->tm_mday = CMOS_READ(RTC_DAY_OF_MONTH); - time->tm_mon = CMOS_READ(RTC_MONTH); - time->tm_year = CMOS_READ(RTC_YEAR); -#ifdef CONFIG_MACH_DECSTATION - real_year = CMOS_READ(RTC_DEC_YEAR); -#endif -#ifdef CONFIG_ACPI - if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && - acpi_gbl_FADT.century) - century = CMOS_READ(acpi_gbl_FADT.century); -#endif - ctrl = CMOS_READ(RTC_CONTROL); - spin_unlock_irqrestore(&rtc_lock, flags); - - if (!(ctrl & RTC_DM_BINARY) || RTC_ALWAYS_BCD) - { - time->tm_sec = bcd2bin(time->tm_sec); - time->tm_min = bcd2bin(time->tm_min); - time->tm_hour = bcd2bin(time->tm_hour); - time->tm_mday = bcd2bin(time->tm_mday); - time->tm_mon = bcd2bin(time->tm_mon); - time->tm_year = bcd2bin(time->tm_year); - century = bcd2bin(century); - } - -#ifdef CONFIG_MACH_DECSTATION - time->tm_year += real_year - 72; -#endif - - if (century) - time->tm_year += (century - 19) * 100; - - /* - * Account for differences between how the RTC uses the values - * and how they are defined in a struct rtc_time; - */ - if (time->tm_year <= 69) - time->tm_year += 100; - - time->tm_mon--; - - return RTC_24H; -} - -#ifndef get_rtc_time -#define get_rtc_time __get_rtc_time -#endif - -/* Set the current date and time in the real time clock. */ -static inline int __set_rtc_time(struct rtc_time *time) -{ - unsigned long flags; - unsigned char mon, day, hrs, min, sec; - unsigned char save_control, save_freq_select; - unsigned int yrs; -#ifdef CONFIG_MACH_DECSTATION - unsigned int real_yrs, leap_yr; -#endif - unsigned char century = 0; - - yrs = time->tm_year; - mon = time->tm_mon + 1; /* tm_mon starts at zero */ - day = time->tm_mday; - hrs = time->tm_hour; - min = time->tm_min; - sec = time->tm_sec; - - if (yrs > 255) /* They are unsigned */ - return -EINVAL; - - spin_lock_irqsave(&rtc_lock, flags); -#ifdef CONFIG_MACH_DECSTATION - real_yrs = yrs; - leap_yr = ((!((yrs + 1900) % 4) && ((yrs + 1900) % 100)) || - !((yrs + 1900) % 400)); - yrs = 72; - - /* - * We want to keep the year set to 73 until March - * for non-leap years, so that Feb, 29th is handled - * correctly. - */ - if (!leap_yr && mon < 3) { - real_yrs--; - yrs = 73; - } -#endif - -#ifdef CONFIG_ACPI - if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && - acpi_gbl_FADT.century) { - century = (yrs + 1900) / 100; - yrs %= 100; - } -#endif - - /* These limits and adjustments are independent of - * whether the chip is in binary mode or not. - */ - if (yrs > 169) { - spin_unlock_irqrestore(&rtc_lock, flags); - return -EINVAL; - } - - if (yrs >= 100) - yrs -= 100; - - if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) - || RTC_ALWAYS_BCD) { - sec = bin2bcd(sec); - min = bin2bcd(min); - hrs = bin2bcd(hrs); - day = bin2bcd(day); - mon = bin2bcd(mon); - yrs = bin2bcd(yrs); - century = bin2bcd(century); - } - - save_control = CMOS_READ(RTC_CONTROL); - CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL); - save_freq_select = CMOS_READ(RTC_FREQ_SELECT); - CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT); - -#ifdef CONFIG_MACH_DECSTATION - CMOS_WRITE(real_yrs, RTC_DEC_YEAR); -#endif - CMOS_WRITE(yrs, RTC_YEAR); - CMOS_WRITE(mon, RTC_MONTH); - CMOS_WRITE(day, RTC_DAY_OF_MONTH); - CMOS_WRITE(hrs, RTC_HOURS); - CMOS_WRITE(min, RTC_MINUTES); - CMOS_WRITE(sec, RTC_SECONDS); -#ifdef CONFIG_ACPI - if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && - acpi_gbl_FADT.century) - CMOS_WRITE(century, acpi_gbl_FADT.century); -#endif - - CMOS_WRITE(save_control, RTC_CONTROL); - CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); - - spin_unlock_irqrestore(&rtc_lock, flags); - - return 0; -} - -#ifndef set_rtc_time -#define set_rtc_time __set_rtc_time -#endif - static inline unsigned int get_rtc_ss(void) { struct rtc_time h; diff --git a/include/linux/mc146818rtc.h b/include/linux/mc146818rtc.h index 433e0c74d643..e9e346b37846 100644 --- a/include/linux/mc146818rtc.h +++ b/include/linux/mc146818rtc.h @@ -14,6 +14,12 @@ #include #include /* get the user-level API */ #include /* register access macros */ +#include +#include + +#ifdef CONFIG_ACPI +#include +#endif #ifdef __KERNEL__ #include /* spinlock_t */ @@ -120,4 +126,192 @@ struct cmos_rtc_board_info { #define RTC_IO_EXTENT_USED RTC_IO_EXTENT #endif /* ARCH_RTC_LOCATION */ +/* + * Returns true if a clock update is in progress + */ +static inline unsigned char mc146818_is_updating(void) +{ + unsigned char uip; + unsigned long flags; + + spin_lock_irqsave(&rtc_lock, flags); + uip = (CMOS_READ(RTC_FREQ_SELECT) & RTC_UIP); + spin_unlock_irqrestore(&rtc_lock, flags); + return uip; +} + +static inline unsigned int mc146818_get_time(struct rtc_time *time) +{ + unsigned char ctrl; + unsigned long flags; + unsigned char century = 0; + +#ifdef CONFIG_MACH_DECSTATION + unsigned int real_year; +#endif + + /* + * read RTC once any update in progress is done. The update + * can take just over 2ms. We wait 20ms. There is no need to + * to poll-wait (up to 1s - eeccch) for the falling edge of RTC_UIP. + * If you need to know *exactly* when a second has started, enable + * periodic update complete interrupts, (via ioctl) and then + * immediately read /dev/rtc which will block until you get the IRQ. + * Once the read clears, read the RTC time (again via ioctl). Easy. + */ + if (mc146818_is_updating()) + mdelay(20); + + /* + * Only the values that we read from the RTC are set. We leave + * tm_wday, tm_yday and tm_isdst untouched. Even though the + * RTC has RTC_DAY_OF_WEEK, we ignore it, as it is only updated + * by the RTC when initially set to a non-zero value. + */ + spin_lock_irqsave(&rtc_lock, flags); + time->tm_sec = CMOS_READ(RTC_SECONDS); + time->tm_min = CMOS_READ(RTC_MINUTES); + time->tm_hour = CMOS_READ(RTC_HOURS); + time->tm_mday = CMOS_READ(RTC_DAY_OF_MONTH); + time->tm_mon = CMOS_READ(RTC_MONTH); + time->tm_year = CMOS_READ(RTC_YEAR); +#ifdef CONFIG_MACH_DECSTATION + real_year = CMOS_READ(RTC_DEC_YEAR); +#endif +#ifdef CONFIG_ACPI + if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && + acpi_gbl_FADT.century) + century = CMOS_READ(acpi_gbl_FADT.century); +#endif + ctrl = CMOS_READ(RTC_CONTROL); + spin_unlock_irqrestore(&rtc_lock, flags); + + if (!(ctrl & RTC_DM_BINARY) || RTC_ALWAYS_BCD) + { + time->tm_sec = bcd2bin(time->tm_sec); + time->tm_min = bcd2bin(time->tm_min); + time->tm_hour = bcd2bin(time->tm_hour); + time->tm_mday = bcd2bin(time->tm_mday); + time->tm_mon = bcd2bin(time->tm_mon); + time->tm_year = bcd2bin(time->tm_year); + century = bcd2bin(century); + } + +#ifdef CONFIG_MACH_DECSTATION + time->tm_year += real_year - 72; +#endif + + if (century) + time->tm_year += (century - 19) * 100; + + /* + * Account for differences between how the RTC uses the values + * and how they are defined in a struct rtc_time; + */ + if (time->tm_year <= 69) + time->tm_year += 100; + + time->tm_mon--; + + return RTC_24H; +} + +/* Set the current date and time in the real time clock. */ +static inline int mc146818_set_time(struct rtc_time *time) +{ + unsigned long flags; + unsigned char mon, day, hrs, min, sec; + unsigned char save_control, save_freq_select; + unsigned int yrs; +#ifdef CONFIG_MACH_DECSTATION + unsigned int real_yrs, leap_yr; +#endif + unsigned char century = 0; + + yrs = time->tm_year; + mon = time->tm_mon + 1; /* tm_mon starts at zero */ + day = time->tm_mday; + hrs = time->tm_hour; + min = time->tm_min; + sec = time->tm_sec; + + if (yrs > 255) /* They are unsigned */ + return -EINVAL; + + spin_lock_irqsave(&rtc_lock, flags); +#ifdef CONFIG_MACH_DECSTATION + real_yrs = yrs; + leap_yr = ((!((yrs + 1900) % 4) && ((yrs + 1900) % 100)) || + !((yrs + 1900) % 400)); + yrs = 72; + + /* + * We want to keep the year set to 73 until March + * for non-leap years, so that Feb, 29th is handled + * correctly. + */ + if (!leap_yr && mon < 3) { + real_yrs--; + yrs = 73; + } +#endif + +#ifdef CONFIG_ACPI + if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && + acpi_gbl_FADT.century) { + century = (yrs + 1900) / 100; + yrs %= 100; + } +#endif + + /* These limits and adjustments are independent of + * whether the chip is in binary mode or not. + */ + if (yrs > 169) { + spin_unlock_irqrestore(&rtc_lock, flags); + return -EINVAL; + } + + if (yrs >= 100) + yrs -= 100; + + if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) + || RTC_ALWAYS_BCD) { + sec = bin2bcd(sec); + min = bin2bcd(min); + hrs = bin2bcd(hrs); + day = bin2bcd(day); + mon = bin2bcd(mon); + yrs = bin2bcd(yrs); + century = bin2bcd(century); + } + + save_control = CMOS_READ(RTC_CONTROL); + CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL); + save_freq_select = CMOS_READ(RTC_FREQ_SELECT); + CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT); + +#ifdef CONFIG_MACH_DECSTATION + CMOS_WRITE(real_yrs, RTC_DEC_YEAR); +#endif + CMOS_WRITE(yrs, RTC_YEAR); + CMOS_WRITE(mon, RTC_MONTH); + CMOS_WRITE(day, RTC_DAY_OF_MONTH); + CMOS_WRITE(hrs, RTC_HOURS); + CMOS_WRITE(min, RTC_MINUTES); + CMOS_WRITE(sec, RTC_SECONDS); +#ifdef CONFIG_ACPI + if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && + acpi_gbl_FADT.century) + CMOS_WRITE(century, acpi_gbl_FADT.century); +#endif + + CMOS_WRITE(save_control, RTC_CONTROL); + CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); + + spin_unlock_irqrestore(&rtc_lock, flags); + + return 0; +} + #endif /* _MC146818RTC_H */ -- cgit v1.2.3 From 463a86304cae92e10277b47180ac59cf93982e5b Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:51 +0200 Subject: char/genrtc: x86: remove remnants of asm/rtc.h Commit 3195ef59cb42 ("x86: Do full rtc synchronization with ntp") had the side-effect of unconditionally enabling the RTC_LIB symbol on x86, which in turn disables the selection of the CONFIG_RTC and CONFIG_GEN_RTC drivers that contain a two older implementations of the CONFIG_RTC_DRV_CMOS driver. This removes x86 from the list for genrtc, and changes all references to the asm/rtc.h header to instead point to the interfaces from linux/mc146818rtc.h. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/x86/include/asm/rtc.h | 1 - arch/x86/kernel/hpet.c | 3 +-- arch/x86/kernel/rtc.c | 3 +-- arch/x86/platform/efi/efi.c | 1 - drivers/acpi/acpi_cmos_rtc.c | 2 +- drivers/base/power/trace.c | 6 +++--- drivers/char/Kconfig | 2 +- drivers/rtc/rtc-mrst.c | 2 +- 8 files changed, 8 insertions(+), 12 deletions(-) delete mode 100644 arch/x86/include/asm/rtc.h diff --git a/arch/x86/include/asm/rtc.h b/arch/x86/include/asm/rtc.h deleted file mode 100644 index f71c3b0ed360..000000000000 --- a/arch/x86/include/asm/rtc.h +++ /dev/null @@ -1 +0,0 @@ -#include diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index f112af7aa62e..357900aad62c 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -1020,7 +1020,6 @@ void hpet_disable(void) */ #include #include -#include #define DEFAULT_RTC_INT_FREQ 64 #define DEFAULT_RTC_SHIFT 6 @@ -1244,7 +1243,7 @@ irqreturn_t hpet_rtc_interrupt(int irq, void *dev_id) memset(&curr_time, 0, sizeof(struct rtc_time)); if (hpet_rtc_flags & (RTC_UIE | RTC_AIE)) - get_rtc_time(&curr_time); + mc146818_set_time(&curr_time); if (hpet_rtc_flags & RTC_UIE && curr_time.tm_sec != hpet_prev_update_sec) { diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c index eceaa082ec3f..79c6311cd912 100644 --- a/arch/x86/kernel/rtc.c +++ b/arch/x86/kernel/rtc.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #ifdef CONFIG_X86_32 @@ -47,7 +46,7 @@ int mach_set_rtc_mmss(const struct timespec *now) rtc_time_to_tm(nowtime, &tm); if (!rtc_valid_tm(&tm)) { - retval = set_rtc_time(&tm); + retval = mc146818_set_time(&tm); if (retval) printk(KERN_ERR "%s: RTC write failed with error %d\n", __func__, retval); diff --git a/arch/x86/platform/efi/efi.c b/arch/x86/platform/efi/efi.c index f93545e7dc54..56400d03409f 100644 --- a/arch/x86/platform/efi/efi.c +++ b/arch/x86/platform/efi/efi.c @@ -51,7 +51,6 @@ #include #include #include -#include #include static struct efi efi_phys __initdata; diff --git a/drivers/acpi/acpi_cmos_rtc.c b/drivers/acpi/acpi_cmos_rtc.c index 81dc75033f15..0980a133916f 100644 --- a/drivers/acpi/acpi_cmos_rtc.c +++ b/drivers/acpi/acpi_cmos_rtc.c @@ -14,7 +14,7 @@ #include #include #include -#include +#include #include "internal.h" diff --git a/drivers/base/power/trace.c b/drivers/base/power/trace.c index a6975795e7f3..efec10b49d59 100644 --- a/drivers/base/power/trace.c +++ b/drivers/base/power/trace.c @@ -11,7 +11,7 @@ #include #include -#include +#include #include "power.h" @@ -103,7 +103,7 @@ static int set_magic_time(unsigned int user, unsigned int file, unsigned int dev n /= 24; time.tm_min = (n % 20) * 3; n /= 20; - set_rtc_time(&time); + mc146818_set_time(&time); return n ? -1 : 0; } @@ -112,7 +112,7 @@ static unsigned int read_magic_time(void) struct rtc_time time; unsigned int val; - get_rtc_time(&time); + mc146818_get_time(&time); pr_info("RTC time: %2d:%02d:%02d, date: %02d/%02d/%02d\n", time.tm_hour, time.tm_min, time.tm_sec, time.tm_mon + 1, time.tm_mday, time.tm_year % 100); diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 601f64fcc890..31b5861649ee 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -328,7 +328,7 @@ config JS_RTC config GEN_RTC tristate "Generic /dev/rtc emulation" depends on RTC!=y - depends on ALPHA || M68K || MN10300 || PARISC || PPC || X86 + depends on ALPHA || M68K || MN10300 || PARISC || PPC ---help--- If you say Y here and create a character special file /dev/rtc with major number 10 and minor number 135 using mknod ("man mknod"), you diff --git a/drivers/rtc/rtc-mrst.c b/drivers/rtc/rtc-mrst.c index 0094d9bdd1e6..fe840c0edfef 100644 --- a/drivers/rtc/rtc-mrst.c +++ b/drivers/rtc/rtc-mrst.c @@ -32,11 +32,11 @@ #include #include #include +#include #include #include #include -#include #include #include #include -- cgit v1.2.3 From d4db68741d381b04d603bc9e75fe2a8a29097a84 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:52 +0200 Subject: rtc: sh: provide rtc_class_ops directly The rtc-generic driver provides an architecture specific wrapper on top of the generic rtc_class_ops abstraction, and on sh, that goes through another indirection using the rtc_sh_get_time/rtc_sh_set_time functions. This changes the sh rtc-generic device to provide its rtc_class_ops directly, skipping one of the abstraction levels. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Acked-by: Rich Felker Signed-off-by: Alexandre Belloni --- arch/sh/include/asm/rtc.h | 11 ----------- arch/sh/kernel/time.c | 33 ++++++++++++++++++++------------- drivers/rtc/rtc-generic.c | 2 +- 3 files changed, 21 insertions(+), 25 deletions(-) diff --git a/arch/sh/include/asm/rtc.h b/arch/sh/include/asm/rtc.h index 52b0c2dba979..f7b010d48af7 100644 --- a/arch/sh/include/asm/rtc.h +++ b/arch/sh/include/asm/rtc.h @@ -6,17 +6,6 @@ extern void (*board_time_init)(void); extern void (*rtc_sh_get_time)(struct timespec *); extern int (*rtc_sh_set_time)(const time_t); -/* some dummy definitions */ -#define RTC_BATT_BAD 0x100 /* battery bad */ -#define RTC_SQWE 0x08 /* enable square-wave output */ -#define RTC_DM_BINARY 0x04 /* all time/date values are BCD if clear */ -#define RTC_24H 0x02 /* 24 hour mode - else hours bit 7 means pm */ -#define RTC_DST_EN 0x01 /* auto switch DST - works f. USA only */ - -struct rtc_time; -unsigned int get_rtc_time(struct rtc_time *); -int set_rtc_time(struct rtc_time *); - #define RTC_CAP_4_DIGIT_YEAR (1 << 0) struct sh_rtc_platform_info { diff --git a/arch/sh/kernel/time.c b/arch/sh/kernel/time.c index d6d0a986c6e9..a4a7862b489a 100644 --- a/arch/sh/kernel/time.c +++ b/arch/sh/kernel/time.c @@ -50,27 +50,31 @@ int update_persistent_clock(struct timespec now) } #endif -unsigned int get_rtc_time(struct rtc_time *tm) +static int rtc_generic_get_time(struct device *dev, struct rtc_time *tm) { - if (rtc_sh_get_time != null_rtc_get_time) { - struct timespec tv; + struct timespec tv; - rtc_sh_get_time(&tv); - rtc_time_to_tm(tv.tv_sec, tm); - } - - return RTC_24H; + rtc_sh_get_time(&tv); + rtc_time_to_tm(tv.tv_sec, tm); + return 0; } -EXPORT_SYMBOL(get_rtc_time); -int set_rtc_time(struct rtc_time *tm) +static int rtc_generic_set_time(struct device *dev, struct rtc_time *tm) { unsigned long secs; rtc_tm_to_time(tm, &secs); - return rtc_sh_set_time(secs); + if ((rtc_sh_set_time == null_rtc_set_time) || + (rtc_sh_set_time(secs) < 0)) + return -EOPNOTSUPP; + + return 0; } -EXPORT_SYMBOL(set_rtc_time); + +static const struct rtc_class_ops rtc_generic_ops = { + .read_time = rtc_generic_get_time, + .set_time = rtc_generic_set_time, +}; static int __init rtc_generic_init(void) { @@ -79,7 +83,10 @@ static int __init rtc_generic_init(void) if (rtc_sh_get_time == null_rtc_get_time) return -ENODEV; - pdev = platform_device_register_simple("rtc-generic", -1, NULL, 0); + pdev = platform_device_register_data(NULL, "rtc-generic", -1, + &rtc_generic_ops, + sizeof(rtc_generic_ops)); + return PTR_ERR_OR_ZERO(pdev); } diff --git a/drivers/rtc/rtc-generic.c b/drivers/rtc/rtc-generic.c index d726c6aa96a8..3958e87a05fa 100644 --- a/drivers/rtc/rtc-generic.c +++ b/drivers/rtc/rtc-generic.c @@ -10,7 +10,7 @@ #include #if defined(CONFIG_M68K) || defined(CONFIG_PARISC) || \ - defined(CONFIG_PPC) || defined(CONFIG_SUPERH32) + defined(CONFIG_PPC) #include static int generic_get_time(struct device *dev, struct rtc_time *tm) -- cgit v1.2.3 From 1cb1e35327b0cb95f33ab39365bc745e8399f384 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:53 +0200 Subject: char/genrtc: remove alpha support The genrtc driver serves no purpose on Alpha because it drives the same hardware as the original rtc.c driver, and the newer rtc-generic.c or rtc-cmos.c drivers on architectures that use the asm-generic/rtc.h header. The defconfig uses CONFIG_RTC=y, so this driver is not used by default. At one point it was used to abstract a quirk for the "Marvel" platform, but it does not do this any more after the code was moved into yet another driver in arch/alpha/kernel/rtc.c. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/alpha/include/asm/rtc.h | 1 - arch/alpha/kernel/core_marvel.c | 1 - drivers/char/Kconfig | 2 +- 3 files changed, 1 insertion(+), 3 deletions(-) delete mode 100644 arch/alpha/include/asm/rtc.h diff --git a/arch/alpha/include/asm/rtc.h b/arch/alpha/include/asm/rtc.h deleted file mode 100644 index f71c3b0ed360..000000000000 --- a/arch/alpha/include/asm/rtc.h +++ /dev/null @@ -1 +0,0 @@ -#include diff --git a/arch/alpha/kernel/core_marvel.c b/arch/alpha/kernel/core_marvel.c index 53dd2f1a53aa..d5f0580746a5 100644 --- a/arch/alpha/kernel/core_marvel.c +++ b/arch/alpha/kernel/core_marvel.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include "proto.h" diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 31b5861649ee..e3f2f3e293e5 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -328,7 +328,7 @@ config JS_RTC config GEN_RTC tristate "Generic /dev/rtc emulation" depends on RTC!=y - depends on ALPHA || M68K || MN10300 || PARISC || PPC + depends on M68K || MN10300 || PARISC || PPC ---help--- If you say Y here and create a character special file /dev/rtc with major number 10 and minor number 135 using mknod ("man mknod"), you -- cgit v1.2.3 From 7ee942179fd8d485cedaf1875f81c5ee7a3ebeae Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:54 +0200 Subject: char/genrtc: remove mn10300 support The genrtc driver serves no purpose on mn10300 because it drives the same hardware as the original rtc.c driver, and the newer rtc-generic.c or rtc-cmos.c drivers on architectures that use the asm-generic/rtc.h header. I assume it was initially only added for completeness when the mn10300 port was done, but the older rtc.c driver was always used instead. We can also stop include asm-generic/rtc.h now, because we just call mc146818_set_time() directly. It would be nice to change the architecture to use the rtc-cmos driver next, and remove support for the old rtc driver as well. [linux@roeck-us.net: Add missing include file to proc-init.c] Signed-off-by: Arnd Bergmann Signed-off-by: Guenter Roeck Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/mn10300/include/asm/rtc.h | 2 -- arch/mn10300/kernel/rtc.c | 2 +- arch/mn10300/proc-mn103e010/proc-init.c | 2 ++ arch/mn10300/proc-mn2ws0050/proc-init.c | 1 + drivers/char/Kconfig | 2 +- 5 files changed, 5 insertions(+), 4 deletions(-) diff --git a/arch/mn10300/include/asm/rtc.h b/arch/mn10300/include/asm/rtc.h index 6c14bb1d0d9b..07dc87656197 100644 --- a/arch/mn10300/include/asm/rtc.h +++ b/arch/mn10300/include/asm/rtc.h @@ -25,6 +25,4 @@ static inline void calibrate_clock(void) #endif /* !CONFIG_MN10300_RTC */ -#include - #endif /* _ASM_RTC_H */ diff --git a/arch/mn10300/kernel/rtc.c b/arch/mn10300/kernel/rtc.c index 48d7058b3295..77e0be4d92ea 100644 --- a/arch/mn10300/kernel/rtc.c +++ b/arch/mn10300/kernel/rtc.c @@ -27,7 +27,7 @@ void read_persistent_clock(struct timespec *ts) { struct rtc_time tm; - get_rtc_time(&tm); + mc146818_set_time(&tm); ts->tv_nsec = 0; ts->tv_sec = mktime(tm.tm_year, tm.tm_mon, tm.tm_mday, diff --git a/arch/mn10300/proc-mn103e010/proc-init.c b/arch/mn10300/proc-mn103e010/proc-init.c index 27b97980dca4..94641c420b3f 100644 --- a/arch/mn10300/proc-mn103e010/proc-init.c +++ b/arch/mn10300/proc-mn103e010/proc-init.c @@ -9,7 +9,9 @@ * 2 of the Licence, or (at your option) any later version. */ #include +#include #include +#include #include #include diff --git a/arch/mn10300/proc-mn2ws0050/proc-init.c b/arch/mn10300/proc-mn2ws0050/proc-init.c index ee6d03dbc8d8..950cc8dbb284 100644 --- a/arch/mn10300/proc-mn2ws0050/proc-init.c +++ b/arch/mn10300/proc-mn2ws0050/proc-init.c @@ -14,6 +14,7 @@ #include #include +#include #include #include #include diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index e3f2f3e293e5..1d9e3063373b 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -328,7 +328,7 @@ config JS_RTC config GEN_RTC tristate "Generic /dev/rtc emulation" depends on RTC!=y - depends on M68K || MN10300 || PARISC || PPC + depends on M68K || PARISC || PPC ---help--- If you say Y here and create a character special file /dev/rtc with major number 10 and minor number 135 using mknod ("man mknod"), you -- cgit v1.2.3 From ca6da801878635bfb851088e1a4eaa3745720582 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:55 +0200 Subject: rtc: parisc: provide rtc_class_ops directly The rtc-generic driver provides an architecture specific wrapper on top of the generic rtc_class_ops abstraction, and on pa-risc, that is implemented using an open-coded version of rtc_time_to_tm/rtc_tm_to_time. This changes the parisc rtc-generic device to provide its rtc_class_ops directly, using the normal helper functions, which makes this y2038 safe (on 32-bit) and simplifies the implementation. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/parisc/kernel/time.c | 36 +++++++++++++++++++++++++++++++++++- drivers/rtc/rtc-generic.c | 3 +-- 2 files changed, 36 insertions(+), 3 deletions(-) diff --git a/arch/parisc/kernel/time.c b/arch/parisc/kernel/time.c index 58dd6801f5be..744878789752 100644 --- a/arch/parisc/kernel/time.c +++ b/arch/parisc/kernel/time.c @@ -12,6 +12,7 @@ */ #include #include +#include #include #include #include @@ -248,14 +249,47 @@ void __init start_cpu_itimer(void) per_cpu(cpu_data, cpu).it_value = next_tick; } +#if IS_ENABLED(CONFIG_RTC_DRV_GENERIC) +static int rtc_generic_get_time(struct device *dev, struct rtc_time *tm) +{ + struct pdc_tod tod_data; + + memset(tm, 0, sizeof(*tm)); + if (pdc_tod_read(&tod_data) < 0) + return -EOPNOTSUPP; + + /* we treat tod_sec as unsigned, so this can work until year 2106 */ + rtc_time64_to_tm(tod_data.tod_sec, tm); + return rtc_valid_tm(tm); +} + +static int rtc_generic_set_time(struct device *dev, struct rtc_time *tm) +{ + time64_t secs = rtc_tm_to_time64(tm); + + if (pdc_tod_set(secs, 0) < 0) + return -EOPNOTSUPP; + + return 0; +} + +static const struct rtc_class_ops rtc_generic_ops = { + .read_time = rtc_generic_get_time, + .set_time = rtc_generic_set_time, +}; + static int __init rtc_init(void) { struct platform_device *pdev; - pdev = platform_device_register_simple("rtc-generic", -1, NULL, 0); + pdev = platform_device_register_data(NULL, "rtc-generic", -1, + &rtc_generic_ops, + sizeof(rtc_generic_ops)); + return PTR_ERR_OR_ZERO(pdev); } device_initcall(rtc_init); +#endif void read_persistent_clock(struct timespec *ts) { diff --git a/drivers/rtc/rtc-generic.c b/drivers/rtc/rtc-generic.c index 3958e87a05fa..92b5e02a3750 100644 --- a/drivers/rtc/rtc-generic.c +++ b/drivers/rtc/rtc-generic.c @@ -9,8 +9,7 @@ #include #include -#if defined(CONFIG_M68K) || defined(CONFIG_PARISC) || \ - defined(CONFIG_PPC) +#if defined(CONFIG_M68K) || defined(CONFIG_PPC) #include static int generic_get_time(struct device *dev, struct rtc_time *tm) -- cgit v1.2.3 From 8bbe6b6f002a00134d3b64d826d7ff596caeda0c Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:56 +0200 Subject: char/genrtc: remove parisc support This architecture selects RTC_CLASS unconditionally, so the GEN_RTC has not worked here for a long time. Now we can remove both the asm/rtc.h header and the Kconfig dependency for CONFIG_GEN_RTC. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/parisc/include/asm/rtc.h | 131 ------------------------------------------ drivers/char/Kconfig | 2 +- 2 files changed, 1 insertion(+), 132 deletions(-) delete mode 100644 arch/parisc/include/asm/rtc.h diff --git a/arch/parisc/include/asm/rtc.h b/arch/parisc/include/asm/rtc.h deleted file mode 100644 index 099d641a42c2..000000000000 --- a/arch/parisc/include/asm/rtc.h +++ /dev/null @@ -1,131 +0,0 @@ -/* - * include/asm-parisc/rtc.h - * - * Copyright 2002 Randolph CHung - * - * Based on: include/asm-ppc/rtc.h and the genrtc driver in the - * 2.4 parisc linux tree - */ - -#ifndef __ASM_RTC_H__ -#define __ASM_RTC_H__ - -#ifdef __KERNEL__ - -#include - -#include - -#define SECS_PER_HOUR (60 * 60) -#define SECS_PER_DAY (SECS_PER_HOUR * 24) - - -#define RTC_PIE 0x40 /* periodic interrupt enable */ -#define RTC_AIE 0x20 /* alarm interrupt enable */ -#define RTC_UIE 0x10 /* update-finished interrupt enable */ - -#define RTC_BATT_BAD 0x100 /* battery bad */ - -/* some dummy definitions */ -#define RTC_SQWE 0x08 /* enable square-wave output */ -#define RTC_DM_BINARY 0x04 /* all time/date values are BCD if clear */ -#define RTC_24H 0x02 /* 24 hour mode - else hours bit 7 means pm */ -#define RTC_DST_EN 0x01 /* auto switch DST - works f. USA only */ - -# define __isleap(year) \ - ((year) % 4 == 0 && ((year) % 100 != 0 || (year) % 400 == 0)) - -/* How many days come before each month (0-12). */ -static const unsigned short int __mon_yday[2][13] = -{ - /* Normal years. */ - { 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365 }, - /* Leap years. */ - { 0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366 } -}; - -static inline unsigned int get_rtc_time(struct rtc_time *wtime) -{ - struct pdc_tod tod_data; - long int days, rem, y; - const unsigned short int *ip; - - memset(wtime, 0, sizeof(*wtime)); - if (pdc_tod_read(&tod_data) < 0) - return RTC_24H | RTC_BATT_BAD; - - // most of the remainder of this function is: -// Copyright (C) 1991, 1993, 1997, 1998 Free Software Foundation, Inc. -// This was originally a part of the GNU C Library. -// It is distributed under the GPL, and was swiped from offtime.c - - - days = tod_data.tod_sec / SECS_PER_DAY; - rem = tod_data.tod_sec % SECS_PER_DAY; - - wtime->tm_hour = rem / SECS_PER_HOUR; - rem %= SECS_PER_HOUR; - wtime->tm_min = rem / 60; - wtime->tm_sec = rem % 60; - - y = 1970; - -#define DIV(a, b) ((a) / (b) - ((a) % (b) < 0)) -#define LEAPS_THRU_END_OF(y) (DIV (y, 4) - DIV (y, 100) + DIV (y, 400)) - - while (days < 0 || days >= (__isleap (y) ? 366 : 365)) - { - /* Guess a corrected year, assuming 365 days per year. */ - long int yg = y + days / 365 - (days % 365 < 0); - - /* Adjust DAYS and Y to match the guessed year. */ - days -= ((yg - y) * 365 - + LEAPS_THRU_END_OF (yg - 1) - - LEAPS_THRU_END_OF (y - 1)); - y = yg; - } - wtime->tm_year = y - 1900; - - ip = __mon_yday[__isleap(y)]; - for (y = 11; days < (long int) ip[y]; --y) - continue; - days -= ip[y]; - wtime->tm_mon = y; - wtime->tm_mday = days + 1; - - return RTC_24H; -} - -static int set_rtc_time(struct rtc_time *wtime) -{ - u_int32_t secs; - - secs = mktime(wtime->tm_year + 1900, wtime->tm_mon + 1, wtime->tm_mday, - wtime->tm_hour, wtime->tm_min, wtime->tm_sec); - - if(pdc_tod_set(secs, 0) < 0) - return -1; - else - return 0; - -} - -static inline unsigned int get_rtc_ss(void) -{ - struct rtc_time h; - - get_rtc_time(&h); - return h.tm_sec; -} - -static inline int get_rtc_pll(struct rtc_pll_info *pll) -{ - return -EINVAL; -} -static inline int set_rtc_pll(struct rtc_pll_info *pll) -{ - return -EINVAL; -} - -#endif /* __KERNEL__ */ -#endif /* __ASM_RTC_H__ */ diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 1d9e3063373b..074bc394e31b 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -328,7 +328,7 @@ config JS_RTC config GEN_RTC tristate "Generic /dev/rtc emulation" depends on RTC!=y - depends on M68K || PARISC || PPC + depends on M68K || PPC ---help--- If you say Y here and create a character special file /dev/rtc with major number 10 and minor number 135 using mknod ("man mknod"), you -- cgit v1.2.3 From 923c904c33023b168baad724d12c7e5260660210 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:57 +0200 Subject: rtc: m68k: provide rtc_class_ops directly The rtc-generic driver provides an architecture specific wrapper on top of the generic rtc_class_ops abstraction, and m68k has another abstraction on top, which is a bit silly. This changes the m68k rtc-generic device to provide its rtc_class_ops directly, to reduce the number of layers by one. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Tested-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/m68k/kernel/time.c | 23 +++++++++++++++++++++-- drivers/rtc/rtc-generic.c | 2 +- 2 files changed, 22 insertions(+), 3 deletions(-) diff --git a/arch/m68k/kernel/time.c b/arch/m68k/kernel/time.c index 3857737e3958..bd6417d38d5a 100644 --- a/arch/m68k/kernel/time.c +++ b/arch/m68k/kernel/time.c @@ -86,7 +86,24 @@ void read_persistent_clock(struct timespec *ts) } } -#ifdef CONFIG_ARCH_USES_GETTIMEOFFSET +#if defined(CONFIG_ARCH_USES_GETTIMEOFFSET) && IS_ENABLED(CONFIG_RTC_DRV_GENERIC) +static int rtc_generic_get_time(struct device *dev, struct rtc_time *tm) +{ + mach_hwclk(0, tm); + return rtc_valid_tm(tm); +} + +static int rtc_generic_set_time(struct device *dev, struct rtc_time *tm) +{ + if (mach_hwclk(1, tm) < 0) + return -EOPNOTSUPP; + return 0; +} + +static const struct rtc_class_ops generic_rtc_ops = { + .read_time = rtc_generic_get_time, + .set_time = rtc_generic_set_time, +}; static int __init rtc_init(void) { @@ -95,7 +112,9 @@ static int __init rtc_init(void) if (!mach_hwclk) return -ENODEV; - pdev = platform_device_register_simple("rtc-generic", -1, NULL, 0); + pdev = platform_device_register_data(NULL, "rtc-generic", -1, + &generic_rtc_ops, + sizeof(generic_rtc_ops)); return PTR_ERR_OR_ZERO(pdev); } diff --git a/drivers/rtc/rtc-generic.c b/drivers/rtc/rtc-generic.c index 92b5e02a3750..5c82bae73b9c 100644 --- a/drivers/rtc/rtc-generic.c +++ b/drivers/rtc/rtc-generic.c @@ -9,7 +9,7 @@ #include #include -#if defined(CONFIG_M68K) || defined(CONFIG_PPC) +#if defined(CONFIG_PPC) #include static int generic_get_time(struct device *dev, struct rtc_time *tm) -- cgit v1.2.3 From 598e8a1fb5e123894686cce5500775c2ae8b57fc Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:58 +0200 Subject: rtc: m68k: provide ioctl for q40 The q40 platform is the only machine in the kernel that provides RTC_PLL_GET/RTC_PLL_SET ioctl commands in its rtc through the mach_get_rtc_pll/mach_set_rtc_pll callbacks. However, this currenctly works only in the old-style genrtc driver, not the (somewhat) modern rtc-generic driver replacing it. This adds an ioctl implementation to the m68k generic_rtc_ops in order to let both drivers provide the same API. After this, we should be able to remove support for genrtc from the m68k architecture. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Tested-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/m68k/kernel/time.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/arch/m68k/kernel/time.c b/arch/m68k/kernel/time.c index bd6417d38d5a..4e5aa2f4f522 100644 --- a/arch/m68k/kernel/time.c +++ b/arch/m68k/kernel/time.c @@ -100,7 +100,32 @@ static int rtc_generic_set_time(struct device *dev, struct rtc_time *tm) return 0; } +static int rtc_ioctl(struct device *dev, unsigned int cmd, unsigned long arg) +{ + struct rtc_pll_info pll; + struct rtc_pll_info __user *argp = (void __user *)arg; + + switch (cmd) { + case RTC_PLL_GET: + if (!mach_get_rtc_pll || mach_get_rtc_pll(&pll)) + return -EINVAL; + return copy_to_user(argp, &pll, sizeof pll) ? -EFAULT : 0; + + case RTC_PLL_SET: + if (!mach_set_rtc_pll) + return -EINVAL; + if (!capable(CAP_SYS_TIME)) + return -EACCES; + if (copy_from_user(&pll, argp, sizeof(pll))) + return -EFAULT; + return mach_set_rtc_pll(&pll); + } + + return -ENOIOCTLCMD; +} + static const struct rtc_class_ops generic_rtc_ops = { + .ioctl = rtc_ioctl, .read_time = rtc_generic_get_time, .set_time = rtc_generic_set_time, }; -- cgit v1.2.3 From 084b3600e2d98ebbab968f91b8b8f48ffbbf2ecb Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:57:59 +0200 Subject: char/genrtc: remove m68k support The asm/rtc.h header is only used for the old gen_rtc driver that has been replaced by rtc-generic. According to Geert Uytterhoeven, nobody has used the old driver on m68k for a long time, so we can now just remove the header file and disallow the driver in Kconfig. All files that used to include asm/rtc.h are now changed so they include the headers that were used implicitly through asm/rtc.h. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Tested-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/m68k/amiga/config.c | 1 - arch/m68k/apollo/config.c | 1 - arch/m68k/bvme6000/config.c | 1 - arch/m68k/hp300/config.c | 2 +- arch/m68k/include/asm/rtc.h | 79 --------------------------------------------- arch/m68k/mac/config.c | 3 +- arch/m68k/mac/misc.c | 1 - arch/m68k/mvme147/config.c | 1 - arch/m68k/mvme16x/config.c | 1 - arch/m68k/q40/config.c | 2 +- arch/m68k/sun3/config.c | 1 - arch/m68k/sun3/intersil.c | 2 +- arch/m68k/sun3x/time.c | 2 +- drivers/char/Kconfig | 2 +- 14 files changed, 7 insertions(+), 92 deletions(-) delete mode 100644 arch/m68k/include/asm/rtc.h diff --git a/arch/m68k/amiga/config.c b/arch/m68k/amiga/config.c index 01693df7f2f6..ec9cc1fdd237 100644 --- a/arch/m68k/amiga/config.c +++ b/arch/m68k/amiga/config.c @@ -35,7 +35,6 @@ #include #include #include -#include #include #include diff --git a/arch/m68k/apollo/config.c b/arch/m68k/apollo/config.c index 6e62d66c396e..432bc8bacfc2 100644 --- a/arch/m68k/apollo/config.c +++ b/arch/m68k/apollo/config.c @@ -15,7 +15,6 @@ #include #include #include -#include #include u_long sio01_physaddr; diff --git a/arch/m68k/bvme6000/config.c b/arch/m68k/bvme6000/config.c index 478623dbb209..611d4d9ea2bd 100644 --- a/arch/m68k/bvme6000/config.c +++ b/arch/m68k/bvme6000/config.c @@ -34,7 +34,6 @@ #include #include #include -#include #include #include diff --git a/arch/m68k/hp300/config.c b/arch/m68k/hp300/config.c index a9befe65adc4..7cfab158fb61 100644 --- a/arch/m68k/hp300/config.c +++ b/arch/m68k/hp300/config.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include @@ -20,7 +21,6 @@ #include #include /* readb() and writeb() */ #include -#include #include "time.h" diff --git a/arch/m68k/include/asm/rtc.h b/arch/m68k/include/asm/rtc.h deleted file mode 100644 index a4d08ea122ee..000000000000 --- a/arch/m68k/include/asm/rtc.h +++ /dev/null @@ -1,79 +0,0 @@ -/* include/asm-m68k/rtc.h - * - * Copyright Richard Zidlicky - * implementation details for genrtc/q40rtc driver - */ -/* permission is hereby granted to copy, modify and redistribute this code - * in terms of the GNU Library General Public License, Version 2 or later, - * at your option. - */ - -#ifndef _ASM_RTC_H -#define _ASM_RTC_H - -#ifdef __KERNEL__ - -#include -#include -#include - -#define RTC_PIE 0x40 /* periodic interrupt enable */ -#define RTC_AIE 0x20 /* alarm interrupt enable */ -#define RTC_UIE 0x10 /* update-finished interrupt enable */ - -/* some dummy definitions */ -#define RTC_BATT_BAD 0x100 /* battery bad */ -#define RTC_SQWE 0x08 /* enable square-wave output */ -#define RTC_DM_BINARY 0x04 /* all time/date values are BCD if clear */ -#define RTC_24H 0x02 /* 24 hour mode - else hours bit 7 means pm */ -#define RTC_DST_EN 0x01 /* auto switch DST - works f. USA only */ - -static inline unsigned int get_rtc_time(struct rtc_time *time) -{ - /* - * Only the values that we read from the RTC are set. We leave - * tm_wday, tm_yday and tm_isdst untouched. Even though the - * RTC has RTC_DAY_OF_WEEK, we ignore it, as it is only updated - * by the RTC when initially set to a non-zero value. - */ - if (mach_hwclk) - mach_hwclk(0, time); - return RTC_24H; -} - -static inline int set_rtc_time(struct rtc_time *time) -{ - if (mach_hwclk) - return mach_hwclk(1, time); - return -EINVAL; -} - -static inline unsigned int get_rtc_ss(void) -{ - if (mach_get_ss) - return mach_get_ss(); - else{ - struct rtc_time h; - - get_rtc_time(&h); - return h.tm_sec; - } -} - -static inline int get_rtc_pll(struct rtc_pll_info *pll) -{ - if (mach_get_rtc_pll) - return mach_get_rtc_pll(pll); - else - return -EINVAL; -} -static inline int set_rtc_pll(struct rtc_pll_info *pll) -{ - if (mach_set_rtc_pll) - return mach_set_rtc_pll(pll); - else - return -EINVAL; -} -#endif /* __KERNEL__ */ - -#endif /* _ASM__RTC_H */ diff --git a/arch/m68k/mac/config.c b/arch/m68k/mac/config.c index 689b47d292ac..2f33a33001e5 100644 --- a/arch/m68k/mac/config.c +++ b/arch/m68k/mac/config.c @@ -10,6 +10,7 @@ * Miscellaneous linux stuff */ +#include #include #include #include @@ -25,6 +26,7 @@ #include #include #include +#include #include #include @@ -34,7 +36,6 @@ #include #include #include -#include #include #include diff --git a/arch/m68k/mac/misc.c b/arch/m68k/mac/misc.c index 707b61aea203..0fb54a90eac2 100644 --- a/arch/m68k/mac/misc.c +++ b/arch/m68k/mac/misc.c @@ -18,7 +18,6 @@ #include #include -#include #include #include #include diff --git a/arch/m68k/mvme147/config.c b/arch/m68k/mvme147/config.c index e6a3b56c6481..c11d38dfad08 100644 --- a/arch/m68k/mvme147/config.c +++ b/arch/m68k/mvme147/config.c @@ -32,7 +32,6 @@ #include #include #include -#include #include #include diff --git a/arch/m68k/mvme16x/config.c b/arch/m68k/mvme16x/config.c index a53803cc66cd..58e240939d26 100644 --- a/arch/m68k/mvme16x/config.c +++ b/arch/m68k/mvme16x/config.c @@ -35,7 +35,6 @@ #include #include #include -#include #include #include diff --git a/arch/m68k/q40/config.c b/arch/m68k/q40/config.c index e90fe903613e..fcb7f05b60b6 100644 --- a/arch/m68k/q40/config.c +++ b/arch/m68k/q40/config.c @@ -12,6 +12,7 @@ * for more details. */ +#include #include #include #include @@ -27,7 +28,6 @@ #include #include -#include #include #include #include diff --git a/arch/m68k/sun3/config.c b/arch/m68k/sun3/config.c index 71884bf01d72..3af34fa3a344 100644 --- a/arch/m68k/sun3/config.c +++ b/arch/m68k/sun3/config.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/m68k/sun3/intersil.c b/arch/m68k/sun3/intersil.c index 889829e11f1d..2cd0bcbe6f30 100644 --- a/arch/m68k/sun3/intersil.c +++ b/arch/m68k/sun3/intersil.c @@ -14,8 +14,8 @@ #include #include -#include #include +#include /* bits to set for start/run of the intersil */ diff --git a/arch/m68k/sun3x/time.c b/arch/m68k/sun3x/time.c index c8eb08add6b0..431d3c4306dd 100644 --- a/arch/m68k/sun3x/time.c +++ b/arch/m68k/sun3x/time.c @@ -15,10 +15,10 @@ #include #include +#include #include #include #include -#include #include "time.h" diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 074bc394e31b..fccac76257c9 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -328,7 +328,7 @@ config JS_RTC config GEN_RTC tristate "Generic /dev/rtc emulation" depends on RTC!=y - depends on M68K || PPC + depends on PPC ---help--- If you say Y here and create a character special file /dev/rtc with major number 10 and minor number 135 using mknod ("man mknod"), you -- cgit v1.2.3 From 169047f447e8e7a56b459e13d86e1fc06457461a Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:58:00 +0200 Subject: rtc: powerpc: provide rtc_class_ops directly The rtc-generic driver provides an architecture specific wrapper on top of the generic rtc_class_ops abstraction, and powerpc has another abstraction on top, which is a bit silly. This changes the powerpc rtc-generic device to provide its rtc_class_ops directly, to reduce the number of layers by one. Signed-off-by: Arnd Bergmann Acked-by: Michael Ellerman Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/powerpc/kernel/time.c | 29 ++++++++++++++++++++++++++++- drivers/rtc/rtc-generic.c | 2 +- 2 files changed, 29 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index 3ed9a5a21d77..7a482a7f4d8d 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -56,6 +56,7 @@ #include #include #include +#include #include #include @@ -1081,6 +1082,29 @@ void calibrate_delay(void) loops_per_jiffy = tb_ticks_per_jiffy; } +#if IS_ENABLED(CONFIG_RTC_DRV_GENERIC) +static int rtc_generic_get_time(struct device *dev, struct rtc_time *tm) +{ + ppc_md.get_rtc_time(tm); + return rtc_valid_tm(tm); +} + +static int rtc_generic_set_time(struct device *dev, struct rtc_time *tm) +{ + if (!ppc_md.set_rtc_time) + return -EOPNOTSUPP; + + if (ppc_md.set_rtc_time(tm) < 0) + return -EOPNOTSUPP; + + return 0; +} + +static const struct rtc_class_ops rtc_generic_ops = { + .read_time = rtc_generic_get_time, + .set_time = rtc_generic_set_time, +}; + static int __init rtc_init(void) { struct platform_device *pdev; @@ -1088,9 +1112,12 @@ static int __init rtc_init(void) if (!ppc_md.get_rtc_time) return -ENODEV; - pdev = platform_device_register_simple("rtc-generic", -1, NULL, 0); + pdev = platform_device_register_data(NULL, "rtc-generic", -1, + &rtc_generic_ops, + sizeof(rtc_generic_ops)); return PTR_ERR_OR_ZERO(pdev); } device_initcall(rtc_init); +#endif diff --git a/drivers/rtc/rtc-generic.c b/drivers/rtc/rtc-generic.c index 5c82bae73b9c..efcb9833cac8 100644 --- a/drivers/rtc/rtc-generic.c +++ b/drivers/rtc/rtc-generic.c @@ -9,7 +9,7 @@ #include #include -#if defined(CONFIG_PPC) +#if 0 #include static int generic_get_time(struct device *dev, struct rtc_time *tm) -- cgit v1.2.3 From 835ea93e9d266dee03b11926656bb52d99b3d274 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:58:01 +0200 Subject: char/genrtc: remove powerpc support PowerPC is the last architecture using the GEN_RTC driver on some machines, but we can migrate them all to using the RTC_DRV_GENERIC driver instead now. This moves over the CONFIG_GEN_RTC option from drivers/char into arch/powerpc/platforms/Kconfig and makes it just select the replacement driver instead, for the only reason of not breaking existing defconfig and .config files that users may have. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/powerpc/include/asm/rtc.h | 78 --------------------------------------- arch/powerpc/platforms/Kconfig | 11 ++++++ arch/powerpc/platforms/ps3/time.c | 2 +- drivers/char/Kconfig | 4 +- drivers/char/Makefile | 2 +- 5 files changed, 15 insertions(+), 82 deletions(-) delete mode 100644 arch/powerpc/include/asm/rtc.h diff --git a/arch/powerpc/include/asm/rtc.h b/arch/powerpc/include/asm/rtc.h deleted file mode 100644 index f5802926b6c0..000000000000 --- a/arch/powerpc/include/asm/rtc.h +++ /dev/null @@ -1,78 +0,0 @@ -/* - * Real-time clock definitions and interfaces - * - * Author: Tom Rini - * - * 2002 (c) MontaVista, Software, Inc. This file is licensed under - * the terms of the GNU General Public License version 2. This program - * is licensed "as is" without any warranty of any kind, whether express - * or implied. - * - * Based on: - * include/asm-m68k/rtc.h - * - * Copyright Richard Zidlicky - * implementation details for genrtc/q40rtc driver - * - * And the old drivers/macintosh/rtc.c which was heavily based on: - * Linux/SPARC Real Time Clock Driver - * Copyright (C) 1996 Thomas K. Dyas (tdyas@eden.rutgers.edu) - * - * With additional work by Paul Mackerras and Franz Sirl. - */ - -#ifndef __ASM_POWERPC_RTC_H__ -#define __ASM_POWERPC_RTC_H__ - -#ifdef __KERNEL__ - -#include - -#include -#include - -#define RTC_PIE 0x40 /* periodic interrupt enable */ -#define RTC_AIE 0x20 /* alarm interrupt enable */ -#define RTC_UIE 0x10 /* update-finished interrupt enable */ - -/* some dummy definitions */ -#define RTC_BATT_BAD 0x100 /* battery bad */ -#define RTC_SQWE 0x08 /* enable square-wave output */ -#define RTC_DM_BINARY 0x04 /* all time/date values are BCD if clear */ -#define RTC_24H 0x02 /* 24 hour mode - else hours bit 7 means pm */ -#define RTC_DST_EN 0x01 /* auto switch DST - works f. USA only */ - -static inline unsigned int get_rtc_time(struct rtc_time *time) -{ - if (ppc_md.get_rtc_time) - ppc_md.get_rtc_time(time); - return RTC_24H; -} - -/* Set the current date and time in the real time clock. */ -static inline int set_rtc_time(struct rtc_time *time) -{ - if (ppc_md.set_rtc_time) - return ppc_md.set_rtc_time(time); - return -EINVAL; -} - -static inline unsigned int get_rtc_ss(void) -{ - struct rtc_time h; - - get_rtc_time(&h); - return h.tm_sec; -} - -static inline int get_rtc_pll(struct rtc_pll_info *pll) -{ - return -EINVAL; -} -static inline int set_rtc_pll(struct rtc_pll_info *pll) -{ - return -EINVAL; -} - -#endif /* __KERNEL__ */ -#endif /* __ASM_POWERPC_RTC_H__ */ diff --git a/arch/powerpc/platforms/Kconfig b/arch/powerpc/platforms/Kconfig index 46a3533d3acb..863aa4341390 100644 --- a/arch/powerpc/platforms/Kconfig +++ b/arch/powerpc/platforms/Kconfig @@ -321,6 +321,17 @@ config OF_RTC Uses information from the OF or flattened device tree to instantiate platform devices for direct mapped RTC chips like the DS1742 or DS1743. +config GEN_RTC + bool "Use the platform RTC operations from user space" + select RTC_CLASS + select RTC_DRV_GENERIC + help + This option provides backwards compatibility with the old gen_rtc.ko + module that was traditionally used for old PowerPC machines. + Platforms should migrate to enabling the RTC_DRV_GENERIC by hand + replacing their get_rtc_time/set_rtc_time callbacks with + a proper RTC device driver. + config SIMPLE_GPIO bool "Support for simple, memory-mapped GPIO controllers" depends on PPC diff --git a/arch/powerpc/platforms/ps3/time.c b/arch/powerpc/platforms/ps3/time.c index 791c6142c4a7..11b45b58c81b 100644 --- a/arch/powerpc/platforms/ps3/time.c +++ b/arch/powerpc/platforms/ps3/time.c @@ -20,9 +20,9 @@ #include #include +#include #include -#include #include #include diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index fccac76257c9..153e90f8a445 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -325,10 +325,10 @@ config JS_RTC To compile this driver as a module, choose M here: the module will be called js-rtc. -config GEN_RTC +config OLD_GEN_RTC tristate "Generic /dev/rtc emulation" depends on RTC!=y - depends on PPC + depends on BROKEN ---help--- If you say Y here and create a character special file /dev/rtc with major number 10 and minor number 135 using mknod ("man mknod"), you diff --git a/drivers/char/Makefile b/drivers/char/Makefile index d8a7579300d2..fc0905aa7373 100644 --- a/drivers/char/Makefile +++ b/drivers/char/Makefile @@ -25,7 +25,7 @@ obj-$(CONFIG_APPLICOM) += applicom.o obj-$(CONFIG_SONYPI) += sonypi.o obj-$(CONFIG_RTC) += rtc.o obj-$(CONFIG_HPET) += hpet.o -obj-$(CONFIG_GEN_RTC) += genrtc.o +obj-$(CONFIG_OLD_GEN_RTC) += genrtc.o obj-$(CONFIG_EFI_RTC) += efirtc.o obj-$(CONFIG_DS1302) += ds1302.o obj-$(CONFIG_XILINX_HWICAP) += xilinx_hwicap/ -- cgit v1.2.3 From 4273b49a5211194e947c08867cf7fa5973d34a14 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:58:02 +0200 Subject: rtc: generic: remove get_rtc_time/set_rtc_time wrappers All architectures using this driver are now converted to provide their own operations, so this one can be turned into a trivial stub driver relying on its platform data. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-generic.c | 35 +---------------------------------- 1 file changed, 1 insertion(+), 34 deletions(-) diff --git a/drivers/rtc/rtc-generic.c b/drivers/rtc/rtc-generic.c index efcb9833cac8..1bf5d2347928 100644 --- a/drivers/rtc/rtc-generic.c +++ b/drivers/rtc/rtc-generic.c @@ -9,43 +9,10 @@ #include #include -#if 0 -#include - -static int generic_get_time(struct device *dev, struct rtc_time *tm) -{ - unsigned int ret = get_rtc_time(tm); - - if (ret & RTC_BATT_BAD) - return -EOPNOTSUPP; - - return rtc_valid_tm(tm); -} - -static int generic_set_time(struct device *dev, struct rtc_time *tm) -{ - if (set_rtc_time(tm) < 0) - return -EOPNOTSUPP; - - return 0; -} - -static const struct rtc_class_ops generic_rtc_ops = { - .read_time = generic_get_time, - .set_time = generic_set_time, -}; -#else -#define generic_rtc_ops *(struct rtc_class_ops*)NULL -#endif - static int __init generic_rtc_probe(struct platform_device *dev) { struct rtc_device *rtc; - const struct rtc_class_ops *ops; - - ops = dev_get_platdata(&dev->dev); - if (!ops) - ops = &generic_rtc_ops; + const struct rtc_class_ops *ops = dev_get_platdata(&dev->dev); rtc = devm_rtc_device_register(&dev->dev, "rtc-generic", ops, THIS_MODULE); -- cgit v1.2.3 From f9a8097a835c504197512f47a3b47823f6432ee1 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:58:03 +0200 Subject: char/genrtc: remove asm-generic/rtc.h from mips arch/mips/sni/time.c includes asm-generic/rtc.h for no apparent reason, and it works fine without that header, so lets remove the inclusion in preparation of deleting the file. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- arch/mips/sni/time.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/mips/sni/time.c b/arch/mips/sni/time.c index fb4b3520cdc6..7ee14f41fc25 100644 --- a/arch/mips/sni/time.c +++ b/arch/mips/sni/time.c @@ -8,7 +8,6 @@ #include #include -#include #define SNI_CLOCK_TICK_RATE 3686400 #define SNI_COUNTER2_DIV 64 -- cgit v1.2.3 From 6705fdb34c7ddc2154f616aba42bc84d4e4a05a2 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 30 May 2016 20:58:04 +0200 Subject: char/genrtc: remove the rest of the driver No architecture uses the genrtc driver any more, so let's kill it off for good. This now also includes asm-generic/rtc.h, which is otherwise completely unused. Signed-off-by: Arnd Bergmann Acked-by: Geert Uytterhoeven Signed-off-by: Alexandre Belloni --- drivers/char/Kconfig | 26 --- drivers/char/Makefile | 1 - drivers/char/genrtc.c | 539 ---------------------------------------------- include/asm-generic/rtc.h | 51 ----- 4 files changed, 617 deletions(-) delete mode 100644 drivers/char/genrtc.c delete mode 100644 include/asm-generic/rtc.h diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 153e90f8a445..0ec91cc6c2fb 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -325,32 +325,6 @@ config JS_RTC To compile this driver as a module, choose M here: the module will be called js-rtc. -config OLD_GEN_RTC - tristate "Generic /dev/rtc emulation" - depends on RTC!=y - depends on BROKEN - ---help--- - If you say Y here and create a character special file /dev/rtc with - major number 10 and minor number 135 using mknod ("man mknod"), you - will get access to the real time clock (or hardware clock) built - into your computer. - - It reports status information via the file /proc/driver/rtc and its - behaviour is set by various ioctls on /dev/rtc. If you enable the - "extended RTC operation" below it will also provide an emulation - for RTC_UIE which is required by some programs and may improve - precision in some cases. - - To compile this driver as a module, choose M here: the - module will be called genrtc. - -config GEN_RTC_X - bool "Extended RTC operation" - depends on GEN_RTC - help - Provides an emulation for RTC_UIE which is required by some programs - and may improve precision of the generic RTC support in some cases. - config EFI_RTC bool "EFI Real Time Clock Services" depends on IA64 diff --git a/drivers/char/Makefile b/drivers/char/Makefile index fc0905aa7373..f33f5e35df9c 100644 --- a/drivers/char/Makefile +++ b/drivers/char/Makefile @@ -25,7 +25,6 @@ obj-$(CONFIG_APPLICOM) += applicom.o obj-$(CONFIG_SONYPI) += sonypi.o obj-$(CONFIG_RTC) += rtc.o obj-$(CONFIG_HPET) += hpet.o -obj-$(CONFIG_OLD_GEN_RTC) += genrtc.o obj-$(CONFIG_EFI_RTC) += efirtc.o obj-$(CONFIG_DS1302) += ds1302.o obj-$(CONFIG_XILINX_HWICAP) += xilinx_hwicap/ diff --git a/drivers/char/genrtc.c b/drivers/char/genrtc.c deleted file mode 100644 index 4f943759d376..000000000000 --- a/drivers/char/genrtc.c +++ /dev/null @@ -1,539 +0,0 @@ -/* - * Real Time Clock interface for - * - q40 and other m68k machines, - * - HP PARISC machines - * - PowerPC machines - * emulate some RTC irq capabilities in software - * - * Copyright (C) 1999 Richard Zidlicky - * - * based on Paul Gortmaker's rtc.c device and - * Sam Creasey Generic rtc driver - * - * This driver allows use of the real time clock (built into - * nearly all computers) from user space. It exports the /dev/rtc - * interface supporting various ioctl() and also the /proc/driver/rtc - * pseudo-file for status information. - * - * The ioctls can be used to set the interrupt behaviour where - * supported. - * - * The /dev/rtc interface will block on reads until an interrupt - * has been received. If a RTC interrupt has already happened, - * it will output an unsigned long and then block. The output value - * contains the interrupt status in the low byte and the number of - * interrupts since the last read in the remaining high bytes. The - * /dev/rtc interface can also be used with the select(2) call. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - * - - * 1.01 fix for 2.3.X rz@linux-m68k.org - * 1.02 merged with code from genrtc.c rz@linux-m68k.org - * 1.03 make it more portable zippel@linux-m68k.org - * 1.04 removed useless timer code rz@linux-m68k.org - * 1.05 portable RTC_UIE emulation rz@linux-m68k.org - * 1.06 set_rtc_time can return an error trini@kernel.crashing.org - * 1.07 ported to HP PARISC (hppa) Helge Deller - */ - -#define RTC_VERSION "1.07" - -#include -#include -#include -#include -#include - -#include -#include -#include -#include -#include -#include -#include - -#include -#include - -/* - * We sponge a minor off of the misc major. No need slurping - * up another valuable major dev number for this. If you add - * an ioctl, make sure you don't conflict with SPARC's RTC - * ioctls. - */ - -static DEFINE_MUTEX(gen_rtc_mutex); -static DECLARE_WAIT_QUEUE_HEAD(gen_rtc_wait); - -/* - * Bits in gen_rtc_status. - */ - -#define RTC_IS_OPEN 0x01 /* means /dev/rtc is in use */ - -static unsigned char gen_rtc_status; /* bitmapped status byte. */ -static unsigned long gen_rtc_irq_data; /* our output to the world */ - -/* months start at 0 now */ -static unsigned char days_in_mo[] = -{31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}; - -static int irq_active; - -#ifdef CONFIG_GEN_RTC_X -static struct work_struct genrtc_task; -static struct timer_list timer_task; - -static unsigned int oldsecs; -static int lostint; -static unsigned long tt_exp; - -static void gen_rtc_timer(unsigned long data); - -static volatile int stask_active; /* schedule_work */ -static volatile int ttask_active; /* timer_task */ -static int stop_rtc_timers; /* don't requeue tasks */ -static DEFINE_SPINLOCK(gen_rtc_lock); - -static void gen_rtc_interrupt(unsigned long arg); - -/* - * Routine to poll RTC seconds field for change as often as possible, - * after first RTC_UIE use timer to reduce polling - */ -static void genrtc_troutine(struct work_struct *work) -{ - unsigned int tmp = get_rtc_ss(); - - if (stop_rtc_timers) { - stask_active = 0; - return; - } - - if (oldsecs != tmp){ - oldsecs = tmp; - - timer_task.function = gen_rtc_timer; - timer_task.expires = jiffies + HZ - (HZ/10); - tt_exp=timer_task.expires; - ttask_active=1; - stask_active=0; - add_timer(&timer_task); - - gen_rtc_interrupt(0); - } else if (schedule_work(&genrtc_task) == 0) - stask_active = 0; -} - -static void gen_rtc_timer(unsigned long data) -{ - lostint = get_rtc_ss() - oldsecs ; - if (lostint<0) - lostint = 60 - lostint; - if (time_after(jiffies, tt_exp)) - printk(KERN_INFO "genrtc: timer task delayed by %ld jiffies\n", - jiffies-tt_exp); - ttask_active=0; - stask_active=1; - if ((schedule_work(&genrtc_task) == 0)) - stask_active = 0; -} - -/* - * call gen_rtc_interrupt function to signal an RTC_UIE, - * arg is unused. - * Could be invoked either from a real interrupt handler or - * from some routine that periodically (eg 100HZ) monitors - * whether RTC_SECS changed - */ -static void gen_rtc_interrupt(unsigned long arg) -{ - /* We store the status in the low byte and the number of - * interrupts received since the last read in the remainder - * of rtc_irq_data. */ - - gen_rtc_irq_data += 0x100; - gen_rtc_irq_data &= ~0xff; - gen_rtc_irq_data |= RTC_UIE; - - if (lostint){ - printk("genrtc: system delaying clock ticks?\n"); - /* increment count so that userspace knows something is wrong */ - gen_rtc_irq_data += ((lostint-1)<<8); - lostint = 0; - } - - wake_up_interruptible(&gen_rtc_wait); -} - -/* - * Now all the various file operations that we export. - */ -static ssize_t gen_rtc_read(struct file *file, char __user *buf, - size_t count, loff_t *ppos) -{ - unsigned long data; - ssize_t retval; - - if (count != sizeof (unsigned int) && count != sizeof (unsigned long)) - return -EINVAL; - - if (file->f_flags & O_NONBLOCK && !gen_rtc_irq_data) - return -EAGAIN; - - retval = wait_event_interruptible(gen_rtc_wait, - (data = xchg(&gen_rtc_irq_data, 0))); - if (retval) - goto out; - - /* first test allows optimizer to nuke this case for 32-bit machines */ - if (sizeof (int) != sizeof (long) && count == sizeof (unsigned int)) { - unsigned int uidata = data; - retval = put_user(uidata, (unsigned int __user *)buf) ?: - sizeof(unsigned int); - } - else { - retval = put_user(data, (unsigned long __user *)buf) ?: - sizeof(unsigned long); - } -out: - return retval; -} - -static unsigned int gen_rtc_poll(struct file *file, - struct poll_table_struct *wait) -{ - poll_wait(file, &gen_rtc_wait, wait); - if (gen_rtc_irq_data != 0) - return POLLIN | POLLRDNORM; - return 0; -} - -#endif - -/* - * Used to disable/enable interrupts, only RTC_UIE supported - * We also clear out any old irq data after an ioctl() that - * meddles with the interrupt enable/disable bits. - */ - -static inline void gen_clear_rtc_irq_bit(unsigned char bit) -{ -#ifdef CONFIG_GEN_RTC_X - stop_rtc_timers = 1; - if (ttask_active){ - del_timer_sync(&timer_task); - ttask_active = 0; - } - while (stask_active) - schedule(); - - spin_lock(&gen_rtc_lock); - irq_active = 0; - spin_unlock(&gen_rtc_lock); -#endif -} - -static inline int gen_set_rtc_irq_bit(unsigned char bit) -{ -#ifdef CONFIG_GEN_RTC_X - spin_lock(&gen_rtc_lock); - if ( !irq_active ) { - irq_active = 1; - stop_rtc_timers = 0; - lostint = 0; - INIT_WORK(&genrtc_task, genrtc_troutine); - oldsecs = get_rtc_ss(); - init_timer(&timer_task); - - stask_active = 1; - if (schedule_work(&genrtc_task) == 0){ - stask_active = 0; - } - } - spin_unlock(&gen_rtc_lock); - gen_rtc_irq_data = 0; - return 0; -#else - return -EINVAL; -#endif -} - -static int gen_rtc_ioctl(struct file *file, - unsigned int cmd, unsigned long arg) -{ - struct rtc_time wtime; - struct rtc_pll_info pll; - void __user *argp = (void __user *)arg; - - switch (cmd) { - - case RTC_PLL_GET: - if (get_rtc_pll(&pll)) - return -EINVAL; - else - return copy_to_user(argp, &pll, sizeof pll) ? -EFAULT : 0; - - case RTC_PLL_SET: - if (!capable(CAP_SYS_TIME)) - return -EACCES; - if (copy_from_user(&pll, argp, sizeof(pll))) - return -EFAULT; - return set_rtc_pll(&pll); - - case RTC_UIE_OFF: /* disable ints from RTC updates. */ - gen_clear_rtc_irq_bit(RTC_UIE); - return 0; - - case RTC_UIE_ON: /* enable ints for RTC updates. */ - return gen_set_rtc_irq_bit(RTC_UIE); - - case RTC_RD_TIME: /* Read the time/date from RTC */ - /* this doesn't get week-day, who cares */ - memset(&wtime, 0, sizeof(wtime)); - get_rtc_time(&wtime); - - return copy_to_user(argp, &wtime, sizeof(wtime)) ? -EFAULT : 0; - - case RTC_SET_TIME: /* Set the RTC */ - { - int year; - unsigned char leap_yr; - - if (!capable(CAP_SYS_TIME)) - return -EACCES; - - if (copy_from_user(&wtime, argp, sizeof(wtime))) - return -EFAULT; - - year = wtime.tm_year + 1900; - leap_yr = ((!(year % 4) && (year % 100)) || - !(year % 400)); - - if ((wtime.tm_mon < 0 || wtime.tm_mon > 11) || (wtime.tm_mday < 1)) - return -EINVAL; - - if (wtime.tm_mday < 0 || wtime.tm_mday > - (days_in_mo[wtime.tm_mon] + ((wtime.tm_mon == 1) && leap_yr))) - return -EINVAL; - - if (wtime.tm_hour < 0 || wtime.tm_hour >= 24 || - wtime.tm_min < 0 || wtime.tm_min >= 60 || - wtime.tm_sec < 0 || wtime.tm_sec >= 60) - return -EINVAL; - - return set_rtc_time(&wtime); - } - } - - return -EINVAL; -} - -static long gen_rtc_unlocked_ioctl(struct file *file, unsigned int cmd, - unsigned long arg) -{ - int ret; - - mutex_lock(&gen_rtc_mutex); - ret = gen_rtc_ioctl(file, cmd, arg); - mutex_unlock(&gen_rtc_mutex); - - return ret; -} - -/* - * We enforce only one user at a time here with the open/close. - * Also clear the previous interrupt data on an open, and clean - * up things on a close. - */ - -static int gen_rtc_open(struct inode *inode, struct file *file) -{ - mutex_lock(&gen_rtc_mutex); - if (gen_rtc_status & RTC_IS_OPEN) { - mutex_unlock(&gen_rtc_mutex); - return -EBUSY; - } - - gen_rtc_status |= RTC_IS_OPEN; - gen_rtc_irq_data = 0; - irq_active = 0; - mutex_unlock(&gen_rtc_mutex); - - return 0; -} - -static int gen_rtc_release(struct inode *inode, struct file *file) -{ - /* - * Turn off all interrupts once the device is no longer - * in use and clear the data. - */ - - gen_clear_rtc_irq_bit(RTC_PIE|RTC_AIE|RTC_UIE); - - gen_rtc_status &= ~RTC_IS_OPEN; - return 0; -} - - -#ifdef CONFIG_PROC_FS - -/* - * Info exported via "/proc/driver/rtc". - */ - -static int gen_rtc_proc_show(struct seq_file *m, void *v) -{ - struct rtc_time tm; - unsigned int flags; - struct rtc_pll_info pll; - - flags = get_rtc_time(&tm); - - seq_printf(m, - "rtc_time\t: %02d:%02d:%02d\n" - "rtc_date\t: %04d-%02d-%02d\n" - "rtc_epoch\t: %04u\n", - tm.tm_hour, tm.tm_min, tm.tm_sec, - tm.tm_year + 1900, tm.tm_mon + 1, tm.tm_mday, 1900); - - tm.tm_hour = tm.tm_min = tm.tm_sec = 0; - - seq_puts(m, "alarm\t\t: "); - if (tm.tm_hour <= 24) - seq_printf(m, "%02d:", tm.tm_hour); - else - seq_puts(m, "**:"); - - if (tm.tm_min <= 59) - seq_printf(m, "%02d:", tm.tm_min); - else - seq_puts(m, "**:"); - - if (tm.tm_sec <= 59) - seq_printf(m, "%02d\n", tm.tm_sec); - else - seq_puts(m, "**\n"); - - seq_printf(m, - "DST_enable\t: %s\n" - "BCD\t\t: %s\n" - "24hr\t\t: %s\n" - "square_wave\t: %s\n" - "alarm_IRQ\t: %s\n" - "update_IRQ\t: %s\n" - "periodic_IRQ\t: %s\n" - "periodic_freq\t: %ld\n" - "batt_status\t: %s\n", - (flags & RTC_DST_EN) ? "yes" : "no", - (flags & RTC_DM_BINARY) ? "no" : "yes", - (flags & RTC_24H) ? "yes" : "no", - (flags & RTC_SQWE) ? "yes" : "no", - (flags & RTC_AIE) ? "yes" : "no", - irq_active ? "yes" : "no", - (flags & RTC_PIE) ? "yes" : "no", - 0L /* freq */, - (flags & RTC_BATT_BAD) ? "bad" : "okay"); - if (!get_rtc_pll(&pll)) - seq_printf(m, - "PLL adjustment\t: %d\n" - "PLL max +ve adjustment\t: %d\n" - "PLL max -ve adjustment\t: %d\n" - "PLL +ve adjustment factor\t: %d\n" - "PLL -ve adjustment factor\t: %d\n" - "PLL frequency\t: %ld\n", - pll.pll_value, - pll.pll_max, - pll.pll_min, - pll.pll_posmult, - pll.pll_negmult, - pll.pll_clock); - return 0; -} - -static int gen_rtc_proc_open(struct inode *inode, struct file *file) -{ - return single_open(file, gen_rtc_proc_show, NULL); -} - -static const struct file_operations gen_rtc_proc_fops = { - .open = gen_rtc_proc_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -static int __init gen_rtc_proc_init(void) -{ - struct proc_dir_entry *r; - - r = proc_create("driver/rtc", 0, NULL, &gen_rtc_proc_fops); - if (!r) - return -ENOMEM; - return 0; -} -#else -static inline int gen_rtc_proc_init(void) { return 0; } -#endif /* CONFIG_PROC_FS */ - - -/* - * The various file operations we support. - */ - -static const struct file_operations gen_rtc_fops = { - .owner = THIS_MODULE, -#ifdef CONFIG_GEN_RTC_X - .read = gen_rtc_read, - .poll = gen_rtc_poll, -#endif - .unlocked_ioctl = gen_rtc_unlocked_ioctl, - .open = gen_rtc_open, - .release = gen_rtc_release, - .llseek = noop_llseek, -}; - -static struct miscdevice rtc_gen_dev = -{ - .minor = RTC_MINOR, - .name = "rtc", - .fops = &gen_rtc_fops, -}; - -static int __init rtc_generic_init(void) -{ - int retval; - - printk(KERN_INFO "Generic RTC Driver v%s\n", RTC_VERSION); - - retval = misc_register(&rtc_gen_dev); - if (retval < 0) - return retval; - - retval = gen_rtc_proc_init(); - if (retval) { - misc_deregister(&rtc_gen_dev); - return retval; - } - - return 0; -} - -static void __exit rtc_generic_exit(void) -{ - remove_proc_entry ("driver/rtc", NULL); - misc_deregister(&rtc_gen_dev); -} - - -module_init(rtc_generic_init); -module_exit(rtc_generic_exit); - -MODULE_AUTHOR("Richard Zidlicky"); -MODULE_LICENSE("GPL"); -MODULE_ALIAS_MISCDEV(RTC_MINOR); diff --git a/include/asm-generic/rtc.h b/include/asm-generic/rtc.h deleted file mode 100644 index 4fcff22cd707..000000000000 --- a/include/asm-generic/rtc.h +++ /dev/null @@ -1,51 +0,0 @@ -/* - * include/asm-generic/rtc.h - * - * Author: Tom Rini - * - * Based on: - * drivers/char/rtc.c - * - * Please read the COPYING file for all license details. - */ - -#ifndef __ASM_RTC_H__ -#define __ASM_RTC_H__ - -#include - -#ifndef get_rtc_time -#include -#define get_rtc_time mc146818_get_time -#define set_rtc_time mc146818_set_time -#endif - -#define RTC_PIE 0x40 /* periodic interrupt enable */ -#define RTC_AIE 0x20 /* alarm interrupt enable */ -#define RTC_UIE 0x10 /* update-finished interrupt enable */ - -/* some dummy definitions */ -#define RTC_BATT_BAD 0x100 /* battery bad */ -#define RTC_SQWE 0x08 /* enable square-wave output */ -#define RTC_DM_BINARY 0x04 /* all time/date values are BCD if clear */ -#define RTC_24H 0x02 /* 24 hour mode - else hours bit 7 means pm */ -#define RTC_DST_EN 0x01 /* auto switch DST - works f. USA only */ - -static inline unsigned int get_rtc_ss(void) -{ - struct rtc_time h; - - get_rtc_time(&h); - return h.tm_sec; -} - -static inline int get_rtc_pll(struct rtc_pll_info *pll) -{ - return -EINVAL; -} -static inline int set_rtc_pll(struct rtc_pll_info *pll) -{ - return -EINVAL; -} - -#endif /* __ASM_RTC_H__ */ -- cgit v1.2.3 From 279da1495dae0d8ceee0cbe26187188ee27b7853 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sat, 4 Jun 2016 11:36:44 +0200 Subject: sparc32: remove stale RTC_PORT definition sparc32:allmodconfig fails to build in next-20160602 as follows. In file included from drivers/block/floppy.c:185:0: include/linux/mc146818rtc.h: In function 'mc146818_is_updating': include/linux/mc146818rtc.h:138:9: error: 'rtc_port' undeclared (first use in this function) include/linux/mc146818rtc.h:138:9: note: each undeclared identifier is reported only once for each function it appears in include/linux/mc146818rtc.h: In function 'mc146818_get_time': include/linux/mc146818rtc.h:172:17: error: 'rtc_port' undeclared (first use in this function) include/linux/mc146818rtc.h: In function 'mc146818_set_time': include/linux/mc146818rtc.h:278:8: error: 'rtc_port' undeclared (first use in this function) scripts/Makefile.build:295: recipe for target 'drivers/block/floppy.o' failed The reason is a duplicate definition of the RTC_PORT macro. The one in arch/sparc/include/asm/io_32.h was apparently used a long time ago for the drivers/char/rtc.c driver that is not available on SPARC any more, since we now select 'RTC_CLASS' unconditionally. Removing the macro fixes the build problem, and for consistency, this also removes the RTC_ALWAYS_BCD macro and the comment for both. Signed-off-by: Arnd Bergmann Reported-by: Guenter Roeck Fixes: fd09cc80165c ("rtc: cmos: move mc146818rtc code out of asm-generic/rtc.h") Signed-off-by: Alexandre Belloni --- arch/sparc/include/asm/io_32.h | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/arch/sparc/include/asm/io_32.h b/arch/sparc/include/asm/io_32.h index 57f26c398dc9..4dd268a3a8b0 100644 --- a/arch/sparc/include/asm/io_32.h +++ b/arch/sparc/include/asm/io_32.h @@ -140,16 +140,6 @@ void ioport_unmap(void __iomem *); struct pci_dev; void pci_iounmap(struct pci_dev *dev, void __iomem *); - - -/* - * At the moment, we do not use CMOS_READ anywhere outside of rtc.c, - * so rtc_port is static in it. This should not change unless a new - * hardware pops up. - */ -#define RTC_PORT(x) (rtc_port + (x)) -#define RTC_ALWAYS_BCD 0 - static inline int sbus_can_dma_64bit(void) { return 0; /* actually, sparc_cpu_model==sun4d */ -- cgit v1.2.3 From 2b2f5ff00f63847d95adad6289bd8b05f5983dd5 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Mon, 16 May 2016 17:22:54 +0100 Subject: rtc: interface: ignore expired timers when enqueuing new timers This patch fixes a RTC wakealarm issue, namely, the event fires during hibernate and is not cleared from the list, causing hwclock to block. The current enqueuing does not trigger an alarm if any expired timers already exist on the timerqueue. This can occur when a RTC wake alarm is used to wake a machine out of hibernate and the resumed state has old expired timers that have not been removed from the timer queue. This fix skips over any expired timers and triggers an alarm if there are no pending timers on the timerqueue. Note that the skipped expired timer will get reaped later on, so there is no need to clean it up immediately. The issue can be reproduced by putting a machine into hibernate and waking it with the RTC wakealarm. Running the example RTC test program from tools/testing/selftests/timers/rtctest.c after the hibernate will block indefinitely. With the fix, it no longer blocks after the hibernate resume. BugLink: http://bugs.launchpad.net/bugs/1333569 Signed-off-by: Colin Ian King Signed-off-by: Alexandre Belloni --- drivers/rtc/interface.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) diff --git a/drivers/rtc/interface.c b/drivers/rtc/interface.c index 99475908e556..7cafd4d33ed3 100644 --- a/drivers/rtc/interface.c +++ b/drivers/rtc/interface.c @@ -758,9 +758,23 @@ EXPORT_SYMBOL_GPL(rtc_irq_set_freq); */ static int rtc_timer_enqueue(struct rtc_device *rtc, struct rtc_timer *timer) { + struct timerqueue_node *next = timerqueue_getnext(&rtc->timerqueue); + struct rtc_time tm; + ktime_t now; + timer->enabled = 1; + __rtc_read_time(rtc, &tm); + now = rtc_tm_to_ktime(tm); + + /* Skip over expired timers */ + while (next) { + if (next->expires.tv64 >= now.tv64) + break; + next = timerqueue_iterate_next(next); + } + timerqueue_add(&rtc->timerqueue, &timer->node); - if (&timer->node == timerqueue_getnext(&rtc->timerqueue)) { + if (!next) { struct rtc_wkalrm alarm; int err; alarm.time = rtc_ktime_to_tm(timer->node.expires); -- cgit v1.2.3 From c5776dec8f2a90a0aa9e197c2900f68fda8dd978 Mon Sep 17 00:00:00 2001 From: Heinrich Schuchardt Date: Sun, 22 May 2016 00:18:55 +0200 Subject: rtc: ds1685: correct day of month checking The day of month is checked in ds1685_rtc_read_alarm and ds1685_rtc_set_alarm. Multiple errors exist in the day of month check. Operator ! has a higher priority than &&. (!(mday >= 1) && (mday <= 31)) is false for mday == 32. When verifying the day of month the binary and the BCD mode have to be considered. Signed-off-by: Heinrich Schuchardt Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-ds1685.c | 34 +++++++++++++++++++++++++++++----- 1 file changed, 29 insertions(+), 5 deletions(-) diff --git a/drivers/rtc/rtc-ds1685.c b/drivers/rtc/rtc-ds1685.c index b3ce3c652fcd..68197042feb7 100644 --- a/drivers/rtc/rtc-ds1685.c +++ b/drivers/rtc/rtc-ds1685.c @@ -102,6 +102,26 @@ ds1685_rtc_bin2bcd(struct ds1685_priv *rtc, u8 val, u8 bin_mask, u8 bcd_mask) return (val & bin_mask); } +/** + * s1685_rtc_check_mday - check validity of the day of month. + * @rtc: pointer to the ds1685 rtc structure. + * @mday: day of month. + * + * Returns -EDOM if the day of month is not within 1..31 range. + */ +static inline int +ds1685_rtc_check_mday(struct ds1685_priv *rtc, u8 mday) +{ + if (rtc->bcd_mode) { + if (mday < 0x01 || mday > 0x31 || (mday & 0x0f) > 0x09) + return -EDOM; + } else { + if (mday < 1 || mday > 31) + return -EDOM; + } + return 0; +} + /** * ds1685_rtc_switch_to_bank0 - switch the rtc to bank 0. * @rtc: pointer to the ds1685 rtc structure. @@ -377,6 +397,7 @@ ds1685_rtc_read_alarm(struct device *dev, struct rtc_wkalrm *alrm) struct platform_device *pdev = to_platform_device(dev); struct ds1685_priv *rtc = platform_get_drvdata(pdev); u8 seconds, minutes, hours, mday, ctrlb, ctrlc; + int ret; /* Fetch the alarm info from the RTC alarm registers. */ ds1685_rtc_begin_data_access(rtc); @@ -388,9 +409,10 @@ ds1685_rtc_read_alarm(struct device *dev, struct rtc_wkalrm *alrm) ctrlc = rtc->read(rtc, RTC_CTRL_C); ds1685_rtc_end_data_access(rtc); - /* Check month date. */ - if (!(mday >= 1) && (mday <= 31)) - return -EDOM; + /* Check the month date for validity. */ + ret = ds1685_rtc_check_mday(rtc, mday); + if (ret) + return ret; /* * Check the three alarm bytes. @@ -445,6 +467,7 @@ ds1685_rtc_set_alarm(struct device *dev, struct rtc_wkalrm *alrm) struct platform_device *pdev = to_platform_device(dev); struct ds1685_priv *rtc = platform_get_drvdata(pdev); u8 ctrlb, seconds, minutes, hours, mday; + int ret; /* Fetch the alarm info and convert to BCD. */ seconds = ds1685_rtc_bin2bcd(rtc, alrm->time.tm_sec, @@ -461,8 +484,9 @@ ds1685_rtc_set_alarm(struct device *dev, struct rtc_wkalrm *alrm) RTC_MDAY_BCD_MASK); /* Check the month date for validity. */ - if (!(mday >= 1) && (mday <= 31)) - return -EDOM; + ret = ds1685_rtc_check_mday(rtc, mday); + if (ret) + return ret; /* * Check the three alarm bytes. -- cgit v1.2.3 From 82df3e045d8cce5eaeb9bf1ed1d8d8841b04a41a Mon Sep 17 00:00:00 2001 From: Martin Kepplinger Date: Mon, 18 Apr 2016 12:17:44 +0200 Subject: rtc: pcf2123: use sign_extend32() for sign extension Use sign_extend32() instead of open coding sign extension. Signed-off-by: Martin Kepplinger Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-pcf2123.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/rtc/rtc-pcf2123.c b/drivers/rtc/rtc-pcf2123.c index f22e060709e5..b4478cc92b55 100644 --- a/drivers/rtc/rtc-pcf2123.c +++ b/drivers/rtc/rtc-pcf2123.c @@ -96,7 +96,7 @@ #define CD_TMR_TE BIT(3) /* Countdown timer enable */ /* PCF2123_REG_OFFSET BITS */ -#define OFFSET_SIGN_BIT BIT(6) /* 2's complement sign bit */ +#define OFFSET_SIGN_BIT 6 /* 2's complement sign bit */ #define OFFSET_COARSE BIT(7) /* Coarse mode offset */ #define OFFSET_STEP (2170) /* Offset step in parts per billion */ @@ -217,7 +217,7 @@ static int pcf2123_read_offset(struct device *dev, long *offset) if (reg & OFFSET_COARSE) reg <<= 1; /* multiply by 2 and sign extend */ else - reg |= (reg & OFFSET_SIGN_BIT) << 1; /* sign extend only */ + reg = sign_extend32(reg, OFFSET_SIGN_BIT); *offset = ((long)reg) * OFFSET_STEP; -- cgit v1.2.3 From 06776c89214a14735fed5519340e2cea88b5b91f Mon Sep 17 00:00:00 2001 From: Venkat Prashanth B U Date: Wed, 1 Jun 2016 11:59:54 +0530 Subject: rtc: add support for Maxim max6916 Add support for Maxim max6916 RTC. Signed-off-by: Venkat Prashanth B U Signed-off-by: Alexandre Belloni --- drivers/rtc/Kconfig | 12 ++++ drivers/rtc/Makefile | 1 + drivers/rtc/rtc-max6916.c | 164 ++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 177 insertions(+) create mode 100644 drivers/rtc/rtc-max6916.c diff --git a/drivers/rtc/Kconfig b/drivers/rtc/Kconfig index 18639e0cb6e2..9f8b59d94b1d 100644 --- a/drivers/rtc/Kconfig +++ b/drivers/rtc/Kconfig @@ -670,6 +670,18 @@ config RTC_DRV_DS1390 This driver can also be built as a module. If so, the module will be called rtc-ds1390. +config RTC_DRV_MAX6916 + tristate "Maxim MAX6916" + help + If you say yes here you will get support for the + Maxim MAX6916 SPI RTC chip. + + This driver only supports the RTC feature, and not other chip + features such as alarms. + + This driver can also be built as a module. If so, the module + will be called rtc-max6916. + config RTC_DRV_R9701 tristate "Epson RTC-9701JE" help diff --git a/drivers/rtc/Makefile b/drivers/rtc/Makefile index ea2833723fa9..024da2723d86 100644 --- a/drivers/rtc/Makefile +++ b/drivers/rtc/Makefile @@ -85,6 +85,7 @@ obj-$(CONFIG_RTC_DRV_M48T59) += rtc-m48t59.o obj-$(CONFIG_RTC_DRV_M48T86) += rtc-m48t86.o obj-$(CONFIG_RTC_DRV_MAX6900) += rtc-max6900.o obj-$(CONFIG_RTC_DRV_MAX6902) += rtc-max6902.o +obj-$(CONFIG_RTC_DRV_MAX6916) += rtc-max6916.o obj-$(CONFIG_RTC_DRV_MAX77686) += rtc-max77686.o obj-$(CONFIG_RTC_DRV_MAX8907) += rtc-max8907.o obj-$(CONFIG_RTC_DRV_MAX8925) += rtc-max8925.o diff --git a/drivers/rtc/rtc-max6916.c b/drivers/rtc/rtc-max6916.c new file mode 100644 index 000000000000..623ab27b2757 --- /dev/null +++ b/drivers/rtc/rtc-max6916.c @@ -0,0 +1,164 @@ +/* rtc-max6916.c + * + * Driver for MAXIM max6916 Low Current, SPI Compatible + * Real Time Clock + * + * Author : Venkat Prashanth B U + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#include +#include +#include +#include +#include +#include +#include + +/* Registers in max6916 rtc */ + +#define MAX6916_SECONDS_REG 0x01 +#define MAX6916_MINUTES_REG 0x02 +#define MAX6916_HOURS_REG 0x03 +#define MAX6916_DATE_REG 0x04 +#define MAX6916_MONTH_REG 0x05 +#define MAX6916_DAY_REG 0x06 +#define MAX6916_YEAR_REG 0x07 +#define MAX6916_CONTROL_REG 0x08 +#define MAX6916_STATUS_REG 0x0C +#define MAX6916_CLOCK_BURST 0x3F + +static int max6916_read_reg(struct device *dev, unsigned char address, + unsigned char *data) +{ + struct spi_device *spi = to_spi_device(dev); + + *data = address | 0x80; + + return spi_write_then_read(spi, data, 1, data, 1); +} + +static int max6916_write_reg(struct device *dev, unsigned char address, + unsigned char data) +{ + struct spi_device *spi = to_spi_device(dev); + unsigned char buf[2]; + + buf[0] = address & 0x7F; + buf[1] = data; + + return spi_write_then_read(spi, buf, 2, NULL, 0); +} + +static int max6916_read_time(struct device *dev, struct rtc_time *dt) +{ + struct spi_device *spi = to_spi_device(dev); + int err; + unsigned char buf[8]; + + buf[0] = MAX6916_CLOCK_BURST | 0x80; + + err = spi_write_then_read(spi, buf, 1, buf, 8); + + if (err) + return err; + + dt->tm_sec = bcd2bin(buf[0]); + dt->tm_min = bcd2bin(buf[1]); + dt->tm_hour = bcd2bin(buf[2] & 0x3F); + dt->tm_mday = bcd2bin(buf[3]); + dt->tm_mon = bcd2bin(buf[4]) - 1; + dt->tm_wday = bcd2bin(buf[5]) - 1; + dt->tm_year = bcd2bin(buf[6]) + 100; + + return rtc_valid_tm(dt); +} + +static int max6916_set_time(struct device *dev, struct rtc_time *dt) +{ + struct spi_device *spi = to_spi_device(dev); + unsigned char buf[9]; + + if (dt->tm_year < 100 || dt->tm_year > 199) { + dev_err(&spi->dev, "Year must be between 2000 and 2099. It's %d.\n", + dt->tm_year + 1900); + return -EINVAL; + } + + buf[0] = MAX6916_CLOCK_BURST & 0x7F; + buf[1] = bin2bcd(dt->tm_sec); + buf[2] = bin2bcd(dt->tm_min); + buf[3] = (bin2bcd(dt->tm_hour) & 0X3F); + buf[4] = bin2bcd(dt->tm_mday); + buf[5] = bin2bcd(dt->tm_mon + 1); + buf[6] = bin2bcd(dt->tm_wday + 1); + buf[7] = bin2bcd(dt->tm_year % 100); + buf[8] = bin2bcd(0x00); + + /* write the rtc settings */ + return spi_write_then_read(spi, buf, 9, NULL, 0); +} + +static const struct rtc_class_ops max6916_rtc_ops = { + .read_time = max6916_read_time, + .set_time = max6916_set_time, +}; + +static int max6916_probe(struct spi_device *spi) +{ + struct rtc_device *rtc; + unsigned char data; + int res; + + /* spi setup with max6916 in mode 3 and bits per word as 8 */ + spi->mode = SPI_MODE_3; + spi->bits_per_word = 8; + spi_setup(spi); + + /* RTC Settings */ + res = max6916_read_reg(&spi->dev, MAX6916_SECONDS_REG, &data); + if (res) + return res; + + /* Disable the write protect of rtc */ + max6916_read_reg(&spi->dev, MAX6916_CONTROL_REG, &data); + data = data & ~(1 << 7); + max6916_write_reg(&spi->dev, MAX6916_CONTROL_REG, data); + + /*Enable oscillator,disable oscillator stop flag, glitch filter*/ + max6916_read_reg(&spi->dev, MAX6916_STATUS_REG, &data); + data = data & 0x1B; + max6916_write_reg(&spi->dev, MAX6916_STATUS_REG, data); + + /* display the settings */ + max6916_read_reg(&spi->dev, MAX6916_CONTROL_REG, &data); + dev_info(&spi->dev, "MAX6916 RTC CTRL Reg = 0x%02x\n", data); + + max6916_read_reg(&spi->dev, MAX6916_STATUS_REG, &data); + dev_info(&spi->dev, "MAX6916 RTC Status Reg = 0x%02x\n", data); + + rtc = devm_rtc_device_register(&spi->dev, "max6916", + &max6916_rtc_ops, THIS_MODULE); + if (IS_ERR(rtc)) + return PTR_ERR(rtc); + + spi_set_drvdata(spi, rtc); + + return 0; +} + +static struct spi_driver max6916_driver = { + .driver = { + .name = "max6916", + }, + .probe = max6916_probe, +}; +module_spi_driver(max6916_driver); + +MODULE_DESCRIPTION("MAX6916 SPI RTC DRIVER"); +MODULE_AUTHOR("Venkat Prashanth B U "); +MODULE_LICENSE("GPL v2"); -- cgit v1.2.3 From 7368c69c033ff29e8e766f481748113c7adc1d1b Mon Sep 17 00:00:00 2001 From: Alexander Graf Date: Sun, 5 Jun 2016 11:35:56 +0200 Subject: rtc: efi: Fail probing if RTC reads don't work While the EFI spec mandates an RTC, not every implementation actually adheres to that rule (or can adhere to it - some systems just don't have an RTC). For those, we really don't want to probe the EFI RTC driver at all, because if we do we'd get a non-functional driver that does nothing useful but only spills our kernel log with warnings. Signed-off-by: Alexander Graf Acked-by: Ard Biesheuvel Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-efi.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/rtc/rtc-efi.c b/drivers/rtc/rtc-efi.c index 96d38609d803..0130afd7fe88 100644 --- a/drivers/rtc/rtc-efi.c +++ b/drivers/rtc/rtc-efi.c @@ -259,6 +259,12 @@ static const struct rtc_class_ops efi_rtc_ops = { static int __init efi_rtc_probe(struct platform_device *dev) { struct rtc_device *rtc; + efi_time_t eft; + efi_time_cap_t cap; + + /* First check if the RTC is usable */ + if (efi.get_time(&eft, &cap) != EFI_SUCCESS) + return -ENODEV; rtc = devm_rtc_device_register(&dev->dev, "rtc-efi", &efi_rtc_ops, THIS_MODULE); -- cgit v1.2.3 From c361db5c2c64f1b7ffed5e9cc42e5062432238f2 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Mon, 6 Jun 2016 16:09:52 +0200 Subject: x86: include linux/ratelimit.h in nmi.c When building random configurations, we now occasionally get a new build error: In file included from include/linux/kernel.h:13:0, from include/linux/list.h:8, from include/linux/preempt.h:10, from include/linux/spinlock.h:50, from arch/x86/kernel/nmi.c:13: arch/x86/kernel/nmi.c: In function 'nmi_max_handler': include/linux/printk.h:375:9: error: type defaults to 'int' in declaration of 'DEFINE_RATELIMIT_STATE' [-Werror=implicit-int] static DEFINE_RATELIMIT_STATE(_rs, \ ^ arch/x86/kernel/nmi.c:110:2: note: in expansion of macro 'printk_ratelimited' printk_ratelimited(KERN_INFO ^~~~~~~~~~~~~~~~~~ This was working before the rtc rework series because linux/ratelimit.h was included implictly through asm/mach_traps.h -> asm/mc146818rtc.h -> linux/mc146818rtc.h -> linux/rtc.h -> linux/device.h. We clearly shouldn't rely on this indirect inclusion, so this adds an explicit #include in the file that needs it. Signed-off-by: Arnd Bergmann Reported-by: kbuild test robot Acked-by: Thomas Gleixner Fixes: 5ab788d73832 ("rtc: cmos: move mc146818rtc code out of asm-generic/rtc.h") Signed-off-by: Alexandre Belloni --- arch/x86/kernel/nmi.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c index 04b132a767f1..bfe4d6c96fbd 100644 --- a/arch/x86/kernel/nmi.c +++ b/arch/x86/kernel/nmi.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include -- cgit v1.2.3 From 3a3d1a4e32ab47323d7b8c8b7631a8d36a3098b2 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Wed, 8 Jun 2016 10:21:23 +0100 Subject: pwm: Add PWM capture support Supply a PWM capture callback op in order to pass back information obtained by running analysis on a PWM signal. This would normally (at least during testing) be called from the sysfs routines with a view to printing out PWM capture data which has been encoded into a string. Signed-off-by: Lee Jones [thierry.reding@gmail.com: make capture data unsigned int for symmetry] Signed-off-by: Thierry Reding --- drivers/pwm/core.c | 27 +++++++++++++++++++++++++++ include/linux/pwm.h | 24 ++++++++++++++++++++++++ 2 files changed, 51 insertions(+) diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c index dba3843c53b8..8f40604046d6 100644 --- a/drivers/pwm/core.c +++ b/drivers/pwm/core.c @@ -524,6 +524,33 @@ int pwm_apply_state(struct pwm_device *pwm, struct pwm_state *state) } EXPORT_SYMBOL_GPL(pwm_apply_state); +/** + * pwm_capture() - capture and report a PWM signal + * @pwm: PWM device + * @result: structure to fill with capture result + * @timeout: time to wait, in milliseconds, before giving up on capture + * + * Returns: 0 on success or a negative error code on failure. + */ +int pwm_capture(struct pwm_device *pwm, struct pwm_capture *result, + unsigned long timeout) +{ + int err; + + if (!pwm || !pwm->chip->ops) + return -EINVAL; + + if (!pwm->chip->ops->capture) + return -ENOSYS; + + mutex_lock(&pwm_lock); + err = pwm->chip->ops->capture(pwm->chip, pwm, result, timeout); + mutex_unlock(&pwm_lock); + + return err; +} +EXPORT_SYMBOL_GPL(pwm_capture); + /** * pwm_adjust_config() - adjust the current PWM config to the PWM arguments * @pwm: PWM device diff --git a/include/linux/pwm.h b/include/linux/pwm.h index 17018f3c066e..8402b5dd3e06 100644 --- a/include/linux/pwm.h +++ b/include/linux/pwm.h @@ -5,7 +5,9 @@ #include #include +struct pwm_capture; struct seq_file; + struct pwm_chip; /** @@ -153,6 +155,7 @@ static inline void pwm_get_args(const struct pwm_device *pwm, * @free: optional hook for freeing a PWM * @config: configure duty cycles and period length for this PWM * @set_polarity: configure the polarity of this PWM + * @capture: capture and report PWM signal * @enable: enable PWM output toggling * @disable: disable PWM output toggling * @apply: atomically apply a new PWM config. The state argument @@ -172,6 +175,8 @@ struct pwm_ops { int duty_ns, int period_ns); int (*set_polarity)(struct pwm_chip *chip, struct pwm_device *pwm, enum pwm_polarity polarity); + int (*capture)(struct pwm_chip *chip, struct pwm_device *pwm, + struct pwm_capture *result, unsigned long timeout); int (*enable)(struct pwm_chip *chip, struct pwm_device *pwm); void (*disable)(struct pwm_chip *chip, struct pwm_device *pwm); int (*apply)(struct pwm_chip *chip, struct pwm_device *pwm, @@ -212,6 +217,16 @@ struct pwm_chip { bool can_sleep; }; +/** + * struct pwm_capture - PWM capture data + * @period: period of the PWM signal (in nanoseconds) + * @duty_cycle: duty cycle of the PWM signal (in nanoseconds) + */ +struct pwm_capture { + unsigned int period; + unsigned int duty_cycle; +}; + #if IS_ENABLED(CONFIG_PWM) /* PWM user APIs */ struct pwm_device *pwm_request(int pwm_id, const char *label); @@ -322,6 +337,8 @@ static inline void pwm_disable(struct pwm_device *pwm) /* PWM provider APIs */ +int pwm_capture(struct pwm_device *pwm, struct pwm_capture *result, + unsigned long timeout); int pwm_set_chip_data(struct pwm_device *pwm, void *data); void *pwm_get_chip_data(struct pwm_device *pwm); @@ -373,6 +390,13 @@ static inline int pwm_config(struct pwm_device *pwm, int duty_ns, return -EINVAL; } +static inline int pwm_capture(struct pwm_device *pwm, + struct pwm_capture *result, + unsigned long timeout) +{ + return -EINVAL; +} + static inline int pwm_set_polarity(struct pwm_device *pwm, enum pwm_polarity polarity) { -- cgit v1.2.3 From 1a366fe9153f445e950a7a344932b7419aa83094 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Wed, 8 Jun 2016 10:21:25 +0100 Subject: pwm: sysfs: Add PWM capture support Allow a user to read PWM capture results from sysfs. To start a capture and read the result, simply read the file: $ cat $PWMCHIP/capture The output format is " ". Signed-off-by: Lee Jones Signed-off-by: Thierry Reding --- Documentation/ABI/testing/sysfs-class-pwm | 9 +++++++++ drivers/pwm/sysfs.c | 17 +++++++++++++++++ 2 files changed, 26 insertions(+) diff --git a/Documentation/ABI/testing/sysfs-class-pwm b/Documentation/ABI/testing/sysfs-class-pwm index c479d77b67c5..c20e61354561 100644 --- a/Documentation/ABI/testing/sysfs-class-pwm +++ b/Documentation/ABI/testing/sysfs-class-pwm @@ -77,3 +77,12 @@ Description: Enable/disable the PWM signal. 0 is disabled 1 is enabled + +What: /sys/class/pwm/pwmchipN/pwmX/capture +Date: June 2016 +KernelVersion: 4.8 +Contact: Lee Jones +Description: + Capture information about a PWM signal. The output format is a + pair unsigned integers (period and duty cycle), separated by a + single space. diff --git a/drivers/pwm/sysfs.c b/drivers/pwm/sysfs.c index d98599249a05..c3b1b563480e 100644 --- a/drivers/pwm/sysfs.c +++ b/drivers/pwm/sysfs.c @@ -208,16 +208,33 @@ static ssize_t polarity_store(struct device *child, return ret ? : size; } +static ssize_t capture_show(struct device *child, + struct device_attribute *attr, + char *buf) +{ + struct pwm_device *pwm = child_to_pwm_device(child); + struct pwm_capture result; + int ret; + + ret = pwm_capture(pwm, &result, jiffies_to_msecs(HZ)); + if (ret) + return ret; + + return sprintf(buf, "%u %u\n", result.period, result.duty_cycle); +} + static DEVICE_ATTR_RW(period); static DEVICE_ATTR_RW(duty_cycle); static DEVICE_ATTR_RW(enable); static DEVICE_ATTR_RW(polarity); +static DEVICE_ATTR_RO(capture); static struct attribute *pwm_attrs[] = { &dev_attr_period.attr, &dev_attr_duty_cycle.attr, &dev_attr_enable.attr, &dev_attr_polarity.attr, + &dev_attr_capture.attr, NULL }; ATTRIBUTE_GROUPS(pwm); -- cgit v1.2.3 From 3d4e79949cf9e8032f0cd1f91564c6ce3304755c Mon Sep 17 00:00:00 2001 From: Hariprasad S Date: Fri, 10 Jun 2016 01:05:11 +0530 Subject: RDMA/iw_cxgb4: only read markers_enabled mod param once markers_enabled should be read only once during MPA negotiation. The present code does read markers_enabled twice during negotiation which results in setting wrong recv/xmit markers if the markers_enabled is changed in the middle of negotiation. With this change the markers_enabled is read only once during MPA negotiation. recv markers are set based on markers enabled module parameter and xmit markers are set based on markers flag from the MPA_START_REQ/MPA_START_REP. Signed-off-by: Steve Wise Signed-off-by: Hariprasad Shenai Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/cm.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index a3a67216bce6..2ba3403d8101 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -992,9 +992,19 @@ static int send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb, mpa = (struct mpa_message *)(req + 1); memcpy(mpa->key, MPA_KEY_REQ, sizeof(mpa->key)); - mpa->flags = (crc_enabled ? MPA_CRC : 0) | - (markers_enabled ? MPA_MARKERS : 0) | - (mpa_rev_to_use == 2 ? MPA_ENHANCED_RDMA_CONN : 0); + + mpa->flags = 0; + if (crc_enabled) + mpa->flags |= MPA_CRC; + if (markers_enabled) { + mpa->flags |= MPA_MARKERS; + ep->mpa_attr.recv_marker_enabled = 1; + } else { + ep->mpa_attr.recv_marker_enabled = 0; + } + if (mpa_rev_to_use == 2) + mpa->flags |= MPA_ENHANCED_RDMA_CONN; + mpa->private_data_size = htons(ep->plen); mpa->revision = mpa_rev_to_use; if (mpa_rev_to_use == 1) { @@ -1169,8 +1179,11 @@ static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen) mpa = (struct mpa_message *)(req + 1); memset(mpa, 0, sizeof(*mpa)); memcpy(mpa->key, MPA_KEY_REP, sizeof(mpa->key)); - mpa->flags = (ep->mpa_attr.crc_enabled ? MPA_CRC : 0) | - (markers_enabled ? MPA_MARKERS : 0); + mpa->flags = 0; + if (ep->mpa_attr.crc_enabled) + mpa->flags |= MPA_CRC; + if (ep->mpa_attr.recv_marker_enabled) + mpa->flags |= MPA_MARKERS; mpa->revision = ep->mpa_attr.version; mpa->private_data_size = htons(plen); @@ -1555,7 +1568,6 @@ static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) */ __state_set(&ep->com, FPDU_MODE); ep->mpa_attr.crc_enabled = (mpa->flags & MPA_CRC) | crc_enabled ? 1 : 0; - ep->mpa_attr.recv_marker_enabled = markers_enabled; ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0; ep->mpa_attr.version = mpa->revision; ep->mpa_attr.p2p_type = FW_RI_INIT_P2PTYPE_DISABLED; -- cgit v1.2.3 From 68cebcab59fbfbab5827fcc44e04cd31da02041b Mon Sep 17 00:00:00 2001 From: Hariprasad S Date: Fri, 10 Jun 2016 01:05:12 +0530 Subject: RDMA/iw_cxgb4: allocate enough space for debugfs "qps" dump With IPv6 addresses, the "qps" debugfs is running out of space and truncating the output. Bump the required size accordingly. Signed-off-by: Steve Wise Signed-off-by: Hariprasad Shenai Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c index ae2e8b23d2dd..071d7332ec06 100644 --- a/drivers/infiniband/hw/cxgb4/device.c +++ b/drivers/infiniband/hw/cxgb4/device.c @@ -317,7 +317,7 @@ static int qp_open(struct inode *inode, struct file *file) idr_for_each(&qpd->devp->qpidr, count_idrs, &count); spin_unlock_irq(&qpd->devp->lock); - qpd->bufsize = count * 128; + qpd->bufsize = count * 180; qpd->buf = vmalloc(qpd->bufsize); if (!qpd->buf) { kfree(qpd); -- cgit v1.2.3 From bce2841f5a0dd9caf327fb1bc0051ea09a9e7dfb Mon Sep 17 00:00:00 2001 From: Hariprasad S Date: Fri, 10 Jun 2016 01:05:13 +0530 Subject: RDMA/iw_cxgb4: clean up c4iw_reject_cr() Get rid of unneeded code, and refactor things a bit. For MPA version 0 we abort the connection. For > 0, we attempt to send an MPA_START/REJECT Reply, and then disconnect gracefully. If the send of the MPA message fails, then we abort the connection. We can ignore c4iw_ep_disconnect() errors here because it will clean up the endpoint if there are failures. Signed-off-by: Steve Wise Signed-off-by: Hariprasad Shenai Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/cm.c | 19 ++++++++----------- 1 file changed, 8 insertions(+), 11 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index 2ba3403d8101..25236fc7dc92 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -3037,9 +3037,9 @@ out: int c4iw_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) { - int err = 0; - int disconnect = 0; + int abort; struct c4iw_ep *ep = to_ep(cm_id); + PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); mutex_lock(&ep->com.mutex); @@ -3050,16 +3050,13 @@ int c4iw_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) } set_bit(ULP_REJECT, &ep->com.history); if (mpa_rev == 0) - disconnect = 2; - else { - err = send_mpa_reject(ep, pdata, pdata_len); - disconnect = 1; - } + abort = 1; + else + abort = send_mpa_reject(ep, pdata, pdata_len); mutex_unlock(&ep->com.mutex); - if (disconnect) { - stop_ep_timer(ep); - err = c4iw_ep_disconnect(ep, disconnect == 2, GFP_KERNEL); - } + + stop_ep_timer(ep); + c4iw_ep_disconnect(ep, abort != 0, GFP_KERNEL); c4iw_put_ep(&ep->com); return 0; } -- cgit v1.2.3 From 4c72efefd9af015459ed63b164ff4b13cbfa3f61 Mon Sep 17 00:00:00 2001 From: Hariprasad S Date: Fri, 10 Jun 2016 01:05:14 +0530 Subject: RDMA/iw_cxgb4: Add missing error codes for act open cmd Signed-off-by: Steve Wise Signed-off-by: Hariprasad Shenai Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/cm.c | 11 ++++++++--- drivers/net/ethernet/chelsio/cxgb4/t4_msg.h | 2 ++ 2 files changed, 10 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index 25236fc7dc92..4819314ef510 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -2016,12 +2016,17 @@ static int send_fw_act_open_req(struct c4iw_ep *ep, unsigned int atid) } /* - * Return whether a failed active open has allocated a TID + * Some of the error codes above implicitly indicate that there is no TID + * allocated with the result of an ACT_OPEN. We use this predicate to make + * that explicit. */ static inline int act_open_has_tid(int status) { - return status != CPL_ERR_TCAM_FULL && status != CPL_ERR_CONN_EXIST && - status != CPL_ERR_ARP_MISS; + return (status != CPL_ERR_TCAM_PARITY && + status != CPL_ERR_TCAM_MISS && + status != CPL_ERR_TCAM_FULL && + status != CPL_ERR_CONN_EXIST_SYNRECV && + status != CPL_ERR_CONN_EXIST); } /* Returns whether a CPL status conveys negative advice. diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_msg.h b/drivers/net/ethernet/chelsio/cxgb4/t4_msg.h index 4705e2dea423..e0ebe1378cb2 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/t4_msg.h +++ b/drivers/net/ethernet/chelsio/cxgb4/t4_msg.h @@ -104,6 +104,8 @@ enum { enum CPL_error { CPL_ERR_NONE = 0, + CPL_ERR_TCAM_PARITY = 1, + CPL_ERR_TCAM_MISS = 2, CPL_ERR_TCAM_FULL = 3, CPL_ERR_BAD_LENGTH = 15, CPL_ERR_BAD_ROUTE = 18, -- cgit v1.2.3 From 4a740838bf44ca4b03e26ff79aedb86b9d8cace4 Mon Sep 17 00:00:00 2001 From: Hariprasad S Date: Fri, 10 Jun 2016 01:05:15 +0530 Subject: RDMA/iw_cxgb4: Low resource fixes for connection manager Pre-allocate buffers for sending various control messages to close connection, abort connection, etc so that we gracefully handle connections when system is running out of memory. Signed-off-by: Steve Wise Signed-off-by: Hariprasad Shenai Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/cm.c | 127 +++++++++++++++++++++++---------- drivers/infiniband/hw/cxgb4/iw_cxgb4.h | 19 +++++ drivers/infiniband/hw/cxgb4/qp.c | 10 +-- 3 files changed, 114 insertions(+), 42 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index 4819314ef510..81211dbb76df 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -294,6 +294,25 @@ static void state_set(struct c4iw_ep_common *epc, enum c4iw_ep_state new) return; } +static int alloc_ep_skb_list(struct sk_buff_head *ep_skb_list, int size) +{ + struct sk_buff *skb; + unsigned int i; + size_t len; + + len = roundup(sizeof(union cpl_wr_size), 16); + for (i = 0; i < size; i++) { + skb = alloc_skb(len, GFP_KERNEL); + if (!skb) + goto fail; + skb_queue_tail(ep_skb_list, skb); + } + return 0; +fail: + skb_queue_purge(ep_skb_list); + return -ENOMEM; +} + static void *alloc_ep(int size, gfp_t gfp) { struct c4iw_ep_common *epc; @@ -384,6 +403,8 @@ void _c4iw_free_ep(struct kref *kref) if (ep->mpa_skb) kfree_skb(ep->mpa_skb); } + if (!skb_queue_empty(&ep->com.ep_skb_list)) + skb_queue_purge(&ep->com.ep_skb_list); kfree(ep); } @@ -620,25 +641,27 @@ static void abort_arp_failure(void *handle, struct sk_buff *skb) } } -static int send_flowc(struct c4iw_ep *ep, struct sk_buff *skb) +static int send_flowc(struct c4iw_ep *ep) { - unsigned int flowclen = 80; struct fw_flowc_wr *flowc; + struct sk_buff *skb = skb_dequeue(&ep->com.ep_skb_list); int i; u16 vlan = ep->l2t->vlan; int nparams; + if (WARN_ON(!skb)) + return -ENOMEM; + if (vlan == CPL_L2T_VLAN_NONE) nparams = 8; else nparams = 9; - skb = get_skb(skb, flowclen, GFP_KERNEL); - flowc = (struct fw_flowc_wr *)__skb_put(skb, flowclen); + flowc = (struct fw_flowc_wr *)__skb_put(skb, FLOWC_LEN); flowc->op_to_nparams = cpu_to_be32(FW_WR_OP_V(FW_FLOWC_WR) | FW_FLOWC_WR_NPARAMS_V(nparams)); - flowc->flowid_len16 = cpu_to_be32(FW_WR_LEN16_V(DIV_ROUND_UP(flowclen, + flowc->flowid_len16 = cpu_to_be32(FW_WR_LEN16_V(DIV_ROUND_UP(FLOWC_LEN, 16)) | FW_WR_FLOWID_V(ep->hwtid)); flowc->mnemval[0].mnemonic = FW_FLOWC_MNEM_PFNVFN; @@ -679,18 +702,16 @@ static int send_flowc(struct c4iw_ep *ep, struct sk_buff *skb) return c4iw_ofld_send(&ep->com.dev->rdev, skb); } -static int send_halfclose(struct c4iw_ep *ep, gfp_t gfp) +static int send_halfclose(struct c4iw_ep *ep) { struct cpl_close_con_req *req; - struct sk_buff *skb; + struct sk_buff *skb = skb_dequeue(&ep->com.ep_skb_list); int wrlen = roundup(sizeof *req, 16); PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); - skb = get_skb(NULL, wrlen, gfp); - if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__); + if (WARN_ON(!skb)) return -ENOMEM; - } + set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx); t4_set_arp_err_handler(skb, NULL, arp_failure_discard); req = (struct cpl_close_con_req *) skb_put(skb, wrlen); @@ -701,26 +722,24 @@ static int send_halfclose(struct c4iw_ep *ep, gfp_t gfp) return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); } -static int send_abort(struct c4iw_ep *ep, struct sk_buff *skb, gfp_t gfp) +static int send_abort(struct c4iw_ep *ep) { struct cpl_abort_req *req; int wrlen = roundup(sizeof *req, 16); + struct sk_buff *req_skb = skb_dequeue(&ep->com.ep_skb_list); PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); - skb = get_skb(skb, wrlen, gfp); - if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb.\n", - __func__); + if (WARN_ON(!req_skb)) return -ENOMEM; - } - set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx); - t4_set_arp_err_handler(skb, ep, abort_arp_failure); - req = (struct cpl_abort_req *) skb_put(skb, wrlen); + + set_wr_txq(req_skb, CPL_PRIORITY_DATA, ep->txq_idx); + t4_set_arp_err_handler(req_skb, ep, abort_arp_failure); + req = (struct cpl_abort_req *)skb_put(req_skb, wrlen); memset(req, 0, wrlen); INIT_TP_WR(req, ep->hwtid); OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_ABORT_REQ, ep->hwtid)); req->cmd = CPL_ABORT_SEND_RST; - return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); + return c4iw_l2t_send(&ep->com.dev->rdev, req_skb, ep->l2t); } static void best_mtu(const unsigned short *mtus, unsigned short mtu, @@ -1261,7 +1280,7 @@ static int act_establish(struct c4iw_dev *dev, struct sk_buff *skb) set_bit(ACT_ESTAB, &ep->com.history); /* start MPA negotiation */ - ret = send_flowc(ep, NULL); + ret = send_flowc(ep); if (ret) goto err; if (ep->retry_with_mpa_v1) @@ -2147,6 +2166,7 @@ out: static int c4iw_reconnect(struct c4iw_ep *ep) { int err = 0; + int size = 0; struct sockaddr_in *laddr = (struct sockaddr_in *) &ep->com.cm_id->m_local_addr; struct sockaddr_in *raddr = (struct sockaddr_in *) @@ -2162,6 +2182,21 @@ static int c4iw_reconnect(struct c4iw_ep *ep) init_timer(&ep->timer); c4iw_init_wr_wait(&ep->com.wr_wait); + /* When MPA revision is different on nodes, the node with MPA_rev=2 + * tries to reconnect with MPA_rev 1 for the same EP through + * c4iw_reconnect(), where the same EP is assigned with new tid for + * further connection establishment. As we are using the same EP pointer + * for reconnect, few skbs are used during the previous c4iw_connect(), + * which leaves the EP with inadequate skbs for further + * c4iw_reconnect(), Further causing an assert BUG_ON() due to empty + * skb_list() during peer_abort(). Allocate skbs which is already used. + */ + size = (CN_MAX_CON_BUF - skb_queue_len(&ep->com.ep_skb_list)); + if (alloc_ep_skb_list(&ep->com.ep_skb_list, size)) { + err = -ENOMEM; + goto fail1; + } + /* * Allocate an active TID to initiate a TCP connection. */ @@ -2227,6 +2262,7 @@ fail2: * response of 1st connect request. */ connect_reply_upcall(ep, -ECONNRESET); +fail1: c4iw_put_ep(&ep->com); out: return err; @@ -2593,6 +2629,10 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) if (peer_mss && child_ep->mtu > (peer_mss + hdrs)) child_ep->mtu = peer_mss + hdrs; + skb_queue_head_init(&child_ep->com.ep_skb_list); + if (alloc_ep_skb_list(&child_ep->com.ep_skb_list, CN_MAX_CON_BUF)) + goto fail; + state_set(&child_ep->com, CONNECTING); child_ep->com.dev = dev; child_ep->com.cm_id = NULL; @@ -2657,6 +2697,8 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) (const u32 *)&sin6->sin6_addr.s6_addr, 1); } goto out; +fail: + c4iw_put_ep(&child_ep->com); reject: reject_cr(dev, hwtid, skb); if (parent_ep) @@ -2687,7 +2729,7 @@ static int pass_establish(struct c4iw_dev *dev, struct sk_buff *skb) ep->com.state = MPA_REQ_WAIT; start_ep_timer(ep); set_bit(PASS_ESTAB, &ep->com.history); - ret = send_flowc(ep, skb); + ret = send_flowc(ep); mutex_unlock(&ep->com.mutex); if (ret) c4iw_ep_disconnect(ep, 1, GFP_KERNEL); @@ -2888,10 +2930,8 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) } mutex_unlock(&ep->com.mutex); - rpl_skb = get_skb(skb, sizeof(*rpl), GFP_KERNEL); - if (!rpl_skb) { - printk(KERN_ERR MOD "%s - cannot allocate skb!\n", - __func__); + rpl_skb = skb_dequeue(&ep->com.ep_skb_list); + if (WARN_ON(!rpl_skb)) { release = 1; goto out; } @@ -3262,6 +3302,13 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) err = -ENOMEM; goto out; } + + skb_queue_head_init(&ep->com.ep_skb_list); + if (alloc_ep_skb_list(&ep->com.ep_skb_list, CN_MAX_CON_BUF)) { + err = -ENOMEM; + goto fail1; + } + init_timer(&ep->timer); ep->plen = conn_param->private_data_len; if (ep->plen) @@ -3280,7 +3327,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (!ep->com.qp) { PDBG("%s qpn 0x%x not found!\n", __func__, conn_param->qpn); err = -EINVAL; - goto fail1; + goto fail2; } ref_qp(ep); PDBG("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn, @@ -3293,7 +3340,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (ep->atid == -1) { printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__); err = -ENOMEM; - goto fail1; + goto fail2; } insert_handle(dev, &dev->atid_idr, ep, ep->atid); @@ -3317,7 +3364,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (raddr->sin_addr.s_addr == htonl(INADDR_ANY)) { err = pick_local_ipaddrs(dev, cm_id); if (err) - goto fail1; + goto fail2; } /* find a route */ @@ -3337,7 +3384,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (ipv6_addr_type(&raddr6->sin6_addr) == IPV6_ADDR_ANY) { err = pick_local_ip6addrs(dev, cm_id); if (err) - goto fail1; + goto fail2; } /* find a route */ @@ -3353,14 +3400,14 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (!ep->dst) { printk(KERN_ERR MOD "%s - cannot find route.\n", __func__); err = -EHOSTUNREACH; - goto fail2; + goto fail3; } err = import_ep(ep, iptype, ra, ep->dst, ep->com.dev, true, ep->com.dev->rdev.lldi.adapter_type, cm_id->tos); if (err) { printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __func__); - goto fail3; + goto fail4; } PDBG("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n", @@ -3376,13 +3423,15 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) goto out; cxgb4_l2t_release(ep->l2t); -fail3: +fail4: dst_release(ep->dst); -fail2: +fail3: remove_handle(ep->com.dev, &ep->com.dev->atid_idr, ep->atid); cxgb4_free_atid(ep->com.dev->rdev.lldi.tids, ep->atid); -fail1: +fail2: + skb_queue_purge(&ep->com.ep_skb_list); deref_cm_id(&ep->com); +fail1: c4iw_put_ep(&ep->com); out: return err; @@ -3475,6 +3524,7 @@ int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog) err = -ENOMEM; goto fail1; } + skb_queue_head_init(&ep->com.ep_skb_list); PDBG("%s ep %p\n", __func__, ep); ep->com.cm_id = cm_id; ref_cm_id(&ep->com); @@ -3591,6 +3641,7 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) case MPA_REQ_RCVD: case MPA_REP_SENT: case FPDU_MODE: + case CONNECTING: close = 1; if (abrupt) ep->com.state = ABORTING; @@ -3625,10 +3676,10 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) if (abrupt) { set_bit(EP_DISC_ABORT, &ep->com.history); close_complete_upcall(ep, -ECONNRESET); - ret = send_abort(ep, NULL, gfp); + ret = send_abort(ep); } else { set_bit(EP_DISC_CLOSE, &ep->com.history); - ret = send_halfclose(ep, gfp); + ret = send_halfclose(ep); } if (ret) { set_bit(EP_DISC_FAIL, &ep->com.history); diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h index f6f34a75af27..ce42d09ef9f7 100644 --- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h +++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h @@ -789,10 +789,29 @@ enum c4iw_ep_history { CM_ID_DEREFED = 28, }; +enum conn_pre_alloc_buffers { + CN_ABORT_REQ_BUF, + CN_ABORT_RPL_BUF, + CN_CLOSE_CON_REQ_BUF, + CN_DESTROY_BUF, + CN_FLOWC_BUF, + CN_MAX_CON_BUF +}; + +#define FLOWC_LEN 80 +union cpl_wr_size { + struct cpl_abort_req abrt_req; + struct cpl_abort_rpl abrt_rpl; + struct fw_ri_wr ri_req; + struct cpl_close_con_req close_req; + char flowc_buf[FLOWC_LEN]; +}; + struct c4iw_ep_common { struct iw_cm_id *cm_id; struct c4iw_qp *qp; struct c4iw_dev *dev; + struct sk_buff_head ep_skb_list; enum c4iw_ep_state state; struct kref kref; struct mutex mutex; diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c index e8993e49b8b3..b3441af69573 100644 --- a/drivers/infiniband/hw/cxgb4/qp.c +++ b/drivers/infiniband/hw/cxgb4/qp.c @@ -1081,9 +1081,10 @@ static void post_terminate(struct c4iw_qp *qhp, struct t4_cqe *err_cqe, PDBG("%s qhp %p qid 0x%x tid %u\n", __func__, qhp, qhp->wq.sq.qid, qhp->ep->hwtid); - skb = alloc_skb(sizeof *wqe, gfp); - if (!skb) + skb = skb_dequeue(&qhp->ep->com.ep_skb_list); + if (WARN_ON(!skb)) return; + set_wr_txq(skb, CPL_PRIORITY_DATA, qhp->ep->txq_idx); wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe)); @@ -1202,9 +1203,10 @@ static int rdma_fini(struct c4iw_dev *rhp, struct c4iw_qp *qhp, PDBG("%s qhp %p qid 0x%x tid %u\n", __func__, qhp, qhp->wq.sq.qid, ep->hwtid); - skb = alloc_skb(sizeof *wqe, GFP_KERNEL); - if (!skb) + skb = skb_dequeue(&ep->com.ep_skb_list); + if (WARN_ON(!skb)) return -ENOMEM; + set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx); wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe)); -- cgit v1.2.3 From 0f8ab0b6e91b4d5302ceee5ec12ce54d81297635 Mon Sep 17 00:00:00 2001 From: Hariprasad S Date: Fri, 10 Jun 2016 01:05:16 +0530 Subject: RDMA/iw_cxgb4: Low resource fixes for Memory registration Pre-allocate buffers for deregistering memory region and memory window during RDMA connection close, when system is running out of memory. Signed-off-by: Steve Wise Signed-off-by: Hariprasad Shenai Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/iw_cxgb4.h | 2 + drivers/infiniband/hw/cxgb4/mem.c | 111 ++++++++++++++++++++++----------- 2 files changed, 76 insertions(+), 37 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h index ce42d09ef9f7..e94ec0d766bc 100644 --- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h +++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h @@ -384,6 +384,7 @@ struct c4iw_mr { struct ib_mr ibmr; struct ib_umem *umem; struct c4iw_dev *rhp; + struct sk_buff *dereg_skb; u64 kva; struct tpt_attributes attr; u64 *mpl; @@ -400,6 +401,7 @@ static inline struct c4iw_mr *to_c4iw_mr(struct ib_mr *ibmr) struct c4iw_mw { struct ib_mw ibmw; struct c4iw_dev *rhp; + struct sk_buff *dereg_skb; u64 kva; struct tpt_attributes attr; }; diff --git a/drivers/infiniband/hw/cxgb4/mem.c b/drivers/infiniband/hw/cxgb4/mem.c index 55d0651ee4de..5d0aa55e82d4 100644 --- a/drivers/infiniband/hw/cxgb4/mem.c +++ b/drivers/infiniband/hw/cxgb4/mem.c @@ -59,9 +59,9 @@ static int mr_exceeds_hw_limits(struct c4iw_dev *dev, u64 length) } static int _c4iw_write_mem_dma_aligned(struct c4iw_rdev *rdev, u32 addr, - u32 len, dma_addr_t data, int wait) + u32 len, dma_addr_t data, + int wait, struct sk_buff *skb) { - struct sk_buff *skb; struct ulp_mem_io *req; struct ulptx_sgl *sgl; u8 wr_len; @@ -74,9 +74,11 @@ static int _c4iw_write_mem_dma_aligned(struct c4iw_rdev *rdev, u32 addr, c4iw_init_wr_wait(&wr_wait); wr_len = roundup(sizeof(*req) + sizeof(*sgl), 16); - skb = alloc_skb(wr_len, GFP_KERNEL); - if (!skb) - return -ENOMEM; + if (!skb) { + skb = alloc_skb(wr_len, GFP_KERNEL | __GFP_NOFAIL); + if (!skb) + return -ENOMEM; + } set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); req = (struct ulp_mem_io *)__skb_put(skb, wr_len); @@ -108,9 +110,8 @@ static int _c4iw_write_mem_dma_aligned(struct c4iw_rdev *rdev, u32 addr, } static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len, - void *data) + void *data, struct sk_buff *skb) { - struct sk_buff *skb; struct ulp_mem_io *req; struct ulptx_idata *sc; u8 wr_len, *to_dp, *from_dp; @@ -134,9 +135,11 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len, wr_len = roundup(sizeof *req + sizeof *sc + roundup(copy_len, T4_ULPTX_MIN_IO), 16); - skb = alloc_skb(wr_len, GFP_KERNEL); - if (!skb) - return -ENOMEM; + if (!skb) { + skb = alloc_skb(wr_len, GFP_KERNEL | __GFP_NOFAIL); + if (!skb) + return -ENOMEM; + } set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); req = (struct ulp_mem_io *)__skb_put(skb, wr_len); @@ -173,6 +176,7 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len, memset(to_dp + copy_len, 0, T4_ULPTX_MIN_IO - (copy_len % T4_ULPTX_MIN_IO)); ret = c4iw_ofld_send(rdev, skb); + skb = NULL; if (ret) return ret; len -= C4IW_MAX_INLINE_SIZE; @@ -182,7 +186,8 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len, return ret; } -static int _c4iw_write_mem_dma(struct c4iw_rdev *rdev, u32 addr, u32 len, void *data) +static int _c4iw_write_mem_dma(struct c4iw_rdev *rdev, u32 addr, u32 len, + void *data, struct sk_buff *skb) { u32 remain = len; u32 dmalen; @@ -205,7 +210,7 @@ static int _c4iw_write_mem_dma(struct c4iw_rdev *rdev, u32 addr, u32 len, void * dmalen = T4_ULPTX_MAX_DMA; remain -= dmalen; ret = _c4iw_write_mem_dma_aligned(rdev, addr, dmalen, daddr, - !remain); + !remain, skb); if (ret) goto out; addr += dmalen >> 5; @@ -213,7 +218,7 @@ static int _c4iw_write_mem_dma(struct c4iw_rdev *rdev, u32 addr, u32 len, void * daddr += dmalen; } if (remain) - ret = _c4iw_write_mem_inline(rdev, addr, remain, data); + ret = _c4iw_write_mem_inline(rdev, addr, remain, data, skb); out: dma_unmap_single(&rdev->lldi.pdev->dev, save, len, DMA_TO_DEVICE); return ret; @@ -224,23 +229,25 @@ out: * If data is NULL, clear len byte of memory to zero. */ static int write_adapter_mem(struct c4iw_rdev *rdev, u32 addr, u32 len, - void *data) + void *data, struct sk_buff *skb) { if (is_t5(rdev->lldi.adapter_type) && use_dsgl) { if (len > inline_threshold) { - if (_c4iw_write_mem_dma(rdev, addr, len, data)) { + if (_c4iw_write_mem_dma(rdev, addr, len, data, skb)) { printk_ratelimited(KERN_WARNING "%s: dma map" " failure (non fatal)\n", pci_name(rdev->lldi.pdev)); return _c4iw_write_mem_inline(rdev, addr, len, - data); - } else + data, skb); + } else { return 0; + } } else - return _c4iw_write_mem_inline(rdev, addr, len, data); + return _c4iw_write_mem_inline(rdev, addr, + len, data, skb); } else - return _c4iw_write_mem_inline(rdev, addr, len, data); + return _c4iw_write_mem_inline(rdev, addr, len, data, skb); } /* @@ -253,7 +260,8 @@ static int write_tpt_entry(struct c4iw_rdev *rdev, u32 reset_tpt_entry, u32 *stag, u8 stag_state, u32 pdid, enum fw_ri_stag_type type, enum fw_ri_mem_perms perm, int bind_enabled, u32 zbva, u64 to, - u64 len, u8 page_size, u32 pbl_size, u32 pbl_addr) + u64 len, u8 page_size, u32 pbl_size, u32 pbl_addr, + struct sk_buff *skb) { int err; struct fw_ri_tpte tpt; @@ -307,7 +315,7 @@ static int write_tpt_entry(struct c4iw_rdev *rdev, u32 reset_tpt_entry, } err = write_adapter_mem(rdev, stag_idx + (rdev->lldi.vr->stag.start >> 5), - sizeof(tpt), &tpt); + sizeof(tpt), &tpt, skb); if (reset_tpt_entry) { c4iw_put_resource(&rdev->resource.tpt_table, stag_idx); @@ -327,28 +335,29 @@ static int write_pbl(struct c4iw_rdev *rdev, __be64 *pbl, __func__, pbl_addr, rdev->lldi.vr->pbl.start, pbl_size); - err = write_adapter_mem(rdev, pbl_addr >> 5, pbl_size << 3, pbl); + err = write_adapter_mem(rdev, pbl_addr >> 5, pbl_size << 3, pbl, NULL); return err; } static int dereg_mem(struct c4iw_rdev *rdev, u32 stag, u32 pbl_size, - u32 pbl_addr) + u32 pbl_addr, struct sk_buff *skb) { return write_tpt_entry(rdev, 1, &stag, 0, 0, 0, 0, 0, 0, 0UL, 0, 0, - pbl_size, pbl_addr); + pbl_size, pbl_addr, skb); } static int allocate_window(struct c4iw_rdev *rdev, u32 * stag, u32 pdid) { *stag = T4_STAG_UNSET; return write_tpt_entry(rdev, 0, stag, 0, pdid, FW_RI_STAG_MW, 0, 0, 0, - 0UL, 0, 0, 0, 0); + 0UL, 0, 0, 0, 0, NULL); } -static int deallocate_window(struct c4iw_rdev *rdev, u32 stag) +static int deallocate_window(struct c4iw_rdev *rdev, u32 stag, + struct sk_buff *skb) { return write_tpt_entry(rdev, 1, &stag, 0, 0, 0, 0, 0, 0, 0UL, 0, 0, 0, - 0); + 0, skb); } static int allocate_stag(struct c4iw_rdev *rdev, u32 *stag, u32 pdid, @@ -356,7 +365,7 @@ static int allocate_stag(struct c4iw_rdev *rdev, u32 *stag, u32 pdid, { *stag = T4_STAG_UNSET; return write_tpt_entry(rdev, 0, stag, 0, pdid, FW_RI_STAG_NSMR, 0, 0, 0, - 0UL, 0, 0, pbl_size, pbl_addr); + 0UL, 0, 0, pbl_size, pbl_addr, NULL); } static int finish_mem_reg(struct c4iw_mr *mhp, u32 stag) @@ -383,14 +392,16 @@ static int register_mem(struct c4iw_dev *rhp, struct c4iw_pd *php, mhp->attr.mw_bind_enable, mhp->attr.zbva, mhp->attr.va_fbo, mhp->attr.len ? mhp->attr.len : -1, shift - 12, - mhp->attr.pbl_size, mhp->attr.pbl_addr); + mhp->attr.pbl_size, mhp->attr.pbl_addr, NULL); if (ret) return ret; ret = finish_mem_reg(mhp, stag); - if (ret) + if (ret) { dereg_mem(&rhp->rdev, mhp->attr.stag, mhp->attr.pbl_size, - mhp->attr.pbl_addr); + mhp->attr.pbl_addr, mhp->dereg_skb); + mhp->dereg_skb = NULL; + } return ret; } @@ -423,6 +434,12 @@ struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc) if (!mhp) return ERR_PTR(-ENOMEM); + mhp->dereg_skb = alloc_skb(SGE_MAX_WR_LEN, GFP_KERNEL); + if (!mhp->dereg_skb) { + ret = -ENOMEM; + goto err0; + } + mhp->rhp = rhp; mhp->attr.pdid = php->pdid; mhp->attr.perms = c4iw_ib_to_tpt_access(acc); @@ -435,7 +452,8 @@ struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc) ret = write_tpt_entry(&rhp->rdev, 0, &stag, 1, php->pdid, FW_RI_STAG_NSMR, mhp->attr.perms, - mhp->attr.mw_bind_enable, 0, 0, ~0ULL, 0, 0, 0); + mhp->attr.mw_bind_enable, 0, 0, ~0ULL, 0, 0, 0, + NULL); if (ret) goto err1; @@ -445,8 +463,10 @@ struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc) return &mhp->ibmr; err2: dereg_mem(&rhp->rdev, mhp->attr.stag, mhp->attr.pbl_size, - mhp->attr.pbl_addr); + mhp->attr.pbl_addr, mhp->dereg_skb); err1: + kfree_skb(mhp->dereg_skb); +err0: kfree(mhp); return ERR_PTR(ret); } @@ -481,11 +501,18 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, if (!mhp) return ERR_PTR(-ENOMEM); + mhp->dereg_skb = alloc_skb(SGE_MAX_WR_LEN, GFP_KERNEL); + if (!mhp->dereg_skb) { + kfree(mhp); + return ERR_PTR(-ENOMEM); + } + mhp->rhp = rhp; mhp->umem = ib_umem_get(pd->uobject->context, start, length, acc, 0); if (IS_ERR(mhp->umem)) { err = PTR_ERR(mhp->umem); + kfree_skb(mhp->dereg_skb); kfree(mhp); return ERR_PTR(err); } @@ -550,6 +577,7 @@ err_pbl: err: ib_umem_release(mhp->umem); + kfree_skb(mhp->dereg_skb); kfree(mhp); return ERR_PTR(err); } @@ -572,8 +600,16 @@ struct ib_mw *c4iw_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, mhp = kzalloc(sizeof(*mhp), GFP_KERNEL); if (!mhp) return ERR_PTR(-ENOMEM); + + mhp->dereg_skb = alloc_skb(SGE_MAX_WR_LEN, GFP_KERNEL); + if (!mhp->dereg_skb) { + kfree(mhp); + return ERR_PTR(-ENOMEM); + } + ret = allocate_window(&rhp->rdev, &stag, php->pdid); if (ret) { + kfree(mhp->dereg_skb); kfree(mhp); return ERR_PTR(ret); } @@ -584,7 +620,8 @@ struct ib_mw *c4iw_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, mmid = (stag) >> 8; mhp->ibmw.rkey = stag; if (insert_handle(rhp, &rhp->mmidr, mhp, mmid)) { - deallocate_window(&rhp->rdev, mhp->attr.stag); + deallocate_window(&rhp->rdev, mhp->attr.stag, mhp->dereg_skb); + kfree(mhp->dereg_skb); kfree(mhp); return ERR_PTR(-ENOMEM); } @@ -602,7 +639,7 @@ int c4iw_dealloc_mw(struct ib_mw *mw) rhp = mhp->rhp; mmid = (mw->rkey) >> 8; remove_handle(rhp, &rhp->mmidr, mmid); - deallocate_window(&rhp->rdev, mhp->attr.stag); + deallocate_window(&rhp->rdev, mhp->attr.stag, mhp->dereg_skb); kfree(mhp); PDBG("%s ib_mw %p mmid 0x%x ptr %p\n", __func__, mw, mmid, mhp); return 0; @@ -666,7 +703,7 @@ struct ib_mr *c4iw_alloc_mr(struct ib_pd *pd, return &(mhp->ibmr); err3: dereg_mem(&rhp->rdev, stag, mhp->attr.pbl_size, - mhp->attr.pbl_addr); + mhp->attr.pbl_addr, mhp->dereg_skb); err2: c4iw_pblpool_free(&mhp->rhp->rdev, mhp->attr.pbl_addr, mhp->attr.pbl_size << 3); @@ -717,7 +754,7 @@ int c4iw_dereg_mr(struct ib_mr *ib_mr) dma_free_coherent(&mhp->rhp->rdev.lldi.pdev->dev, mhp->max_mpl_len, mhp->mpl, mhp->mpl_addr); dereg_mem(&rhp->rdev, mhp->attr.stag, mhp->attr.pbl_size, - mhp->attr.pbl_addr); + mhp->attr.pbl_addr, mhp->dereg_skb); if (mhp->attr.pbl_size) c4iw_pblpool_free(&mhp->rhp->rdev, mhp->attr.pbl_addr, mhp->attr.pbl_size << 3); -- cgit v1.2.3 From dd6b0241260d0f16c4051da1c5b09d4da8992ef2 Mon Sep 17 00:00:00 2001 From: Hariprasad S Date: Fri, 10 Jun 2016 01:05:17 +0530 Subject: RDMA/iw_cxgb4: Low resource fixes for Completion queue Pre-allocate buffers to deallocate completion queue, so that completion queue is deallocated during RDMA termination when system is running out of memory. Signed-off-by: Steve Wise Signed-off-by: Hariprasad Shenai Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/cq.c | 42 ++++++++++++++++++++-------------- drivers/infiniband/hw/cxgb4/iw_cxgb4.h | 1 + 2 files changed, 26 insertions(+), 17 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c index b0b955724458..812ab7278b8e 100644 --- a/drivers/infiniband/hw/cxgb4/cq.c +++ b/drivers/infiniband/hw/cxgb4/cq.c @@ -33,19 +33,15 @@ #include "iw_cxgb4.h" static int destroy_cq(struct c4iw_rdev *rdev, struct t4_cq *cq, - struct c4iw_dev_ucontext *uctx) + struct c4iw_dev_ucontext *uctx, struct sk_buff *skb) { struct fw_ri_res_wr *res_wr; struct fw_ri_res *res; int wr_len; struct c4iw_wr_wait wr_wait; - struct sk_buff *skb; int ret; wr_len = sizeof *res_wr + sizeof *res; - skb = alloc_skb(wr_len, GFP_KERNEL); - if (!skb) - return -ENOMEM; set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); res_wr = (struct fw_ri_res_wr *)__skb_put(skb, wr_len); @@ -863,7 +859,9 @@ int c4iw_destroy_cq(struct ib_cq *ib_cq) ucontext = ib_cq->uobject ? to_c4iw_ucontext(ib_cq->uobject->context) : NULL; destroy_cq(&chp->rhp->rdev, &chp->cq, - ucontext ? &ucontext->uctx : &chp->cq.rdev->uctx); + ucontext ? &ucontext->uctx : &chp->cq.rdev->uctx, + chp->destroy_skb); + chp->destroy_skb = NULL; kfree(chp); return 0; } @@ -879,7 +877,7 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, struct c4iw_cq *chp; struct c4iw_create_cq_resp uresp; struct c4iw_ucontext *ucontext = NULL; - int ret; + int ret, wr_len; size_t memsize, hwentries; struct c4iw_mm_entry *mm, *mm2; @@ -896,6 +894,13 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, if (!chp) return ERR_PTR(-ENOMEM); + wr_len = sizeof(struct fw_ri_res_wr) + sizeof(struct fw_ri_res); + chp->destroy_skb = alloc_skb(wr_len, GFP_KERNEL); + if (!chp->destroy_skb) { + ret = -ENOMEM; + goto err1; + } + if (ib_context) ucontext = to_c4iw_ucontext(ib_context); @@ -936,7 +941,7 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, ret = create_cq(&rhp->rdev, &chp->cq, ucontext ? &ucontext->uctx : &rhp->rdev.uctx); if (ret) - goto err1; + goto err2; chp->rhp = rhp; chp->cq.size--; /* status page */ @@ -947,15 +952,15 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, init_waitqueue_head(&chp->wait); ret = insert_handle(rhp, &rhp->cqidr, chp, chp->cq.cqid); if (ret) - goto err2; + goto err3; if (ucontext) { mm = kmalloc(sizeof *mm, GFP_KERNEL); if (!mm) - goto err3; + goto err4; mm2 = kmalloc(sizeof *mm2, GFP_KERNEL); if (!mm2) - goto err4; + goto err5; uresp.qid_mask = rhp->rdev.cqmask; uresp.cqid = chp->cq.cqid; @@ -970,7 +975,7 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, ret = ib_copy_to_udata(udata, &uresp, sizeof(uresp) - sizeof(uresp.reserved)); if (ret) - goto err5; + goto err6; mm->key = uresp.key; mm->addr = virt_to_phys(chp->cq.queue); @@ -986,15 +991,18 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, __func__, chp->cq.cqid, chp, chp->cq.size, chp->cq.memsize, (unsigned long long) chp->cq.dma_addr); return &chp->ibcq; -err5: +err6: kfree(mm2); -err4: +err5: kfree(mm); -err3: +err4: remove_handle(rhp, &rhp->cqidr, chp->cq.cqid); -err2: +err3: destroy_cq(&chp->rhp->rdev, &chp->cq, - ucontext ? &ucontext->uctx : &rhp->rdev.uctx); + ucontext ? &ucontext->uctx : &rhp->rdev.uctx, + chp->destroy_skb); +err2: + kfree_skb(chp->destroy_skb); err1: kfree(chp); return ERR_PTR(ret); diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h index e94ec0d766bc..c5f5881ab80c 100644 --- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h +++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h @@ -414,6 +414,7 @@ static inline struct c4iw_mw *to_c4iw_mw(struct ib_mw *ibmw) struct c4iw_cq { struct ib_cq ibcq; struct c4iw_dev *rhp; + struct sk_buff *destroy_skb; struct t4_cq cq; spinlock_t lock; spinlock_t comp_handler_lock; -- cgit v1.2.3 From 16bd020147abeb37dd32cc6442cee2d32b1c1af0 Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:47 +0300 Subject: net/mlx5: Export required core functions to support RSS In order to support RSS QPs, we need to create Ethernet based objects. This is done by create_rq, destroy_rq, create_rqt and destroy_rqt mlx5_core functions. We export these functions. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/net/ethernet/mellanox/mlx5/core/transobj.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/transobj.c b/drivers/net/ethernet/mellanox/mlx5/core/transobj.c index 03a5093ffeb7..28274a6fbafe 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/transobj.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/transobj.c @@ -85,6 +85,7 @@ int mlx5_core_create_rq(struct mlx5_core_dev *dev, u32 *in, int inlen, u32 *rqn) return err; } +EXPORT_SYMBOL(mlx5_core_create_rq); int mlx5_core_modify_rq(struct mlx5_core_dev *dev, u32 rqn, u32 *in, int inlen) { @@ -110,6 +111,7 @@ void mlx5_core_destroy_rq(struct mlx5_core_dev *dev, u32 rqn) mlx5_cmd_exec_check_status(dev, in, sizeof(in), out, sizeof(out)); } +EXPORT_SYMBOL(mlx5_core_destroy_rq); int mlx5_core_query_rq(struct mlx5_core_dev *dev, u32 rqn, u32 *out) { @@ -430,6 +432,7 @@ int mlx5_core_create_rqt(struct mlx5_core_dev *dev, u32 *in, int inlen, return err; } +EXPORT_SYMBOL(mlx5_core_create_rqt); int mlx5_core_modify_rqt(struct mlx5_core_dev *dev, u32 rqtn, u32 *in, int inlen) @@ -455,3 +458,4 @@ void mlx5_core_destroy_rqt(struct mlx5_core_dev *dev, u32 rqtn) mlx5_cmd_exec_check_status(dev, in, sizeof(in), out, sizeof(out)); } +EXPORT_SYMBOL(mlx5_core_destroy_rqt); -- cgit v1.2.3 From 5fd251c8b4c52da0d0916470a67fbb77b972125e Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:48 +0300 Subject: IB/core: Introduce Work Queue object and its verbs Introduce Work Queue object and its create/destroy/modify verbs. QP can be created without internal WQs "packaged" inside it, this QP can be configured to use "external" WQ object as its receive/send queue. WQ is a necessary component for RSS technology since RSS mechanism is supposed to distribute the traffic between multiple Receive Work Queues. WQ associated (many to one) with Completion Queue and it owns WQ properties (PD, WQ size, etc.). WQ has a type, this patch introduces the IB_WQT_RQ (i.e.receive queue), it may be extend to others such as IB_WQT_SQ. (send queue). WQ from type IB_WQT_RQ contains receive work requests. PD is an attribute of a work queue (i.e. send/receive queue), it's used by the hardware for security validation before scattering to a memory region which is pointed by the WQ. For that, an external WQ object needs a PD, letting the hardware makes that validation. When accessing a memory region that is pointed by the WQ its PD is used and not the QP's PD, this behavior is similar to a SRQ and a QP. WQ context is subject to a well-defined state transitions done by the modify_wq verb. When WQ is created its initial state becomes IB_WQS_RESET. >From IB_WQS_RESET it can be modified to itself or to IB_WQS_RDY. >From IB_WQS_RDY it can be modified to itself, to IB_WQS_RESET or to IB_WQS_ERR. >From IB_WQS_ERR it can be modified to IB_WQS_RESET. Note: transition to IB_WQS_ERR might occur implicitly in case there was some HW error. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/core/verbs.c | 82 +++++++++++++++++++++++++++++++++++++++++ include/rdma/ib_verbs.h | 56 +++++++++++++++++++++++++++- 2 files changed, 137 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 1d7d4cf442e3..c096cadc6e23 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -1554,6 +1554,88 @@ int ib_dealloc_xrcd(struct ib_xrcd *xrcd) } EXPORT_SYMBOL(ib_dealloc_xrcd); +/** + * ib_create_wq - Creates a WQ associated with the specified protection + * domain. + * @pd: The protection domain associated with the WQ. + * @wq_init_attr: A list of initial attributes required to create the + * WQ. If WQ creation succeeds, then the attributes are updated to + * the actual capabilities of the created WQ. + * + * wq_init_attr->max_wr and wq_init_attr->max_sge determine + * the requested size of the WQ, and set to the actual values allocated + * on return. + * If ib_create_wq() succeeds, then max_wr and max_sge will always be + * at least as large as the requested values. + */ +struct ib_wq *ib_create_wq(struct ib_pd *pd, + struct ib_wq_init_attr *wq_attr) +{ + struct ib_wq *wq; + + if (!pd->device->create_wq) + return ERR_PTR(-ENOSYS); + + wq = pd->device->create_wq(pd, wq_attr, NULL); + if (!IS_ERR(wq)) { + wq->event_handler = wq_attr->event_handler; + wq->wq_context = wq_attr->wq_context; + wq->wq_type = wq_attr->wq_type; + wq->cq = wq_attr->cq; + wq->device = pd->device; + wq->pd = pd; + wq->uobject = NULL; + atomic_inc(&pd->usecnt); + atomic_inc(&wq_attr->cq->usecnt); + atomic_set(&wq->usecnt, 0); + } + return wq; +} +EXPORT_SYMBOL(ib_create_wq); + +/** + * ib_destroy_wq - Destroys the specified WQ. + * @wq: The WQ to destroy. + */ +int ib_destroy_wq(struct ib_wq *wq) +{ + int err; + struct ib_cq *cq = wq->cq; + struct ib_pd *pd = wq->pd; + + if (atomic_read(&wq->usecnt)) + return -EBUSY; + + err = wq->device->destroy_wq(wq); + if (!err) { + atomic_dec(&pd->usecnt); + atomic_dec(&cq->usecnt); + } + return err; +} +EXPORT_SYMBOL(ib_destroy_wq); + +/** + * ib_modify_wq - Modifies the specified WQ. + * @wq: The WQ to modify. + * @wq_attr: On input, specifies the WQ attributes to modify. + * @wq_attr_mask: A bit-mask used to specify which attributes of the WQ + * are being modified. + * On output, the current values of selected WQ attributes are returned. + */ +int ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, + u32 wq_attr_mask) +{ + int err; + + if (!wq->device->modify_wq) + return -ENOSYS; + + err = wq->device->modify_wq(wq, wq_attr, wq_attr_mask, NULL); + return err; +} +EXPORT_SYMBOL(ib_modify_wq); + struct ib_flow *ib_create_flow(struct ib_qp *qp, struct ib_flow_attr *flow_attr, int domain) diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 7e440d41487a..f2d954ac42f6 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1428,6 +1428,48 @@ struct ib_srq { } ext; }; +enum ib_wq_type { + IB_WQT_RQ +}; + +enum ib_wq_state { + IB_WQS_RESET, + IB_WQS_RDY, + IB_WQS_ERR +}; + +struct ib_wq { + struct ib_device *device; + struct ib_uobject *uobject; + void *wq_context; + void (*event_handler)(struct ib_event *, void *); + struct ib_pd *pd; + struct ib_cq *cq; + u32 wq_num; + enum ib_wq_state state; + enum ib_wq_type wq_type; + atomic_t usecnt; +}; + +struct ib_wq_init_attr { + void *wq_context; + enum ib_wq_type wq_type; + u32 max_wr; + u32 max_sge; + struct ib_cq *cq; + void (*event_handler)(struct ib_event *, void *); +}; + +enum ib_wq_attr_mask { + IB_WQ_STATE = 1 << 0, + IB_WQ_CUR_STATE = 1 << 1, +}; + +struct ib_wq_attr { + enum ib_wq_state wq_state; + enum ib_wq_state curr_wq_state; +}; + struct ib_qp { struct ib_device *device; struct ib_pd *pd; @@ -1921,7 +1963,14 @@ struct ib_device { struct ifla_vf_stats *stats); int (*set_vf_guid)(struct ib_device *device, int vf, u8 port, u64 guid, int type); - + struct ib_wq * (*create_wq)(struct ib_pd *pd, + struct ib_wq_init_attr *init_attr, + struct ib_udata *udata); + int (*destroy_wq)(struct ib_wq *wq); + int (*modify_wq)(struct ib_wq *wq, + struct ib_wq_attr *attr, + u32 wq_attr_mask, + struct ib_udata *udata); struct ib_dma_mapping_ops *dma_ops; struct module *owner; @@ -3167,6 +3216,11 @@ int ib_check_mr_status(struct ib_mr *mr, u32 check_mask, struct net_device *ib_get_net_dev_by_params(struct ib_device *dev, u8 port, u16 pkey, const union ib_gid *gid, const struct sockaddr *addr); +struct ib_wq *ib_create_wq(struct ib_pd *pd, + struct ib_wq_init_attr *init_attr); +int ib_destroy_wq(struct ib_wq *wq); +int ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *attr, + u32 wq_attr_mask); int ib_map_mr_sg(struct ib_mr *mr, struct scatterlist *sg, int sg_nents, unsigned int *sg_offset, unsigned int page_size); -- cgit v1.2.3 From f213c05272100f385912372fff678d0af4d7f8ad Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:49 +0300 Subject: IB/uverbs: Add WQ support User space applications which use RSS functionality need to create a work queue object (WQ). The lifetime of such an object is: * Create a WQ * Modify the WQ from reset to init state. * Use the WQ (by downstream patches). * Destroy the WQ. These commands are added to the uverbs API. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/core/uverbs.h | 9 ++ drivers/infiniband/core/uverbs_cmd.c | 243 ++++++++++++++++++++++++++++++++++ drivers/infiniband/core/uverbs_main.c | 25 ++++ include/rdma/ib_verbs.h | 3 + include/uapi/rdma/ib_user_verbs.h | 41 ++++++ 5 files changed, 321 insertions(+) diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h index 612ccfd39bf9..74776c6531f4 100644 --- a/drivers/infiniband/core/uverbs.h +++ b/drivers/infiniband/core/uverbs.h @@ -162,6 +162,10 @@ struct ib_uqp_object { struct ib_uxrcd_object *uxrcd; }; +struct ib_uwq_object { + struct ib_uevent_object uevent; +}; + struct ib_ucq_object { struct ib_uobject uobject; struct ib_uverbs_file *uverbs_file; @@ -181,6 +185,7 @@ extern struct idr ib_uverbs_qp_idr; extern struct idr ib_uverbs_srq_idr; extern struct idr ib_uverbs_xrcd_idr; extern struct idr ib_uverbs_rule_idr; +extern struct idr ib_uverbs_wq_idr; void idr_remove_uobj(struct idr *idp, struct ib_uobject *uobj); @@ -199,6 +204,7 @@ void ib_uverbs_release_uevent(struct ib_uverbs_file *file, void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context); void ib_uverbs_cq_event_handler(struct ib_event *event, void *context_ptr); void ib_uverbs_qp_event_handler(struct ib_event *event, void *context_ptr); +void ib_uverbs_wq_event_handler(struct ib_event *event, void *context_ptr); void ib_uverbs_srq_event_handler(struct ib_event *event, void *context_ptr); void ib_uverbs_event_handler(struct ib_event_handler *handler, struct ib_event *event); @@ -275,5 +281,8 @@ IB_UVERBS_DECLARE_EX_CMD(destroy_flow); IB_UVERBS_DECLARE_EX_CMD(query_device); IB_UVERBS_DECLARE_EX_CMD(create_cq); IB_UVERBS_DECLARE_EX_CMD(create_qp); +IB_UVERBS_DECLARE_EX_CMD(create_wq); +IB_UVERBS_DECLARE_EX_CMD(modify_wq); +IB_UVERBS_DECLARE_EX_CMD(destroy_wq); #endif /* UVERBS_H */ diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 1a8babb8ee3c..22e617391657 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -57,6 +57,7 @@ static struct uverbs_lock_class ah_lock_class = { .name = "AH-uobj" }; static struct uverbs_lock_class srq_lock_class = { .name = "SRQ-uobj" }; static struct uverbs_lock_class xrcd_lock_class = { .name = "XRCD-uobj" }; static struct uverbs_lock_class rule_lock_class = { .name = "RULE-uobj" }; +static struct uverbs_lock_class wq_lock_class = { .name = "WQ-uobj" }; /* * The ib_uobject locking scheme is as follows: @@ -243,6 +244,16 @@ static struct ib_qp *idr_read_qp(int qp_handle, struct ib_ucontext *context) return idr_read_obj(&ib_uverbs_qp_idr, qp_handle, context, 0); } +static struct ib_wq *idr_read_wq(int wq_handle, struct ib_ucontext *context) +{ + return idr_read_obj(&ib_uverbs_wq_idr, wq_handle, context, 0); +} + +static void put_wq_read(struct ib_wq *wq) +{ + put_uobj_read(wq->uobject); +} + static struct ib_qp *idr_write_qp(int qp_handle, struct ib_ucontext *context) { struct ib_uobject *uobj; @@ -326,6 +337,7 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file, INIT_LIST_HEAD(&ucontext->qp_list); INIT_LIST_HEAD(&ucontext->srq_list); INIT_LIST_HEAD(&ucontext->ah_list); + INIT_LIST_HEAD(&ucontext->wq_list); INIT_LIST_HEAD(&ucontext->xrcd_list); INIT_LIST_HEAD(&ucontext->rule_list); rcu_read_lock(); @@ -3056,6 +3068,237 @@ static int kern_spec_to_ib_spec(struct ib_uverbs_flow_spec *kern_spec, return 0; } +int ib_uverbs_ex_create_wq(struct ib_uverbs_file *file, + struct ib_device *ib_dev, + struct ib_udata *ucore, + struct ib_udata *uhw) +{ + struct ib_uverbs_ex_create_wq cmd = {}; + struct ib_uverbs_ex_create_wq_resp resp = {}; + struct ib_uwq_object *obj; + int err = 0; + struct ib_cq *cq; + struct ib_pd *pd; + struct ib_wq *wq; + struct ib_wq_init_attr wq_init_attr = {}; + size_t required_cmd_sz; + size_t required_resp_len; + + required_cmd_sz = offsetof(typeof(cmd), max_sge) + sizeof(cmd.max_sge); + required_resp_len = offsetof(typeof(resp), wqn) + sizeof(resp.wqn); + + if (ucore->inlen < required_cmd_sz) + return -EINVAL; + + if (ucore->outlen < required_resp_len) + return -ENOSPC; + + if (ucore->inlen > sizeof(cmd) && + !ib_is_udata_cleared(ucore, sizeof(cmd), + ucore->inlen - sizeof(cmd))) + return -EOPNOTSUPP; + + err = ib_copy_from_udata(&cmd, ucore, min(sizeof(cmd), ucore->inlen)); + if (err) + return err; + + if (cmd.comp_mask) + return -EOPNOTSUPP; + + obj = kmalloc(sizeof(*obj), GFP_KERNEL); + if (!obj) + return -ENOMEM; + + init_uobj(&obj->uevent.uobject, cmd.user_handle, file->ucontext, + &wq_lock_class); + down_write(&obj->uevent.uobject.mutex); + pd = idr_read_pd(cmd.pd_handle, file->ucontext); + if (!pd) { + err = -EINVAL; + goto err_uobj; + } + + cq = idr_read_cq(cmd.cq_handle, file->ucontext, 0); + if (!cq) { + err = -EINVAL; + goto err_put_pd; + } + + wq_init_attr.cq = cq; + wq_init_attr.max_sge = cmd.max_sge; + wq_init_attr.max_wr = cmd.max_wr; + wq_init_attr.wq_context = file; + wq_init_attr.wq_type = cmd.wq_type; + wq_init_attr.event_handler = ib_uverbs_wq_event_handler; + obj->uevent.events_reported = 0; + INIT_LIST_HEAD(&obj->uevent.event_list); + wq = pd->device->create_wq(pd, &wq_init_attr, uhw); + if (IS_ERR(wq)) { + err = PTR_ERR(wq); + goto err_put_cq; + } + + wq->uobject = &obj->uevent.uobject; + obj->uevent.uobject.object = wq; + wq->wq_type = wq_init_attr.wq_type; + wq->cq = cq; + wq->pd = pd; + wq->device = pd->device; + wq->wq_context = wq_init_attr.wq_context; + atomic_set(&wq->usecnt, 0); + atomic_inc(&pd->usecnt); + atomic_inc(&cq->usecnt); + wq->uobject = &obj->uevent.uobject; + obj->uevent.uobject.object = wq; + err = idr_add_uobj(&ib_uverbs_wq_idr, &obj->uevent.uobject); + if (err) + goto destroy_wq; + + memset(&resp, 0, sizeof(resp)); + resp.wq_handle = obj->uevent.uobject.id; + resp.max_sge = wq_init_attr.max_sge; + resp.max_wr = wq_init_attr.max_wr; + resp.wqn = wq->wq_num; + resp.response_length = required_resp_len; + err = ib_copy_to_udata(ucore, + &resp, resp.response_length); + if (err) + goto err_copy; + + put_pd_read(pd); + put_cq_read(cq); + + mutex_lock(&file->mutex); + list_add_tail(&obj->uevent.uobject.list, &file->ucontext->wq_list); + mutex_unlock(&file->mutex); + + obj->uevent.uobject.live = 1; + up_write(&obj->uevent.uobject.mutex); + return 0; + +err_copy: + idr_remove_uobj(&ib_uverbs_wq_idr, &obj->uevent.uobject); +destroy_wq: + ib_destroy_wq(wq); +err_put_cq: + put_cq_read(cq); +err_put_pd: + put_pd_read(pd); +err_uobj: + put_uobj_write(&obj->uevent.uobject); + + return err; +} + +int ib_uverbs_ex_destroy_wq(struct ib_uverbs_file *file, + struct ib_device *ib_dev, + struct ib_udata *ucore, + struct ib_udata *uhw) +{ + struct ib_uverbs_ex_destroy_wq cmd = {}; + struct ib_uverbs_ex_destroy_wq_resp resp = {}; + struct ib_wq *wq; + struct ib_uobject *uobj; + struct ib_uwq_object *obj; + size_t required_cmd_sz; + size_t required_resp_len; + int ret; + + required_cmd_sz = offsetof(typeof(cmd), wq_handle) + sizeof(cmd.wq_handle); + required_resp_len = offsetof(typeof(resp), reserved) + sizeof(resp.reserved); + + if (ucore->inlen < required_cmd_sz) + return -EINVAL; + + if (ucore->outlen < required_resp_len) + return -ENOSPC; + + if (ucore->inlen > sizeof(cmd) && + !ib_is_udata_cleared(ucore, sizeof(cmd), + ucore->inlen - sizeof(cmd))) + return -EOPNOTSUPP; + + ret = ib_copy_from_udata(&cmd, ucore, min(sizeof(cmd), ucore->inlen)); + if (ret) + return ret; + + if (cmd.comp_mask) + return -EOPNOTSUPP; + + resp.response_length = required_resp_len; + uobj = idr_write_uobj(&ib_uverbs_wq_idr, cmd.wq_handle, + file->ucontext); + if (!uobj) + return -EINVAL; + + wq = uobj->object; + obj = container_of(uobj, struct ib_uwq_object, uevent.uobject); + ret = ib_destroy_wq(wq); + if (!ret) + uobj->live = 0; + + put_uobj_write(uobj); + if (ret) + return ret; + + idr_remove_uobj(&ib_uverbs_wq_idr, uobj); + + mutex_lock(&file->mutex); + list_del(&uobj->list); + mutex_unlock(&file->mutex); + + ib_uverbs_release_uevent(file, &obj->uevent); + resp.events_reported = obj->uevent.events_reported; + put_uobj(uobj); + + ret = ib_copy_to_udata(ucore, &resp, resp.response_length); + if (ret) + return ret; + + return 0; +} + +int ib_uverbs_ex_modify_wq(struct ib_uverbs_file *file, + struct ib_device *ib_dev, + struct ib_udata *ucore, + struct ib_udata *uhw) +{ + struct ib_uverbs_ex_modify_wq cmd = {}; + struct ib_wq *wq; + struct ib_wq_attr wq_attr = {}; + size_t required_cmd_sz; + int ret; + + required_cmd_sz = offsetof(typeof(cmd), curr_wq_state) + sizeof(cmd.curr_wq_state); + if (ucore->inlen < required_cmd_sz) + return -EINVAL; + + if (ucore->inlen > sizeof(cmd) && + !ib_is_udata_cleared(ucore, sizeof(cmd), + ucore->inlen - sizeof(cmd))) + return -EOPNOTSUPP; + + ret = ib_copy_from_udata(&cmd, ucore, min(sizeof(cmd), ucore->inlen)); + if (ret) + return ret; + + if (!cmd.attr_mask) + return -EINVAL; + + if (cmd.attr_mask > (IB_WQ_STATE | IB_WQ_CUR_STATE)) + return -EINVAL; + + wq = idr_read_wq(cmd.wq_handle, file->ucontext); + if (!wq) + return -EINVAL; + + wq_attr.curr_wq_state = cmd.curr_wq_state; + wq_attr.wq_state = cmd.wq_state; + ret = wq->device->modify_wq(wq, &wq_attr, cmd.attr_mask, uhw); + put_wq_read(wq); + return ret; +} + int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file, struct ib_device *ib_dev, struct ib_udata *ucore, diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c index 31f422a70623..91cb36f66ea7 100644 --- a/drivers/infiniband/core/uverbs_main.c +++ b/drivers/infiniband/core/uverbs_main.c @@ -76,6 +76,7 @@ DEFINE_IDR(ib_uverbs_qp_idr); DEFINE_IDR(ib_uverbs_srq_idr); DEFINE_IDR(ib_uverbs_xrcd_idr); DEFINE_IDR(ib_uverbs_rule_idr); +DEFINE_IDR(ib_uverbs_wq_idr); static DEFINE_SPINLOCK(map_lock); static DECLARE_BITMAP(dev_map, IB_UVERBS_MAX_DEVICES); @@ -130,6 +131,9 @@ static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file, [IB_USER_VERBS_EX_CMD_QUERY_DEVICE] = ib_uverbs_ex_query_device, [IB_USER_VERBS_EX_CMD_CREATE_CQ] = ib_uverbs_ex_create_cq, [IB_USER_VERBS_EX_CMD_CREATE_QP] = ib_uverbs_ex_create_qp, + [IB_USER_VERBS_EX_CMD_CREATE_WQ] = ib_uverbs_ex_create_wq, + [IB_USER_VERBS_EX_CMD_MODIFY_WQ] = ib_uverbs_ex_modify_wq, + [IB_USER_VERBS_EX_CMD_DESTROY_WQ] = ib_uverbs_ex_destroy_wq, }; static void ib_uverbs_add_one(struct ib_device *device); @@ -265,6 +269,17 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file, kfree(uqp); } + list_for_each_entry_safe(uobj, tmp, &context->wq_list, list) { + struct ib_wq *wq = uobj->object; + struct ib_uwq_object *uwq = + container_of(uobj, struct ib_uwq_object, uevent.uobject); + + idr_remove_uobj(&ib_uverbs_wq_idr, uobj); + ib_destroy_wq(wq); + ib_uverbs_release_uevent(file, &uwq->uevent); + kfree(uwq); + } + list_for_each_entry_safe(uobj, tmp, &context->srq_list, list) { struct ib_srq *srq = uobj->object; struct ib_uevent_object *uevent = @@ -568,6 +583,16 @@ void ib_uverbs_qp_event_handler(struct ib_event *event, void *context_ptr) &uobj->events_reported); } +void ib_uverbs_wq_event_handler(struct ib_event *event, void *context_ptr) +{ + struct ib_uevent_object *uobj = container_of(event->element.wq->uobject, + struct ib_uevent_object, uobject); + + ib_uverbs_async_handler(context_ptr, uobj->uobject.user_handle, + event->event, &uobj->event_list, + &uobj->events_reported); +} + void ib_uverbs_srq_event_handler(struct ib_event *event, void *context_ptr) { struct ib_uevent_object *uobj; diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index f2d954ac42f6..0c1956a98573 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -562,6 +562,7 @@ enum ib_event_type { IB_EVENT_QP_LAST_WQE_REACHED, IB_EVENT_CLIENT_REREGISTER, IB_EVENT_GID_CHANGE, + IB_EVENT_WQ_FATAL, }; const char *__attribute_const__ ib_event_msg(enum ib_event_type event); @@ -572,6 +573,7 @@ struct ib_event { struct ib_cq *cq; struct ib_qp *qp; struct ib_srq *srq; + struct ib_wq *wq; u8 port_num; } element; enum ib_event_type event; @@ -1323,6 +1325,7 @@ struct ib_ucontext { struct list_head ah_list; struct list_head xrcd_list; struct list_head rule_list; + struct list_head wq_list; int closing; struct pid *tgid; diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h index b6543d73d20a..c9470e511e03 100644 --- a/include/uapi/rdma/ib_user_verbs.h +++ b/include/uapi/rdma/ib_user_verbs.h @@ -95,6 +95,9 @@ enum { IB_USER_VERBS_EX_CMD_CREATE_QP = IB_USER_VERBS_CMD_CREATE_QP, IB_USER_VERBS_EX_CMD_CREATE_FLOW = IB_USER_VERBS_CMD_THRESHOLD, IB_USER_VERBS_EX_CMD_DESTROY_FLOW, + IB_USER_VERBS_EX_CMD_CREATE_WQ, + IB_USER_VERBS_EX_CMD_MODIFY_WQ, + IB_USER_VERBS_EX_CMD_DESTROY_WQ, }; /* @@ -946,4 +949,42 @@ struct ib_uverbs_destroy_srq_resp { __u32 events_reported; }; +struct ib_uverbs_ex_create_wq { + __u32 comp_mask; + __u32 wq_type; + __u64 user_handle; + __u32 pd_handle; + __u32 cq_handle; + __u32 max_wr; + __u32 max_sge; +}; + +struct ib_uverbs_ex_create_wq_resp { + __u32 comp_mask; + __u32 response_length; + __u32 wq_handle; + __u32 max_wr; + __u32 max_sge; + __u32 wqn; +}; + +struct ib_uverbs_ex_destroy_wq { + __u32 comp_mask; + __u32 wq_handle; +}; + +struct ib_uverbs_ex_destroy_wq_resp { + __u32 comp_mask; + __u32 response_length; + __u32 events_reported; + __u32 reserved; +}; + +struct ib_uverbs_ex_modify_wq { + __u32 attr_mask; + __u32 wq_handle; + __u32 wq_state; + __u32 curr_wq_state; +}; + #endif /* IB_USER_VERBS_H */ -- cgit v1.2.3 From 79b20a6c3014c789253fcb1ac4f09f8bdee2e94b Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:50 +0300 Subject: IB/mlx5: Add receive Work Queue verbs A QP can be created without internal WQs "packaged" inside it, this QP can be configured to use "external" WQ object as its receive/send queue. WQ is a necessary component for RSS technology since RSS mechanism is supposed to distribute the traffic between multiple Receive Work Queues Receive WQs are implemented by RQs. Implement the WQ creation, modification and destruction verbs. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 8 +- drivers/infiniband/hw/mlx5/mlx5_ib.h | 35 ++++ drivers/infiniband/hw/mlx5/qp.c | 306 +++++++++++++++++++++++++++++++++++ drivers/infiniband/hw/mlx5/user.h | 25 +++ 4 files changed, 373 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index b48ad85315dc..b3589b7e4c19 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -2450,9 +2450,15 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) IB_LINK_LAYER_ETHERNET) { dev->ib_dev.create_flow = mlx5_ib_create_flow; dev->ib_dev.destroy_flow = mlx5_ib_destroy_flow; + dev->ib_dev.create_wq = mlx5_ib_create_wq; + dev->ib_dev.modify_wq = mlx5_ib_modify_wq; + dev->ib_dev.destroy_wq = mlx5_ib_destroy_wq; dev->ib_dev.uverbs_ex_cmd_mask |= (1ull << IB_USER_VERBS_EX_CMD_CREATE_FLOW) | - (1ull << IB_USER_VERBS_EX_CMD_DESTROY_FLOW); + (1ull << IB_USER_VERBS_EX_CMD_DESTROY_FLOW) | + (1ull << IB_USER_VERBS_EX_CMD_CREATE_WQ) | + (1ull << IB_USER_VERBS_EX_CMD_MODIFY_WQ) | + (1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ); } err = init_node_data(dev); if (err) diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index c4a9825828bc..62d4e137593a 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -217,12 +217,36 @@ struct mlx5_ib_wq { void *qend; }; +struct mlx5_ib_rwq { + struct ib_wq ibwq; + u32 rqn; + u32 rq_num_pas; + u32 log_rq_stride; + u32 log_rq_size; + u32 rq_page_offset; + u32 log_page_size; + struct ib_umem *umem; + size_t buf_size; + unsigned int page_shift; + int create_type; + struct mlx5_db db; + u32 user_index; + u32 wqe_count; + u32 wqe_shift; + int wq_sig; +}; + enum { MLX5_QP_USER, MLX5_QP_KERNEL, MLX5_QP_EMPTY }; +enum { + MLX5_WQ_USER, + MLX5_WQ_KERNEL +}; + /* * Connect-IB can trigger up to four concurrent pagefaults * per-QP. @@ -628,6 +652,11 @@ static inline struct mlx5_ib_qp *to_mqp(struct ib_qp *ibqp) return container_of(ibqp, struct mlx5_ib_qp, ibqp); } +static inline struct mlx5_ib_rwq *to_mrwq(struct ib_wq *ibwq) +{ + return container_of(ibwq, struct mlx5_ib_rwq, ibwq); +} + static inline struct mlx5_ib_srq *to_mibsrq(struct mlx5_core_srq *msrq) { return container_of(msrq, struct mlx5_ib_srq, msrq); @@ -762,6 +791,12 @@ int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev); int mlx5_mr_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift); int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask, struct ib_mr_status *mr_status); +struct ib_wq *mlx5_ib_create_wq(struct ib_pd *pd, + struct ib_wq_init_attr *init_attr, + struct ib_udata *udata); +int mlx5_ib_destroy_wq(struct ib_wq *wq); +int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, + u32 wq_attr_mask, struct ib_udata *udata); #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING extern struct workqueue_struct *mlx5_ib_page_fault_wq; diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index ce434228a5ea..43d45e3042e4 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -649,6 +649,71 @@ err_umem: return err; } +static void destroy_user_rq(struct ib_pd *pd, struct mlx5_ib_rwq *rwq) +{ + struct mlx5_ib_ucontext *context; + + context = to_mucontext(pd->uobject->context); + mlx5_ib_db_unmap_user(context, &rwq->db); + if (rwq->umem) + ib_umem_release(rwq->umem); +} + +static int create_user_rq(struct mlx5_ib_dev *dev, struct ib_pd *pd, + struct mlx5_ib_rwq *rwq, + struct mlx5_ib_create_wq *ucmd) +{ + struct mlx5_ib_ucontext *context; + int page_shift = 0; + int npages; + u32 offset = 0; + int ncont = 0; + int err; + + if (!ucmd->buf_addr) + return -EINVAL; + + context = to_mucontext(pd->uobject->context); + rwq->umem = ib_umem_get(pd->uobject->context, ucmd->buf_addr, + rwq->buf_size, 0, 0); + if (IS_ERR(rwq->umem)) { + mlx5_ib_dbg(dev, "umem_get failed\n"); + err = PTR_ERR(rwq->umem); + return err; + } + + mlx5_ib_cont_pages(rwq->umem, ucmd->buf_addr, &npages, &page_shift, + &ncont, NULL); + err = mlx5_ib_get_buf_offset(ucmd->buf_addr, page_shift, + &rwq->rq_page_offset); + if (err) { + mlx5_ib_warn(dev, "bad offset\n"); + goto err_umem; + } + + rwq->rq_num_pas = ncont; + rwq->page_shift = page_shift; + rwq->log_page_size = page_shift - MLX5_ADAPTER_PAGE_SHIFT; + rwq->wq_sig = !!(ucmd->flags & MLX5_WQ_FLAG_SIGNATURE); + + mlx5_ib_dbg(dev, "addr 0x%llx, size %zd, npages %d, page_shift %d, ncont %d, offset %d\n", + (unsigned long long)ucmd->buf_addr, rwq->buf_size, + npages, page_shift, ncont, offset); + + err = mlx5_ib_db_map_user(context, ucmd->db_addr, &rwq->db); + if (err) { + mlx5_ib_dbg(dev, "map failed\n"); + goto err_umem; + } + + rwq->create_type = MLX5_WQ_USER; + return 0; + +err_umem: + ib_umem_release(rwq->umem); + return err; +} + static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, struct mlx5_ib_qp *qp, struct ib_udata *udata, struct ib_qp_init_attr *attr, @@ -4163,3 +4228,244 @@ int mlx5_ib_dealloc_xrcd(struct ib_xrcd *xrcd) return 0; } + +static int create_rq(struct mlx5_ib_rwq *rwq, struct ib_pd *pd, + struct ib_wq_init_attr *init_attr) +{ + struct mlx5_ib_dev *dev; + __be64 *rq_pas0; + void *in; + void *rqc; + void *wq; + int inlen; + int err; + + dev = to_mdev(pd->device); + + inlen = MLX5_ST_SZ_BYTES(create_rq_in) + sizeof(u64) * rwq->rq_num_pas; + in = mlx5_vzalloc(inlen); + if (!in) + return -ENOMEM; + + rqc = MLX5_ADDR_OF(create_rq_in, in, ctx); + MLX5_SET(rqc, rqc, mem_rq_type, + MLX5_RQC_MEM_RQ_TYPE_MEMORY_RQ_INLINE); + MLX5_SET(rqc, rqc, user_index, rwq->user_index); + MLX5_SET(rqc, rqc, cqn, to_mcq(init_attr->cq)->mcq.cqn); + MLX5_SET(rqc, rqc, state, MLX5_RQC_STATE_RST); + MLX5_SET(rqc, rqc, flush_in_error_en, 1); + wq = MLX5_ADDR_OF(rqc, rqc, wq); + MLX5_SET(wq, wq, wq_type, MLX5_WQ_TYPE_CYCLIC); + MLX5_SET(wq, wq, end_padding_mode, MLX5_WQ_END_PAD_MODE_ALIGN); + MLX5_SET(wq, wq, log_wq_stride, rwq->log_rq_stride); + MLX5_SET(wq, wq, log_wq_sz, rwq->log_rq_size); + MLX5_SET(wq, wq, pd, to_mpd(pd)->pdn); + MLX5_SET(wq, wq, page_offset, rwq->rq_page_offset); + MLX5_SET(wq, wq, log_wq_pg_sz, rwq->log_page_size); + MLX5_SET(wq, wq, wq_signature, rwq->wq_sig); + MLX5_SET64(wq, wq, dbr_addr, rwq->db.dma); + rq_pas0 = (__be64 *)MLX5_ADDR_OF(wq, wq, pas); + mlx5_ib_populate_pas(dev, rwq->umem, rwq->page_shift, rq_pas0, 0); + err = mlx5_core_create_rq(dev->mdev, in, inlen, &rwq->rqn); + kvfree(in); + return err; +} + +static int set_user_rq_size(struct mlx5_ib_dev *dev, + struct ib_wq_init_attr *wq_init_attr, + struct mlx5_ib_create_wq *ucmd, + struct mlx5_ib_rwq *rwq) +{ + /* Sanity check RQ size before proceeding */ + if (wq_init_attr->max_wr > (1 << MLX5_CAP_GEN(dev->mdev, log_max_wq_sz))) + return -EINVAL; + + if (!ucmd->rq_wqe_count) + return -EINVAL; + + rwq->wqe_count = ucmd->rq_wqe_count; + rwq->wqe_shift = ucmd->rq_wqe_shift; + rwq->buf_size = (rwq->wqe_count << rwq->wqe_shift); + rwq->log_rq_stride = rwq->wqe_shift; + rwq->log_rq_size = ilog2(rwq->wqe_count); + return 0; +} + +static int prepare_user_rq(struct ib_pd *pd, + struct ib_wq_init_attr *init_attr, + struct ib_udata *udata, + struct mlx5_ib_rwq *rwq) +{ + struct mlx5_ib_dev *dev = to_mdev(pd->device); + struct mlx5_ib_create_wq ucmd = {}; + int err; + size_t required_cmd_sz; + + required_cmd_sz = offsetof(typeof(ucmd), reserved) + sizeof(ucmd.reserved); + if (udata->inlen < required_cmd_sz) { + mlx5_ib_dbg(dev, "invalid inlen\n"); + return -EINVAL; + } + + if (udata->inlen > sizeof(ucmd) && + !ib_is_udata_cleared(udata, sizeof(ucmd), + udata->inlen - sizeof(ucmd))) { + mlx5_ib_dbg(dev, "inlen is not supported\n"); + return -EOPNOTSUPP; + } + + if (ib_copy_from_udata(&ucmd, udata, min(sizeof(ucmd), udata->inlen))) { + mlx5_ib_dbg(dev, "copy failed\n"); + return -EFAULT; + } + + if (ucmd.comp_mask) { + mlx5_ib_dbg(dev, "invalid comp mask\n"); + return -EOPNOTSUPP; + } + + if (ucmd.reserved) { + mlx5_ib_dbg(dev, "invalid reserved\n"); + return -EOPNOTSUPP; + } + + err = set_user_rq_size(dev, init_attr, &ucmd, rwq); + if (err) { + mlx5_ib_dbg(dev, "err %d\n", err); + return err; + } + + err = create_user_rq(dev, pd, rwq, &ucmd); + if (err) { + mlx5_ib_dbg(dev, "err %d\n", err); + if (err) + return err; + } + + rwq->user_index = ucmd.user_index; + return 0; +} + +struct ib_wq *mlx5_ib_create_wq(struct ib_pd *pd, + struct ib_wq_init_attr *init_attr, + struct ib_udata *udata) +{ + struct mlx5_ib_dev *dev; + struct mlx5_ib_rwq *rwq; + struct mlx5_ib_create_wq_resp resp = {}; + size_t min_resp_len; + int err; + + if (!udata) + return ERR_PTR(-ENOSYS); + + min_resp_len = offsetof(typeof(resp), reserved) + sizeof(resp.reserved); + if (udata->outlen && udata->outlen < min_resp_len) + return ERR_PTR(-EINVAL); + + dev = to_mdev(pd->device); + switch (init_attr->wq_type) { + case IB_WQT_RQ: + rwq = kzalloc(sizeof(*rwq), GFP_KERNEL); + if (!rwq) + return ERR_PTR(-ENOMEM); + err = prepare_user_rq(pd, init_attr, udata, rwq); + if (err) + goto err; + err = create_rq(rwq, pd, init_attr); + if (err) + goto err_user_rq; + break; + default: + mlx5_ib_dbg(dev, "unsupported wq type %d\n", + init_attr->wq_type); + return ERR_PTR(-EINVAL); + } + + rwq->ibwq.wq_num = rwq->rqn; + rwq->ibwq.state = IB_WQS_RESET; + if (udata->outlen) { + resp.response_length = offsetof(typeof(resp), response_length) + + sizeof(resp.response_length); + err = ib_copy_to_udata(udata, &resp, resp.response_length); + if (err) + goto err_copy; + } + + return &rwq->ibwq; + +err_copy: + mlx5_core_destroy_rq(dev->mdev, rwq->rqn); +err_user_rq: + destroy_user_rq(pd, rwq); +err: + kfree(rwq); + return ERR_PTR(err); +} + +int mlx5_ib_destroy_wq(struct ib_wq *wq) +{ + struct mlx5_ib_dev *dev = to_mdev(wq->device); + struct mlx5_ib_rwq *rwq = to_mrwq(wq); + + mlx5_core_destroy_rq(dev->mdev, rwq->rqn); + destroy_user_rq(wq->pd, rwq); + kfree(rwq); + + return 0; +} + +int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, + u32 wq_attr_mask, struct ib_udata *udata) +{ + struct mlx5_ib_dev *dev = to_mdev(wq->device); + struct mlx5_ib_rwq *rwq = to_mrwq(wq); + struct mlx5_ib_modify_wq ucmd = {}; + size_t required_cmd_sz; + int curr_wq_state; + int wq_state; + int inlen; + int err; + void *rqc; + void *in; + + required_cmd_sz = offsetof(typeof(ucmd), reserved) + sizeof(ucmd.reserved); + if (udata->inlen < required_cmd_sz) + return -EINVAL; + + if (udata->inlen > sizeof(ucmd) && + !ib_is_udata_cleared(udata, sizeof(ucmd), + udata->inlen - sizeof(ucmd))) + return -EOPNOTSUPP; + + if (ib_copy_from_udata(&ucmd, udata, min(sizeof(ucmd), udata->inlen))) + return -EFAULT; + + if (ucmd.comp_mask || ucmd.reserved) + return -EOPNOTSUPP; + + inlen = MLX5_ST_SZ_BYTES(modify_rq_in); + in = mlx5_vzalloc(inlen); + if (!in) + return -ENOMEM; + + rqc = MLX5_ADDR_OF(modify_rq_in, in, ctx); + + curr_wq_state = (wq_attr_mask & IB_WQ_CUR_STATE) ? + wq_attr->curr_wq_state : wq->state; + wq_state = (wq_attr_mask & IB_WQ_STATE) ? + wq_attr->wq_state : curr_wq_state; + if (curr_wq_state == IB_WQS_ERR) + curr_wq_state = MLX5_RQC_STATE_ERR; + if (wq_state == IB_WQS_ERR) + wq_state = MLX5_RQC_STATE_ERR; + MLX5_SET(modify_rq_in, in, rq_state, curr_wq_state); + MLX5_SET(rqc, rqc, state, wq_state); + + err = mlx5_core_modify_rq(dev->mdev, rwq->rqn, in, inlen); + kvfree(in); + if (!err) + rwq->ibwq.state = (wq_state == MLX5_RQC_STATE_ERR) ? IB_WQS_ERR : wq_state; + + return err; +} diff --git a/drivers/infiniband/hw/mlx5/user.h b/drivers/infiniband/hw/mlx5/user.h index 61bc308bb802..3e66f935e4c8 100644 --- a/drivers/infiniband/hw/mlx5/user.h +++ b/drivers/infiniband/hw/mlx5/user.h @@ -46,6 +46,10 @@ enum { MLX5_SRQ_FLAG_SIGNATURE = 1 << 0, }; +enum { + MLX5_WQ_FLAG_SIGNATURE = 1 << 0, +}; + /* Increment this value if any changes that break userspace ABI * compatibility are made. @@ -159,6 +163,27 @@ struct mlx5_ib_alloc_mw { __u16 reserved2; }; +struct mlx5_ib_create_wq { + __u64 buf_addr; + __u64 db_addr; + __u32 rq_wqe_count; + __u32 rq_wqe_shift; + __u32 user_index; + __u32 flags; + __u32 comp_mask; + __u32 reserved; +}; + +struct mlx5_ib_create_wq_resp { + __u32 response_length; + __u32 reserved; +}; + +struct mlx5_ib_modify_wq { + __u32 comp_mask; + __u32 reserved; +}; + static inline int get_qp_user_index(struct mlx5_ib_ucontext *ucontext, struct mlx5_ib_create_qp *ucmd, int inlen, -- cgit v1.2.3 From 6d39786bf116e476d75eca91f7cfa22586a32e5f Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:51 +0300 Subject: IB/core: Introduce Receive Work Queue indirection table Introduce Receive Work Queue (WQ) indirection table. This object can be used to spread incoming traffic to different receive Work Queues. A Receive WQ indirection table points to variable size of WQs. This table is given to a QP in downstream patches. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/core/verbs.c | 62 +++++++++++++++++++++++++++++++++++++++++ include/rdma/ib_verbs.h | 23 +++++++++++++++ 2 files changed, 85 insertions(+) diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index c096cadc6e23..6b548d7f8753 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -1636,6 +1636,68 @@ int ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, } EXPORT_SYMBOL(ib_modify_wq); +/* + * ib_create_rwq_ind_table - Creates a RQ Indirection Table. + * @device: The device on which to create the rwq indirection table. + * @ib_rwq_ind_table_init_attr: A list of initial attributes required to + * create the Indirection Table. + * + * Note: The life time of ib_rwq_ind_table_init_attr->ind_tbl is not less + * than the created ib_rwq_ind_table object and the caller is responsible + * for its memory allocation/free. + */ +struct ib_rwq_ind_table *ib_create_rwq_ind_table(struct ib_device *device, + struct ib_rwq_ind_table_init_attr *init_attr) +{ + struct ib_rwq_ind_table *rwq_ind_table; + int i; + u32 table_size; + + if (!device->create_rwq_ind_table) + return ERR_PTR(-ENOSYS); + + table_size = (1 << init_attr->log_ind_tbl_size); + rwq_ind_table = device->create_rwq_ind_table(device, + init_attr, NULL); + if (IS_ERR(rwq_ind_table)) + return rwq_ind_table; + + rwq_ind_table->ind_tbl = init_attr->ind_tbl; + rwq_ind_table->log_ind_tbl_size = init_attr->log_ind_tbl_size; + rwq_ind_table->device = device; + rwq_ind_table->uobject = NULL; + atomic_set(&rwq_ind_table->usecnt, 0); + + for (i = 0; i < table_size; i++) + atomic_inc(&rwq_ind_table->ind_tbl[i]->usecnt); + + return rwq_ind_table; +} +EXPORT_SYMBOL(ib_create_rwq_ind_table); + +/* + * ib_destroy_rwq_ind_table - Destroys the specified Indirection Table. + * @wq_ind_table: The Indirection Table to destroy. +*/ +int ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *rwq_ind_table) +{ + int err, i; + u32 table_size = (1 << rwq_ind_table->log_ind_tbl_size); + struct ib_wq **ind_tbl = rwq_ind_table->ind_tbl; + + if (atomic_read(&rwq_ind_table->usecnt)) + return -EBUSY; + + err = rwq_ind_table->device->destroy_rwq_ind_table(rwq_ind_table); + if (!err) { + for (i = 0; i < table_size; i++) + atomic_dec(&ind_tbl[i]->usecnt); + } + + return err; +} +EXPORT_SYMBOL(ib_destroy_rwq_ind_table); + struct ib_flow *ib_create_flow(struct ib_qp *qp, struct ib_flow_attr *flow_attr, int domain) diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 0c1956a98573..fa2e0184dcc5 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1473,6 +1473,21 @@ struct ib_wq_attr { enum ib_wq_state curr_wq_state; }; +struct ib_rwq_ind_table { + struct ib_device *device; + struct ib_uobject *uobject; + atomic_t usecnt; + u32 ind_tbl_num; + u32 log_ind_tbl_size; + struct ib_wq **ind_tbl; +}; + +struct ib_rwq_ind_table_init_attr { + u32 log_ind_tbl_size; + /* Each entry is a pointer to Receive Work Queue */ + struct ib_wq **ind_tbl; +}; + struct ib_qp { struct ib_device *device; struct ib_pd *pd; @@ -1974,6 +1989,10 @@ struct ib_device { struct ib_wq_attr *attr, u32 wq_attr_mask, struct ib_udata *udata); + struct ib_rwq_ind_table * (*create_rwq_ind_table)(struct ib_device *device, + struct ib_rwq_ind_table_init_attr *init_attr, + struct ib_udata *udata); + int (*destroy_rwq_ind_table)(struct ib_rwq_ind_table *wq_ind_table); struct ib_dma_mapping_ops *dma_ops; struct module *owner; @@ -3224,6 +3243,10 @@ struct ib_wq *ib_create_wq(struct ib_pd *pd, int ib_destroy_wq(struct ib_wq *wq); int ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *attr, u32 wq_attr_mask); +struct ib_rwq_ind_table *ib_create_rwq_ind_table(struct ib_device *device, + struct ib_rwq_ind_table_init_attr* + wq_ind_table_init_attr); +int ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table); int ib_map_mr_sg(struct ib_mr *mr, struct scatterlist *sg, int sg_nents, unsigned int *sg_offset, unsigned int page_size); -- cgit v1.2.3 From de019a94049d579608a5511f8c50652faf125182 Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:52 +0300 Subject: IB/uverbs: Introduce RWQ Indirection table User applications that want to spread traffic on several WQs, need to create an indirection table, by using already created WQs. Adding uverbs API in order to create and destroy this table. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/core/uverbs.h | 3 + drivers/infiniband/core/uverbs_cmd.c | 210 ++++++++++++++++++++++++++++++++++ drivers/infiniband/core/uverbs_main.c | 13 +++ include/rdma/ib_verbs.h | 1 + include/uapi/rdma/ib_user_verbs.h | 26 +++++ 5 files changed, 253 insertions(+) diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h index 74776c6531f4..6c2292397cd6 100644 --- a/drivers/infiniband/core/uverbs.h +++ b/drivers/infiniband/core/uverbs.h @@ -186,6 +186,7 @@ extern struct idr ib_uverbs_srq_idr; extern struct idr ib_uverbs_xrcd_idr; extern struct idr ib_uverbs_rule_idr; extern struct idr ib_uverbs_wq_idr; +extern struct idr ib_uverbs_rwq_ind_tbl_idr; void idr_remove_uobj(struct idr *idp, struct ib_uobject *uobj); @@ -284,5 +285,7 @@ IB_UVERBS_DECLARE_EX_CMD(create_qp); IB_UVERBS_DECLARE_EX_CMD(create_wq); IB_UVERBS_DECLARE_EX_CMD(modify_wq); IB_UVERBS_DECLARE_EX_CMD(destroy_wq); +IB_UVERBS_DECLARE_EX_CMD(create_rwq_ind_table); +IB_UVERBS_DECLARE_EX_CMD(destroy_rwq_ind_table); #endif /* UVERBS_H */ diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 22e617391657..327a56cccc27 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -58,6 +58,7 @@ static struct uverbs_lock_class srq_lock_class = { .name = "SRQ-uobj" }; static struct uverbs_lock_class xrcd_lock_class = { .name = "XRCD-uobj" }; static struct uverbs_lock_class rule_lock_class = { .name = "RULE-uobj" }; static struct uverbs_lock_class wq_lock_class = { .name = "WQ-uobj" }; +static struct uverbs_lock_class rwq_ind_table_lock_class = { .name = "IND_TBL-uobj" }; /* * The ib_uobject locking scheme is as follows: @@ -338,6 +339,7 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file, INIT_LIST_HEAD(&ucontext->srq_list); INIT_LIST_HEAD(&ucontext->ah_list); INIT_LIST_HEAD(&ucontext->wq_list); + INIT_LIST_HEAD(&ucontext->rwq_ind_tbl_list); INIT_LIST_HEAD(&ucontext->xrcd_list); INIT_LIST_HEAD(&ucontext->rule_list); rcu_read_lock(); @@ -3299,6 +3301,214 @@ int ib_uverbs_ex_modify_wq(struct ib_uverbs_file *file, return ret; } +int ib_uverbs_ex_create_rwq_ind_table(struct ib_uverbs_file *file, + struct ib_device *ib_dev, + struct ib_udata *ucore, + struct ib_udata *uhw) +{ + struct ib_uverbs_ex_create_rwq_ind_table cmd = {}; + struct ib_uverbs_ex_create_rwq_ind_table_resp resp = {}; + struct ib_uobject *uobj; + int err = 0; + struct ib_rwq_ind_table_init_attr init_attr = {}; + struct ib_rwq_ind_table *rwq_ind_tbl; + struct ib_wq **wqs = NULL; + u32 *wqs_handles = NULL; + struct ib_wq *wq = NULL; + int i, j, num_read_wqs; + u32 num_wq_handles; + u32 expected_in_size; + size_t required_cmd_sz_header; + size_t required_resp_len; + + required_cmd_sz_header = offsetof(typeof(cmd), log_ind_tbl_size) + sizeof(cmd.log_ind_tbl_size); + required_resp_len = offsetof(typeof(resp), ind_tbl_num) + sizeof(resp.ind_tbl_num); + + if (ucore->inlen < required_cmd_sz_header) + return -EINVAL; + + if (ucore->outlen < required_resp_len) + return -ENOSPC; + + err = ib_copy_from_udata(&cmd, ucore, required_cmd_sz_header); + if (err) + return err; + + ucore->inbuf += required_cmd_sz_header; + ucore->inlen -= required_cmd_sz_header; + + if (cmd.comp_mask) + return -EOPNOTSUPP; + + if (cmd.log_ind_tbl_size > IB_USER_VERBS_MAX_LOG_IND_TBL_SIZE) + return -EINVAL; + + num_wq_handles = 1 << cmd.log_ind_tbl_size; + expected_in_size = num_wq_handles * sizeof(__u32); + if (num_wq_handles == 1) + /* input size for wq handles is u64 aligned */ + expected_in_size += sizeof(__u32); + + if (ucore->inlen < expected_in_size) + return -EINVAL; + + if (ucore->inlen > expected_in_size && + !ib_is_udata_cleared(ucore, expected_in_size, + ucore->inlen - expected_in_size)) + return -EOPNOTSUPP; + + wqs_handles = kcalloc(num_wq_handles, sizeof(*wqs_handles), + GFP_KERNEL); + if (!wqs_handles) + return -ENOMEM; + + err = ib_copy_from_udata(wqs_handles, ucore, + num_wq_handles * sizeof(__u32)); + if (err) + goto err_free; + + wqs = kcalloc(num_wq_handles, sizeof(*wqs), GFP_KERNEL); + if (!wqs) { + err = -ENOMEM; + goto err_free; + } + + for (num_read_wqs = 0; num_read_wqs < num_wq_handles; + num_read_wqs++) { + wq = idr_read_wq(wqs_handles[num_read_wqs], file->ucontext); + if (!wq) { + err = -EINVAL; + goto put_wqs; + } + + wqs[num_read_wqs] = wq; + } + + uobj = kmalloc(sizeof(*uobj), GFP_KERNEL); + if (!uobj) { + err = -ENOMEM; + goto put_wqs; + } + + init_uobj(uobj, 0, file->ucontext, &rwq_ind_table_lock_class); + down_write(&uobj->mutex); + init_attr.log_ind_tbl_size = cmd.log_ind_tbl_size; + init_attr.ind_tbl = wqs; + rwq_ind_tbl = ib_dev->create_rwq_ind_table(ib_dev, &init_attr, uhw); + + if (IS_ERR(rwq_ind_tbl)) { + err = PTR_ERR(rwq_ind_tbl); + goto err_uobj; + } + + rwq_ind_tbl->ind_tbl = wqs; + rwq_ind_tbl->log_ind_tbl_size = init_attr.log_ind_tbl_size; + rwq_ind_tbl->uobject = uobj; + uobj->object = rwq_ind_tbl; + rwq_ind_tbl->device = ib_dev; + atomic_set(&rwq_ind_tbl->usecnt, 0); + + for (i = 0; i < num_wq_handles; i++) + atomic_inc(&wqs[i]->usecnt); + + err = idr_add_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); + if (err) + goto destroy_ind_tbl; + + resp.ind_tbl_handle = uobj->id; + resp.ind_tbl_num = rwq_ind_tbl->ind_tbl_num; + resp.response_length = required_resp_len; + + err = ib_copy_to_udata(ucore, + &resp, resp.response_length); + if (err) + goto err_copy; + + kfree(wqs_handles); + + for (j = 0; j < num_read_wqs; j++) + put_wq_read(wqs[j]); + + mutex_lock(&file->mutex); + list_add_tail(&uobj->list, &file->ucontext->rwq_ind_tbl_list); + mutex_unlock(&file->mutex); + + uobj->live = 1; + + up_write(&uobj->mutex); + return 0; + +err_copy: + idr_remove_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); +destroy_ind_tbl: + ib_destroy_rwq_ind_table(rwq_ind_tbl); +err_uobj: + put_uobj_write(uobj); +put_wqs: + for (j = 0; j < num_read_wqs; j++) + put_wq_read(wqs[j]); +err_free: + kfree(wqs_handles); + kfree(wqs); + return err; +} + +int ib_uverbs_ex_destroy_rwq_ind_table(struct ib_uverbs_file *file, + struct ib_device *ib_dev, + struct ib_udata *ucore, + struct ib_udata *uhw) +{ + struct ib_uverbs_ex_destroy_rwq_ind_table cmd = {}; + struct ib_rwq_ind_table *rwq_ind_tbl; + struct ib_uobject *uobj; + int ret; + struct ib_wq **ind_tbl; + size_t required_cmd_sz; + + required_cmd_sz = offsetof(typeof(cmd), ind_tbl_handle) + sizeof(cmd.ind_tbl_handle); + + if (ucore->inlen < required_cmd_sz) + return -EINVAL; + + if (ucore->inlen > sizeof(cmd) && + !ib_is_udata_cleared(ucore, sizeof(cmd), + ucore->inlen - sizeof(cmd))) + return -EOPNOTSUPP; + + ret = ib_copy_from_udata(&cmd, ucore, min(sizeof(cmd), ucore->inlen)); + if (ret) + return ret; + + if (cmd.comp_mask) + return -EOPNOTSUPP; + + uobj = idr_write_uobj(&ib_uverbs_rwq_ind_tbl_idr, cmd.ind_tbl_handle, + file->ucontext); + if (!uobj) + return -EINVAL; + rwq_ind_tbl = uobj->object; + ind_tbl = rwq_ind_tbl->ind_tbl; + + ret = ib_destroy_rwq_ind_table(rwq_ind_tbl); + if (!ret) + uobj->live = 0; + + put_uobj_write(uobj); + + if (ret) + return ret; + + idr_remove_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); + + mutex_lock(&file->mutex); + list_del(&uobj->list); + mutex_unlock(&file->mutex); + + put_uobj(uobj); + kfree(ind_tbl); + return ret; +} + int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file, struct ib_device *ib_dev, struct ib_udata *ucore, diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c index 91cb36f66ea7..426e0ac203fc 100644 --- a/drivers/infiniband/core/uverbs_main.c +++ b/drivers/infiniband/core/uverbs_main.c @@ -77,6 +77,7 @@ DEFINE_IDR(ib_uverbs_srq_idr); DEFINE_IDR(ib_uverbs_xrcd_idr); DEFINE_IDR(ib_uverbs_rule_idr); DEFINE_IDR(ib_uverbs_wq_idr); +DEFINE_IDR(ib_uverbs_rwq_ind_tbl_idr); static DEFINE_SPINLOCK(map_lock); static DECLARE_BITMAP(dev_map, IB_UVERBS_MAX_DEVICES); @@ -134,6 +135,8 @@ static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file, [IB_USER_VERBS_EX_CMD_CREATE_WQ] = ib_uverbs_ex_create_wq, [IB_USER_VERBS_EX_CMD_MODIFY_WQ] = ib_uverbs_ex_modify_wq, [IB_USER_VERBS_EX_CMD_DESTROY_WQ] = ib_uverbs_ex_destroy_wq, + [IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL] = ib_uverbs_ex_create_rwq_ind_table, + [IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL] = ib_uverbs_ex_destroy_rwq_ind_table, }; static void ib_uverbs_add_one(struct ib_device *device); @@ -269,6 +272,16 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file, kfree(uqp); } + list_for_each_entry_safe(uobj, tmp, &context->rwq_ind_tbl_list, list) { + struct ib_rwq_ind_table *rwq_ind_tbl = uobj->object; + struct ib_wq **ind_tbl = rwq_ind_tbl->ind_tbl; + + idr_remove_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); + ib_destroy_rwq_ind_table(rwq_ind_tbl); + kfree(ind_tbl); + kfree(uobj); + } + list_for_each_entry_safe(uobj, tmp, &context->wq_list, list) { struct ib_wq *wq = uobj->object; struct ib_uwq_object *uwq = diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index fa2e0184dcc5..e305c9a36663 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1326,6 +1326,7 @@ struct ib_ucontext { struct list_head xrcd_list; struct list_head rule_list; struct list_head wq_list; + struct list_head rwq_ind_tbl_list; int closing; struct pid *tgid; diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h index c9470e511e03..2cf7c95860d8 100644 --- a/include/uapi/rdma/ib_user_verbs.h +++ b/include/uapi/rdma/ib_user_verbs.h @@ -98,6 +98,8 @@ enum { IB_USER_VERBS_EX_CMD_CREATE_WQ, IB_USER_VERBS_EX_CMD_MODIFY_WQ, IB_USER_VERBS_EX_CMD_DESTROY_WQ, + IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL, + IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL }; /* @@ -987,4 +989,28 @@ struct ib_uverbs_ex_modify_wq { __u32 curr_wq_state; }; +/* Prevent memory allocation rather than max expected size */ +#define IB_USER_VERBS_MAX_LOG_IND_TBL_SIZE 0x0d +struct ib_uverbs_ex_create_rwq_ind_table { + __u32 comp_mask; + __u32 log_ind_tbl_size; + /* Following are the wq handles according to log_ind_tbl_size + * wq_handle1 + * wq_handle2 + */ + __u32 wq_handles[0]; +}; + +struct ib_uverbs_ex_create_rwq_ind_table_resp { + __u32 comp_mask; + __u32 response_length; + __u32 ind_tbl_handle; + __u32 ind_tbl_num; +}; + +struct ib_uverbs_ex_destroy_rwq_ind_table { + __u32 comp_mask; + __u32 ind_tbl_handle; +}; + #endif /* IB_USER_VERBS_H */ -- cgit v1.2.3 From c5f9092936fe88b39e2eddccedeb1c51883fcd31 Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:53 +0300 Subject: IB/mlx5: Add Receive Work Queue Indirection table operations Some mlx5 based hardwares support a RQ table object. This RQ table points to a few RQ objects. We implement the receive work queue indirection table API (create and destroy) by using this hardware object. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 6 ++- drivers/infiniband/hw/mlx5/mlx5_ib.h | 14 +++++++ drivers/infiniband/hw/mlx5/qp.c | 78 ++++++++++++++++++++++++++++++++++++ drivers/infiniband/hw/mlx5/user.h | 5 +++ 4 files changed, 102 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index b3589b7e4c19..82be19447a3f 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -2453,12 +2453,16 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->ib_dev.create_wq = mlx5_ib_create_wq; dev->ib_dev.modify_wq = mlx5_ib_modify_wq; dev->ib_dev.destroy_wq = mlx5_ib_destroy_wq; + dev->ib_dev.create_rwq_ind_table = mlx5_ib_create_rwq_ind_table; + dev->ib_dev.destroy_rwq_ind_table = mlx5_ib_destroy_rwq_ind_table; dev->ib_dev.uverbs_ex_cmd_mask |= (1ull << IB_USER_VERBS_EX_CMD_CREATE_FLOW) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_FLOW) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_WQ) | (1ull << IB_USER_VERBS_EX_CMD_MODIFY_WQ) | - (1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ); + (1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ) | + (1ull << IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL) | + (1ull << IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL); } err = init_node_data(dev); if (err) diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 62d4e137593a..cd3d6203ca1c 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -247,6 +247,11 @@ enum { MLX5_WQ_KERNEL }; +struct mlx5_ib_rwq_ind_table { + struct ib_rwq_ind_table ib_rwq_ind_tbl; + u32 rqtn; +}; + /* * Connect-IB can trigger up to four concurrent pagefaults * per-QP. @@ -657,6 +662,11 @@ static inline struct mlx5_ib_rwq *to_mrwq(struct ib_wq *ibwq) return container_of(ibwq, struct mlx5_ib_rwq, ibwq); } +static inline struct mlx5_ib_rwq_ind_table *to_mrwq_ind_table(struct ib_rwq_ind_table *ib_rwq_ind_tbl) +{ + return container_of(ib_rwq_ind_tbl, struct mlx5_ib_rwq_ind_table, ib_rwq_ind_tbl); +} + static inline struct mlx5_ib_srq *to_mibsrq(struct mlx5_core_srq *msrq) { return container_of(msrq, struct mlx5_ib_srq, msrq); @@ -797,6 +807,10 @@ struct ib_wq *mlx5_ib_create_wq(struct ib_pd *pd, int mlx5_ib_destroy_wq(struct ib_wq *wq); int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, u32 wq_attr_mask, struct ib_udata *udata); +struct ib_rwq_ind_table *mlx5_ib_create_rwq_ind_table(struct ib_device *device, + struct ib_rwq_ind_table_init_attr *init_attr, + struct ib_udata *udata); +int mlx5_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table); #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING extern struct workqueue_struct *mlx5_ib_page_fault_wq; diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 43d45e3042e4..1c0e332146b5 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -4415,6 +4415,84 @@ int mlx5_ib_destroy_wq(struct ib_wq *wq) return 0; } +struct ib_rwq_ind_table *mlx5_ib_create_rwq_ind_table(struct ib_device *device, + struct ib_rwq_ind_table_init_attr *init_attr, + struct ib_udata *udata) +{ + struct mlx5_ib_dev *dev = to_mdev(device); + struct mlx5_ib_rwq_ind_table *rwq_ind_tbl; + int sz = 1 << init_attr->log_ind_tbl_size; + struct mlx5_ib_create_rwq_ind_tbl_resp resp = {}; + size_t min_resp_len; + int inlen; + int err; + int i; + u32 *in; + void *rqtc; + + if (udata->inlen > 0 && + !ib_is_udata_cleared(udata, 0, + udata->inlen)) + return ERR_PTR(-EOPNOTSUPP); + + min_resp_len = offsetof(typeof(resp), reserved) + sizeof(resp.reserved); + if (udata->outlen && udata->outlen < min_resp_len) + return ERR_PTR(-EINVAL); + + rwq_ind_tbl = kzalloc(sizeof(*rwq_ind_tbl), GFP_KERNEL); + if (!rwq_ind_tbl) + return ERR_PTR(-ENOMEM); + + inlen = MLX5_ST_SZ_BYTES(create_rqt_in) + sizeof(u32) * sz; + in = mlx5_vzalloc(inlen); + if (!in) { + err = -ENOMEM; + goto err; + } + + rqtc = MLX5_ADDR_OF(create_rqt_in, in, rqt_context); + + MLX5_SET(rqtc, rqtc, rqt_actual_size, sz); + MLX5_SET(rqtc, rqtc, rqt_max_size, sz); + + for (i = 0; i < sz; i++) + MLX5_SET(rqtc, rqtc, rq_num[i], init_attr->ind_tbl[i]->wq_num); + + err = mlx5_core_create_rqt(dev->mdev, in, inlen, &rwq_ind_tbl->rqtn); + kvfree(in); + + if (err) + goto err; + + rwq_ind_tbl->ib_rwq_ind_tbl.ind_tbl_num = rwq_ind_tbl->rqtn; + if (udata->outlen) { + resp.response_length = offsetof(typeof(resp), response_length) + + sizeof(resp.response_length); + err = ib_copy_to_udata(udata, &resp, resp.response_length); + if (err) + goto err_copy; + } + + return &rwq_ind_tbl->ib_rwq_ind_tbl; + +err_copy: + mlx5_core_destroy_rqt(dev->mdev, rwq_ind_tbl->rqtn); +err: + kfree(rwq_ind_tbl); + return ERR_PTR(err); +} + +int mlx5_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *ib_rwq_ind_tbl) +{ + struct mlx5_ib_rwq_ind_table *rwq_ind_tbl = to_mrwq_ind_table(ib_rwq_ind_tbl); + struct mlx5_ib_dev *dev = to_mdev(ib_rwq_ind_tbl->device); + + mlx5_core_destroy_rqt(dev->mdev, rwq_ind_tbl->rqtn); + + kfree(rwq_ind_tbl); + return 0; +} + int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, u32 wq_attr_mask, struct ib_udata *udata) { diff --git a/drivers/infiniband/hw/mlx5/user.h b/drivers/infiniband/hw/mlx5/user.h index 3e66f935e4c8..0f8795546471 100644 --- a/drivers/infiniband/hw/mlx5/user.h +++ b/drivers/infiniband/hw/mlx5/user.h @@ -179,6 +179,11 @@ struct mlx5_ib_create_wq_resp { __u32 reserved; }; +struct mlx5_ib_create_rwq_ind_tbl_resp { + __u32 response_length; + __u32 reserved; +}; + struct mlx5_ib_modify_wq { __u32 comp_mask; __u32 reserved; -- cgit v1.2.3 From a9017e232ff9eaabeb50eb89841d99310cfc98dc Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:54 +0300 Subject: IB/core: Extend create QP to get indirection table Extend create QP to get Receive Work Queue (WQ) indirection table. QP can be created with external Receive Work Queue indirection table, in that case it is ready to receive immediately. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/core/verbs.c | 19 +++++++++++++++++-- include/rdma/ib_verbs.h | 2 ++ 2 files changed, 19 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 6b548d7f8753..6916d5c5920b 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -754,6 +754,12 @@ struct ib_qp *ib_create_qp(struct ib_pd *pd, struct ib_qp *qp; int ret; + if (qp_init_attr->rwq_ind_tbl && + (qp_init_attr->recv_cq || + qp_init_attr->srq || qp_init_attr->cap.max_recv_wr || + qp_init_attr->cap.max_recv_sge)) + return ERR_PTR(-EINVAL); + /* * If the callers is using the RDMA API calculate the resources * needed for the RDMA READ/WRITE operations. @@ -771,6 +777,7 @@ struct ib_qp *ib_create_qp(struct ib_pd *pd, qp->real_qp = qp; qp->uobject = NULL; qp->qp_type = qp_init_attr->qp_type; + qp->rwq_ind_tbl = qp_init_attr->rwq_ind_tbl; atomic_set(&qp->usecnt, 0); qp->mrs_used = 0; @@ -788,7 +795,8 @@ struct ib_qp *ib_create_qp(struct ib_pd *pd, qp->srq = NULL; } else { qp->recv_cq = qp_init_attr->recv_cq; - atomic_inc(&qp_init_attr->recv_cq->usecnt); + if (qp_init_attr->recv_cq) + atomic_inc(&qp_init_attr->recv_cq->usecnt); qp->srq = qp_init_attr->srq; if (qp->srq) atomic_inc(&qp_init_attr->srq->usecnt); @@ -799,7 +807,10 @@ struct ib_qp *ib_create_qp(struct ib_pd *pd, qp->xrcd = NULL; atomic_inc(&pd->usecnt); - atomic_inc(&qp_init_attr->send_cq->usecnt); + if (qp_init_attr->send_cq) + atomic_inc(&qp_init_attr->send_cq->usecnt); + if (qp_init_attr->rwq_ind_tbl) + atomic_inc(&qp->rwq_ind_tbl->usecnt); if (qp_init_attr->cap.max_rdma_ctxs) { ret = rdma_rw_init_mrs(qp, qp_init_attr); @@ -1279,6 +1290,7 @@ int ib_destroy_qp(struct ib_qp *qp) struct ib_pd *pd; struct ib_cq *scq, *rcq; struct ib_srq *srq; + struct ib_rwq_ind_table *ind_tbl; int ret; WARN_ON_ONCE(qp->mrs_used > 0); @@ -1293,6 +1305,7 @@ int ib_destroy_qp(struct ib_qp *qp) scq = qp->send_cq; rcq = qp->recv_cq; srq = qp->srq; + ind_tbl = qp->rwq_ind_tbl; if (!qp->uobject) rdma_rw_cleanup_mrs(qp); @@ -1307,6 +1320,8 @@ int ib_destroy_qp(struct ib_qp *qp) atomic_dec(&rcq->usecnt); if (srq) atomic_dec(&srq->usecnt); + if (ind_tbl) + atomic_dec(&ind_tbl->usecnt); } return ret; diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index e305c9a36663..9b2fafe5eb38 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1017,6 +1017,7 @@ struct ib_qp_init_attr { * Only needed for special QP types, or when using the RW API. */ u8 port_num; + struct ib_rwq_ind_table *rwq_ind_tbl; }; struct ib_qp_open_attr { @@ -1511,6 +1512,7 @@ struct ib_qp { void *qp_context; u32 qp_num; enum ib_qp_type qp_type; + struct ib_rwq_ind_table *rwq_ind_tbl; }; struct ib_mr { -- cgit v1.2.3 From c70285f880e88cb4f73effb722065a182ba5936f Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:55 +0300 Subject: IB/uverbs: Extend create QP to get RWQ indirection table User applications that want to spread incoming traffic between several WQs should create a QP which contains an indirection table. When such a QP is created other receive side parameters are not valid and should not be given. Its send side is optional and assumed active based on max_send_wr capability value. Extend create QP to work accordingly. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/core/uverbs_cmd.c | 75 ++++++++++++++++++++++++++++++------ include/uapi/rdma/ib_user_verbs.h | 10 +++++ 2 files changed, 73 insertions(+), 12 deletions(-) diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 327a56cccc27..65ab2097cd81 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -255,6 +255,17 @@ static void put_wq_read(struct ib_wq *wq) put_uobj_read(wq->uobject); } +static struct ib_rwq_ind_table *idr_read_rwq_indirection_table(int ind_table_handle, + struct ib_ucontext *context) +{ + return idr_read_obj(&ib_uverbs_rwq_ind_tbl_idr, ind_table_handle, context, 0); +} + +static void put_rwq_indirection_table_read(struct ib_rwq_ind_table *ind_table) +{ + put_uobj_read(ind_table->uobject); +} + static struct ib_qp *idr_write_qp(int qp_handle, struct ib_ucontext *context) { struct ib_uobject *uobj; @@ -1761,9 +1772,11 @@ static int create_qp(struct ib_uverbs_file *file, struct ib_srq *srq = NULL; struct ib_qp *qp; char *buf; - struct ib_qp_init_attr attr; + struct ib_qp_init_attr attr = {}; struct ib_uverbs_ex_create_qp_resp resp; int ret; + struct ib_rwq_ind_table *ind_tbl = NULL; + bool has_sq = true; if (cmd->qp_type == IB_QPT_RAW_PACKET && !capable(CAP_NET_RAW)) return -EPERM; @@ -1775,6 +1788,32 @@ static int create_qp(struct ib_uverbs_file *file, init_uobj(&obj->uevent.uobject, cmd->user_handle, file->ucontext, &qp_lock_class); down_write(&obj->uevent.uobject.mutex); + if (cmd_sz >= offsetof(typeof(*cmd), rwq_ind_tbl_handle) + + sizeof(cmd->rwq_ind_tbl_handle) && + (cmd->comp_mask & IB_UVERBS_CREATE_QP_MASK_IND_TABLE)) { + ind_tbl = idr_read_rwq_indirection_table(cmd->rwq_ind_tbl_handle, + file->ucontext); + if (!ind_tbl) { + ret = -EINVAL; + goto err_put; + } + + attr.rwq_ind_tbl = ind_tbl; + } + + if ((cmd_sz >= offsetof(typeof(*cmd), reserved1) + + sizeof(cmd->reserved1)) && cmd->reserved1) { + ret = -EOPNOTSUPP; + goto err_put; + } + + if (ind_tbl && (cmd->max_recv_wr || cmd->max_recv_sge || cmd->is_srq)) { + ret = -EINVAL; + goto err_put; + } + + if (ind_tbl && !cmd->max_send_wr) + has_sq = false; if (cmd->qp_type == IB_QPT_XRC_TGT) { xrcd = idr_read_xrcd(cmd->pd_handle, file->ucontext, @@ -1798,20 +1837,24 @@ static int create_qp(struct ib_uverbs_file *file, } } - if (cmd->recv_cq_handle != cmd->send_cq_handle) { - rcq = idr_read_cq(cmd->recv_cq_handle, - file->ucontext, 0); - if (!rcq) { - ret = -EINVAL; - goto err_put; + if (!ind_tbl) { + if (cmd->recv_cq_handle != cmd->send_cq_handle) { + rcq = idr_read_cq(cmd->recv_cq_handle, + file->ucontext, 0); + if (!rcq) { + ret = -EINVAL; + goto err_put; + } } } } - scq = idr_read_cq(cmd->send_cq_handle, file->ucontext, !!rcq); - rcq = rcq ?: scq; + if (has_sq) + scq = idr_read_cq(cmd->send_cq_handle, file->ucontext, !!rcq); + if (!ind_tbl) + rcq = rcq ?: scq; pd = idr_read_pd(cmd->pd_handle, file->ucontext); - if (!pd || !scq) { + if (!pd || (!scq && has_sq)) { ret = -EINVAL; goto err_put; } @@ -1878,16 +1921,20 @@ static int create_qp(struct ib_uverbs_file *file, qp->send_cq = attr.send_cq; qp->recv_cq = attr.recv_cq; qp->srq = attr.srq; + qp->rwq_ind_tbl = ind_tbl; qp->event_handler = attr.event_handler; qp->qp_context = attr.qp_context; qp->qp_type = attr.qp_type; atomic_set(&qp->usecnt, 0); atomic_inc(&pd->usecnt); - atomic_inc(&attr.send_cq->usecnt); + if (attr.send_cq) + atomic_inc(&attr.send_cq->usecnt); if (attr.recv_cq) atomic_inc(&attr.recv_cq->usecnt); if (attr.srq) atomic_inc(&attr.srq->usecnt); + if (ind_tbl) + atomic_inc(&ind_tbl->usecnt); } qp->uobject = &obj->uevent.uobject; @@ -1927,6 +1974,8 @@ static int create_qp(struct ib_uverbs_file *file, put_cq_read(rcq); if (srq) put_srq_read(srq); + if (ind_tbl) + put_rwq_indirection_table_read(ind_tbl); mutex_lock(&file->mutex); list_add_tail(&obj->uevent.uobject.list, &file->ucontext->qp_list); @@ -1954,6 +2003,8 @@ err_put: put_cq_read(rcq); if (srq) put_srq_read(srq); + if (ind_tbl) + put_rwq_indirection_table_read(ind_tbl); put_uobj_write(&obj->uevent.uobject); return ret; @@ -2047,7 +2098,7 @@ int ib_uverbs_ex_create_qp(struct ib_uverbs_file *file, if (err) return err; - if (cmd.comp_mask) + if (cmd.comp_mask & ~IB_UVERBS_CREATE_QP_SUP_COMP_MASK) return -EINVAL; if (cmd.reserved) diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h index 2cf7c95860d8..2c8bca8cd2c2 100644 --- a/include/uapi/rdma/ib_user_verbs.h +++ b/include/uapi/rdma/ib_user_verbs.h @@ -523,6 +523,14 @@ struct ib_uverbs_create_qp { __u64 driver_data[0]; }; +enum ib_uverbs_create_qp_mask { + IB_UVERBS_CREATE_QP_MASK_IND_TABLE = 1UL << 0, +}; + +enum { + IB_UVERBS_CREATE_QP_SUP_COMP_MASK = IB_UVERBS_CREATE_QP_MASK_IND_TABLE, +}; + struct ib_uverbs_ex_create_qp { __u64 user_handle; __u32 pd_handle; @@ -540,6 +548,8 @@ struct ib_uverbs_ex_create_qp { __u8 reserved; __u32 comp_mask; __u32 create_flags; + __u32 rwq_ind_tbl_handle; + __u32 reserved1; }; struct ib_uverbs_open_qp { -- cgit v1.2.3 From 28d6137008b2aa09e35750c604394e363dbfca94 Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Mon, 23 May 2016 15:20:56 +0300 Subject: IB/mlx5: Add RSS QP support Add support for Raw Ethernet RX HASH QP. Currently, creation and destruction of such a QP are supported. This QP is implemented as a simple TIR object which points to the receive RQ indirection table. The given hashing configuration is used to configure the TIR and by that it chooses the right RQ from the RQ indirection table. Signed-off-by: Yishai Hadas Signed-off-by: Matan Barak Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/mlx5_ib.h | 5 + drivers/infiniband/hw/mlx5/qp.c | 200 +++++++++++++++++++++++++++++++++++ drivers/infiniband/hw/mlx5/user.h | 34 ++++++ 3 files changed, 239 insertions(+) diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index cd3d6203ca1c..7ac4647f7382 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -295,6 +295,10 @@ struct mlx5_ib_qp_trans { u8 resp_depth; }; +struct mlx5_ib_rss_qp { + u32 tirn; +}; + struct mlx5_ib_rq { struct mlx5_ib_qp_base base; struct mlx5_ib_wq *rq; @@ -323,6 +327,7 @@ struct mlx5_ib_qp { union { struct mlx5_ib_qp_trans trans_qp; struct mlx5_ib_raw_packet_qp raw_packet_qp; + struct mlx5_ib_rss_qp rss_qp; }; struct mlx5_buf buf; diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 1c0e332146b5..f9df4b527a11 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -1266,6 +1266,187 @@ static void raw_packet_qp_copy_info(struct mlx5_ib_qp *qp, rq->doorbell = &qp->db; } +static void destroy_rss_raw_qp_tir(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) +{ + mlx5_core_destroy_tir(dev->mdev, qp->rss_qp.tirn); +} + +static int create_rss_raw_qp_tir(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, + struct ib_pd *pd, + struct ib_qp_init_attr *init_attr, + struct ib_udata *udata) +{ + struct ib_uobject *uobj = pd->uobject; + struct ib_ucontext *ucontext = uobj->context; + struct mlx5_ib_ucontext *mucontext = to_mucontext(ucontext); + struct mlx5_ib_create_qp_resp resp = {}; + int inlen; + int err; + u32 *in; + void *tirc; + void *hfso; + u32 selected_fields = 0; + size_t min_resp_len; + u32 tdn = mucontext->tdn; + struct mlx5_ib_create_qp_rss ucmd = {}; + size_t required_cmd_sz; + + if (init_attr->qp_type != IB_QPT_RAW_PACKET) + return -EOPNOTSUPP; + + if (init_attr->create_flags || init_attr->send_cq) + return -EINVAL; + + min_resp_len = offsetof(typeof(resp), uuar_index) + sizeof(resp.uuar_index); + if (udata->outlen < min_resp_len) + return -EINVAL; + + required_cmd_sz = offsetof(typeof(ucmd), reserved1) + sizeof(ucmd.reserved1); + if (udata->inlen < required_cmd_sz) { + mlx5_ib_dbg(dev, "invalid inlen\n"); + return -EINVAL; + } + + if (udata->inlen > sizeof(ucmd) && + !ib_is_udata_cleared(udata, sizeof(ucmd), + udata->inlen - sizeof(ucmd))) { + mlx5_ib_dbg(dev, "inlen is not supported\n"); + return -EOPNOTSUPP; + } + + if (ib_copy_from_udata(&ucmd, udata, min(sizeof(ucmd), udata->inlen))) { + mlx5_ib_dbg(dev, "copy failed\n"); + return -EFAULT; + } + + if (ucmd.comp_mask) { + mlx5_ib_dbg(dev, "invalid comp mask\n"); + return -EOPNOTSUPP; + } + + if (memchr_inv(ucmd.reserved, 0, sizeof(ucmd.reserved)) || ucmd.reserved1) { + mlx5_ib_dbg(dev, "invalid reserved\n"); + return -EOPNOTSUPP; + } + + err = ib_copy_to_udata(udata, &resp, min_resp_len); + if (err) { + mlx5_ib_dbg(dev, "copy failed\n"); + return -EINVAL; + } + + inlen = MLX5_ST_SZ_BYTES(create_tir_in); + in = mlx5_vzalloc(inlen); + if (!in) + return -ENOMEM; + + tirc = MLX5_ADDR_OF(create_tir_in, in, ctx); + MLX5_SET(tirc, tirc, disp_type, + MLX5_TIRC_DISP_TYPE_INDIRECT); + MLX5_SET(tirc, tirc, indirect_table, + init_attr->rwq_ind_tbl->ind_tbl_num); + MLX5_SET(tirc, tirc, transport_domain, tdn); + + hfso = MLX5_ADDR_OF(tirc, tirc, rx_hash_field_selector_outer); + switch (ucmd.rx_hash_function) { + case MLX5_RX_HASH_FUNC_TOEPLITZ: + { + void *rss_key = MLX5_ADDR_OF(tirc, tirc, rx_hash_toeplitz_key); + size_t len = MLX5_FLD_SZ_BYTES(tirc, rx_hash_toeplitz_key); + + if (len != ucmd.rx_key_len) { + err = -EINVAL; + goto err; + } + + MLX5_SET(tirc, tirc, rx_hash_fn, MLX5_RX_HASH_FN_TOEPLITZ); + MLX5_SET(tirc, tirc, rx_hash_symmetric, 1); + memcpy(rss_key, ucmd.rx_hash_key, len); + break; + } + default: + err = -EOPNOTSUPP; + goto err; + } + + if (!ucmd.rx_hash_fields_mask) { + /* special case when this TIR serves as steering entry without hashing */ + if (!init_attr->rwq_ind_tbl->log_ind_tbl_size) + goto create_tir; + err = -EINVAL; + goto err; + } + + if (((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_IPV4) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_IPV4)) && + ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_IPV6) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_IPV6))) { + err = -EINVAL; + goto err; + } + + /* If none of IPV4 & IPV6 SRC/DST was set - this bit field is ignored */ + if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_IPV4) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_IPV4)) + MLX5_SET(rx_hash_field_select, hfso, l3_prot_type, + MLX5_L3_PROT_TYPE_IPV4); + else if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_IPV6) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_IPV6)) + MLX5_SET(rx_hash_field_select, hfso, l3_prot_type, + MLX5_L3_PROT_TYPE_IPV6); + + if (((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_PORT_TCP) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_PORT_TCP)) && + ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_PORT_UDP) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_PORT_UDP))) { + err = -EINVAL; + goto err; + } + + /* If none of TCP & UDP SRC/DST was set - this bit field is ignored */ + if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_PORT_TCP) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_PORT_TCP)) + MLX5_SET(rx_hash_field_select, hfso, l4_prot_type, + MLX5_L4_PROT_TYPE_TCP); + else if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_PORT_UDP) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_PORT_UDP)) + MLX5_SET(rx_hash_field_select, hfso, l4_prot_type, + MLX5_L4_PROT_TYPE_UDP); + + if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_IPV4) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_IPV6)) + selected_fields |= MLX5_HASH_FIELD_SEL_SRC_IP; + + if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_IPV4) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_IPV6)) + selected_fields |= MLX5_HASH_FIELD_SEL_DST_IP; + + if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_PORT_TCP) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_SRC_PORT_UDP)) + selected_fields |= MLX5_HASH_FIELD_SEL_L4_SPORT; + + if ((ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_PORT_TCP) || + (ucmd.rx_hash_fields_mask & MLX5_RX_HASH_DST_PORT_UDP)) + selected_fields |= MLX5_HASH_FIELD_SEL_L4_DPORT; + + MLX5_SET(rx_hash_field_select, hfso, selected_fields, selected_fields); + +create_tir: + err = mlx5_core_create_tir(dev->mdev, in, inlen, &qp->rss_qp.tirn); + + if (err) + goto err; + + kvfree(in); + /* qpn is reserved for that QP */ + qp->trans_qp.base.mqp.qpn = 0; + return 0; + +err: + kvfree(in); + return err; +} + static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, struct ib_qp_init_attr *init_attr, struct ib_udata *udata, struct mlx5_ib_qp *qp) @@ -1292,6 +1473,14 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, spin_lock_init(&qp->sq.lock); spin_lock_init(&qp->rq.lock); + if (init_attr->rwq_ind_tbl) { + if (!udata) + return -ENOSYS; + + err = create_rss_raw_qp_tir(dev, qp, pd, init_attr, udata); + return err; + } + if (init_attr->create_flags & IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK) { if (!MLX5_CAP_GEN(mdev, block_lb_mc)) { mlx5_ib_dbg(dev, "block multicast loopback isn't supported\n"); @@ -1644,6 +1833,11 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) struct mlx5_modify_qp_mbox_in *in; int err; + if (qp->ibqp.rwq_ind_tbl) { + destroy_rss_raw_qp_tir(dev, qp); + return; + } + base = qp->ibqp.qp_type == IB_QPT_RAW_PACKET ? &qp->raw_packet_qp.rq.base : &qp->trans_qp.base; @@ -2504,6 +2698,9 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int port; enum rdma_link_layer ll = IB_LINK_LAYER_UNSPECIFIED; + if (ibqp->rwq_ind_tbl) + return -ENOSYS; + if (unlikely(ibqp->qp_type == IB_QPT_GSI)) return mlx5_ib_gsi_modify_qp(ibqp, attr, attr_mask); @@ -4119,6 +4316,9 @@ int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int err = 0; u8 raw_packet_qp_state; + if (ibqp->rwq_ind_tbl) + return -ENOSYS; + if (unlikely(ibqp->qp_type == IB_QPT_GSI)) return mlx5_ib_gsi_query_qp(ibqp, qp_attr, qp_attr_mask, qp_init_attr); diff --git a/drivers/infiniband/hw/mlx5/user.h b/drivers/infiniband/hw/mlx5/user.h index 0f8795546471..33c54fbab1e7 100644 --- a/drivers/infiniband/hw/mlx5/user.h +++ b/drivers/infiniband/hw/mlx5/user.h @@ -152,6 +152,40 @@ struct mlx5_ib_create_qp { __u64 sq_buf_addr; }; +/* RX Hash function flags */ +enum mlx5_rx_hash_function_flags { + MLX5_RX_HASH_FUNC_TOEPLITZ = 1 << 0, +}; + +/* + * RX Hash flags, these flags allows to set which incoming packet's field should + * participates in RX Hash. Each flag represent certain packet's field, + * when the flag is set the field that is represented by the flag will + * participate in RX Hash calculation. + * Note: *IPV4 and *IPV6 flags can't be enabled together on the same QP + * and *TCP and *UDP flags can't be enabled together on the same QP. +*/ +enum mlx5_rx_hash_fields { + MLX5_RX_HASH_SRC_IPV4 = 1 << 0, + MLX5_RX_HASH_DST_IPV4 = 1 << 1, + MLX5_RX_HASH_SRC_IPV6 = 1 << 2, + MLX5_RX_HASH_DST_IPV6 = 1 << 3, + MLX5_RX_HASH_SRC_PORT_TCP = 1 << 4, + MLX5_RX_HASH_DST_PORT_TCP = 1 << 5, + MLX5_RX_HASH_SRC_PORT_UDP = 1 << 6, + MLX5_RX_HASH_DST_PORT_UDP = 1 << 7 +}; + +struct mlx5_ib_create_qp_rss { + __u64 rx_hash_fields_mask; /* enum mlx5_rx_hash_fields */ + __u8 rx_hash_function; /* enum mlx5_rx_hash_function_flags */ + __u8 rx_key_len; /* valid only for Toeplitz */ + __u8 reserved[6]; + __u8 rx_hash_key[128]; /* valid only for Toeplitz */ + __u32 comp_mask; + __u32 reserved1; +}; + struct mlx5_ib_create_qp_resp { __u32 uuar_index; }; -- cgit v1.2.3 From 7c2344c3bbf97eb5dfa732d5098285d15d3bf9bf Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Fri, 17 Jun 2016 14:56:44 +0300 Subject: IB/mlx5: Implements disassociate_ucontext API Implements the IB core disassociate_ucontext API. The driver detaches the HW resources for a given user context to prevent a dependency between application termination and device disconnect. This is done by managing the VMAs that were mapped to the HW bars such as doorbell and blueflame. When need to detach, remap them to an arbitrary kernel page returned by the zap API. Signed-off-by: Maor Gottlieb Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 133 +++++++++++++++++++++++++++++++++-- drivers/infiniband/hw/mlx5/mlx5_ib.h | 6 ++ 2 files changed, 135 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index 82be19447a3f..21acee41f0ec 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -42,11 +42,13 @@ #include #endif #include +#include #include #include #include #include #include +#include #include #include #include @@ -983,6 +985,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, goto out_uars; } + INIT_LIST_HEAD(&context->vma_private_list); INIT_LIST_HEAD(&context->db_page_list); mutex_init(&context->db_page_mutex); @@ -1086,6 +1089,125 @@ static int get_index(unsigned long offset) return get_arg(offset); } +static void mlx5_ib_vma_open(struct vm_area_struct *area) +{ + /* vma_open is called when a new VMA is created on top of our VMA. This + * is done through either mremap flow or split_vma (usually due to + * mlock, madvise, munmap, etc.) We do not support a clone of the VMA, + * as this VMA is strongly hardware related. Therefore we set the + * vm_ops of the newly created/cloned VMA to NULL, to prevent it from + * calling us again and trying to do incorrect actions. We assume that + * the original VMA size is exactly a single page, and therefore all + * "splitting" operation will not happen to it. + */ + area->vm_ops = NULL; +} + +static void mlx5_ib_vma_close(struct vm_area_struct *area) +{ + struct mlx5_ib_vma_private_data *mlx5_ib_vma_priv_data; + + /* It's guaranteed that all VMAs opened on a FD are closed before the + * file itself is closed, therefore no sync is needed with the regular + * closing flow. (e.g. mlx5 ib_dealloc_ucontext) + * However need a sync with accessing the vma as part of + * mlx5_ib_disassociate_ucontext. + * The close operation is usually called under mm->mmap_sem except when + * process is exiting. + * The exiting case is handled explicitly as part of + * mlx5_ib_disassociate_ucontext. + */ + mlx5_ib_vma_priv_data = (struct mlx5_ib_vma_private_data *)area->vm_private_data; + + /* setting the vma context pointer to null in the mlx5_ib driver's + * private data, to protect a race condition in + * mlx5_ib_disassociate_ucontext(). + */ + mlx5_ib_vma_priv_data->vma = NULL; + list_del(&mlx5_ib_vma_priv_data->list); + kfree(mlx5_ib_vma_priv_data); +} + +static const struct vm_operations_struct mlx5_ib_vm_ops = { + .open = mlx5_ib_vma_open, + .close = mlx5_ib_vma_close +}; + +static int mlx5_ib_set_vma_data(struct vm_area_struct *vma, + struct mlx5_ib_ucontext *ctx) +{ + struct mlx5_ib_vma_private_data *vma_prv; + struct list_head *vma_head = &ctx->vma_private_list; + + vma_prv = kzalloc(sizeof(*vma_prv), GFP_KERNEL); + if (!vma_prv) + return -ENOMEM; + + vma_prv->vma = vma; + vma->vm_private_data = vma_prv; + vma->vm_ops = &mlx5_ib_vm_ops; + + list_add(&vma_prv->list, vma_head); + + return 0; +} + +static void mlx5_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) +{ + int ret; + struct vm_area_struct *vma; + struct mlx5_ib_vma_private_data *vma_private, *n; + struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); + struct task_struct *owning_process = NULL; + struct mm_struct *owning_mm = NULL; + + owning_process = get_pid_task(ibcontext->tgid, PIDTYPE_PID); + if (!owning_process) + return; + + owning_mm = get_task_mm(owning_process); + if (!owning_mm) { + pr_info("no mm, disassociate ucontext is pending task termination\n"); + while (1) { + put_task_struct(owning_process); + usleep_range(1000, 2000); + owning_process = get_pid_task(ibcontext->tgid, + PIDTYPE_PID); + if (!owning_process || + owning_process->state == TASK_DEAD) { + pr_info("disassociate ucontext done, task was terminated\n"); + /* in case task was dead need to release the + * task struct. + */ + if (owning_process) + put_task_struct(owning_process); + return; + } + } + } + + /* need to protect from a race on closing the vma as part of + * mlx5_ib_vma_close. + */ + down_read(&owning_mm->mmap_sem); + list_for_each_entry_safe(vma_private, n, &context->vma_private_list, + list) { + vma = vma_private->vma; + ret = zap_vma_ptes(vma, vma->vm_start, + PAGE_SIZE); + WARN_ONCE(ret, "%s: zap_vma_ptes failed", __func__); + /* context going to be destroyed, should + * not access ops any more. + */ + vma->vm_ops = NULL; + list_del(&vma_private->list); + kfree(vma_private); + } + up_read(&owning_mm->mmap_sem); + mmput(owning_mm); + put_task_struct(owning_process); +} + static inline char *mmap_cmd2str(enum mlx5_ib_mmap_cmd cmd) { switch (cmd) { @@ -1101,8 +1223,10 @@ static inline char *mmap_cmd2str(enum mlx5_ib_mmap_cmd cmd) } static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, - struct vm_area_struct *vma, struct mlx5_uuar_info *uuari) + struct vm_area_struct *vma, + struct mlx5_ib_ucontext *context) { + struct mlx5_uuar_info *uuari = &context->uuari; int err; unsigned long idx; phys_addr_t pfn, pa; @@ -1152,14 +1276,13 @@ static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, mlx5_ib_dbg(dev, "mapped %s at 0x%lx, PA %pa\n", mmap_cmd2str(cmd), vma->vm_start, &pa); - return 0; + return mlx5_ib_set_vma_data(vma, context); } static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vma) { struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct mlx5_ib_dev *dev = to_mdev(ibcontext->device); - struct mlx5_uuar_info *uuari = &context->uuari; unsigned long command; phys_addr_t pfn; @@ -1168,7 +1291,7 @@ static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vm case MLX5_IB_MMAP_WC_PAGE: case MLX5_IB_MMAP_NC_PAGE: case MLX5_IB_MMAP_REGULAR_PAGE: - return uar_mmap(dev, command, vma, uuari); + return uar_mmap(dev, command, vma, context); case MLX5_IB_MMAP_GET_CONTIGUOUS_PAGES: return -ENOSYS; @@ -2428,6 +2551,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->ib_dev.set_vf_guid = mlx5_ib_set_vf_guid; } + dev->ib_dev.disassociate_ucontext = mlx5_ib_disassociate_ucontext; + mlx5_ib_internal_fill_odp_caps(dev); if (MLX5_CAP_GEN(mdev, imaicl)) { diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 7ac4647f7382..391588e01317 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -105,6 +105,11 @@ enum { MLX5_CQE_VERSION_V1, }; +struct mlx5_ib_vma_private_data { + struct list_head list; + struct vm_area_struct *vma; +}; + struct mlx5_ib_ucontext { struct ib_ucontext ibucontext; struct list_head db_page_list; @@ -116,6 +121,7 @@ struct mlx5_ib_ucontext { u8 cqe_version; /* Transport Domain number */ u32 tdn; + struct list_head vma_private_list; }; static inline struct mlx5_ib_ucontext *to_mucontext(struct ib_ucontext *ibucontext) -- cgit v1.2.3 From 89ea94a7b6c40eb423c144aef1caceebaff79c8d Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Fri, 17 Jun 2016 15:01:38 +0300 Subject: IB/mlx5: Reset flow support for IB kernel ULPs The driver exposes interfaces that directly relate to HW state. Upon fatal error, consumers of these interfaces (ULPs) that rely on completion of all their posted work-request could hang, thereby introducing dependencies in shutdown order. To prevent this from happening, we manage the relevant resources (CQs, QPs) that are used by the device. Upon a fatal error, we now generate simulated completions for outstanding WQEs that were not completed at the time the HW was reset. It includes invoking the completion event handler for all involved CQs so that the ULPs will poll those CQs. When polled we return simulated CQEs with IB_WC_WR_FLUSH_ERR return code enabling ULPs to clean up their resources and not wait forever for completions upon receiving remove_one. The above change requires an extra check in the data path to make sure that when device is in error state, the simulated CQEs will be returned and no further WQEs will be posted. Signed-off-by: Maor Gottlieb Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/cq.c | 87 +++++++++++++++++++++++++++++++- drivers/infiniband/hw/mlx5/main.c | 62 +++++++++++++++++++++++ drivers/infiniband/hw/mlx5/mlx5_ib.h | 8 +++ drivers/infiniband/hw/mlx5/mr.c | 4 ++ drivers/infiniband/hw/mlx5/qp.c | 98 +++++++++++++++++++++++++++++------- drivers/infiniband/hw/mlx5/srq.c | 10 +++- include/linux/mlx5/cq.h | 2 + 7 files changed, 250 insertions(+), 21 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c index 9c0e67bd2ba7..308a358e5b46 100644 --- a/drivers/infiniband/hw/mlx5/cq.c +++ b/drivers/infiniband/hw/mlx5/cq.c @@ -424,6 +424,83 @@ static void get_sig_err_item(struct mlx5_sig_err_cqe *cqe, item->key = be32_to_cpu(cqe->mkey); } +static void sw_send_comp(struct mlx5_ib_qp *qp, int num_entries, + struct ib_wc *wc, int *npolled) +{ + struct mlx5_ib_wq *wq; + unsigned int cur; + unsigned int idx; + int np; + int i; + + wq = &qp->sq; + cur = wq->head - wq->tail; + np = *npolled; + + if (cur == 0) + return; + + for (i = 0; i < cur && np < num_entries; i++) { + idx = wq->last_poll & (wq->wqe_cnt - 1); + wc->wr_id = wq->wrid[idx]; + wc->status = IB_WC_WR_FLUSH_ERR; + wc->vendor_err = MLX5_CQE_SYNDROME_WR_FLUSH_ERR; + wq->tail++; + np++; + wc->qp = &qp->ibqp; + wc++; + wq->last_poll = wq->w_list[idx].next; + } + *npolled = np; +} + +static void sw_recv_comp(struct mlx5_ib_qp *qp, int num_entries, + struct ib_wc *wc, int *npolled) +{ + struct mlx5_ib_wq *wq; + unsigned int cur; + int np; + int i; + + wq = &qp->rq; + cur = wq->head - wq->tail; + np = *npolled; + + if (cur == 0) + return; + + for (i = 0; i < cur && np < num_entries; i++) { + wc->wr_id = wq->wrid[wq->tail & (wq->wqe_cnt - 1)]; + wc->status = IB_WC_WR_FLUSH_ERR; + wc->vendor_err = MLX5_CQE_SYNDROME_WR_FLUSH_ERR; + wq->tail++; + np++; + wc->qp = &qp->ibqp; + wc++; + } + *npolled = np; +} + +static void mlx5_ib_poll_sw_comp(struct mlx5_ib_cq *cq, int num_entries, + struct ib_wc *wc, int *npolled) +{ + struct mlx5_ib_qp *qp; + + *npolled = 0; + /* Find uncompleted WQEs belonging to that cq and retrun mmics ones */ + list_for_each_entry(qp, &cq->list_send_qp, cq_send_list) { + sw_send_comp(qp, num_entries, wc + *npolled, npolled); + if (*npolled >= num_entries) + return; + } + + list_for_each_entry(qp, &cq->list_recv_qp, cq_recv_list) { + sw_recv_comp(qp, num_entries, wc + *npolled, npolled); + if (*npolled >= num_entries) + return; + } +} + static int mlx5_poll_one(struct mlx5_ib_cq *cq, struct mlx5_ib_qp **cur_qp, struct ib_wc *wc) @@ -594,12 +671,18 @@ int mlx5_ib_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc) { struct mlx5_ib_cq *cq = to_mcq(ibcq); struct mlx5_ib_qp *cur_qp = NULL; + struct mlx5_ib_dev *dev = to_mdev(cq->ibcq.device); + struct mlx5_core_dev *mdev = dev->mdev; unsigned long flags; int soft_polled = 0; int npolled; int err = 0; spin_lock_irqsave(&cq->lock, flags); + if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) { + mlx5_ib_poll_sw_comp(cq, num_entries, wc, &npolled); + goto out; + } if (unlikely(!list_empty(&cq->wc_list))) soft_polled = poll_soft_wc(cq, num_entries, wc); @@ -612,7 +695,7 @@ int mlx5_ib_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc) if (npolled) mlx5_cq_set_ci(&cq->mcq); - +out: spin_unlock_irqrestore(&cq->lock, flags); if (err == 0 || err == -EAGAIN) @@ -843,6 +926,8 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, cq->resize_buf = NULL; cq->resize_umem = NULL; cq->create_flags = attr->flags; + INIT_LIST_HEAD(&cq->list_send_qp); + INIT_LIST_HEAD(&cq->list_recv_qp); if (context) { err = create_cq_user(dev, udata, context, cq, entries, diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index 21acee41f0ec..9b6d2838072d 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -1980,6 +1980,65 @@ static void pkey_change_handler(struct work_struct *work) mutex_unlock(&ports->devr->mutex); } +static void mlx5_ib_handle_internal_error(struct mlx5_ib_dev *ibdev) +{ + struct mlx5_ib_qp *mqp; + struct mlx5_ib_cq *send_mcq, *recv_mcq; + struct mlx5_core_cq *mcq; + struct list_head cq_armed_list; + unsigned long flags_qp; + unsigned long flags_cq; + unsigned long flags; + + INIT_LIST_HEAD(&cq_armed_list); + + /* Go over qp list reside on that ibdev, sync with create/destroy qp.*/ + spin_lock_irqsave(&ibdev->reset_flow_resource_lock, flags); + list_for_each_entry(mqp, &ibdev->qp_list, qps_list) { + spin_lock_irqsave(&mqp->sq.lock, flags_qp); + if (mqp->sq.tail != mqp->sq.head) { + send_mcq = to_mcq(mqp->ibqp.send_cq); + spin_lock_irqsave(&send_mcq->lock, flags_cq); + if (send_mcq->mcq.comp && + mqp->ibqp.send_cq->comp_handler) { + if (!send_mcq->mcq.reset_notify_added) { + send_mcq->mcq.reset_notify_added = 1; + list_add_tail(&send_mcq->mcq.reset_notify, + &cq_armed_list); + } + } + spin_unlock_irqrestore(&send_mcq->lock, flags_cq); + } + spin_unlock_irqrestore(&mqp->sq.lock, flags_qp); + spin_lock_irqsave(&mqp->rq.lock, flags_qp); + /* no handling is needed for SRQ */ + if (!mqp->ibqp.srq) { + if (mqp->rq.tail != mqp->rq.head) { + recv_mcq = to_mcq(mqp->ibqp.recv_cq); + spin_lock_irqsave(&recv_mcq->lock, flags_cq); + if (recv_mcq->mcq.comp && + mqp->ibqp.recv_cq->comp_handler) { + if (!recv_mcq->mcq.reset_notify_added) { + recv_mcq->mcq.reset_notify_added = 1; + list_add_tail(&recv_mcq->mcq.reset_notify, + &cq_armed_list); + } + } + spin_unlock_irqrestore(&recv_mcq->lock, + flags_cq); + } + } + spin_unlock_irqrestore(&mqp->rq.lock, flags_qp); + } + /*At that point all inflight post send were put to be executed as of we + * lock/unlock above locks Now need to arm all involved CQs. + */ + list_for_each_entry(mcq, &cq_armed_list, reset_notify) { + mcq->comp(mcq); + } + spin_unlock_irqrestore(&ibdev->reset_flow_resource_lock, flags); +} + static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, enum mlx5_dev_event event, unsigned long param) { @@ -1992,6 +2051,7 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, case MLX5_DEV_EVENT_SYS_ERROR: ibdev->ib_active = false; ibev.event = IB_EVENT_DEVICE_FATAL; + mlx5_ib_handle_internal_error(ibdev); break; case MLX5_DEV_EVENT_PORT_UP: @@ -2595,6 +2655,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) mutex_init(&dev->flow_db.lock); mutex_init(&dev->cap_mask_mutex); + INIT_LIST_HEAD(&dev->qp_list); + spin_lock_init(&dev->reset_flow_resource_lock); if (ll == IB_LINK_LAYER_ETHERNET) { err = mlx5_enable_roce(dev); diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 391588e01317..0001ed51e1bd 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -380,6 +380,9 @@ struct mlx5_ib_qp { spinlock_t disable_page_faults_lock; struct mlx5_ib_pfault pagefaults[MLX5_IB_PAGEFAULT_CONTEXTS]; #endif + struct list_head qps_list; + struct list_head cq_recv_list; + struct list_head cq_send_list; }; struct mlx5_ib_cq_buf { @@ -441,6 +444,8 @@ struct mlx5_ib_cq { struct mlx5_ib_cq_buf *resize_buf; struct ib_umem *resize_umem; int cqe_size; + struct list_head list_send_qp; + struct list_head list_recv_qp; u32 create_flags; struct list_head wc_list; enum ib_cq_notify_flags notify_flags; @@ -621,6 +626,9 @@ struct mlx5_ib_dev { struct srcu_struct mr_srcu; #endif struct mlx5_ib_flow_db flow_db; + /* protect resources needed as part of reset flow */ + spinlock_t reset_flow_resource_lock; + struct list_head qp_list; }; static inline struct mlx5_ib_cq *to_mibcq(struct mlx5_core_cq *mcq) diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c index 8cf2ce50511f..4b021305c321 100644 --- a/drivers/infiniband/hw/mlx5/mr.c +++ b/drivers/infiniband/hw/mlx5/mr.c @@ -1193,12 +1193,16 @@ error: static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { + struct mlx5_core_dev *mdev = dev->mdev; struct umr_common *umrc = &dev->umrc; struct mlx5_ib_umr_context umr_context; struct mlx5_umr_wr umrwr = {}; struct ib_send_wr *bad; int err; + if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) + return 0; + mlx5_ib_init_umr_context(&umr_context); umrwr.wr.wr_cqe = &umr_context.cqe; diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index f9df4b527a11..9004905d3d93 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -77,6 +77,10 @@ struct mlx5_wqe_eth_pad { u8 rsvd0[16]; }; +static void get_cqs(enum ib_qp_type qp_type, + struct ib_cq *ib_send_cq, struct ib_cq *ib_recv_cq, + struct mlx5_ib_cq **send_cq, struct mlx5_ib_cq **recv_cq); + static int is_qp0(enum ib_qp_type qp_type) { return qp_type == IB_QPT_SMI; @@ -609,6 +613,11 @@ static int to_mlx5_st(enum ib_qp_type type) } } +static void mlx5_ib_lock_cqs(struct mlx5_ib_cq *send_cq, + struct mlx5_ib_cq *recv_cq); +static void mlx5_ib_unlock_cqs(struct mlx5_ib_cq *send_cq, + struct mlx5_ib_cq *recv_cq); + static int uuarn_to_uar_index(struct mlx5_uuar_info *uuari, int uuarn) { return uuari->uars[uuarn / MLX5_BF_REGS_PER_PAGE].index; @@ -1457,6 +1466,9 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, struct mlx5_ib_create_qp_resp resp; struct mlx5_create_qp_mbox_in *in; struct mlx5_ib_create_qp ucmd; + struct mlx5_ib_cq *send_cq; + struct mlx5_ib_cq *recv_cq; + unsigned long flags; int inlen = sizeof(*in); int err; u32 uidx = MLX5_IB_DEFAULT_UIDX; @@ -1714,6 +1726,23 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, base->container_mibqp = qp; base->mqp.event = mlx5_ib_qp_event; + get_cqs(init_attr->qp_type, init_attr->send_cq, init_attr->recv_cq, + &send_cq, &recv_cq); + spin_lock_irqsave(&dev->reset_flow_resource_lock, flags); + mlx5_ib_lock_cqs(send_cq, recv_cq); + /* Maintain device to QPs access, needed for further handling via reset + * flow + */ + list_add_tail(&qp->qps_list, &dev->qp_list); + /* Maintain CQ to QPs access, needed for further handling via reset flow + */ + if (send_cq) + list_add_tail(&qp->cq_send_list, &send_cq->list_send_qp); + if (recv_cq) + list_add_tail(&qp->cq_recv_list, &recv_cq->list_recv_qp); + mlx5_ib_unlock_cqs(send_cq, recv_cq); + spin_unlock_irqrestore(&dev->reset_flow_resource_lock, flags); + return 0; err_create: @@ -1732,23 +1761,23 @@ static void mlx5_ib_lock_cqs(struct mlx5_ib_cq *send_cq, struct mlx5_ib_cq *recv if (send_cq) { if (recv_cq) { if (send_cq->mcq.cqn < recv_cq->mcq.cqn) { - spin_lock_irq(&send_cq->lock); + spin_lock(&send_cq->lock); spin_lock_nested(&recv_cq->lock, SINGLE_DEPTH_NESTING); } else if (send_cq->mcq.cqn == recv_cq->mcq.cqn) { - spin_lock_irq(&send_cq->lock); + spin_lock(&send_cq->lock); __acquire(&recv_cq->lock); } else { - spin_lock_irq(&recv_cq->lock); + spin_lock(&recv_cq->lock); spin_lock_nested(&send_cq->lock, SINGLE_DEPTH_NESTING); } } else { - spin_lock_irq(&send_cq->lock); + spin_lock(&send_cq->lock); __acquire(&recv_cq->lock); } } else if (recv_cq) { - spin_lock_irq(&recv_cq->lock); + spin_lock(&recv_cq->lock); __acquire(&send_cq->lock); } else { __acquire(&send_cq->lock); @@ -1763,21 +1792,21 @@ static void mlx5_ib_unlock_cqs(struct mlx5_ib_cq *send_cq, struct mlx5_ib_cq *re if (recv_cq) { if (send_cq->mcq.cqn < recv_cq->mcq.cqn) { spin_unlock(&recv_cq->lock); - spin_unlock_irq(&send_cq->lock); + spin_unlock(&send_cq->lock); } else if (send_cq->mcq.cqn == recv_cq->mcq.cqn) { __release(&recv_cq->lock); - spin_unlock_irq(&send_cq->lock); + spin_unlock(&send_cq->lock); } else { spin_unlock(&send_cq->lock); - spin_unlock_irq(&recv_cq->lock); + spin_unlock(&recv_cq->lock); } } else { __release(&recv_cq->lock); - spin_unlock_irq(&send_cq->lock); + spin_unlock(&send_cq->lock); } } else if (recv_cq) { __release(&send_cq->lock); - spin_unlock_irq(&recv_cq->lock); + spin_unlock(&recv_cq->lock); } else { __release(&recv_cq->lock); __release(&send_cq->lock); @@ -1789,17 +1818,18 @@ static struct mlx5_ib_pd *get_pd(struct mlx5_ib_qp *qp) return to_mpd(qp->ibqp.pd); } -static void get_cqs(struct mlx5_ib_qp *qp, +static void get_cqs(enum ib_qp_type qp_type, + struct ib_cq *ib_send_cq, struct ib_cq *ib_recv_cq, struct mlx5_ib_cq **send_cq, struct mlx5_ib_cq **recv_cq) { - switch (qp->ibqp.qp_type) { + switch (qp_type) { case IB_QPT_XRC_TGT: *send_cq = NULL; *recv_cq = NULL; break; case MLX5_IB_QPT_REG_UMR: case IB_QPT_XRC_INI: - *send_cq = to_mcq(qp->ibqp.send_cq); + *send_cq = ib_send_cq ? to_mcq(ib_send_cq) : NULL; *recv_cq = NULL; break; @@ -1811,8 +1841,8 @@ static void get_cqs(struct mlx5_ib_qp *qp, case IB_QPT_RAW_IPV6: case IB_QPT_RAW_ETHERTYPE: case IB_QPT_RAW_PACKET: - *send_cq = to_mcq(qp->ibqp.send_cq); - *recv_cq = to_mcq(qp->ibqp.recv_cq); + *send_cq = ib_send_cq ? to_mcq(ib_send_cq) : NULL; + *recv_cq = ib_recv_cq ? to_mcq(ib_recv_cq) : NULL; break; case IB_QPT_MAX: @@ -1831,6 +1861,7 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) struct mlx5_ib_cq *send_cq, *recv_cq; struct mlx5_ib_qp_base *base = &qp->trans_qp.base; struct mlx5_modify_qp_mbox_in *in; + unsigned long flags; int err; if (qp->ibqp.rwq_ind_tbl) { @@ -1861,17 +1892,28 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) base->mqp.qpn); } - get_cqs(qp, &send_cq, &recv_cq); + get_cqs(qp->ibqp.qp_type, qp->ibqp.send_cq, qp->ibqp.recv_cq, + &send_cq, &recv_cq); + + spin_lock_irqsave(&dev->reset_flow_resource_lock, flags); + mlx5_ib_lock_cqs(send_cq, recv_cq); + /* del from lists under both locks above to protect reset flow paths */ + list_del(&qp->qps_list); + if (send_cq) + list_del(&qp->cq_send_list); + + if (recv_cq) + list_del(&qp->cq_recv_list); if (qp->create_type == MLX5_QP_KERNEL) { - mlx5_ib_lock_cqs(send_cq, recv_cq); __mlx5_ib_cq_clean(recv_cq, base->mqp.qpn, qp->ibqp.srq ? to_msrq(qp->ibqp.srq) : NULL); if (send_cq != recv_cq) __mlx5_ib_cq_clean(send_cq, base->mqp.qpn, NULL); - mlx5_ib_unlock_cqs(send_cq, recv_cq); } + mlx5_ib_unlock_cqs(send_cq, recv_cq); + spin_unlock_irqrestore(&dev->reset_flow_resource_lock, flags); if (qp->ibqp.qp_type == IB_QPT_RAW_PACKET) { destroy_raw_packet_qp(dev, qp); @@ -2559,7 +2601,8 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, } pd = get_pd(qp); - get_cqs(qp, &send_cq, &recv_cq); + get_cqs(qp->ibqp.qp_type, qp->ibqp.send_cq, qp->ibqp.recv_cq, + &send_cq, &recv_cq); context->flags_pd = cpu_to_be32(pd ? pd->pdn : to_mpd(dev->devr.p0)->pdn); context->cqn_send = send_cq ? cpu_to_be32(send_cq->mcq.cqn) : 0; @@ -3658,6 +3701,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, { struct mlx5_wqe_ctrl_seg *ctrl = NULL; /* compiler warning */ struct mlx5_ib_dev *dev = to_mdev(ibqp->device); + struct mlx5_core_dev *mdev = dev->mdev; struct mlx5_ib_qp *qp; struct mlx5_ib_mr *mr; struct mlx5_wqe_data_seg *dpseg; @@ -3685,6 +3729,13 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, spin_lock_irqsave(&qp->sq.lock, flags); + if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) { + err = -EIO; + *bad_wr = wr; + nreq = 0; + goto out; + } + for (nreq = 0; wr; nreq++, wr = wr->next) { if (unlikely(wr->opcode >= ARRAY_SIZE(mlx5_ib_opcode))) { mlx5_ib_warn(dev, "\n"); @@ -3986,6 +4037,8 @@ int mlx5_ib_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, struct mlx5_ib_qp *qp = to_mqp(ibqp); struct mlx5_wqe_data_seg *scat; struct mlx5_rwqe_sig *sig; + struct mlx5_ib_dev *dev = to_mdev(ibqp->device); + struct mlx5_core_dev *mdev = dev->mdev; unsigned long flags; int err = 0; int nreq; @@ -3997,6 +4050,13 @@ int mlx5_ib_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, spin_lock_irqsave(&qp->rq.lock, flags); + if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) { + err = -EIO; + *bad_wr = wr; + nreq = 0; + goto out; + } + ind = qp->rq.head & (qp->rq.wqe_cnt - 1); for (nreq = 0; wr; nreq++, wr = wr->next) { diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c index 3b2ddd64a371..55efb343137b 100644 --- a/drivers/infiniband/hw/mlx5/srq.c +++ b/drivers/infiniband/hw/mlx5/srq.c @@ -458,6 +458,8 @@ int mlx5_ib_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, struct mlx5_ib_srq *srq = to_msrq(ibsrq); struct mlx5_wqe_srq_next_seg *next; struct mlx5_wqe_data_seg *scat; + struct mlx5_ib_dev *dev = to_mdev(ibsrq->device); + struct mlx5_core_dev *mdev = dev->mdev; unsigned long flags; int err = 0; int nreq; @@ -465,6 +467,12 @@ int mlx5_ib_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, spin_lock_irqsave(&srq->lock, flags); + if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) { + err = -EIO; + *bad_wr = wr; + goto out; + } + for (nreq = 0; wr; nreq++, wr = wr->next) { if (unlikely(wr->num_sge > srq->msrq.max_gs)) { err = -EINVAL; @@ -507,7 +515,7 @@ int mlx5_ib_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, *srq->db.db = cpu_to_be32(srq->wqe_ctr); } - +out: spin_unlock_irqrestore(&srq->lock, flags); return err; diff --git a/include/linux/mlx5/cq.h b/include/linux/mlx5/cq.h index 2be976dd4966..2566f6d6444f 100644 --- a/include/linux/mlx5/cq.h +++ b/include/linux/mlx5/cq.h @@ -58,6 +58,8 @@ struct mlx5_core_cq { void (*comp)(struct mlx5_core_cq *); void *priv; } tasklet_ctx; + int reset_notify_added; + struct list_head reset_notify; }; -- cgit v1.2.3 From 4c2aae712cb024f9d30a1fa62e3ba2ff785c6a3e Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Fri, 17 Jun 2016 15:14:50 +0300 Subject: IB/core: Add IPv6 support to flow steering Add IPv6 flow specification support. Signed-off-by: Maor Gottlieb Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/core/uverbs.h | 1 + drivers/infiniband/core/uverbs_cmd.c | 9 +++++++++ include/rdma/ib_verbs.h | 14 ++++++++++++++ include/uapi/rdma/ib_user_verbs.h | 18 ++++++++++++++++++ 4 files changed, 42 insertions(+) diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h index 6c2292397cd6..b7f3b8d359b9 100644 --- a/drivers/infiniband/core/uverbs.h +++ b/drivers/infiniband/core/uverbs.h @@ -226,6 +226,7 @@ struct ib_uverbs_flow_spec { struct ib_uverbs_flow_spec_eth eth; struct ib_uverbs_flow_spec_ipv4 ipv4; struct ib_uverbs_flow_spec_tcp_udp tcp_udp; + struct ib_uverbs_flow_spec_ipv6 ipv6; }; }; diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 65ab2097cd81..f6647318138d 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -3105,6 +3105,15 @@ static int kern_spec_to_ib_spec(struct ib_uverbs_flow_spec *kern_spec, memcpy(&ib_spec->ipv4.mask, &kern_spec->ipv4.mask, sizeof(struct ib_flow_ipv4_filter)); break; + case IB_FLOW_SPEC_IPV6: + ib_spec->ipv6.size = sizeof(struct ib_flow_spec_ipv6); + if (ib_spec->ipv6.size != kern_spec->ipv6.size) + return -EINVAL; + memcpy(&ib_spec->ipv6.val, &kern_spec->ipv6.val, + sizeof(struct ib_flow_ipv6_filter)); + memcpy(&ib_spec->ipv6.mask, &kern_spec->ipv6.mask, + sizeof(struct ib_flow_ipv6_filter)); + break; case IB_FLOW_SPEC_TCP: case IB_FLOW_SPEC_UDP: ib_spec->tcp_udp.size = sizeof(struct ib_flow_spec_tcp_udp); diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 9b2fafe5eb38..9bbca6887f7f 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1569,6 +1569,7 @@ enum ib_flow_spec_type { IB_FLOW_SPEC_IB = 0x22, /* L3 header*/ IB_FLOW_SPEC_IPV4 = 0x30, + IB_FLOW_SPEC_IPV6 = 0x31, /* L4 headers*/ IB_FLOW_SPEC_TCP = 0x40, IB_FLOW_SPEC_UDP = 0x41 @@ -1630,6 +1631,18 @@ struct ib_flow_spec_ipv4 { struct ib_flow_ipv4_filter mask; }; +struct ib_flow_ipv6_filter { + u8 src_ip[16]; + u8 dst_ip[16]; +}; + +struct ib_flow_spec_ipv6 { + enum ib_flow_spec_type type; + u16 size; + struct ib_flow_ipv6_filter val; + struct ib_flow_ipv6_filter mask; +}; + struct ib_flow_tcp_udp_filter { __be16 dst_port; __be16 src_port; @@ -1651,6 +1664,7 @@ union ib_flow_spec { struct ib_flow_spec_ib ib; struct ib_flow_spec_ipv4 ipv4; struct ib_flow_spec_tcp_udp tcp_udp; + struct ib_flow_spec_ipv6 ipv6; }; struct ib_flow_attr { diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h index 2c8bca8cd2c2..7f035f4b53b0 100644 --- a/include/uapi/rdma/ib_user_verbs.h +++ b/include/uapi/rdma/ib_user_verbs.h @@ -867,6 +867,24 @@ struct ib_uverbs_flow_spec_tcp_udp { struct ib_uverbs_flow_tcp_udp_filter mask; }; +struct ib_uverbs_flow_ipv6_filter { + __u8 src_ip[16]; + __u8 dst_ip[16]; +}; + +struct ib_uverbs_flow_spec_ipv6 { + union { + struct ib_uverbs_flow_spec_hdr hdr; + struct { + __u32 type; + __u16 size; + __u16 reserved; + }; + }; + struct ib_uverbs_flow_ipv6_filter val; + struct ib_uverbs_flow_ipv6_filter mask; +}; + struct ib_uverbs_flow_attr { __u32 type; __u16 size; -- cgit v1.2.3 From 026bae0cb428102228d110780d90e6ae44bbe4c7 Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Fri, 17 Jun 2016 15:14:51 +0300 Subject: IB/mlx5: Enable flow steering for IPv6 traffic Enable flow steering for IPv6 traffic by using an IPv6 spec. Signed-off-by: Maor Gottlieb Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index 9b6d2838072d..8f832fc20341 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -1454,6 +1454,32 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, &ib_spec->ipv4.val.dst_ip, sizeof(ib_spec->ipv4.val.dst_ip)); break; + case IB_FLOW_SPEC_IPV6: + if (ib_spec->size != sizeof(ib_spec->ipv6)) + return -EINVAL; + + MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, + ethertype, 0xffff); + MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, + ethertype, ETH_P_IPV6); + + memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, + src_ipv4_src_ipv6.ipv6_layout.ipv6), + &ib_spec->ipv6.mask.src_ip, + sizeof(ib_spec->ipv6.mask.src_ip)); + memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, + src_ipv4_src_ipv6.ipv6_layout.ipv6), + &ib_spec->ipv6.val.src_ip, + sizeof(ib_spec->ipv6.val.src_ip)); + memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, + dst_ipv4_dst_ipv6.ipv6_layout.ipv6), + &ib_spec->ipv6.mask.dst_ip, + sizeof(ib_spec->ipv6.mask.dst_ip)); + memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, + dst_ipv4_dst_ipv6.ipv6_layout.ipv6), + &ib_spec->ipv6.val.dst_ip, + sizeof(ib_spec->ipv6.val.dst_ip)); + break; case IB_FLOW_SPEC_TCP: if (ib_spec->size != sizeof(ib_spec->tcp_udp)) return -EINVAL; -- cgit v1.2.3 From 402ca53644ff1928f2ea68cce6fe41b2b5f38510 Mon Sep 17 00:00:00 2001 From: Bodong Wang Date: Fri, 17 Jun 2016 15:02:20 +0300 Subject: IB/mlx5: Report mlx5 TSO capabilities when querying device Enable mlx5 based hardware to report TCP segmentation offload (TSO) capabilities from kernel to user space. A TSO enabled NIC will accept big chunks of data with sizes greater than MTU for TCP traffic. The TSO engine will break the data into separate packets and will insert headers automatically. The capabilities are exposed to user space through query_device by uhw directly. The following capabilities are reported: 1. The maximum payload size in bytes supported for segmentation by TSO engine. 2. Bitmap showing which QP types are supported by TSO operation. The bitmap is built by members from 'enmu ib_qp_type'. For example, similar code should be performed if UD QP is supported: supported_qpts |= 1 << IB_QPT_UD; To make user-space library aware of whether kernel supports uhw or not, a new flag: cmds_supp_uhw will be returned back to user-space through alloc_ucontext. Signed-off-by: Bodong Wang Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 41 ++++++++++++++++++++++++++++++++++----- drivers/infiniband/hw/mlx5/user.h | 24 +++++++++++++++++++++-- 2 files changed, 58 insertions(+), 7 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index 8f832fc20341..a7cc6d74d3b0 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -459,8 +459,17 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, int max_rq_sg; int max_sq_sg; u64 min_page_size = 1ull << MLX5_CAP_GEN(mdev, log_pg_sz); + struct mlx5_ib_query_device_resp resp = {}; + size_t resp_len; + u64 max_tso; - if (uhw->inlen || uhw->outlen) + resp_len = sizeof(resp.comp_mask) + sizeof(resp.response_length); + if (uhw->outlen && uhw->outlen < resp_len) + return -EINVAL; + else + resp.response_length = resp_len; + + if (uhw->inlen && !ib_is_udata_cleared(uhw, 0, uhw->inlen)) return -EINVAL; memset(props, 0, sizeof(*props)); @@ -513,10 +522,21 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, if (MLX5_CAP_GEN(mdev, block_lb_mc)) props->device_cap_flags |= IB_DEVICE_BLOCK_MULTICAST_LOOPBACK; - if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && - (MLX5_CAP_ETH(dev->mdev, csum_cap))) + if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads)) { + if (MLX5_CAP_ETH(mdev, csum_cap)) props->device_cap_flags |= IB_DEVICE_RAW_IP_CSUM; + if (field_avail(typeof(resp), tso_caps, uhw->outlen)) { + max_tso = MLX5_CAP_ETH(mdev, max_lso_cap); + if (max_tso) { + resp.tso_caps.max_tso = 1 << max_tso; + resp.tso_caps.supported_qpts |= + 1 << IB_QPT_RAW_PACKET; + resp.response_length += sizeof(resp.tso_caps); + } + } + } + if (MLX5_CAP_GEN(mdev, ipoib_basic_offloads)) { props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM; props->device_cap_flags |= IB_DEVICE_UD_TSO; @@ -578,6 +598,13 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, if (!mlx5_core_is_pf(mdev)) props->device_cap_flags |= IB_DEVICE_VIRTUAL_FUNCTION; + if (uhw->outlen) { + err = ib_copy_to_udata(uhw, &resp, resp.response_length); + + if (err) + return err; + } + return 0; } @@ -995,6 +1022,11 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, if (field_avail(typeof(resp), cqe_version, udata->outlen)) resp.response_length += sizeof(resp.cqe_version); + if (field_avail(typeof(resp), cmds_supp_uhw, udata->outlen)) { + resp.cmds_supp_uhw |= MLX5_USER_CMDS_SUPP_UHW_QUERY_DEVICE; + resp.response_length += sizeof(resp.cmds_supp_uhw); + } + /* * We don't want to expose information from the PCI bar that is located * after 4096 bytes, so if the arch only supports larger pages, let's @@ -1009,8 +1041,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, offsetof(struct mlx5_init_seg, internal_timer_h) % PAGE_SIZE; resp.response_length += sizeof(resp.hca_core_clock_offset) + - sizeof(resp.reserved2) + - sizeof(resp.reserved3); + sizeof(resp.reserved2); } err = ib_copy_to_udata(udata, &resp, resp.response_length); diff --git a/drivers/infiniband/hw/mlx5/user.h b/drivers/infiniband/hw/mlx5/user.h index 33c54fbab1e7..188dac4301b5 100644 --- a/drivers/infiniband/hw/mlx5/user.h +++ b/drivers/infiniband/hw/mlx5/user.h @@ -83,6 +83,10 @@ enum mlx5_ib_alloc_ucontext_resp_mask { MLX5_IB_ALLOC_UCONTEXT_RESP_MASK_CORE_CLOCK_OFFSET = 1UL << 0, }; +enum mlx5_user_cmds_supp_uhw { + MLX5_USER_CMDS_SUPP_UHW_QUERY_DEVICE = 1 << 0, +}; + struct mlx5_ib_alloc_ucontext_resp { __u32 qp_tab_size; __u32 bf_reg_size; @@ -98,8 +102,8 @@ struct mlx5_ib_alloc_ucontext_resp { __u32 comp_mask; __u32 response_length; __u8 cqe_version; - __u8 reserved2; - __u16 reserved3; + __u8 cmds_supp_uhw; + __u16 reserved2; __u64 hca_core_clock_offset; }; @@ -107,6 +111,22 @@ struct mlx5_ib_alloc_pd_resp { __u32 pdn; }; +struct mlx5_ib_tso_caps { + __u32 max_tso; /* Maximum tso payload size in bytes */ + + /* Corresponding bit will be set if qp type from + * 'enum ib_qp_type' is supported, e.g. + * supported_qpts |= 1 << IB_QPT_UD + */ + __u32 supported_qpts; +}; + +struct mlx5_ib_query_device_resp { + __u32 comp_mask; + __u32 response_length; + struct mlx5_ib_tso_caps tso_caps; +}; + struct mlx5_ib_create_cq { __u64 buf_addr; __u64 db_addr; -- cgit v1.2.3 From e3353c268b06236d6c40fa1714c114f21f44451c Mon Sep 17 00:00:00 2001 From: Artemy Kovalyov Date: Fri, 17 Jun 2016 15:33:31 +0300 Subject: IB/mlx5: Fix MODIFY_QP command input structure Make MODIFY_QP command input structure compliant to specification Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters') Signed-off-by: Artemy Kovalyov Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- include/linux/mlx5/qp.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/linux/mlx5/qp.h b/include/linux/mlx5/qp.h index 266320feb160..acb28c989dee 100644 --- a/include/linux/mlx5/qp.h +++ b/include/linux/mlx5/qp.h @@ -555,9 +555,9 @@ struct mlx5_destroy_qp_mbox_out { struct mlx5_modify_qp_mbox_in { struct mlx5_inbox_hdr hdr; __be32 qpn; - u8 rsvd1[4]; - __be32 optparam; u8 rsvd0[4]; + __be32 optparam; + u8 rsvd1[4]; struct mlx5_qp_context ctx; u8 rsvd2[16]; }; -- cgit v1.2.3 From af1ba291c5e498973cc325c501dd8da80b234571 Mon Sep 17 00:00:00 2001 From: Artemy Kovalyov Date: Fri, 17 Jun 2016 15:33:32 +0300 Subject: {net, IB}/mlx5: Refactor internal SRQ API Currently, the SRQ API uses the obsolete mlx5_*_srq_mbox_{in,out} structs which limit the ability to pass the SRQ attributes between net and IB parts of the driver. This patch changes the SRQ API so as to use auto-generated structs and provides a better way to pass attributes which will be in use by coming features. Signed-off-by: Artemy Kovalyov Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/srq.c | 102 ++++------ drivers/net/ethernet/mellanox/mlx5/core/srq.c | 265 ++++++++++++++------------ include/linux/mlx5/driver.h | 6 +- include/linux/mlx5/srq.h | 25 +++ 4 files changed, 215 insertions(+), 183 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c index 55efb343137b..ed6ac52355f1 100644 --- a/drivers/infiniband/hw/mlx5/srq.c +++ b/drivers/infiniband/hw/mlx5/srq.c @@ -74,14 +74,12 @@ static void mlx5_ib_srq_event(struct mlx5_core_srq *srq, enum mlx5_event type) } static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq, - struct mlx5_create_srq_mbox_in **in, - struct ib_udata *udata, int buf_size, int *inlen, - int is_xrc) + struct mlx5_srq_attr *in, + struct ib_udata *udata, int buf_size) { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct mlx5_ib_create_srq ucmd = {}; size_t ucmdlen; - void *xsrqc; int err; int npages; int page_shift; @@ -104,7 +102,7 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq, udata->inlen - sizeof(ucmd))) return -EINVAL; - if (is_xrc) { + if (in->type == IB_SRQT_XRC) { err = get_srq_user_index(to_mucontext(pd->uobject->context), &ucmd, udata->inlen, &uidx); if (err) @@ -130,14 +128,13 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq, goto err_umem; } - *inlen = sizeof(**in) + sizeof(*(*in)->pas) * ncont; - *in = mlx5_vzalloc(*inlen); - if (!(*in)) { + in->pas = mlx5_vzalloc(sizeof(*in->pas) * ncont); + if (!in->pas) { err = -ENOMEM; goto err_umem; } - mlx5_ib_populate_pas(dev, srq->umem, page_shift, (*in)->pas, 0); + mlx5_ib_populate_pas(dev, srq->umem, page_shift, in->pas, 0); err = mlx5_ib_db_map_user(to_mucontext(pd->uobject->context), ucmd.db_addr, &srq->db); @@ -146,20 +143,16 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq, goto err_in; } - (*in)->ctx.log_pg_sz = page_shift - MLX5_ADAPTER_PAGE_SHIFT; - (*in)->ctx.pgoff_cqn = cpu_to_be32(offset << 26); - - if ((MLX5_CAP_GEN(dev->mdev, cqe_version) == MLX5_CQE_VERSION_V1) && - is_xrc){ - xsrqc = MLX5_ADDR_OF(create_xrc_srq_in, *in, - xrc_srq_context_entry); - MLX5_SET(xrc_srqc, xsrqc, user_index, uidx); - } + in->log_page_size = page_shift - MLX5_ADAPTER_PAGE_SHIFT; + in->page_offset = offset; + if (MLX5_CAP_GEN(dev->mdev, cqe_version) == MLX5_CQE_VERSION_V1 && + in->type == IB_SRQT_XRC) + in->user_index = uidx; return 0; err_in: - kvfree(*in); + kvfree(in->pas); err_umem: ib_umem_release(srq->umem); @@ -168,15 +161,13 @@ err_umem: } static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq, - struct mlx5_create_srq_mbox_in **in, int buf_size, - int *inlen, int is_xrc) + struct mlx5_srq_attr *in, int buf_size) { int err; int i; struct mlx5_wqe_srq_next_seg *next; int page_shift; int npages; - void *xsrqc; err = mlx5_db_alloc(dev->mdev, &srq->db); if (err) { @@ -204,13 +195,12 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq, npages = DIV_ROUND_UP(srq->buf.npages, 1 << (page_shift - PAGE_SHIFT)); mlx5_ib_dbg(dev, "buf_size %d, page_shift %d, npages %d, calc npages %d\n", buf_size, page_shift, srq->buf.npages, npages); - *inlen = sizeof(**in) + sizeof(*(*in)->pas) * npages; - *in = mlx5_vzalloc(*inlen); - if (!*in) { + in->pas = mlx5_vzalloc(sizeof(*in->pas) * npages); + if (!in->pas) { err = -ENOMEM; goto err_buf; } - mlx5_fill_page_array(&srq->buf, (*in)->pas); + mlx5_fill_page_array(&srq->buf, in->pas); srq->wrid = kmalloc(srq->msrq.max * sizeof(u64), GFP_KERNEL); if (!srq->wrid) { @@ -221,20 +211,15 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq, } srq->wq_sig = !!srq_signature; - (*in)->ctx.log_pg_sz = page_shift - MLX5_ADAPTER_PAGE_SHIFT; - - if ((MLX5_CAP_GEN(dev->mdev, cqe_version) == MLX5_CQE_VERSION_V1) && - is_xrc){ - xsrqc = MLX5_ADDR_OF(create_xrc_srq_in, *in, - xrc_srq_context_entry); - /* 0xffffff means we ask to work with cqe version 0 */ - MLX5_SET(xrc_srqc, xsrqc, user_index, MLX5_IB_DEFAULT_UIDX); - } + in->log_page_size = page_shift - MLX5_ADAPTER_PAGE_SHIFT; + if (MLX5_CAP_GEN(dev->mdev, cqe_version) == MLX5_CQE_VERSION_V1 && + in->type == IB_SRQT_XRC) + in->user_index = MLX5_IB_DEFAULT_UIDX; return 0; err_in: - kvfree(*in); + kvfree(in->pas); err_buf: mlx5_buf_free(dev->mdev, &srq->buf); @@ -267,10 +252,7 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd, int desc_size; int buf_size; int err; - struct mlx5_create_srq_mbox_in *uninitialized_var(in); - int uninitialized_var(inlen); - int is_xrc; - u32 flgs, xrcdn; + struct mlx5_srq_attr in = {0}; __u32 max_srq_wqes = 1 << MLX5_CAP_GEN(dev->mdev, log_max_srq_sz); /* Sanity check SRQ size before proceeding */ @@ -302,14 +284,10 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd, desc_size, init_attr->attr.max_wr, srq->msrq.max, srq->msrq.max_gs, srq->msrq.max_avail_gather); - is_xrc = (init_attr->srq_type == IB_SRQT_XRC); - if (pd->uobject) - err = create_srq_user(pd, srq, &in, udata, buf_size, &inlen, - is_xrc); + err = create_srq_user(pd, srq, &in, udata, buf_size); else - err = create_srq_kernel(dev, srq, &in, buf_size, &inlen, - is_xrc); + err = create_srq_kernel(dev, srq, &in, buf_size); if (err) { mlx5_ib_warn(dev, "create srq %s failed, err %d\n", @@ -317,23 +295,23 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd, goto err_srq; } - in->ctx.state_log_sz = ilog2(srq->msrq.max); - flgs = ((srq->msrq.wqe_shift - 4) | (is_xrc << 5) | (srq->wq_sig << 7)) << 24; - xrcdn = 0; - if (is_xrc) { - xrcdn = to_mxrcd(init_attr->ext.xrc.xrcd)->xrcdn; - in->ctx.pgoff_cqn |= cpu_to_be32(to_mcq(init_attr->ext.xrc.cq)->mcq.cqn); + in.type = init_attr->srq_type; + in.log_size = ilog2(srq->msrq.max); + in.wqe_shift = srq->msrq.wqe_shift - 4; + if (srq->wq_sig) + in.flags |= MLX5_SRQ_FLAG_WQ_SIG; + if (init_attr->srq_type == IB_SRQT_XRC) { + in.xrcd = to_mxrcd(init_attr->ext.xrc.xrcd)->xrcdn; + in.cqn = to_mcq(init_attr->ext.xrc.cq)->mcq.cqn; } else if (init_attr->srq_type == IB_SRQT_BASIC) { - xrcdn = to_mxrcd(dev->devr.x0)->xrcdn; - in->ctx.pgoff_cqn |= cpu_to_be32(to_mcq(dev->devr.c0)->mcq.cqn); + in.xrcd = to_mxrcd(dev->devr.x0)->xrcdn; + in.cqn = to_mcq(dev->devr.c0)->mcq.cqn; } - in->ctx.flags_xrcd = cpu_to_be32((flgs & 0xFF000000) | (xrcdn & 0xFFFFFF)); - - in->ctx.pd = cpu_to_be32(to_mpd(pd)->pdn); - in->ctx.db_record = cpu_to_be64(srq->db.dma); - err = mlx5_core_create_srq(dev->mdev, &srq->msrq, in, inlen, is_xrc); - kvfree(in); + in.pd = to_mpd(pd)->pdn; + in.db_record = srq->db.dma; + err = mlx5_core_create_srq(dev->mdev, &srq->msrq, &in); + kvfree(in.pas); if (err) { mlx5_ib_dbg(dev, "create SRQ failed, err %d\n", err); goto err_usr_kern_srq; @@ -401,7 +379,7 @@ int mlx5_ib_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *srq_attr) struct mlx5_ib_dev *dev = to_mdev(ibsrq->device); struct mlx5_ib_srq *srq = to_msrq(ibsrq); int ret; - struct mlx5_query_srq_mbox_out *out; + struct mlx5_srq_attr *out; out = kzalloc(sizeof(*out), GFP_KERNEL); if (!out) @@ -411,7 +389,7 @@ int mlx5_ib_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *srq_attr) if (ret) goto out_box; - srq_attr->srq_limit = be16_to_cpu(out->ctx.lwm); + srq_attr->srq_limit = out->lwm; srq_attr->max_wr = srq->msrq.max - 1; srq_attr->max_sge = srq->msrq.max_gs; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/srq.c b/drivers/net/ethernet/mellanox/mlx5/core/srq.c index 04bc522605a0..c07f4d01b70e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/srq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/srq.c @@ -63,12 +63,12 @@ void mlx5_srq_event(struct mlx5_core_dev *dev, u32 srqn, int event_type) complete(&srq->free); } -static int get_pas_size(void *srqc) +static int get_pas_size(struct mlx5_srq_attr *in) { - u32 log_page_size = MLX5_GET(srqc, srqc, log_page_size) + 12; - u32 log_srq_size = MLX5_GET(srqc, srqc, log_srq_size); - u32 log_rq_stride = MLX5_GET(srqc, srqc, log_rq_stride); - u32 page_offset = MLX5_GET(srqc, srqc, page_offset); + u32 log_page_size = in->log_page_size + 12; + u32 log_srq_size = in->log_size; + u32 log_rq_stride = in->wqe_shift; + u32 page_offset = in->page_offset; u32 po_quanta = 1 << (log_page_size - 6); u32 rq_sz = 1 << (log_srq_size + 4 + log_rq_stride); u32 page_size = 1 << log_page_size; @@ -78,57 +78,58 @@ static int get_pas_size(void *srqc) return rq_num_pas * sizeof(u64); } -static void rmpc_srqc_reformat(void *srqc, void *rmpc, bool srqc_to_rmpc) +static void set_wq(void *wq, struct mlx5_srq_attr *in) { - void *wq = MLX5_ADDR_OF(rmpc, rmpc, wq); - - if (srqc_to_rmpc) { - switch (MLX5_GET(srqc, srqc, state)) { - case MLX5_SRQC_STATE_GOOD: - MLX5_SET(rmpc, rmpc, state, MLX5_RMPC_STATE_RDY); - break; - case MLX5_SRQC_STATE_ERROR: - MLX5_SET(rmpc, rmpc, state, MLX5_RMPC_STATE_ERR); - break; - default: - pr_warn("%s: %d: Unknown srq state = 0x%x\n", __func__, - __LINE__, MLX5_GET(srqc, srqc, state)); - MLX5_SET(rmpc, rmpc, state, MLX5_GET(srqc, srqc, state)); - } - - MLX5_SET(wq, wq, wq_signature, MLX5_GET(srqc, srqc, wq_signature)); - MLX5_SET(wq, wq, log_wq_pg_sz, MLX5_GET(srqc, srqc, log_page_size)); - MLX5_SET(wq, wq, log_wq_stride, MLX5_GET(srqc, srqc, log_rq_stride) + 4); - MLX5_SET(wq, wq, log_wq_sz, MLX5_GET(srqc, srqc, log_srq_size)); - MLX5_SET(wq, wq, page_offset, MLX5_GET(srqc, srqc, page_offset)); - MLX5_SET(wq, wq, lwm, MLX5_GET(srqc, srqc, lwm)); - MLX5_SET(wq, wq, pd, MLX5_GET(srqc, srqc, pd)); - MLX5_SET64(wq, wq, dbr_addr, MLX5_GET64(srqc, srqc, dbr_addr)); - } else { - switch (MLX5_GET(rmpc, rmpc, state)) { - case MLX5_RMPC_STATE_RDY: - MLX5_SET(srqc, srqc, state, MLX5_SRQC_STATE_GOOD); - break; - case MLX5_RMPC_STATE_ERR: - MLX5_SET(srqc, srqc, state, MLX5_SRQC_STATE_ERROR); - break; - default: - pr_warn("%s: %d: Unknown rmp state = 0x%x\n", - __func__, __LINE__, - MLX5_GET(rmpc, rmpc, state)); - MLX5_SET(srqc, srqc, state, - MLX5_GET(rmpc, rmpc, state)); - } - - MLX5_SET(srqc, srqc, wq_signature, MLX5_GET(wq, wq, wq_signature)); - MLX5_SET(srqc, srqc, log_page_size, MLX5_GET(wq, wq, log_wq_pg_sz)); - MLX5_SET(srqc, srqc, log_rq_stride, MLX5_GET(wq, wq, log_wq_stride) - 4); - MLX5_SET(srqc, srqc, log_srq_size, MLX5_GET(wq, wq, log_wq_sz)); - MLX5_SET(srqc, srqc, page_offset, MLX5_GET(wq, wq, page_offset)); - MLX5_SET(srqc, srqc, lwm, MLX5_GET(wq, wq, lwm)); - MLX5_SET(srqc, srqc, pd, MLX5_GET(wq, wq, pd)); - MLX5_SET64(srqc, srqc, dbr_addr, MLX5_GET64(wq, wq, dbr_addr)); - } + MLX5_SET(wq, wq, wq_signature, !!(in->flags + & MLX5_SRQ_FLAG_WQ_SIG)); + MLX5_SET(wq, wq, log_wq_pg_sz, in->log_page_size); + MLX5_SET(wq, wq, log_wq_stride, in->wqe_shift + 4); + MLX5_SET(wq, wq, log_wq_sz, in->log_size); + MLX5_SET(wq, wq, page_offset, in->page_offset); + MLX5_SET(wq, wq, lwm, in->lwm); + MLX5_SET(wq, wq, pd, in->pd); + MLX5_SET64(wq, wq, dbr_addr, in->db_record); +} + +static void set_srqc(void *srqc, struct mlx5_srq_attr *in) +{ + MLX5_SET(srqc, srqc, wq_signature, !!(in->flags + & MLX5_SRQ_FLAG_WQ_SIG)); + MLX5_SET(srqc, srqc, log_page_size, in->log_page_size); + MLX5_SET(srqc, srqc, log_rq_stride, in->wqe_shift); + MLX5_SET(srqc, srqc, log_srq_size, in->log_size); + MLX5_SET(srqc, srqc, page_offset, in->page_offset); + MLX5_SET(srqc, srqc, lwm, in->lwm); + MLX5_SET(srqc, srqc, pd, in->pd); + MLX5_SET64(srqc, srqc, dbr_addr, in->db_record); + MLX5_SET(srqc, srqc, xrcd, in->xrcd); + MLX5_SET(srqc, srqc, cqn, in->cqn); +} + +static void get_wq(void *wq, struct mlx5_srq_attr *in) +{ + if (MLX5_GET(wq, wq, wq_signature)) + in->flags &= MLX5_SRQ_FLAG_WQ_SIG; + in->log_page_size = MLX5_GET(wq, wq, log_wq_pg_sz); + in->wqe_shift = MLX5_GET(wq, wq, log_wq_stride) - 4; + in->log_size = MLX5_GET(wq, wq, log_wq_sz); + in->page_offset = MLX5_GET(wq, wq, page_offset); + in->lwm = MLX5_GET(wq, wq, lwm); + in->pd = MLX5_GET(wq, wq, pd); + in->db_record = MLX5_GET64(wq, wq, dbr_addr); +} + +static void get_srqc(void *srqc, struct mlx5_srq_attr *in) +{ + if (MLX5_GET(srqc, srqc, wq_signature)) + in->flags &= MLX5_SRQ_FLAG_WQ_SIG; + in->log_page_size = MLX5_GET(srqc, srqc, log_page_size); + in->wqe_shift = MLX5_GET(srqc, srqc, log_rq_stride); + in->log_size = MLX5_GET(srqc, srqc, log_srq_size); + in->page_offset = MLX5_GET(srqc, srqc, page_offset); + in->lwm = MLX5_GET(srqc, srqc, lwm); + in->pd = MLX5_GET(srqc, srqc, pd); + in->db_record = MLX5_GET64(srqc, srqc, dbr_addr); } struct mlx5_core_srq *mlx5_core_get_srq(struct mlx5_core_dev *dev, u32 srqn) @@ -149,19 +150,36 @@ struct mlx5_core_srq *mlx5_core_get_srq(struct mlx5_core_dev *dev, u32 srqn) EXPORT_SYMBOL(mlx5_core_get_srq); static int create_srq_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_create_srq_mbox_in *in, int inlen) + struct mlx5_srq_attr *in) { - struct mlx5_create_srq_mbox_out out; + u32 create_out[MLX5_ST_SZ_DW(create_srq_out)] = {0}; + void *create_in; + void *srqc; + void *pas; + int pas_size; + int inlen; int err; - memset(&out, 0, sizeof(out)); + pas_size = get_pas_size(in); + inlen = MLX5_ST_SZ_BYTES(create_srq_in) + pas_size; + create_in = mlx5_vzalloc(inlen); + if (!create_in) + return -ENOMEM; + + srqc = MLX5_ADDR_OF(create_srq_in, create_in, srq_context_entry); + pas = MLX5_ADDR_OF(create_srq_in, create_in, pas); - in->hdr.opcode = cpu_to_be16(MLX5_CMD_OP_CREATE_SRQ); + set_srqc(srqc, in); + memcpy(pas, in->pas, pas_size); - err = mlx5_cmd_exec_check_status(dev, (u32 *)in, inlen, (u32 *)(&out), - sizeof(out)); + MLX5_SET(create_srq_in, create_in, opcode, + MLX5_CMD_OP_CREATE_SRQ); - srq->srqn = be32_to_cpu(out.srqn) & 0xffffff; + err = mlx5_cmd_exec_check_status(dev, create_in, inlen, create_out, + sizeof(create_out)); + kvfree(create_in); + if (!err) + srq->srqn = MLX5_GET(create_srq_out, create_out, srqn); return err; } @@ -169,67 +187,75 @@ static int create_srq_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, static int destroy_srq_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq) { - struct mlx5_destroy_srq_mbox_in in; - struct mlx5_destroy_srq_mbox_out out; + u32 srq_in[MLX5_ST_SZ_DW(destroy_srq_in)] = {0}; + u32 srq_out[MLX5_ST_SZ_DW(destroy_srq_out)] = {0}; - memset(&in, 0, sizeof(in)); - memset(&out, 0, sizeof(out)); - in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_DESTROY_SRQ); - in.srqn = cpu_to_be32(srq->srqn); + MLX5_SET(destroy_srq_in, srq_in, opcode, + MLX5_CMD_OP_DESTROY_SRQ); + MLX5_SET(destroy_srq_in, srq_in, srqn, srq->srqn); - return mlx5_cmd_exec_check_status(dev, (u32 *)(&in), sizeof(in), - (u32 *)(&out), sizeof(out)); + return mlx5_cmd_exec_check_status(dev, srq_in, sizeof(srq_in), + srq_out, sizeof(srq_out)); } static int arm_srq_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, u16 lwm, int is_srq) { - struct mlx5_arm_srq_mbox_in in; - struct mlx5_arm_srq_mbox_out out; - - memset(&in, 0, sizeof(in)); - memset(&out, 0, sizeof(out)); + /* arm_srq structs missing using identical xrc ones */ + u32 srq_in[MLX5_ST_SZ_DW(arm_xrc_srq_in)] = {0}; + u32 srq_out[MLX5_ST_SZ_DW(arm_xrc_srq_out)] = {0}; - in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_ARM_RQ); - in.hdr.opmod = cpu_to_be16(!!is_srq); - in.srqn = cpu_to_be32(srq->srqn); - in.lwm = cpu_to_be16(lwm); + MLX5_SET(arm_xrc_srq_in, srq_in, opcode, MLX5_CMD_OP_ARM_XRC_SRQ); + MLX5_SET(arm_xrc_srq_in, srq_in, xrc_srqn, srq->srqn); + MLX5_SET(arm_xrc_srq_in, srq_in, lwm, lwm); - return mlx5_cmd_exec_check_status(dev, (u32 *)(&in), - sizeof(in), (u32 *)(&out), - sizeof(out)); + return mlx5_cmd_exec_check_status(dev, srq_in, sizeof(srq_in), + srq_out, sizeof(srq_out)); } static int query_srq_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_query_srq_mbox_out *out) + struct mlx5_srq_attr *out) { - struct mlx5_query_srq_mbox_in in; + u32 srq_in[MLX5_ST_SZ_DW(query_srq_in)] = {0}; + u32 *srq_out; + void *srqc; + int err; - memset(&in, 0, sizeof(in)); + srq_out = mlx5_vzalloc(MLX5_ST_SZ_BYTES(query_srq_out)); + if (!srq_out) + return -ENOMEM; - in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_SRQ); - in.srqn = cpu_to_be32(srq->srqn); + MLX5_SET(query_srq_in, srq_in, opcode, + MLX5_CMD_OP_QUERY_SRQ); + MLX5_SET(query_srq_in, srq_in, srqn, srq->srqn); + err = mlx5_cmd_exec_check_status(dev, srq_in, sizeof(srq_in), + srq_out, + MLX5_ST_SZ_BYTES(query_srq_out)); + if (err) + goto out; - return mlx5_cmd_exec_check_status(dev, (u32 *)(&in), sizeof(in), - (u32 *)out, sizeof(*out)); + srqc = MLX5_ADDR_OF(query_srq_out, srq_out, srq_context_entry); + get_srqc(srqc, out); + if (MLX5_GET(srqc, srqc, state) != MLX5_SRQC_STATE_GOOD) + out->flags |= MLX5_SRQ_FLAG_ERR; +out: + kvfree(srq_out); + return err; } static int create_xrc_srq_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_create_srq_mbox_in *in, - int srq_inlen) + struct mlx5_srq_attr *in) { u32 create_out[MLX5_ST_SZ_DW(create_xrc_srq_out)]; void *create_in; - void *srqc; void *xrc_srqc; void *pas; int pas_size; int inlen; int err; - srqc = MLX5_ADDR_OF(create_srq_in, in, srq_context_entry); - pas_size = get_pas_size(srqc); + pas_size = get_pas_size(in); inlen = MLX5_ST_SZ_BYTES(create_xrc_srq_in) + pas_size; create_in = mlx5_vzalloc(inlen); if (!create_in) @@ -239,7 +265,8 @@ static int create_xrc_srq_cmd(struct mlx5_core_dev *dev, xrc_srq_context_entry); pas = MLX5_ADDR_OF(create_xrc_srq_in, create_in, pas); - memcpy(xrc_srqc, srqc, MLX5_ST_SZ_BYTES(srqc)); + set_srqc(xrc_srqc, in); + MLX5_SET(xrc_srqc, xrc_srqc, user_index, in->user_index); memcpy(pas, in->pas, pas_size); MLX5_SET(create_xrc_srq_in, create_in, opcode, MLX5_CMD_OP_CREATE_XRC_SRQ); @@ -293,11 +320,10 @@ static int arm_xrc_srq_cmd(struct mlx5_core_dev *dev, static int query_xrc_srq_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_query_srq_mbox_out *out) + struct mlx5_srq_attr *out) { u32 xrcsrq_in[MLX5_ST_SZ_DW(query_xrc_srq_in)]; u32 *xrcsrq_out; - void *srqc; void *xrc_srqc; int err; @@ -317,8 +343,9 @@ static int query_xrc_srq_cmd(struct mlx5_core_dev *dev, xrc_srqc = MLX5_ADDR_OF(query_xrc_srq_out, xrcsrq_out, xrc_srq_context_entry); - srqc = MLX5_ADDR_OF(query_srq_out, out, srq_context_entry); - memcpy(srqc, xrc_srqc, MLX5_ST_SZ_BYTES(srqc)); + get_srqc(xrc_srqc, out); + if (MLX5_GET(xrc_srqc, xrc_srqc, state) != MLX5_XRC_SRQC_STATE_GOOD) + out->flags |= MLX5_SRQ_FLAG_ERR; out: kvfree(xrcsrq_out); @@ -326,26 +353,27 @@ out: } static int create_rmp_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_create_srq_mbox_in *in, int srq_inlen) + struct mlx5_srq_attr *in) { void *create_in; void *rmpc; - void *srqc; + void *wq; int pas_size; int inlen; int err; - srqc = MLX5_ADDR_OF(create_srq_in, in, srq_context_entry); - pas_size = get_pas_size(srqc); + pas_size = get_pas_size(in); inlen = MLX5_ST_SZ_BYTES(create_rmp_in) + pas_size; create_in = mlx5_vzalloc(inlen); if (!create_in) return -ENOMEM; rmpc = MLX5_ADDR_OF(create_rmp_in, create_in, ctx); + wq = MLX5_ADDR_OF(rmpc, rmpc, wq); + MLX5_SET(rmpc, rmpc, state, MLX5_RMPC_STATE_RDY); + set_wq(wq, in); memcpy(MLX5_ADDR_OF(rmpc, rmpc, wq.pas), in->pas, pas_size); - rmpc_srqc_reformat(srqc, rmpc, true); err = mlx5_core_create_rmp(dev, create_in, inlen, &srq->srqn); @@ -390,11 +418,10 @@ static int arm_rmp_cmd(struct mlx5_core_dev *dev, } static int query_rmp_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_query_srq_mbox_out *out) + struct mlx5_srq_attr *out) { u32 *rmp_out; void *rmpc; - void *srqc; int err; rmp_out = mlx5_vzalloc(MLX5_ST_SZ_BYTES(query_rmp_out)); @@ -405,9 +432,10 @@ static int query_rmp_cmd(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, if (err) goto out; - srqc = MLX5_ADDR_OF(query_srq_out, out, srq_context_entry); rmpc = MLX5_ADDR_OF(query_rmp_out, rmp_out, rmp_context); - rmpc_srqc_reformat(srqc, rmpc, false); + get_wq(MLX5_ADDR_OF(rmpc, rmpc, wq), out); + if (MLX5_GET(rmpc, rmpc, state) != MLX5_RMPC_STATE_RDY) + out->flags |= MLX5_SRQ_FLAG_ERR; out: kvfree(rmp_out); @@ -416,15 +444,14 @@ out: static int create_srq_split(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_create_srq_mbox_in *in, - int inlen, int is_xrc) + struct mlx5_srq_attr *in) { if (!dev->issi) - return create_srq_cmd(dev, srq, in, inlen); + return create_srq_cmd(dev, srq, in); else if (srq->common.res == MLX5_RES_XSRQ) - return create_xrc_srq_cmd(dev, srq, in, inlen); + return create_xrc_srq_cmd(dev, srq, in); else - return create_rmp_cmd(dev, srq, in, inlen); + return create_rmp_cmd(dev, srq, in); } static int destroy_srq_split(struct mlx5_core_dev *dev, @@ -439,15 +466,17 @@ static int destroy_srq_split(struct mlx5_core_dev *dev, } int mlx5_core_create_srq(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_create_srq_mbox_in *in, int inlen, - int is_xrc) + struct mlx5_srq_attr *in) { int err; struct mlx5_srq_table *table = &dev->priv.srq_table; - srq->common.res = is_xrc ? MLX5_RES_XSRQ : MLX5_RES_SRQ; + if (in->type == IB_SRQT_XRC) + srq->common.res = MLX5_RES_XSRQ; + else + srq->common.res = MLX5_RES_SRQ; - err = create_srq_split(dev, srq, in, inlen, is_xrc); + err = create_srq_split(dev, srq, in); if (err) return err; @@ -502,7 +531,7 @@ int mlx5_core_destroy_srq(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq) EXPORT_SYMBOL(mlx5_core_destroy_srq); int mlx5_core_query_srq(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_query_srq_mbox_out *out) + struct mlx5_srq_attr *out) { if (!dev->issi) return query_srq_cmd(dev, srq, out); diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index 80776d0c52dc..ba933335772c 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -46,6 +46,7 @@ #include #include +#include enum { MLX5_RQ_BITMASK_VSD = 1 << 1, @@ -772,11 +773,10 @@ struct mlx5_cmd_mailbox *mlx5_alloc_cmd_mailbox_chain(struct mlx5_core_dev *dev, void mlx5_free_cmd_mailbox_chain(struct mlx5_core_dev *dev, struct mlx5_cmd_mailbox *head); int mlx5_core_create_srq(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_create_srq_mbox_in *in, int inlen, - int is_xrc); + struct mlx5_srq_attr *in); int mlx5_core_destroy_srq(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq); int mlx5_core_query_srq(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, - struct mlx5_query_srq_mbox_out *out); + struct mlx5_srq_attr *out); int mlx5_core_arm_srq(struct mlx5_core_dev *dev, struct mlx5_core_srq *srq, u16 lwm, int is_srq); void mlx5_init_mkey_table(struct mlx5_core_dev *dev); diff --git a/include/linux/mlx5/srq.h b/include/linux/mlx5/srq.h index f43ed054a3e0..33c97dc900f8 100644 --- a/include/linux/mlx5/srq.h +++ b/include/linux/mlx5/srq.h @@ -35,6 +35,31 @@ #include +enum { + MLX5_SRQ_FLAG_ERR = (1 << 0), + MLX5_SRQ_FLAG_WQ_SIG = (1 << 1), +}; + +struct mlx5_srq_attr { + u32 type; + u32 flags; + u32 log_size; + u32 wqe_shift; + u32 log_page_size; + u32 wqe_cnt; + u32 srqn; + u32 xrcd; + u32 page_offset; + u32 cqn; + u32 pd; + u32 lwm; + u32 user_index; + u64 db_record; + u64 *pas; +}; + +struct mlx5_core_dev; + void mlx5_init_srq_table(struct mlx5_core_dev *dev); void mlx5_cleanup_srq_table(struct mlx5_core_dev *dev); -- cgit v1.2.3 From 0837e86a7a3422b85aa45c6f4631f6a3f74cbd01 Mon Sep 17 00:00:00 2001 From: Mark Bloch Date: Fri, 17 Jun 2016 15:10:55 +0300 Subject: IB/mlx5: Add per port counters In order to support statistics for ports, we attach each QP to a counter set which is dedicate to this port. Signed-off-by: Mark Bloch Reviewed-by: Leon Romanovsky Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 56 ++++++++++++++++++++++++++++++++++-- drivers/infiniband/hw/mlx5/mlx5_ib.h | 6 ++++ drivers/infiniband/hw/mlx5/qp.c | 9 ++++++ 3 files changed, 69 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index a7cc6d74d3b0..b29b84134c4c 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -2541,6 +2541,41 @@ static void mlx5_disable_roce(struct mlx5_ib_dev *dev) unregister_netdevice_notifier(&dev->roce.nb); } +static void mlx5_ib_dealloc_q_counters(struct mlx5_ib_dev *dev) +{ + unsigned int i; + + for (i = 0; i < dev->num_ports; i++) + mlx5_core_dealloc_q_counter(dev->mdev, + dev->port[i].q_cnt_id); +} + +static int mlx5_ib_alloc_q_counters(struct mlx5_ib_dev *dev) +{ + int i; + int ret; + + for (i = 0; i < dev->num_ports; i++) { + ret = mlx5_core_alloc_q_counter(dev->mdev, + &dev->port[i].q_cnt_id); + if (ret) { + mlx5_ib_warn(dev, + "couldn't allocate queue counter for port %d, err %d\n", + i + 1, ret); + goto dealloc_counters; + } + } + + return 0; + +dealloc_counters: + while (--i >= 0) + mlx5_core_dealloc_q_counter(dev->mdev, + dev->port[i].q_cnt_id); + + return ret; +} + static void *mlx5_ib_add(struct mlx5_core_dev *mdev) { struct mlx5_ib_dev *dev; @@ -2563,10 +2598,15 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->mdev = mdev; + dev->port = kcalloc(MLX5_CAP_GEN(mdev, num_ports), sizeof(*dev->port), + GFP_KERNEL); + if (!dev->port) + goto err_dealloc; + rwlock_init(&dev->roce.netdev_lock); err = get_port_caps(dev); if (err) - goto err_dealloc; + goto err_free_port; if (mlx5_use_mad_ifc(dev)) get_ext_port_caps(dev); @@ -2729,10 +2769,14 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) if (err) goto err_rsrc; - err = ib_register_device(&dev->ib_dev, NULL); + err = mlx5_ib_alloc_q_counters(dev); if (err) goto err_odp; + err = ib_register_device(&dev->ib_dev, NULL); + if (err) + goto err_q_cnt; + err = create_umr_res(dev); if (err) goto err_dev; @@ -2754,6 +2798,9 @@ err_umrc: err_dev: ib_unregister_device(&dev->ib_dev); +err_q_cnt: + mlx5_ib_dealloc_q_counters(dev); + err_odp: mlx5_ib_odp_remove_one(dev); @@ -2764,6 +2811,9 @@ err_disable_roce: if (ll == IB_LINK_LAYER_ETHERNET) mlx5_disable_roce(dev); +err_free_port: + kfree(dev->port); + err_dealloc: ib_dealloc_device((struct ib_device *)dev); @@ -2776,11 +2826,13 @@ static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context) enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev, 1); ib_unregister_device(&dev->ib_dev); + mlx5_ib_dealloc_q_counters(dev); destroy_umrc_res(dev); mlx5_ib_odp_remove_one(dev); destroy_dev_resources(&dev->devr); if (ll == IB_LINK_LAYER_ETHERNET) mlx5_disable_roce(dev); + kfree(dev->port); ib_dealloc_device(&dev->ib_dev); } diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 0001ed51e1bd..372385d0f993 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -591,6 +591,10 @@ struct mlx5_ib_resources { struct mutex mutex; }; +struct mlx5_ib_port { + u16 q_cnt_id; +}; + struct mlx5_roce { /* Protect mlx5_ib_get_netdev from invoking dev_hold() with a NULL * netdev pointer @@ -629,6 +633,8 @@ struct mlx5_ib_dev { /* protect resources needed as part of reset flow */ spinlock_t reset_flow_resource_lock; struct list_head qp_list; + /* Array with num_ports elements */ + struct mlx5_ib_port *port; }; static inline struct mlx5_ib_cq *to_mibcq(struct mlx5_core_cq *mcq) diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 9004905d3d93..5ca14a22868f 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -2651,6 +2651,15 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, else sqd_event = 0; + if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) { + u8 port_num = (attr_mask & IB_QP_PORT ? attr->port_num : + qp->port) - 1; + struct mlx5_ib_port *mibport = &dev->port[port_num]; + + context->qp_counter_set_usr_page |= + cpu_to_be32(mibport->q_cnt_id << 16); + } + if (!ibqp->uobject && cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) context->sq_crq_size |= cpu_to_be16(1 << 4); -- cgit v1.2.3 From 0ad17a8f7fa000cbfc51eedd7fddd20f7664e4b6 Mon Sep 17 00:00:00 2001 From: Mark Bloch Date: Fri, 17 Jun 2016 15:10:56 +0300 Subject: IB/mlx5: Add port protocol stats Expose new counters using the get protocol stats callback. We expose the following counters: |------------------------------------------------------------------------| | Name | IB | EN | Description | |------------------------------------------------------------------------| |rx_write_requests | + | - | Number of received WRITE requests for | | | | | the associated QP. | |------------------------------------------------------------------------| |rx_read_requests | + | - | Number of received READ requests for | | | | | the associated QP. | |------------------------------------------------------------------------| |rx_atomic_requests | + | - | Number of received ATOMIC requests for | | | | | the associated QP. | |------------------------------------------------------------------------| |out_of_buffer | + | + | Number of drops occurred due to lack | | | | | of WQE for the associated QPs/RQs. | |------------------------------------------------------------------------| |out_of_sequence | + | - | Number of errors in the packet | | | | | transport sequence number | |------------------------------------------------------------------------| |duplicate_request | + | + | Number of received duplicated packets. | | | | | A request that previously executed is | | | | | named duplicated. | |------------------------------------------------------------------------| |rnr_nak_retry_err | + | + | Number of received RNR NAC packets. | | | | | The QP retry limit did not exceed. | |------------------------------------------------------------------------| |packet_seq_err | + | + | Number of received NAK - sequence error| | | | | packets. The QP retry limit did not | | | | | exceed. | |------------------------------------------------------------------------| |implied_nak_err | + | + | Number of times the requester detected | | | | | an ACK with a PSN larger than expected | | | | | PSN for RDMA READ or ATOMIC response | | | | | The QP retry limit did not exceed. | |------------------------------------------------------------------------| |local_ack_timeout_err| + | - | Number of NO ACK responses from | | | | | responder within timer interval. | | | | | The QP retry limit did not exceed. | |------------------------------------------------------------------------| Counters are available if all of them are supported. Signed-off-by: Mark Bloch Signed-off-by: Leon Romanovsky Signed-off-by: Leon Romanovsky Reviewed-by: Christoph Lameter Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 78 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 78 insertions(+) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index b29b84134c4c..11e48665d953 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -2576,6 +2576,78 @@ dealloc_counters: return ret; } +static const char const *names[] = { + "rx_write_requests", + "rx_read_requests", + "rx_atomic_requests", + "out_of_buffer", + "out_of_sequence", + "duplicate_request", + "rnr_nak_retry_err", + "packet_seq_err", + "implied_nak_seq_err", + "local_ack_timeout_err", +}; + +static const size_t stats_offsets[] = { + MLX5_BYTE_OFF(query_q_counter_out, rx_write_requests), + MLX5_BYTE_OFF(query_q_counter_out, rx_read_requests), + MLX5_BYTE_OFF(query_q_counter_out, rx_atomic_requests), + MLX5_BYTE_OFF(query_q_counter_out, out_of_buffer), + MLX5_BYTE_OFF(query_q_counter_out, out_of_sequence), + MLX5_BYTE_OFF(query_q_counter_out, duplicate_request), + MLX5_BYTE_OFF(query_q_counter_out, rnr_nak_retry_err), + MLX5_BYTE_OFF(query_q_counter_out, packet_seq_err), + MLX5_BYTE_OFF(query_q_counter_out, implied_nak_seq_err), + MLX5_BYTE_OFF(query_q_counter_out, local_ack_timeout_err), +}; + +static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev, + u8 port_num) +{ + BUILD_BUG_ON(ARRAY_SIZE(names) != ARRAY_SIZE(stats_offsets)); + + /* We support only per port stats */ + if (port_num == 0) + return NULL; + + return rdma_alloc_hw_stats_struct(names, ARRAY_SIZE(names), + RDMA_HW_STATS_DEFAULT_LIFESPAN); +} + +static int mlx5_ib_get_hw_stats(struct ib_device *ibdev, + struct rdma_hw_stats *stats, + u8 port, int index) +{ + struct mlx5_ib_dev *dev = to_mdev(ibdev); + int outlen = MLX5_ST_SZ_BYTES(query_q_counter_out); + void *out; + __be32 val; + int ret; + int i; + + if (!port || !stats) + return -ENOSYS; + + out = mlx5_vzalloc(outlen); + if (!out) + return -ENOMEM; + + ret = mlx5_core_query_q_counter(dev->mdev, + dev->port[port - 1].q_cnt_id, 0, + out, outlen); + if (ret) + goto free; + + for (i = 0; i < ARRAY_SIZE(names); i++) { + val = *(__be32 *)(out + stats_offsets[i]); + stats->value[i] = (u64)be32_to_cpu(val); + } +free: + kvfree(out); + return ARRAY_SIZE(names); +} + static void *mlx5_ib_add(struct mlx5_core_dev *mdev) { struct mlx5_ib_dev *dev; @@ -2720,6 +2792,12 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) (1ull << IB_USER_VERBS_CMD_DEALLOC_MW); } + if (MLX5_CAP_GEN(dev->mdev, out_of_seq_cnt) && + MLX5_CAP_GEN(dev->mdev, retransmission_q_counters)) { + dev->ib_dev.get_hw_stats = mlx5_ib_get_hw_stats; + dev->ib_dev.alloc_hw_stats = mlx5_ib_alloc_hw_stats; + } + if (MLX5_CAP_GEN(mdev, xrc)) { dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd; dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd; -- cgit v1.2.3 From 5fa76c20458518ed6181adddef2e31c5afc0745c Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:21:56 -0400 Subject: IB/core: Add get FW version string to the core Allow for a common core function to get firmware version strings from the individual devices. In later patches this format can then then be used to pass a properly formated version string through the IPoIB layer. The problem with the current code in the IPoIB layer is that it is specific to certain hardware types. Furthermore, this gives us a common function through which the core can provide a common sysfs entry. Eventually we may want to remove the sysfs export but this provides for user space backwards compatibility. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/core/device.c | 9 +++++++++ include/rdma/ib_verbs.h | 3 +++ 2 files changed, 12 insertions(+) diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 5c155fa91eec..760ef603a468 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -311,6 +311,15 @@ static int read_port_immutable(struct ib_device *device) return 0; } +void ib_get_device_fw_str(struct ib_device *dev, char *str, size_t str_len) +{ + if (dev->get_dev_fw_str) + dev->get_dev_fw_str(dev, str, str_len); + else + str[0] = '\0'; +} +EXPORT_SYMBOL(ib_get_device_fw_str); + /** * ib_register_device - Register an IB device with IB core * @device:Device to register diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 7e440d41487a..1dc3d0d90202 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1956,6 +1956,7 @@ struct ib_device { * in fast paths. */ int (*get_port_immutable)(struct ib_device *, u8, struct ib_port_immutable *); + void (*get_dev_fw_str)(struct ib_device *, char *str, size_t str_len); }; struct ib_client { @@ -1991,6 +1992,8 @@ struct ib_client { struct ib_device *ib_alloc_device(size_t size); void ib_dealloc_device(struct ib_device *device); +void ib_get_device_fw_str(struct ib_device *device, char *str, size_t str_len); + int ib_register_device(struct ib_device *device, int (*port_callback)(struct ib_device *, u8, struct kobject *)); -- cgit v1.2.3 From e18036942444f5c6e4befd08e2794a50a93a6e34 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:21:57 -0400 Subject: IB/cxgb3: Support device FW version string Also remove fw_ver sysfs to be replaced by the common core one. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Reviewed-by: Steve Wise Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb3/iwch_provider.c | 27 +++++++++++++-------------- 1 file changed, 13 insertions(+), 14 deletions(-) diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c index bb1a839d4d6d..3edb80644b53 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_provider.c +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c @@ -1183,18 +1183,6 @@ static ssize_t show_rev(struct device *dev, struct device_attribute *attr, return sprintf(buf, "%d\n", iwch_dev->rdev.t3cdev_p->type); } -static ssize_t show_fw_ver(struct device *dev, struct device_attribute *attr, char *buf) -{ - struct iwch_dev *iwch_dev = container_of(dev, struct iwch_dev, - ibdev.dev); - struct ethtool_drvinfo info; - struct net_device *lldev = iwch_dev->rdev.t3cdev_p->lldev; - - PDBG("%s dev 0x%p\n", __func__, dev); - lldev->ethtool_ops->get_drvinfo(lldev, &info); - return sprintf(buf, "%s\n", info.fw_version); -} - static ssize_t show_hca(struct device *dev, struct device_attribute *attr, char *buf) { @@ -1334,13 +1322,11 @@ static int iwch_get_mib(struct ib_device *ibdev, struct rdma_hw_stats *stats, } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static struct device_attribute *iwch_class_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type, &dev_attr_board_id, }; @@ -1362,6 +1348,18 @@ static int iwch_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void get_dev_fw_ver_str(struct ib_device *ibdev, char *str, + size_t str_len) +{ + struct iwch_dev *iwch_dev = to_iwch_dev(ibdev); + struct ethtool_drvinfo info; + struct net_device *lldev = iwch_dev->rdev.t3cdev_p->lldev; + + PDBG("%s dev 0x%p\n", __func__, iwch_dev); + lldev->ethtool_ops->get_drvinfo(lldev, &info); + snprintf(str, str_len, "%s", info.fw_version); +} + int iwch_register_device(struct iwch_dev *dev) { int ret; @@ -1437,6 +1435,7 @@ int iwch_register_device(struct iwch_dev *dev) dev->ibdev.get_hw_stats = iwch_get_mib; dev->ibdev.uverbs_abi_ver = IWCH_UVERBS_ABI_VERSION; dev->ibdev.get_port_immutable = iwch_port_immutable; + dev->ibdev.get_dev_fw_str = get_dev_fw_ver_str; dev->ibdev.iwcm = kmalloc(sizeof(struct iw_cm_verbs), GFP_KERNEL); if (!dev->ibdev.iwcm) -- cgit v1.2.3 From ce1922435de3763e6097ffaa21f5f85d091818fd Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:21:58 -0400 Subject: IB/cxgb4: Support device FW version string And remove sysfs fw_ver in favor of the core. Reviewed-by: Steve Wise Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/cxgb4/provider.c | 31 +++++++++++++++---------------- 1 file changed, 15 insertions(+), 16 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c index dd8a86b726d2..df127ce6b6ec 100644 --- a/drivers/infiniband/hw/cxgb4/provider.c +++ b/drivers/infiniband/hw/cxgb4/provider.c @@ -409,20 +409,6 @@ static ssize_t show_rev(struct device *dev, struct device_attribute *attr, CHELSIO_CHIP_RELEASE(c4iw_dev->rdev.lldi.adapter_type)); } -static ssize_t show_fw_ver(struct device *dev, struct device_attribute *attr, - char *buf) -{ - struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev, - ibdev.dev); - PDBG("%s dev 0x%p\n", __func__, dev); - - return sprintf(buf, "%u.%u.%u.%u\n", - FW_HDR_FW_VER_MAJOR_G(c4iw_dev->rdev.lldi.fw_vers), - FW_HDR_FW_VER_MINOR_G(c4iw_dev->rdev.lldi.fw_vers), - FW_HDR_FW_VER_MICRO_G(c4iw_dev->rdev.lldi.fw_vers), - FW_HDR_FW_VER_BUILD_G(c4iw_dev->rdev.lldi.fw_vers)); -} - static ssize_t show_hca(struct device *dev, struct device_attribute *attr, char *buf) { @@ -502,13 +488,11 @@ static int c4iw_get_mib(struct ib_device *ibdev, } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static struct device_attribute *c4iw_class_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type, &dev_attr_board_id, }; @@ -530,6 +514,20 @@ static int c4iw_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void get_dev_fw_str(struct ib_device *dev, char *str, + size_t str_len) +{ + struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev, + ibdev); + PDBG("%s dev 0x%p\n", __func__, dev); + + snprintf(str, str_len, "%u.%u.%u.%u", + FW_HDR_FW_VER_MAJOR_G(c4iw_dev->rdev.lldi.fw_vers), + FW_HDR_FW_VER_MINOR_G(c4iw_dev->rdev.lldi.fw_vers), + FW_HDR_FW_VER_MICRO_G(c4iw_dev->rdev.lldi.fw_vers), + FW_HDR_FW_VER_BUILD_G(c4iw_dev->rdev.lldi.fw_vers)); +} + int c4iw_register_device(struct c4iw_dev *dev) { int ret; @@ -605,6 +603,7 @@ int c4iw_register_device(struct c4iw_dev *dev) dev->ibdev.get_hw_stats = c4iw_get_mib; dev->ibdev.uverbs_abi_ver = C4IW_UVERBS_ABI_VERSION; dev->ibdev.get_port_immutable = c4iw_port_immutable; + dev->ibdev.get_dev_fw_str = get_dev_fw_str; dev->ibdev.drain_sq = c4iw_drain_sq; dev->ibdev.drain_rq = c4iw_drain_rq; -- cgit v1.2.3 From f65c52ca23550ab44b08f243a9a4067a911e9fd3 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:21:59 -0400 Subject: IB/i40iw: Support device FW version string And remove sysfs support in favor of the core version. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/i40iw/i40iw_verbs.c | 24 ++++++++++-------------- 1 file changed, 10 insertions(+), 14 deletions(-) diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c index 02a735b64208..39096a246863 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c @@ -1972,18 +1972,6 @@ static ssize_t i40iw_show_rev(struct device *dev, return sprintf(buf, "%x\n", hw_rev); } -/** - * i40iw_show_fw_ver - */ -static ssize_t i40iw_show_fw_ver(struct device *dev, - struct device_attribute *attr, char *buf) -{ - u32 firmware_version = I40IW_FW_VERSION; - - return sprintf(buf, "%u.%u\n", firmware_version, - (firmware_version & 0x000000ff)); -} - /** * i40iw_show_hca */ @@ -2004,13 +1992,11 @@ static ssize_t i40iw_show_board(struct device *dev, } static DEVICE_ATTR(hw_rev, S_IRUGO, i40iw_show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, i40iw_show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, i40iw_show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, i40iw_show_board, NULL); static struct device_attribute *i40iw_dev_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type, &dev_attr_board_id }; @@ -2427,6 +2413,15 @@ static const char * const i40iw_hw_stat_names[] = { "iwRdmaInv" }; +static void i40iw_get_dev_fw_str(struct ib_device *dev, char *str, + size_t str_len) +{ + u32 firmware_version = I40IW_FW_VERSION; + + snprintf(str, str_len, "%u.%u", firmware_version, + (firmware_version & 0x000000ff)); +} + /** * i40iw_alloc_hw_stats - Allocate a hw stats structure * @ibdev: device pointer from stack @@ -2650,6 +2645,7 @@ static struct i40iw_ib_device *i40iw_init_rdma_device(struct i40iw_device *iwdev memcpy(iwibdev->ibdev.iwcm->ifname, netdev->name, sizeof(iwibdev->ibdev.iwcm->ifname)); iwibdev->ibdev.get_port_immutable = i40iw_port_immutable; + iwibdev->ibdev.get_dev_fw_str = i40iw_get_dev_fw_str; iwibdev->ibdev.poll_cq = i40iw_poll_cq; iwibdev->ibdev.req_notify_cq = i40iw_req_notify_cq; iwibdev->ibdev.post_send = i40iw_post_send; -- cgit v1.2.3 From e9db59fcd28ad6d7aceb74c2fb65a44499fdbf9a Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:00 -0400 Subject: IB/mlx4: Support device FW version string And remove the sysfs in favor of common core version. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx4/main.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index 0eb09e104542..1cbd075ce7e4 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -2022,16 +2022,6 @@ static ssize_t show_hca(struct device *device, struct device_attribute *attr, return sprintf(buf, "MT%d\n", dev->dev->persist->pdev->device); } -static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, - char *buf) -{ - struct mlx4_ib_dev *dev = - container_of(device, struct mlx4_ib_dev, ib_dev.dev); - return sprintf(buf, "%d.%d.%d\n", (int) (dev->dev->caps.fw_ver >> 32), - (int) (dev->dev->caps.fw_ver >> 16) & 0xffff, - (int) dev->dev->caps.fw_ver & 0xffff); -} - static ssize_t show_rev(struct device *device, struct device_attribute *attr, char *buf) { @@ -2050,13 +2040,11 @@ static ssize_t show_board(struct device *device, struct device_attribute *attr, } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static struct device_attribute *mlx4_class_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type, &dev_attr_board_id }; @@ -2277,6 +2265,17 @@ static int mlx4_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void get_fw_ver_str(struct ib_device *device, char *str, + size_t str_len) +{ + struct mlx4_ib_dev *dev = + container_of(device, struct mlx4_ib_dev, ib_dev); + snprintf(str, str_len, "%d.%d.%d", + (int) (dev->dev->caps.fw_ver >> 32), + (int) (dev->dev->caps.fw_ver >> 16) & 0xffff, + (int) dev->dev->caps.fw_ver & 0xffff); +} + static void *mlx4_ib_add(struct mlx4_dev *dev) { struct mlx4_ib_dev *ibdev; @@ -2410,6 +2409,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev) ibdev->ib_dev.detach_mcast = mlx4_ib_mcg_detach; ibdev->ib_dev.process_mad = mlx4_ib_process_mad; ibdev->ib_dev.get_port_immutable = mlx4_port_immutable; + ibdev->ib_dev.get_dev_fw_str = get_fw_ver_str; ibdev->ib_dev.disassociate_ucontext = mlx4_ib_disassociate_ucontext; if (!mlx4_is_slave(ibdev->dev)) { -- cgit v1.2.3 From c73428230d98d1352bcc69cd8306c292a85e1e42 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:01 -0400 Subject: IB/mlx5: Support device FW version string And remove sysfs entry in favor of the common code. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mlx5/main.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index b48ad85315dc..3fa8557eddbd 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -1804,15 +1804,6 @@ static ssize_t show_hca(struct device *device, struct device_attribute *attr, return sprintf(buf, "MT%d\n", dev->mdev->pdev->device); } -static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, - char *buf) -{ - struct mlx5_ib_dev *dev = - container_of(device, struct mlx5_ib_dev, ib_dev.dev); - return sprintf(buf, "%d.%d.%04d\n", fw_rev_maj(dev->mdev), - fw_rev_min(dev->mdev), fw_rev_sub(dev->mdev)); -} - static ssize_t show_rev(struct device *device, struct device_attribute *attr, char *buf) { @@ -1831,7 +1822,6 @@ static ssize_t show_board(struct device *device, struct device_attribute *attr, } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static DEVICE_ATTR(fw_pages, S_IRUGO, show_fw_pages, NULL); @@ -1839,7 +1829,6 @@ static DEVICE_ATTR(reg_pages, S_IRUGO, show_reg_pages, NULL); static struct device_attribute *mlx5_class_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type, &dev_attr_board_id, &dev_attr_fw_pages, @@ -2275,6 +2264,15 @@ static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void get_dev_fw_str(struct ib_device *ibdev, char *str, + size_t str_len) +{ + struct mlx5_ib_dev *dev = + container_of(ibdev, struct mlx5_ib_dev, ib_dev); + snprintf(str, str_len, "%d.%d.%04d", fw_rev_maj(dev->mdev), + fw_rev_min(dev->mdev), fw_rev_sub(dev->mdev)); +} + static int mlx5_enable_roce(struct mlx5_ib_dev *dev) { int err; @@ -2421,6 +2419,7 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->ib_dev.map_mr_sg = mlx5_ib_map_mr_sg; dev->ib_dev.check_mr_status = mlx5_ib_check_mr_status; dev->ib_dev.get_port_immutable = mlx5_port_immutable; + dev->ib_dev.get_dev_fw_str = get_dev_fw_str; if (mlx5_core_is_pf(mdev)) { dev->ib_dev.get_vf_config = mlx5_ib_get_vf_config; dev->ib_dev.set_vf_link_state = mlx5_ib_set_vf_link_state; -- cgit v1.2.3 From 51ed03978e2c14e6cbe5458cd94f46fed108fda3 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:02 -0400 Subject: IB/mthca: Supprot device FW version string And remove the sysfs entry in favor of the core support. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/mthca/mthca_provider.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c index 9866c35cc977..da2335f7f7c3 100644 --- a/drivers/infiniband/hw/mthca/mthca_provider.c +++ b/drivers/infiniband/hw/mthca/mthca_provider.c @@ -1081,16 +1081,6 @@ static ssize_t show_rev(struct device *device, struct device_attribute *attr, return sprintf(buf, "%x\n", dev->rev_id); } -static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, - char *buf) -{ - struct mthca_dev *dev = - container_of(device, struct mthca_dev, ib_dev.dev); - return sprintf(buf, "%d.%d.%d\n", (int) (dev->fw_ver >> 32), - (int) (dev->fw_ver >> 16) & 0xffff, - (int) dev->fw_ver & 0xffff); -} - static ssize_t show_hca(struct device *device, struct device_attribute *attr, char *buf) { @@ -1120,13 +1110,11 @@ static ssize_t show_board(struct device *device, struct device_attribute *attr, } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static struct device_attribute *mthca_dev_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type, &dev_attr_board_id }; @@ -1187,6 +1175,17 @@ static int mthca_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void get_dev_fw_str(struct ib_device *device, char *str, + size_t str_len) +{ + struct mthca_dev *dev = + container_of(device, struct mthca_dev, ib_dev); + snprintf(str, str_len, "%d.%d.%d", + (int) (dev->fw_ver >> 32), + (int) (dev->fw_ver >> 16) & 0xffff, + (int) dev->fw_ver & 0xffff); +} + int mthca_register_device(struct mthca_dev *dev) { int ret; @@ -1266,6 +1265,7 @@ int mthca_register_device(struct mthca_dev *dev) dev->ib_dev.reg_user_mr = mthca_reg_user_mr; dev->ib_dev.dereg_mr = mthca_dereg_mr; dev->ib_dev.get_port_immutable = mthca_port_immutable; + dev->ib_dev.get_dev_fw_str = get_dev_fw_str; if (dev->mthca_flags & MTHCA_FLAG_FMR) { dev->ib_dev.alloc_fmr = mthca_alloc_fmr; -- cgit v1.2.3 From 96357454eb8427732e8deea9f75a95e4c36a5e3b Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:03 -0400 Subject: IB/nes: Support device FW version string And remove the sysfs in favor of the core version. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/nes/nes_verbs.c | 33 ++++++++++++++------------------- 1 file changed, 14 insertions(+), 19 deletions(-) diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c index 464d6da5fe91..bd69125731c1 100644 --- a/drivers/infiniband/hw/nes/nes_verbs.c +++ b/drivers/infiniband/hw/nes/nes_verbs.c @@ -2605,23 +2605,6 @@ static ssize_t show_rev(struct device *dev, struct device_attribute *attr, } -/** - * show_fw_ver - */ -static ssize_t show_fw_ver(struct device *dev, struct device_attribute *attr, - char *buf) -{ - struct nes_ib_device *nesibdev = - container_of(dev, struct nes_ib_device, ibdev.dev); - struct nes_vnic *nesvnic = nesibdev->nesvnic; - - nes_debug(NES_DBG_INIT, "\n"); - return sprintf(buf, "%u.%u\n", - (nesvnic->nesdev->nesadapter->firmware_version >> 16), - (nesvnic->nesdev->nesadapter->firmware_version & 0x000000ff)); -} - - /** * show_hca */ @@ -2645,13 +2628,11 @@ static ssize_t show_board(struct device *dev, struct device_attribute *attr, static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static struct device_attribute *nes_dev_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type, &dev_attr_board_id }; @@ -3703,6 +3684,19 @@ static int nes_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void get_dev_fw_str(struct ib_device *dev, char *str, + size_t str_len) +{ + struct nes_ib_device *nesibdev = + container_of(dev, struct nes_ib_device, ibdev); + struct nes_vnic *nesvnic = nesibdev->nesvnic; + + nes_debug(NES_DBG_INIT, "\n"); + snprintf(str, str_len, "%u.%u", + (nesvnic->nesdev->nesadapter->firmware_version >> 16), + (nesvnic->nesdev->nesadapter->firmware_version & 0x000000ff)); +} + /** * nes_init_ofa_device */ @@ -3802,6 +3796,7 @@ struct nes_ib_device *nes_init_ofa_device(struct net_device *netdev) nesibdev->ibdev.iwcm->create_listen = nes_create_listen; nesibdev->ibdev.iwcm->destroy_listen = nes_destroy_listen; nesibdev->ibdev.get_port_immutable = nes_port_immutable; + nesibdev->ibdev.get_dev_fw_str = get_dev_fw_str; memcpy(nesibdev->ibdev.iwcm->ifname, netdev->name, sizeof(nesibdev->ibdev.iwcm->ifname)); -- cgit v1.2.3 From bd395005d2d9c2a4df59a861e359866b6c6f28f6 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:04 -0400 Subject: IB/ocrdma: Support device FW version string And remove sysfs in favor of the core support. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/ocrdma/ocrdma_main.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_main.c b/drivers/infiniband/hw/ocrdma/ocrdma_main.c index 3d75f65ce87e..07d0c6c5b046 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_main.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_main.c @@ -107,6 +107,14 @@ static int ocrdma_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void get_dev_fw_str(struct ib_device *device, char *str, + size_t str_len) +{ + struct ocrdma_dev *dev = get_ocrdma_dev(device); + + snprintf(str, str_len, "%s", &dev->attr.fw_ver[0]); +} + static int ocrdma_register_device(struct ocrdma_dev *dev) { strlcpy(dev->ibdev.name, "ocrdma%d", IB_DEVICE_NAME_MAX); @@ -193,6 +201,7 @@ static int ocrdma_register_device(struct ocrdma_dev *dev) dev->ibdev.process_mad = ocrdma_process_mad; dev->ibdev.get_port_immutable = ocrdma_port_immutable; + dev->ibdev.get_dev_fw_str = get_dev_fw_str; if (ocrdma_get_asic_type(dev) == OCRDMA_ASIC_GEN_SKH_R) { dev->ibdev.uverbs_cmd_mask |= @@ -262,14 +271,6 @@ static ssize_t show_rev(struct device *device, struct device_attribute *attr, return scnprintf(buf, PAGE_SIZE, "0x%x\n", dev->nic_info.pdev->vendor); } -static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, - char *buf) -{ - struct ocrdma_dev *dev = dev_get_drvdata(device); - - return scnprintf(buf, PAGE_SIZE, "%s\n", &dev->attr.fw_ver[0]); -} - static ssize_t show_hca_type(struct device *device, struct device_attribute *attr, char *buf) { @@ -279,12 +280,10 @@ static ssize_t show_hca_type(struct device *device, } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); -static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca_type, NULL); static struct device_attribute *ocrdma_attributes[] = { &dev_attr_hw_rev, - &dev_attr_fw_ver, &dev_attr_hca_type }; -- cgit v1.2.3 From 15453e857a50b3fffa78151e2c84f325b776c0e0 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:05 -0400 Subject: IB/usnic: Support device FW version string And remove sysfs file in favor of the common core. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/usnic/usnic_ib_main.c | 16 ++++++++++++++++ drivers/infiniband/hw/usnic/usnic_ib_sysfs.c | 17 ----------------- 2 files changed, 16 insertions(+), 17 deletions(-) diff --git a/drivers/infiniband/hw/usnic/usnic_ib_main.c b/drivers/infiniband/hw/usnic/usnic_ib_main.c index 565c881a44ba..c229b9f4a52d 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_main.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_main.c @@ -331,6 +331,21 @@ static int usnic_port_immutable(struct ib_device *ibdev, u8 port_num, return 0; } +static void usnic_get_dev_fw_str(struct ib_device *device, + char *str, + size_t str_len) +{ + struct usnic_ib_dev *us_ibdev = + container_of(device, struct usnic_ib_dev, ib_dev); + struct ethtool_drvinfo info; + + mutex_lock(&us_ibdev->usdev_lock); + us_ibdev->netdev->ethtool_ops->get_drvinfo(us_ibdev->netdev, &info); + mutex_unlock(&us_ibdev->usdev_lock); + + snprintf(str, str_len, "%s", info.fw_version); +} + /* Start of PF discovery section */ static void *usnic_ib_device_add(struct pci_dev *dev) { @@ -414,6 +429,7 @@ static void *usnic_ib_device_add(struct pci_dev *dev) us_ibdev->ib_dev.req_notify_cq = usnic_ib_req_notify_cq; us_ibdev->ib_dev.get_dma_mr = usnic_ib_get_dma_mr; us_ibdev->ib_dev.get_port_immutable = usnic_port_immutable; + us_ibdev->ib_dev.get_dev_fw_str = usnic_get_dev_fw_str; if (ib_register_device(&us_ibdev->ib_dev, NULL)) diff --git a/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c b/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c index 3412ea06116e..80ef3f8998c8 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c @@ -45,21 +45,6 @@ #include "usnic_ib_verbs.h" #include "usnic_log.h" -static ssize_t usnic_ib_show_fw_ver(struct device *device, - struct device_attribute *attr, - char *buf) -{ - struct usnic_ib_dev *us_ibdev = - container_of(device, struct usnic_ib_dev, ib_dev.dev); - struct ethtool_drvinfo info; - - mutex_lock(&us_ibdev->usdev_lock); - us_ibdev->netdev->ethtool_ops->get_drvinfo(us_ibdev->netdev, &info); - mutex_unlock(&us_ibdev->usdev_lock); - - return scnprintf(buf, PAGE_SIZE, "%s\n", info.fw_version); -} - static ssize_t usnic_ib_show_board(struct device *device, struct device_attribute *attr, char *buf) @@ -192,7 +177,6 @@ usnic_ib_show_cq_per_vf(struct device *device, struct device_attribute *attr, us_ibdev->vf_res_cnt[USNIC_VNIC_RES_TYPE_CQ]); } -static DEVICE_ATTR(fw_ver, S_IRUGO, usnic_ib_show_fw_ver, NULL); static DEVICE_ATTR(board_id, S_IRUGO, usnic_ib_show_board, NULL); static DEVICE_ATTR(config, S_IRUGO, usnic_ib_show_config, NULL); static DEVICE_ATTR(iface, S_IRUGO, usnic_ib_show_iface, NULL); @@ -201,7 +185,6 @@ static DEVICE_ATTR(qp_per_vf, S_IRUGO, usnic_ib_show_qp_per_vf, NULL); static DEVICE_ATTR(cq_per_vf, S_IRUGO, usnic_ib_show_cq_per_vf, NULL); static struct device_attribute *usnic_class_attributes[] = { - &dev_attr_fw_ver, &dev_attr_board_id, &dev_attr_config, &dev_attr_iface, -- cgit v1.2.3 From 1a8632121a5e5bdebef00e59bd13951355682bab Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:06 -0400 Subject: IB/ipoib: Use new device FW version string Using this allows for devices to specify the format of their firmware version rather than forcing a format. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/ulp/ipoib/ipoib_ethtool.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c b/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c index 1502199c8e56..7b6d40ff1acf 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c @@ -62,10 +62,8 @@ static void ipoib_get_drvinfo(struct net_device *netdev, { struct ipoib_dev_priv *priv = netdev_priv(netdev); - snprintf(drvinfo->fw_version, sizeof(drvinfo->fw_version), - "%d.%d.%d", (int)(priv->ca->attrs.fw_ver >> 32), - (int)(priv->ca->attrs.fw_ver >> 16) & 0xffff, - (int)priv->ca->attrs.fw_ver & 0xffff); + ib_get_device_fw_str(priv->ca, drvinfo->fw_version, + sizeof(drvinfo->fw_version)); strlcpy(drvinfo->bus_info, dev_name(priv->ca->dma_device), sizeof(drvinfo->bus_info)); -- cgit v1.2.3 From 41a6ae1ebd51d074a43d608b8ecfc9dd2b323d5e Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:07 -0400 Subject: IB/core: Export a common fw_ver sysfs entry Now that all the devices have stopped exporting their own sysfs entry points we can have the core export this on their behalf. Eventually this may be removed but this provides for backwards compatibility. Reviewed-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/core/sysfs.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c index a5793c8f1590..0d1ab73f8186 100644 --- a/drivers/infiniband/core/sysfs.c +++ b/drivers/infiniband/core/sysfs.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include @@ -1196,16 +1197,28 @@ static ssize_t set_node_desc(struct device *device, return count; } +static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, + char *buf) +{ + struct ib_device *dev = container_of(device, struct ib_device, dev); + + ib_get_device_fw_str(dev, buf, PAGE_SIZE); + strlcat(buf, "\n", PAGE_SIZE); + return strlen(buf); +} + static DEVICE_ATTR(node_type, S_IRUGO, show_node_type, NULL); static DEVICE_ATTR(sys_image_guid, S_IRUGO, show_sys_image_guid, NULL); static DEVICE_ATTR(node_guid, S_IRUGO, show_node_guid, NULL); static DEVICE_ATTR(node_desc, S_IRUGO | S_IWUSR, show_node_desc, set_node_desc); +static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL); static struct device_attribute *ib_class_attributes[] = { &dev_attr_node_type, &dev_attr_sys_image_guid, &dev_attr_node_guid, - &dev_attr_node_desc + &dev_attr_node_desc, + &dev_attr_fw_ver, }; static void free_port_list_attributes(struct ib_device *device) -- cgit v1.2.3 From 939b6ca873e7bbef4b26a1cffacf1a1bf2f3f362 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Wed, 15 Jun 2016 02:22:08 -0400 Subject: IB/hfi1: Add device FW version string Export the firmware version through the core. Acked-by: Dennis Dalessandro Signed-off-by: Ira Weiny Signed-off-by: Doug Ledford --- drivers/infiniband/hw/hfi1/hfi.h | 2 ++ drivers/infiniband/hw/hfi1/verbs.c | 15 +++++++++++++++ 2 files changed, 17 insertions(+) diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index 4417a0fd3ef9..49a71e24a8f0 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1174,6 +1174,8 @@ struct hfi1_devdata { /* 8051 firmware version helper */ #define dc8051_ver(a, b) ((a) << 8 | (b)) +#define dc8051_ver_maj(a) ((a & 0xff00) >> 8) +#define dc8051_ver_min(a) (a & 0x00ff) /* f_put_tid types */ #define PT_EXPECTED 0 diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index 849c4b9399d4..dd4be3c2b225 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -1291,9 +1291,12 @@ int hfi1_verbs_send(struct rvt_qp *qp, struct hfi1_pkt_state *ps) static void hfi1_fill_device_attr(struct hfi1_devdata *dd) { struct rvt_dev_info *rdi = &dd->verbs_dev.rdi; + u16 ver = dd->dc8051_ver; memset(&rdi->dparms.props, 0, sizeof(rdi->dparms.props)); + rdi->dparms.props.fw_ver = ((u64)(dc8051_ver_maj(ver)) << 16) | + (u64)dc8051_ver_min(ver); rdi->dparms.props.device_cap_flags = IB_DEVICE_BAD_PKEY_CNTR | IB_DEVICE_BAD_QKEY_CNTR | IB_DEVICE_SHUTDOWN_PORT | IB_DEVICE_SYS_IMAGE_GUID | IB_DEVICE_RC_RNR_NAK_GEN | @@ -1567,6 +1570,17 @@ static void init_ibport(struct hfi1_pportdata *ppd) RCU_INIT_POINTER(ibp->rvp.qp[1], NULL); } +static void hfi1_get_dev_fw_str(struct ib_device *ibdev, char *str, + size_t str_len) +{ + struct rvt_dev_info *rdi = ib_to_rvt(ibdev); + struct hfi1_ibdev *dev = dev_from_rdi(rdi); + u16 ver = dd_from_dev(dev)->dc8051_ver; + + snprintf(str, str_len, "%u.%u", dc8051_ver_maj(ver), + dc8051_ver_min(ver)); +} + /** * hfi1_register_ib_device - register our device with the infiniband core * @dd: the device data structure @@ -1613,6 +1627,7 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) /* keep process mad in the driver */ ibdev->process_mad = hfi1_process_mad; + ibdev->get_dev_fw_str = hfi1_get_dev_fw_str; strncpy(ibdev->node_desc, init_utsname()->nodename, sizeof(ibdev->node_desc)); -- cgit v1.2.3 From 5ee98ab3a8ea94fe24e288dc02c6b489889cf06a Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Wed, 1 Jun 2016 16:46:22 +0200 Subject: mn10300: use RTC_DRV_CMOS instead of CONFIG_RTC nn10300 has a dependency on mc146818_get_time/mc146818_set_time, which we want to move from the mc146818rtc.h header into the rtc subsystem, which in turn is not usable on mn10300. This changes mn10300 to use the modern rtc-cmos driver instead of the old RTC driver, and that in turn lets us completely remove the read_persistent_clock/update_persistent_clock callbacks. Signed-off-by: Arnd Bergmann Signed-off-by: Alexandre Belloni --- arch/mn10300/Kconfig | 4 +- arch/mn10300/include/asm/rtc-regs.h | 4 +- arch/mn10300/kernel/rtc.c | 104 +++----------------------------- arch/mn10300/proc-mn103e010/proc-init.c | 1 + drivers/char/Kconfig | 2 +- drivers/rtc/Kconfig | 2 +- drivers/rtc/rtc-cmos.c | 2 +- 7 files changed, 18 insertions(+), 101 deletions(-) diff --git a/arch/mn10300/Kconfig b/arch/mn10300/Kconfig index 9627e81a6cbb..38e3494bfb63 100644 --- a/arch/mn10300/Kconfig +++ b/arch/mn10300/Kconfig @@ -236,7 +236,9 @@ source "kernel/Kconfig.hz" config MN10300_RTC bool "Using MN10300 RTC" depends on MN10300_PROC_MN103E010 || MN10300_PROC_MN2WS0050 - select GENERIC_CMOS_UPDATE + select RTC_CLASS + select RTC_DRV_CMOS + select RTC_SYSTOHC default n help This option enables support for the RTC, thus enabling time to be diff --git a/arch/mn10300/include/asm/rtc-regs.h b/arch/mn10300/include/asm/rtc-regs.h index c42deefaec11..c81cacecb6e3 100644 --- a/arch/mn10300/include/asm/rtc-regs.h +++ b/arch/mn10300/include/asm/rtc-regs.h @@ -75,9 +75,9 @@ #define RTC_PORT(x) 0xd8600000 #define RTC_ALWAYS_BCD 1 /* RTC operates in binary mode */ -#define CMOS_READ(addr) __SYSREG(0xd8600000 + (addr), u8) +#define CMOS_READ(addr) __SYSREG(0xd8600000 + (u32)(addr), u8) #define CMOS_WRITE(val, addr) \ - do { __SYSREG(0xd8600000 + (addr), u8) = val; } while (0) + do { __SYSREG(0xd8600000 + (u32)(addr), u8) = val; } while (0) #define RTC_IRQ RTIRQ diff --git a/arch/mn10300/kernel/rtc.c b/arch/mn10300/kernel/rtc.c index 77e0be4d92ea..f81f37025072 100644 --- a/arch/mn10300/kernel/rtc.c +++ b/arch/mn10300/kernel/rtc.c @@ -12,107 +12,19 @@ #include #include #include -#include -#include +#include +#include + #include #include DEFINE_SPINLOCK(rtc_lock); EXPORT_SYMBOL(rtc_lock); -/* - * Read the current RTC time - */ -void read_persistent_clock(struct timespec *ts) -{ - struct rtc_time tm; - - mc146818_set_time(&tm); - - ts->tv_nsec = 0; - ts->tv_sec = mktime(tm.tm_year, tm.tm_mon, tm.tm_mday, - tm.tm_hour, tm.tm_min, tm.tm_sec); - - /* if rtc is way off in the past, set something reasonable */ - if (ts->tv_sec < 0) - ts->tv_sec = mktime(2009, 1, 1, 12, 0, 0); -} - -/* - * In order to set the CMOS clock precisely, set_rtc_mmss has to be called 500 - * ms after the second nowtime has started, because when nowtime is written - * into the registers of the CMOS clock, it will jump to the next second - * precisely 500 ms later. Check the Motorola MC146818A or Dallas DS12887 data - * sheet for details. - * - * BUG: This routine does not handle hour overflow properly; it just - * sets the minutes. Usually you'll only notice that after reboot! - */ -static int set_rtc_mmss(unsigned long nowtime) -{ - unsigned char save_control, save_freq_select; - int retval = 0; - int real_seconds, real_minutes, cmos_minutes; - - /* gets recalled with irq locally disabled */ - spin_lock(&rtc_lock); - save_control = CMOS_READ(RTC_CONTROL); /* tell the clock it's being - * set */ - CMOS_WRITE(save_control | RTC_SET, RTC_CONTROL); - - save_freq_select = CMOS_READ(RTC_FREQ_SELECT); /* stop and reset - * prescaler */ - CMOS_WRITE(save_freq_select | RTC_DIV_RESET2, RTC_FREQ_SELECT); - - cmos_minutes = CMOS_READ(RTC_MINUTES); - if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD) - cmos_minutes = bcd2bin(cmos_minutes); - - /* - * since we're only adjusting minutes and seconds, - * don't interfere with hour overflow. This avoids - * messing with unknown time zones but requires your - * RTC not to be off by more than 15 minutes - */ - real_seconds = nowtime % 60; - real_minutes = nowtime / 60; - if (((abs(real_minutes - cmos_minutes) + 15) / 30) & 1) - /* correct for half hour time zone */ - real_minutes += 30; - real_minutes %= 60; - - if (abs(real_minutes - cmos_minutes) < 30) { - if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD) { - real_seconds = bin2bcd(real_seconds); - real_minutes = bin2bcd(real_minutes); - } - CMOS_WRITE(real_seconds, RTC_SECONDS); - CMOS_WRITE(real_minutes, RTC_MINUTES); - } else { - printk_once(KERN_NOTICE - "set_rtc_mmss: can't update from %d to %d\n", - cmos_minutes, real_minutes); - retval = -1; - } - - /* The following flags have to be released exactly in this order, - * otherwise the DS12887 (popular MC146818A clone with integrated - * battery and quartz) will not reset the oscillator and will not - * update precisely 500 ms later. You won't find this mentioned in - * the Dallas Semiconductor data sheets, but who believes data - * sheets anyway ... -- Markus Kuhn - */ - CMOS_WRITE(save_control, RTC_CONTROL); - CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); - spin_unlock(&rtc_lock); - - return retval; -} - -int update_persistent_clock(struct timespec now) -{ - return set_rtc_mmss(now.tv_sec); -} +static const __initdata struct resource res[] = { + DEFINE_RES_IO(RTC_PORT(0), RTC_IO_EXTENT), + DEFINE_RES_IRQ(RTC_IRQ), +}; /* * calibrate the TSC clock against the RTC @@ -129,4 +41,6 @@ void __init calibrate_clock(void) RTCRA |= RTCRA_DVR; RTCRA &= ~RTCRA_DVR; RTCRB &= ~RTCRB_SET; + + platform_device_register_simple("rtc_cmos", -1, res, ARRAY_SIZE(res)); } diff --git a/arch/mn10300/proc-mn103e010/proc-init.c b/arch/mn10300/proc-mn103e010/proc-init.c index 94641c420b3f..102d86a6ae56 100644 --- a/arch/mn10300/proc-mn103e010/proc-init.c +++ b/arch/mn10300/proc-mn103e010/proc-init.c @@ -9,6 +9,7 @@ * 2 of the Licence, or (at your option) any later version. */ #include +#include #include #include #include diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index 0ec91cc6c2fb..273b5492fbcf 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -279,7 +279,7 @@ if RTC_LIB=n config RTC tristate "Enhanced Real Time Clock Support (legacy PC RTC driver)" - depends on ALPHA || (MIPS && MACH_LOONGSON64) || MN10300 + depends on ALPHA || (MIPS && MACH_LOONGSON64) ---help--- If you say Y here and create a character special file /dev/rtc with major number 10 and minor number 135 using mknod ("man mknod"), you diff --git a/drivers/rtc/Kconfig b/drivers/rtc/Kconfig index 9f8b59d94b1d..13128a89657b 100644 --- a/drivers/rtc/Kconfig +++ b/drivers/rtc/Kconfig @@ -807,7 +807,7 @@ comment "Platform RTC drivers" config RTC_DRV_CMOS tristate "PC-style 'CMOS'" - depends on X86 || ARM || M32R || PPC || MIPS || SPARC64 + depends on X86 || ARM || M32R || PPC || MIPS || SPARC64 || MN10300 default y if X86 help Say "yes" here to get direct support for the real time clock diff --git a/drivers/rtc/rtc-cmos.c b/drivers/rtc/rtc-cmos.c index cf8eb98382ce..f9fdb6fc641c 100644 --- a/drivers/rtc/rtc-cmos.c +++ b/drivers/rtc/rtc-cmos.c @@ -630,7 +630,7 @@ cmos_do_probe(struct device *dev, struct resource *ports, int rtc_irq) address_space = 64; #elif defined(__i386__) || defined(__x86_64__) || defined(__arm__) \ || defined(__sparc__) || defined(__mips__) \ - || defined(__powerpc__) + || defined(__powerpc__) || defined(CONFIG_MN10300) address_space = 128; #else #warning Assuming 128 bytes of RTC+NVRAM address space, not 64 bytes. -- cgit v1.2.3 From d6faca40f40b62aca8ea8c29289c7bf7456172bb Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Wed, 1 Jun 2016 16:46:23 +0200 Subject: rtc: move mc146818 helper functions out-of-line The mc146818_get_time/mc146818_set_time functions are rather large inline functions in a global header file and are used in several drivers and in x86 specific code. Here we move them into a separate .c file that is compiled whenever any of the users require it. This also lets us remove the linux/acpi.h header inclusion from mc146818rtc.h, which in turn avoids some warnings about duplicate definition of the TRUE/FALSE macros. Signed-off-by: Arnd Bergmann Signed-off-by: Alexandre Belloni --- arch/x86/Kconfig | 1 + drivers/rtc/Kconfig | 6 ++ drivers/rtc/Makefile | 1 + drivers/rtc/rtc-mc146818-lib.c | 198 +++++++++++++++++++++++++++++++++++++++++ include/linux/mc146818rtc.h | 193 +-------------------------------------- 5 files changed, 208 insertions(+), 191 deletions(-) create mode 100644 drivers/rtc/rtc-mc146818-lib.c diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 0a7b885964ba..54d46c5c04d0 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -151,6 +151,7 @@ config X86 select OLD_SIGSUSPEND3 if X86_32 || IA32_EMULATION select PERF_EVENTS select RTC_LIB + select RTC_MC146818_LIB select SPARSE_IRQ select SRCU select SYSCTL_EXCEPTION_TRACE diff --git a/drivers/rtc/Kconfig b/drivers/rtc/Kconfig index 13128a89657b..8526f1cded08 100644 --- a/drivers/rtc/Kconfig +++ b/drivers/rtc/Kconfig @@ -5,6 +5,10 @@ config RTC_LIB bool +config RTC_MC146818_LIB + bool + select RTC_LIB + menuconfig RTC_CLASS bool "Real Time Clock" default n @@ -809,6 +813,7 @@ config RTC_DRV_CMOS tristate "PC-style 'CMOS'" depends on X86 || ARM || M32R || PPC || MIPS || SPARC64 || MN10300 default y if X86 + select RTC_MC146818_LIB help Say "yes" here to get direct support for the real time clock found in every PC or ACPI-based system, and some other boards. @@ -827,6 +832,7 @@ config RTC_DRV_CMOS config RTC_DRV_ALPHA bool "Alpha PC-style CMOS" depends on ALPHA + select RTC_MC146818_LIB default y help Direct support for the real-time clock found on every Alpha diff --git a/drivers/rtc/Makefile b/drivers/rtc/Makefile index 024da2723d86..7cf7ad559c79 100644 --- a/drivers/rtc/Makefile +++ b/drivers/rtc/Makefile @@ -8,6 +8,7 @@ obj-$(CONFIG_RTC_LIB) += rtc-lib.o obj-$(CONFIG_RTC_HCTOSYS) += hctosys.o obj-$(CONFIG_RTC_SYSTOHC) += systohc.o obj-$(CONFIG_RTC_CLASS) += rtc-core.o +obj-$(CONFIG_RTC_MC146818_LIB) += rtc-mc146818-lib.o rtc-core-y := class.o interface.o ifdef CONFIG_RTC_DRV_EFI diff --git a/drivers/rtc/rtc-mc146818-lib.c b/drivers/rtc/rtc-mc146818-lib.c new file mode 100644 index 000000000000..2f1772a358ca --- /dev/null +++ b/drivers/rtc/rtc-mc146818-lib.c @@ -0,0 +1,198 @@ +#include +#include +#include +#include + +#ifdef CONFIG_ACPI +#include +#endif + +/* + * Returns true if a clock update is in progress + */ +static inline unsigned char mc146818_is_updating(void) +{ + unsigned char uip; + unsigned long flags; + + spin_lock_irqsave(&rtc_lock, flags); + uip = (CMOS_READ(RTC_FREQ_SELECT) & RTC_UIP); + spin_unlock_irqrestore(&rtc_lock, flags); + return uip; +} + +unsigned int mc146818_get_time(struct rtc_time *time) +{ + unsigned char ctrl; + unsigned long flags; + unsigned char century = 0; + +#ifdef CONFIG_MACH_DECSTATION + unsigned int real_year; +#endif + + /* + * read RTC once any update in progress is done. The update + * can take just over 2ms. We wait 20ms. There is no need to + * to poll-wait (up to 1s - eeccch) for the falling edge of RTC_UIP. + * If you need to know *exactly* when a second has started, enable + * periodic update complete interrupts, (via ioctl) and then + * immediately read /dev/rtc which will block until you get the IRQ. + * Once the read clears, read the RTC time (again via ioctl). Easy. + */ + if (mc146818_is_updating()) + mdelay(20); + + /* + * Only the values that we read from the RTC are set. We leave + * tm_wday, tm_yday and tm_isdst untouched. Even though the + * RTC has RTC_DAY_OF_WEEK, we ignore it, as it is only updated + * by the RTC when initially set to a non-zero value. + */ + spin_lock_irqsave(&rtc_lock, flags); + time->tm_sec = CMOS_READ(RTC_SECONDS); + time->tm_min = CMOS_READ(RTC_MINUTES); + time->tm_hour = CMOS_READ(RTC_HOURS); + time->tm_mday = CMOS_READ(RTC_DAY_OF_MONTH); + time->tm_mon = CMOS_READ(RTC_MONTH); + time->tm_year = CMOS_READ(RTC_YEAR); +#ifdef CONFIG_MACH_DECSTATION + real_year = CMOS_READ(RTC_DEC_YEAR); +#endif +#ifdef CONFIG_ACPI + if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && + acpi_gbl_FADT.century) + century = CMOS_READ(acpi_gbl_FADT.century); +#endif + ctrl = CMOS_READ(RTC_CONTROL); + spin_unlock_irqrestore(&rtc_lock, flags); + + if (!(ctrl & RTC_DM_BINARY) || RTC_ALWAYS_BCD) + { + time->tm_sec = bcd2bin(time->tm_sec); + time->tm_min = bcd2bin(time->tm_min); + time->tm_hour = bcd2bin(time->tm_hour); + time->tm_mday = bcd2bin(time->tm_mday); + time->tm_mon = bcd2bin(time->tm_mon); + time->tm_year = bcd2bin(time->tm_year); + century = bcd2bin(century); + } + +#ifdef CONFIG_MACH_DECSTATION + time->tm_year += real_year - 72; +#endif + + if (century) + time->tm_year += (century - 19) * 100; + + /* + * Account for differences between how the RTC uses the values + * and how they are defined in a struct rtc_time; + */ + if (time->tm_year <= 69) + time->tm_year += 100; + + time->tm_mon--; + + return RTC_24H; +} +EXPORT_SYMBOL_GPL(mc146818_get_time); + +/* Set the current date and time in the real time clock. */ +int mc146818_set_time(struct rtc_time *time) +{ + unsigned long flags; + unsigned char mon, day, hrs, min, sec; + unsigned char save_control, save_freq_select; + unsigned int yrs; +#ifdef CONFIG_MACH_DECSTATION + unsigned int real_yrs, leap_yr; +#endif + unsigned char century = 0; + + yrs = time->tm_year; + mon = time->tm_mon + 1; /* tm_mon starts at zero */ + day = time->tm_mday; + hrs = time->tm_hour; + min = time->tm_min; + sec = time->tm_sec; + + if (yrs > 255) /* They are unsigned */ + return -EINVAL; + + spin_lock_irqsave(&rtc_lock, flags); +#ifdef CONFIG_MACH_DECSTATION + real_yrs = yrs; + leap_yr = ((!((yrs + 1900) % 4) && ((yrs + 1900) % 100)) || + !((yrs + 1900) % 400)); + yrs = 72; + + /* + * We want to keep the year set to 73 until March + * for non-leap years, so that Feb, 29th is handled + * correctly. + */ + if (!leap_yr && mon < 3) { + real_yrs--; + yrs = 73; + } +#endif + +#ifdef CONFIG_ACPI + if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && + acpi_gbl_FADT.century) { + century = (yrs + 1900) / 100; + yrs %= 100; + } +#endif + + /* These limits and adjustments are independent of + * whether the chip is in binary mode or not. + */ + if (yrs > 169) { + spin_unlock_irqrestore(&rtc_lock, flags); + return -EINVAL; + } + + if (yrs >= 100) + yrs -= 100; + + if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) + || RTC_ALWAYS_BCD) { + sec = bin2bcd(sec); + min = bin2bcd(min); + hrs = bin2bcd(hrs); + day = bin2bcd(day); + mon = bin2bcd(mon); + yrs = bin2bcd(yrs); + century = bin2bcd(century); + } + + save_control = CMOS_READ(RTC_CONTROL); + CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL); + save_freq_select = CMOS_READ(RTC_FREQ_SELECT); + CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT); + +#ifdef CONFIG_MACH_DECSTATION + CMOS_WRITE(real_yrs, RTC_DEC_YEAR); +#endif + CMOS_WRITE(yrs, RTC_YEAR); + CMOS_WRITE(mon, RTC_MONTH); + CMOS_WRITE(day, RTC_DAY_OF_MONTH); + CMOS_WRITE(hrs, RTC_HOURS); + CMOS_WRITE(min, RTC_MINUTES); + CMOS_WRITE(sec, RTC_SECONDS); +#ifdef CONFIG_ACPI + if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && + acpi_gbl_FADT.century) + CMOS_WRITE(century, acpi_gbl_FADT.century); +#endif + + CMOS_WRITE(save_control, RTC_CONTROL); + CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); + + spin_unlock_irqrestore(&rtc_lock, flags); + + return 0; +} +EXPORT_SYMBOL_GPL(mc146818_set_time); diff --git a/include/linux/mc146818rtc.h b/include/linux/mc146818rtc.h index e9e346b37846..a585b4b5fa0e 100644 --- a/include/linux/mc146818rtc.h +++ b/include/linux/mc146818rtc.h @@ -17,10 +17,6 @@ #include #include -#ifdef CONFIG_ACPI -#include -#endif - #ifdef __KERNEL__ #include /* spinlock_t */ extern spinlock_t rtc_lock; /* serialize CMOS RAM access */ @@ -126,192 +122,7 @@ struct cmos_rtc_board_info { #define RTC_IO_EXTENT_USED RTC_IO_EXTENT #endif /* ARCH_RTC_LOCATION */ -/* - * Returns true if a clock update is in progress - */ -static inline unsigned char mc146818_is_updating(void) -{ - unsigned char uip; - unsigned long flags; - - spin_lock_irqsave(&rtc_lock, flags); - uip = (CMOS_READ(RTC_FREQ_SELECT) & RTC_UIP); - spin_unlock_irqrestore(&rtc_lock, flags); - return uip; -} - -static inline unsigned int mc146818_get_time(struct rtc_time *time) -{ - unsigned char ctrl; - unsigned long flags; - unsigned char century = 0; - -#ifdef CONFIG_MACH_DECSTATION - unsigned int real_year; -#endif - - /* - * read RTC once any update in progress is done. The update - * can take just over 2ms. We wait 20ms. There is no need to - * to poll-wait (up to 1s - eeccch) for the falling edge of RTC_UIP. - * If you need to know *exactly* when a second has started, enable - * periodic update complete interrupts, (via ioctl) and then - * immediately read /dev/rtc which will block until you get the IRQ. - * Once the read clears, read the RTC time (again via ioctl). Easy. - */ - if (mc146818_is_updating()) - mdelay(20); - - /* - * Only the values that we read from the RTC are set. We leave - * tm_wday, tm_yday and tm_isdst untouched. Even though the - * RTC has RTC_DAY_OF_WEEK, we ignore it, as it is only updated - * by the RTC when initially set to a non-zero value. - */ - spin_lock_irqsave(&rtc_lock, flags); - time->tm_sec = CMOS_READ(RTC_SECONDS); - time->tm_min = CMOS_READ(RTC_MINUTES); - time->tm_hour = CMOS_READ(RTC_HOURS); - time->tm_mday = CMOS_READ(RTC_DAY_OF_MONTH); - time->tm_mon = CMOS_READ(RTC_MONTH); - time->tm_year = CMOS_READ(RTC_YEAR); -#ifdef CONFIG_MACH_DECSTATION - real_year = CMOS_READ(RTC_DEC_YEAR); -#endif -#ifdef CONFIG_ACPI - if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && - acpi_gbl_FADT.century) - century = CMOS_READ(acpi_gbl_FADT.century); -#endif - ctrl = CMOS_READ(RTC_CONTROL); - spin_unlock_irqrestore(&rtc_lock, flags); - - if (!(ctrl & RTC_DM_BINARY) || RTC_ALWAYS_BCD) - { - time->tm_sec = bcd2bin(time->tm_sec); - time->tm_min = bcd2bin(time->tm_min); - time->tm_hour = bcd2bin(time->tm_hour); - time->tm_mday = bcd2bin(time->tm_mday); - time->tm_mon = bcd2bin(time->tm_mon); - time->tm_year = bcd2bin(time->tm_year); - century = bcd2bin(century); - } - -#ifdef CONFIG_MACH_DECSTATION - time->tm_year += real_year - 72; -#endif - - if (century) - time->tm_year += (century - 19) * 100; - - /* - * Account for differences between how the RTC uses the values - * and how they are defined in a struct rtc_time; - */ - if (time->tm_year <= 69) - time->tm_year += 100; - - time->tm_mon--; - - return RTC_24H; -} - -/* Set the current date and time in the real time clock. */ -static inline int mc146818_set_time(struct rtc_time *time) -{ - unsigned long flags; - unsigned char mon, day, hrs, min, sec; - unsigned char save_control, save_freq_select; - unsigned int yrs; -#ifdef CONFIG_MACH_DECSTATION - unsigned int real_yrs, leap_yr; -#endif - unsigned char century = 0; - - yrs = time->tm_year; - mon = time->tm_mon + 1; /* tm_mon starts at zero */ - day = time->tm_mday; - hrs = time->tm_hour; - min = time->tm_min; - sec = time->tm_sec; - - if (yrs > 255) /* They are unsigned */ - return -EINVAL; - - spin_lock_irqsave(&rtc_lock, flags); -#ifdef CONFIG_MACH_DECSTATION - real_yrs = yrs; - leap_yr = ((!((yrs + 1900) % 4) && ((yrs + 1900) % 100)) || - !((yrs + 1900) % 400)); - yrs = 72; - - /* - * We want to keep the year set to 73 until March - * for non-leap years, so that Feb, 29th is handled - * correctly. - */ - if (!leap_yr && mon < 3) { - real_yrs--; - yrs = 73; - } -#endif - -#ifdef CONFIG_ACPI - if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && - acpi_gbl_FADT.century) { - century = (yrs + 1900) / 100; - yrs %= 100; - } -#endif - - /* These limits and adjustments are independent of - * whether the chip is in binary mode or not. - */ - if (yrs > 169) { - spin_unlock_irqrestore(&rtc_lock, flags); - return -EINVAL; - } - - if (yrs >= 100) - yrs -= 100; - - if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) - || RTC_ALWAYS_BCD) { - sec = bin2bcd(sec); - min = bin2bcd(min); - hrs = bin2bcd(hrs); - day = bin2bcd(day); - mon = bin2bcd(mon); - yrs = bin2bcd(yrs); - century = bin2bcd(century); - } - - save_control = CMOS_READ(RTC_CONTROL); - CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL); - save_freq_select = CMOS_READ(RTC_FREQ_SELECT); - CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT); - -#ifdef CONFIG_MACH_DECSTATION - CMOS_WRITE(real_yrs, RTC_DEC_YEAR); -#endif - CMOS_WRITE(yrs, RTC_YEAR); - CMOS_WRITE(mon, RTC_MONTH); - CMOS_WRITE(day, RTC_DAY_OF_MONTH); - CMOS_WRITE(hrs, RTC_HOURS); - CMOS_WRITE(min, RTC_MINUTES); - CMOS_WRITE(sec, RTC_SECONDS); -#ifdef CONFIG_ACPI - if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && - acpi_gbl_FADT.century) - CMOS_WRITE(century, acpi_gbl_FADT.century); -#endif - - CMOS_WRITE(save_control, RTC_CONTROL); - CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); - - spin_unlock_irqrestore(&rtc_lock, flags); - - return 0; -} +unsigned int mc146818_get_time(struct rtc_time *time); +int mc146818_set_time(struct rtc_time *time); #endif /* _MC146818RTC_H */ -- cgit v1.2.3 From cde0fe2acd7cf6f76e926d4e2ce14439681fd5be Mon Sep 17 00:00:00 2001 From: Alexandre Belloni Date: Tue, 7 Jun 2016 20:44:05 +0200 Subject: rtc: rv8803: broaden workaround The previous workaround may still fail as there are actually 4 retries to be done to ensure the communication succeed. Also, some I2C adapter drivers may return -EIO instead of -ENXIO. Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-rv8803.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/rtc/rtc-rv8803.c b/drivers/rtc/rtc-rv8803.c index f623038e586e..022ef26bef47 100644 --- a/drivers/rtc/rtc-rv8803.c +++ b/drivers/rtc/rtc-rv8803.c @@ -68,7 +68,7 @@ static irqreturn_t rv8803_handle_irq(int irq, void *dev_id) do { flags = i2c_smbus_read_byte_data(client, RV8803_FLAG); try++; - } while ((flags == -ENXIO) && (try < 3)); + } while (((flags == -ENXIO) || (flags == -EIO)) && (try < 4)); if (flags <= 0) { mutex_unlock(&rv8803->flags_lock); return IRQ_NONE; @@ -452,7 +452,7 @@ static int rv8803_probe(struct i2c_client *client, do { flags = i2c_smbus_read_byte_data(client, RV8803_FLAG); try++; - } while ((flags == -ENXIO) && (try < 3)); + } while (((flags == -ENXIO) || (flags == -EIO)) && (try < 4)); if (flags < 0) return flags; @@ -493,7 +493,7 @@ static int rv8803_probe(struct i2c_client *client, err = i2c_smbus_write_byte_data(rv8803->client, RV8803_EXT, RV8803_EXT_WADA); try++; - } while ((err == -ENXIO) && (try < 3)); + } while (((err == -ENXIO) || (flags == -EIO)) && (try < 4)); if (err) return err; -- cgit v1.2.3 From cd26fca202f9a720d90e00ac2dbb7c4d0544f748 Mon Sep 17 00:00:00 2001 From: Andrey Smirnov Date: Tue, 21 Jun 2016 00:22:48 -0700 Subject: rtc: rtctest: Change alarm IRQ support detection For old style drivers, call a call to ioctl(..., RTC_ALM_SET, ...): - char/ds1302.c will always return -EINVAL - char/genrtc.c: will always return -EINVAL - char/rtc.c will succeed regardless if IRQs are supported or not - char/efirtc.c will always return -EINVAL - input/misc/hp_sdc_rtc.c ... that ioctl code is a good lesson about ifdefing code out and punting implementation ... and it will always return -EINVAL For new style rtc drivers, a call to ioctl(..., RTC_ALM_SET, ...) never results in a call to __rtc_set_alarm, since struct rtc_wkalarm passed to rtc_set_alarm has 'enabled' field set to 0. This means that rtc->ops->set_alarm driver hook is never called in that ioctl. Since no driver code interaction happens as a part of that call, using its results to ascertain properties of the driver is not going to work. To remedy this - use the result of RTC_AIE_ON to make the judgement. This patch also changes ENOTTY to EINVAL as an error code value that would tell us that IRQs are not supported. There are three reason for this: - As mentioned above old style driver never returns ENOTTY for this ioctl - In it's code __rtc_set_alarm() returns -EINVAL if rtc->ops->set_alarm method is not provided by the driver, so one reason for change is to be consistent with that code path. - A call to ioctl(..., RTC_UIE_ON, ...) will result in a call to rtc_update_irq_enable() and then __rtc_set_alarm(), which, if IRQs are not supported by the driver, will result in a non-zero error code. Returning ENOTTY in that case would: a) Not be consistent with other codepaths of rtc_update_irq_enable, for example the check of rtc->uie_unsupported b) Would break update IRQ emulation code since that codpath expects EINVAL c) Would break test's logic for feature support detection in the case of RTC_UIE_ON ioctl Signed-off-by: Andrey Smirnov Signed-off-by: Alexandre Belloni --- tools/testing/selftests/timers/rtctest.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/tools/testing/selftests/timers/rtctest.c b/tools/testing/selftests/timers/rtctest.c index 624bce51b27d..0cb46285ccf4 100644 --- a/tools/testing/selftests/timers/rtctest.c +++ b/tools/testing/selftests/timers/rtctest.c @@ -144,11 +144,12 @@ test_READ: retval = ioctl(fd, RTC_ALM_SET, &rtc_tm); if (retval == -1) { - if (errno == ENOTTY) { + if (errno == EINVAL) { fprintf(stderr, "\n...Alarm IRQs not supported.\n"); goto test_PIE; } + perror("RTC_ALM_SET ioctl"); exit(errno); } @@ -166,6 +167,12 @@ test_READ: /* Enable alarm interrupts */ retval = ioctl(fd, RTC_AIE_ON, 0); if (retval == -1) { + if (errno == EINVAL) { + fprintf(stderr, + "\n...Alarm IRQs not supported.\n"); + goto test_PIE; + } + perror("RTC_AIE_ON ioctl"); exit(errno); } -- cgit v1.2.3 From 519efa98051717479ff3d0b6be996e112e7a87ff Mon Sep 17 00:00:00 2001 From: Andrey Smirnov Date: Tue, 21 Jun 2016 00:22:49 -0700 Subject: rtc: rtctest: Change no IRQ detection for RTC_IRQP_READ A call to ioctl(..., RTC_IRQP_READ, ...) should never result in ENOTTY. All new style RTC drivers implement it and all of the old style drivers return EINVAL when they don't support periodic IRQs. Signed-off-by: Andrey Smirnov Signed-off-by: Alexandre Belloni --- tools/testing/selftests/timers/rtctest.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/timers/rtctest.c b/tools/testing/selftests/timers/rtctest.c index 0cb46285ccf4..97beadf548b0 100644 --- a/tools/testing/selftests/timers/rtctest.c +++ b/tools/testing/selftests/timers/rtctest.c @@ -200,7 +200,7 @@ test_PIE: retval = ioctl(fd, RTC_IRQP_READ, &tmp); if (retval == -1) { /* not all RTCs support periodic IRQs */ - if (errno == ENOTTY) { + if (errno == EINVAL) { fprintf(stderr, "\nNo periodic IRQ support\n"); goto done; } -- cgit v1.2.3 From 0a553cbabd369c52b921fced5e1a1ea3e60e0e0d Mon Sep 17 00:00:00 2001 From: Andrey Smirnov Date: Tue, 21 Jun 2016 00:22:50 -0700 Subject: rtc: rtctest: Change no IRQ detection for RTC_IRQP_SET A call to ioctl(..., RTC_IRQP_SET, ...) should never result in ENOTTY. All new style RTC drivers implement it and all of the old style drivers return EINVAL when they don't support periodic IRQs. Signed-off-by: Andrey Smirnov Signed-off-by: Alexandre Belloni --- tools/testing/selftests/timers/rtctest.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/timers/rtctest.c b/tools/testing/selftests/timers/rtctest.c index 97beadf548b0..4230d3052e5d 100644 --- a/tools/testing/selftests/timers/rtctest.c +++ b/tools/testing/selftests/timers/rtctest.c @@ -218,7 +218,7 @@ test_PIE: retval = ioctl(fd, RTC_IRQP_SET, tmp); if (retval == -1) { /* not all RTCs can change their periodic IRQ rate */ - if (errno == ENOTTY) { + if (errno == EINVAL) { fprintf(stderr, "\n...Periodic IRQ rate is fixed\n"); goto done; -- cgit v1.2.3 From d68fa32dccc136bb4e092d53606bb6f5515fa972 Mon Sep 17 00:00:00 2001 From: Alexandre Belloni Date: Sun, 26 Jun 2016 23:20:37 +0200 Subject: rtc: ds17287: remove unused header ds17287rtc.h is unused since 15beb694c661 ("mips: ip32: add platform data hooks to use DS1685 driver"), remove it. Signed-off-by: Alexandre Belloni --- include/linux/ds17287rtc.h | 66 ---------------------------------------------- 1 file changed, 66 deletions(-) delete mode 100644 include/linux/ds17287rtc.h diff --git a/include/linux/ds17287rtc.h b/include/linux/ds17287rtc.h deleted file mode 100644 index d85d3f497b96..000000000000 --- a/include/linux/ds17287rtc.h +++ /dev/null @@ -1,66 +0,0 @@ -/* - * ds17287rtc.h - register definitions for the ds1728[57] RTC / CMOS RAM - * - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * (C) 2003 Guido Guenther - */ -#ifndef __LINUX_DS17287RTC_H -#define __LINUX_DS17287RTC_H - -#include /* get the user-level API */ -#include - -/* Register A */ -#define DS_REGA_DV2 0x40 /* countdown chain */ -#define DS_REGA_DV1 0x20 /* oscillator enable */ -#define DS_REGA_DV0 0x10 /* bank select */ - -/* bank 1 registers */ -#define DS_B1_MODEL 0x40 /* model number byte */ -#define DS_B1_SN1 0x41 /* serial number byte 1 */ -#define DS_B1_SN2 0x42 /* serial number byte 2 */ -#define DS_B1_SN3 0x43 /* serial number byte 3 */ -#define DS_B1_SN4 0x44 /* serial number byte 4 */ -#define DS_B1_SN5 0x45 /* serial number byte 5 */ -#define DS_B1_SN6 0x46 /* serial number byte 6 */ -#define DS_B1_CRC 0x47 /* CRC byte */ -#define DS_B1_CENTURY 0x48 /* Century byte */ -#define DS_B1_DALARM 0x49 /* date alarm */ -#define DS_B1_XCTRL4A 0x4a /* extendec control register 4a */ -#define DS_B1_XCTRL4B 0x4b /* extendec control register 4b */ -#define DS_B1_RTCADDR2 0x4e /* rtc address 2 */ -#define DS_B1_RTCADDR3 0x4f /* rtc address 3 */ -#define DS_B1_RAMLSB 0x50 /* extended ram LSB */ -#define DS_B1_RAMMSB 0x51 /* extended ram MSB */ -#define DS_B1_RAMDPORT 0x53 /* extended ram data port */ - -/* register details */ -/* extended control register 4a */ -#define DS_XCTRL4A_VRT2 0x80 /* valid ram and time */ -#define DS_XCTRL4A_INCR 0x40 /* increment progress status */ -#define DS_XCTRL4A_BME 0x20 /* burst mode enable */ -#define DS_XCTRL4A_PAB 0x08 /* power active bar ctrl */ -#define DS_XCTRL4A_RF 0x04 /* ram clear flag */ -#define DS_XCTRL4A_WF 0x02 /* wake up alarm flag */ -#define DS_XCTRL4A_KF 0x01 /* kickstart flag */ - -/* interrupt causes */ -#define DS_XCTRL4A_IFS (DS_XCTRL4A_RF|DS_XCTRL4A_WF|DS_XCTRL4A_KF) - -/* extended control register 4b */ -#define DS_XCTRL4B_ABE 0x80 /* auxiliary battery enable */ -#define DS_XCTRL4B_E32K 0x40 /* enable 32.768 kHz Output */ -#define DS_XCTRL4B_CS 0x20 /* crystal select */ -#define DS_XCTRL4B_RCE 0x10 /* ram clear enable */ -#define DS_XCTRL4B_PRS 0x08 /* PAB resec select */ -#define DS_XCTRL4B_RIE 0x04 /* ram clear interrupt enable */ -#define DS_XCTRL4B_WFE 0x02 /* wake up alarm interrupt enable */ -#define DS_XCTRL4B_KFE 0x01 /* kickstart interrupt enable */ - -/* interrupt enable bits */ -#define DS_XCTRL4B_IFES (DS_XCTRL4B_RIE|DS_XCTRL4B_WFE|DS_XCTRL4B_KFE) - -#endif /* __LINUX_DS17287RTC_H */ -- cgit v1.2.3 From 10c2a2e71301deb4d7760c140606bc61841aacd3 Mon Sep 17 00:00:00 2001 From: Alexandre Belloni Date: Sun, 26 Jun 2016 23:38:44 +0200 Subject: rtc: ds2404: move rtc-ds2404.h to platform_data rtc-ds2404.h belongs to include/linux/platform_data/ Signed-off-by: Alexandre Belloni --- drivers/rtc/rtc-ds2404.c | 2 +- include/linux/platform_data/rtc-ds2404.h | 20 ++++++++++++++++++++ include/linux/rtc-ds2404.h | 20 -------------------- 3 files changed, 21 insertions(+), 21 deletions(-) create mode 100644 include/linux/platform_data/rtc-ds2404.h delete mode 100644 include/linux/rtc-ds2404.h diff --git a/drivers/rtc/rtc-ds2404.c b/drivers/rtc/rtc-ds2404.c index 16310fe79d76..9a1582ed7070 100644 --- a/drivers/rtc/rtc-ds2404.c +++ b/drivers/rtc/rtc-ds2404.c @@ -13,7 +13,7 @@ #include #include #include -#include +#include #include #include #include diff --git a/include/linux/platform_data/rtc-ds2404.h b/include/linux/platform_data/rtc-ds2404.h new file mode 100644 index 000000000000..22c53825528f --- /dev/null +++ b/include/linux/platform_data/rtc-ds2404.h @@ -0,0 +1,20 @@ +/* + * ds2404.h - platform data structure for the DS2404 RTC. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2012 Sven Schnelle + */ + +#ifndef __LINUX_DS2404_H +#define __LINUX_DS2404_H + +struct ds2404_platform_data { + + unsigned int gpio_rst; + unsigned int gpio_clk; + unsigned int gpio_dq; +}; +#endif diff --git a/include/linux/rtc-ds2404.h b/include/linux/rtc-ds2404.h deleted file mode 100644 index 22c53825528f..000000000000 --- a/include/linux/rtc-ds2404.h +++ /dev/null @@ -1,20 +0,0 @@ -/* - * ds2404.h - platform data structure for the DS2404 RTC. - * - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Copyright (C) 2012 Sven Schnelle - */ - -#ifndef __LINUX_DS2404_H -#define __LINUX_DS2404_H - -struct ds2404_platform_data { - - unsigned int gpio_rst; - unsigned int gpio_clk; - unsigned int gpio_dq; -}; -#endif -- cgit v1.2.3 From 803bb30145df3132f9c8c8704d11a9e6732340c7 Mon Sep 17 00:00:00 2001 From: Alexandre Belloni Date: Sun, 26 Jun 2016 22:57:52 +0200 Subject: rtc: m48t86: move m48t86.h to platform_data m48t86.h belongs to include/linux/platform_data/ Acked-by: Jason Cooper Acked-by: H Hartley Sweeten Acked-by: Alexander Clouter Signed-off-by: Alexandre Belloni --- arch/arm/mach-ep93xx/ts72xx.c | 2 +- arch/arm/mach-orion5x/ts78xx-setup.c | 2 +- drivers/rtc/rtc-m48t86.c | 2 +- include/linux/m48t86.h | 16 ---------------- include/linux/platform_data/rtc-m48t86.h | 16 ++++++++++++++++ 5 files changed, 19 insertions(+), 19 deletions(-) delete mode 100644 include/linux/m48t86.h create mode 100644 include/linux/platform_data/rtc-m48t86.h diff --git a/arch/arm/mach-ep93xx/ts72xx.c b/arch/arm/mach-ep93xx/ts72xx.c index 45b81a2bcd4b..3b39ea353d30 100644 --- a/arch/arm/mach-ep93xx/ts72xx.c +++ b/arch/arm/mach-ep93xx/ts72xx.c @@ -16,7 +16,7 @@ #include #include #include -#include +#include #include #include diff --git a/arch/arm/mach-orion5x/ts78xx-setup.c b/arch/arm/mach-orion5x/ts78xx-setup.c index 3a58a5d4a28a..8d597267d0c4 100644 --- a/arch/arm/mach-orion5x/ts78xx-setup.c +++ b/arch/arm/mach-orion5x/ts78xx-setup.c @@ -16,7 +16,7 @@ #include #include #include -#include +#include #include #include #include diff --git a/drivers/rtc/rtc-m48t86.c b/drivers/rtc/rtc-m48t86.c index f72b91f2501f..0eeb5714c00f 100644 --- a/drivers/rtc/rtc-m48t86.c +++ b/drivers/rtc/rtc-m48t86.c @@ -16,7 +16,7 @@ #include #include #include -#include +#include #include #define M48T86_REG_SEC 0x00 diff --git a/include/linux/m48t86.h b/include/linux/m48t86.h deleted file mode 100644 index 915d6b4f0f89..000000000000 --- a/include/linux/m48t86.h +++ /dev/null @@ -1,16 +0,0 @@ -/* - * ST M48T86 / Dallas DS12887 RTC driver - * Copyright (c) 2006 Tower Technologies - * - * Author: Alessandro Zummo - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. -*/ - -struct m48t86_ops -{ - void (*writebyte)(unsigned char value, unsigned long addr); - unsigned char (*readbyte)(unsigned long addr); -}; diff --git a/include/linux/platform_data/rtc-m48t86.h b/include/linux/platform_data/rtc-m48t86.h new file mode 100644 index 000000000000..915d6b4f0f89 --- /dev/null +++ b/include/linux/platform_data/rtc-m48t86.h @@ -0,0 +1,16 @@ +/* + * ST M48T86 / Dallas DS12887 RTC driver + * Copyright (c) 2006 Tower Technologies + * + * Author: Alessandro Zummo + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. +*/ + +struct m48t86_ops +{ + void (*writebyte)(unsigned char value, unsigned long addr); + unsigned char (*readbyte)(unsigned long addr); +}; -- cgit v1.2.3 From fe23c336125c50a8ef90619f8bfab414d211662d Mon Sep 17 00:00:00 2001 From: Alexandre Belloni Date: Mon, 27 Jun 2016 21:58:40 +0200 Subject: rtc: adjust MAINTAINERS entry Add missing files to the RTC entry of MAINTAINERS Signed-off-by: Alexandre Belloni --- MAINTAINERS | 3 +++ 1 file changed, 3 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 72c28f4520d5..e86cf021113c 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9563,6 +9563,9 @@ F: Documentation/rtc.txt F: drivers/rtc/ F: include/linux/rtc.h F: include/uapi/linux/rtc.h +F: include/linux/rtc/ +F: include/linux/platform_data/rtc-* +F: tools/testing/selftests/timers/rtctest.c REALTEK AUDIO CODECS M: Bard Liao -- cgit v1.2.3 From db1bb44c4c7e8d49ed674dc59e5222d99c698088 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 24 Jun 2016 10:55:44 -0400 Subject: SUNRPC: Don't allocate a full sockaddr_storage for tracing We're always tracing IPv4 or IPv6 addresses, so we can save a lot of space on the ringbuffer by allocating the correct sockaddr size. Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org Fixes: 83a712e0afef "sunrpc: add some tracepoints around ..." Signed-off-by: J. Bruce Fields --- include/trace/events/sunrpc.h | 47 +++++++++++++++++++++++++++++++------------ 1 file changed, 34 insertions(+), 13 deletions(-) diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 003dca933803..5664ca07c9c7 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -529,20 +529,27 @@ TRACE_EVENT(svc_xprt_do_enqueue, TP_STRUCT__entry( __field(struct svc_xprt *, xprt) - __field_struct(struct sockaddr_storage, ss) __field(int, pid) __field(unsigned long, flags) + __dynamic_array(unsigned char, addr, xprt != NULL ? + xprt->xpt_remotelen : 0) ), TP_fast_assign( __entry->xprt = xprt; - xprt ? memcpy(&__entry->ss, &xprt->xpt_remote, sizeof(__entry->ss)) : memset(&__entry->ss, 0, sizeof(__entry->ss)); __entry->pid = rqst? rqst->rq_task->pid : 0; - __entry->flags = xprt ? xprt->xpt_flags : 0; + if (xprt) { + memcpy(__get_dynamic_array(addr), + &xprt->xpt_remote, + xprt->xpt_remotelen); + __entry->flags = xprt->xpt_flags; + } else + __entry->flags = 0; ), TP_printk("xprt=0x%p addr=%pIScp pid=%d flags=%s", __entry->xprt, - (struct sockaddr *)&__entry->ss, + __get_dynamic_array_len(addr) != 0 ? + (struct sockaddr *)__get_dynamic_array(addr) : NULL, __entry->pid, show_svc_xprt_flags(__entry->flags)) ); @@ -553,18 +560,25 @@ TRACE_EVENT(svc_xprt_dequeue, TP_STRUCT__entry( __field(struct svc_xprt *, xprt) - __field_struct(struct sockaddr_storage, ss) __field(unsigned long, flags) + __dynamic_array(unsigned char, addr, xprt != NULL ? + xprt->xpt_remotelen : 0) ), TP_fast_assign( - __entry->xprt = xprt, - xprt ? memcpy(&__entry->ss, &xprt->xpt_remote, sizeof(__entry->ss)) : memset(&__entry->ss, 0, sizeof(__entry->ss)); - __entry->flags = xprt ? xprt->xpt_flags : 0; + __entry->xprt = xprt; + if (xprt) { + memcpy(__get_dynamic_array(addr), + &xprt->xpt_remote, + xprt->xpt_remotelen); + __entry->flags = xprt->xpt_flags; + } else + __entry->flags = 0; ), TP_printk("xprt=0x%p addr=%pIScp flags=%s", __entry->xprt, - (struct sockaddr *)&__entry->ss, + __get_dynamic_array_len(addr) != 0 ? + (struct sockaddr *)__get_dynamic_array(addr) : NULL, show_svc_xprt_flags(__entry->flags)) ); @@ -592,19 +606,26 @@ TRACE_EVENT(svc_handle_xprt, TP_STRUCT__entry( __field(struct svc_xprt *, xprt) __field(int, len) - __field_struct(struct sockaddr_storage, ss) __field(unsigned long, flags) + __dynamic_array(unsigned char, addr, xprt != NULL ? + xprt->xpt_remotelen : 0) ), TP_fast_assign( __entry->xprt = xprt; - xprt ? memcpy(&__entry->ss, &xprt->xpt_remote, sizeof(__entry->ss)) : memset(&__entry->ss, 0, sizeof(__entry->ss)); __entry->len = len; - __entry->flags = xprt ? xprt->xpt_flags : 0; + if (xprt) { + memcpy(__get_dynamic_array(addr), + &xprt->xpt_remote, + xprt->xpt_remotelen); + __entry->flags = xprt->xpt_flags; + } else + __entry->flags = 0; ), TP_printk("xprt=0x%p addr=%pIScp len=%d flags=%s", __entry->xprt, - (struct sockaddr *)&__entry->ss, + __get_dynamic_array_len(addr) != 0 ? + (struct sockaddr *)__get_dynamic_array(addr) : NULL, __entry->len, show_svc_xprt_flags(__entry->flags)) ); #endif /* _TRACE_SUNRPC_H */ -- cgit v1.2.3 From c683ffe2193b31bba64650fc74d1d6ba254eaafd Mon Sep 17 00:00:00 2001 From: Hugh Sipière Date: Sat, 4 Jun 2016 17:17:01 +0100 Subject: SSB: Change bare unsigned to unsigned int to suit coding style MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit These lines just have unsigned gpio rather than unsigned int gpio. I changed it to suit the coding style. Michael Buesch told me to send this to the MIPS tree. Signed-off-by: Hugh Sipière Acked-by: Michael Buesch Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/13460/ Signed-off-by: Ralf Baechle --- drivers/ssb/driver_gpio.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/drivers/ssb/driver_gpio.c b/drivers/ssb/driver_gpio.c index 180e027b1c8a..796e22037bc4 100644 --- a/drivers/ssb/driver_gpio.c +++ b/drivers/ssb/driver_gpio.c @@ -23,7 +23,7 @@ **************************************************/ #if IS_ENABLED(CONFIG_SSB_EMBEDDED) -static int ssb_gpio_to_irq(struct gpio_chip *chip, unsigned gpio) +static int ssb_gpio_to_irq(struct gpio_chip *chip, unsigned int gpio) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -38,14 +38,14 @@ static int ssb_gpio_to_irq(struct gpio_chip *chip, unsigned gpio) * ChipCommon **************************************************/ -static int ssb_gpio_chipco_get_value(struct gpio_chip *chip, unsigned gpio) +static int ssb_gpio_chipco_get_value(struct gpio_chip *chip, unsigned int gpio) { struct ssb_bus *bus = gpiochip_get_data(chip); return !!ssb_chipco_gpio_in(&bus->chipco, 1 << gpio); } -static void ssb_gpio_chipco_set_value(struct gpio_chip *chip, unsigned gpio, +static void ssb_gpio_chipco_set_value(struct gpio_chip *chip, unsigned int gpio, int value) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -54,7 +54,7 @@ static void ssb_gpio_chipco_set_value(struct gpio_chip *chip, unsigned gpio, } static int ssb_gpio_chipco_direction_input(struct gpio_chip *chip, - unsigned gpio) + unsigned int gpio) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -63,7 +63,7 @@ static int ssb_gpio_chipco_direction_input(struct gpio_chip *chip, } static int ssb_gpio_chipco_direction_output(struct gpio_chip *chip, - unsigned gpio, int value) + unsigned int gpio, int value) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -72,7 +72,7 @@ static int ssb_gpio_chipco_direction_output(struct gpio_chip *chip, return 0; } -static int ssb_gpio_chipco_request(struct gpio_chip *chip, unsigned gpio) +static int ssb_gpio_chipco_request(struct gpio_chip *chip, unsigned int gpio) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -85,7 +85,7 @@ static int ssb_gpio_chipco_request(struct gpio_chip *chip, unsigned gpio) return 0; } -static void ssb_gpio_chipco_free(struct gpio_chip *chip, unsigned gpio) +static void ssb_gpio_chipco_free(struct gpio_chip *chip, unsigned int gpio) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -256,14 +256,14 @@ static int ssb_gpio_chipco_init(struct ssb_bus *bus) #ifdef CONFIG_SSB_DRIVER_EXTIF -static int ssb_gpio_extif_get_value(struct gpio_chip *chip, unsigned gpio) +static int ssb_gpio_extif_get_value(struct gpio_chip *chip, unsigned int gpio) { struct ssb_bus *bus = gpiochip_get_data(chip); return !!ssb_extif_gpio_in(&bus->extif, 1 << gpio); } -static void ssb_gpio_extif_set_value(struct gpio_chip *chip, unsigned gpio, +static void ssb_gpio_extif_set_value(struct gpio_chip *chip, unsigned int gpio, int value) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -272,7 +272,7 @@ static void ssb_gpio_extif_set_value(struct gpio_chip *chip, unsigned gpio, } static int ssb_gpio_extif_direction_input(struct gpio_chip *chip, - unsigned gpio) + unsigned int gpio) { struct ssb_bus *bus = gpiochip_get_data(chip); @@ -281,7 +281,7 @@ static int ssb_gpio_extif_direction_input(struct gpio_chip *chip, } static int ssb_gpio_extif_direction_output(struct gpio_chip *chip, - unsigned gpio, int value) + unsigned int gpio, int value) { struct ssb_bus *bus = gpiochip_get_data(chip); -- cgit v1.2.3 From c00ab4896ed5f7d89af6f90b809e2c0197c6d170 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Sat, 2 Jul 2016 10:38:05 +0200 Subject: MIPS: Remove cpu_has_safe_index_cacheops Very early versions of the 1004K had an hardware issue that made index cache ops unsafe so they had to be avoided and hit ops be used instead. This may significantly slow down cache maintenance operations. Only very early FPGA versions of the 1004K were affected so let's get rid of the workaround which was only implemented for the DMA cache maintenance operations anyway. Signed-off-by: Ralf Baechle --- arch/mips/mm/c-r4k.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c index ef7f925dd1b0..645c69c95c9c 100644 --- a/arch/mips/mm/c-r4k.c +++ b/arch/mips/mm/c-r4k.c @@ -66,12 +66,6 @@ static inline void r4k_on_each_cpu(void (*func) (void *info), void *info) preempt_enable(); } -#if defined(CONFIG_MIPS_CMP) || defined(CONFIG_MIPS_CPS) -#define cpu_has_safe_index_cacheops 0 -#else -#define cpu_has_safe_index_cacheops 1 -#endif - /* * Must die. */ @@ -744,7 +738,7 @@ static void r4k_dma_cache_wback_inv(unsigned long addr, unsigned long size) * subset property so we have to flush the primary caches * explicitly */ - if (cpu_has_safe_index_cacheops && size >= dcache_size) { + if (size >= dcache_size) { r4k_blast_dcache(); } else { R4600_HIT_CACHEOP_WAR_IMPL; @@ -781,7 +775,7 @@ static void r4k_dma_cache_inv(unsigned long addr, unsigned long size) return; } - if (cpu_has_safe_index_cacheops && size >= dcache_size) { + if (size >= dcache_size) { r4k_blast_dcache(); } else { R4600_HIT_CACHEOP_WAR_IMPL; @@ -861,7 +855,7 @@ static inline void local_r4k_flush_kernel_vmap_range(void *args) * Aliases only affect the primary caches so don't bother with * S-caches or T-caches. */ - if (cpu_has_safe_index_cacheops && size >= dcache_size) + if (size >= dcache_size) r4k_blast_dcache(); else { R4600_HIT_CACHEOP_WAR_IMPL; -- cgit v1.2.3 From acd168c0bf2ce709f056a6b1bf21634b1207d7a5 Mon Sep 17 00:00:00 2001 From: David Daney Date: Thu, 16 Jun 2016 15:50:31 -0700 Subject: MIPS: Fix page table corruption on THP permission changes. When the core THP code is modifying the permissions of a huge page it calls pmd_modify(), which unfortunately was clearing the _PAGE_HUGE bit of the page table entry. The result can be kernel messages like: mm/memory.c:397: bad pmd 000000040080004d. mm/memory.c:397: bad pmd 00000003ff00004d. mm/memory.c:397: bad pmd 000000040100004d. or: ------------[ cut here ]------------ WARNING: at mm/mmap.c:3200 exit_mmap+0x150/0x158() Modules linked in: ipv6 at24 octeon3_ethernet octeon_srio_nexus m25p80 CPU: 12 PID: 1295 Comm: pmderr Not tainted 3.10.87-rt80-Cavium-Octeon #4 Stack : 0000000040808000 0000000014009ce1 0000000000400004 ffffffff81076ba0 0000000000000000 0000000000000000 ffffffff85110000 0000000000000119 0000000000000004 0000000000000000 0000000000000119 43617669756d2d4f 0000000000000000 ffffffff850fda40 ffffffff85110000 0000000000000000 0000000000000000 0000000000000009 ffffffff809207a0 0000000000000c80 ffffffff80f1bf20 0000000000000001 000000ffeca36828 0000000000000001 0000000000000000 0000000000000001 000000ffeca7e700 ffffffff80886924 80000003fd7a0000 80000003fd7a39b0 80000003fdea8000 ffffffff80885780 80000003fdea8000 ffffffff80f12218 000000000000000c 000000000000050f 0000000000000000 ffffffff80865c4c 0000000000000000 0000000000000000 ... Call Trace: [] show_stack+0x6c/0xf8 [] warn_slowpath_common+0x78/0xa8 [] exit_mmap+0x150/0x158 [] mmput+0x5c/0x110 [] do_exit+0x230/0xa68 [] do_group_exit+0x54/0x1d0 [] __wake_up_parent+0x0/0x18 ---[ end trace c7b38293191c57dc ]--- BUG: Bad rss-counter state mm:80000003fa168000 idx:1 val:1536 Fix by not clearing _PAGE_HUGE bit. Signed-off-by: David Daney Tested-by: Aaro Koskinen Cc: stable@vger.kernel.org Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/13687/ Signed-off-by: Ralf Baechle --- arch/mips/include/asm/pgtable.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h index f53816744d60..7d44e888134f 100644 --- a/arch/mips/include/asm/pgtable.h +++ b/arch/mips/include/asm/pgtable.h @@ -633,7 +633,7 @@ static inline struct page *pmd_page(pmd_t pmd) static inline pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot) { - pmd_val(pmd) = (pmd_val(pmd) & _PAGE_CHG_MASK) | + pmd_val(pmd) = (pmd_val(pmd) & (_PAGE_CHG_MASK | _PAGE_HUGE)) | (pgprot_val(newprot) & ~_PAGE_CHG_MASK); return pmd; } -- cgit v1.2.3 From d4a5f6d71e8bd32f268e738ed39b31a15334cdf7 Mon Sep 17 00:00:00 2001 From: Alexandre Belloni Date: Mon, 27 Jun 2016 00:03:00 +0200 Subject: rtc: ds1286: move header to linux/rtc Move ds1286.h to rtc specific folder. Signed-off-by: Alexandre Belloni --- arch/mips/sgi-ip22/ip22-reset.c | 2 +- drivers/rtc/rtc-ds1286.c | 2 +- include/linux/ds1286.h | 52 ----------------------------------------- include/linux/rtc/ds1286.h | 52 +++++++++++++++++++++++++++++++++++++++++ 4 files changed, 54 insertions(+), 54 deletions(-) delete mode 100644 include/linux/ds1286.h create mode 100644 include/linux/rtc/ds1286.h diff --git a/arch/mips/sgi-ip22/ip22-reset.c b/arch/mips/sgi-ip22/ip22-reset.c index 063c2dd31e72..2f45b0357021 100644 --- a/arch/mips/sgi-ip22/ip22-reset.c +++ b/arch/mips/sgi-ip22/ip22-reset.c @@ -7,7 +7,7 @@ */ #include #include -#include +#include #include #include #include diff --git a/drivers/rtc/rtc-ds1286.c b/drivers/rtc/rtc-ds1286.c index 756e509f6ed2..ef75c349dff9 100644 --- a/drivers/rtc/rtc-ds1286.c +++ b/drivers/rtc/rtc-ds1286.c @@ -16,7 +16,7 @@ #include #include #include -#include +#include #include #include diff --git a/include/linux/ds1286.h b/include/linux/ds1286.h deleted file mode 100644 index 45ea0aa0aeb9..000000000000 --- a/include/linux/ds1286.h +++ /dev/null @@ -1,52 +0,0 @@ -/* - * Copyright (C) 1998, 1999, 2003 Ralf Baechle - * - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - */ -#ifndef __LINUX_DS1286_H -#define __LINUX_DS1286_H - -/********************************************************************** - * register summary - **********************************************************************/ -#define RTC_HUNDREDTH_SECOND 0 -#define RTC_SECONDS 1 -#define RTC_MINUTES 2 -#define RTC_MINUTES_ALARM 3 -#define RTC_HOURS 4 -#define RTC_HOURS_ALARM 5 -#define RTC_DAY 6 -#define RTC_DAY_ALARM 7 -#define RTC_DATE 8 -#define RTC_MONTH 9 -#define RTC_YEAR 10 -#define RTC_CMD 11 -#define RTC_WHSEC 12 -#define RTC_WSEC 13 -#define RTC_UNUSED 14 - -/* RTC_*_alarm is always true if 2 MSBs are set */ -# define RTC_ALARM_DONT_CARE 0xC0 - - -/* - * Bits in the month register - */ -#define RTC_EOSC 0x80 -#define RTC_ESQW 0x40 - -/* - * Bits in the Command register - */ -#define RTC_TDF 0x01 -#define RTC_WAF 0x02 -#define RTC_TDM 0x04 -#define RTC_WAM 0x08 -#define RTC_PU_LVL 0x10 -#define RTC_IBH_LO 0x20 -#define RTC_IPSW 0x40 -#define RTC_TE 0x80 - -#endif /* __LINUX_DS1286_H */ diff --git a/include/linux/rtc/ds1286.h b/include/linux/rtc/ds1286.h new file mode 100644 index 000000000000..45ea0aa0aeb9 --- /dev/null +++ b/include/linux/rtc/ds1286.h @@ -0,0 +1,52 @@ +/* + * Copyright (C) 1998, 1999, 2003 Ralf Baechle + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + */ +#ifndef __LINUX_DS1286_H +#define __LINUX_DS1286_H + +/********************************************************************** + * register summary + **********************************************************************/ +#define RTC_HUNDREDTH_SECOND 0 +#define RTC_SECONDS 1 +#define RTC_MINUTES 2 +#define RTC_MINUTES_ALARM 3 +#define RTC_HOURS 4 +#define RTC_HOURS_ALARM 5 +#define RTC_DAY 6 +#define RTC_DAY_ALARM 7 +#define RTC_DATE 8 +#define RTC_MONTH 9 +#define RTC_YEAR 10 +#define RTC_CMD 11 +#define RTC_WHSEC 12 +#define RTC_WSEC 13 +#define RTC_UNUSED 14 + +/* RTC_*_alarm is always true if 2 MSBs are set */ +# define RTC_ALARM_DONT_CARE 0xC0 + + +/* + * Bits in the month register + */ +#define RTC_EOSC 0x80 +#define RTC_ESQW 0x40 + +/* + * Bits in the Command register + */ +#define RTC_TDF 0x01 +#define RTC_WAF 0x02 +#define RTC_TDM 0x04 +#define RTC_WAM 0x08 +#define RTC_PU_LVL 0x10 +#define RTC_IBH_LO 0x20 +#define RTC_IPSW 0x40 +#define RTC_TE 0x80 + +#endif /* __LINUX_DS1286_H */ -- cgit v1.2.3 From a6a0dbbcfa469cf3e5c4d9522106c0b7b9e9e373 Mon Sep 17 00:00:00 2001 From: Boris Brezillon Date: Tue, 14 Jun 2016 11:13:09 +0200 Subject: pwm: Add a helper to prepare a new PWM state The pwm_init_state() helper prepares a new state object containing the current PWM state except for the polarity and period fields which are set to the reference values (those in struct pwm_args). This is particularly useful for PWM users who want to apply a new duty- cycle expressed relatively to the reference period without changing the enable state. Signed-off-by: Boris Brezillon Tested-by: Heiko Stuebner Signed-off-by: Thierry Reding --- include/linux/pwm.h | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) diff --git a/include/linux/pwm.h b/include/linux/pwm.h index 17018f3c066e..a100f6e80738 100644 --- a/include/linux/pwm.h +++ b/include/linux/pwm.h @@ -147,6 +147,39 @@ static inline void pwm_get_args(const struct pwm_device *pwm, *args = pwm->args; } +/** + * pwm_init_state() - prepare a new state to be applied with pwm_apply_state() + * @pwm: PWM device + * @state: state to fill with the prepared PWM state + * + * This functions prepares a state that can later be tweaked and applied + * to the PWM device with pwm_apply_state(). This is a convenient function + * that first retrieves the current PWM state and the replaces the period + * and polarity fields with the reference values defined in pwm->args. + * Once the function returns, you can adjust the ->enabled and ->duty_cycle + * fields according to your needs before calling pwm_apply_state(). + * + * ->duty_cycle is initially set to zero to avoid cases where the current + * ->duty_cycle value exceed the pwm_args->period one, which would trigger + * an error if the user calls pwm_apply_state() without adjusting ->duty_cycle + * first. + */ +static inline void pwm_init_state(const struct pwm_device *pwm, + struct pwm_state *state) +{ + struct pwm_args args; + + /* First get the current state. */ + pwm_get_state(pwm, state); + + /* Then fill it with the reference config */ + pwm_get_args(pwm, &args); + + state->period = args.period; + state->polarity = args.polarity; + state->duty_cycle = 0; +} + /** * struct pwm_ops - PWM controller operations * @request: optional hook for requesting a PWM -- cgit v1.2.3 From f6f3bddf7b2b994a927808fcc5a3d07069c35956 Mon Sep 17 00:00:00 2001 From: Boris Brezillon Date: Tue, 14 Jun 2016 11:13:10 +0200 Subject: pwm: Add relative duty cycle manipulation helpers The PWM framework expects PWM users to configure the duty cycle in nano- seconds, but many users want to express the duty cycle relatively to the period value (i.e. duty_cycle = 33% of the period). Add the pwm_{get,set}_relative_duty_cycle() helpers to ease this kind of conversion. Signed-off-by: Boris Brezillon Tested-by: Heiko Stuebner Signed-off-by: Thierry Reding --- include/linux/pwm.h | 55 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 55 insertions(+) diff --git a/include/linux/pwm.h b/include/linux/pwm.h index a100f6e80738..fd1092729ed6 100644 --- a/include/linux/pwm.h +++ b/include/linux/pwm.h @@ -180,6 +180,61 @@ static inline void pwm_init_state(const struct pwm_device *pwm, state->duty_cycle = 0; } +/** + * pwm_get_relative_duty_cycle() - Get a relative duty cycle value + * @state: PWM state to extract the duty cycle from + * @scale: target scale of the relative duty cycle + * + * This functions converts the absolute duty cycle stored in @state (expressed + * in nanosecond) into a value relative to the period. + * + * For example if you want to get the duty_cycle expressed in percent, call: + * + * pwm_get_state(pwm, &state); + * duty = pwm_get_relative_duty_cycle(&state, 100); + */ +static inline unsigned int +pwm_get_relative_duty_cycle(const struct pwm_state *state, unsigned int scale) +{ + if (!state->period) + return 0; + + return DIV_ROUND_CLOSEST_ULL((u64)state->duty_cycle * scale, + state->period); +} + +/** + * pwm_set_relative_duty_cycle() - Set a relative duty cycle value + * @state: PWM state to fill + * @duty_cycle: relative duty cycle value + * @scale: scale in which @duty_cycle is expressed + * + * This functions converts a relative into an absolute duty cycle (expressed + * in nanoseconds), and puts the result in state->duty_cycle. + * + * For example if you want to configure a 50% duty cycle, call: + * + * pwm_init_state(pwm, &state); + * pwm_set_relative_duty_cycle(&state, 50, 100); + * pwm_apply_state(pwm, &state); + * + * This functions returns -EINVAL if @duty_cycle and/or @scale are + * inconsistent (@scale == 0 or @duty_cycle > @scale). + */ +static inline int +pwm_set_relative_duty_cycle(struct pwm_state *state, unsigned int duty_cycle, + unsigned int scale) +{ + if (!scale || duty_cycle > scale) + return -EINVAL; + + state->duty_cycle = DIV_ROUND_CLOSEST_ULL((u64)duty_cycle * + state->period, + scale); + + return 0; +} + /** * struct pwm_ops - PWM controller operations * @request: optional hook for requesting a PWM -- cgit v1.2.3 From e1aaf89a5415ab6514cb6798148c7fe04c0ee138 Mon Sep 17 00:00:00 2001 From: "Cooper Jr., Franklin" Date: Thu, 17 Mar 2016 20:15:23 -0500 Subject: pwm: pwm-tiehrpwm: Update DT binding document to use generic node name Now that the node name has been changed from ehrpwm to pwm the document should show this proper usage. Also change the unit address in the example from 0 to the proper physical address value that should be used. Signed-off-by: Franklin S Cooper Jr Acked-by: Rob Herring Signed-off-by: Thierry Reding --- Documentation/devicetree/bindings/pwm/pwm-tiehrpwm.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Documentation/devicetree/bindings/pwm/pwm-tiehrpwm.txt b/Documentation/devicetree/bindings/pwm/pwm-tiehrpwm.txt index 9c100b2c5b23..0f9ba509ebd6 100644 --- a/Documentation/devicetree/bindings/pwm/pwm-tiehrpwm.txt +++ b/Documentation/devicetree/bindings/pwm/pwm-tiehrpwm.txt @@ -15,14 +15,14 @@ Optional properties: Example: -ehrpwm0: ehrpwm@0 { /* EHRPWM on am33xx */ +ehrpwm0: pwm@48300200 { /* EHRPWM on am33xx */ compatible = "ti,am33xx-ehrpwm"; #pwm-cells = <3>; reg = <0x48300200 0x100>; ti,hwmods = "ehrpwm0"; }; -ehrpwm0: ehrpwm@0 { /* EHRPWM on da850 */ +ehrpwm0: pwm@300000 { /* EHRPWM on da850 */ compatible = "ti,da850-ehrpwm", "ti,am33xx-ehrpwm"; #pwm-cells = <3>; reg = <0x300000 0x2000>; -- cgit v1.2.3 From ae5200d2271691b28aff8d0409fbefd4644ff507 Mon Sep 17 00:00:00 2001 From: "Cooper Jr., Franklin" Date: Tue, 3 May 2016 10:56:52 -0500 Subject: pwm: pwm-ti*: Get the clock from the PWMSS parent when using old bindings When using the old eCAP and ePWM bindings for AM335x and AM437x the clock can be retrieved from the PWMSS parent. Newer bindings will insure that this clock is provided via device tree. Therefore, update this driver to support the newer and older bindings. In the case of the older binding being used give a warning. Signed-off-by: Franklin S Cooper Jr [thierry.reding@gmail.com: rewrite slightly for readability] Signed-off-by: Thierry Reding --- drivers/pwm/pwm-tiecap.c | 9 +++++++++ drivers/pwm/pwm-tiehrpwm.c | 9 +++++++++ 2 files changed, 18 insertions(+) diff --git a/drivers/pwm/pwm-tiecap.c b/drivers/pwm/pwm-tiecap.c index 616af764a276..0dc5b97a7524 100644 --- a/drivers/pwm/pwm-tiecap.c +++ b/drivers/pwm/pwm-tiecap.c @@ -195,6 +195,7 @@ static const struct pwm_ops ecap_pwm_ops = { }; static const struct of_device_id ecap_of_match[] = { + { .compatible = "ti,am3352-ecap" }, { .compatible = "ti,am33xx-ecap" }, {}, }; @@ -202,6 +203,7 @@ MODULE_DEVICE_TABLE(of, ecap_of_match); static int ecap_pwm_probe(struct platform_device *pdev) { + struct device_node *np = pdev->dev.of_node; int ret; struct resource *r; struct clk *clk; @@ -213,6 +215,13 @@ static int ecap_pwm_probe(struct platform_device *pdev) return -ENOMEM; clk = devm_clk_get(&pdev->dev, "fck"); + if (IS_ERR(clk)) { + if (of_device_is_compatible(np, "ti,am33xx-ecap")) { + dev_warn(&pdev->dev, "Binding is obsolete.\n"); + clk = devm_clk_get(pdev->dev.parent, "fck"); + } + } + if (IS_ERR(clk)) { dev_err(&pdev->dev, "failed to get clock\n"); return PTR_ERR(clk); diff --git a/drivers/pwm/pwm-tiehrpwm.c b/drivers/pwm/pwm-tiehrpwm.c index 6a41e66015b6..c791720d27b0 100644 --- a/drivers/pwm/pwm-tiehrpwm.c +++ b/drivers/pwm/pwm-tiehrpwm.c @@ -426,6 +426,7 @@ static const struct pwm_ops ehrpwm_pwm_ops = { }; static const struct of_device_id ehrpwm_of_match[] = { + { .compatible = "ti,am3352-ehrpwm" }, { .compatible = "ti,am33xx-ehrpwm" }, {}, }; @@ -433,6 +434,7 @@ MODULE_DEVICE_TABLE(of, ehrpwm_of_match); static int ehrpwm_pwm_probe(struct platform_device *pdev) { + struct device_node *np = pdev->dev.of_node; int ret; struct resource *r; struct clk *clk; @@ -444,6 +446,13 @@ static int ehrpwm_pwm_probe(struct platform_device *pdev) return -ENOMEM; clk = devm_clk_get(&pdev->dev, "fck"); + if (IS_ERR(clk)) { + if (of_device_is_compatible(np, "ti,am33xx-ecap")) { + dev_warn(&pdev->dev, "Binding is obsolete.\n"); + clk = devm_clk_get(pdev->dev.parent, "fck"); + } + } + if (IS_ERR(clk)) { dev_err(&pdev->dev, "failed to get clock\n"); return PTR_ERR(clk); -- cgit v1.2.3 From 2b77487f2e8ff7e6496a7f5a08839de9bbb39ab3 Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Fri, 10 Jun 2016 15:49:53 +0200 Subject: pwm: Remove gratuitous blank line Commit 5ec803edcb70 ("pwm: Add core infrastructure to allow atomic updates") introduced this double blank line by mistake. Signed-off-by: Thierry Reding --- include/linux/pwm.h | 1 - 1 file changed, 1 deletion(-) diff --git a/include/linux/pwm.h b/include/linux/pwm.h index fd1092729ed6..83d8bcb7e1de 100644 --- a/include/linux/pwm.h +++ b/include/linux/pwm.h @@ -408,7 +408,6 @@ static inline void pwm_disable(struct pwm_device *pwm) pwm_apply_state(pwm, &state); } - /* PWM provider APIs */ int pwm_set_chip_data(struct pwm_device *pwm, void *data); void *pwm_get_chip_data(struct pwm_device *pwm); -- cgit v1.2.3 From cd9b518b98d3e989f523e63b2ffda78467a3679e Mon Sep 17 00:00:00 2001 From: Alexandre Belloni Date: Mon, 27 Jun 2016 00:03:03 +0200 Subject: rtc: v3020: move rtc-v3020.h to platform_data rtc-v3020.h belongs to include/linux/platform_data/ Acked-by: Robert Jarzmik Signed-off-by: Alexandre Belloni --- arch/arm/mach-pxa/cm-x270.c | 2 +- arch/arm/mach-pxa/cm-x300.c | 2 +- arch/arm/mach-pxa/em-x270.c | 2 +- drivers/rtc/rtc-v3020.c | 2 +- include/linux/platform_data/rtc-v3020.h | 41 +++++++++++++++++++++++++++++++++ include/linux/rtc-v3020.h | 41 --------------------------------- 6 files changed, 45 insertions(+), 45 deletions(-) create mode 100644 include/linux/platform_data/rtc-v3020.h delete mode 100644 include/linux/rtc-v3020.h diff --git a/arch/arm/mach-pxa/cm-x270.c b/arch/arm/mach-pxa/cm-x270.c index fa5f51d633a3..be4a66166d61 100644 --- a/arch/arm/mach-pxa/cm-x270.c +++ b/arch/arm/mach-pxa/cm-x270.c @@ -14,7 +14,7 @@ #include #include -#include +#include #include