summaryrefslogtreecommitdiffstats
path: root/drivers
AgeCommit message (Collapse)AuthorFilesLines
2018-01-26i2c: mv64xxx: Remove useless test before clk_disable_unprepareGregory CLEMENT1-6/+2
clk_disable_unprepare() already checks that the clock pointer is valid. No need to test it before calling it. Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-26i2c: mxs: use true and false for boolean valuesGustavo A. R. Silva1-2/+2
Assign true or false to boolean variables instead of an integer value. This issue was detected with the help of Coccinelle. Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-26i2c: meson: update doc description to fix build warningsYixun Lan1-1/+1
Add description for 'data' parameter and drop unused 'irq' memeber. Here is the warnings: drivers/i2c/busses/i2c-meson.c:103: warning: No description found for parameter 'data' drivers/i2c/busses/i2c-meson.c:103: warning: Excess struct member 'irq' description in 'meson_i2c' Reported-by: Wolfram Sang <wsa@the-dreams.de> Signed-off-by: Yixun Lan <yixun.lan@amlogic.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-24i2c: meson: add configurable divider factorsJian Hu1-4/+28
This patch try to add support for I2C controller in Meson-AXG SoC, Due to the IP changes between I2C controller, we need to introduce a compatible data to make the divider factor configurable. Reviewed-by: Neil Armstrong <narmstrong@baylibre.com> Signed-off-by: Jian Hu <jian.hu@amlogic.com> Signed-off-by: Yixun Lan <yixun.lan@amlogic.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-24i2c: imx-lpi2c: add runtime pm supportFugang Duan1-17/+51
Add runtime pm support to dynamically manage the clock to avoid enable/disable clock in frequently that can improve the i2c bus transfer performance. And use pm_runtime_force_suspend/resume() instead of lpi2c_imx_suspend/resume(). Signed-off-by: Fugang Duan <fugang.duan@nxp.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-24i2c: rcar: fix some trivial typos in commentsWolfram Sang1-3/+3
Nothing big, but they get annoying after a while ;) Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-24i2c: davinci: fix the cpufreq transitionBartosz Golaszewski1-8/+4
i2c_davinci_cpufreq_transition() is implemented in a way that will block if it ever gets called while no transfer is in progress. Not only that, but reinit_completion() is never called for xfr_complete. Use the fact that cpufreq uses an srcu_notifier (running in process context) for transitions and that the bus_lock is taken during the call to master_xfer() and simplify the code by removing the transfer completion entirely and protecting i2c_davinci_cpufreq_transition() with i2c_lock/unlock_adapter(). Reported-by: David Lechner <david@lechnology.com> Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com> Reviewed-by: Sekhar Nori <nsekhar@ti.com> Tested-by: David Lechner <david@lechnology.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-18i2c: rk3x: add proper kerneldoc headerWolfram Sang1-0/+1
gcc noticed the kerneldoc was wrongly formatted. Fix it! drivers/i2c/busses/i2c-rk3x.c:164: warning: Cannot understand * @grf_offset: ... on line 164 - I thought it was a doc line Signed-off-by: Wolfram Sang <wsa@the-dreams.de> Reviewed-by: Heiko Stuebner <heiko@sntech.de>
2018-01-18i2c: rk3x: account for const type of of_device_id.dataJulia Lawall1-8/+8
This driver creates a number of const structures that it stores in the data field of an of_device_id array. The data field of an of_device_id structure has type const void *, so there is no need for a const-discarding cast when putting const values into such a structure. Furthermore, adding const to the declaration of the location that receives a const value from such a field ensures that the compiler will continue to check that the value is not modified. The const-discarding cast on the extraction from the data field is thus no longer needed. Done using Coccinelle. Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> Reviewed-by: Heiko Stuebner <heiko@sntech.de> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-18i2c: acorn: remove outdated path from file headerWolfram Sang1-3/+1
That path has gone away for a long time. Move the HW name upwards for a proper header. Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-18i2c: acorn: add MODULE_LICENSE tagArnd Bergmann1-0/+4
As of v4.15, Kbuild warns about missing MODULE_LICENSE tags: WARNING: modpost: missing MODULE_LICENSE() in drivers/i2c/busses/i2c-acorn.o This adds a license, author and description tag, matching the comment at the start of the acorn i2c driver. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-16i2c: rcar: implement bus recoveryWolfram Sang1-2/+52
We can force levels of SCL and SDA, so we can use that for bus recovery. Note that we cannot read SDA back, because we will only get the internal state of the bus free detection. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-16i2c: send STOP after successful bus recoveryWolfram Sang1-0/+12
If we managed to get a client release SDA again, send a STOP afterwards to make sure we have a consistent state on the bus again. Tested-by: Phil Reid <preid@electromag.com.au> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-16i2c: ensure SDA is released in recovery if SDA is controllableWolfram Sang1-0/+2
If we have a function to control SDA, we should ensure that SDA is not held down by us. So, release the GPIO in this case. Tested-by: Phil Reid <preid@electromag.com.au> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-16i2c: add 'set_sda' to bus_recovery_infoWolfram Sang1-1/+10
This will be needed when we want to create STOP conditions, too, later. Create the needed fields and populate them for the GPIO case if the GPIO is set to output. Tested-by: Phil Reid <preid@electromag.com.au> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Reviewed-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: exynos5: change internal transmission timeout to 100msAndrzej Hajda1-1/+1
Exynos-I2C uses default timeout of 1 second for the whole transaction, including re-transmissions due to arbitration lost errors (-EAGAIN). To allow re-transmissions driver's internal timeout should be significantly lower, 100ms seems to be good candidate. Signed-off-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: ismt: 16-byte align the DMA buffer addressRadu Rendec1-18/+19
Use only a portion of the data buffer for DMA transfers, which is always 16-byte aligned. This makes the DMA buffer address 16-byte aligned and compensates for spurious hardware parity errors that may appear when the DMA buffer address is not 16-byte aligned. The data buffer is enlarged in order to accommodate any possible 16-byte alignment offset and changes the DMA code to only use a portion of the data buffer, which is 16-byte aligned. The symptom of the hardware issue is the same as the one addressed in v3.12-rc2-5-gbf41691 and manifests by transfers failing with EIO, with bit 9 being set in the ERRSTS register. Signed-off-by: Radu Rendec <radu.rendec@gmail.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: mpc: always determine I2C clock prescaler at runtimeArseny Solokha1-22/+15
Remove the facility for setting the prescaler value at compile time entirely. It was only used for two SoCs, duplicating the actual value for one of them and setting sometimes bogus value for another. Make all MPC8xxx SoCs obtain their actual I2C clock prescaler from a single place in the code. Changes from v2: - left Device Tree compatibles in place Signed-off-by: Arseny Solokha <asolokha@kb.kras.ru> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: mpc: fix PORDEVSR2 mask for MPC8533/44Arseny Solokha1-2/+6
According to the reference manuals for the corresponding SoCs, SEC frequency ratio configuration is indicated by bit 26 of the POR Device Status Register 2. Consequently, SEC_CFG bit should be tested by mask 0x20, not 0x80. Testing the wrong bit leads to selection of wrong I2C clock prescaler on those SoCs. Signed-off-by: Arseny Solokha <asolokha@kb.kras.ru> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: mpc: unify obtaining the MPC8533/44 I2C clock prescaler w/ MPC8xxxArseny Solokha1-3/+9
Commit 8ce795cb0c6b ("i2c: mpc: assign the correct prescaler from SVR") introduced the common helper function for obtaining the actual clock prescaler value for MPC85xx. However, getting the prescaler for MPC8544 which depends on the SEC frequency ratio on this platform, has been always performed separately based on the corresponding Device Tree configuration. Move special handling of MPC8544 into that common helper. Make it dependent on the SoC version and not on Device Tree compatible node, as is the case with all other SoCs. Handle MPC8533 the same way which is similar to MPC8544 in this regard, according to AN2919 "Determining the I2C Frequency Divider Ratio for SCL". Signed-off-by: Arseny Solokha <asolokha@kb.kras.ru> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: mpc: get MPC8xxx I2C clock prescaler before using it in calculationsArseny Solokha1-6/+6
Obtaining the actual I2C clock prescaler value in mpc_i2c_setup_8xxx() only happens when the clock parameter is set to something other than MPC_I2C_CLOCK_LEGACY. When the clock parameter is exactly MPC_I2C_CLOCK_LEGACY, the prescaler parameter is used in arithmetic division as provided by the caller, resulting in a division by zero for the majority of processors supported by the module. Avoid division by zero by obtaining the actual I2C clock prescaler in mpc_i2c_setup_8xxx() unconditionally regardless of the passed clock value. Signed-off-by: Arseny Solokha <asolokha@kb.kras.ru> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: ismt: dump registers at the end of transactionsRadu Rendec1-0/+2
This patch dumps general and master registers at the end of transactions when debugging is enabled. Previously, registers were only dumped before submitting new descriptors (at the beginning of transactions). This helps debugging if some registers change as result of a failed transaction (e.g. bits are set in the ERRSTS general register). Signed-off-by: Radu Rendec <radu.rendec@gmail.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: sh_mobile: let r8a7790 (R-Car H2) use the new formulaWolfram Sang1-1/+1
Make use of the new formula for more precise bus frequencies. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: sh_mobile: add new frequency calculation for later SoCWolfram Sang1-1/+18
The formula to generate the desired bus speeds has changed a little over time. Implement the new formula and allow drivers to opt-in by changing to this new config set. Ensure in probe that we don't divide by zero. The returned values on a R-Car H2 (r8a7790/Lager board) match the suggested values in the datasheet. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: sh_mobile: add helper to check frequency calculationsWolfram Sang1-22/+27
Because we will add a second formula soon, put the sanity checks for the computed results into a separate function. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: sh_mobile: let RuntimePM do the clock handlingWolfram Sang1-18/+17
Start RuntimePM a bit earlier, so we can use it to enable the clock during probe for frequency calculations. Make sure it is enabled before calling setup(). Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: sh_mobile: require setup callbackWolfram Sang1-9/+6
Require the setup callback and move the frequency calculation into it. This is in preparation for supporting multiple formulas. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: sh_mobile: allow setup callback to return errnoWolfram Sang1-4/+9
The setup callback will be more generic and, thus, need to be able to return error codes. Change the return type to 'int' for that. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-15i2c: sh_mobile: move type detection upwardsWolfram Sang1-6/+4
For refactoring reasons, we will need this information before the setup callback. Also, simplify the comment to a oneliner. Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-04i2c: gpio: Enable working over slow can_sleep GPIOsJan Kundrát1-4/+7
"Slow" GPIOs (usually those connected over an SPI or an I2C bus) are, well, slow in their operation. It is generally a good idea to avoid using them for time-critical operation, but sometimes the hardware just sucks, and the software has to cope. In addition to that, the I2C bus itself does not actually define any strict timing limits; the bus is free to go all the way down to DC. The timeouts (and therefore the slowest acceptable frequency) are present only in SMBus. The `can_sleep` is IMHO a wrong concept to use here. My SPI-to-quad-UART chip (MAX14830) is connected via a 26MHz SPI bus, and it happily drives SCL at 200kHz (5µs pulses) during my benchmarks. That's faster than the maximal allowed speed of the traditional I2C. The previous version of this code did not really block operation over slow GPIO pins, anyway. Instead, it just resorted to printing a warning with a backtrace each time a GPIO pin was accessed, thereby slowing things down even more. Finally, it's not just me. A similar patch was originally submitted in 2015 [1]. [1] https://patchwork.ozlabs.org/patch/450956/ Signed-off-by: Jan Kundrát <jan.kundrat@cesnet.cz> Acked-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-04i2c: mediatek: Enable i2c module clock before i2c registers access.Jun Gao1-0/+9
Make sure i2c module clock has been enabled before i2c registers access. Signed-off-by: Jun Gao <jun.gao@mediatek.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-04i2c: mediatek: Add i2c compatible for MediaTek MT2712Jun Gao1-4/+27
Add i2c compatible for MT2712. Compare to MT8173 i2c controller, internal divider of i2c source clock need to be configured for MT2712 i2c speed calculation. Signed-off-by: Jun Gao <jun.gao@mediatek.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
2018-01-03Merge tag 'at24-4.16-updates-for-wolfram' of ↵Wolfram Sang753-5516/+8537
git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux into i2c/for-4.16 "AT24 updates for 4.16 merge window The driver has been converted to using regmap instead of raw i2c and smbus calls which shrank the code significantly. Device tree binding document has been cleaned up. Device tree support in the driver has been improved and we now support all at24 models as well as two new DT properties (no-read-rollover and wp-gpios). We no longer user unreadable magic values for driver data as the way it was implemented caused problems for some EEPROM models - we switched to regular structs. Aside from that, there's a bunch of coding style fixes and minor improvements all over the place."
2018-01-02eeprom: at24: extend the list of chips supported in DTBartosz Golaszewski1-0/+9
Add all supported at24 variants to the of_match table. Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl> Reviewed-by: Javier Martinez Canillas <javierm@redhat.com>
2018-01-01eeprom: at24: add support for the write-protect pinBartosz Golaszewski1-0/+11
AT24 EEPROMs have a write-protect pin, which - when pulled high - inhibits writes to the upper quadrant of memory (although it has been observed that on some chips it disables writing to the entire memory range). On some boards, this pin is connected to a GPIO and pulled high by default, which forces the user to manually change its state before writing. On linux this means that we either need to hog the line all the time, or set the GPIO value before writing from outside of the at24 driver. Make the driver check if the write-protect GPIO was defined in the device tree and pull it low whenever writing to the EEPROM. Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
2018-01-01eeprom: at24: remove temporary fix for at24mac402 sizeSven Van Asbroeck1-10/+0
The chip size passed via devicetree, i2c, or acpi device ids is now no longer limited to a power of two. So the temporary fix can be removed. Signed-off-by: Sven Van Asbroeck <svendev@arcx.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: convert magic numbers to structsSven Van Asbroeck1-121/+100
Fundamental properties such as capacity and page size differ among at24-type chips. But these chips do not have an id register, so this can't be discovered at runtime. Traditionally, at24-type eeprom properties were determined in two ways: - by passing a 'struct at24_platform_data' via platform_data, or - by naming the chip type in the devicetree, which passes a 'magic number' to probe(), which is then converted to a 'struct at24_platform_data'. Recently a bug was discovered because the magic number rounds down all chip sizes to the lowest power of two. This was addressed by a work-around commit 5478e478eee3 ("eeprom: at24: correctly set the size for at24mac402"), with the wish that magic numbers should over time be converted to structs. This patch replaces the magic numbers with 'struct at24_chip_data'. Signed-off-by: Sven Van Asbroeck <svendev@arcx.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: code shrinkBartosz Golaszewski1-17/+6
A regmap_config struct is pretty big and declaring two of them statically just to tweak the reg_bits value adds unnecessary bloat. Declare the regmap config locally in at24_probe() instead. Bloat-o-meter output for ARM: add/remove: 0/2 grow/shrink: 1/0 up/down: 4/-272 (-268) Function old new delta at24_probe 1560 1564 +4 regmap_config_8 136 - -136 regmap_config_16 136 - -136 Total: Before=7012, After=6744, chg -3.82% Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: use a common prefix for all symbols in at24.cBartosz Golaszewski1-16/+18
There are a couple symbols defined in the driver source file which are missing the at24_ prefix. This patch fixes that. For module params: use module_param_named() in order to not break userspace. Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: fix coding style issuesBartosz Golaszewski1-9/+9
Fix issues reported by checkpatch for at24.c. Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: support eeproms that do not auto-rollover readsSven Van Asbroeck1-12/+27
Some multi-address eeproms in the at24 family may not automatically roll-over reads to the next slave address. On those eeproms, reads that straddle slave boundaries will not work correctly. Solution: Mark such eeproms with a flag that prevents reads straddling slave boundaries. Add the AT24_FLAG_NO_RDROL flag to the eeprom entry in the device_id table, or add 'no-read-rollover' to the eeprom devicetree entry. Note that I have not personally enountered an at24 chip that does not support read rollovers. They may or may not exist. However, my hardware requires this functionality because of a quirk. Signed-off-by: Sven Van Asbroeck <svendev@arcx.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: remove now unneeded smbus-related codeHeiner Kallweit1-48/+8
Remove remaining now unneeded code dealing with SMBUS details. Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: remove old read functionsHeiner Kallweit1-186/+0
Remove the old and now unused read functions. Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: add regmap-based read functionHeiner Kallweit1-1/+56
Add regmap-based read function and instead of using three different read functions (standard, mac, serial) use just one and factor out the read offset adjustment for mac and serial to at24_adjust_read_offset. Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: remove old write functionsHeiner Kallweit1-117/+0
Remove the old and now unused write functions. Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: add regmap-based write functionHeiner Kallweit1-1/+26
Add a regmap-based write function. Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: change at24_translate_offset return typeHeiner Kallweit1-10/+24
Change return type of at24_translate_offset to *at24_client to make member regmap accessible for subsequent patches of this series. Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2018-01-01eeprom: at24: add basic regmap_i2c supportHeiner Kallweit2-13/+45
This patch adds basic regmap support to be used by subsequent patches of this series. Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: Bartosz Golaszewski <brgl@bgdev.pl>
2017-12-31Merge branch 'x86/urgent' of ↵Linus Torvalds1-6/+2
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull x86 fixes from Thomas Gleixner: "A couple of fixlets for x86: - Fix the ESPFIX double fault handling for 5-level pagetables - Fix the commandline parsing for 'apic=' on 32bit systems and update documentation - Make zombie stack traces reliable - Fix kexec with stack canary - Fix the delivery mode for APICs which was missed when the x86 vector management was converted to single target delivery. Caused a regression due to the broken hardware which ignores affinity settings in lowest prio delivery mode. - Unbreak modules when AMD memory encryption is enabled - Remove an unused parameter of prepare_switch_to" * 'x86/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/apic: Switch all APICs to Fixed delivery mode x86/apic: Update the 'apic=' description of setting APIC driver x86/apic: Avoid wrong warning when parsing 'apic=' in X86-32 case x86-32: Fix kexec with stack canary (CONFIG_CC_STACKPROTECTOR) x86: Remove unused parameter of prepare_switch_to x86/stacktrace: Make zombie stack traces reliable x86/mm: Unbreak modules that use the DMA API x86/build: Make isoimage work on Debian x86/espfix/64: Fix espfix double-fault handling on 5-level systems
2017-12-31Merge branch 'irq-urgent-for-linus' of ↵Linus Torvalds12-21/+44
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip Pull irq fixes from Thomas Gleixner: "A rather large update after the kaisered maintainer finally found time to handle regression reports. - The larger part addresses a regression caused by the x86 vector management rework. The reservation based model does not work reliably for MSI interrupts, if they cannot be masked (yes, yet another hw engineering trainwreck). The reason is that the reservation mode assigns a dummy vector when the interrupt is allocated and switches to a real vector when the interrupt is requested. If the MSI entry cannot be masked then the initialization might raise an interrupt before the interrupt is requested, which ends up as spurious interrupt and causes device malfunction and worse. The fix is to exclude MSI interrupts which do not support masking from reservation mode and assign a real vector right away. - Extend the extra lockdep class setup for nested interrupts with a class for the recently added irq_desc::request_mutex so lockdep can differeniate and does not emit false positive warnings. - A ratelimit guard for the bad irq printout so in case a bad irq comes back immediately the system does not drown in dmesg spam" * 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: genirq/msi, x86/vector: Prevent reservation mode for non maskable MSI genirq/irqdomain: Rename early argument of irq_domain_activate_irq() x86/vector: Use IRQD_CAN_RESERVE flag genirq: Introduce IRQD_CAN_RESERVE flag genirq/msi: Handle reactivation only on success gpio: brcmstb: Make really use of the new lockdep class genirq: Guard handle_bad_irq log messages kernel/irq: Extend lockdep class for request mutex