From ae9b9403644f3ecc76867af042e7e1cfd5c099d0 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 30 Oct 2008 17:43:57 +0100 Subject: AMD IOMMU: fix detection of NP capable IOMMUs This patch changes the code to use IOMMU_CAP_NPCACHE as a shift and not as a mask. Signed-off-by: Joerg Roedel --- arch/x86/kernel/amd_iommu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 3b346c6f5514..38e88d40ab10 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -50,7 +50,7 @@ static int dma_ops_unity_map(struct dma_ops_domain *dma_dom, /* returns !0 if the IOMMU is caching non-present entries in its TLB */ static int iommu_has_npcache(struct amd_iommu *iommu) { - return iommu->cap & IOMMU_CAP_NPCACHE; + return iommu->cap & (1UL << IOMMU_CAP_NPCACHE); } /**************************************************************************** -- cgit v1.2.3 From 73557af5bf32c3db973050de1fb73423e8fc873e Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Fri, 31 Oct 2008 13:59:49 -0400 Subject: x86, voyager: fix smp_intr_init() compile breakage Impact: fix x86/Voyager build Looks like this became static on the rest of x86. Fix it up by adding an external definition to mach-voyager/setup.c Signed-off-by: Ingo Molnar --- arch/x86/include/asm/voyager.h | 1 + arch/x86/mach-voyager/setup.c | 2 +- arch/x86/mach-voyager/voyager_smp.c | 2 +- 3 files changed, 3 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/voyager.h b/arch/x86/include/asm/voyager.h index 9c811d2e6f91..b3e647307625 100644 --- a/arch/x86/include/asm/voyager.h +++ b/arch/x86/include/asm/voyager.h @@ -520,6 +520,7 @@ extern void voyager_restart(void); extern void voyager_cat_power_off(void); extern void voyager_cat_do_common_interrupt(void); extern void voyager_handle_nmi(void); +extern void voyager_smp_intr_init(void); /* Commands for the following are */ #define VOYAGER_PSI_READ 0 #define VOYAGER_PSI_WRITE 1 diff --git a/arch/x86/mach-voyager/setup.c b/arch/x86/mach-voyager/setup.c index 6bbdd633864c..a580b9562e76 100644 --- a/arch/x86/mach-voyager/setup.c +++ b/arch/x86/mach-voyager/setup.c @@ -27,7 +27,7 @@ static struct irqaction irq2 = { void __init intr_init_hook(void) { #ifdef CONFIG_SMP - smp_intr_init(); + voyager_smp_intr_init(); #endif setup_irq(2, &irq2); diff --git a/arch/x86/mach-voyager/voyager_smp.c b/arch/x86/mach-voyager/voyager_smp.c index 7f4c6af14351..0e331652681e 100644 --- a/arch/x86/mach-voyager/voyager_smp.c +++ b/arch/x86/mach-voyager/voyager_smp.c @@ -1258,7 +1258,7 @@ static void handle_vic_irq(unsigned int irq, struct irq_desc *desc) #define QIC_SET_GATE(cpi, vector) \ set_intr_gate((cpi) + QIC_DEFAULT_CPI_BASE, (vector)) -void __init smp_intr_init(void) +void __init voyager_smp_intr_init(void) { int i; -- cgit v1.2.3 From 70de9a97049e0ba79dc040868564408d5ce697f9 Mon Sep 17 00:00:00 2001 From: Alok Kataria Date: Mon, 3 Nov 2008 11:18:47 -0800 Subject: x86: don't use tsc_khz to calculate lpj if notsc is passed Impact: fix udelay when "notsc" boot parameter is passed With notsc passed on commandline, tsc may not be used for udelays, make sure that we do not use tsc_khz to calculate the lpj value in such cases. Reported-by: Bartlomiej Zolnierkiewicz Signed-off-by: Alok N Kataria Cc: Signed-off-by: Ingo Molnar --- arch/x86/kernel/tsc.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 62348e4fd8d1..2ef80e301925 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -813,10 +813,6 @@ void __init tsc_init(void) cpu_khz = calibrate_cpu(); #endif - lpj = ((u64)tsc_khz * 1000); - do_div(lpj, HZ); - lpj_fine = lpj; - printk("Detected %lu.%03lu MHz processor.\n", (unsigned long)cpu_khz / 1000, (unsigned long)cpu_khz % 1000); @@ -836,6 +832,10 @@ void __init tsc_init(void) /* now allow native_sched_clock() to use rdtsc */ tsc_disabled = 0; + lpj = ((u64)tsc_khz * 1000); + do_div(lpj, HZ); + lpj_fine = lpj; + use_tsc_delay(); /* Check and install the TSC clocksource */ dmi_check_system(bad_tsc_dmi_table); -- cgit v1.2.3 From 9fcd18c9e63e325dbd2b4c726623f760788d5aa8 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Wed, 5 Nov 2008 16:52:08 +0100 Subject: sched: re-tune balancing Impact: improve wakeup affinity on NUMA systems, tweak SMP systems Given the fixes+tweaks to the wakeup-buddy code, re-tweak the domain balancing defaults on NUMA and SMP systems. Turn on SD_WAKE_AFFINE which was off on x86 NUMA - there's no reason why we would not want to have wakeup affinity across nodes as well. (we already do this in the standard NUMA template.) lat_ctx on a NUMA box is particularly happy about this change: before: | phoenix:~/l> ./lat_ctx -s 0 2 | "size=0k ovr=2.60 | 2 5.70 after: | phoenix:~/l> ./lat_ctx -s 0 2 | "size=0k ovr=2.65 | 2 2.07 a 2.75x speedup. pipe-test is similarly happy about it too: | phoenix:~/sched-tests> ./pipe-test | 18.26 usecs/loop. | 14.70 usecs/loop. | 14.38 usecs/loop. | 10.55 usecs/loop. # +WAKE_AFFINE on domain0+domain1 | 8.63 usecs/loop. | 8.59 usecs/loop. | 9.03 usecs/loop. | 8.94 usecs/loop. | 8.96 usecs/loop. | 8.63 usecs/loop. Also: - disable SD_BALANCE_NEWIDLE on NUMA and SMP domains (keep it for siblings) - enable SD_WAKE_BALANCE on SMP domains Sysbench+postgresql improves all around the board, quite significantly: .28-rc3-11474e2c .28-rc3-11474e2c-tune ------------------------------------------------- 1: 571 688 +17.08% 2: 1236 1206 -2.55% 4: 2381 2642 +9.89% 8: 4958 5164 +3.99% 16: 9580 9574 -0.07% 32: 7128 8118 +12.20% 64: 7342 8266 +11.18% 128: 7342 8064 +8.95% 256: 7519 7884 +4.62% 512: 7350 7731 +4.93% ------------------------------------------------- SUM: 55412 59341 +6.62% So it's a win both for the runup portion, the peak area and the tail. Signed-off-by: Ingo Molnar --- arch/x86/include/asm/topology.h | 7 ++++--- include/linux/topology.h | 4 ++-- 2 files changed, 6 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index 90ac7718469a..4850e4b02b61 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -154,7 +154,7 @@ extern unsigned long node_remap_size[]; #endif -/* sched_domains SD_NODE_INIT for NUMAQ machines */ +/* sched_domains SD_NODE_INIT for NUMA machines */ #define SD_NODE_INIT (struct sched_domain) { \ .min_interval = 8, \ .max_interval = 32, \ @@ -169,8 +169,9 @@ extern unsigned long node_remap_size[]; .flags = SD_LOAD_BALANCE \ | SD_BALANCE_EXEC \ | SD_BALANCE_FORK \ - | SD_SERIALIZE \ - | SD_WAKE_BALANCE, \ + | SD_WAKE_AFFINE \ + | SD_WAKE_BALANCE \ + | SD_SERIALIZE, \ .last_balance = jiffies, \ .balance_interval = 1, \ } diff --git a/include/linux/topology.h b/include/linux/topology.h index 2158fc0d5a56..34a7ee0ebed2 100644 --- a/include/linux/topology.h +++ b/include/linux/topology.h @@ -146,10 +146,10 @@ void arch_update_cpu_topology(void); .wake_idx = 1, \ .forkexec_idx = 1, \ .flags = SD_LOAD_BALANCE \ - | SD_BALANCE_NEWIDLE \ - | SD_BALANCE_FORK \ | SD_BALANCE_EXEC \ + | SD_BALANCE_FORK \ | SD_WAKE_AFFINE \ + | SD_WAKE_BALANCE \ | BALANCE_FOR_PKG_POWER,\ .last_balance = jiffies, \ .balance_interval = 1, \ -- cgit v1.2.3 From c78d0cf2925bffae8a6f00e7d9b8e971b0392edd Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Wed, 5 Nov 2008 12:04:46 +0000 Subject: x86: don't allow nr_irqs > NR_IRQS Impact: fix boot hang on 32-bit systems with more than 224 IO-APIC pins On some 32-bit systems with a lot of IO-APICs probe_nr_irqs() can return a value larger than NR_IRQS. This will lead to probe_irq_on() overrunning the irq_desc array. I hit this when running net-next-2.6 (close to 2.6.28-rc3) on a Supermicro dual Xeon system. NR_IRQS is 224 but probe_nr_irqs() detects 5 IOAPICs and returns 240. Here are the log messages: Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x01] address[0xfec00000] gsi_base[0]) Tue Nov 4 16:53:47 2008 IOAPIC[0]: apic_id 1, version 32, address 0xfec00000, GSI 0-23 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x02] address[0xfec81000] gsi_base[24]) Tue Nov 4 16:53:47 2008 IOAPIC[1]: apic_id 2, version 32, address 0xfec81000, GSI 24-47 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x03] address[0xfec81400] gsi_base[48]) Tue Nov 4 16:53:47 2008 IOAPIC[2]: apic_id 3, version 32, address 0xfec81400, GSI 48-71 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x04] address[0xfec82000] gsi_base[72]) Tue Nov 4 16:53:47 2008 IOAPIC[3]: apic_id 4, version 32, address 0xfec82000, GSI 72-95 Tue Nov 4 16:53:47 2008 ACPI: IOAPIC (id[0x05] address[0xfec82400] gsi_base[96]) Tue Nov 4 16:53:47 2008 IOAPIC[4]: apic_id 5, version 32, address 0xfec82400, GSI 96-119 Tue Nov 4 16:53:47 2008 ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 high edge) Tue Nov 4 16:53:47 2008 ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 high level) Tue Nov 4 16:53:47 2008 Enabling APIC mode: Flat. Using 5 I/O APICs Signed-off-by: Ben Hutchings Acked-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/kernel/io_apic.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/io_apic.c b/arch/x86/kernel/io_apic.c index b764d7429c61..7a3f2028e2eb 100644 --- a/arch/x86/kernel/io_apic.c +++ b/arch/x86/kernel/io_apic.c @@ -3611,6 +3611,8 @@ int __init probe_nr_irqs(void) /* something wrong ? */ if (nr < nr_min) nr = nr_min; + if (WARN_ON(nr > NR_IRQS)) + nr = NR_IRQS; return nr; } -- cgit v1.2.3 From 1b4897688011cd05e07f00dcfe6af3331eb36a3c Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Tue, 4 Nov 2008 14:10:13 -0800 Subject: x86: size NR_IRQS on 32-bit systems the same way as 64-bit Impact: make NR_IRQS big enough for system with lots of apic/pins If lots of IO_APIC's are there (or can be there), size the same way as 64-bit, depending on MAX_IO_APICS and NR_CPUS. This fixes the boot problem reported by Ben Hutchings on a 32-bit server with 5 IO-APICs and 240 IO-APIC pins. Signed-off-by: Yinghai Tested-by: Ben Hutchings Signed-off-by: Ingo Molnar --- arch/x86/include/asm/irq_vectors.h | 20 ++++++-------------- 1 file changed, 6 insertions(+), 14 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h index d843ed0e9b2e..503aadc4ad35 100644 --- a/arch/x86/include/asm/irq_vectors.h +++ b/arch/x86/include/asm/irq_vectors.h @@ -101,30 +101,22 @@ #define LAST_VM86_IRQ 15 #define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15) -#ifdef CONFIG_X86_64 +#if defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_PARAVIRT) && !defined(CONFIG_X86_VISWS) && !defined(CONFIG_X86_VOYAGER) # if NR_CPUS < MAX_IO_APICS # define NR_IRQS (NR_VECTORS + (32 * NR_CPUS)) # else # define NR_IRQS (NR_VECTORS + (32 * MAX_IO_APICS)) # endif -#elif !defined(CONFIG_X86_VOYAGER) +#elif defined(CONFIG_PARAVIRT) || defined(CONFIG_X86_VISWS) || defined(CONFIG_X86_VOYAGER) -# if defined(CONFIG_X86_IO_APIC) || defined(CONFIG_PARAVIRT) || defined(CONFIG_X86_VISWS) - -# define NR_IRQS 224 - -# else /* IO_APIC || PARAVIRT */ - -# define NR_IRQS 16 - -# endif +# define NR_IRQS 224 -#else /* !VISWS && !VOYAGER */ +#else /* IO_APIC || PARAVIRT */ -# define NR_IRQS 224 +# define NR_IRQS 16 -#endif /* VISWS */ +#endif /* Voyager specific defines */ /* These define the CPIs we use in linux */ -- cgit v1.2.3 From da85f865b1dcec0853c48b763ed312441ce0c7df Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Wed, 5 Nov 2008 13:37:27 -0600 Subject: x86: mention ACPI in top-level Kconfig menu Impact: clarify menuconfig text Mention ACPI in the top-level menu to give a clue as to where it lives. This matches what ia64 does. Signed-off-by: Bjorn Helgaas Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 6f20718d3156..5d6aa4013dc7 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1494,7 +1494,7 @@ config HAVE_ARCH_EARLY_PFN_TO_NID def_bool X86_64 depends on NUMA -menu "Power management options" +menu "Power management and ACPI options" depends on !X86_VOYAGER config ARCH_HIBERNATION_HEADER -- cgit v1.2.3 From 7db282fa67b58daff8a57f9e1c93d4474b5908ff Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 5 Nov 2008 23:36:48 -0800 Subject: x86: remove VISWS and PARAVIRT around NR_IRQS puzzle Impact: fix warning message when PARAVIRT is set in config Remove stale #ifdef components from our IRQ sizing logic. x86/Voyager is the only holdout. Signed-off-by: Yinghai Lu Signed-off-by: Ingo Molnar --- arch/x86/include/asm/irq_vectors.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h index 503aadc4ad35..0005adb0f941 100644 --- a/arch/x86/include/asm/irq_vectors.h +++ b/arch/x86/include/asm/irq_vectors.h @@ -101,18 +101,18 @@ #define LAST_VM86_IRQ 15 #define invalid_vm86_irq(irq) ((irq) < 3 || (irq) > 15) -#if defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_PARAVIRT) && !defined(CONFIG_X86_VISWS) && !defined(CONFIG_X86_VOYAGER) +#if defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_X86_VOYAGER) # if NR_CPUS < MAX_IO_APICS # define NR_IRQS (NR_VECTORS + (32 * NR_CPUS)) # else # define NR_IRQS (NR_VECTORS + (32 * MAX_IO_APICS)) # endif -#elif defined(CONFIG_PARAVIRT) || defined(CONFIG_X86_VISWS) || defined(CONFIG_X86_VOYAGER) +#elif defined(CONFIG_X86_VOYAGER) # define NR_IRQS 224 -#else /* IO_APIC || PARAVIRT */ +#else /* IO_APIC || VOYAGER */ # define NR_IRQS 16 -- cgit v1.2.3 From d6f0f39b7d05e62b347c4352d070e4afb3ade4b5 Mon Sep 17 00:00:00 2001 From: Suresh Siddha Date: Tue, 4 Nov 2008 13:53:04 -0800 Subject: x86: add smp_mb() before sending INVALIDATE_TLB_VECTOR Impact: fix rare x2apic hang On x86, x2apic mode accesses for sending IPI's don't have serializing semantics. If the IPI receivner refers(in lock-free fashion) to some memory setup by the sender, the need for smp_mb() before sending the IPI becomes critical in x2apic mode. Add the smp_mb() in native_flush_tlb_others() before sending the IPI. Signed-off-by: Suresh Siddha Signed-off-by: Ingo Molnar --- arch/x86/kernel/tlb_32.c | 6 ++++++ arch/x86/kernel/tlb_64.c | 5 +++++ 2 files changed, 11 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/kernel/tlb_32.c b/arch/x86/kernel/tlb_32.c index e00534b33534..f4049f3513b6 100644 --- a/arch/x86/kernel/tlb_32.c +++ b/arch/x86/kernel/tlb_32.c @@ -154,6 +154,12 @@ void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm, flush_mm = mm; flush_va = va; cpus_or(flush_cpumask, cpumask, flush_cpumask); + + /* + * Make the above memory operations globally visible before + * sending the IPI. + */ + smp_mb(); /* * We have to send the IPI only to * CPUs affected. diff --git a/arch/x86/kernel/tlb_64.c b/arch/x86/kernel/tlb_64.c index dcbf7a1159ea..8f919ca69494 100644 --- a/arch/x86/kernel/tlb_64.c +++ b/arch/x86/kernel/tlb_64.c @@ -182,6 +182,11 @@ void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm, f->flush_va = va; cpus_or(f->flush_cpumask, cpumask, f->flush_cpumask); + /* + * Make the above memory operations globally visible before + * sending the IPI. + */ + smp_mb(); /* * We have to send the IPI only to * CPUs affected. -- cgit v1.2.3 From 80be308dfa3798c7bad0fc81760b2faf83870e91 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 6 Nov 2008 14:59:05 +0100 Subject: AMD IOMMU: fix lazy IO/TLB flushing in unmap path Lazy flushing needs to take care of the unmap path too which is not yet implemented and leads to stale IO/TLB entries. This is fixed by this patch. Signed-off-by: Joerg Roedel --- arch/x86/kernel/amd_iommu.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 38e88d40ab10..4755bbc7ae5b 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -526,6 +526,9 @@ static void dma_ops_free_addresses(struct dma_ops_domain *dom, { address >>= PAGE_SHIFT; iommu_area_free(dom->bitmap, address, pages); + + if (address + pages >= dom->next_bit) + dom->need_flush = true; } /**************************************************************************** @@ -981,8 +984,10 @@ static void __unmap_single(struct amd_iommu *iommu, dma_ops_free_addresses(dma_dom, dma_addr, pages); - if (amd_iommu_unmap_flush) + if (amd_iommu_unmap_flush || dma_dom->need_flush) { iommu_flush_pages(iommu, dma_dom->domain.id, dma_addr, size); + dma_dom->need_flush = false; + } } /* -- cgit v1.2.3 From b9c3bfc24e1088d260de4091b2b41808c7398355 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 6 Nov 2008 12:05:40 +0000 Subject: x86: align DirectMap in /proc/meminfo Impact: right-align /proc/meminfo consistent with other fields When the split-LRU patches added Inactive(anon) and Inactive(file) lines to /proc/meminfo, all counts were moved two columns rightwards to fit in. Now move x86's DirectMap lines two columns rightwards to line up. Signed-off-by: Hugh Dickins Signed-off-by: Ingo Molnar --- arch/x86/mm/pageattr.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index f1dc1b75d166..e89d24815f26 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -67,18 +67,18 @@ static void split_page_count(int level) void arch_report_meminfo(struct seq_file *m) { - seq_printf(m, "DirectMap4k: %8lu kB\n", + seq_printf(m, "DirectMap4k: %8lu kB\n", direct_pages_count[PG_LEVEL_4K] << 2); #if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) - seq_printf(m, "DirectMap2M: %8lu kB\n", + seq_printf(m, "DirectMap2M: %8lu kB\n", direct_pages_count[PG_LEVEL_2M] << 11); #else - seq_printf(m, "DirectMap4M: %8lu kB\n", + seq_printf(m, "DirectMap4M: %8lu kB\n", direct_pages_count[PG_LEVEL_2M] << 12); #endif #ifdef CONFIG_X86_64 if (direct_gbpages) - seq_printf(m, "DirectMap1G: %8lu kB\n", + seq_printf(m, "DirectMap1G: %8lu kB\n", direct_pages_count[PG_LEVEL_1G] << 20); #endif } -- cgit v1.2.3 From 8d00450d296dedec9ada38d43b83e79cca6fd5a3 Mon Sep 17 00:00:00 2001 From: Eduardo Habkost Date: Tue, 4 Nov 2008 12:52:44 -0200 Subject: Revert "x86: default to reboot via ACPI" This reverts commit c7ffa6c26277b403920e2255d10df849bd613380. the assumptio of this change was that this would not break any existing machine. Andrey Borzenkov reported troubles with the ACPI reboot method: the system would hang on reboot, necessiating a power cycle. Probably more systems are affected as well. Also, there are patches queued up for v2.6.29 to disable virtualization on emergency_restart() - which was the original motivation of this change. Reported-by: Andrey Borzenkov Bisected-by: Andrey Borzenkov Signed-off-by: Eduardo Habkost Acked-by: Avi Kivity Signed-off-by: Ingo Molnar --- arch/x86/kernel/reboot.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index f4c93f1cfc19..724adfc63cb9 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -29,11 +29,7 @@ EXPORT_SYMBOL(pm_power_off); static const struct desc_ptr no_idt = {}; static int reboot_mode; -/* - * Keyboard reset and triple fault may result in INIT, not RESET, which - * doesn't work when we're in vmx root mode. Try ACPI first. - */ -enum reboot_type reboot_type = BOOT_ACPI; +enum reboot_type reboot_type = BOOT_KBD; int reboot_force; #if defined(CONFIG_X86_32) && defined(CONFIG_SMP) -- cgit v1.2.3 From 47cb2ed9df2789fc4a3fe1201e475078f93c4839 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Thu, 6 Nov 2008 13:48:24 -0800 Subject: x86, xen: fix use of pgd_page now that it really does return a page Impact: fix 32-bit Xen guest boot crash On 32-bit PAE, pud_page, for no good reason, didn't really return a struct page *. Since Jan Beulich's fix "i386/PAE: fix pud_page()", pud_page does return a struct page *. Because PAE has 3 pagetable levels, the pud level is folded into the pgd level, so pgd_page() is the same as pud_page(), and now returns a struct page *. Update the xen/mmu.c code which uses pgd_page() accordingly. Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Ingo Molnar --- arch/x86/xen/mmu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index aba77b2b7d18..49697d86c6a5 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -874,7 +874,7 @@ static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd) #else /* CONFIG_X86_32 */ #ifdef CONFIG_X86_PAE /* Need to make sure unshared kernel PMD is pinnable */ - xen_pin_page(mm, virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), + xen_pin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]), PT_PMD); #endif xen_do_pin(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(pgd))); @@ -991,7 +991,7 @@ static void __xen_pgd_unpin(struct mm_struct *mm, pgd_t *pgd) #ifdef CONFIG_X86_PAE /* Need to make sure unshared kernel PMD is unpinned */ - xen_unpin_page(mm, virt_to_page(pgd_page(pgd[pgd_index(TASK_SIZE)])), + xen_unpin_page(mm, pgd_page(pgd[pgd_index(TASK_SIZE)]), PT_PMD); #endif -- cgit v1.2.3 From d05fdf316067cd311d5e7add08da26ded8a58080 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Tue, 28 Oct 2008 19:23:06 +1100 Subject: xen: make sure stray alias mappings are gone before pinning Xen requires that all mappings of pagetable pages are read-only, so that they can't be updated illegally. As a result, if a page is being turned into a pagetable page, we need to make sure all its mappings are RO. If the page had been used for ioremap or vmalloc, it may still have left over mappings as a result of not having been lazily unmapped. This change makes sure we explicitly mop them all up before pinning the page. Unlike aliases created by kmap, the there can be vmalloc aliases even for non-high pages, so we must do the flush unconditionally. Signed-off-by: Jeremy Fitzhardinge Cc: Linux Memory Management List Cc: Nick Piggin Signed-off-by: Ingo Molnar --- arch/x86/xen/enlighten.c | 5 +++-- arch/x86/xen/mmu.c | 9 ++++++--- 2 files changed, 9 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index b61534c7a4c4..5e4686d70f62 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -863,15 +863,16 @@ static void xen_alloc_ptpage(struct mm_struct *mm, unsigned long pfn, unsigned l if (PagePinned(virt_to_page(mm->pgd))) { SetPagePinned(page); + vm_unmap_aliases(); if (!PageHighMem(page)) { make_lowmem_page_readonly(__va(PFN_PHYS((unsigned long)pfn))); if (level == PT_PTE && USE_SPLIT_PTLOCKS) pin_pagetable_pfn(MMUEXT_PIN_L1_TABLE, pfn); - } else + } else { /* make sure there are no stray mappings of this page */ kmap_flush_unused(); - vm_unmap_aliases(); + } } } diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index aba77b2b7d18..89f3b6edc65a 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -850,13 +850,16 @@ static int xen_pin_page(struct mm_struct *mm, struct page *page, read-only, and can be pinned. */ static void __xen_pgd_pin(struct mm_struct *mm, pgd_t *pgd) { + vm_unmap_aliases(); + xen_mc_batch(); - if (xen_pgd_walk(mm, xen_pin_page, USER_LIMIT)) { - /* re-enable interrupts for kmap_flush_unused */ + if (xen_pgd_walk(mm, xen_pin_page, USER_LIMIT)) { + /* re-enable interrupts for flushing */ xen_mc_issue(0); + kmap_flush_unused(); - vm_unmap_aliases(); + xen_mc_batch(); } -- cgit v1.2.3 From 7c64ade53a6f977d73f16243865c42ceae999aea Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Fri, 7 Nov 2008 14:02:49 +0100 Subject: oprofile: Fix p6 counter overflow check Fix the counter overflow check for CPUs with counter width > 32 I had a similar change in a different patch that I didn't submit and I didn't notice the problem earlier because it was always tested together. Signed-off-by: Andi Kleen Signed-off-by: Robert Richter --- arch/x86/oprofile/op_model_ppro.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index 0620d6d45f7d..3f1b81a83e2e 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -27,8 +27,7 @@ static int num_counters = 2; static int counter_width = 32; #define CTR_IS_RESERVED(msrs, c) (msrs->counters[(c)].addr ? 1 : 0) -#define CTR_READ(l, h, msrs, c) do {rdmsr(msrs->counters[(c)].addr, (l), (h)); } while (0) -#define CTR_OVERFLOWED(n) (!((n) & (1U<<(counter_width-1)))) +#define CTR_OVERFLOWED(n) (!((n) & (1ULL<<(counter_width-1)))) #define CTRL_IS_RESERVED(msrs, c) (msrs->controls[(c)].addr ? 1 : 0) #define CTRL_READ(l, h, msrs, c) do {rdmsr((msrs->controls[(c)].addr), (l), (h)); } while (0) @@ -124,14 +123,14 @@ static void ppro_setup_ctrs(struct op_msrs const * const msrs) static int ppro_check_ctrs(struct pt_regs * const regs, struct op_msrs const * const msrs) { - unsigned int low, high; + u64 val; int i; for (i = 0 ; i < num_counters; ++i) { if (!reset_value[i]) continue; - CTR_READ(low, high, msrs, i); - if (CTR_OVERFLOWED(low)) { + rdmsrl(msrs->counters[i].addr, val); + if (CTR_OVERFLOWED(val)) { oprofile_add_sample(regs, i); wrmsrl(msrs->counters[i].addr, -reset_value[i]); } -- cgit v1.2.3 From 0d12cdd5f883f508d33b85c1bae98fa28987c8c7 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 8 Nov 2008 16:19:55 +0100 Subject: sched: improve sched_clock() performance in scheduler-intense workloads native_read_tsc() overhead accounts for 20% of the system overhead: 659567 system_call 41222.9375 686796 schedule 435.7843 718382 __switch_to 665.1685 823875 switch_mm 4526.7857 1883122 native_read_tsc 55385.9412 9761990 total 2.8468 this is large part due to the rdtsc_barrier() that is done before and after reading the TSC. But sched_clock() is not a precise clock in the GTOD sense, using such barriers is completely pointless. So remove the barriers and only use them in vget_cycles(). This improves lat_ctx performance by about 5%. Signed-off-by: Ingo Molnar --- arch/x86/include/asm/msr.h | 2 -- arch/x86/include/asm/tsc.h | 8 +++++++- 2 files changed, 7 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/msr.h b/arch/x86/include/asm/msr.h index 46be2fa7ac26..c2a812ebde89 100644 --- a/arch/x86/include/asm/msr.h +++ b/arch/x86/include/asm/msr.h @@ -108,9 +108,7 @@ static __always_inline unsigned long long __native_read_tsc(void) { DECLARE_ARGS(val, low, high); - rdtsc_barrier(); asm volatile("rdtsc" : EAX_EDX_RET(val, low, high)); - rdtsc_barrier(); return EAX_EDX_VAL(val, low, high); } diff --git a/arch/x86/include/asm/tsc.h b/arch/x86/include/asm/tsc.h index 38ae163cc91b..9cd83a8e40d5 100644 --- a/arch/x86/include/asm/tsc.h +++ b/arch/x86/include/asm/tsc.h @@ -34,6 +34,8 @@ static inline cycles_t get_cycles(void) static __always_inline cycles_t vget_cycles(void) { + cycles_t cycles; + /* * We only do VDSOs on TSC capable CPUs, so this shouldnt * access boot_cpu_data (which is not VDSO-safe): @@ -42,7 +44,11 @@ static __always_inline cycles_t vget_cycles(void) if (!cpu_has_tsc) return 0; #endif - return (cycles_t)__native_read_tsc(); + rdtsc_barrier(); + cycles = (cycles_t)__native_read_tsc(); + rdtsc_barrier(); + + return cycles; } extern void tsc_init(void); -- cgit v1.2.3 From 7cbaef9c83e58bbd4bdd534b09052b6c5ec457d5 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 8 Nov 2008 17:05:38 +0100 Subject: sched: optimize sched_clock() a bit sched_clock() uses cycles_2_ns() needlessly - which is an irq-disabling variant of __cycles_2_ns(). Most of the time sched_clock() is called with irqs disabled already. The few places that call it with irqs enabled need to be updated. Signed-off-by: Ingo Molnar --- arch/x86/kernel/tsc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index 2ef80e301925..424093b157d3 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -55,7 +55,7 @@ u64 native_sched_clock(void) rdtscll(this_offset); /* return the value in ns */ - return cycles_2_ns(this_offset); + return __cycles_2_ns(this_offset); } /* We need to define a real function for sched_clock, to override the -- cgit v1.2.3