From c40e3641670eb6ebfdb71d4b0c775416ef95f4f0 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Mon, 18 Mar 2013 19:44:14 +0100 Subject: ARM: 7679/1: Clear IDIVT hwcap if CONFIG_ARM_THUMB=n Don't advertise support for the SDIV/UDIV thumb instructions if the kernel is not compiled with support for thumb userspace. This is in line with how we remove the THUMB hwcap in these configurations. Acked-by: Will Deacon Cc: Stepan Moskovchenko Signed-off-by: Stephen Boyd Signed-off-by: Russell King --- arch/arm/kernel/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/arm/kernel') diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index 3f6cbb2e3eda..e2c8bbffb0b1 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -484,7 +484,7 @@ static void __init setup_processor(void) list->elf_name, ENDIANNESS); elf_hwcap = list->elf_hwcap; #ifndef CONFIG_ARM_THUMB - elf_hwcap &= ~HWCAP_THUMB; + elf_hwcap &= ~(HWCAP_THUMB | HWCAP_IDIVT); #endif feat_v6_fixup(); -- cgit v1.2.3 From 8164f7af88d9ad3a757bd14f634b23997ee77f6b Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Mon, 18 Mar 2013 19:44:15 +0100 Subject: ARM: 7680/1: Detect support for SDIV/UDIV from ISAR0 register The ISAR0 register indicates support for the SDIV and UDIV instructions in both the Thumb and ARM instruction set. Read the register to detect the supported instructions and update the elf_hwcap mask as appropriate. This is better than adding more and more cpuid checks in proc-v7.S for each new cpu variant that supports these instructions. Acked-by: Will Deacon Cc: Stepan Moskovchenko Signed-off-by: Stephen Boyd Signed-off-by: Russell King --- arch/arm/kernel/setup.c | 20 ++++++++++++++++++++ arch/arm/mm/proc-v7.S | 4 ++-- 2 files changed, 22 insertions(+), 2 deletions(-) (limited to 'arch/arm/kernel') diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index e2c8bbffb0b1..f3ac13f69b7a 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -353,6 +353,23 @@ void __init early_print(const char *str, ...) printk("%s", buf); } +static void __init cpuid_init_hwcaps(void) +{ + unsigned int divide_instrs; + + if (cpu_architecture() < CPU_ARCH_ARMv7) + return; + + divide_instrs = (read_cpuid_ext(CPUID_EXT_ISAR0) & 0x0f000000) >> 24; + + switch (divide_instrs) { + case 2: + elf_hwcap |= HWCAP_IDIVA; + case 1: + elf_hwcap |= HWCAP_IDIVT; + } +} + static void __init feat_v6_fixup(void) { int id = read_cpuid_id(); @@ -483,6 +500,9 @@ static void __init setup_processor(void) snprintf(elf_platform, ELF_PLATFORM_SIZE, "%s%c", list->elf_name, ENDIANNESS); elf_hwcap = list->elf_hwcap; + + cpuid_init_hwcaps(); + #ifndef CONFIG_ARM_THUMB elf_hwcap &= ~(HWCAP_THUMB | HWCAP_IDIVT); #endif diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 3a3c015f8d5c..bcd3d48922fb 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -420,7 +420,7 @@ __v7_pj4b_proc_info: __v7_ca7mp_proc_info: .long 0x410fc070 .long 0xff0ffff0 - __v7_proc __v7_ca7mp_setup, hwcaps = HWCAP_IDIV + __v7_proc __v7_ca7mp_setup .size __v7_ca7mp_proc_info, . - __v7_ca7mp_proc_info /* @@ -430,7 +430,7 @@ __v7_ca7mp_proc_info: __v7_ca15mp_proc_info: .long 0x410fc0f0 .long 0xff0ffff0 - __v7_proc __v7_ca15mp_setup, hwcaps = HWCAP_IDIV + __v7_proc __v7_ca15mp_setup .size __v7_ca15mp_proc_info, . - __v7_ca15mp_proc_info /* -- cgit v1.2.3 From 68a154fc53ddd3f7b33e482847a411bf54a50855 Mon Sep 17 00:00:00 2001 From: Santosh Shilimkar Date: Wed, 20 Mar 2013 17:30:30 +0100 Subject: ARM: 7681/1: hw_breakpoint: use warn_once to avoid spam from reset_ctrl_regs() CPU debug features like hardware break, watchpoints can be used only when the debug mode is enabled and available. Unfortunately on OMAP4 based devices, after a CPU power cycle, the debug feature gets disabled which leads to a flood of messages coming from reset_ctrl_regs() which gets called on every CPU_PM_EXIT with CPUidle enabled. So make use of warn_once() so that system is usable. Thanks to Will for pointers and Lokesh for the analysis of the issue. Tested-by: Lokesh Vutla Signed-off-by: Santosh Shilimkar Signed-off-by: Will Deacon Signed-off-by: Russell King --- arch/arm/kernel/hw_breakpoint.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch/arm/kernel') diff --git a/arch/arm/kernel/hw_breakpoint.c b/arch/arm/kernel/hw_breakpoint.c index 96093b75ab90..5dc1aa6f0f7d 100644 --- a/arch/arm/kernel/hw_breakpoint.c +++ b/arch/arm/kernel/hw_breakpoint.c @@ -966,7 +966,7 @@ static void reset_ctrl_regs(void *unused) } if (err) { - pr_warning("CPU %d debug is powered down!\n", cpu); + pr_warn_once("CPU %d debug is powered down!\n", cpu); cpumask_or(&debug_err_mask, &debug_err_mask, cpumask_of(cpu)); return; } @@ -987,7 +987,7 @@ clear_vcr: isb(); if (cpumask_intersects(&debug_err_mask, cpumask_of(cpu))) { - pr_warning("CPU %d failed to disable vector catch\n", cpu); + pr_warn_once("CPU %d failed to disable vector catch\n", cpu); return; } @@ -1007,7 +1007,7 @@ clear_vcr: } if (cpumask_intersects(&debug_err_mask, cpumask_of(cpu))) { - pr_warning("CPU %d failed to clear debug register pairs\n", cpu); + pr_warn_once("CPU %d failed to clear debug register pairs\n", cpu); return; } -- cgit v1.2.3 From 93dc68876b608da041fe40ed39424b0fcd5aa2fb Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Tue, 26 Mar 2013 23:35:04 +0100 Subject: ARM: 7684/1: errata: Workaround for Cortex-A15 erratum 798181 (TLBI/DSB operations) On Cortex-A15 (r0p0..r3p2) the TLBI/DSB are not adequately shooting down all use of the old entries. This patch implements the erratum workaround which consists of: 1. Dummy TLBIMVAIS and DSB on the CPU doing the TLBI operation. 2. Send IPI to the CPUs that are running the same mm (and ASID) as the one being invalidated (or all the online CPUs for global pages). 3. CPU receiving the IPI executes a DMB and CLREX (part of the exception return code already). Signed-off-by: Catalin Marinas Signed-off-by: Russell King --- arch/arm/Kconfig | 10 ++++++ arch/arm/include/asm/highmem.h | 7 ++++ arch/arm/include/asm/mmu_context.h | 2 ++ arch/arm/include/asm/tlbflush.h | 15 +++++++++ arch/arm/kernel/smp_tlb.c | 66 ++++++++++++++++++++++++++++++++++++++ arch/arm/mm/context.c | 3 +- 6 files changed, 102 insertions(+), 1 deletion(-) (limited to 'arch/arm/kernel') diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 12ea3b3d49a9..1cacda426a0e 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1439,6 +1439,16 @@ config ARM_ERRATA_775420 to deadlock. This workaround puts DSB before executing ISB if an abort may occur on cache maintenance. +config ARM_ERRATA_798181 + bool "ARM errata: TLBI/DSB failure on Cortex-A15" + depends on CPU_V7 && SMP + help + On Cortex-A15 (r0p0..r3p2) the TLBI*IS/DSB operations are not + adequately shooting down all use of the old entries. This + option enables the Linux kernel workaround for this erratum + which sends an IPI to the CPUs that are running the same ASID + as the one being invalidated. + endmenu source "arch/arm/common/Kconfig" diff --git a/arch/arm/include/asm/highmem.h b/arch/arm/include/asm/highmem.h index 8c5e828f484d..91b99abe7a95 100644 --- a/arch/arm/include/asm/highmem.h +++ b/arch/arm/include/asm/highmem.h @@ -41,6 +41,13 @@ extern void kunmap_high(struct page *page); #endif #endif +/* + * Needed to be able to broadcast the TLB invalidation for kmap. + */ +#ifdef CONFIG_ARM_ERRATA_798181 +#undef ARCH_NEEDS_KMAP_HIGH_GET +#endif + #ifdef ARCH_NEEDS_KMAP_HIGH_GET extern void *kmap_high_get(struct page *page); #else diff --git a/arch/arm/include/asm/mmu_context.h b/arch/arm/include/asm/mmu_context.h index 863a6611323c..a7b85e0d0cc1 100644 --- a/arch/arm/include/asm/mmu_context.h +++ b/arch/arm/include/asm/mmu_context.h @@ -27,6 +27,8 @@ void __check_vmalloc_seq(struct mm_struct *mm); void check_and_switch_context(struct mm_struct *mm, struct task_struct *tsk); #define init_new_context(tsk,mm) ({ atomic64_set(&mm->context.id, 0); 0; }) +DECLARE_PER_CPU(atomic64_t, active_asids); + #else /* !CONFIG_CPU_HAS_ASID */ #ifdef CONFIG_MMU diff --git a/arch/arm/include/asm/tlbflush.h b/arch/arm/include/asm/tlbflush.h index 4db8c8820f0d..9e9c041358ca 100644 --- a/arch/arm/include/asm/tlbflush.h +++ b/arch/arm/include/asm/tlbflush.h @@ -450,6 +450,21 @@ static inline void local_flush_bp_all(void) isb(); } +#ifdef CONFIG_ARM_ERRATA_798181 +static inline void dummy_flush_tlb_a15_erratum(void) +{ + /* + * Dummy TLBIMVAIS. Using the unmapped address 0 and ASID 0. + */ + asm("mcr p15, 0, %0, c8, c3, 1" : : "r" (0)); + dsb(); +} +#else +static inline void dummy_flush_tlb_a15_erratum(void) +{ +} +#endif + /* * flush_pmd_entry * diff --git a/arch/arm/kernel/smp_tlb.c b/arch/arm/kernel/smp_tlb.c index bd0300531399..e82e1d248772 100644 --- a/arch/arm/kernel/smp_tlb.c +++ b/arch/arm/kernel/smp_tlb.c @@ -12,6 +12,7 @@ #include #include +#include /**********************************************************************/ @@ -69,12 +70,72 @@ static inline void ipi_flush_bp_all(void *ignored) local_flush_bp_all(); } +#ifdef CONFIG_ARM_ERRATA_798181 +static int erratum_a15_798181(void) +{ + unsigned int midr = read_cpuid_id(); + + /* Cortex-A15 r0p0..r3p2 affected */ + if ((midr & 0xff0ffff0) != 0x410fc0f0 || midr > 0x413fc0f2) + return 0; + return 1; +} +#else +static int erratum_a15_798181(void) +{ + return 0; +} +#endif + +static void ipi_flush_tlb_a15_erratum(void *arg) +{ + dmb(); +} + +static void broadcast_tlb_a15_erratum(void) +{ + if (!erratum_a15_798181()) + return; + + dummy_flush_tlb_a15_erratum(); + smp_call_function_many(cpu_online_mask, ipi_flush_tlb_a15_erratum, + NULL, 1); +} + +static void broadcast_tlb_mm_a15_erratum(struct mm_struct *mm) +{ + int cpu; + cpumask_t mask = { CPU_BITS_NONE }; + + if (!erratum_a15_798181()) + return; + + dummy_flush_tlb_a15_erratum(); + for_each_online_cpu(cpu) { + if (cpu == smp_processor_id()) + continue; + /* + * We only need to send an IPI if the other CPUs are running + * the same ASID as the one being invalidated. There is no + * need for locking around the active_asids check since the + * switch_mm() function has at least one dmb() (as required by + * this workaround) in case a context switch happens on + * another CPU after the condition below. + */ + if (atomic64_read(&mm->context.id) == + atomic64_read(&per_cpu(active_asids, cpu))) + cpumask_set_cpu(cpu, &mask); + } + smp_call_function_many(&mask, ipi_flush_tlb_a15_erratum, NULL, 1); +} + void flush_tlb_all(void) { if (tlb_ops_need_broadcast()) on_each_cpu(ipi_flush_tlb_all, NULL, 1); else local_flush_tlb_all(); + broadcast_tlb_a15_erratum(); } void flush_tlb_mm(struct mm_struct *mm) @@ -83,6 +144,7 @@ void flush_tlb_mm(struct mm_struct *mm) on_each_cpu_mask(mm_cpumask(mm), ipi_flush_tlb_mm, mm, 1); else local_flush_tlb_mm(mm); + broadcast_tlb_mm_a15_erratum(mm); } void flush_tlb_page(struct vm_area_struct *vma, unsigned long uaddr) @@ -95,6 +157,7 @@ void flush_tlb_page(struct vm_area_struct *vma, unsigned long uaddr) &ta, 1); } else local_flush_tlb_page(vma, uaddr); + broadcast_tlb_mm_a15_erratum(vma->vm_mm); } void flush_tlb_kernel_page(unsigned long kaddr) @@ -105,6 +168,7 @@ void flush_tlb_kernel_page(unsigned long kaddr) on_each_cpu(ipi_flush_tlb_kernel_page, &ta, 1); } else local_flush_tlb_kernel_page(kaddr); + broadcast_tlb_a15_erratum(); } void flush_tlb_range(struct vm_area_struct *vma, @@ -119,6 +183,7 @@ void flush_tlb_range(struct vm_area_struct *vma, &ta, 1); } else local_flush_tlb_range(vma, start, end); + broadcast_tlb_mm_a15_erratum(vma->vm_mm); } void flush_tlb_kernel_range(unsigned long start, unsigned long end) @@ -130,6 +195,7 @@ void flush_tlb_kernel_range(unsigned long start, unsigned long end) on_each_cpu(ipi_flush_tlb_kernel_range, &ta, 1); } else local_flush_tlb_kernel_range(start, end); + broadcast_tlb_a15_erratum(); } void flush_bp_all(void) diff --git a/arch/arm/mm/context.c b/arch/arm/mm/context.c index a5a4b2bc42ba..2ac37372ef52 100644 --- a/arch/arm/mm/context.c +++ b/arch/arm/mm/context.c @@ -48,7 +48,7 @@ static DEFINE_RAW_SPINLOCK(cpu_asid_lock); static atomic64_t asid_generation = ATOMIC64_INIT(ASID_FIRST_VERSION); static DECLARE_BITMAP(asid_map, NUM_USER_ASIDS); -static DEFINE_PER_CPU(atomic64_t, active_asids); +DEFINE_PER_CPU(atomic64_t, active_asids); static DEFINE_PER_CPU(u64, reserved_asids); static cpumask_t tlb_flush_pending; @@ -215,6 +215,7 @@ void check_and_switch_context(struct mm_struct *mm, struct task_struct *tsk) if (cpumask_test_and_clear_cpu(cpu, &tlb_flush_pending)) { local_flush_bp_all(); local_flush_tlb_all(); + dummy_flush_tlb_a15_erratum(); } atomic64_set(&per_cpu(active_asids, cpu), asid); -- cgit v1.2.3 From 6f3d90e55660ba42301b5e9c7eed332cc9f70fd7 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Thu, 28 Mar 2013 11:17:55 +0100 Subject: ARM: 7685/1: delay: use private ticks_per_jiffy field for timer-based delay ops Commit 70264367a243 ("ARM: 7653/2: do not scale loops_per_jiffy when using a constant delay clock") fixed a problem with our timer-based delay loop, where loops_per_jiffy is scaled by cpufreq yet used directly by the timer delay ops. This patch fixes the problem in a more elegant way by keeping a private ticks_per_jiffy field in the delay ops, independent of loops_per_jiffy and therefore not subject to scaling. The loop-based delay continues to use loops_per_jiffy directly, as it should. Acked-by: Nicolas Pitre Signed-off-by: Will Deacon Signed-off-by: Russell King --- arch/arm/include/asm/delay.h | 2 +- arch/arm/kernel/smp.c | 3 --- arch/arm/lib/delay.c | 8 +++++--- 3 files changed, 6 insertions(+), 7 deletions(-) (limited to 'arch/arm/kernel') diff --git a/arch/arm/include/asm/delay.h b/arch/arm/include/asm/delay.h index 720799fd3a81..dff714d886d5 100644 --- a/arch/arm/include/asm/delay.h +++ b/arch/arm/include/asm/delay.h @@ -24,7 +24,7 @@ extern struct arm_delay_ops { void (*delay)(unsigned long); void (*const_udelay)(unsigned long); void (*udelay)(unsigned long); - bool const_clock; + unsigned long ticks_per_jiffy; } arm_delay_ops; #define __delay(n) arm_delay_ops.delay(n) diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c index 79078edbb9bc..1f2ccccaf009 100644 --- a/arch/arm/kernel/smp.c +++ b/arch/arm/kernel/smp.c @@ -673,9 +673,6 @@ static int cpufreq_callback(struct notifier_block *nb, if (freq->flags & CPUFREQ_CONST_LOOPS) return NOTIFY_OK; - if (arm_delay_ops.const_clock) - return NOTIFY_OK; - if (!per_cpu(l_p_j_ref, cpu)) { per_cpu(l_p_j_ref, cpu) = per_cpu(cpu_data, cpu).loops_per_jiffy; diff --git a/arch/arm/lib/delay.c b/arch/arm/lib/delay.c index 6b93f6a1a3c7..64dbfa57204a 100644 --- a/arch/arm/lib/delay.c +++ b/arch/arm/lib/delay.c @@ -58,7 +58,7 @@ static void __timer_delay(unsigned long cycles) static void __timer_const_udelay(unsigned long xloops) { unsigned long long loops = xloops; - loops *= loops_per_jiffy; + loops *= arm_delay_ops.ticks_per_jiffy; __timer_delay(loops >> UDELAY_SHIFT); } @@ -73,11 +73,13 @@ void __init register_current_timer_delay(const struct delay_timer *timer) pr_info("Switching to timer-based delay loop\n"); delay_timer = timer; lpj_fine = timer->freq / HZ; - loops_per_jiffy = lpj_fine; + + /* cpufreq may scale loops_per_jiffy, so keep a private copy */ + arm_delay_ops.ticks_per_jiffy = lpj_fine; arm_delay_ops.delay = __timer_delay; arm_delay_ops.const_udelay = __timer_const_udelay; arm_delay_ops.udelay = __timer_udelay; - arm_delay_ops.const_clock = true; + delay_calibrated = true; } else { pr_info("Ignoring duplicate/late registration of read_current_timer delay\n"); -- cgit v1.2.3 From b21e023ba4003fe4b5c32540e4aee5991c019e92 Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Tue, 2 Apr 2013 22:11:46 +0100 Subject: ARM: 7689/1: add unwind annotations to ftrace asm Add unwind annotations to the ftrace assembly code so that the function tracer's stacktracing options (func_stack_trace, etc.) work when CONFIG_ARM_UNWIND is enabled. Signed-off-by: Rabin Vincent Signed-off-by: Russell King --- arch/arm/kernel/entry-common.S | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'arch/arm/kernel') diff --git a/arch/arm/kernel/entry-common.S b/arch/arm/kernel/entry-common.S index 3248cde504ed..fefd7f971437 100644 --- a/arch/arm/kernel/entry-common.S +++ b/arch/arm/kernel/entry-common.S @@ -276,7 +276,13 @@ ENDPROC(ftrace_graph_caller_old) */ .macro mcount_enter +/* + * This pad compensates for the push {lr} at the call site. Note that we are + * unable to unwind through a function which does not otherwise save its lr. + */ + UNWIND(.pad #4) stmdb sp!, {r0-r3, lr} + UNWIND(.save {r0-r3, lr}) .endm .macro mcount_get_lr reg @@ -289,6 +295,7 @@ ENDPROC(ftrace_graph_caller_old) .endm ENTRY(__gnu_mcount_nc) +UNWIND(.fnstart) #ifdef CONFIG_DYNAMIC_FTRACE mov ip, lr ldmia sp!, {lr} @@ -296,17 +303,22 @@ ENTRY(__gnu_mcount_nc) #else __mcount #endif +UNWIND(.fnend) ENDPROC(__gnu_mcount_nc) #ifdef CONFIG_DYNAMIC_FTRACE ENTRY(ftrace_caller) +UNWIND(.fnstart) __ftrace_caller +UNWIND(.fnend) ENDPROC(ftrace_caller) #endif #ifdef CONFIG_FUNCTION_GRAPH_TRACER ENTRY(ftrace_graph_caller) +UNWIND(.fnstart) __ftrace_graph_caller +UNWIND(.fnend) ENDPROC(ftrace_graph_caller) #endif -- cgit v1.2.3 From 4e1db26a0b42e2b6e27c05d68adcc01709c2eed2 Mon Sep 17 00:00:00 2001 From: Paul Bolle Date: Wed, 3 Apr 2013 12:24:45 +0100 Subject: ARM: 7690/1: mm: fix CONFIG_LPAE typos CONFIG_LPAE doesn't exist: the correct option is CONFIG_ARM_LPAE, so fix up the two typos under arch/arm/. The fix to head.S is slightly scary, but this is just for setting up an early io-mapping for the serial port when running on a big-endian, LPAE system. Since these systems don't exist in the wild (at least, I have no access to one outside of kvmtool, which doesn't provide a serial port suitable for earlyprintk), then we can revisit the code later if it causes any problems. Signed-off-by: Paul Bolle Signed-off-by: Will Deacon Signed-off-by: Russell King --- arch/arm/kernel/head.S | 2 +- arch/arm/kernel/setup.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/arm/kernel') diff --git a/arch/arm/kernel/head.S b/arch/arm/kernel/head.S index e0eb9a1cae77..8bac553fe213 100644 --- a/arch/arm/kernel/head.S +++ b/arch/arm/kernel/head.S @@ -267,7 +267,7 @@ __create_page_tables: addne r6, r6, #1 << SECTION_SHIFT strne r6, [r3] -#if defined(CONFIG_LPAE) && defined(CONFIG_CPU_ENDIAN_BE8) +#if defined(CONFIG_ARM_LPAE) && defined(CONFIG_CPU_ENDIAN_BE8) sub r4, r4, #4 @ Fixup page table pointer @ for 64-bit descriptors #endif diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index f3ac13f69b7a..d343a6c3a6d1 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -544,7 +544,7 @@ int __init arm_add_memory(phys_addr_t start, phys_addr_t size) size -= start & ~PAGE_MASK; bank->start = PAGE_ALIGN(start); -#ifndef CONFIG_LPAE +#ifndef CONFIG_ARM_LPAE if (bank->start + size < bank->start) { printk(KERN_CRIT "Truncating memory at 0x%08llx to fit in " "32-bit physical address space\n", (long long)start); -- cgit v1.2.3