From d430d3d7e646eb1eac2bb4aa244a644312e67c76 Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Wed, 16 Mar 2011 17:29:47 -0400 Subject: jump label: Introduce static_branch() interface Introduce: static __always_inline bool static_branch(struct jump_label_key *key); instead of the old JUMP_LABEL(key, label) macro. In this way, jump labels become really easy to use: Define: struct jump_label_key jump_key; Can be used as: if (static_branch(&jump_key)) do unlikely code enable/disale via: jump_label_inc(&jump_key); jump_label_dec(&jump_key); that's it! For the jump labels disabled case, the static_branch() becomes an atomic_read(), and jump_label_inc()/dec() are simply atomic_inc(), atomic_dec() operations. We show testing results for this change below. Thanks to H. Peter Anvin for suggesting the 'static_branch()' construct. Since we now require a 'struct jump_label_key *key', we can store a pointer into the jump table addresses. In this way, we can enable/disable jump labels, in basically constant time. This change allows us to completely remove the previous hashtable scheme. Thanks to Peter Zijlstra for this re-write. Testing: I ran a series of 'tbench 20' runs 5 times (with reboots) for 3 configurations, where tracepoints were disabled. jump label configured in avg: 815.6 jump label *not* configured in (using atomic reads) avg: 800.1 jump label *not* configured in (regular reads) avg: 803.4 Signed-off-by: Peter Zijlstra LKML-Reference: <20110316212947.GA8792@redhat.com> Signed-off-by: Jason Baron Suggested-by: H. Peter Anvin Tested-by: David Daney Acked-by: Ralf Baechle Acked-by: David S. Miller Acked-by: Mathieu Desnoyers Signed-off-by: Steven Rostedt --- arch/x86/include/asm/alternative.h | 3 +-- arch/x86/include/asm/jump_label.h | 26 +++++++++++++++----------- 2 files changed, 16 insertions(+), 13 deletions(-) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h index 13009d1af99a..8cdd1e247975 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -4,7 +4,6 @@ #include #include #include -#include #include /* @@ -191,7 +190,7 @@ extern void *text_poke(void *addr, const void *opcode, size_t len); extern void *text_poke_smp(void *addr, const void *opcode, size_t len); extern void text_poke_smp_batch(struct text_poke_param *params, int n); -#if defined(CONFIG_DYNAMIC_FTRACE) || defined(HAVE_JUMP_LABEL) +#if defined(CONFIG_DYNAMIC_FTRACE) || defined(CONFIG_JUMP_LABEL) #define IDEAL_NOP_SIZE_5 5 extern unsigned char ideal_nop5[IDEAL_NOP_SIZE_5]; extern void arch_init_ideal_nop5(void); diff --git a/arch/x86/include/asm/jump_label.h b/arch/x86/include/asm/jump_label.h index 574dbc22893a..f217cee86533 100644 --- a/arch/x86/include/asm/jump_label.h +++ b/arch/x86/include/asm/jump_label.h @@ -5,20 +5,24 @@ #include #include +#include #define JUMP_LABEL_NOP_SIZE 5 -# define JUMP_LABEL_INITIAL_NOP ".byte 0xe9 \n\t .long 0\n\t" - -# define JUMP_LABEL(key, label) \ - do { \ - asm goto("1:" \ - JUMP_LABEL_INITIAL_NOP \ - ".pushsection __jump_table, \"aw\" \n\t"\ - _ASM_PTR "1b, %l[" #label "], %c0 \n\t" \ - ".popsection \n\t" \ - : : "i" (key) : : label); \ - } while (0) +#define JUMP_LABEL_INITIAL_NOP ".byte 0xe9 \n\t .long 0\n\t" + +static __always_inline bool arch_static_branch(struct jump_label_key *key) +{ + asm goto("1:" + JUMP_LABEL_INITIAL_NOP + ".pushsection __jump_table, \"aw\" \n\t" + _ASM_PTR "1b, %l[l_yes], %c0 \n\t" + ".popsection \n\t" + : : "i" (key) : : l_yes); + return false; +l_yes: + return true; +} #endif /* __KERNEL__ */ -- cgit v1.2.3 From ef64789413c73f32faa5e5f1bc393e5843b0aa51 Mon Sep 17 00:00:00 2001 From: Jason Baron Date: Wed, 16 Mar 2011 15:58:27 -0400 Subject: jump label: Add _ASM_ALIGN for x86 and x86_64 The linker should not be adding holes to word size aligned pointers, but out of paranoia we are explicitly specifying that alignment. I have not seen any holes in the jump label section in practice. Signed-off-by: Jason Baron LKML-Reference: Acked-by: Peter Zijlstra Acked-by: Mathieu Desnoyers Signed-off-by: Steven Rostedt --- arch/x86/include/asm/jump_label.h | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/jump_label.h b/arch/x86/include/asm/jump_label.h index f217cee86533..a32b18ce6ead 100644 --- a/arch/x86/include/asm/jump_label.h +++ b/arch/x86/include/asm/jump_label.h @@ -16,6 +16,7 @@ static __always_inline bool arch_static_branch(struct jump_label_key *key) asm goto("1:" JUMP_LABEL_INITIAL_NOP ".pushsection __jump_table, \"aw\" \n\t" + _ASM_ALIGN "\n\t" _ASM_PTR "1b, %l[l_yes], %c0 \n\t" ".popsection \n\t" : : "i" (key) : : l_yes); -- cgit v1.2.3 From 449a66fd1fa75d36dca917704827c40c8f416bca Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Thu, 12 May 2011 15:11:12 +0200 Subject: x86: Remove warning and warning_symbol from struct stacktrace_ops Both warning and warning_symbol are nowhere used. Let's get rid of them. Signed-off-by: Richard Weinberger Cc: Oleg Nesterov Cc: Andrew Morton Cc: Huang Ying Cc: Soeren Sandmann Pedersen Cc: Namhyung Kim Cc: x86 Cc: H. Peter Anvin Cc: Thomas Gleixner Cc: Robert Richter Cc: Paul Mundt Link: http://lkml.kernel.org/r/1305205872-10321-2-git-send-email-richard@nod.at Signed-off-by: Frederic Weisbecker --- arch/x86/include/asm/stacktrace.h | 3 --- arch/x86/kernel/cpu/perf_event.c | 13 ------------- arch/x86/kernel/dumpstack.c | 16 ---------------- arch/x86/kernel/stacktrace.c | 13 ------------- arch/x86/oprofile/backtrace.c | 13 ------------- 5 files changed, 58 deletions(-) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/stacktrace.h b/arch/x86/include/asm/stacktrace.h index d7e89c83645d..70bbe39043a9 100644 --- a/arch/x86/include/asm/stacktrace.h +++ b/arch/x86/include/asm/stacktrace.h @@ -37,9 +37,6 @@ print_context_stack_bp(struct thread_info *tinfo, /* Generic stack tracer with callbacks */ struct stacktrace_ops { - void (*warning)(void *data, char *msg); - /* msg must contain %s for the symbol */ - void (*warning_symbol)(void *data, char *msg, unsigned long symbol); void (*address)(void *data, unsigned long address, int reliable); /* On negative return stop dumping */ int (*stack)(void *data, char *name); diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 0de6b2b31f61..3a0338b4b179 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1773,17 +1773,6 @@ static struct pmu pmu = { * callchain support */ -static void -backtrace_warning_symbol(void *data, char *msg, unsigned long symbol) -{ - /* Ignore warnings */ -} - -static void backtrace_warning(void *data, char *msg) -{ - /* Ignore warnings */ -} - static int backtrace_stack(void *data, char *name) { return 0; @@ -1797,8 +1786,6 @@ static void backtrace_address(void *data, unsigned long addr, int reliable) } static const struct stacktrace_ops backtrace_ops = { - .warning = backtrace_warning, - .warning_symbol = backtrace_warning_symbol, .stack = backtrace_stack, .address = backtrace_address, .walk_stack = print_context_stack_bp, diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index e2a3f0606da4..f478ff6877ef 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -135,20 +135,6 @@ print_context_stack_bp(struct thread_info *tinfo, } EXPORT_SYMBOL_GPL(print_context_stack_bp); - -static void -print_trace_warning_symbol(void *data, char *msg, unsigned long symbol) -{ - printk(data); - print_symbol(msg, symbol); - printk("\n"); -} - -static void print_trace_warning(void *data, char *msg) -{ - printk("%s%s\n", (char *)data, msg); -} - static int print_trace_stack(void *data, char *name) { printk("%s <%s> ", (char *)data, name); @@ -166,8 +152,6 @@ static void print_trace_address(void *data, unsigned long addr, int reliable) } static const struct stacktrace_ops print_trace_ops = { - .warning = print_trace_warning, - .warning_symbol = print_trace_warning_symbol, .stack = print_trace_stack, .address = print_trace_address, .walk_stack = print_context_stack, diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c index 6515733a289d..55d9bc03f696 100644 --- a/arch/x86/kernel/stacktrace.c +++ b/arch/x86/kernel/stacktrace.c @@ -9,15 +9,6 @@ #include #include -static void save_stack_warning(void *data, char *msg) -{ -} - -static void -save_stack_warning_symbol(void *data, char *msg, unsigned long symbol) -{ -} - static int save_stack_stack(void *data, char *name) { return 0; @@ -53,16 +44,12 @@ save_stack_address_nosched(void *data, unsigned long addr, int reliable) } static const struct stacktrace_ops save_stack_ops = { - .warning = save_stack_warning, - .warning_symbol = save_stack_warning_symbol, .stack = save_stack_stack, .address = save_stack_address, .walk_stack = print_context_stack, }; static const struct stacktrace_ops save_stack_ops_nosched = { - .warning = save_stack_warning, - .warning_symbol = save_stack_warning_symbol, .stack = save_stack_stack, .address = save_stack_address_nosched, .walk_stack = print_context_stack, diff --git a/arch/x86/oprofile/backtrace.c b/arch/x86/oprofile/backtrace.c index 2d49d4e19a36..a5b64ab4cd6e 100644 --- a/arch/x86/oprofile/backtrace.c +++ b/arch/x86/oprofile/backtrace.c @@ -16,17 +16,6 @@ #include #include -static void backtrace_warning_symbol(void *data, char *msg, - unsigned long symbol) -{ - /* Ignore warnings */ -} - -static void backtrace_warning(void *data, char *msg) -{ - /* Ignore warnings */ -} - static int backtrace_stack(void *data, char *name) { /* Yes, we want all stacks */ @@ -42,8 +31,6 @@ static void backtrace_address(void *data, unsigned long addr, int reliable) } static struct stacktrace_ops backtrace_ops = { - .warning = backtrace_warning, - .warning_symbol = backtrace_warning_symbol, .stack = backtrace_stack, .address = backtrace_address, .walk_stack = print_context_stack, -- cgit v1.2.3 From 2895cd2ab81dfb7bc22637bc110857db44a30b4a Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 13 Apr 2011 16:43:29 -0400 Subject: ftrace/x86: Do not trace .discard.text section The section called .discard.text has tracing attached to it and is currently ignored by ftrace. But it does include a call to the mcount stub. Adding a notrace to the code keeps gcc from adding the useless mcount caller to it. Link: http://lkml.kernel.org/r/20110421023739.243651696@goodmis.org Signed-off-by: Steven Rostedt --- arch/x86/include/asm/setup.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/setup.h b/arch/x86/include/asm/setup.h index db8aa19a08a2..647d8a06ce4f 100644 --- a/arch/x86/include/asm/setup.h +++ b/arch/x86/include/asm/setup.h @@ -88,7 +88,7 @@ void *extend_brk(size_t size, size_t align); * executable.) */ #define RESERVE_BRK(name,sz) \ - static void __section(.discard.text) __used \ + static void __section(.discard.text) __used notrace \ __brk_reservation_fn_##name##__(void) { \ asm volatile ( \ ".pushsection .brk_reservation,\"aw\",@nobits;" \ -- cgit v1.2.3 From 521ccb5c4aece609311bfa7157910a8f0c942af5 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Tue, 10 May 2011 10:10:41 +0200 Subject: ftrace/x86: mcount offset calculation Do the mcount offset adjustment in the recordmcount.pl/recordmcount.[ch] at compile time and not in ftrace_call_adjust at run time. Signed-off-by: Martin Schwidefsky Signed-off-by: Steven Rostedt --- arch/x86/include/asm/ftrace.h | 7 +++---- scripts/recordmcount.c | 2 ++ scripts/recordmcount.pl | 2 ++ 3 files changed, 7 insertions(+), 4 deletions(-) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/ftrace.h b/arch/x86/include/asm/ftrace.h index db24c2278be0..268c783ab1c0 100644 --- a/arch/x86/include/asm/ftrace.h +++ b/arch/x86/include/asm/ftrace.h @@ -38,11 +38,10 @@ extern void mcount(void); static inline unsigned long ftrace_call_adjust(unsigned long addr) { /* - * call mcount is "e8 <4 byte offset>" - * The addr points to the 4 byte offset and the caller of this - * function wants the pointer to e8. Simply subtract one. + * addr is the address of the mcount call instruction. + * recordmcount does the necessary offset calculation. */ - return addr - 1; + return addr; } #ifdef CONFIG_DYNAMIC_FTRACE diff --git a/scripts/recordmcount.c b/scripts/recordmcount.c index 0e18975824f7..7648a5d11154 100644 --- a/scripts/recordmcount.c +++ b/scripts/recordmcount.c @@ -335,6 +335,7 @@ do_file(char const *const fname) reltype = R_386_32; make_nop = make_nop_x86; ideal_nop = ideal_nop5_x86_32; + mcount_adjust_32 = -1; break; case EM_ARM: reltype = R_ARM_ABS32; altmcount = "__gnu_mcount_nc"; @@ -350,6 +351,7 @@ do_file(char const *const fname) make_nop = make_nop_x86; ideal_nop = ideal_nop5_x86_64; reltype = R_X86_64_64; + mcount_adjust_64 = -1; break; } /* end switch */ diff --git a/scripts/recordmcount.pl b/scripts/recordmcount.pl index a871cd414055..414e7f5e42ec 100755 --- a/scripts/recordmcount.pl +++ b/scripts/recordmcount.pl @@ -223,6 +223,7 @@ if ($arch eq "x86_64") { $mcount_regex = "^\\s*([0-9a-fA-F]+):.*\\smcount([+-]0x[0-9a-zA-Z]+)?\$"; $type = ".quad"; $alignment = 8; + $mcount_adjust = -1; # force flags for this arch $ld .= " -m elf_x86_64"; @@ -232,6 +233,7 @@ if ($arch eq "x86_64") { } elsif ($arch eq "i386") { $alignment = 4; + $mcount_adjust = -1; # force flags for this arch $ld .= " -m elf_i386"; -- cgit v1.2.3 From 724a92ee45c04cb9d82884a856b03b1e594d9de1 Mon Sep 17 00:00:00 2001 From: Fenghua Yu Date: Tue, 17 May 2011 15:29:10 -0700 Subject: x86, cpufeature: Add CPU feature bit for enhanced REP MOVSB/STOSB Intel processors are adding enhancements to REP MOVSB/STOSB and the use of REP MOVSB/STOSB for optimal memcpy/memset or similar functions is recommended. Enhancement availability is indicated by CPUID.7.0.EBX[9] (Enhanced REP MOVSB/ STOSB). Signed-off-by: Fenghua Yu Link: http://lkml.kernel.org/r/1305671358-14478-2-git-send-email-fenghua.yu@intel.com Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/cpufeature.h | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index 91f3e087cf21..7f2f7b123293 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -195,6 +195,7 @@ /* Intel-defined CPU features, CPUID level 0x00000007:0 (ebx), word 9 */ #define X86_FEATURE_FSGSBASE (9*32+ 0) /* {RD/WR}{FS/GS}BASE instructions*/ +#define X86_FEATURE_ERMS (9*32+ 9) /* Enhanced REP MOVSB/STOSB */ #if defined(__KERNEL__) && !defined(__ASSEMBLY__) -- cgit v1.2.3 From 9072d11da15a71e086eab3b5085184f2c1d06913 Mon Sep 17 00:00:00 2001 From: Fenghua Yu Date: Tue, 17 May 2011 15:29:13 -0700 Subject: x86, alternative: Add altinstruction_entry macro Add altinstruction_entry macro to generate .altinstructions section entries from assembly code. This should be less failure-prone than open-coding. Signed-off-by: Fenghua Yu Link: http://lkml.kernel.org/r/1305671358-14478-5-git-send-email-fenghua.yu@intel.com Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/alternative-asm.h | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/alternative-asm.h b/arch/x86/include/asm/alternative-asm.h index a63a68be1cce..94d420b360d1 100644 --- a/arch/x86/include/asm/alternative-asm.h +++ b/arch/x86/include/asm/alternative-asm.h @@ -15,4 +15,13 @@ .endm #endif +.macro altinstruction_entry orig alt feature orig_len alt_len + .align 8 + .quad \orig + .quad \alt + .word \feature + .byte \orig_len + .byte \alt_len +.endm + #endif /* __ASSEMBLY__ */ -- cgit v1.2.3 From 26afb7c661080ae3f1f13ddf7f0c58c4f931c22b Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Thu, 12 May 2011 16:30:30 +0200 Subject: x86, 64-bit: Fix copy_[to/from]_user() checks for the userspace address limit As reported in BZ #30352: https://bugzilla.kernel.org/show_bug.cgi?id=30352 there's a kernel bug related to reading the last allowed page on x86_64. The _copy_to_user() and _copy_from_user() functions use the following check for address limit: if (buf + size >= limit) fail(); while it should be more permissive: if (buf + size > limit) fail(); That's because the size represents the number of bytes being read/write from/to buf address AND including the buf address. So the copy function will actually never touch the limit address even if "buf + size == limit". Following program fails to use the last page as buffer due to the wrong limit check: #include #include #include #define PAGE_SIZE (4096) #define LAST_PAGE ((void*)(0x7fffffffe000)) int main() { int fds[2], err; void * ptr = mmap(LAST_PAGE, PAGE_SIZE, PROT_READ | PROT_WRITE, MAP_ANONYMOUS | MAP_PRIVATE | MAP_FIXED, -1, 0); assert(ptr == LAST_PAGE); err = socketpair(AF_LOCAL, SOCK_STREAM, 0, fds); assert(err == 0); err = send(fds[0], ptr, PAGE_SIZE, 0); perror("send"); assert(err == PAGE_SIZE); err = recv(fds[1], ptr, PAGE_SIZE, MSG_WAITALL); perror("recv"); assert(err == PAGE_SIZE); return 0; } The other place checking the addr limit is the access_ok() function, which is working properly. There's just a misleading comment for the __range_not_ok() macro - which this patch fixes as well. The last page of the user-space address range is a guard page and Brian Gerst observed that the guard page itself due to an erratum on K8 cpus (#121 Sequential Execution Across Non-Canonical Boundary Causes Processor Hang). However, the test code is using the last valid page before the guard page. The bug is that the last byte before the guard page can't be read because of the off-by-one error. The guard page is left in place. This bug would normally not show up because the last page is part of the process stack and never accessed via syscalls. Signed-off-by: Jiri Olsa Acked-by: Brian Gerst Acked-by: Linus Torvalds Cc: Link: http://lkml.kernel.org/r/1305210630-7136-1-git-send-email-jolsa@redhat.com Signed-off-by: Ingo Molnar --- arch/x86/include/asm/uaccess.h | 2 +- arch/x86/lib/copy_user_64.S | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'arch/x86/include/asm') diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h index abd3e0ea762a..99f0ad753f32 100644 --- a/arch/x86/include/asm/uaccess.h +++ b/arch/x86/include/asm/uaccess.h @@ -42,7 +42,7 @@ * Returns 0 if the range is valid, nonzero otherwise. * * This is equivalent to the following test: - * (u33)addr + (u33)size >= (u33)current->addr_limit.seg (u65 for x86_64) + * (u33)addr + (u33)size > (u33)current->addr_limit.seg (u65 for x86_64) * * This needs 33-bit (65-bit for x86_64) arithmetic. We have a carry... */ diff --git a/arch/x86/lib/copy_user_64.S b/arch/x86/lib/copy_user_64.S index d17a1170bbf5..024840266ba0 100644 --- a/arch/x86/lib/copy_user_64.S +++ b/arch/x86/lib/copy_user_64.S @@ -79,7 +79,7 @@ ENTRY(_copy_to_user) addq %rdx,%rcx jc bad_to_user cmpq TI_addr_limit(%rax),%rcx - jae bad_to_user + ja bad_to_user ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,X86_FEATURE_ERMS, \ copy_user_generic_unrolled,copy_user_generic_string, \ copy_user_enhanced_fast_string @@ -94,7 +94,7 @@ ENTRY(_copy_from_user) addq %rdx,%rcx jc bad_from_user cmpq TI_addr_limit(%rax),%rcx - jae bad_from_user + ja bad_from_user ALTERNATIVE_JUMP X86_FEATURE_REP_GOOD,X86_FEATURE_ERMS, \ copy_user_generic_unrolled,copy_user_generic_string, \ copy_user_enhanced_fast_string -- cgit v1.2.3