From 890252823766e562301e61340f3187a14033d045 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Mon, 26 Jan 2009 18:28:02 +0300 Subject: x86: ftrace - simplify wait_for_nmi Get rid of 'waited' stack variable. Signed-off-by: Cyrill Gorcunov Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 1b43086b097..4d33224c055 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -133,15 +133,14 @@ void ftrace_nmi_exit(void) static void wait_for_nmi(void) { - int waited = 0; + if (!atomic_read(&in_nmi)) + return; - while (atomic_read(&in_nmi)) { - waited = 1; + do { cpu_relax(); - } + } while(atomic_read(&in_nmi)); - if (waited) - nmi_wait_count++; + nmi_wait_count++; } static int -- cgit v1.2.3 From 78d904b46a72fcf15ea6a39672bbef92953876b5 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 5 Feb 2009 18:43:07 -0500 Subject: ring-buffer: add NMI protection for spinlocks Impact: prevent deadlock in NMI The ring buffers are not yet totally lockless with writing to the buffer. When a writer crosses a page, it grabs a per cpu spinlock to protect against a reader. The spinlocks taken by a writer are not to protect against other writers, since a writer can only write to its own per cpu buffer. The spinlocks protect against readers that can touch any cpu buffer. The writers are made to be reentrant with the spinlocks disabling interrupts. The problem arises when an NMI writes to the buffer, and that write crosses a page boundary. If it grabs a spinlock, it can be racing with another writer (since disabling interrupts does not protect against NMIs) or with a reader on the same CPU. Luckily, most of the users are not reentrant and protects against this issue. But if a user of the ring buffer becomes reentrant (which is what the ring buffers do allow), if the NMI also writes to the ring buffer then we risk the chance of a deadlock. This patch moves the ftrace_nmi_enter called by nmi_enter() to the ring buffer code. It replaces the current ftrace_nmi_enter that is used by arch specific code to arch_ftrace_nmi_enter and updates the Kconfig to handle it. When an NMI is called, it will set a per cpu variable in the ring buffer code and will clear it when the NMI exits. If a write to the ring buffer crosses page boundaries inside an NMI, a trylock is used on the spin lock instead. If the spinlock fails to be acquired, then the entry is discarded. This bug appeared in the ftrace work in the RT tree, where event tracing is reentrant. This workaround solved the deadlocks that appeared there. Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 4d33224c055..4c683587055 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -113,7 +113,7 @@ static void ftrace_mod_code(void) MCOUNT_INSN_SIZE); } -void ftrace_nmi_enter(void) +void arch_ftrace_nmi_enter(void) { atomic_inc(&in_nmi); /* Must have in_nmi seen before reading write flag */ @@ -124,7 +124,7 @@ void ftrace_nmi_enter(void) } } -void ftrace_nmi_exit(void) +void arch_ftrace_nmi_exit(void) { /* Finish all executions before clearing in_nmi */ smp_wmb(); @@ -376,12 +376,12 @@ int ftrace_disable_ftrace_graph_caller(void) */ static atomic_t in_nmi; -void ftrace_nmi_enter(void) +void arch_ftrace_nmi_enter(void) { atomic_inc(&in_nmi); } -void ftrace_nmi_exit(void) +void arch_ftrace_nmi_exit(void) { atomic_dec(&in_nmi); } -- cgit v1.2.3 From 4e6ea1440c67de32d7c89aacf233472dfc3bce82 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 5 Feb 2009 22:30:07 -0500 Subject: ftrace, x86: rename in_nmi variable Impact: clean up The in_nmi variable in x86 arch ftrace.c is a misnomer. Andrew Morton pointed out that the in_nmi variable is incremented by all CPUS. It can be set when another CPU is running an NMI. Since this is actually intentional, the fix is to rename it to what it really is: "nmi_running" Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 4c683587055..e3fad2ef622 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -82,7 +82,7 @@ static unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) * are the same as what exists. */ -static atomic_t in_nmi = ATOMIC_INIT(0); +static atomic_t nmi_running = ATOMIC_INIT(0); static int mod_code_status; /* holds return value of text write */ static int mod_code_write; /* set when NMI should do the write */ static void *mod_code_ip; /* holds the IP to write to */ @@ -115,8 +115,8 @@ static void ftrace_mod_code(void) void arch_ftrace_nmi_enter(void) { - atomic_inc(&in_nmi); - /* Must have in_nmi seen before reading write flag */ + atomic_inc(&nmi_running); + /* Must have nmi_running seen before reading write flag */ smp_mb(); if (mod_code_write) { ftrace_mod_code(); @@ -126,19 +126,19 @@ void arch_ftrace_nmi_enter(void) void arch_ftrace_nmi_exit(void) { - /* Finish all executions before clearing in_nmi */ + /* Finish all executions before clearing nmi_running */ smp_wmb(); - atomic_dec(&in_nmi); + atomic_dec(&nmi_running); } static void wait_for_nmi(void) { - if (!atomic_read(&in_nmi)) + if (!atomic_read(&nmi_running)) return; do { cpu_relax(); - } while(atomic_read(&in_nmi)); + } while (atomic_read(&nmi_running)); nmi_wait_count++; } @@ -374,16 +374,16 @@ int ftrace_disable_ftrace_graph_caller(void) * this page for dynamic ftrace. They have been * simplified to ignore all traces in NMI context. */ -static atomic_t in_nmi; +static atomic_t nmi_running; void arch_ftrace_nmi_enter(void) { - atomic_inc(&in_nmi); + atomic_inc(&nmi_running); } void arch_ftrace_nmi_exit(void) { - atomic_dec(&in_nmi); + atomic_dec(&nmi_running); } #endif /* !CONFIG_DYNAMIC_FTRACE */ @@ -475,7 +475,7 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) &return_to_handler; /* Nmi's are currently unsupported */ - if (unlikely(atomic_read(&in_nmi))) + if (unlikely(atomic_read(&nmi_running))) return; if (unlikely(atomic_read(¤t->tracing_graph_pause))) -- cgit v1.2.3 From 9a5fd902273d01170fd033691bd70b142baa7309 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 6 Feb 2009 01:14:26 -0500 Subject: ftrace: change function graph tracer to use new in_nmi The function graph tracer piggy backed onto the dynamic ftracer to use the in_nmi custom code for dynamic tracing. The problem was (as Andrew Morton pointed out) it really only wanted to bail out if the context of the current CPU was in NMI context. But the dynamic ftrace in_nmi custom code was true if _any_ CPU happened to be in NMI context. Now that we have a generic in_nmi interface, this patch changes the function graph code to use it instead of the dynamic ftarce custom code. Reported-by: Andrew Morton Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 21 +-------------------- 1 file changed, 1 insertion(+), 20 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index e3fad2ef622..918073c6681 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -367,25 +367,6 @@ int ftrace_disable_ftrace_graph_caller(void) return ftrace_mod_jmp(ip, old_offset, new_offset); } -#else /* CONFIG_DYNAMIC_FTRACE */ - -/* - * These functions are picked from those used on - * this page for dynamic ftrace. They have been - * simplified to ignore all traces in NMI context. - */ -static atomic_t nmi_running; - -void arch_ftrace_nmi_enter(void) -{ - atomic_inc(&nmi_running); -} - -void arch_ftrace_nmi_exit(void) -{ - atomic_dec(&nmi_running); -} - #endif /* !CONFIG_DYNAMIC_FTRACE */ /* Add a function return address to the trace stack on thread info.*/ @@ -475,7 +456,7 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) &return_to_handler; /* Nmi's are currently unsupported */ - if (unlikely(atomic_read(&nmi_running))) + if (unlikely(in_nmi())) return; if (unlikely(atomic_read(¤t->tracing_graph_pause))) -- cgit v1.2.3 From a81bd80a0b0a405dc0483e2c428332d69da2c79f Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 6 Feb 2009 01:45:16 -0500 Subject: ring-buffer: use generic version of in_nmi Impact: clean up Now that a generic in_nmi is available, this patch removes the special code in the ring_buffer and implements the in_nmi generic version instead. With this change, I was also able to rename the "arch_ftrace_nmi_enter" back to "ftrace_nmi_enter" and remove the code from the ring buffer. Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 918073c6681..d74d75e0952 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -113,7 +113,7 @@ static void ftrace_mod_code(void) MCOUNT_INSN_SIZE); } -void arch_ftrace_nmi_enter(void) +void ftrace_nmi_enter(void) { atomic_inc(&nmi_running); /* Must have nmi_running seen before reading write flag */ @@ -124,7 +124,7 @@ void arch_ftrace_nmi_enter(void) } } -void arch_ftrace_nmi_exit(void) +void ftrace_nmi_exit(void) { /* Finish all executions before clearing nmi_running */ smp_wmb(); -- cgit v1.2.3 From 3861a17bcc0af815f684c6178bc9ec2d790c350e Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 8 Feb 2009 00:04:02 +0100 Subject: tracing/function-graph-tracer: drop the kernel_text_address check When the function graph tracer picks a return address, it ensures this address is really a kernel text one by calling __kernel_text_address() Actually this path has never been taken.Its role was more likely to debug the tracer on the beginning of its development but this function is wasteful since it is called for every traced function. The fault check is already sufficient. Signed-off-by: Frederic Weisbecker Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 7 ------- 1 file changed, 7 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index d74d75e0952..18828aee878 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -491,13 +491,6 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) return; } - if (unlikely(!__kernel_text_address(old))) { - ftrace_graph_stop(); - *parent = old; - WARN_ON(1); - return; - } - calltime = cpu_clock(raw_smp_processor_id()); if (push_return_trace(old, calltime, -- cgit v1.2.3 From 966657883fdc3a2883a5e641ca4ec8f79ffb8ecd Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 10 Feb 2009 11:53:23 -0500 Subject: tracing, x86: fix constraint for parent variable The constraint used for retrieving and restoring the parent function pointer is incorrect. The parent variable is a pointer, and the address of the pointer is modified by the asm statement and not the pointer itself. It is incorrect to pass it in as an output constraint since the asm will never update the pointer. Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 18828aee878..370bafaa43a 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -468,8 +468,8 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) * ignore such a protection. */ asm volatile( - "1: " _ASM_MOV " (%[parent_old]), %[old]\n" - "2: " _ASM_MOV " %[return_hooker], (%[parent_replaced])\n" + "1: " _ASM_MOV " (%[parent]), %[old]\n" + "2: " _ASM_MOV " %[return_hooker], (%[parent])\n" " movl $0, %[faulted]\n" ".section .fixup, \"ax\"\n" @@ -479,9 +479,8 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) _ASM_EXTABLE(1b, 3b) _ASM_EXTABLE(2b, 3b) - : [parent_replaced] "=r" (parent), [old] "=r" (old), - [faulted] "=r" (faulted) - : [parent_old] "0" (parent), [return_hooker] "r" (return_hooker) + : [old] "=r" (old), [faulted] "=r" (faulted) + : [parent] "r" (parent), [return_hooker] "r" (return_hooker) : "memory" ); -- cgit v1.2.3 From 16239630974516a8879a3695ee9b4dc661f79f96 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 17 Feb 2009 17:57:30 -0500 Subject: ftrace, x86: make kernel text writable only for conversions Impact: keep kernel text read only Because dynamic ftrace converts the calls to mcount into and out of nops at run time, we needed to always keep the kernel text writable. But this defeats the point of CONFIG_DEBUG_RODATA. This patch converts the kernel code to writable before ftrace modifies the text, and converts it back to read only afterward. The kernel text is converted to read/write, stop_machine is called to modify the code, then the kernel text is converted back to read only. The original version used SYSTEM_STATE to determine when it was OK or not to change the code to rw or ro. Andrew Morton pointed out that using SYSTEM_STATE is a bad idea since there is no guarantee to what its state will actually be. Instead, I moved the check into the set_kernel_text_* functions themselves, and use a local variable to determine when it is OK to change the kernel text RW permissions. [ Update: Ingo Molnar suggested moving the prototypes to cacheflush.h ] Reviewed-by: Andrew Morton Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 231bdd3c5b1..77857d4f7d0 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -18,6 +18,7 @@ #include #include +#include #include #include #include @@ -26,6 +27,18 @@ #ifdef CONFIG_DYNAMIC_FTRACE +int ftrace_arch_code_modify_prepare(void) +{ + set_kernel_text_rw(); + return 0; +} + +int ftrace_arch_code_modify_post_process(void) +{ + set_kernel_text_ro(); + return 0; +} + union ftrace_code_union { char code[MCOUNT_INSN_SIZE]; struct { -- cgit v1.2.3 From 90c7ac49aa819feb9433b5310089fca6399881c0 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 19 Feb 2009 13:32:57 -0500 Subject: ftrace: immediately stop code modification if failure is detected Impact: fix to prevent NMI lockup If the page fault handler produces a WARN_ON in the modifying of text, and the system is setup to have a high frequency of NMIs, we can lock up the system on a failure to modify code. The modifying of code with NMIs allows all NMIs to modify the code if it is about to run. This prevents a modifier on one CPU from modifying code running in NMI context on another CPU. The modifying is done through stop_machine, so only NMIs must be considered. But if the write causes the page fault handler to produce a warning, the print can slow it down enough that as soon as it is done it will take another NMI before going back to the process context. The new NMI will perform the write again causing another print and this will hang the box. This patch turns off the writing as soon as a failure is detected and does not wait for it to be turned off by the process context. This will keep NMIs from getting stuck in this back and forth of print outs. Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 77857d4f7d0..c56d7389432 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -124,6 +124,10 @@ static void ftrace_mod_code(void) */ mod_code_status = probe_kernel_write(mod_code_ip, mod_code_newcode, MCOUNT_INSN_SIZE); + + /* if we fail, then kill any new writers */ + if (mod_code_status) + mod_code_write = 0; } void ftrace_nmi_enter(void) -- cgit v1.2.3 From 0012693ad4f636c720fed3802027f9427962f540 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 5 Mar 2009 01:49:22 +0100 Subject: tracing/function-graph-tracer: use the more lightweight local clock Impact: decrease hangs risks with the graph tracer on slow systems Since the function graph tracer can spend too much time on timer interrupts, it's better now to use the more lightweight local clock. Anyway, the function graph traces are more reliable on a per cpu trace. Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt Cc: Peter Zijlstra LKML-Reference: <49af243d.06e9300a.53ad.ffff840c@mx.google.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 3925ec0184b..a85da1764b1 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -436,7 +436,7 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) return; } - calltime = cpu_clock(raw_smp_processor_id()); + calltime = trace_clock_local(); if (ftrace_push_return_trace(old, calltime, self_addr, &trace.depth) == -EBUSY) { -- cgit v1.2.3 From f58ba100678f421bdcb000a3c71793f432dfab93 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 13 Mar 2009 15:42:12 +0100 Subject: tracing/syscalls: support for syscalls tracing on x86 Extend x86 architecture syscall tracing support with syscall metadata table details. (The upcoming core syscall tracing modifications rely on this.) Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt LKML-Reference: <1236955332-10133-3-git-send-email-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 63 insertions(+) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index a85da1764b1..1d0d7f42efe 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -453,3 +453,66 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) } } #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ + +#ifdef CONFIG_FTRACE_SYSCALLS + +extern unsigned long __start_syscalls_metadata[]; +extern unsigned long __stop_syscalls_metadata[]; +extern unsigned long *sys_call_table; + +static struct syscall_metadata **syscalls_metadata; + +static struct syscall_metadata *find_syscall_meta(unsigned long *syscall) +{ + struct syscall_metadata *start; + struct syscall_metadata *stop; + char str[KSYM_SYMBOL_LEN]; + + + start = (struct syscall_metadata *)__start_syscalls_metadata; + stop = (struct syscall_metadata *)__stop_syscalls_metadata; + kallsyms_lookup((unsigned long) syscall, NULL, NULL, NULL, str); + + for ( ; start < stop; start++) { + if (start->name && !strcmp(start->name, str)) + return start; + } + return NULL; +} + +struct syscall_metadata *syscall_nr_to_meta(int nr) +{ + if (!syscalls_metadata || nr >= FTRACE_SYSCALL_MAX || nr < 0) + return NULL; + + return syscalls_metadata[nr]; +} + +void arch_init_ftrace_syscalls(void) +{ + int i; + struct syscall_metadata *meta; + unsigned long **psys_syscall_table = &sys_call_table; + static atomic_t refs; + + if (atomic_inc_return(&refs) != 1) + goto end; + + syscalls_metadata = kzalloc(sizeof(*syscalls_metadata) * + FTRACE_SYSCALL_MAX, GFP_KERNEL); + if (!syscalls_metadata) { + WARN_ON(1); + return; + } + + for (i = 0; i < FTRACE_SYSCALL_MAX; i++) { + meta = find_syscall_meta(psys_syscall_table[i]); + syscalls_metadata[i] = meta; + } + return; + + /* Paranoid: avoid overflow */ +end: + atomic_dec(&refs); +} +#endif -- cgit v1.2.3 From e9d9df44736d116726f4596f7e2f9ce2764ffc0a Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Wed, 18 Mar 2009 16:42:57 +0800 Subject: ftrace: protect running nmi (V3) When I review the sensitive code ftrace_nmi_enter(), I found the atomic variable nmi_running does protect NMI VS do_ftrace_mod_code(), but it can not protects NMI(entered nmi) VS NMI(ftrace_nmi_enter()). cpu#1 | cpu#2 | cpu#3 ftrace_nmi_enter() | do_ftrace_mod_code() | not modify | | ------------------------|-----------------------|-- executing | set mod_code_write = 1| executing --|-----------------------|-------------------- executing | | ftrace_nmi_enter() executing | | do modify ------------------------|-----------------------|----------------- ftrace_nmi_exit() | | cpu#3 may be being modified the code which is still being executed on cpu#1, it will have undefined results and possibly take a GPF, this patch prevents it occurred. Signed-off-by: Lai Jiangshan LKML-Reference: <49C0B411.30003@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 63 ++++++++++++++++++++++++++++++------------------ 1 file changed, 40 insertions(+), 23 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 1d0d7f42efe..57b33edb7ce 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -79,11 +79,11 @@ static unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) * * 1) Put the instruction pointer into the IP buffer * and the new code into the "code" buffer. - * 2) Set a flag that says we are modifying code - * 3) Wait for any running NMIs to finish. - * 4) Write the code - * 5) clear the flag. - * 6) Wait for any running NMIs to finish. + * 2) Wait for any running NMIs to finish and set a flag that says + * we are modifying code, it is done in an atomic operation. + * 3) Write the code + * 4) clear the flag. + * 5) Wait for any running NMIs to finish. * * If an NMI is executed, the first thing it does is to call * "ftrace_nmi_enter". This will check if the flag is set to write @@ -95,9 +95,9 @@ static unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) * are the same as what exists. */ +#define MOD_CODE_WRITE_FLAG (1 << 31) /* set when NMI should do the write */ static atomic_t nmi_running = ATOMIC_INIT(0); static int mod_code_status; /* holds return value of text write */ -static int mod_code_write; /* set when NMI should do the write */ static void *mod_code_ip; /* holds the IP to write to */ static void *mod_code_newcode; /* holds the text to write to the IP */ @@ -114,6 +114,20 @@ int ftrace_arch_read_dyn_info(char *buf, int size) return r; } +static void clear_mod_flag(void) +{ + int old = atomic_read(&nmi_running); + + for (;;) { + int new = old & ~MOD_CODE_WRITE_FLAG; + + if (old == new) + break; + + old = atomic_cmpxchg(&nmi_running, old, new); + } +} + static void ftrace_mod_code(void) { /* @@ -127,27 +141,39 @@ static void ftrace_mod_code(void) /* if we fail, then kill any new writers */ if (mod_code_status) - mod_code_write = 0; + clear_mod_flag(); } void ftrace_nmi_enter(void) { - atomic_inc(&nmi_running); - /* Must have nmi_running seen before reading write flag */ - smp_mb(); - if (mod_code_write) { + if (atomic_inc_return(&nmi_running) & MOD_CODE_WRITE_FLAG) { + smp_rmb(); ftrace_mod_code(); atomic_inc(&nmi_update_count); } + /* Must have previous changes seen before executions */ + smp_mb(); } void ftrace_nmi_exit(void) { /* Finish all executions before clearing nmi_running */ - smp_wmb(); + smp_mb(); atomic_dec(&nmi_running); } +static void wait_for_nmi_and_set_mod_flag(void) +{ + if (!atomic_cmpxchg(&nmi_running, 0, MOD_CODE_WRITE_FLAG)) + return; + + do { + cpu_relax(); + } while (atomic_cmpxchg(&nmi_running, 0, MOD_CODE_WRITE_FLAG)); + + nmi_wait_count++; +} + static void wait_for_nmi(void) { if (!atomic_read(&nmi_running)) @@ -167,14 +193,9 @@ do_ftrace_mod_code(unsigned long ip, void *new_code) mod_code_newcode = new_code; /* The buffers need to be visible before we let NMIs write them */ - smp_wmb(); - - mod_code_write = 1; - - /* Make sure write bit is visible before we wait on NMIs */ smp_mb(); - wait_for_nmi(); + wait_for_nmi_and_set_mod_flag(); /* Make sure all running NMIs have finished before we write the code */ smp_mb(); @@ -182,13 +203,9 @@ do_ftrace_mod_code(unsigned long ip, void *new_code) ftrace_mod_code(); /* Make sure the write happens before clearing the bit */ - smp_wmb(); - - mod_code_write = 0; - - /* make sure NMIs see the cleared bit */ smp_mb(); + clear_mod_flag(); wait_for_nmi(); return mod_code_status; -- cgit v1.2.3 From 5d1a03dc541dc6672e60e57249ed22f40654ca47 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Mon, 23 Mar 2009 23:38:49 -0400 Subject: function-graph: moved the timestamp from arch to generic code This patch move the timestamp from happening in the arch specific code into the general code. This allows for better control by the tracer to time manipulation. Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) (limited to 'arch/x86/kernel/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 57b33edb7ce..61df7753212 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -410,7 +410,6 @@ int ftrace_disable_ftrace_graph_caller(void) void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) { unsigned long old; - unsigned long long calltime; int faulted; struct ftrace_graph_ent trace; unsigned long return_hooker = (unsigned long) @@ -453,10 +452,7 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) return; } - calltime = trace_clock_local(); - - if (ftrace_push_return_trace(old, calltime, - self_addr, &trace.depth) == -EBUSY) { + if (ftrace_push_return_trace(old, self_addr, &trace.depth) == -EBUSY) { *parent = old; return; } -- cgit v1.2.3