From 0c54dd341fb701928b8e5dca91ced1870c55b05b Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 25 Feb 2010 08:42:06 -0500 Subject: ftrace: Remove memory barriers from NMI code when not needed The code in stop_machine that modifies the kernel text has a bit of logic to handle the case of NMIs. stop_machine does not prevent NMIs from executing, and if an NMI were to trigger on another CPU as the modifying CPU is changing the NMI text, a GPF could result. To prevent the GPF, the NMI calls ftrace_nmi_enter() which may modify the code first, then any other NMIs will just change the text to the same content which will do no harm. The code that stop_machine called must wait for NMIs to finish while it changes each location in the kernel. That code may also change the text to what the NMI changed it to. The key is that the text will never change content while another CPU is executing it. To make the above work, the call to ftrace_nmi_enter() must also do a smp_mb() as well as atomic_inc(). But for applications like perf that require a high number of NMIs for profiling, this can have a dramatic effect on the system. Not only is it doing a full memory barrier on both nmi_enter() as well as nmi_exit() it is also modifying a global variable with an atomic operation. This kills performance on large SMP machines. Since the memory barriers are only needed when ftrace is in the process of modifying the text (which is seldom), this patch adds a "modifying_code" variable that gets set before stop machine is executed and cleared afterwards. The NMIs will check this variable and store it in a per CPU "save_modifying_code" variable that it will use to check if it needs to do the memory barriers and atomic dec on NMI exit. Acked-by: Peter Zijlstra Signed-off-by: Steven Rostedt --- arch/x86/kernel/ftrace.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 309689245431..605ef196fdd6 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -30,14 +30,32 @@ #ifdef CONFIG_DYNAMIC_FTRACE +/* + * modifying_code is set to notify NMIs that they need to use + * memory barriers when entering or exiting. But we don't want + * to burden NMIs with unnecessary memory barriers when code + * modification is not being done (which is most of the time). + * + * A mutex is already held when ftrace_arch_code_modify_prepare + * and post_process are called. No locks need to be taken here. + * + * Stop machine will make sure currently running NMIs are done + * and new NMIs will see the updated variable before we need + * to worry about NMIs doing memory barriers. + */ +static int modifying_code __read_mostly; +static DEFINE_PER_CPU(int, save_modifying_code); + int ftrace_arch_code_modify_prepare(void) { set_kernel_text_rw(); + modifying_code = 1; return 0; } int ftrace_arch_code_modify_post_process(void) { + modifying_code = 0; set_kernel_text_ro(); return 0; } @@ -149,6 +167,11 @@ static void ftrace_mod_code(void) void ftrace_nmi_enter(void) { + __get_cpu_var(save_modifying_code) = modifying_code; + + if (!__get_cpu_var(save_modifying_code)) + return; + if (atomic_inc_return(&nmi_running) & MOD_CODE_WRITE_FLAG) { smp_rmb(); ftrace_mod_code(); @@ -160,6 +183,9 @@ void ftrace_nmi_enter(void) void ftrace_nmi_exit(void) { + if (!__get_cpu_var(save_modifying_code)) + return; + /* Finish all executions before clearing nmi_running */ smp_mb(); atomic_dec(&nmi_running); -- cgit v1.2.3 From 86c38a31aa7f2dd6e74a262710bf8ebf7455acc5 Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Wed, 24 Feb 2010 13:59:23 -0500 Subject: tracing: Fix ftrace_event_call alignment for use with gcc 4.5 GCC 4.5 introduces behavior that forces the alignment of structures to use the largest possible value. The default value is 32 bytes, so if some structures are defined with a 4-byte alignment and others aren't declared with an alignment constraint at all - it will align at 32-bytes. For things like the ftrace events, this results in a non-standard array. When initializing the ftrace subsystem, we traverse the _ftrace_events section and call the initialization callback for each event. When the structures are misaligned, we could be treating another part of the structure (or the zeroed out space between them) as a function pointer. This patch forces the alignment for all the ftrace_event_call structures to 4 bytes. Without this patch, the kernel fails to boot very early when built with gcc 4.5. It's trivial to check the alignment of the members of the array, so it might be worthwhile to add something to the build system to do that automatically. Unfortunately, that only covers this case. I've asked one of the gcc developers about adding a warning when this condition is seen. Cc: stable@kernel.org Signed-off-by: Jeff Mahoney LKML-Reference: <4B85770B.6010901@suse.com> Signed-off-by: Steven Rostedt --- include/linux/syscalls.h | 6 ++++-- include/trace/ftrace.h | 3 ++- kernel/trace/trace.h | 3 ++- 3 files changed, 8 insertions(+), 4 deletions(-) diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 7b219696ad24..91bd7d78a07d 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -132,7 +132,8 @@ struct perf_event_attr; #define SYSCALL_TRACE_ENTER_EVENT(sname) \ static const struct syscall_metadata __syscall_meta_##sname; \ - static struct ftrace_event_call event_enter_##sname; \ + static struct ftrace_event_call \ + __attribute__((__aligned__(4))) event_enter_##sname; \ static struct trace_event enter_syscall_print_##sname = { \ .trace = print_syscall_enter, \ }; \ @@ -153,7 +154,8 @@ struct perf_event_attr; #define SYSCALL_TRACE_EXIT_EVENT(sname) \ static const struct syscall_metadata __syscall_meta_##sname; \ - static struct ftrace_event_call event_exit_##sname; \ + static struct ftrace_event_call \ + __attribute__((__aligned__(4))) event_exit_##sname; \ static struct trace_event exit_syscall_print_##sname = { \ .trace = print_syscall_exit, \ }; \ diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 09fd9afc0859..f23a0ca6910a 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -65,7 +65,8 @@ }; #undef DEFINE_EVENT #define DEFINE_EVENT(template, name, proto, args) \ - static struct ftrace_event_call event_##name + static struct ftrace_event_call \ + __attribute__((__aligned__(4))) event_##name #undef DEFINE_EVENT_PRINT #define DEFINE_EVENT_PRINT(template, name, proto, args, print) \ diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index b477fce41edf..fd05bcaf91b0 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -792,7 +792,8 @@ extern const char *__stop___trace_bprintk_fmt[]; #undef FTRACE_ENTRY #define FTRACE_ENTRY(call, struct_name, id, tstruct, print) \ - extern struct ftrace_event_call event_##call; + extern struct ftrace_event_call \ + __attribute__((__aligned__(4))) event_##call; #undef FTRACE_ENTRY_DUP #define FTRACE_ENTRY_DUP(call, struct_name, id, tstruct, print) \ FTRACE_ENTRY(call, struct_name, id, PARAMS(tstruct), PARAMS(print)) -- cgit v1.2.3 From 1ab83a89411556c4d5914dcf0d5da190178ae7db Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Feb 2010 15:28:14 +0800 Subject: tracing: Remove CONFIG_TRACE_POWER from kernel config The power tracer has been converted to power trace events. Acked-by: Frederic Weisbecker Signed-off-by: Li Zefan LKML-Reference: <4B84D50E.4070806@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/Kconfig | 9 --------- 1 file changed, 9 deletions(-) diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 6c22d8a2f289..ca2d3a8778b6 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -330,15 +330,6 @@ config BRANCH_TRACER Say N if unsure. -config POWER_TRACER - bool "Trace power consumption behavior" - depends on X86 - select GENERIC_TRACER - help - This tracer helps developers to analyze and optimize the kernel's - power management decisions, specifically the C-state and P-state - behavior. - config KSYM_TRACER bool "Trace read and write access on kernel memory locations" depends on HAVE_HW_BREAKPOINT -- cgit v1.2.3 From 6574658b3bc7c408581629de5efb809f125cce8c Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:22 +0800 Subject: tracing: Fix typo in prof_sysexit_enable() Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-1-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_syscalls.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 49cea70fbf6d..8cdda95da81a 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -603,7 +603,7 @@ int prof_sysexit_enable(struct ftrace_event_call *call) ret = register_trace_sys_exit(prof_syscall_exit); if (ret) { pr_info("event trace: Could not activate" - "syscall entry trace point"); + "syscall exit trace point"); } else { set_bit(num, enabled_prof_exit_syscalls); sys_prof_refcount_exit++; -- cgit v1.2.3 From a5efd925115cbc1f90195dca9a25f7b8daa10c37 Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:23 +0800 Subject: tracing: Fix typo of info text in trace_kprobe.c Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-2-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_kprobe.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index c99029916c76..8d4bd16d6f70 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -651,12 +651,12 @@ static int create_trace_probe(int argc, char **argv) event = strchr(group, '/') + 1; event[-1] = '\0'; if (strlen(group) == 0) { - pr_info("Group name is not specifiled\n"); + pr_info("Group name is not specified\n"); return -EINVAL; } } if (strlen(event) == 0) { - pr_info("Event name is not specifiled\n"); + pr_info("Event name is not specified\n"); return -EINVAL; } } -- cgit v1.2.3 From c85f3a91f84d5a85f179c2504bb7a39370c82b41 Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:24 +0800 Subject: tracing: Remove unnecessary variable in print_graph_return The "cpu" variable is declared at the start of the function and also within a branch, with the exact same initialization. Remove the local variable of the same name in the branch. Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-3-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_functions_graph.c | 1 - 1 file changed, 1 deletion(-) diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 616b135c9eb9..112561df2a0a 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -855,7 +855,6 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, int i; if (data) { - int cpu = iter->cpu; int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); /* -- cgit v1.2.3 From 7b60997f73865b019e595720185c85285ca3df9a Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:26 +0800 Subject: tracing: Simplify memory recycle of trace_define_field Discard freeing field->type since it is not necessary. Reviewed-by: Li Zefan Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-5-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_events.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index c2a3077b7353..3f972ad98d04 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -60,10 +60,8 @@ int trace_define_field(struct ftrace_event_call *call, const char *type, return 0; err: - if (field) { + if (field) kfree(field->name); - kfree(field->type); - } kfree(field); return -ENOMEM; -- cgit v1.2.3