diff options
author | Steven Rostedt <srostedt@redhat.com> | 2010-02-26 17:08:16 -0500 |
---|---|---|
committer | Steven Rostedt <rostedt@goodmis.org> | 2010-02-26 19:25:53 -0500 |
commit | f1c7f517a5dc23bce07efa5ed55e2c074ed9d4ba (patch) | |
tree | 187be4d66aaa35db2eb1cd77b49520868e999b67 /kernel | |
parent | 281b3714e91162b66add1cfac404cf7b81e3e2f2 (diff) |
ftrace: Add function names to dangling } in function graph tracer
The function graph tracer is currently the most invasive tracer
in the ftrace family. It can easily overflow the buffer even with
10megs per CPU. This means that events can often be lost.
On start up, or after events are lost, if the function return is
recorded but the function enter was lost, all we get to see is the
exiting '}'.
Here is how a typical trace output starts:
[tracing] cat trace
# tracer: function_graph
#
# CPU DURATION FUNCTION CALLS
# | | | | | | |
0) + 91.897 us | }
0) ! 567.961 us | }
0) <========== |
0) ! 579.083 us | _raw_spin_lock_irqsave();
0) 4.694 us | _raw_spin_unlock_irqrestore();
0) ! 594.862 us | }
0) ! 603.361 us | }
0) ! 613.574 us | }
0) ! 623.554 us | }
0) 3.653 us | fget_light();
0) | sock_poll() {
There are a series of '}' with no matching "func() {". There's no information
to what functions these ending brackets belong to.
This patch adds a stack on the per cpu structure used in outputting
the function graph tracer to keep track of what function was outputted.
Then on a function exit event, it checks the depth to see if the
function exit has a matching entry event. If it does, then it only
prints the '}', otherwise it adds the function name after the '}'.
This allows function exit events to show what function they belong to
at trace output startup, when the entry was lost due to ring buffer
overflow, or even after a new task is scheduled in.
Here is what the above trace will look like after this patch:
[tracing] cat trace
# tracer: function_graph
#
# CPU DURATION FUNCTION CALLS
# | | | | | | |
0) + 91.897 us | } (irq_exit)
0) ! 567.961 us | } (smp_apic_timer_interrupt)
0) <========== |
0) ! 579.083 us | _raw_spin_lock_irqsave();
0) 4.694 us | _raw_spin_unlock_irqrestore();
0) ! 594.862 us | } (add_wait_queue)
0) ! 603.361 us | } (__pollwait)
0) ! 613.574 us | } (tcp_poll)
0) ! 623.554 us | } (sock_poll)
0) 3.653 us | fget_light();
0) | sock_poll() {
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Diffstat (limited to 'kernel')
-rw-r--r-- | kernel/trace/trace_functions_graph.c | 52 |
1 files changed, 43 insertions, 9 deletions
diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 112561df2a0a..e998a824e9db 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c | |||
@@ -18,6 +18,7 @@ struct fgraph_cpu_data { | |||
18 | pid_t last_pid; | 18 | pid_t last_pid; |
19 | int depth; | 19 | int depth; |
20 | int ignore; | 20 | int ignore; |
21 | unsigned long enter_funcs[FTRACE_RETFUNC_DEPTH]; | ||
21 | }; | 22 | }; |
22 | 23 | ||
23 | struct fgraph_data { | 24 | struct fgraph_data { |
@@ -670,15 +671,21 @@ print_graph_entry_leaf(struct trace_iterator *iter, | |||
670 | duration = graph_ret->rettime - graph_ret->calltime; | 671 | duration = graph_ret->rettime - graph_ret->calltime; |
671 | 672 | ||
672 | if (data) { | 673 | if (data) { |
674 | struct fgraph_cpu_data *cpu_data; | ||
673 | int cpu = iter->cpu; | 675 | int cpu = iter->cpu; |
674 | int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); | 676 | |
677 | cpu_data = per_cpu_ptr(data->cpu_data, cpu); | ||
675 | 678 | ||
676 | /* | 679 | /* |
677 | * Comments display at + 1 to depth. Since | 680 | * Comments display at + 1 to depth. Since |
678 | * this is a leaf function, keep the comments | 681 | * this is a leaf function, keep the comments |
679 | * equal to this depth. | 682 | * equal to this depth. |
680 | */ | 683 | */ |
681 | *depth = call->depth - 1; | 684 | cpu_data->depth = call->depth - 1; |
685 | |||
686 | /* No need to keep this function around for this depth */ | ||
687 | if (call->depth < FTRACE_RETFUNC_DEPTH) | ||
688 | cpu_data->enter_funcs[call->depth] = 0; | ||
682 | } | 689 | } |
683 | 690 | ||
684 | /* Overhead */ | 691 | /* Overhead */ |
@@ -718,10 +725,15 @@ print_graph_entry_nested(struct trace_iterator *iter, | |||
718 | int i; | 725 | int i; |
719 | 726 | ||
720 | if (data) { | 727 | if (data) { |
728 | struct fgraph_cpu_data *cpu_data; | ||
721 | int cpu = iter->cpu; | 729 | int cpu = iter->cpu; |
722 | int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); | ||
723 | 730 | ||
724 | *depth = call->depth; | 731 | cpu_data = per_cpu_ptr(data->cpu_data, cpu); |
732 | cpu_data->depth = call->depth; | ||
733 | |||
734 | /* Save this function pointer to see if the exit matches */ | ||
735 | if (call->depth < FTRACE_RETFUNC_DEPTH) | ||
736 | cpu_data->enter_funcs[call->depth] = call->func; | ||
725 | } | 737 | } |
726 | 738 | ||
727 | /* No overhead */ | 739 | /* No overhead */ |
@@ -851,18 +863,28 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, | |||
851 | struct fgraph_data *data = iter->private; | 863 | struct fgraph_data *data = iter->private; |
852 | pid_t pid = ent->pid; | 864 | pid_t pid = ent->pid; |
853 | int cpu = iter->cpu; | 865 | int cpu = iter->cpu; |
866 | int func_match = 1; | ||
854 | int ret; | 867 | int ret; |
855 | int i; | 868 | int i; |
856 | 869 | ||
857 | if (data) { | 870 | if (data) { |
858 | int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); | 871 | struct fgraph_cpu_data *cpu_data; |
872 | int cpu = iter->cpu; | ||
873 | |||
874 | cpu_data = per_cpu_ptr(data->cpu_data, cpu); | ||
859 | 875 | ||
860 | /* | 876 | /* |
861 | * Comments display at + 1 to depth. This is the | 877 | * Comments display at + 1 to depth. This is the |
862 | * return from a function, we now want the comments | 878 | * return from a function, we now want the comments |
863 | * to display at the same level of the bracket. | 879 | * to display at the same level of the bracket. |
864 | */ | 880 | */ |
865 | *depth = trace->depth - 1; | 881 | cpu_data->depth = trace->depth - 1; |
882 | |||
883 | if (trace->depth < FTRACE_RETFUNC_DEPTH) { | ||
884 | if (cpu_data->enter_funcs[trace->depth] != trace->func) | ||
885 | func_match = 0; | ||
886 | cpu_data->enter_funcs[trace->depth] = 0; | ||
887 | } | ||
866 | } | 888 | } |
867 | 889 | ||
868 | if (print_graph_prologue(iter, s, 0, 0)) | 890 | if (print_graph_prologue(iter, s, 0, 0)) |
@@ -887,9 +909,21 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, | |||
887 | return TRACE_TYPE_PARTIAL_LINE; | 909 | return TRACE_TYPE_PARTIAL_LINE; |
888 | } | 910 | } |
889 | 911 | ||
890 | ret = trace_seq_printf(s, "}\n"); | 912 | /* |
891 | if (!ret) | 913 | * If the return function does not have a matching entry, |
892 | return TRACE_TYPE_PARTIAL_LINE; | 914 | * then the entry was lost. Instead of just printing |
915 | * the '}' and letting the user guess what function this | ||
916 | * belongs to, write out the function name. | ||
917 | */ | ||
918 | if (func_match) { | ||
919 | ret = trace_seq_printf(s, "}\n"); | ||
920 | if (!ret) | ||
921 | return TRACE_TYPE_PARTIAL_LINE; | ||
922 | } else { | ||
923 | ret = trace_seq_printf(s, "} (%ps)\n", (void *)trace->func); | ||
924 | if (!ret) | ||
925 | return TRACE_TYPE_PARTIAL_LINE; | ||
926 | } | ||
893 | 927 | ||
894 | /* Overrun */ | 928 | /* Overrun */ |
895 | if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERRUN) { | 929 | if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERRUN) { |