aboutsummaryrefslogtreecommitdiffstats
path: root/scripts
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2016-10-06 14:48:41 -0400
committerLinus Torvalds <torvalds@linux-foundation.org>2016-10-06 14:48:41 -0400
commit95107b30be68953e3a4f1c3994c2233500502ccf (patch)
treeed1824ce842d9813f2ee8a5fa8683e0d6b13f662 /scripts
parent541efb7632642cab55361178d73d544f025b593c (diff)
parenta0d0c6216afad4b2b1704a72bd76bea259e07655 (diff)
Merge tag 'trace-v4.9' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing updates from Steven Rostedt: "This release cycle is rather small. Just a few fixes to tracing. The big change is the addition of the hwlat tracer. It not only detects SMIs, but also other latency that's caused by the hardware. I have detected some latency from large boxes having bus contention" * tag 'trace-v4.9' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: tracing: Call traceoff trigger after event is recorded ftrace/scripts: Add helper script to bisect function tracing problem functions tracing: Have max_latency be defined for HWLAT_TRACER as well tracing: Add NMI tracing in hwlat detector tracing: Have hwlat trace migrate across tracing_cpumask CPUs tracing: Add documentation for hwlat_detector tracer tracing: Added hardware latency tracer ftrace: Access ret_stack->subtime only in the function profiler function_graph: Handle TRACE_BPUTS in print_graph_comment tracing/uprobe: Drop isdigit() check in create_trace_uprobe
Diffstat (limited to 'scripts')
-rwxr-xr-xscripts/tracing/ftrace-bisect.sh115
1 files changed, 115 insertions, 0 deletions
diff --git a/scripts/tracing/ftrace-bisect.sh b/scripts/tracing/ftrace-bisect.sh
new file mode 100755
index 000000000000..9ff8ac5fc53c
--- /dev/null
+++ b/scripts/tracing/ftrace-bisect.sh
@@ -0,0 +1,115 @@
1#!/bin/bash
2#
3# Here's how to use this:
4#
5# This script is used to help find functions that are being traced by function
6# tracer or function graph tracing that causes the machine to reboot, hang, or
7# crash. Here's the steps to take.
8#
9# First, determine if function tracing is working with a single function:
10#
11# (note, if this is a problem with function_graph tracing, then simply
12# replace "function" with "function_graph" in the following steps).
13#
14# # cd /sys/kernel/debug/tracing
15# # echo schedule > set_ftrace_filter
16# # echo function > current_tracer
17#
18# If this works, then we know that something is being traced that shouldn't be.
19#
20# # echo nop > current_tracer
21#
22# # cat available_filter_functions > ~/full-file
23# # ftrace-bisect ~/full-file ~/test-file ~/non-test-file
24# # cat ~/test-file > set_ftrace_filter
25#
26# *** Note *** this will take several minutes. Setting multiple functions is
27# an O(n^2) operation, and we are dealing with thousands of functions. So go
28# have coffee, talk with your coworkers, read facebook. And eventually, this
29# operation will end.
30#
31# # echo function > current_tracer
32#
33# If it crashes, we know that ~/test-file has a bad function.
34#
35# Reboot back to test kernel.
36#
37# # cd /sys/kernel/debug/tracing
38# # mv ~/test-file ~/full-file
39#
40# If it didn't crash.
41#
42# # echo nop > current_tracer
43# # mv ~/non-test-file ~/full-file
44#
45# Get rid of the other test file from previous run (or save them off somewhere).
46# # rm -f ~/test-file ~/non-test-file
47#
48# And start again:
49#
50# # ftrace-bisect ~/full-file ~/test-file ~/non-test-file
51#
52# The good thing is, because this cuts the number of functions in ~/test-file
53# by half, the cat of it into set_ftrace_filter takes half as long each
54# iteration, so don't talk so much at the water cooler the second time.
55#
56# Eventually, if you did this correctly, you will get down to the problem
57# function, and all we need to do is to notrace it.
58#
59# The way to figure out if the problem function is bad, just do:
60#
61# # echo <problem-function> > set_ftrace_notrace
62# # echo > set_ftrace_filter
63# # echo function > current_tracer
64#
65# And if it doesn't crash, we are done.
66#
67# If it does crash, do this again (there's more than one problem function)
68# but you need to echo the problem function(s) into set_ftrace_notrace before
69# enabling function tracing in the above steps. Or if you can compile the
70# kernel, annotate the problem functions with "notrace" and start again.
71#
72
73
74if [ $# -ne 3 ]; then
75 echo 'usage: ftrace-bisect full-file test-file non-test-file'
76 exit
77fi
78
79full=$1
80test=$2
81nontest=$3
82
83x=`cat $full | wc -l`
84if [ $x -eq 1 ]; then
85 echo "There's only one function left, must be the bad one"
86 cat $full
87 exit 0
88fi
89
90let x=$x/2
91let y=$x+1
92
93if [ ! -f $full ]; then
94 echo "$full does not exist"
95 exit 1
96fi
97
98if [ -f $test ]; then
99 echo -n "$test exists, delete it? [y/N]"
100 read a
101 if [ "$a" != "y" -a "$a" != "Y" ]; then
102 exit 1
103 fi
104fi
105
106if [ -f $nontest ]; then
107 echo -n "$nontest exists, delete it? [y/N]"
108 read a
109 if [ "$a" != "y" -a "$a" != "Y" ]; then
110 exit 1
111 fi
112fi
113
114sed -ne "1,${x}p" $full > $test
115sed -ne "$y,\$p" $full > $nontest