building libtraceevent with clang
PATCH 1/1 building libtraceevent with clang
Hi Steven,
I tried building perf (and thus libtraceevent) with clang and
got this one:
kbuffer-parse.c:312:7: warning: variable 'length' is used uninitialized whenever switch case is taken [-Wsometimes-uninitialized]
case OLD_RINGBUF_TYPE_TIME_EXTEND:
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
kbuffer-parse.c:339:29: note: uninitialized use occurs here
kbuf->next = kbuf->index + length;
^~~~~~
kbuffer-parse.c:297:21: note: initialize the variable 'length' to silence this warning
unsigned int length;
^
= 01 warning generated.
Please take a look if the following is what should be done:
tools/lib/traceevent/kbuffer-parse.c +8 -8
@@ -309,20 +309,20 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) kbuf->next = kbuf->size;return 0 ; -case OLD_RINGBUF_TYPE_TIME_EXTEND : - extend =read_4 (kbuf, ptr); - extend <<=TS_SHIFT ; - extend += delta; - delta = extend; - ptr +=4 ; -break ; -case OLD_RINGBUF_TYPE_TIME_STAMP :/* should never happen! */ kbuf->curr = kbuf->size; kbuf->next = kbuf->size; kbuf->index = kbuf->size;return -1 ; + +case OLD_RINGBUF_TYPE_TIME_EXTEND : + extend =read_4 (kbuf, ptr); + extend <<=TS_SHIFT ; + extend += delta; + delta = extend; + ptr +=4 ; +/* Fall through */ default :if (len) length = len *4 ;
Quoted 15 lines · On Fri, 10 Feb 2017 14:03:17 -0300 Arnaldo Carvalho de Melo <[email protected]>…
Hi Steven,
I tried building perf (and thus libtraceevent) with clang and
got this one:kbuffer-parse.c:312:7: warning: variable 'length' is used uninitialized whenever switch case is taken [-Wsometimes-uninitialized]
case OLD_RINGBUF_TYPE_TIME_EXTEND: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~ kbuffer-parse.c:339:29: note: uninitialized use occurs here kbuf->next = kbuf->index + length; ^~~~~~ kbuffer-parse.c:297:21: note: initialize the variable 'length' to silence this warning unsigned int length; ^ = 0
Quoted 1 line
1 warning generated.
Hmm, I'm surprised gcc never complained.
Quoted 24 lines
Please take a look if the following is what should be done:
tools/lib/traceevent/kbuffer-parse.c +0
-8@@ -309,20 +309,20 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) kbuf->next = kbuf->size;return 0 ; -case OLD_RINGBUF_TYPE_TIME_EXTEND : - extend =read_4 (kbuf, ptr); - extend <<=TS_SHIFT ; - extend += delta; - delta = extend; - ptr +=4 ; -break ; -case OLD_RINGBUF_TYPE_TIME_STAMP :/* should never happen! */ kbuf->curr = kbuf->size; kbuf->next = kbuf->size;
tools/lib/traceevent/kbuffer-parse.c static unsigned int old_update_pointers(struct kbuffer *kbuf) · line 325
Quoted 10 lines
tools/lib/traceevent/kbuffer-parse.c +8
-0kbuf->index = kbuf->size;return -1 ; + +case OLD_RINGBUF_TYPE_TIME_EXTEND : + extend =read_4 (kbuf, ptr); + extend <<=TS_SHIFT ; + extend += delta; + delta = extend; + ptr +=4 ; +/* Fall through */
No, actually, the length should be zeroed here.
length = 0;
As the ptr is placed passed the metadata. At the end of this case statement, the ptr will be after the metadata. The length is to represent the length of the data that is part of the event. The index is the ptr - start_data + length. As there's no data to this type of event, it should be zero.
When I add length = 0, I get more events from my old 2.6.30 trace.dat file :-/ I'll have to update my tests.
Care to send another patch?
-- Steve
Quoted 3 lines
tools/lib/traceevent/kbuffer-parse.c +0
-0default :if (len) length = len *4 ;
Em Fri, Feb 10, 2017 at 02:14:08PM -0500, Steven Rostedt escreveu:
Quoted 20 lines
Review
Quoted 7 lines · On Fri, 10 Feb 2017 14:03:17 -0300 Arnaldo Carvalho de Melo <[email protected]>…
(quoted diff) +7
-0+ case OLD_RINGBUF_TYPE_TIME_EXTEND: + extend = read_4(kbuf, ptr); + extend <<= TS_SHIFT; + extend += delta; + delta = extend; + ptr += 4; + /* Fall through */No, actually, the length should be zeroed here.
length = 0;As the ptr is placed passed the metadata. At the end of this case statement, the ptr will be after the metadata. The length is to represent the length of the data that is part of the event. The index is the ptr -
start_data+ length. As there's no data to this type of event, it should be zero.
Quoted 4 lines
When I add length = 0, I get more events from my old 2.6.30 trace.dat file :-/ I'll have to update my tests.
Care to send another patch?
I have this now in my tree, and since the patch is basically yours, I just made myself the reporter and you the author, please ack.
commit 1fb82b6cba967afde11957099a9ef440cf8c2248
Author: Steven Rostedt <[email protected]> Date: Mon Feb 13 13:18:29 2017 -0300
tools lib traceevent: Initialize lenght on OLD_RING_BUFFER_TYPE_TIME_STAMP
A undefined value was being used for the OLD_RING_BUFFER_TYPE_TIME_STAMP case entry, as the 'lenght' variable was not being initialized, fix it.
Caught by the reporter when building tools/perf/ using clang, which emmitted this warning:
kbuffer-parse.c:312:7: warning: variable 'length' is used uninitialized whenever switch case is taken [-Wsometimes-uninitialized]
case OLD_RINGBUF_TYPE_TIME_EXTEND:
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
kbuffer-parse.c:339:29: note: uninitialized use occurs here
kbuf->next = kbuf->index + length;
^~~~~~
kbuffer-parse.c:297:21: note: initialize the variable 'length' to silence this warning
unsigned int length;
^
= 0
tools/lib/traceevent/kbuffer-parse.c +1 -0
@@ -315,6 +315,7 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) extend += delta; delta = extend; ptr +=4 ; + lenght =0 ;break ;case OLD_RINGBUF_TYPE_TIME_STAMP :
Em Mon, Feb 13, 2017 at 01:24:55PM -0300, Arnaldo Carvalho de Melo escreveu:
Quoted 67 lines
Em Fri, Feb 10, 2017 at 02:14:08PM -0500, Steven Rostedt escreveu:
Quoted 20 lines
Review
Quoted 7 lines · On Fri, 10 Feb 2017 14:03:17 -0300 Arnaldo Carvalho de Melo <[email protected]>…
(quoted diff) +7
-0+ case OLD_RINGBUF_TYPE_TIME_EXTEND: + extend = read_4(kbuf, ptr); + extend <<= TS_SHIFT; + extend += delta; + delta = extend; + ptr += 4; + /* Fall through */No, actually, the length should be zeroed here.
length = 0;As the ptr is placed passed the metadata. At the end of this case statement, the ptr will be after the metadata. The length is to represent the length of the data that is part of the event. The index is the ptr -
start_data+ length. As there's no data to this type of event, it should be zero.Quoted 4 lines
When I add length = 0, I get more events from my old 2.6.30 trace.dat file :-/ I'll have to update my tests.
Care to send another patch?
I have this now in my tree, and since the patch is basically yours, I just made myself the reporter and you the author, please ack.
commit
1fb82b6cba967afde11957099a9ef440cf8c2248Author: Steven Rostedt <[email protected]> Date: Mon Feb 13 13:18:29 2017 -0300tools lib traceevent: Initialize lenght on
OLD_RING_BUFFER_TYPE_TIME_STAMPA undefined value was being used for the
OLD_RING_BUFFER_TYPE_TIME_STAMPcase entry, as the 'lenght' variable was not being initialized, fix it.Caught by the reporter when building
tools/perf/using clang, which emmitted this warning:kbuffer-parse.c:312:7: warning: variable 'length' is used uninitialized whenever switch case is taken [-Wsometimes-uninitialized]
caseOLD_RINGBUF_TYPE_TIME_EXTEND:
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
kbuffer-parse.c:339:29: note: uninitialized use occurs here
kbuf->next=kbuf->index+ length;
^~~~~~
kbuffer-parse.c:297:21: note: initialize the variable 'length' to silence this warning
unsigned int length;
^
= 0Reported-by Arnaldo Carvalho de Melo <[email protected]>Cc Adrian Hunter <[email protected]>Cc David Ahern <[email protected]>Cc Jiri Olsa <[email protected]>Cc Namhyung Kim <[email protected]>Cc Wang Nan <[email protected]>Signed-off-by Steven Rostedt <[email protected]>tools/lib/traceevent/kbuffer-parse.c +0
-0
tools/lib/traceevent/kbuffer-parse.c static unsigned int old_update_pointers(struct kbuffer *kbuf) · line 318
Quoted 4 lines
tools/lib/traceevent/kbuffer-parse.c +1
-0@@ -315,6 +315,7 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) extend += delta; delta = extend; ptr +=4 ; + lenght =0 ;
ouch, 'length' :-)
clang provides a really nice error message:
kbuffer-parse.c:318:3: error: use of undeclared identifier 'lenght'; did you mean 'length'?
lenght = 0;
^~~~~~
length
kbuffer-parse.c:297:15: note: 'length' declared here
unsigned int length;
^1 error generated.
I only had to test compile it :-)
Quoted 4 lines
tools/lib/traceevent/kbuffer-parse.c +0
-0break ;case OLD_RINGBUF_TYPE_TIME_STAMP :
Quoted 21 lines · On Mon, 13 Feb 2017 13:26:22 -0300 Arnaldo Carvalho de Melo <[email protected]>…
tools/lib/traceevent/kbuffer-parse.c static unsigned int old_update_pointers(struct kbuffer *kbuf) · line 318
Quoted 9 lines
tools/lib/traceevent/kbuffer-parse.c +1
-0@@ -315,6 +315,7 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) extend += delta; delta = extend; ptr +=4 ; + lenght =0 ;ouch, 'length' :-)
clang provides a really nice error message:
kbuffer-parse.c:318:3: error: use of undeclared identifier 'lenght'; did you mean 'length'?
lenght = 0; ^~~~~~ length kbuffer-parse.c:297:15: note: 'length' declared here unsigned int length; ^
Quoted 8 lines
1 error generated.
I only had to test compile it :-)
Quoted 4 lines
tools/lib/traceevent/kbuffer-parse.c +0
-0break ;case OLD_RINGBUF_TYPE_TIME_STAMP :
OK, I pulled the patch into git, and made it official ;-) I also fixed the "lenght" in the change log too.
Could you use this instead.
-- Steve
Quoted 1 line
From
4f060a0577156acfdc0524a634f4afcc1657c929Mon Sep 17 00:00:00 2001
From: "Steven Rostedt (VMware)" <[email protected]>
Date: Mon, 13 Feb 2017 12:11:44 -0500
Subject: [PATCH] tools lib traceevent: Initialize lenght on
OLD_RING_BUFFER_TYPE_TIME_STAMP
A undefined value was being used for the OLD_RING_BUFFER_TYPE_TIME_STAMP case entry, as the 'length' variable was not being initialized, fix it.
Caught by the reporter when building tools/perf/ using clang, which emmitted this warning:
kbuffer-parse.c:312:7: warning: variable 'length' is used uninitialized whenever switch case is taken [-Wsometimes-uninitialized]
case OLD_RINGBUF_TYPE_TIME_EXTEND:
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
kbuffer-parse.c:339:29: note: uninitialized use occurs here
kbuf->next = kbuf->index + length;
^~~~~~
kbuffer-parse.c:297:21: note: initialize the variable 'length' to silence this warning
unsigned int length;
^
= 0--- tools/lib/traceevent/kbuffer-parse.c | 1 + 1 file changed, 1 insertion(+)
tools/lib/traceevent/kbuffer-parse.c +1 -0
@@ -315,6 +315,7 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) extend += delta; delta = extend; ptr +=4 ; + length =0 ;break ;case OLD_RINGBUF_TYPE_TIME_STAMP :
Signature · 1 line
2.9.3
Em Mon, Feb 13, 2017 at 12:14:18PM -0500, Steven Rostedt escreveu:
Quoted 31 lines
Quoted 27 lines · On Mon, 13 Feb 2017 13:26:22 -0300 Arnaldo Carvalho de Melo <[email protected]>…
tools/lib/traceevent/kbuffer-parse.c static unsigned int old_update_pointers(struct kbuffer *kbuf) · line 318
Quoted 9 lines
tools/lib/traceevent/kbuffer-parse.c +1
-0@@ -315,6 +315,7 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) extend += delta; delta = extend; ptr +=4 ; + lenght =0 ;ouch, 'length' :-)
clang provides a really nice error message:
kbuffer-parse.c:318:3: error: use of undeclared identifier 'lenght'; did you mean 'length'?
lenght = 0; ^~~~~~ length kbuffer-parse.c:297:15: note: 'length' declared here unsigned int length; ^1 error generated.
I only had to test compile it :-)
Quoted 2 lines
tools/lib/traceevent/kbuffer-parse.c +0
-0break ;
Quoted 7 lines
Quoted 2 lines
tools/lib/traceevent/kbuffer-parse.c static unsigned int old_update_pointers(struct kbuffer *kbuf) · line 321
Quoted 2 lines
tools/lib/traceevent/kbuffer-parse.c +0
-0case OLD_RINGBUF_TYPE_TIME_STAMP :OK, I pulled the patch into git, and made it official ;-) I also fixed the "lenght" in the change log too.
Could you use this instead.
Sure, now take a look at this another one:
commit 6401e4361df183bd9953dce56f7c51d8ef28b11e
Author: Arnaldo Carvalho de Melo <[email protected]> Date: Mon Feb 13 13:33:57 2017 -0300
tools lib traceevent plugin function: Initialize 'index' variable
Detected with clang:
CC /tmp/build/perf/plugin_function.o
plugin_function.c:145:6: warning: variable 'index' is used uninitialized whenever 'if' condition is false [-Wsometimes-uninitialized]
if (parent && ftrace_indent->set)
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
plugin_function.c:148:29: note: uninitialized use occurs here
trace_seq_printf(s, "%*s", index*3, "");
^~~~~
plugin_function.c:145:2: note: remove the 'if' if its condition is always true
if (parent && ftrace_indent->set)
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
plugin_function.c:145:6: warning: variable 'index' is used uninitialized whenever '&&' condition is false [-Wsometimes-uninitialized]
if (parent && ftrace_indent->set)
^~~~~~
plugin_function.c:148:29: note: uninitialized use occurs here
trace_seq_printf(s, "%*s", index*3, "");
^~~~~
plugin_function.c:145:6: note: remove the '&&' if its condition is always true
if (parent && ftrace_indent->set)
^~~~~~~~~
plugin_function.c:133:11: note: initialize the variable 'index' to silence this warning
int index;
^
= 0
2 warnings generated.
Cc: Adrian Hunter <[email protected]>
Cc: David Ahern <[email protected]>
Cc: Jiri Olsa <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Wang Nan <[email protected]>
Link: http://lkml.kernel.org/n/[email protected]
Signed-off-by: Arnaldo Carvalho de Melo <[email protected]>tools/lib/traceevent/plugin_function.c +1 -1
@@ -130,7 +130,7 @@ static int function_handler(struct trace_seq *s, struct pevent_record *record,unsigned long long pfunction;const char *func;const char *parent; -int index; +int index =0 ;if (pevent_get_field_val (s, event,"ip" , record, &function,1 ))return trace_seq_putc (s,'!' );
Quoted 52 lines
-- Steve
From
4f060a0577156acfdc0524a634f4afcc1657c929Mon Sep 17 00:00:00 2001
From: "Steven Rostedt (VMware)" <[email protected]>
Date: Mon, 13 Feb 2017 12:11:44 -0500
Subject: [PATCH] tools lib traceevent: Initialize lenght on
OLD_RING_BUFFER_TYPE_TIME_STAMPA undefined value was being used for the
OLD_RING_BUFFER_TYPE_TIME_STAMPcase entry, as the 'length' variable was not being initialized, fix it.Caught by the reporter when building
tools/perf/using clang, which emmitted this warning:kbuffer-parse.c:312:7: warning: variable 'length' is used uninitialized whenever switch case is taken [-Wsometimes-uninitialized]
case OLD_RINGBUF_TYPE_TIME_EXTEND: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~ kbuffer-parse.c:339:29: note: uninitialized use occurs here kbuf->next = kbuf->index + length; ^~~~~~ kbuffer-parse.c:297:21: note: initialize the variable 'length' to silence this warning unsigned int length; ^ = 0Reported-by Arnaldo Carvalho de Melo <[email protected]>Cc Adrian Hunter <[email protected]>Cc David Ahern <[email protected]>Cc Jiri Olsa <[email protected]>Cc Namhyung Kim <[email protected]>Cc Wang Nan <[email protected]>Signed-off-by Steven Rostedt (VMware) <[email protected]>--- tools/lib/traceevent/kbuffer-parse.c | 1 + 1 file changed, 1 insertion(+)tools/lib/traceevent/kbuffer-parse.c +1
-0@@ -315,6 +315,7 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) extend += delta; delta = extend; ptr +=4 ; + length =0 ;break ;case OLD_RINGBUF_TYPE_TIME_STAMP :Signature · 1 line
2.9.3
Steven Rostedt
Reviewed-byQuoted 26 lines · On Mon, 13 Feb 2017 14:20:20 -0300 Arnaldo Carvalho de Melo <[email protected]>…
Sure, now take a look at this another one:
commit
6401e4361df183bd9953dce56f7c51d8ef28b11eAuthor: Arnaldo Carvalho de Melo <[email protected]> Date: Mon Feb 13 13:33:57 2017 -0300tools lib traceevent plugin function: Initialize 'index' variable Detected with clang: CC /tmp/build/perf/plugin_function.o plugin_function.c:145:6: warning: variable 'index' is used uninitialized whenever 'if' condition is false [-Wsometimes-uninitialized] if (parent && ftrace_indent->set) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~ plugin_function.c:148:29: note: uninitialized use occurs here trace_seq_printf(s, "%*s", index*3, ""); ^~~~~ plugin_function.c:145:2: note: remove the 'if' if its condition is always true if (parent && ftrace_indent->set) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ plugin_function.c:145:6: warning: variable 'index' is used uninitialized whenever '&&' condition is false [-Wsometimes-uninitialized] if (parent && ftrace_indent->set) ^~~~~~ plugin_function.c:148:29: note: uninitialized use occurs here trace_seq_printf(s, "%*s", index*3, ""); ^~~~~
Quoted 8 lines
plugin_function.c:145:6: note: remove the '&&' if its condition is always true
if (parent && ftrace_indent->set)
^~~~~~~~~
plugin_function.c:133:11: note: initialize the variable 'index' to silence this warning
int index;
^
= 0
2 warnings generated.I wonder why gcc isn't catching these. These look pretty obvious to me.
Quoted 15 lines
Cc Adrian Hunter <[email protected]>Cc David Ahern <[email protected]>Cc Jiri Olsa <[email protected]>Cc Namhyung Kim <[email protected]>Cc Wang Nan <[email protected]>Signed-off-by Arnaldo Carvalho de Melo <[email protected]>tools/lib/traceevent/plugin_function.c +0
-0
tools/lib/traceevent/plugin_function.c static int function_handler(struct trace_seq *s, struct pevent_record *record, · line 136
Quoted 8 lines
tools/lib/traceevent/plugin_function.c +1
-1@@ -130,7 +130,7 @@ static int function_handler(struct trace_seq *s, struct pevent_record *record,unsigned long long pfunction;const char *func;const char *parent; -int index; +int index =0 ;if (pevent_get_field_val (s, event,"ip" , record, &function,1 ))return trace_seq_putc (s,'!' );
-- Steve
tip-bot for Steven Rostedt (VMware)
merge bot[tip:perf/core] tools lib traceevent: Initialize lenght on OLD_RING_BUFFER_TYPE_TIME_STAMP
In reply to Steven Rostedttools lib traceevent: Initialize lenght on OLD_RING_BUFFER_TYPE_TIME_STAMP
Commit-ID: 14e4d7e0abfdefabea2b8796c5a8b2b9c77b5326 Gitweb: http://git.kernel.org/tip/14e4d7e0abfdefabea2b8796c5a8b2b9c77b5326 Author: Steven Rostedt (VMware) <[email protected]> AuthorDate: Mon, 13 Feb 2017 12:11:44 -0500 Committer: Arnaldo Carvalho de Melo <[email protected]> CommitDate: Mon, 13 Feb 2017 17:22:32 -0300
tools lib traceevent: Initialize lenght on OLD_RING_BUFFER_TYPE_TIME_STAMP
A undefined value was being used for the OLD_RING_BUFFER_TYPE_TIME_STAMP case entry, as the 'length' variable was not being initialized, fix it.
Caught by the reporter when building tools/perf/ using clang, which emmitted this warning:
kbuffer-parse.c:312:7: warning: variable 'length' is used uninitialized whenever switch case is taken [-Wsometimes-uninitialized]
case OLD_RINGBUF_TYPE_TIME_EXTEND:
^~~~~~~~~~~~~~~~~~~~~~~~~~~~
kbuffer-parse.c:339:29: note: uninitialized use occurs here
kbuf->next = kbuf->index + length;
^~~~~~
kbuffer-parse.c:297:21: note: initialize the variable 'length' to silence this warning
unsigned int length;
^
= 0--- tools/lib/traceevent/kbuffer-parse.c | 1 + 1 file changed, 1 insertion(+)
tools/lib/traceevent/kbuffer-parse.c +1 -0
@@ -315,6 +315,7 @@ static unsigned int old_update_pointers(struct kbuffer *kbuf) extend += delta; delta = extend; ptr +=4 ; + length =0 ;break ;case OLD_RINGBUF_TYPE_TIME_STAMP :
Quoted 13 lines · On Mon, 13 Feb 2017 14:20:20 -0300 Arnaldo Carvalho de Melo <[email protected]>…
Cc Adrian Hunter <[email protected]>Cc David Ahern <[email protected]>Cc Jiri Olsa <[email protected]>Cc Namhyung Kim <[email protected]>Cc Wang Nan <[email protected]>Signed-off-by Arnaldo Carvalho de Melo <[email protected]>tools/lib/traceevent/plugin_function.c +0
-0
tools/lib/traceevent/plugin_function.c static int function_handler(struct trace_seq *s, struct pevent_record *record, · line 133
Quoted 5 lines
tools/lib/traceevent/plugin_function.c +1
-1@@ -130,7 +130,7 @@ static int function_handler(struct trace_seq *s, struct pevent_record *record,unsigned long long pfunction;const char *func;const char *parent; -int index; +int index =0 ;
I went to apply this to trace-cmd's version, and it failed to apply. Then I realized that I had this fixed back in 2015, but missed sending it. That's because I don't strictly follow the 80 column limit for some of the if statements (I find the break up ugly), and when doing diffs, this got lost in the diffs of the 80 column limits ones done with the port to the kernel tree. :-/
-- Steve
Quoted 4 lines
tools/lib/traceevent/plugin_function.c +0
-0if (pevent_get_field_val (s, event,"ip" , record, &function,1 ))return trace_seq_putc (s,'!' );
Em Thu, Feb 16, 2017 at 09:00:23PM -0500, Steven Rostedt escreveu:
Quoted 13 lines
tools/lib/traceevent/plugin_function.c static int function_handler(struct trace_seq *s, struct pevent_record *record, · line 133
Quoted 7 lines · On Mon, 13 Feb 2017 14:20:20 -0300 Arnaldo Carvalho de Melo <[email protected]>…
+++
b/tools/lib/traceevent/plugin_function.ctools/lib/traceevent/plugin_function.c +1
-1@@ -130,7 +130,7 @@ static int function_handler(struct trace_seq *s, struct pevent_record *record,unsigned long long pfunction;const char *func;const char *parent; -int index; +int index =0 ;I went to apply this to trace-cmd's version, and it failed to apply. Then I realized that I had this fixed back in 2015, but missed sending it.
Quoted 4 lines
That's because I don't strictly follow the 80 column limit for some of the if statements (I find the break up ugly), and when doing diffs, this got lost in the diffs of the 80 column limits ones done with the port to the kernel tree. :-/
I'm not strict about it either:
$ find tools/perf/ -name "*.[ch]" | xargs cat | wc -l 161984 $ find tools/perf/ -name "*.[ch]" | xargs expand | cut -c80- | grep -v ^$ | wc -l 3706 $
2.29%
You (well, libtracevent) are a bit more strict:
$ find tools/lib/traceevent/ -name "*.[ch]" | xargs cat | wc -l 13814 $ find tools/lib/traceevent/ -name "*.[ch]" | xargs expand | cut -c80- | grep -v ^$ | wc -l 184 $
1.33%
For reference, the whole kernel, modulo tools/
$ find . -name "*.[ch]" | grep -v ^\.\/tools | xargs cat | wc -l 20162880 $ find . -name "*.[ch]" | grep -v ^\.\/tools | xargs expand | cut -c80- | grep -v ^$ | wc -l 439722 $
2.18%
Using functions to avoid too deep if statements usually helps here tho :-)
- Arnaldo
Quoted 7 lines
-- Steve
Quoted 4 lines
tools/lib/traceevent/plugin_function.c +0
-0if (pevent_get_field_val (s, event,"ip" , record, &function,1 ))return trace_seq_putc (s,'!' );