Re: perf backtraces off-by-1

2012-08-28 Thread Arun Sharma
On 8/28/12 9:34 AM, Peter Zijlstra wrote: It used to look like this: http://git.savannah.gnu.org/gitweb/?p=libunwind.git;a=commitdiff;h=92cc7fd78a5a79c4bb5f85bfb7d7fb025df9cd5a Hmm, that's not too bad, but a long stretch from pretty ;-) How would you 'encode' this in the perf callchain data?

Re: perf backtraces off-by-1

2012-08-28 Thread Peter Zijlstra
On Sun, 2012-08-26 at 10:52 -0700, Arun Sharma wrote: > On 8/26/12 9:10 AM, Peter Zijlstra wrote: > > On Fri, 2012-08-24 at 15:13 -0700, Arun Sharma wrote: > > > >> One option is to support > >> this for user mode only, with code to detect signal frames. Any other > >> ideas? > >> > > I guess we'

Re: perf backtraces off-by-1

2012-08-26 Thread Arun Sharma
On 8/26/12 9:10 AM, Peter Zijlstra wrote: On Fri, 2012-08-24 at 15:13 -0700, Arun Sharma wrote: One option is to support this for user mode only, with code to detect signal frames. Any other ideas? I guess we'd need to see what that patch would look like... :-) It used to look like this

Re: perf backtraces off-by-1

2012-08-26 Thread Peter Zijlstra
On Fri, 2012-08-24 at 15:13 -0700, Arun Sharma wrote: > > Wondering if this has been discussed before. Not that I can recall. > One option is to support > this for user mode only, with code to detect signal frames. Any other > ideas? > I guess we'd need to see what that patch would look like.

perf backtraces off-by-1

2012-08-24 Thread Arun Sharma
Some of our language runtimes like to map IP addresses in perf backtrace to specific byte codes. The way things stand now, the addresses on the backtrace are return addresses, rather than the caller. I think this issue may be present for other unusual call/return sequences where the user may be