gprof profiling of multi-threaded Cygwin programs

Corinna Vinschen corinna-cygwin@cygwin.com
Fri Jan 29 18:55:00 GMT 2016


Hi Mark,

On Jan 29 02:20, Mark Geisert wrote:
> I've got alpha-quality code running to support this and am running some
> tests to ensure I'm seeing what I ought to be seeing.  Could I get a quick
> design review to confirm I'm on the right track of implementation and to
> maybe point out odd use cases I might have missed?
> 
> I could have just submitted patches but I'd like to be more sure of my
> direction before patching.

Unless your patches are trivial (no new functionality, less than 10
lines), submitting patches requires a copyright assignement, See
https://cygwin.com/contrib.html especially the "Before you get started"
section.

> I see three areas of implementation...
> (1) Modify profil.c to have the profiling thread it creates sample all
> pthreads' pc values in addition to its current sampling of the main
> thread's.  I plan to have profil.c call a routine in cygheap.cc to get
> handles to the running pthreads that can be passed to GetThreadContext()
> back in profil.c.  (The code is cleaner than this reads.)

Sounds ok to me.  Try to keep it so that code isn't slowed down when
profiling is *not* used.

> (2) Modify mcount.c to make mcount_private() thread-safe.

Yuk.  Thread-safety wasn't on the plate at all when this code has been
written I guess...

> entered on every function call that occurs in an executable compiled with
> the '-pg' option; this to build call path diagrams.  I plan to replace the
> sets and refs of the gmonparam state variable with Interlocked* operations.
> It would be nice to record somewhere how many misses occur; the code doesn't
> queue callers because it's being called on *every* profiled function call.
> That happens so very frequently under profiling that if the state doesn't
> allow entry that call's info is just discarded.

Yeah, sounds fine to me.

> (3) Modify gmon.c to allow for possibility of multiple gmon.out files due to
> fork/exec from a profiled process.  I see there is potentially usable code
> already there but it's #ifdef'd out.  Does anybody know why?

Sorry, no.  The gist of that code is basically written the last millenium.
Some of the ifdef's/ifndef's probably have been added just to simplify the
code (or something didn't work in 1997 or so).  Feel free to change that,
but try to keep it as much aligned to some version of upstream as possible,
with upstream being one of FreeBSD, NetBSD, OpenBSD.

> Thanks for the once-over and if submitting patches really is the way I
> should go for this review, just let me know.

Well, it's kind of the next step, isn't it? ;)


Thanks,
Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Maintainer                 cygwin AT cygwin DOT com
Red Hat
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://cygwin.com/pipermail/cygwin-developers/attachments/20160129/062862a9/attachment.sig>


More information about the Cygwin-developers mailing list