This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] Speed-up indirect-call instrumentation
- From: Markus Trippelsdorf <markus at trippelsdorf dot de>
- To: Martin Liška <mliska at suse dot cz>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>, Jan Hubicka <hubicka at ucw dot cz>
- Date: Fri, 9 Jun 2017 18:56:46 +0200
- Subject: Re: [PATCH] Speed-up indirect-call instrumentation
- Authentication-results: sourceware.org; auth=none
- References: <firstname.lastname@example.org>
On 2017.06.09 at 14:17 +0200, Martin Liška wrote:
> I discussed with Honza possibility to speed-up instrumentation that we do for
> indirect call target tracking. By direct emission of:
> if (__gcov_indirect_call_callee != NULL)
> __gcov_indirect_call_profiler_v2 (profile_id, ¤t_function_decl);
> we can save reduce # of execution of __gcov_indirect_call_profiler_v2 function.
> I measured that tramp3d (-O2 -fprofile-generate) goes from 7.1s to 6.3s.
> Apart from that I slightly change probability for direct instrumentation of
> time profiler.
> Patch can bootstrap on ppc64le-redhat-linux and survives regression tests.
But it crashes on gcc.dg/pr78582.c with -O0.