Re: PING: [PATCH]: Fix -fbranch-probabilities
On 08/27/2011 02:04 AM, Jan Hubicka wrote: Hello, Could I have a review for the trivial patch posted in http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01123.html -fprofile-use sets flag_branch_probabilities. But we should also be able to use -fbranch-probabilities on its own using the information generated by -fprofile-arcs, as documented. OK, thanks! I was under impression that some of gcov tests still use -fprofile-arcs -fbranch-probabilities pair. yes, indeed, this is was the documentation claims: http://gcc.gnu.org/onlinedocs/gccint/C-Tests.html#C-Tests. e.g for : gcc.misc-tests ... bprob*.c Test -fbranch-probabilities using gcc.misc-tests/bprob.exp, ... but bprob.exp sets feedback_options to -fprofile-use It don't seem to be the case, so if you add a testcase, you get extra score ;) I feel more like fixing the bprob.exp discrepancy to have the correct pairing with the following. This will act as the testcase, since those tests fail without the patch. OK ? Index: gcc.misc-tests/bprob.exp === --- gcc.misc-tests/bprob.exp(revision 178096) +++ gcc.misc-tests/bprob.exp(working copy) @@ -48,7 +48,7 @@ load_lib profopt.exp set profile_options -fprofile-arcs -set feedback_options -fprofile-use +set feedback_options -fbranch-probabilities foreach profile_option $profile_options feedback_option $feedback_options { foreach src [lsort [glob -nocomplain $srcdir/$subdir/bprob-*.c]] { - 2011-08-29 Christian Bruel christian.br...@st.com * gcc.misc-tests/bprob.exp (feedback_options): Set -fbranch-probabilities. - Thanks Christian Honza Many thanks Christian
Re: PING: [PATCH]: Fix -fbranch-probabilities
On Mon, Aug 29, 2011 at 1:59 PM, Christian Bruel christian.br...@st.com wrote: On 08/27/2011 02:04 AM, Jan Hubicka wrote: Hello, Could I have a review for the trivial patch posted in http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01123.html -fprofile-use sets flag_branch_probabilities. But we should also be able to use -fbranch-probabilities on its own using the information generated by -fprofile-arcs, as documented. OK, thanks! I was under impression that some of gcov tests still use -fprofile-arcs -fbranch-probabilities pair. yes, indeed, this is was the documentation claims: http://gcc.gnu.org/onlinedocs/gccint/C-Tests.html#C-Tests. e.g for : gcc.misc-tests ... bprob*.c Test -fbranch-probabilities using gcc.misc-tests/bprob.exp, ... but bprob.exp sets feedback_options to -fprofile-use It don't seem to be the case, so if you add a testcase, you get extra score ;) I feel more like fixing the bprob.exp discrepancy to have the correct pairing with the following. This will act as the testcase, since those tests fail without the patch. OK ? If it works, yes. Richard. Index: gcc.misc-tests/bprob.exp === --- gcc.misc-tests/bprob.exp (revision 178096) +++ gcc.misc-tests/bprob.exp (working copy) @@ -48,7 +48,7 @@ load_lib profopt.exp set profile_options -fprofile-arcs -set feedback_options -fprofile-use +set feedback_options -fbranch-probabilities foreach profile_option $profile_options feedback_option $feedback_options { foreach src [lsort [glob -nocomplain $srcdir/$subdir/bprob-*.c]] { - 2011-08-29 Christian Bruel christian.br...@st.com * gcc.misc-tests/bprob.exp (feedback_options): Set -fbranch-probabilities. - Thanks Christian Honza Many thanks Christian
PING: [PATCH]: Fix -fbranch-probabilities
Hello, Could I have a review for the trivial patch posted in http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01123.html -fprofile-use sets flag_branch_probabilities. But we should also be able to use -fbranch-probabilities on its own using the information generated by -fprofile-arcs, as documented. Many thanks Christian
Re: PING: [PATCH]: Fix -fbranch-probabilities
Hello, Could I have a review for the trivial patch posted in http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01123.html -fprofile-use sets flag_branch_probabilities. But we should also be able to use -fbranch-probabilities on its own using the information generated by -fprofile-arcs, as documented. OK, thanks! I was under impression that some of gcov tests still use -fprofile-arcs -fbranch-probabilities pair. It don't seem to be the case, so if you add a testcase, you get extra score ;) Honza Many thanks Christian
[PATCH]: Fix -fbranch-probabilities
Hello, -fbranch-probabilities fails to find the gcda information, because they are initialized only if flag_profile_use. The problem is easily observable by recompiling with -fbranch-probabilities instead of -fprofile-use (after profile information generation) This fails with xxx.gcda not found, execution counts estimated. This trivial patch fixes it. testsuite ok. OK ? Christian 2011-08-12 Christian Bruel christian.br...@st.com * coverage.c (coverage_init): Check flag_branch_probabilities instead of flag_profile_use. Index: gcc/coverage.c === --- gcc/coverage.c (revision 177690) +++ gcc/coverage.c (working copy) @@ -1056,7 +1056,7 @@ strcpy (bbg_file_name, filename); strcat (bbg_file_name, GCOV_NOTE_SUFFIX); - if (flag_profile_use) + if (flag_branch_probabilities) read_counts_file (); }