Re: [PATCH] IPA: reduce what we dump in normal mode

2022-08-03 Thread Martin Liška
On 8/2/22 18:27, Jan Hubicka wrote: > If you disable dumping, you can also disable the collection of stats > which is guarded by if (dump_file) as well. Otherwise the patch is OK. Sure, good point! I'm going to push the following. Thanks, MartinFrom 7585e5ecb47761516b8f397002819f2c95b8c32e Mon

Re: [PATCH] IPA: reduce what we dump in normal mode

2022-08-02 Thread Jan Hubicka via Gcc-patches
> gcc/ChangeLog: > > * profile.cc (compute_branch_probabilities): Dump details only > if TDF_DETAILS. > * symtab.cc (symtab_node::dump_base): Do not dump pointer unless > TDF_ADDRESS is used, it makes comparison harder. > --- > gcc/profile.cc | 2 +- > gcc/symtab.cc | 3 +

Re: [PATCH] IPA: reduce what we dump in normal mode

2022-08-02 Thread Richard Biener via Gcc-patches
On Tue, Aug 2, 2022 at 10:46 AM Martin Liška wrote: OK > gcc/ChangeLog: > > * profile.cc (compute_branch_probabilities): Dump details only > if TDF_DETAILS. > * symtab.cc (symtab_node::dump_base): Do not dump pointer unless > TDF_ADDRESS is used, it makes comparis

[PATCH] IPA: reduce what we dump in normal mode

2022-08-02 Thread Martin Liška
gcc/ChangeLog: * profile.cc (compute_branch_probabilities): Dump details only if TDF_DETAILS. * symtab.cc (symtab_node::dump_base): Do not dump pointer unless TDF_ADDRESS is used, it makes comparison harder. --- gcc/profile.cc | 2 +- gcc/symtab.cc | 3 ++- 2 file