gcov: reduce code quality loss by reproducible topn merging [PR92924]

[Martin Liška]

On 1/30/20 5:13 PM, Jan Hubicka wrote:
> Martin, I welcome your opinion on the patch

Ok, recently we've made quite some experiments with Honza about
TOP N counters. Based on the numbers, it shows that tracking a negative
value per-counter value does not help much. So that I suggest to use
only one global flag (a negative total) in order to distinguish in between
reproducible and non-reproducible profile.

I'm sending patch that simplifies the merging and introduces a new option.

For the future, we may (similarly to LLVM) come up with a dynamic allocation
for TOPN counters. I've git a working patch for that.

Martin
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-Introduce-fprofile-reproducible-and-support-it-with-.patch
Type: text/x-patch
Size: 8103 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20200205/c1d74c06/attachment.bin>