[Bug ipa/68035] [5/6 Regression] ipa performance issue when no procedures are present

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68035

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Created attachment 36655
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=36655&action=edit
Suggested patch

Hello.

Following patch should correctly force calculation of hash values that are
different for all variables and no O(n^2) comparison is processed.

Could you Neil test the patch on the whole source file before I'm going to send
the patch to mailing list?

Thanks,
Martin