https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70594
--- Comment #24 from Tobias Burnus <burnus at gcc dot gnu.org> --- (In reply to Jakub Jelinek from comment #18) > For the testcases in this PR probably: [...[ > [...] would be enough [...] But the #c0 report suggests this affects > more than that, but only Tobias has the reproducer that. I applied that patch - and can confirm that it fixes the LLVM issue (comment 4), but not the original issue (comment 0). Regarding comment 3, using -fdump-final-insns= without vs. with --param ggc-min-expand=0 --param ggc-min-heapsize=0 gives indeed a different result for comment 0's test case; however, that flag makes GCC extremely slow; the compile time goes up from 11s to 15min - which is also not the best for delta.pl reduction. The changes are in items such as 'net_stream.160_55->D.114439._vptr.' where D.114439 changes to D.114447 and the decl_uid in the ";; Function" lines changes. (In reply to Jakub Jelinek from comment #20) > Tobias, could you try your to reproduce your #c0 issue, with a debugging > patch like: [...] I will try this now.