On Mon, 2021-08-16 at 22:27 +0530, Ankur Saini wrote: > > Thanks for the review > > > On 16-Aug-2021, at 4:48 AM, David Malcolm <dmalc...@redhat.com> > > wrote: > > > > Thanks, this is looking promising. Has this been rebased recently > > (e.g. since I merged > > https://gcc.gnu.org/pipermail/gcc-patches/2021-August/576737.html ) > > Yes, The branch is totally up to date with master at the time of > writing this mail. > > - - - > > Here is the updated patch : > > > > - - - > P.S. While adding the new tests, I found some last minute bugs where > analyzer would sometimes try to access a NULL cgraph edge to get the > call stmt and crash. > Although the problem was easily fixed by updating > `callgraph_superedge::get_call_stmt ()`, it lead to a delay in running > the test suite on the final version ( so the patch is still under > testing at the time of writing this mail ).
[...snip...] > From 4ef1658de158c07474605ef4df9d092cfe7962aa Mon Sep 17 00:00:00 2001 > From: Ankur Saini <arse...@sourceware.org> > Date: Sun, 15 Aug 2021 19:19:07 +0530 > Subject: [PATCH 2/2] analyzer: detect and analyze virtual function calls > > 2021-08-15 Ankur Saini <arse...@sourceware.org> > > gcc/analyzer/ChangeLog: > PR analyzer/97114 > * analyzer/region-model.cc (region_model::get_rvalue_1): Add case for > OBJ_TYPE_REF. > > gcc/testsuite/ChangeLog: > *g++.dg/analyzer/vfunc-2.C: New test. > *g++.dg/analyzer/vfunc-3.C: New test. > *g++.dg/analyzer/vfunc-4.C: New test. > *g++.dg/analyzer/vfunc-5.C: New test. ChangeLog nit: the format of the above isn't quite correct; the entries are missing a space between the "*" and the filename. You can use contrib/gcc-changelog/git_check_commit.py to check commit messages before attempting to push them (that script runs server-side and will reject the push). I like the new tests, thanks! Both of these patches are OK to push to trunk, provided that they bootstrap and the testsuite doesn't regress. Dave