#1045 [gcc] ocaml-dune segfault on i386 when inlining | rhbz#2171888
Closed by blockerbot. Opened by blockerbot.

Bug details: ** https://bugzilla.redhat.com/show_bug.cgi?id=2171888 **
Information from BlockerBugs App:
2171888

Current vote summary

Commented but haven't voted yet: jakub

The votes have been last counted at 2023-02-22 18:11 UTC and the last processed comment was #comment-842710

To learn how to vote, see:
https://pagure.io/fedora-qa/blocker-review
A quick example: BetaBlocker +1 (where the tracker name is one of BetaBlocker/FinalBlocker/BetaFE/FinalFE/0Day/PreviousRelease and the vote is one of +1/0/-1)


BetaFE +1

I'm not sure what "miscompiling" means here. I suppose it's the segfaul reported in the original comment. That feels bad, but doesn't violate any criteria that I can see. I'll withhold voting on blocker status in the hopes that getting the FE in makes it a non-issue.

It means you can get similar random crashes on any code that uses vfork that has been compiled in f38 since last Friday, what happens in the vfork child which shares address space with the parent can clobber variables in the parent.

BetaFE +1
AGREED AcceptedBetaFE

@jakub I think the problem we have is we have no idea how common use of vfork is (personally I have no idea what it is). Anyway, making it an FE solves the problem.

The following votes have been closed:

https://pubs.opengroup.org/onlinepubs/009695399/functions/vfork.html

I've done on a f36 box (with admittedly not too many packages installed)
for i in find /usr/bin/* /usr/lib64/lib*.so.* -type f; do nm -D $i 2>/dev/null | grep -q 'U vfork@' && rpm -qf $i; done | sort -u
binutils-2.37-27.fc36.x86_64
binutils-aarch64-linux-gnu-2.37-7.fc36.x86_64
binutils-arm-linux-gnu-2.37-7.fc36.x86_64
binutils-powerpc64le-linux-gnu-2.37-7.fc36.x86_64
binutils-powerpc64-linux-gnu-2.37-7.fc36.x86_64
binutils-riscv64-linux-gnu-2.37-7.fc36.x86_64
binutils-s390x-linux-gnu-2.37-7.fc36.x86_64
cpp-12.1.1-1.fc36.x86_64
crash-8.0.1-1.fc36.x86_64
cvs-1.11.23-59.fc36.x86_64
emacs-27.2-11.fc36.x86_64
execstack-0.5.0-22.fc36.x86_64
file-libs-5.41-4.fc36.x86_64
gcc-12.1.1-1.fc36.x86_64
gcc-c++-12.1.1-1.fc36.x86_64
gcc-gdc-12.1.1-1.fc36.x86_64
gcc-gfortran-12.1.1-1.fc36.x86_64
gcc-go-12.1.1-1.fc36.x86_64
gcc-offload-nvptx-12.1.1-1.fc36.x86_64
gcc-riscv64-linux-gnu-12.1.1-1.fc36.x86_64
gimp-2.10.30-1.fc36.1.x86_64
jimtcl-0.81-3.fc36.x86_64
libgccjit-12.1.1-1.fc36.x86_64
libsemanage-3.3-3.fc36.x86_64
mingw32-binutils-2.37-4.fc36.x86_64
mingw32-cpp-11.2.1-5.fc36.x86_64
mingw32-gcc-11.2.1-5.fc36.x86_64
pocl-1.8-1.fc36.x86_64
rcs-5.10.1-1.fc36.x86_64
rsh-0.17-101.fc36.x86_64
ruby-libs-3.1.2-164.fc36.x86_64
SDL2-2.0.22-2.fc36.x86_64
strace-5.16-1.fc36.x86_64
telnet-0.17-86.fc36.x86_64
texlive-gsftopk-20210325-47.fc36.x86_64
texlive-xdvi-20210325-47.fc36.x86_64

Metadata Update from @blockerbot:
- Issue status updated to: Closed (was: Open)

Release F38 is no longer tracked by BlockerBugs, closing this ticket.

Log in to comment on this ticket.

Metadata