https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94278
Thomas Schwinge <tschwinge at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Ever confirmed|0 |1
Status|UNCONFIRMED |NEW
Last reconfirmed| |2020-04-21
--- Comment #1 from Thomas Schwinge <tschwinge at gcc dot gnu.org> ---
Andrew, Julian, any insights here?
(In reply to Thomas Schwinge from comment #0)
> Stack dump:
> 0. Program arguments:
> [...]/install/offload-amdgcn-amdhsa/amdgcn-amdhsa/bin/as
> -triple=amdgcn--amdhsa -mattr=-code-object-v3 -mcpu=fiji -filetype=obj -o
> /tmp/ccn2bzJ1.o /tmp/cchOR6jK.mkoffload.2.s
> xgcc: internal compiler error: Segmentation fault signal terminated
> program as
> [...]
> mkoffload: fatal error:
> [...]/build-gcc/./gcc/x86_64-pc-linux-gnu-accel-amdgcn-amdhsa-gcc returned 4
> exit status
>
> That's with stock Ubuntu 18.04 LLVM 9 packages:
>
> $ [...]/install/offload-amdgcn-amdhsa/amdgcn-amdhsa/bin/as --version
> LLVM (http://llvm.org/):
> LLVM version 9.0.0
> [...]
> $ readlink [...]/install/offload-amdgcn-amdhsa/amdgcn-amdhsa/bin/as
> /usr/bin/llvm-mc-9
> $ dpkg -S /usr/bin/llvm-mc-9
> llvm-9: /usr/bin/llvm-mc-9
> $ apt-cache policy llvm-9
> llvm-9:
> Installed: 1:9-2~ubuntu18.04.2
> [...]
>
> As this is a problem in an external tool, are we going to declare this "not
> our problem", hoping it'll go away in a later release? Or, do we intent to
> be compatible with this LLVM version, and inspect whether there maybe is an
> actual issue with GCC's amdgcn code generation that triggers this problem?
