Hello Karsten,
thanks for the information - that explains why my emulated pentium
failed already at 0x...6bb because not supporting sse at all.
>From wikipedia [1] the pminud instruction at 0x...6fb got
introduced with sse4.1 which seem not supported from your
flags line (while on the other side
On Mon, Oct 26, 2020 at 11:04:42PM +0100, Bernhard Übelacker wrote:
> From wikipedia [1] the pminud instruction at 0x...6fb got
> introduced with sse4.1 which seem not supported from your
> flags line (while on the other side intel says [2] it is a Penryn).
OTOH, apparently wikipedia knows
Hello Bernhard,
thanks for your work.
I have (also) filed a bug against intel-media-va-driver which
was invovked from VLC. They have forwarded the issue upstream:
https://github.com/intel/libva/issues/466
My CPU is Penryn, so it supports "less" SSE than what's
attempted to be used by
Hello Karsten,
I tried to collect some more information for the maintainer and
could reproduce this (or nearly) the same SIGILL with a qemu VM limited
to a pentium class CPU.
The instruction in question might these below:
0xb78816bb <+75>:movd 0x4(%esi),%xmm2 (here I received the
Package: finch
Version: 2.13.0-2.2+b1
Severity: important
Dear maintainers,
on startup this happens (taken from systemd journal):
kernel: traps: finch[25048] trap invalid opcode ip:ad38b6fb sp:bfb44fc0
error:0 in iHD_drv_video.so[ad0ca000+3c2000]
Running under gdb and backtracing:
5 matches
Mail list logo