OK, now I know why these error messages were very familiar to me.
I had these error messages for kalgebra as well, and I suspect it is
because of SSE2. See more info here:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1016658
Version 2.0.8-2 for lightdm-gtk-greeter is very old. But what is thi
> Can you still reproduce this issue?
Hello,
Yes, I can.
I've downloaded and tried the latest daily testing iso on real
hardware (Athlon XP):
filename: debian-testing-i386-netinst.iso
SHA256: 813e13ad3fe1f89f5c86477e624a006edbc76e5fff8b27f06d2d00f480283d6a
Upload date: 2022-10-10 11:16
Size: 525M
Hello,
MR has been opened to "fix" this issue:
https://salsa.debian.org/pkg-llvm-team/z3/-/merge_requests/6
Eliminating SSE2 from z3 upstream is not feasible:
https://github.com/Z3Prover/z3/issues/6369#issuecomment-1259419466
The dependency chain I have identified:
xorg->libgl1-mesa-dri->libllvm
So, since Xorg requires SSE2, I tried GNOME as it is using Wayland
instead. In the debian installer I kept the default selection for the
desktop environment (GNOME, which comes with gdm3). Everything works
fine. The other DE which comes with Wayland is KDE, but I cannot
install it because that requ
> As far as I see your MR just adds a dependency to sse2-support,
> which I guess just makes the installation abort in case of a CPU
> not supporting sse2, so I guess this would just make
> mesa not installable on your hardware?
Exactly.
It will abort the installation as it wouldn't work anyways.
There are no errors in Xorg log files.
I have already included before the error messages I got from the RS232
cable. Those were assertion errors doing some CSS bezier effect
(transition ease-in probably).
In my current installation the default GNOME DE was selected.
Now I tried installing LXDE vi
> According to the developers of libz4-3, the precision of the FPU x87
> registers is not sufficient therefore a libz4-3 package compiled
> without SSE2 would be non-functional.
I think this is just a made-up excuse so they don't have to waste time
on supporting niche 20+ years old hardware.
But i
Control: tags -1 upstream
Control: forward -1 https://bugzilla.mozilla.org/show_bug.cgi?id=1802798
Control: forwarded -1 https://bugzilla.mozilla.org/show_bug.cgi?id=1802798
Control: tags -1 + unreproducible
The issue was due to faulty RAM. No further action required.
Control: severity -1 minor
Yes, it is still crashing by running it that way.
Alberto Garcia ezt írta (időpont: 2023. márc. 29.,
Sze, 12:58):
>
> On Wed, Mar 29, 2023 at 06:22:25AM -0400, Karo Gyoker wrote:
>
> > If I try to open any page in epiphany-browser, nothing happens. The CPU is
> > at 100%.
> > I can see it in dme
For the others watching (if any): I have sent the core dump to Berto.
Actually, firefox-esr doesn't even work on real x86_32 CPUs (except
Pentium 4) since years.
It requires SSE2 without the sse2-support package, so basically it is
committing a baseline violation on i386.
Currently, due to some random GCC behavior, it doesn't crash instantly
when I try to open any pa
It seems to me that the libz3-4 package was compiled with SSE2 turned on.
$ objdump -d libz3.so.4 | grep pxor | wc -l
2896
$ objdump -d libz3.so.4 | grep movq | wc -l
26380
Wow, what a splendid debugging! I hope you enjoyed it.
It seems that these [1] [2] bugs are all having the same root cause then.
Or maybe [1] is different, it would be needed to debug that similarly to
ensure that if libz3-4 is related or not. But probably it is.
[1]: https://bugs.debian.org/cgi-
I've opened a new MR:
https://salsa.debian.org/pkg-llvm-team/z3/-/merge_requests/8
This MR won't add SSE2 as a requirement. Instead, it will make
the compiler not to emit SSE2 instructions when building for x86_32.
I've tested it on my Athlon XP and it works fine.
objdump -d doesn't show up any SS
Hello,
My MR [1] using SIMDEverywhere has been merged into z3.
It is going to solve the SSE2 issue on i386 on CPUs
like Athlon XP and Pentium Pro or anything before
Pentium 4.
[1]: https://salsa.debian.org/pkg-llvm-team/z3/-/merge_requests/9
I can confirm that with the new package the lightdm login screen works again.
It seems that firefox-esr_102.4.0esr-1 works again. I don't know why though.
Merge request with the patch has been opened here:
https://salsa.debian.org/mozilla-team/firefox/-/merge_requests/9
Package: kalgebra
Version: 4:20.12.1-1
Followup-For: Bug #1016658
There is a dependency chain which leads to sse2-support. And in the
near future it will be sse3-support.
kalgebra needs libqt5webenginewidgets5 which needs
libqt5webenginecore5 which needs sse2-support, but only since
bookworm, the
I've tested the proposed patch below:
https://github.com/amurzeau/debian-autobuild/releases/tag/firefox-esr%2F91.12.0esr-1%2Bnosse1_deb11u1
I've tested it on an Athlon XP 2600+ (Barton) with 3GB RAM.
Everything seems fine. I can even play videos on youtube with smooth
sound and if I play the vide
Some more info:
I was using the Standalone profile with LXQT.
Since then I've tried to install it with LXDE instead, as it doesn't
require SSE2. But SSE2 is still required, most probably because some
other package still requires QT5.
24 matches
Mail list logo
