Hello @jpenalbae, can you also copy Steam's runtime diagnostics information (Steam -> Help -> Steam Runtime Diagnostics) to a gist and share it?
Here you have @kisak-valve: gist
Not sure what is going on here:
https://gist.github.com/jpenalbae/3173ad685b308aa55806350cfaeb9932#file-runtime_info-json-L130
If I check 32 bits vainfo seems to be fine from the terminal...
$ file /usr/bin/vainfo
/usr/bin/vainfo: ELF 32-bit LSB pie executable, Intel 80386, version 1 (SYSV), dynamically linked, interpreter /lib/ld-linux.so.2, BuildID[sha1]=71a41a953e75a4bb1440e780cd3044204e6a55cd, for GNU/Linux 3.2.0, stripped
$ LIBVA_DRIVER_NAME=nvidia vainfo
libva info: VA-API version 1.17.0
libva info: User environment variable requested driver 'nvidia'
libva info: Trying to open /usr/lib/i386-linux-gnu/dri/nvidia_drv_video.so
libva info: Found init function __vaDriverInit_1_0
libva info: va_openDriver() returns 0
vainfo: VA-API version: 1.17 (libva 2.12.0)
vainfo: Driver version: Splitted-Desktop Systems VDPAU backend for VA-API - 0.7.4
vainfo: Supported profile and entrypoints
VAProfileMPEG2Simple : VAEntrypointVLD
VAProfileMPEG2Main : VAEntrypointVLD
VAProfileMPEG4Simple : VAEntrypointVLD
VAProfileMPEG4AdvancedSimple : VAEntrypointVLD
<unknown profile> : VAEntrypointVLD
VAProfileH264Main : VAEntrypointVLD
VAProfileH264High : VAEntrypointVLD
VAProfileVC1Simple : VAEntrypointVLD
VAProfileVC1Main : VAEntrypointVLD
VAProfileVC1Advanced : VAEntrypointVLD
Thanks, unfortunately that looks pretty healthy to my eyes, so I have no quick suggestions to try. We'll need a Steam dev to take a look.
As far as I know, 32 bit va-api isn't used by steamwebhelper, so that anomaly could be unrelated.
Ok, thank for your support.
Hopefully this will get fixed. Since the UI overhaul the linux client has been a bit finicky...
Client crashing and restarts seems to be fixed with the latest June 6th update. From the changelog we can see:
- Fixed a crash when steam windows were closed with HW acceleration enabled on Nvidia GPUs.
- Re-enabled HW acceleration by default for Nvidia GPUs.
I can confirm the crashing has been fixed, but the continuous "X error" messages spam in steamwebhelper.log is still present.
For what it's worth : I have the same errors spamming in steamwebhelper.log (with Steam version 1686880776 and the latest : 1687386907 ). After steamwebhelper has crashed, the Store is an empty black page in the Steam client. But I don't need to fully restart the client to restore the Store. I just need to navigate to the Library, then go back to the Store.
When I uncheck "GPU acceleration rendering in web views" the errors seems to stop in steamwebhelper.log . I'll see if crashes stop too.
Edit : it does not prevent the crash, which happens half a minute after the Steam client has restarted.
Nothing extracted yet.
Your system information
Please describe your issue in as much detail as possible:
The following messages are constantly spammed on
steamwebhelper.logIf
GPU acceleration rendering in web viewsis enabled then also get these messages:After several of those then I get these other messages, a client restart and a minidump is created (Not sure if the minidump gets sent).
Also
/tmp/dump/is full of minidumps:Full log here: steam-full2.log.gz
Output for
minidump-stackwalk /tmp/dumps/assert_20230606202928_29.dmpstackwalk.txt.gz
Steps for reproducing this issue:
GPU acceleration rendering in web viewsenabled3and4until the client restarts (normally repeating the steps 3-4 times is enough)Client also seems to randomly restart sometimes, but the steps described above can be used to force the restart.
Other notes:
The only workaround to avoid client crashing and generating dumps is disabling
GPU acceleration rendering in web viewsalthoughX Error of failed requestmessages spam is still present.This only happens with beta client, if I roll back to stable client, there is no error messages spam neither crashes. I have seen this behavior since the big "UI overhaul" update.
Things I have tried to avoid this problem without success:
.steamand.local/share/SteamI have seen many reports talking about bad nvidia or mesa installations but reinstalling does not fix the problem for me. Also the issue #9164 might be similar but without the crashing.