For me as well, on all 3 computers. My thanks, also.
1 Like
To do the chroot trick with btrfs, this can be useful: [HowTo] Recovering from an interrupted update/upgrade
4 Likes
Phew, so itās not that uncommon. 
I really worried it could be a sudden hardware issue.
As for the freeze, it hasnāt happened again yet but I was really hoping for a quick (minor) stable upgrade with the new kernel because of the introduced regression. There must be more RX6XXX & RX7XXX manjaro users out there potentially affected by this.
I do have a timeshift backup from right before the stable update. Admittedly Iāve never done a timeshift restore before but it doesnāt seem to be very difficult. Nevertheless thanks for the tip about the kernel downgrade. 
Looks like the pre-stable branches are already up to kernel 6.12.50, so stable may see that version or one thatās higher.
https://manjaristas.org/branch_compare?q=linux612
A very good practice and option.
It is a bit difficult only on btrfs.
Itās just that Manjaro is starting a new release cycle āManjaro 25.1.0ā, code-named āAnh-Linhā and from my experience in the past this usually takes longer to arrive in stable than usual.
I assume late October at least for GNOME 49 and the new 6.12 stable kernels which means Iām crossing fingers till then, the freeze doesnāt show up again. 
EDIT:
By the way, Iāve always wondered why such an important tool (in my eyes) like the ābranch compareā isnāt hosted at manjaro.org but a 3rd-party site?
It used to be, but Iām not sure why it was moved to a third party either.
3 Likes
wellā¦ if one was so inclined (i.e. feeling a bit experimental)ā¦ you could pull a newer kernel from Releases Ā· manjaro-kernels/linux612 Ā· GitHub as I see both 6.12.49 and 6.12.50 thereā¦ however, itās difficult to know if thereās also a ādependencyā on a newer linux-firmware or amd-ucodethat should accompany it as well.
2 Likes
Iām not brave enough for such a stunt 
and I also lack the deeper knowledge in case Iād run into any of these dependency problems. Like Iām running VirtualBox occasionally which makes everything even more complicated with those kernel modules. Yikes. 
Actuallyā¦ it looks like all channels are using the same versions of linux-firmware and amd-ucode
so, Iād be less worried about any dependency issues, but we each need to respect our limitsā¦ and calculate the risks that are worth taking on our systems.
Mod edit:- Made links a little more obvious. No charge.
I also encountered some sudden system freezes.
Recently I chucked my nvidia GT710 in the bin and ungraded my CPU to AMD Ryzen 7 5700G with Radeon Graphics.
I did a fresh install with Manjaro 25.0.8 Zetar Gnome flavour for a change, after having run XFCE for several years.
After installation I switched to Kernel 6.18. and everything seemed to work just fine. I should mention that my desktop runs 24/7 and I encountered the freezes after it sat some time idling.
I checked journalctl and could not find any hint to why it happened.
Currently, I switched back to Kernel 6.12. and I am waiting if the freezes occur again.
1 Like
Apache could not serve local files. The error log contained a familiar error message, āā¦because search permissions are missing on a component of the pathā¦ā, but with a new reason introduced with this update.
/etc/systemd/system/httpd.service.d/hardening.conf was not in Apache 2.4.63, but is in Apache 2.4.65. The systemd directive ProtechHome is set to on.
A solution is to change the value for ProtectHome to off. After the change, reload daemon-reload and restart httpd.
1 Like
Just to prevent misunderstantings: it absolutely can serve local files, just not from a users home directory.
There is also ProtectHome=read-only where httpd might read the files but wonāt be able to write any, see:
7 Likes
I had to switch back from Nvidia 580xx to AUR 550xx.
On 580xx, some programs will freeze in the first half-minute.
And sometimes after closing programs, there will be artifacts of does programs stayās on the screen as a glitch.
While that is happening, there is no journal log happening with: sudo journalctl -r |more
Back Story
I decided not to delay this update, but made kernel 6.12.50 part of it to get the additional amdgpu commit mentioned from 6.12.49.
my steps...
Timeshiftbackup- Downloaded https://github.com/manjaro-kernels/linux612/releases/download/6.12.50-1/linux-6.12.50-1.zip to
~/Downloads - Extracted the file contents into a
~/Downloads/linux-6.12.50-1subdirectory - Ran the update as normal via terminalā¦
- confirm mirrors up
pacman-mirrors --status - confirm package metadata updated
sudo pacman -Syy - DOWNLOAD (not install) package updates
sudo pacman -Syuw - Cautious toolchain update command
sudo pacman-static -Syu - Checked for orphans
sudo pacman -Qdt - Checked foreign packages
sudo pacman -Qm - Checked for pacnewās
DIFFPROG=meld pacdiff -s
- While still in terminal, installed kernel
6.12.50-1
- changed to package directory
cd ~/Downloads/linux-6.12.50-1/ - installed the kernel packages manually
sudo pacman -U linux612-6.12.50-1-x86_64.pkg.tar.zst linux612-headers-6.12.50-1-x86_64.pkg.tar.zst
- installed AUR updates, while making sure
pamacdoesnāt downgrade the kernel I just installedpamac update -a --disable-downgrade - Reboot
Only 5-ish hrs of up-time so far, but with 3 hrs of watching videos, an hour of browsing, and an hour of playing Hades (Steam)ā¦ so far so good!
Note: The only time my AMD RX7700S dGPU is being used is while gamingā¦ everything else (desktop, browser, etc) is on the AMD 780M iGPUā¦ might explain why my entire desktop session hasnāt frozen when a game has an issue on the dGPU.
3 Likes
Update: fixed the issue of CP2077 not launching. It had nothing to do with this stable update. Something with Cyberpunks mesa_shader_cache had gone bottoms up. Renaming/deleting it fixed the launch issue.
Has anyone, since 09-26 stable update, been experiencing issues with Steamās latest client update from around 10-4, on AMD GPU systems, either 6.12.48 and or 6.16.8 kernels?
Where some games are no longer launching (PID getting killed by Steam before the gameās exe launches), or having hangs in certain games after a while, where before they have been stable?
Summary
On 10-04 Steamās client got updated from Client version: 1757650573 to 1759461205.
But since Iāve not had time off for a gaming session, since 09-26 stable update, I noticed things having launch and stability issues last Monday and today. Most odd and noticeable CP2077, where itās log is being spammed with the following (see spoiler)
Summary
15098.891:00d0:00d4:trace:seh:dispatch_exception code=6ba (RPC_S_SERVER_UNAVAILABLE) flags=0 addr=00006FFFFFBFD197
15098.891:00d0:00d4:trace:seh:dispatch_exception rip=00006fffffbfd197 rsp=000000000021d570 rbp=000000000021d660 eflags=00000202
15098.891:00d0:00d4:trace:seh:dispatch_exception rax=000000000021d590 rbx=000000000021e420 rcx=000000000021d590 rdx=0000000000000000
15098.891:00d0:00d4:trace:seh:dispatch_exception rsi=0000000000000064 rdi=0000000000826440 r8=0000000000000000 r9=0000000000000000
15098.891:00d0:00d4:trace:seh:dispatch_exception r10=0000000000826580 r11=000000000082fff8 r12=000000000021e420 r13=0000000000000005
15098.891:00d0:00d4:trace:seh:dispatch_exception r14=00006ffffee7cdd4 r15=0000000000826440 mxcsr=00001fa0
15098.891:00d0:00d4:trace:unwind:RtlVirtualUnwind2 type 1 base 6fffffbf0000 rip 6fffffbfd197 rva d197 rsp 21d570
15098.891:00d0:00d4:trace:unwind:dump_unwind_info **** func d128-d19f
15098.891:00d0:00d4:trace:unwind:dump_unwind_info unwind info at 00006FFFFFCB1CE8 flags 0 prolog 0xf bytes function 00006FFFFFBFD128-00006FFFFFBFD19F
15098.891:00d0:00d4:trace:unwind:dump_unwind_info 0xf: subq $0xc8,%rsp
15098.891:00d0:00d4:trace:unwind:RtlVirtualUnwind2 inside epilog.
15098.891:00d0:00d4:trace:unwind:RtlVirtualUnwind2 type 1 base 6fffff8f0000 rip 6fffff920e56 rva 30e56 rsp 21d640
To add: Steam terminal output shows an odd entry being repeated on launch:
ATTENTION: default value of option vk_x11_override_min_image_count overridden by environment
pid 97450 != 97449, skipping destruction (fork without exec?)
Just did a clean reinstalled of CP2077, set to Proton Experimental and no launch arguments. Out of the box Red Launcher opens, but on pressing āplayā & split second black full screen I get dropped back to desktop with the Launcher frozen.
over the past month or so (and prior to applying this update) Iāve been having issues with certain games remaining stable and freezing/crashing with various amdgpu and/or dxvk-submitshader log entries after less than an hour of play. (Average of ~ 20 minutesā¦ low being ~2 min, and high being ~40 minā¦ before game/dGPU freeze/crash)
Hard to know if itās a game developer issue or something going on within our system packages, and/or dxvk in proton.
Curiously when reading Driver support Ā· doitsujin/dxvk Wiki Ā· GitHub, the recommendations are Vulkan 1.3 and RADV (mesa?) 25.1ā¦ which didnāt make sense as the 25 series is Vulkan 1.4. And Iāve also noticed that while dxvk is up to v2.7, proton is still back on the v2.6 seriesā¦ wondering now if GE versions close this gap?
At any rate, prior to applying this update, I had downgraded mesa to 25.1.7, switched between proton 9.04 and proton experimental, tried kernel 6.6, 6.12, and 6.16, and got to the point where things would be stable for ~2 hours before TG:FoA ultimately froze/crashed.
Post this update (applied yesterday) I have only played Hades (so far) and not one of the titles thatās been giving me grief (Tainted Grail: FoA)ā¦ so TBD what the overall experience will be now. Iāve taken the update to the latest mesa 25.2.3 and applied a newer linux612-6.12.50-1 kernel.
Odd, really odd.
On my machine the stable update went really well, no issues that stand out what I can remember. I also read your post and from mkb btw. But I do not know if it is related.
When did Vulkan get updated to 1.4, it was some months back I reckon
The most bizarre thing IMO is Cyberpunk should run out of the box if it is a clean install and set Proton Experimental. No launch arguments set etc.
For me it no longer does that as a clean install, which I do not know how to debug and it is really frustrating.
3 posts were merged into an existing topic: Chromium based browser core dump
