Dear Maintainer,

after mesa upgrade to 25.0.0-1
[UPGRADE] libegl-mesa0:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] libgbm1:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] libgl1-mesa-dri:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] libglx-mesa0:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] libosmesa6:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] libxatracker2:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] mesa-libgallium:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] mesa-va-drivers:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] mesa-vdpau-drivers:amd64 24.3.4-3 -> 25.0.0-1
[UPGRADE] mesa-vulkan-drivers:amd64 24.3.4-3 -> 25.0.0-1

my ffplay crashs most often (but not always) after a few seconds if I try to
view my reolink RLC-18MA cam via
ffplay -hwaccel vulkan -an -sync video  rtsp://admin:${PASS}@cam:554/

[Vulkan Renderer @ 0x55922d42b100]     External API interop:
[Vulkan Renderer @ 0x55922d42b100]       UUID:
00:00:00:00:07:00:00:00:00:00:00:00:00:00:00:00
[Vulkan Renderer @ 0x55922d42b100]       PCI: 0000:07:00:0
[Vulkan Renderer @ 0x55922d42b100]       buf export caps: 0x9
[Vulkan Renderer @ 0x55922d42b100]       buf import caps: 0x19
[Vulkan Renderer @ 0x55922d42b100]       tex export caps: 0x9
[Vulkan Renderer @ 0x55922d42b100]       tex import caps: 0x19
[Vulkan Renderer @ 0x55922d42b100]       sync export caps: 0x1
[Vulkan Renderer @ 0x55922d42b100]       sync import caps: 0x0
[Vulkan Renderer @ 0x55922d42b100] Available surface configurations:
[Vulkan Renderer @ 0x55922d42b100]     0: VK_FORMAT_B8G8R8A8_SRGB
VK_COLOR_SPACE_SRGB_NONLINEAR_KHR
[Vulkan Renderer @ 0x55922d42b100]     1: VK_FORMAT_B8G8R8A8_UNORM
VK_COLOR_SPACE_SRGB_NONLINEAR_KHR
[Vulkan Renderer @ 0x55922d42b100] Picked surface configuration 1:
VK_FORMAT_B8G8R8A8_UNORM + VK_COLOR_SPACE_SRGB_NONLINEAR_KHR
Input #0, rtsp, from 'rtsp://admin:$PASS@cam:554/':
  Metadata:
    title           : Session streamed by "preview"
    comment         :
  Duration: N/A, start: 0.051438, bitrate: N/A
  Stream #0:0: Video: hevc (Main), yuv420p(tv), 3840x2160, 20 fps, 20 tbr, 90k
tbn
  Stream #0:1: Audio: aac (LC), 16000 Hz, mono, fltp
[hevc @ 0x7fe77008c680] PPS changed between slices.
[hevc @ 0x7fe77008c680] Skipping invalid undecodable NALU: 1
[hevc @ 0x7fe77008c680] Could not find ref with POC 22
[Vulkan Renderer @ 0x55922d42b100] Spent 62.268 ms translating SPIR-V
[Vulkan Renderer @ 0x55922d42b100] Spent 41.344 ms allocating slab
[Vulkan Renderer @ 0x55922d42b100] Spent 13.360 ms generating shader LUT
[Vulkan Renderer @ 0x55922d42b100] Dithering to 8 bit depth
Segmentation fault (core dumped)   0KB vq=   75KB sq=    0B

If I download the stream to disk (via vlc) and play it from there with ffplay
it doesn't core dump.

Without -hwaccel it also works fine. I also downgraded above packages to
24.3.4-3 and then it's working again with vulkan hwaccel.
[INSTALL, DEPENDENCIES] libglapi-mesa:amd64 24.3.4-3
[DOWNGRADE] libegl-mesa0:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] libgbm1:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] libgl1-mesa-dri:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] libglx-mesa0:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] libosmesa6:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] libxatracker2:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] mesa-libgallium:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] mesa-va-drivers:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] mesa-vdpau-drivers:amd64 25.0.0-1 -> 24.3.4-3
[DOWNGRADE] mesa-vulkan-drivers:amd64 25.0.0-1 -> 24.3.4-3

I tried my other three reolink cams.

With "E1 Pro",RLC-811A and RLC-522 I don't get a core dump :-o

If it crashs I get the following messages

Feb 25 14:44:08 plasmashell[104950]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:79: TypeError: Cannot read property 'status' of null
Feb 25 14:44:14 xdg-desktop-por[104732]: Realtime error: Could not get pidns: Could not fstatat ns/pid: Not a directory
Feb 25 14:44:14 kwin_wayland_wrapper[104718]: kf.windowsystem: static bool KX11Extras::mapViewport() may only be used on X11
Feb 25 14:44:14 plasmashell[104950]: file:///usr/share/plasma/plasmoids/org.kde.plasma.taskmanager/contents/ui/Task.qml:79: TypeError: Cannot read property 'status' of null
Feb 25 14:44:15 kernel: av:hevc:df5[171082]: segfault at 0 ip 0000000000000000 sp 00007f2074fdf568 error 14 likely on CPU 6 (core 6, socket 0)
Feb 25 14:44:15 kernel: Code: Unable to access opcode bytes at 0xffffffffffffffd6.
Feb 25 14:44:15 systemd-coredump[171138]: Process 171032 (ffplay) of user 1000 terminated abnormally with signal 11/SEGV, processing...
Feb 25 14:44:15 systemd[1]: Started systemd-coredump@13-171138-0.service - Process Core Dump (PID 171138/UID 0).
Feb 25 14:44:15 systemd[1]: Started drkonqi-coredump-processor@13-171138-0.service - Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Feb 25 14:44:16 systemd-coredump[171139]: [🡕] Process 171032 (ffplay) of user 1000 dumped core.

                                           Module libblkid.so.1 from deb util-linux-2.40.4-4.amd64
                                           Module libmount.so.1 from deb util-linux-2.40.4-4.amd64
                                           Module libzstd.so.1 from deb libzstd-1.5.6+dfsg-2.amd64
                                           Module libgfortran.so.5 from deb gcc-14-14.2.0-17.amd64
                                           Module libsystemd.so.0 from deb systemd-257.3-1.amd64
                                           Module libudev.so.1 from deb systemd-257.3-1.amd64
                                           Module libatomic.so.1 from deb gcc-14-14.2.0-17.amd64
                                           Module libgomp.so.1 from deb gcc-14-14.2.0-17.amd64
                                           Module libgcc_s.so.1 from deb gcc-14-14.2.0-17.amd64
                                           Module libstdc++.so.6 from deb gcc-14-14.2.0-17.amd64
                                           Stack trace of thread 171082:
                                           #0  0x0000000000000000 n/a (n/a + 0x0)
                                           ELF object binary architecture: AMD x86-64
Feb 25 14:44:16 systemd[1]: systemd-coredump@13-171138-0.service: Deactivated successfully.
Feb 25 14:44:16 systemd[1]: systemd-coredump@13-171138-0.service: Consumed 972ms CPU time, 932.3M memory peak.
Feb 25 14:44:17 drkonqi-coredump-processor[171140]: "/usr/bin/ffplay" 171032 "/var/lib/systemd/coredump/core.ffplay.1000.fbf0945924de4018a4a103bd7b9f550e.171032.1740491055000000.zst"
Feb 25 14:44:17 systemd[101879]: Started drkonqi-coredump-launcher@27-171140-0.service - Launch DrKonqi for a systemd-coredump crash (PID 171140/UID 0).
Feb 25 14:44:17 systemd[1]: drkonqi-coredump-processor@13-171138-0.service: Deactivated successfully.
Feb 25 14:44:17 drkonqi-coredump-launcher[171150]: Unable to find file for pid 171032 expected at "kcrash-metadata/ffplay.fbf0945924de4018a4a103bd7b9f550e.171032.ini"
Feb 25 14:44:17 drkonqi-coredump-launcher[171150]: Nothing handled the dump :O

Daniel Schröter kirjoitti 25.2.2025 klo 9.38:

Hi,

Please test with 25.0.1 which was uploaded today. If it's still broken,
consider filing this upstream at

https://gitlab.freedesktop.org/mesa/mesa/-/issues

Not fixed in 25.0.1 :-(

Im going to report it upstream next days....

Looks similar to
https://gitlab.freedesktop.org/mesa/mesa/-/issues/12764

Quite hard to report a bug on freedesktop. I hit quite often spam detection :-(

I found a "solution" for me.

If I add /h264Preview_01_main to the URL ffplay does not crash.

Should we close this bug?