Package: src:linux Version: 6.12.95+deb13-amd64 Control: affects -1 libgstreamer1.0-0:amd64
To prepare for Debian “forky”, which no longer allows users to run GNOME+Xorg, we decided to switch from Xorg to Wayland. However, this incurs serious issues; cf. http://bugs.debian.org/1125785. As of now, on Debian stable, with # WaylandEnable=false in the [daemon] section of /etc/gdm3/daemon.conf, Wayland doesn't start automatically (Xorg starts instead). To get Wayland running, we wrote GRUB_CMDLINE_LINUX="nouveau.config=NvGspRm=0" in /etc/default/grub and ran update-grub. (There might be other methods, but I'm unaware of them at the moment.) However, this way we get tons of repeated errors in the journal: $ uptime 02:22:49 up 44 min, 1 user, load average: 0,40, 0,38, 0,43 $ sudo journalctl -b| egrep -i "nouveau.*(error|fault|fail|warn|unknown|not found|miss)" | wc -l 2968 $ sudo journalctl -b| egrep -i "nouveau.*(error|fault|fail|warn|unknown|not found|miss)" | cut -d ' ' -f 7- | sort | uniq 0000:65:00.0: drm: BIT table 'A' not found 0000:65:00.0: drm: BIT table 'L' not found 0000:65:00.0: fifo:000000:0004:0004:[gst-plugin-scan[1684]] errored - disabling channel 0000:65:00.0: fifo:000000:0004:0004:[gst-plugin-scan[1859]] errored - disabling channel 0000:65:00.0: fifo:000000:0008:0008:[gst-plugin-scan[3020]] errored - disabling channel 0000:65:00.0: fifo:000000:0008:0008:[gst-plugin-scan[3062]] errored - disabling channel 0000:65:00.0: fifo:000000:0009:0009:[gst-plugin-scan[3418]] errored - disabling channel 0000:65:00.0: fifo:000000:0009:0009:[gst-plugin-scan[3440]] errored - disabling channel 0000:65:00.0: fifo:000000:0009:0009:[gst-plugin-scan[4982]] errored - disabling channel 0000:65:00.0: fifo:000000:0009:0009:[gst-plugin-scan[4987]] errored - disabling channel 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 000000000001c000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 4 [017f811000 gst-plugin-scan[1684]] 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 000000000001c000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 4 [017f819000 gst-plugin-scan[1859]] 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 000000000001c000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 8 [017f86c000 gst-plugin-scan[3062]] 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 000000000001c000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 9 [017f84b000 gst-plugin-scan[4987]] 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 000000000001c000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 9 [017f85b000 gst-plugin-scan[3440]] 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 00000000007f8000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 8 [017f864000 gst-plugin-scan[3020]] 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 0000000003823000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 9 [017f845000 gst-plugin-scan[4982]] 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 0000000003823000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 9 [017f853000 gst-plugin-scan[3418]] 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 2 [017fe43000 gnome-shell[1546]] subc 0 class c597 mthd 0d78 data 00000004 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 2 [017fe43000 gnome-shell[1546]] subc 0 class c597 mthd 17e0 data 00000012 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 2 [017fe43000 gnome-shell[1546]] subc 0 class c597 mthd 17e0 data 00000018 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 2 [017fe43000 gnome-shell[1546]] subc 0 class c597 mthd 17e0 data 0000002a 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 0d78 data 00000004 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 0d78 data 0000000c 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 0d78 data 00000012 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 0d78 data 00000018 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 0d78 data 00000030 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 17e0 data 00000012 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 17e0 data 00000018 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 17e0 data 00000030 0000:65:00.0: gr: DATA_ERROR 0000009c [] ch 6 [017f885000 gnome-shell[2896]] subc 0 class c597 mthd 17e0 data 00000036 $ sudo lspci -v | grep 65:00.0 65:00.0 VGA compatible controller: NVIDIA Corporation TU116 [GeForce GTX 1660] (rev a1) (prog-if 00 [VGA controller]) The messages “BIT table 'A' not found” and “BIT table 'L' not found” are purely informational (not highlighted as warnings/errors). The graphics card is NVIDIA PΛLiΤ GeForce GTX 1660, to which currently two cables are connected: a DisplayPort cable (with a 1920×1200-capable monitor on the other end) and a DVI cable (with a 1920×1080-capable monitor on the other end). The motherboard is Asus WS C422 PRO SE. firmware-nvidia-graphics 20250410-2 is installed. As these errors clutter the journal (in particular, distract from other issues), they must be dealt with (and, if they cannot be dealt with, they should be silenced or their rate should drop, especially when they repeat). The machine is capable of running Windows 11 without visible graphics-related issues, so this seems to be more a software-related problem in the kernel (perhaps in nouveau, gst-plugin-streamer, or some other culprit calling nouveau). The very first error in nouveau since boot is this: nouveau 0000:65:00.0: fifo: fault 01 [VIRT_WRITE] at 000000000001c000 engine 40 [gr] client 13 [GPC0/PROP_0] reason 02 [PTE] on channel 4 [017f811000 gst-plugin-scan[1684]] nouveau 0000:65:00.0: fifo:000000:0004:[gst-plugin-scan[1684]] rc scheduled nouveau 0000:65:00.0: fifo:000000: rc scheduled nouveau 0000:65:00.0: fifo:000000:0004:0004:[gst-plugin-scan[1684]] errored - disabling channel Though I read that nouveau is reverse-engineered, I'd still be grateful for a fix …
