• Too many crashes in a adventure with nouveau/wayland


    I was adventuring with nouveau/wayland and I found that my gnome-shell was more responsive and faster but as nothing is perfect my system started to crash so often that I went back to proprietary.
    Anyone else is having same issue?

    I think that is related to buffer allocation and in the logs I got this:

    1 - Changed chromium to fullscreen to watch youtube:
    kernel: nouveau 0000:01:00.0: Xwayland[1012]: channel 13 killed!
    kernel: nouveau 0000:01:00.0: fifo: engine 0: scheduled for recovery
    kernel: nouveau 0000:01:00.0: fifo: runlist 0: scheduled for recovery
    kernel: nouveau 0000:01:00.0: fifo: channel 13: killed
    kernel: nouveau 0000:01:00.0: fifo: read fault at 000a5e0000 engine 00 [GR] client 04 [GPC0/T1_1] reason 02 [PTE] on channel 13 [007ed50000 Xwayland[1012]]
    kernel: nouveau 0000:01:00.0: gr: GPC0/TPC0/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: TRAP ch 13 [007ed50000 Xwayland[1012]]
    kernel: nouveau 0000:01:00.0: gr: GPC2/TPC1/TEX: 80000009
    kernel: nouveau 0000:01:00.0: gr: GPC2/TPC0/TEX: 80000009
    kernel: nouveau 0000:01:00.0: gr: GPC1/TPC1/TEX: 80000009
    kernel: nouveau 0000:01:00.0: gr: GPC1/TPC0/TEX: 80000009
    kernel: nouveau 0000:01:00.0: gr: GPC0/TPC1/TEX: 80000009
    kernel: nouveau 0000:01:00.0: gr: GPC0/TPC0/TEX: 80000009

    2 - Random things that I don’t remember:
    kernel: nouveau 0000:01:00.0: Xwayland[1009]: channel 3 killed!
    kernel: nouveau 0000:01:00.0: fifo: engine 0: scheduled for recovery
    kernel: nouveau 0000:01:00.0: fifo: runlist 0: scheduled for recovery
    kernel: nouveau 0000:01:00.0: fifo: channel 3: killed
    kernel: nouveau 0000:01:00.0: fifo: read fault at 0002de2000 engine 00 [GR] client 0f [GPC1/PROP_0] reason 02 [PTE] on channel 3 [007f82e000 Xwayland[1009]]
    kernel: nouveau 0000:01:00.0: gr: GPC2/TPC1/MP trap: global 00000004 [MULTIPLE_WARP_ERRORS] warp 3000d [OOR_REG]
    kernel: nouveau 0000:01:00.0: gr: TRAP ch 3 [007f82e000 Xwayland[1009]]

    3 - 500.000+ lines of same thing just for typing a text and dragging gnome-calculator to other position:
    kernel: nouveau 0000:01:00.0: fifo: SCHED_ERROR 06 []

    4 - New trial after last update:
    kernel: nouveau 0000:01:00.0: fifo: read fault at 0002c22000 engine 00 [GR] client 01 [GPC2/T1_0] reason 02 [PTE] on channel 3 [007f82e000 Xwayland[964]]
    kernel: nouveau 0000:01:00.0: gr: GPC2/TPC1/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC2/TPC0/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC1/TPC1/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC1/TPC0/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC0/TPC1/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC0/TPC0/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: TRAP ch 3 [007f82e000 Xwayland[964]]
    kernel: nouveau 0000:01:00.0: Xwayland[964]: channel 13 killed!
    kernel: nouveau 0000:01:00.0: fifo: engine 0: scheduled for recovery
    kernel: nouveau 0000:01:00.0: fifo: runlist 0: scheduled for recovery
    kernel: nouveau 0000:01:00.0: fifo: channel 13: killed
    kernel: nouveau 0000:01:00.0: fifo: read fault at 0007280000 engine 00 [GR] client 01 [GPC0/T1_0] reason 02 [PTE] on channel 13 [007ed4e000 Xwayland[964]]
    kernel: nouveau 0000:01:00.0: gr: GPC2/TPC1/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC2/TPC0/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC1/TPC1/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC1/TPC0/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC0/TPC1/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: GPC0/TPC0/TEX: 80000049
    kernel: nouveau 0000:01:00.0: gr: TRAP ch 13 [007ed4e000 Xwayland[964]]

  • I can’t use noveau here as it do crashing under GNOME and i get screen flickering, till the change to GNOME 3.26

    [updates once a week] = [90% less problems]
    my#repo
    my-blog#k
    how to add system logs
    i3 GNOME

  • @jeannie____ said in Too many crashes in a adventure with nouveau/wayland:

    Nouveau? Seriously?
    J.

    There is no red team cards available where I live, the cryptocurrency miners took all of them and retailers who still have is asking for a unreasonable salty prices.
    Well, I don’t know if the market will recover from that this year, but I was wanting to play a little with Wayland to see what is working and what not, but after a try out and crash, I got curious if it was just me having this issues.

    At this time for who likes rolling distros, Nvidia sucks, gramps xorg did everything that could do and it’s time to retire it. I hope Nvidia change their mind…

  • @edwin-foss
    If wayland only had advantages for the users.
    J.

  • @jeannie____ said in Too many crashes in a adventure with nouveau/wayland:

    @edwin-foss
    If wayland only had advantages for the users.
    J.

    I see what you are looking for, but I’m not a “benchmarker”, I’m more likely to support new ideas and things, for me learning sooner is better than later and more people using will be more likely to be optimized, well, mainstream effect maybe?. Anyway, servers and supercomputers Linux won the race but in desktop market we are the last ones, if I sit stand and start to bash things in the name of traditional mindset I will never go forward in the future, but this is just my thoughts…
    I wished I could help more but I learned basic and pascal languages only (you can call me a dinosaur 😁 ), C and Assembler wasn’t available in my country when I was young, old habits die hard… 😞

  • I’m a user, and I don’t care what I use as long as it works . From my perspective, new things should do two things:
    Improve something.
    Work as the old thing or better.
    Wayland does not do either of these, it just complicates things and removes functionality.
    J.

  • @jeannie____ said in Too many crashes in a adventure with nouveau/wayland:

    I’m a user, and I don’t care what I use as long as it works . From my perspective, new things should do two things:
    Improve something.
    Work as the old thing or better.
    Wayland does not do either of these, it just complicates things and removes functionality.
    J.

    Sorry but at this time you are wrong, nobody will support old software forever, everything has TTL, everything dies some day, all Xorg team is getting older and sooner or later they will retire and none on younger devs seems interested in maintain that old mess with ancient code.
    Well, I’m getting older too but I still know how is to be young, we love how new things smells, as everything in FLOSS is community centric, so many guys spent many, many hours doing a hard work trying to cut the bloat off and make things more attractive to new devs and I’m so grateful for them doing so.

    Sorry but it’s better that our discussion stop here, it’s a useless waste of time, we are different mindsets 😂

    "The definition of insanity is doing the same thing over and over again and expecting a different result."
    -I don’t know who said that, but certainly wasn’t Albert Einstein-

    Well, thanks anyway,
    Best regards Mr. J.

  • @edwin-foss said in Too many crashes in a adventure with nouveau/wayland:

    Sorry but at this time you are wrong, nobody will support old software forever, everything has TTL, everything dies some day, all Xorg team is getting older and sooner or later they will retire and none on younger devs seems interested in maintain that old mess with ancient code.
    Well, I’m getting older too but I still know how is to be young, we love how new things smells, as everything in FLOSS is community centric, so many guys spent many, many hours doing a hard work trying to cut the bloat off and make things more attractive to new devs and I’m so grateful for them doing so.

    Great words!

    –closing– on user request 😉

    [updates once a week] = [90% less problems]
    my#repo
    my-blog#k
    how to add system logs
    i3 GNOME

crashes16 adventure1 nouveauwayland1 Posts 9Views 183
Log in to reply
Bloom Email Optin Plugin

Looks like your connection to Antergos Community Forum was lost, please wait while we try to reconnect.