Bug report concerning Wine and NMVe drivers

[lavadrop]

I found an ugly bug that now I can point to in a general sort of way, but I don't know if it's Wine or a kernel driver issue that manifests because of running World of Warcraft through Wine.
I was a WoW player for a very long time until Activision/Blizzard pissed me off last year and decided to end my sub and remove their programs for my computer.
I originally started playing on a Mac, but I moved to a new DIY PC I put together with Linux instead of Windows. So like most of you, I distrohopped for a while, it was I think Ubuntu->Ubuntu Budgie-Kubuntu-Solus-OpenSUSE. In the beginning I just had 2 SATA SSDs. I eventually bought 2 NVME drives around the time I hopped to Kubuntu and that's when my problems started, regardless of which NVMe drive is mounted or where the WoW program is ran from.

This is the dmesg log that showed on Kubuntu, Solus and now OpenSUSE, regardless of file system:

Code: Select all

[34852.744346] nvme nvme1: controller is down; will reset: CSTS=0xffffffff, PCI_STATUS=0x10
[34852.744354] nvme nvme1: Does your device have a faulty power saving mode enabled?
[34852.744356] nvme nvme1: Try "nvme_core.default_ps_max_latency_us=0 pcie_aspm=off" and report a bug
[34852.812934] nvme1n1: I/O Cmd(0x2) @ LBA 227547560, 8 blocks, I/O Error (sct 0x3 / sc 0x71) 
[34852.812945] I/O error, dev nvme1n1, sector 227547560 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2
[34852.848347] nvme 0000:25:00.0: enabling device (0000 -> 0002)
[34852.848431] nvme nvme1: Removing after probe failure status: -19
[34852.884354] nvme1n1: detected capacity change from 1953525168 to 0
[34852.884352] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 0, rd 1, flush 0, corrupt 0, gen 0
[34852.884359] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 1, rd 2, flush 0, corrupt 0, gen 0
[34852.884372] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 2, rd 2, flush 0, corrupt 0, gen 0
[34852.884388] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 3, rd 2, flush 0, corrupt 0, gen 0
[34852.884397] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 4, rd 2, flush 0, corrupt 0, gen 0
[34852.884417] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 6, rd 2, flush 0, corrupt 0, gen 0
[34852.884417] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 5, rd 2, flush 0, corrupt 0, gen 0
[34852.884441] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 7, rd 2, flush 0, corrupt 0, gen 0
[34852.884452] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 8, rd 2, flush 0, corrupt 0, gen 0
[34852.884462] BTRFS error (device nvme1n1p1): bdev /dev/nvme1n1p1 errs: wr 9, rd 2, flush 0, corrupt 0, gen 0
[34852.889921] BTRFS: error (device nvme1n1p1) in btrfs_commit_transaction:2436: errno=-5 IO failure (Error while writing out transaction)
[34852.889933] BTRFS info (device nvme1n1p1: state E): forced readonly
[34852.889937] BTRFS warning (device nvme1n1p1: state E): Skipping commit of aborted transaction.
[34852.889939] BTRFS: error (device nvme1n1p1: state EA) in cleanup_transaction:1983: errno=-5 IO failure
[34873.476412] btrfs_dev_stat_inc_and_print: 106 callbacks suppressed
[34873.476416] BTRFS error (device nvme1n1p1: state EA): bdev /dev/nvme1n1p1 errs: wr 116, rd 2, flush 0, corrupt 0, gen 0
[35282.491105] perf: interrupt took too long (3195 > 3127), lowering kernel.perf_event_max_sample_rate to 62500

The system becomes unresposive for a few seconds, background music loops, then the window server becomes responsive again but the WoW program is locked, no way but to kill it and restart it, except that now my system has mounted the NVMe drive WoW was running from as read only and it's removed from /dev. Only recourse is rebooting.

I think this is being triggered by Wine/Lutris because a year ago I removed everything needed to run World of Warcraft from my NVME drive and since then I exclusively ran games using Steam/Proton, without issues. Today I reinstalled WoW under Lutris and after a few minutes of running it, the same thing last year I suffered on Solus came back.

I don't know if I should file a bug report here or on the kernel issue tracker or the lutris issue tracker.

[invisible kid]

I'm probably missing something, but if those errors are getting reported via dmesg wouldn't that mean the problem is separate from wine? This is just an off-the-cuff guess, but could there be a hardware or hardware communication issue going on? Either way, hope you get fixed up.