I think an attempt to add split_lock_detect=off
kernel parameter is worth trying.
[ 42.804085] [ T51] BUG: unable to handle page fault for address: fffff0618f00cfc8
...
[ 42.804835] [ T51] ? kfree+0x70/0x360
[ 42.804842] [ T51] ? _nv049909rm+0x29/0x60 [nvidia 1400000003000000474e55008f4d4216858fba5f]
[ 42.805560] [ T51] _nv047254rm+0x2d/0x50 [nvidia 1400000003000000474e55008f4d4216858fba5f]
[ 42.806062] [ T51] _nv045269rm+0xee/0x1b0 [nvidia 1400000003000000474e55008f4d4216858fba5f]
[ 42.806717] [ T51] _nv047231rm+0x374/0x410 [nvidia 1400000003000000474e55008f4d4216858fba5f]
[ 42.807328] [ T51] _nv014160rm+0x3f1/0x690 [nvidia 1400000003000000474e55008f4d4216858fba5f]
[ 42.807736] [ T51] _nv045408rm+0x26/0x30 [nvidia 1400000003000000474e55008f4d4216858fba5f]
...
NVIDIA binary driver tries to free some invalid memory. While it may be kernel issue, you have zero chances for any kernel developer to investigate it as long as NVIDIA is involved. Try reporting it to NVIDIA on Linux - NVIDIA Developer Forums
1 Like
Thanks for response. I will open topic there too, and update this one if i get answer.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.