Summary: | x11-drivers/nvidia-drivers with kernel 3.18.0 - WARNING: CPU: 1 PID: 2396 at drivers/gpu/drm/drm_ioctl.c:143 drm_setversion+0xd7/0x16b() // No drm_driver.set_busid() implementation provided by nvidia_frontend_exit_module [nvidia]. | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Pavel Volkov <ao> |
Component: | [OLD] Library | Assignee: | Jeroen Roovers (RETIRED) <jer> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | cornicx, markpariente, pmcdermott98, ua_gentoo_bugzilla, yamadharma |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
URL: | https://devtalk.nvidia.com/default/topic/796559/kernel-3-18-warning-no-drm_driver-set_busid-implementation-provided-by-nvidia_frontend_exit_modu/ | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
nvidia drm patch
nvidia drm fix for 346.35 |
Description
Pavel Volkov
2014-12-09 21:15:28 UTC
Everything seems to just work fine despite this error message. Maybe I should downgrade the supported version again just in case? I think it can be reported to upstream if they claim 3.18 is supported. (In reply to Pavel Volkov from comment #2) > I think it can be reported to upstream if they claim 3.18 is supported. Well, that's always tricky. [1] claims "Improved compatibility with recent Linux kernels" and came out on 8 December, a day after kernel 3.18 was released. So far the forums[2] do not mention any 3.18 problems, so maybe you should report it there. [1] http://www.nvidia.com/download/driverResults.aspx/80652/en-us [2] https://devtalk.nvidia.com/default/board/98/ I'm found a better solution: installed a card with AMD chip (switched mostly because of the noise of my old card but the driver openness played its part as well). I'm leaving this bug open since you also see those messages. Created attachment 392698 [details, diff]
nvidia drm patch
I solved the problem with a patch from: https://devtalk.nvidia.com/default/topic/796645/346-22-343-36-regression-no-dp1-2-daisy-chaining-on-gtx770-worked-with-343-22-/ I patched the files manually. Perhaps someone can make a patch for the ebuild. (In reply to Mario Bachmann from comment #7) > I solved the problem with a patch from: > https://devtalk.nvidia.com/default/topic/796645/346-22-343-36-regression-no- > dp1-2-daisy-chaining-on-gtx770-worked-with-343-22-/ Amazing that you solved it after I pointed it out in comment #5. > I patched the files manually. You could have used epatch_user. > Perhaps someone can make a patch for the ebuild. Don't bother. It isn't going into the tree as per the usual policy. *** Bug 536278 has been marked as a duplicate of this bug. *** Created attachment 394292 [details, diff]
nvidia drm fix for 346.35
It looks like for nvidia-drivers 346.35 the inode fixes from the previous patch aren't needed anymore (upstream wrapped access with NV_FILE_INODE which does the proper thing) - only fix needed is the drm part in nv-drm.c.
interesting that I have no message like that my logs. Just the usual taints kernel. Wonder if it is board specific. kernel 3.18.5-gentoo also have this problem :( with x11-drivers/nvidia-drivers-346.35 (In reply to tman from comment #12) > kernel 3.18.5-gentoo > > > also have this problem :( with x11-drivers/nvidia-drivers-346.35 No me-toos required. nvidia_drm_fix_for_346.35 fixed this problem on my system. *** Bug 541184 has been marked as a duplicate of this bug. *** 346.47 no longer has this issue. |