Pi4/fkms kernel warnings: vc4_crtc_atomic_flush and drm_vblank_put #3020
Labels
Comments
|
Can I ask what the thought process was behind going with the firmware kms for Raspbian? Were there some hardware tweaks to the HVS in the Rpi4? I figure there must be for 4k support, if we can get some information on those, I'd love to get those into the mainline KMS. |
|
This is all 3d related rather than HVS. The 3d hardware in Pi0-3 has a GLES2 driver written for the GPU (the original chip is was designed for didn't even have an arm core). Since then an arm side driver has been written (open source) and to use it rather than the GPU driver you use fkms. Pi4 has a newer version of the 3d hardware (to support 4k, and newer standards like GLES3) which has never had a GPU driver written for it. So fkms is the only option that supports 3d on Pi4. |
|
Sorry, I was unclear. In Pi 0-3, you had the old V3D 2.x, but you also had the HVS with full atomic KMS support - no need for fkms. Pi 4 has a V3D 4.2, but does it also have the old HVS from VC4, something else entirely, or no compositor hardware at all anymore? |
|
The HVS is present and very similar to previous Pi models. (Support for 4K and HDR is added). I think you may be talking about the kms driver (arm side HVS/HDMI driver). |
|
(note: the fkms driver is now much more fully featured. It supports multiple displays and mode setting. There's not an really any missing functionality compared to kms). |
|
It's a modified HVS, pixel valve, and a new HDMI block. The firmware drivers for these were written anyway for device bringup. FKMS also has the benefit of retaining DispmanX's ability to add overlays from the camera, omxplayer, etc, therefore avoided a hard requirement of rewriting all those apps (twice, as displaying stuff through X is different from console). |
|
I'm seeing these kernel warnings on a twin headed machine. HDMI-1 fails to come out of screen-save, and sometimes HDMI-0 then goes blank if the screen saver kicks in again. Will disabling the screen saver avoid the problem in the short term ? |
As per the paragraph at the end of the first comment
These warnings are nothing to do with the actual mode switching, just the reporting of vblank events up the stack. Absolutely EVERYONE will see these messages in the kernel log, which is why I created this issue before anyone else reported it. I could have just nuked them, but they do correctly indicate an issue that we haven't resolved, so that would really be the wrong thing to do. |
|
I've not seen any mention of the screen saver behaviour so I was trying to give you more information, but it seems that was unwelcome..... |
|
Very first paragraph of the first post
Your post seemed to be a "me too" which is not helpful. Apologies if I was too terse - things are rather busy at the moment. |
|
I meant I'd not seen any mention when I emailed James, he pointed me here, so I just repeated what I had told him as I assumed that's why he'd pointed me here... Sorry if I misunderstood... |
6by9
added a commit
to 6by9/linux
that referenced
this issue
Jul 25, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. raspberrypi#3020 Signed-off-by: Dave Stevenson <[email protected]>
6by9
added a commit
to 6by9/linux
that referenced
this issue
Aug 6, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. raspberrypi#3020 Signed-off-by: Dave Stevenson <[email protected]>
6by9
added a commit
to 6by9/linux
that referenced
this issue
Aug 6, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. raspberrypi#3020 Signed-off-by: Dave Stevenson <[email protected]>
pelwell
pushed a commit
that referenced
this issue
Aug 6, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
pelwell
pushed a commit
that referenced
this issue
Sep 13, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
pelwell
pushed a commit
that referenced
this issue
Sep 13, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
pelwell
pushed a commit
that referenced
this issue
Sep 13, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
pelwell
pushed a commit
that referenced
this issue
Sep 13, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Sep 17, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
mike-scott
pushed a commit
to mike-scott/linux
that referenced
this issue
Sep 17, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. raspberrypi/linux#3020 Signed-off-by: Dave Stevenson <[email protected]> Signed-off-by: Michael Scott <[email protected]>
pelwell
pushed a commit
that referenced
this issue
Sep 19, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
pelwell
pushed a commit
that referenced
this issue
Sep 19, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Oct 4, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Oct 4, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Oct 11, 2019
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jun 17, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jun 26, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jun 26, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
rsglobal
pushed a commit
to GloDroid/glodroid_forks
that referenced
this issue
Jun 26, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. raspberrypi/linux#3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jul 1, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jul 1, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jul 13, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jul 13, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Jul 20, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Aug 10, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
rsglobal
pushed a commit
to GloDroid/glodroid_forks
that referenced
this issue
Aug 12, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. raspberrypi/linux#3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Aug 12, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Aug 15, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Aug 15, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Aug 19, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Aug 19, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Sep 1, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Sep 1, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
6by9
added a commit
to 6by9/linux
that referenced
this issue
Sep 4, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. raspberrypi#3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Sep 11, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Sep 15, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Sep 28, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Oct 2, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Oct 7, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Oct 16, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Oct 19, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
popcornmix
pushed a commit
that referenced
this issue
Nov 4, 2020
The details over when and how a driver is to service the vblank events are sketchy, and the fkms driver was triggering a kernel warning every time the crtc was enabled or disabled. Copy the event handling as used by the vc4-kms driver slightly more closely, and we avoid the warnings. #3020 Signed-off-by: Dave Stevenson <[email protected]>
|
@6by9 okay to close? |
|
Should be resolved, yes. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This is more a placeholder to say that we are aware of the two kernel warnings that trigger whenever the screen mode changes or blanks (ie screensaver) with the Pi4. eg:
These are harmless and due to the timing on when we are producing vblank events being subtly wrong. We are investigating these already.
The text was updated successfully, but these errors were encountered: