The composer has one switch, shared by the CRC reader and the writeback
jobs, so either of them turns it off for the other.

Split it: vkms_output keeps a count of holders, and a get() raises it while a
put() lowers it, so the vblank reference is taken by the first holder and
dropped by the last.  The CRC source takes and drops one per open()/release(),
which the core always pairs; each writeback job keeps its own record.

Patch 2 keeps the lock over the count and the queue, so the state the timer
read is still the state it queues.

---
Changes in v3:
- the CRC source no longer keeps a flag of its own: it takes and drops one
  reference per open()/release().  With a flag, a second reader skipped its
  get() while the first one was still releasing, and that release then
  dropped the reference for both.
- vkms_composer_get() returns whether it took a reference; a writeback job
  keeps that value and only drops what it took.
- Link to v2: 
https://patch.msgid.link/[email protected]

---
Oarora Etimis (2):
      drm/vkms: count the composer users
      drm/vkms: queue the composer work under the output lock

 Documentation/gpu/vkms.rst            |  5 -----
 drivers/gpu/drm/vkms/vkms_composer.c  | 33 +++++++++++++++++++++++++--------
 drivers/gpu/drm/vkms/vkms_crtc.c      |  5 ++---
 drivers/gpu/drm/vkms/vkms_drv.h       | 29 +++++++++++++++++++++++++----
 drivers/gpu/drm/vkms/vkms_writeback.c |  7 ++++---
 5 files changed, 56 insertions(+), 23 deletions(-)
---
base-commit: 97429bac371273886238bfcf63d78e7405cacb46
change-id: 20261111-composer-cc9620515bcd

Best regards,
--  
Oarora Etimis <[email protected]>

Reply via email to