[PATCH v3] gpu/drm/bridge/cadence: avoid flush_scheduled_work() usage

* [PATCH v3] gpu/drm/bridge/cadence: avoid flush_scheduled_work() usage
@ 2022-07-28  1:40 Tetsuo Handa
  2022-08-28 12:59 ` Tetsuo Handa
  0 siblings, 1 reply; 3+ messages in thread
From: Tetsuo Handa @ 2022-07-28  1:40 UTC (permalink / raw)
  To: Andrzej Hajda, Neil Armstrong, Robert Foss, Laurent Pinchart,
	Jonas Karlman, Jernej Skrabec, David Airlie, Daniel Vetter
  Cc: Andrew Morton, DRI

Like commit c4f135d643823a86 ("workqueue: Wrap flush_workqueue() using a
macro") says, flush_scheduled_work() is dangerous and will be forbidden.
We are on the way for removing all flush_scheduled_work() callers from
the kernel, and this patch is for removing flush_scheduled_work() call
 from cadence driver.

Since cdns-mhdp8546 driver uses 4 works

  mhdp->modeset_retry_work
  mhdp->hpd_work
  mhdp->hdcp.check_work
  mhdp->hdcp.prop_work

I assume that flush_scheduled_work() in cdns_mhdp_remove() needs to wait
for only these 4 works.

Since mhdp->modeset_retry_work already uses cancel_work_sync(), I assume
that flush_scheduled_work() needs to wait for only 3 works. But I came to
wonder whether mhdp->hdcp.check_work should be flushed or cancelled.

While flush_scheduled_work() waits for completion of works which were
already queued to system_wq, mhdp->hdcp.check_work is a delayed work.
That is, this work won't be queued to system_wq unless timeout expires.

Current code will wait for mhdp->hdcp.check_work only if timeout already
expired. If timeout is not expired yet, flush_scheduled_work() will fail
to cancel mhdp->hdcp.check_work, and cdns_mhdp_hdcp_check_work() which is
triggered by mhdp->hdcp.check_work will schedule hdcp->check_work, which
is too late for flush_scheduled_work() to wait for completion of
cdns_mhdp_hdcp_prop_work().

But since I couldn't get comments on how do we want to handle this race
window [1], this patch chose "do nothing" for mhdp->hdcp.check_work and
mhdp->hdcp.prop_work. That is, I assume that flush_scheduled_work() in
cdns_mhdp_remove() needs to wait for only mhdp->hpd_work work.

Link: https://lkml.kernel.org/r/943273cb-c2ec-24e3-5edb-64eacc6e2d30@I-love.SAKURA.ne.jp [1]
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
---
 drivers/gpu/drm/bridge/cadence/cdns-mhdp8546-core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/bridge/cadence/cdns-mhdp8546-core.c b/drivers/gpu/drm/bridge/cadence/cdns-mhdp8546-core.c
index ab63e7b11944..31442a922502 100644
--- a/drivers/gpu/drm/bridge/cadence/cdns-mhdp8546-core.c
+++ b/drivers/gpu/drm/bridge/cadence/cdns-mhdp8546-core.c
@@ -2605,7 +2605,8 @@ static int cdns_mhdp_remove(struct platform_device *pdev)
 	pm_runtime_disable(&pdev->dev);
 
 	cancel_work_sync(&mhdp->modeset_retry_work);
-	flush_scheduled_work();
+	flush_work(&mhdp->hpd_work);
+	/* Ignoring mhdp->hdcp.check_work and mhdp->hdcp.prop_work here. */
 
 	clk_disable_unprepare(mhdp->clk);
 
-- 
2.18.4

^ permalink raw reply related	[flat|nested] 3+ messages in thread