public inbox for intel-xe@lists.freedesktop.org
 help / color / mirror / Atom feed
From: Maarten Lankhorst <dev@lankhorst.se>
To: intel-gfx@lists.freedesktop.org, intel-xe@lists.freedesktop.org
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
	Clark Williams <williams@redhat.com>,
	Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
	Maarten Lankhorst <dev@lankhorst.se>
Subject: [i915-rt v5 12/21] drm/i915/gt: Use spin_lock_irq() instead of local_irq_disable() + spin_lock()
Date: Wed, 21 Jan 2026 14:53:08 +0100	[thread overview]
Message-ID: <20260121135318.651622-13-dev@lankhorst.se> (raw)
In-Reply-To: <20260121135318.651622-1-dev@lankhorst.se>

From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>

execlists_dequeue() is invoked from a function which uses
local_irq_disable() to disable interrupts so the spin_lock() behaves
like spin_lock_irq().
This breaks PREEMPT_RT because local_irq_disable() + spin_lock() is not
the same as spin_lock_irq().

execlists_dequeue_irq() and execlists_dequeue() has each one caller
only. If intel_engine_cs::active::lock is acquired and released with the
_irq suffix then it behaves almost as if execlists_dequeue() would be
invoked with disabled interrupts. The difference is the last part of the
function which is then invoked with enabled interrupts.
I can't tell if this makes a difference. From looking at it, it might
work to move the last unlock at the end of the function as I didn't find
anything that would acquire the lock again.

Reported-by: Clark Williams <williams@redhat.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
---
 .../drm/i915/gt/intel_execlists_submission.c    | 17 +++++------------
 1 file changed, 5 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_execlists_submission.c b/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
index 3df683b0402ad..948975e72d295 100644
--- a/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
+++ b/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
@@ -1300,7 +1300,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	 * and context switches) submission.
 	 */
 
-	spin_lock(&sched_engine->lock);
+	spin_lock_irq(&sched_engine->lock);
 
 	/*
 	 * If the queue is higher priority than the last
@@ -1400,7 +1400,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				 * Even if ELSP[1] is occupied and not worthy
 				 * of timeslices, our queue might be.
 				 */
-				spin_unlock(&sched_engine->lock);
+				spin_unlock_irq(&sched_engine->lock);
 				return;
 			}
 		}
@@ -1426,7 +1426,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 
 		if (last && !can_merge_rq(last, rq)) {
 			spin_unlock(&ve->base.sched_engine->lock);
-			spin_unlock(&engine->sched_engine->lock);
+			spin_unlock_irq(&engine->sched_engine->lock);
 			return; /* leave this for another sibling */
 		}
 
@@ -1588,7 +1588,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	 */
 	sched_engine->queue_priority_hint = queue_prio(sched_engine);
 	i915_sched_engine_reset_on_empty(sched_engine);
-	spin_unlock(&sched_engine->lock);
+	spin_unlock_irq(&sched_engine->lock);
 
 	/*
 	 * We can skip poking the HW if we ended up with exactly the same set
@@ -1614,13 +1614,6 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	}
 }
 
-static void execlists_dequeue_irq(struct intel_engine_cs *engine)
-{
-	local_irq_disable(); /* Suspend interrupts across request submission */
-	execlists_dequeue(engine);
-	local_irq_enable(); /* flush irq_work (e.g. breadcrumb enabling) */
-}
-
 static void clear_ports(struct i915_request **ports, int count)
 {
 	memset_p((void **)ports, NULL, count);
@@ -2475,7 +2468,7 @@ static void execlists_submission_tasklet(struct tasklet_struct *t)
 	}
 
 	if (!engine->execlists.pending[0]) {
-		execlists_dequeue_irq(engine);
+		execlists_dequeue(engine);
 		start_timeslice(engine);
 	}
 
-- 
2.51.0


  parent reply	other threads:[~2026-01-21 13:53 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-21 13:52 [i915-rt v5 00/21] drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe Maarten Lankhorst
2026-01-21 13:52 ` [i915-rt v5 01/21] drm/i915/display: Fix intel_lpe_audio_irq_handler for PREEMPT-RT Maarten Lankhorst
2026-02-02 19:31   ` Shankar, Uma
2026-01-21 13:52 ` [i915-rt v5 02/21] drm/i915/display: Make get_vblank_counter use intel_de_read_fw() Maarten Lankhorst
2026-02-02 19:33   ` Shankar, Uma
2026-01-21 13:52 ` [i915-rt v5 03/21] drm/i915/display: Use intel_de_write_fw in intel_pipe_fastset Maarten Lankhorst
2026-02-02 19:43   ` Shankar, Uma
2026-01-21 13:53 ` [i915-rt v5 04/21] drm/i915/display: Make set_pipeconf use the fw variants Maarten Lankhorst
2026-02-02 19:45   ` Shankar, Uma
2026-01-21 13:53 ` [i915-rt v5 05/21] drm/i915/display: Move vblank put until after critical section Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 06/21] drm/i915/display: Remove locking from intel_vblank_evade " Maarten Lankhorst
2026-02-02 20:17   ` Shankar, Uma
2026-01-21 13:53 ` [i915-rt v5 07/21] drm/i915/display: Handle vlv dsi workaround in scanline_in_safe_range too Maarten Lankhorst
2026-02-02 20:24   ` Shankar, Uma
2026-01-21 13:53 ` [i915-rt v5 08/21] drm/i915/display: Make icl_dsi_frame_update use _fw too Maarten Lankhorst
2026-02-02 20:31   ` Shankar, Uma
2026-01-21 13:53 ` [i915-rt v5 09/21] drm/i915/display: Enable interrupts earlier on PREEMPT_RT Maarten Lankhorst
2026-02-02 20:59   ` Shankar, Uma
2026-01-21 13:53 ` [i915-rt v5 10/21] drm/i915: Use preempt_disable/enable_rt() where recommended Maarten Lankhorst
2026-02-02 21:07   ` Shankar, Uma
2026-01-21 13:53 ` [i915-rt v5 11/21] drm/i915/display: Use intel_de_read/write_fw in colorops Maarten Lankhorst
2026-02-02 21:08   ` Shankar, Uma
2026-01-21 13:53 ` Maarten Lankhorst [this message]
2026-01-21 13:53 ` [i915-rt v5 13/21] drm/i915: Drop the irqs_disabled() check Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 14/21] drm/i915/guc: Consider also RCU depth in busy loop Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 15/21] Revert "drm/i915: Depend on !PREEMPT_RT." Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 16/21] drm/i915/gt: Fix selftests on PREEMPT_RT Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 17/21] drm/i915/gt: Set stop_timeout() correctly on PREEMPT-RT Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 18/21] drm/i915/display: Do not take uncore lock in i915_get_vblank_counter Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 19/21] drm/i915/display: Remove uncore lock from vlv_atomic_update_fifo Maarten Lankhorst
2026-01-21 13:53 ` [i915-rt v5 20/21] PREEMPT_RT injection Maarten Lankhorst
2026-01-21 23:57   ` kernel test robot
2026-01-22  2:30   ` kernel test robot
2026-01-21 13:53 ` [i915-rt v5 21/21] FOR-CI: drm/i915: Reset lockdep after selftest unload Maarten Lankhorst
2026-01-21 16:56   ` [i915-rt v5] " Maarten Lankhorst
2026-01-22  0:04     ` [i915-rt v5] FOR-CI: try bumping MAX_STACK_TRACE_ENTRIES? Maarten Lankhorst
2026-01-21 14:03 ` ✗ CI.checkpatch: warning for drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe. (rev9) Patchwork
2026-01-21 14:04 ` ✓ CI.KUnit: success " Patchwork
2026-01-21 17:19 ` ✗ CI.checkpatch: warning for drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe. (rev10) Patchwork
2026-01-21 17:21 ` ✓ CI.KUnit: success " Patchwork
2026-01-21 17:36 ` ✗ CI.checksparse: warning " Patchwork
2026-01-21 17:59 ` ✓ Xe.CI.BAT: success " Patchwork
2026-01-22  1:48 ` ✗ CI.checkpatch: warning for drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe. (rev11) Patchwork
2026-01-22  1:49 ` ✓ CI.KUnit: success " Patchwork
2026-01-22  2:04 ` ✗ CI.checksparse: warning " Patchwork
2026-01-22  2:27 ` ✓ Xe.CI.BAT: success " Patchwork
2026-01-22  3:03 ` ✗ Xe.CI.Full: failure for drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe. (rev10) Patchwork
2026-01-22 13:27 ` ✗ Xe.CI.Full: failure for drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe. (rev11) Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260121135318.651622-13-dev@lankhorst.se \
    --to=dev@lankhorst.se \
    --cc=bigeasy@linutronix.de \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=maarten.lankhorst@linux.intel.com \
    --cc=williams@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox