* FAILED: patch "[PATCH] drm/amdgpu: publish IH enabled before PSP programs IH_RB_CNTL" failed to apply to 5.15-stable tree
@ 2026-11-10 1:24 Sasha Levin
2026-11-15 16:41 ` [PATCH 5.15.y] drm/amdgpu: publish IH enabled before PSP programs IH_RB_CNTL Sasha Levin
0 siblings, 1 reply; 2+ messages in thread
From: Sasha Levin @ 2026-11-10 1:24 UTC (permalink / raw)
To: milan.milicevic, danijel.slivka, alexander.deucher; +Cc: stable
The patch below does not apply to the 5.15-stable tree.
If someone wants it applied there, or to any other stable or longterm
tree, then please email the backport, including the original git commit
id to <stable@vger.kernel.org>.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y
git checkout FETCH_HEAD
git cherry-pick -x 9d550dd1a3b4b96da49f3560a84b5c2413a88cd3
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<stable@vger.kernel.org>' --in-reply-to '20261109202411-failed-5.15-9d550dd1a3b4@kernel.org' --subject-prefix 'PATCH 5.15.y' 'HEAD^..'
Thanks,
Sasha
------------------ original commit in Linus's tree ------------------
From 9d550dd1a3b4b96da49f3560a84b5c2413a88cd3 Mon Sep 17 00:00:00 2001
From: Milan Milicevic <Milan.Milicevic@amd.com>
Date: Wed, 28 Oct 2026 15:31:55 +0100
Subject: [PATCH] drm/amdgpu: publish IH enabled before PSP programs IH_RB_CNTL
On SR-IOV, programming IH_RB_CNTL through PSP is slow enough for the
first MSI to arrive before ih->enabled is set. amdgpu_ih_process()
then drops the interrupt, rptr stays at 0, and RPTR_REARM never
generates another edge, so every ring hits the fence fallback timer.
Set ih->enabled before psp_reg_program() on the indirect IH path for
Navi10 and IH v6.0, and clear it if programming fails.
Signed-off-by: Danijel Slivka <Danijel.Slivka@amd.com>
Signed-off-by: Milan Milicevic <Milan.Milicevic@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit b2f9434f9d17c94798e8d8957af3adc4a9949016)
Cc: stable@vger.kernel.org
---
drivers/gpu/drm/amd/amdgpu/ih_v6_0.c | 13 ++++++++++++-
drivers/gpu/drm/amd/amdgpu/navi10_ih.c | 13 ++++++++++++-
2 files changed, 24 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c b/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c
index 8dee486496501..478b7169f44c2 100644
--- a/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/ih_v6_0.c
@@ -168,8 +168,19 @@ static int ih_v6_0_toggle_ring_interrupts(struct amdgpu_device *adev,
tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, ENABLE_INTR, (enable ? 1 : 0));
if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) {
- if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp))
+ /*
+ * PSP programming of IH_RB_CNTL is slow enough for MSI to
+ * arrive before this call returns. Publish ih->enabled first
+ * so amdgpu_ih_process() consumes that IV and rearms.
+ */
+ if (enable) {
+ ih->enabled = true;
+ smp_wmb();
+ }
+ if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp)) {
+ ih->enabled = false;
return -ETIMEDOUT;
+ }
} else {
WREG32(ih_regs->ih_rb_cntl, tmp);
}
diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
index ab12ce8da3dc3..fa5148ae526de 100644
--- a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
@@ -194,8 +194,19 @@ static int navi10_ih_toggle_ring_interrupts(struct amdgpu_device *adev,
tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, ENABLE_INTR, (enable ? 1 : 0));
if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) {
- if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp))
+ /*
+ * PSP programming of IH_RB_CNTL is slow enough for MSI to
+ * arrive before this call returns. Publish ih->enabled first
+ * so amdgpu_ih_process() consumes that IV and rearms.
+ */
+ if (enable) {
+ ih->enabled = true;
+ smp_wmb();
+ }
+ if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp)) {
+ ih->enabled = false;
return -ETIMEDOUT;
+ }
} else {
WREG32(ih_regs->ih_rb_cntl, tmp);
}
--
2.53.0
^ permalink raw reply related [flat|nested] 2+ messages in thread
* [PATCH 5.15.y] drm/amdgpu: publish IH enabled before PSP programs IH_RB_CNTL
2026-11-10 1:24 FAILED: patch "[PATCH] drm/amdgpu: publish IH enabled before PSP programs IH_RB_CNTL" failed to apply to 5.15-stable tree Sasha Levin
@ 2026-11-15 16:41 ` Sasha Levin
0 siblings, 0 replies; 2+ messages in thread
From: Sasha Levin @ 2026-11-15 16:41 UTC (permalink / raw)
To: stable; +Cc: Milan Milicevic, Danijel Slivka, Alex Deucher, Sasha Levin
From: Milan Milicevic <Milan.Milicevic@amd.com>
[ Upstream commit 9d550dd1a3b4b96da49f3560a84b5c2413a88cd3 ]
On SR-IOV, programming IH_RB_CNTL through PSP is slow enough for the
first MSI to arrive before ih->enabled is set. amdgpu_ih_process()
then drops the interrupt, rptr stays at 0, and RPTR_REARM never
generates another edge, so every ring hits the fence fallback timer.
Set ih->enabled before psp_reg_program() on the indirect IH path for
Navi10 and IH v6.0, and clear it if programming fails.
Signed-off-by: Danijel Slivka <Danijel.Slivka@amd.com>
Signed-off-by: Milan Milicevic <Milan.Milicevic@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit b2f9434f9d17c94798e8d8957af3adc4a9949016)
Cc: stable@vger.kernel.org
[ sashal: Omitted changes to the absent IH v6.0 driver. ]
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
drivers/gpu/drm/amd/amdgpu/navi10_ih.c | 13 ++++++++++++-
1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
index 9425a9448cf51..352d6f14aa413 100644
--- a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c
@@ -194,8 +194,19 @@ static int navi10_ih_toggle_ring_interrupts(struct amdgpu_device *adev,
tmp = REG_SET_FIELD(tmp, IH_RB_CNTL, ENABLE_INTR, (enable ? 1 : 0));
if (amdgpu_sriov_vf(adev) && amdgpu_sriov_reg_indirect_ih(adev)) {
- if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp))
+ /*
+ * PSP programming of IH_RB_CNTL is slow enough for MSI to
+ * arrive before this call returns. Publish ih->enabled first
+ * so amdgpu_ih_process() consumes that IV and rearms.
+ */
+ if (enable) {
+ ih->enabled = true;
+ smp_wmb();
+ }
+ if (psp_reg_program(&adev->psp, ih_regs->psp_reg_id, tmp)) {
+ ih->enabled = false;
return -ETIMEDOUT;
+ }
} else {
WREG32(ih_regs->ih_rb_cntl, tmp);
}
--
2.53.0
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-11-15 16:41 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-10 1:24 FAILED: patch "[PATCH] drm/amdgpu: publish IH enabled before PSP programs IH_RB_CNTL" failed to apply to 5.15-stable tree Sasha Levin
2026-11-15 16:41 ` [PATCH 5.15.y] drm/amdgpu: publish IH enabled before PSP programs IH_RB_CNTL Sasha Levin
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox