Skip to contentSkip to conversation

SDM660/MSM8953 Fix video core clock status stuck

PATCH Applied Barnabás Czémán 2 participants 4 messages Mar 15, 2025 – Mar 17, 2025

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover SDM660/MSM8953 Fix video core clock status stuck

On SDM660 and MSM8953 video_subcore0 and venus0_core0 clocks are stuck at off. This patch series fixes them.

---
Barnabás Czémán (1):
clk: qcom: mmcc-sdm660: fix stuck video_subcore0 clock

Vladimir Lypak (1):
clk: qcom: gcc-msm8953: fix stuck venus0_core0 clock

 drivers/clk/qcom/gcc-msm8953.c | 2 +-
 drivers/clk/qcom/mmcc-sdm660.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

---
base-commit: da920b7df701770e006928053672147075587fb2
change-id: 20250315-clock-fix-a5f72db5a416

Best regards,

Signature · 1 line
Barnabás Czémán <[email protected]>
#1 lore raw

PATCH 1/2 clk: qcom: mmcc-sdm660: fix stuck video_subcore0 clock

This clock can't be enable with VENUS_CORE0 GDSC turned off. But that GDSC is under HW control so it can be turned off at any moment. Instead of checking the dependent clock we can just vote for it to enable later when GDSC gets turned on.

Fixes 5db3ae8b33de6 ("clk: qcom: Add SDM660 Multimedia Clock Controller (MMCC) driver")
---
 drivers/clk/qcom/mmcc-sdm660.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/qcom/mmcc-sdm660.c +1 -1
index e5bdcc75a36ed24a8c72d1fd6ee57083a7f5c499..e69fc65b13da2b519af413137353693473ee258a 100644
@@ -2540,7 +2540,7 @@ static struct clk_branch video_core_clk = {
 
 static struct clk_branch video_subcore0_clk = {
 	.halt_reg = 0x1048,
-	.halt_check = BRANCH_HALT,
+	.halt_check = BRANCH_HALT_SKIP,
 	.clkr = {
 		.enable_reg = 0x1048,
 		.enable_mask = BIT(0),
Signature · 1 line
2.48.1
#2 lore raw

PATCH 2/2 clk: qcom: gcc-msm8953: fix stuck venus0_core0 clock

In reply to Barnabás Czémán

From: Vladimir Lypak <[email protected]>

This clock can't be enable with VENUS_CORE0 GDSC turned off. But that GDSC is under HW control so it can be turned off at any moment. Instead of checking the dependent clock we can just vote for it to enable later when GDSC gets turned on.

Fixes 9bb6cfc3c77e6 ("clk: qcom: Add Global Clock Controller driver for MSM8953")
---
 drivers/clk/qcom/gcc-msm8953.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/qcom/gcc-msm8953.c +1 -1
index 855a61966f3ef565c48cbc8a7a7fac5e5e66445f..8f29ecc74c50bf50e34e4870b7960b8a0183644b 100644
@@ -3770,7 +3770,7 @@ static struct clk_branch gcc_venus0_axi_clk = {
 
 static struct clk_branch gcc_venus0_core0_vcodec0_clk = {
 	.halt_reg = 0x4c02c,
-	.halt_check = BRANCH_HALT,
+	.halt_check = BRANCH_HALT_SKIP,
 	.clkr = {
 		.enable_reg = 0x4c02c,
 		.enable_mask = BIT(0),
Signature · 1 line
2.48.1
maintainer Applied
#3 lore raw
In reply to Barnabás Czémán
Quoted 4 lines · Barnabás Czémán

On SDM660 and MSM8953 video_subcore0 and venus0_core0 clocks are stuck at off. This patch series fixes them.

Applied, thanks!

[1/2] clk: qcom: mmcc-sdm660: fix stuck video_subcore0 clock

      commit: 000cbe3896c56bf5c625e286ff096533a6b27657
[2/2] clk: qcom: gcc-msm8953: fix stuck venus0_core0 clock
      commit: cdc59600bccf2cb4c483645438a97d4ec55f326b

Best regards,

Signature · 1 line
Bjorn Andersson <[email protected]>