LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk
@ 2020-02-11 12:13 Taniya Das
  2020-02-11 12:13 ` [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk Taniya Das
                   ` (2 more replies)
  0 siblings, 3 replies; 10+ messages in thread
From: Taniya Das @ 2020-02-11 12:13 UTC (permalink / raw)
  To: Stephen Boyd, Michael Turquette  , robh
  Cc: David Brown, Rajendra Nayak, linux-arm-msm, linux-soc, linux-clk,
	linux-kernel, Andy Gross, devicetree, robh+dt, Doug Anderson,
	Taniya Das

The clock disable signal for video_cc_vcodec0_core_clk is tied to
vcodec0_gdsc which is supported in the HW control mode. Thus turning off
the clock would be taken care automatically when the GDSC turns OFF by
hardware and clock driver does not require to poll on the CLK_OFF bit.

Signed-off-by: Taniya Das <tdas@codeaurora.org>
---
 drivers/clk/qcom/videocc-sc7180.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/clk/qcom/videocc-sc7180.c b/drivers/clk/qcom/videocc-sc7180.c
index c363c3c..276e5ec 100644
--- a/drivers/clk/qcom/videocc-sc7180.c
+++ b/drivers/clk/qcom/videocc-sc7180.c
@@ -97,7 +97,7 @@ static struct clk_branch video_cc_vcodec0_axi_clk = {

 static struct clk_branch video_cc_vcodec0_core_clk = {
 	.halt_reg = 0x890,
-	.halt_check = BRANCH_HALT,
+	.halt_check = BRANCH_HALT_VOTED,
 	.clkr = {
 		.enable_reg = 0x890,
 		.enable_mask = BIT(0),
--
Qualcomm INDIA, on behalf of Qualcomm Innovation Center, Inc.is a member
of the Code Aurora Forum, hosted by the  Linux Foundation.

^ permalink raw reply	[flat|nested] 10+ messages in thread

* [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk
  2020-02-11 12:13 [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk Taniya Das
@ 2020-02-11 12:13 ` Taniya Das
  2020-02-12 23:03   ` Stephen Boyd
  2020-03-03 20:16   ` Matthias Kaehlcke
  2020-02-12  0:33 ` [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk Stephen Boyd
  2020-02-12 23:03 ` Stephen Boyd
  2 siblings, 2 replies; 10+ messages in thread
From: Taniya Das @ 2020-02-11 12:13 UTC (permalink / raw)
  To: Stephen Boyd, Michael Turquette  , robh
  Cc: David Brown, Rajendra Nayak, linux-arm-msm, linux-soc, linux-clk,
	linux-kernel, Andy Gross, devicetree, robh+dt, Doug Anderson,
	Taniya Das

The disp_cc_mdss_rscc_ahb_clk is default enabled from hardware and thus
does not require to be marked CRITICAL. This which would allow the RCG to
be turned OFF when the display turns OFF and not blocking XO.

Signed-off-by: Taniya Das <tdas@codeaurora.org>
---
 drivers/clk/qcom/dispcc-sc7180.c | 19 -------------------
 1 file changed, 19 deletions(-)

diff --git a/drivers/clk/qcom/dispcc-sc7180.c b/drivers/clk/qcom/dispcc-sc7180.c
index dd7af41..0a5d395 100644
--- a/drivers/clk/qcom/dispcc-sc7180.c
+++ b/drivers/clk/qcom/dispcc-sc7180.c
@@ -592,24 +592,6 @@ static struct clk_branch disp_cc_mdss_rot_clk = {
 	},
 };

-static struct clk_branch disp_cc_mdss_rscc_ahb_clk = {
-	.halt_reg = 0x400c,
-	.halt_check = BRANCH_HALT,
-	.clkr = {
-		.enable_reg = 0x400c,
-		.enable_mask = BIT(0),
-		.hw.init = &(struct clk_init_data){
-			.name = "disp_cc_mdss_rscc_ahb_clk",
-			.parent_data = &(const struct clk_parent_data){
-				.hw = &disp_cc_mdss_ahb_clk_src.clkr.hw,
-			},
-			.num_parents = 1,
-			.flags = CLK_IS_CRITICAL | CLK_SET_RATE_PARENT,
-			.ops = &clk_branch2_ops,
-		},
-	},
-};
-
 static struct clk_branch disp_cc_mdss_rscc_vsync_clk = {
 	.halt_reg = 0x4008,
 	.halt_check = BRANCH_HALT,
@@ -687,7 +669,6 @@ static struct clk_regmap *disp_cc_sc7180_clocks[] = {
 	[DISP_CC_MDSS_PCLK0_CLK_SRC] = &disp_cc_mdss_pclk0_clk_src.clkr,
 	[DISP_CC_MDSS_ROT_CLK] = &disp_cc_mdss_rot_clk.clkr,
 	[DISP_CC_MDSS_ROT_CLK_SRC] = &disp_cc_mdss_rot_clk_src.clkr,
-	[DISP_CC_MDSS_RSCC_AHB_CLK] = &disp_cc_mdss_rscc_ahb_clk.clkr,
 	[DISP_CC_MDSS_RSCC_VSYNC_CLK] = &disp_cc_mdss_rscc_vsync_clk.clkr,
 	[DISP_CC_MDSS_VSYNC_CLK] = &disp_cc_mdss_vsync_clk.clkr,
 	[DISP_CC_MDSS_VSYNC_CLK_SRC] = &disp_cc_mdss_vsync_clk_src.clkr,
--
Qualcomm INDIA, on behalf of Qualcomm Innovation Center, Inc.is a member
of the Code Aurora Forum, hosted by the  Linux Foundation.

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk
  2020-02-11 12:13 [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk Taniya Das
  2020-02-11 12:13 ` [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk Taniya Das
@ 2020-02-12  0:33 ` Stephen Boyd
  2020-02-12 23:03 ` Stephen Boyd
  2 siblings, 0 replies; 10+ messages in thread
From: Stephen Boyd @ 2020-02-12  0:33 UTC (permalink / raw)
  To: Michael Turquette, Taniya Das, robh
  Cc: David Brown, Rajendra Nayak, linux-arm-msm, linux-soc, linux-clk,
	linux-kernel, Andy Gross, devicetree, robh+dt, Doug Anderson,
	Taniya Das

Quoting Taniya Das (2020-02-11 04:13:55)
> The clock disable signal for video_cc_vcodec0_core_clk is tied to
> vcodec0_gdsc which is supported in the HW control mode. Thus turning off
> the clock would be taken care automatically when the GDSC turns OFF by
> hardware and clock driver does not require to poll on the CLK_OFF bit.
> 
> Signed-off-by: Taniya Das <tdas@codeaurora.org>
> ---
>  drivers/clk/qcom/videocc-sc7180.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/clk/qcom/videocc-sc7180.c b/drivers/clk/qcom/videocc-sc7180.c
> index c363c3c..276e5ec 100644
> --- a/drivers/clk/qcom/videocc-sc7180.c
> +++ b/drivers/clk/qcom/videocc-sc7180.c
> @@ -97,7 +97,7 @@ static struct clk_branch video_cc_vcodec0_axi_clk = {
> 
>  static struct clk_branch video_cc_vcodec0_core_clk = {
>         .halt_reg = 0x890,
> -       .halt_check = BRANCH_HALT,
> +       .halt_check = BRANCH_HALT_VOTED,

Ok. I looked closely and now I notice that some code is using
BRANCH_VOTED and other code is using BRANCH_HALT_VOTED. In the end, it's
the same value. I guess I should remove BRANCH_VOTED from the header
file and make it BIT(7) that's ored in there so that everyone
consistently uses BRANCH_HALT_VOTED.

>         .clkr = {
>                 .enable_reg = 0x890,

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk
  2020-02-11 12:13 [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk Taniya Das
  2020-02-11 12:13 ` [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk Taniya Das
  2020-02-12  0:33 ` [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk Stephen Boyd
@ 2020-02-12 23:03 ` Stephen Boyd
  2 siblings, 0 replies; 10+ messages in thread
From: Stephen Boyd @ 2020-02-12 23:03 UTC (permalink / raw)
  To: Michael Turquette, Taniya Das, robh
  Cc: David Brown, Rajendra Nayak, linux-arm-msm, linux-soc, linux-clk,
	linux-kernel, Andy Gross, devicetree, robh+dt, Doug Anderson,
	Taniya Das

Quoting Taniya Das (2020-02-11 04:13:55)
> The clock disable signal for video_cc_vcodec0_core_clk is tied to
> vcodec0_gdsc which is supported in the HW control mode. Thus turning off
> the clock would be taken care automatically when the GDSC turns OFF by
> hardware and clock driver does not require to poll on the CLK_OFF bit.
> 
> Signed-off-by: Taniya Das <tdas@codeaurora.org>
> ---

Applied to clk-fixes

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk
  2020-02-11 12:13 ` [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk Taniya Das
@ 2020-02-12 23:03   ` Stephen Boyd
  2020-03-03 20:16   ` Matthias Kaehlcke
  1 sibling, 0 replies; 10+ messages in thread
From: Stephen Boyd @ 2020-02-12 23:03 UTC (permalink / raw)
  To: Michael Turquette, Taniya Das, robh
  Cc: David Brown, Rajendra Nayak, linux-arm-msm, linux-soc, linux-clk,
	linux-kernel, Andy Gross, devicetree, robh+dt, Doug Anderson,
	Taniya Das

Quoting Taniya Das (2020-02-11 04:13:56)
> The disp_cc_mdss_rscc_ahb_clk is default enabled from hardware and thus
> does not require to be marked CRITICAL. This which would allow the RCG to
> be turned OFF when the display turns OFF and not blocking XO.
> 
> Signed-off-by: Taniya Das <tdas@codeaurora.org>
> ---

Applied to clk-fixes

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk
  2020-02-11 12:13 ` [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk Taniya Das
  2020-02-12 23:03   ` Stephen Boyd
@ 2020-03-03 20:16   ` Matthias Kaehlcke
  2020-03-04  3:35     ` Taniya Das
  1 sibling, 1 reply; 10+ messages in thread
From: Matthias Kaehlcke @ 2020-03-03 20:16 UTC (permalink / raw)
  To: Taniya Das
  Cc: Stephen Boyd, Michael Turquette  ,
	robh, David Brown, Rajendra Nayak, linux-arm-msm, linux-soc,
	linux-clk, linux-kernel, Andy Gross, devicetree, robh+dt,
	Doug Anderson

On Tue, Feb 11, 2020 at 05:43:56PM +0530, Taniya Das wrote:
> The disp_cc_mdss_rscc_ahb_clk is default enabled from hardware and thus
> does not require to be marked CRITICAL. This which would allow the RCG to
> be turned OFF when the display turns OFF and not blocking XO.
> 
> Signed-off-by: Taniya Das <tdas@codeaurora.org>
> ---
>  drivers/clk/qcom/dispcc-sc7180.c | 19 -------------------
>  1 file changed, 19 deletions(-)
> 
> diff --git a/drivers/clk/qcom/dispcc-sc7180.c b/drivers/clk/qcom/dispcc-sc7180.c
> index dd7af41..0a5d395 100644
> --- a/drivers/clk/qcom/dispcc-sc7180.c
> +++ b/drivers/clk/qcom/dispcc-sc7180.c
> @@ -592,24 +592,6 @@ static struct clk_branch disp_cc_mdss_rot_clk = {
>  	},
>  };
> 
> -static struct clk_branch disp_cc_mdss_rscc_ahb_clk = {
> -	.halt_reg = 0x400c,
> -	.halt_check = BRANCH_HALT,
> -	.clkr = {
> -		.enable_reg = 0x400c,
> -		.enable_mask = BIT(0),
> -		.hw.init = &(struct clk_init_data){
> -			.name = "disp_cc_mdss_rscc_ahb_clk",
> -			.parent_data = &(const struct clk_parent_data){
> -				.hw = &disp_cc_mdss_ahb_clk_src.clkr.hw,
> -			},
> -			.num_parents = 1,
> -			.flags = CLK_IS_CRITICAL | CLK_SET_RATE_PARENT,
> -			.ops = &clk_branch2_ops,
> -		},
> -	},
> -};
> -
>  static struct clk_branch disp_cc_mdss_rscc_vsync_clk = {
>  	.halt_reg = 0x4008,
>  	.halt_check = BRANCH_HALT,
> @@ -687,7 +669,6 @@ static struct clk_regmap *disp_cc_sc7180_clocks[] = {
>  	[DISP_CC_MDSS_PCLK0_CLK_SRC] = &disp_cc_mdss_pclk0_clk_src.clkr,
>  	[DISP_CC_MDSS_ROT_CLK] = &disp_cc_mdss_rot_clk.clkr,
>  	[DISP_CC_MDSS_ROT_CLK_SRC] = &disp_cc_mdss_rot_clk_src.clkr,
> -	[DISP_CC_MDSS_RSCC_AHB_CLK] = &disp_cc_mdss_rscc_ahb_clk.clkr,
>  	[DISP_CC_MDSS_RSCC_VSYNC_CLK] = &disp_cc_mdss_rscc_vsync_clk.clkr,
>  	[DISP_CC_MDSS_VSYNC_CLK] = &disp_cc_mdss_vsync_clk.clkr,
>  	[DISP_CC_MDSS_VSYNC_CLK_SRC] = &disp_cc_mdss_vsync_clk_src.clkr,

We found that this change leads to a panic at boot time on SC7180 devices
without display configuration (e.g. the SC7180 IDP with the current DT):

[    2.412820] SError Interrupt on CPU6, code 0xbe000411 -- SError
[    2.412822] CPU: 6 PID: 1 Comm: swapper/0 Tainted: G S                5.4.22 #103
[    2.412822] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT)
[    2.412823] pstate: 20c00089 (nzCv daIf +PAN +UAO)
[    2.412823] pc : regmap_mmio_read32le+0x28/0x40
[    2.412823] lr : regmap_mmio_read+0x44/0x6c
[    2.412824] sp : ffffffc01005ba90
[    2.412824] x29: ffffffc01005ba90 x28: 0000000000000000
[    2.412825] x27: 0000000000000000 x26: 0000000000000000
[    2.412826] x25: 0000000000000000 x24: ffffffd1f4aed018
[    2.412827] x23: ffffffd1f4c12148 x22: ffffff8177a6c800
[    2.412827] x21: 0000000000002048 x20: ffffff8177489e00
[    2.412828] x19: 0000000000002048 x18: 000000004a746f4b
[    2.412829] x17: 00000000d0e09034 x16: 000000005079b450
[    2.412830] x15: 000000003e3bf7ed x14: 0000000000007fff
[    2.412830] x13: ffffff8177309b40 x12: 0000000000000000
[    2.412831] x11: 0000000000000000 x10: 0000000000000000
[    2.412831] x9 : 0000000000000001 x8 : ffffffc011c02048
[    2.412832] x7 : aaaaaaaaaaaaaaaa x6 : 0000000000000000
[    2.412833] x5 : 0000000000000000 x4 : 0000000000000000
[    2.412834] x3 : 0000000000000000 x2 : ffffffc01005bb84
[    2.412834] x1 : 0000000000002048 x0 : 0000000080000000
[    2.412835] Kernel panic - not syncing: Asynchronous SError Interrupt
[    2.412836] CPU: 6 PID: 1 Comm: swapper/0 Tainted: G S                5.4.22 #103
[    2.412836] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT)
[    2.412836] Call trace:
[    2.412837]  dump_backtrace+0x0/0x150
[    2.412837]  show_stack+0x20/0x2c
[    2.412837]  dump_stack+0xa0/0xd8
[    2.412838]  panic+0x158/0x360
[    2.412838]  panic+0x0/0x360
[    2.412838]  arm64_serror_panic+0x78/0x84
[    2.412839]  do_serror+0x110/0x118
[    2.412839]  el1_error+0x84/0xf8
[    2.412839]  regmap_mmio_read32le+0x28/0x40
[    2.412840]  regmap_mmio_read+0x44/0x6c
[    2.412840]  _regmap_bus_reg_read+0x34/0x44
[    2.412841]  _regmap_read+0x88/0x164
[    2.412841]  regmap_read+0x54/0x78
[    2.412841]  clk_is_enabled_regmap+0x3c/0x8c
[    2.412842]  clk_core_is_enabled+0x68/0xac
[    2.412842]  clk_disable_unused_subtree+0x90/0x22c
[    2.412843]  clk_disable_unused_subtree+0x34/0x22c
[    2.412843]  clk_disable_unused+0x74/0x108
[    2.412843]  do_one_initcall+0x13c/0x2c8
[    2.412844]  do_initcall_level+0x144/0x16c
[    2.412844]  do_basic_setup+0x30/0x48
[    2.412844]  kernel_init_freeable+0xc4/0x140
[    2.412845]  kernel_init+0x14/0x100
[    2.412845]  ret_from_fork+0x10/0x18
[    2.412858] SMP: stopping secondary CPUs
[    2.412859] Kernel Offset: 0x11e3a00000 from 0xffffffc010000000
[    2.412859] PHYS_OFFSET: 0xffffffe780000000
[    2.412860] CPU features: 0x0006,2a80aa18
[    2.412860] Memory Limit: none

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk
  2020-03-03 20:16   ` Matthias Kaehlcke
@ 2020-03-04  3:35     ` Taniya Das
  2020-03-04 17:09       ` Matthias Kaehlcke
  0 siblings, 1 reply; 10+ messages in thread
From: Taniya Das @ 2020-03-04  3:35 UTC (permalink / raw)
  To: Matthias Kaehlcke
  Cc: Stephen Boyd, Michael Turquette, robh, David Brown,
	Rajendra Nayak, linux-arm-msm, linux-soc, linux-clk,
	linux-kernel, Andy Gross, devicetree, robh+dt, Doug Anderson



On 3/4/2020 1:46 AM, Matthias Kaehlcke wrote:
> On Tue, Feb 11, 2020 at 05:43:56PM +0530, Taniya Das wrote:
>> The disp_cc_mdss_rscc_ahb_clk is default enabled from hardware and thus
>> does not require to be marked CRITICAL. This which would allow the RCG to
>> be turned OFF when the display turns OFF and not blocking XO.
>>
>> Signed-off-by: Taniya Das <tdas@codeaurora.org>
>> ---
>>   drivers/clk/qcom/dispcc-sc7180.c | 19 -------------------
>>   1 file changed, 19 deletions(-)
>>
>> diff --git a/drivers/clk/qcom/dispcc-sc7180.c b/drivers/clk/qcom/dispcc-sc7180.c
>> index dd7af41..0a5d395 100644
>> --- a/drivers/clk/qcom/dispcc-sc7180.c
>> +++ b/drivers/clk/qcom/dispcc-sc7180.c
>> @@ -592,24 +592,6 @@ static struct clk_branch disp_cc_mdss_rot_clk = {
>>   	},
>>   };
>>
>> -static struct clk_branch disp_cc_mdss_rscc_ahb_clk = {
>> -	.halt_reg = 0x400c,
>> -	.halt_check = BRANCH_HALT,
>> -	.clkr = {
>> -		.enable_reg = 0x400c,
>> -		.enable_mask = BIT(0),
>> -		.hw.init = &(struct clk_init_data){
>> -			.name = "disp_cc_mdss_rscc_ahb_clk",
>> -			.parent_data = &(const struct clk_parent_data){
>> -				.hw = &disp_cc_mdss_ahb_clk_src.clkr.hw,
>> -			},
>> -			.num_parents = 1,
>> -			.flags = CLK_IS_CRITICAL | CLK_SET_RATE_PARENT,
>> -			.ops = &clk_branch2_ops,
>> -		},
>> -	},
>> -};
>> -
>>   static struct clk_branch disp_cc_mdss_rscc_vsync_clk = {
>>   	.halt_reg = 0x4008,
>>   	.halt_check = BRANCH_HALT,
>> @@ -687,7 +669,6 @@ static struct clk_regmap *disp_cc_sc7180_clocks[] = {
>>   	[DISP_CC_MDSS_PCLK0_CLK_SRC] = &disp_cc_mdss_pclk0_clk_src.clkr,
>>   	[DISP_CC_MDSS_ROT_CLK] = &disp_cc_mdss_rot_clk.clkr,
>>   	[DISP_CC_MDSS_ROT_CLK_SRC] = &disp_cc_mdss_rot_clk_src.clkr,
>> -	[DISP_CC_MDSS_RSCC_AHB_CLK] = &disp_cc_mdss_rscc_ahb_clk.clkr,
>>   	[DISP_CC_MDSS_RSCC_VSYNC_CLK] = &disp_cc_mdss_rscc_vsync_clk.clkr,
>>   	[DISP_CC_MDSS_VSYNC_CLK] = &disp_cc_mdss_vsync_clk.clkr,
>>   	[DISP_CC_MDSS_VSYNC_CLK_SRC] = &disp_cc_mdss_vsync_clk_src.clkr,
> 
> We found that this change leads to a panic at boot time on SC7180 devices
> without display configuration (e.g. the SC7180 IDP with the current DT):
> 
> [    2.412820] SError Interrupt on CPU6, code 0xbe000411 -- SError
> [    2.412822] CPU: 6 PID: 1 Comm: swapper/0 Tainted: G S                5.4.22 #103
> [    2.412822] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT)
> [    2.412823] pstate: 20c00089 (nzCv daIf +PAN +UAO)
> [    2.412823] pc : regmap_mmio_read32le+0x28/0x40
> [    2.412823] lr : regmap_mmio_read+0x44/0x6c
> [    2.412824] sp : ffffffc01005ba90
> [    2.412824] x29: ffffffc01005ba90 x28: 0000000000000000
> [    2.412825] x27: 0000000000000000 x26: 0000000000000000
> [    2.412826] x25: 0000000000000000 x24: ffffffd1f4aed018
> [    2.412827] x23: ffffffd1f4c12148 x22: ffffff8177a6c800
> [    2.412827] x21: 0000000000002048 x20: ffffff8177489e00
> [    2.412828] x19: 0000000000002048 x18: 000000004a746f4b
> [    2.412829] x17: 00000000d0e09034 x16: 000000005079b450
> [    2.412830] x15: 000000003e3bf7ed x14: 0000000000007fff
> [    2.412830] x13: ffffff8177309b40 x12: 0000000000000000
> [    2.412831] x11: 0000000000000000 x10: 0000000000000000
> [    2.412831] x9 : 0000000000000001 x8 : ffffffc011c02048
> [    2.412832] x7 : aaaaaaaaaaaaaaaa x6 : 0000000000000000
> [    2.412833] x5 : 0000000000000000 x4 : 0000000000000000
> [    2.412834] x3 : 0000000000000000 x2 : ffffffc01005bb84
> [    2.412834] x1 : 0000000000002048 x0 : 0000000080000000
> [    2.412835] Kernel panic - not syncing: Asynchronous SError Interrupt
> [    2.412836] CPU: 6 PID: 1 Comm: swapper/0 Tainted: G S                5.4.22 #103
> [    2.412836] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT)
> [    2.412836] Call trace:
> [    2.412837]  dump_backtrace+0x0/0x150
> [    2.412837]  show_stack+0x20/0x2c
> [    2.412837]  dump_stack+0xa0/0xd8
> [    2.412838]  panic+0x158/0x360
> [    2.412838]  panic+0x0/0x360
> [    2.412838]  arm64_serror_panic+0x78/0x84
> [    2.412839]  do_serror+0x110/0x118
> [    2.412839]  el1_error+0x84/0xf8
> [    2.412839]  regmap_mmio_read32le+0x28/0x40
> [    2.412840]  regmap_mmio_read+0x44/0x6c
> [    2.412840]  _regmap_bus_reg_read+0x34/0x44
> [    2.412841]  _regmap_read+0x88/0x164
> [    2.412841]  regmap_read+0x54/0x78
> [    2.412841]  clk_is_enabled_regmap+0x3c/0x8c
> [    2.412842]  clk_core_is_enabled+0x68/0xac
> [    2.412842]  clk_disable_unused_subtree+0x90/0x22c
> [    2.412843]  clk_disable_unused_subtree+0x34/0x22c
> [    2.412843]  clk_disable_unused+0x74/0x108
> [    2.412843]  do_one_initcall+0x13c/0x2c8
> [    2.412844]  do_initcall_level+0x144/0x16c
> [    2.412844]  do_basic_setup+0x30/0x48
> [    2.412844]  kernel_init_freeable+0xc4/0x140
> [    2.412845]  kernel_init+0x14/0x100
> [    2.412845]  ret_from_fork+0x10/0x18
> [    2.412858] SMP: stopping secondary CPUs
> [    2.412859] Kernel Offset: 0x11e3a00000 from 0xffffffc010000000
> [    2.412859] PHYS_OFFSET: 0xffffffe780000000
> [    2.412860] CPU features: 0x0006,2a80aa18
> [    2.412860] Memory Limit: none
> 

Hi Matthias,

The display device node is not present and we encounter this crash, 
would it be possible to add ALWAYS_ON for the MDSS GDSC and give it a try.


-- 
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member
of Code Aurora Forum, hosted by The Linux Foundation.

--

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk
  2020-03-04  3:35     ` Taniya Das
@ 2020-03-04 17:09       ` Matthias Kaehlcke
  2020-03-13 18:01         ` Stephen Boyd
  0 siblings, 1 reply; 10+ messages in thread
From: Matthias Kaehlcke @ 2020-03-04 17:09 UTC (permalink / raw)
  To: Taniya Das
  Cc: Stephen Boyd, Michael Turquette, robh, David Brown,
	Rajendra Nayak, linux-arm-msm, linux-soc, linux-clk,
	linux-kernel, Andy Gross, devicetree, robh+dt, Doug Anderson

Hi Taniya,

On Wed, Mar 04, 2020 at 09:05:20AM +0530, Taniya Das wrote:
> 
> On 3/4/2020 1:46 AM, Matthias Kaehlcke wrote:
> > On Tue, Feb 11, 2020 at 05:43:56PM +0530, Taniya Das wrote:
> > > The disp_cc_mdss_rscc_ahb_clk is default enabled from hardware and thus
> > > does not require to be marked CRITICAL. This which would allow the RCG to
> > > be turned OFF when the display turns OFF and not blocking XO.
> > > 
> > > Signed-off-by: Taniya Das <tdas@codeaurora.org>
> > > ---
> > >   drivers/clk/qcom/dispcc-sc7180.c | 19 -------------------
> > >   1 file changed, 19 deletions(-)
> > > 
> > > diff --git a/drivers/clk/qcom/dispcc-sc7180.c b/drivers/clk/qcom/dispcc-sc7180.c
> > > index dd7af41..0a5d395 100644
> > > --- a/drivers/clk/qcom/dispcc-sc7180.c
> > > +++ b/drivers/clk/qcom/dispcc-sc7180.c
> > > @@ -592,24 +592,6 @@ static struct clk_branch disp_cc_mdss_rot_clk = {
> > >   	},
> > >   };
> > > 
> > > -static struct clk_branch disp_cc_mdss_rscc_ahb_clk = {
> > > -	.halt_reg = 0x400c,
> > > -	.halt_check = BRANCH_HALT,
> > > -	.clkr = {
> > > -		.enable_reg = 0x400c,
> > > -		.enable_mask = BIT(0),
> > > -		.hw.init = &(struct clk_init_data){
> > > -			.name = "disp_cc_mdss_rscc_ahb_clk",
> > > -			.parent_data = &(const struct clk_parent_data){
> > > -				.hw = &disp_cc_mdss_ahb_clk_src.clkr.hw,
> > > -			},
> > > -			.num_parents = 1,
> > > -			.flags = CLK_IS_CRITICAL | CLK_SET_RATE_PARENT,
> > > -			.ops = &clk_branch2_ops,
> > > -		},
> > > -	},
> > > -};
> > > -
> > >   static struct clk_branch disp_cc_mdss_rscc_vsync_clk = {
> > >   	.halt_reg = 0x4008,
> > >   	.halt_check = BRANCH_HALT,
> > > @@ -687,7 +669,6 @@ static struct clk_regmap *disp_cc_sc7180_clocks[] = {
> > >   	[DISP_CC_MDSS_PCLK0_CLK_SRC] = &disp_cc_mdss_pclk0_clk_src.clkr,
> > >   	[DISP_CC_MDSS_ROT_CLK] = &disp_cc_mdss_rot_clk.clkr,
> > >   	[DISP_CC_MDSS_ROT_CLK_SRC] = &disp_cc_mdss_rot_clk_src.clkr,
> > > -	[DISP_CC_MDSS_RSCC_AHB_CLK] = &disp_cc_mdss_rscc_ahb_clk.clkr,
> > >   	[DISP_CC_MDSS_RSCC_VSYNC_CLK] = &disp_cc_mdss_rscc_vsync_clk.clkr,
> > >   	[DISP_CC_MDSS_VSYNC_CLK] = &disp_cc_mdss_vsync_clk.clkr,
> > >   	[DISP_CC_MDSS_VSYNC_CLK_SRC] = &disp_cc_mdss_vsync_clk_src.clkr,
> > 
> > We found that this change leads to a panic at boot time on SC7180 devices
> > without display configuration (e.g. the SC7180 IDP with the current DT):
> > 
> > [    2.412820] SError Interrupt on CPU6, code 0xbe000411 -- SError
> > [    2.412822] CPU: 6 PID: 1 Comm: swapper/0 Tainted: G S                5.4.22 #103
> > [    2.412822] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT)
> > [    2.412823] pstate: 20c00089 (nzCv daIf +PAN +UAO)
> > [    2.412823] pc : regmap_mmio_read32le+0x28/0x40
> > [    2.412823] lr : regmap_mmio_read+0x44/0x6c
> > [    2.412824] sp : ffffffc01005ba90
> > [    2.412824] x29: ffffffc01005ba90 x28: 0000000000000000
> > [    2.412825] x27: 0000000000000000 x26: 0000000000000000
> > [    2.412826] x25: 0000000000000000 x24: ffffffd1f4aed018
> > [    2.412827] x23: ffffffd1f4c12148 x22: ffffff8177a6c800
> > [    2.412827] x21: 0000000000002048 x20: ffffff8177489e00
> > [    2.412828] x19: 0000000000002048 x18: 000000004a746f4b
> > [    2.412829] x17: 00000000d0e09034 x16: 000000005079b450
> > [    2.412830] x15: 000000003e3bf7ed x14: 0000000000007fff
> > [    2.412830] x13: ffffff8177309b40 x12: 0000000000000000
> > [    2.412831] x11: 0000000000000000 x10: 0000000000000000
> > [    2.412831] x9 : 0000000000000001 x8 : ffffffc011c02048
> > [    2.412832] x7 : aaaaaaaaaaaaaaaa x6 : 0000000000000000
> > [    2.412833] x5 : 0000000000000000 x4 : 0000000000000000
> > [    2.412834] x3 : 0000000000000000 x2 : ffffffc01005bb84
> > [    2.412834] x1 : 0000000000002048 x0 : 0000000080000000
> > [    2.412835] Kernel panic - not syncing: Asynchronous SError Interrupt
> > [    2.412836] CPU: 6 PID: 1 Comm: swapper/0 Tainted: G S                5.4.22 #103
> > [    2.412836] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT)
> > [    2.412836] Call trace:
> > [    2.412837]  dump_backtrace+0x0/0x150
> > [    2.412837]  show_stack+0x20/0x2c
> > [    2.412837]  dump_stack+0xa0/0xd8
> > [    2.412838]  panic+0x158/0x360
> > [    2.412838]  panic+0x0/0x360
> > [    2.412838]  arm64_serror_panic+0x78/0x84
> > [    2.412839]  do_serror+0x110/0x118
> > [    2.412839]  el1_error+0x84/0xf8
> > [    2.412839]  regmap_mmio_read32le+0x28/0x40
> > [    2.412840]  regmap_mmio_read+0x44/0x6c
> > [    2.412840]  _regmap_bus_reg_read+0x34/0x44
> > [    2.412841]  _regmap_read+0x88/0x164
> > [    2.412841]  regmap_read+0x54/0x78
> > [    2.412841]  clk_is_enabled_regmap+0x3c/0x8c
> > [    2.412842]  clk_core_is_enabled+0x68/0xac
> > [    2.412842]  clk_disable_unused_subtree+0x90/0x22c
> > [    2.412843]  clk_disable_unused_subtree+0x34/0x22c
> > [    2.412843]  clk_disable_unused+0x74/0x108
> > [    2.412843]  do_one_initcall+0x13c/0x2c8
> > [    2.412844]  do_initcall_level+0x144/0x16c
> > [    2.412844]  do_basic_setup+0x30/0x48
> > [    2.412844]  kernel_init_freeable+0xc4/0x140
> > [    2.412845]  kernel_init+0x14/0x100
> > [    2.412845]  ret_from_fork+0x10/0x18
> > [    2.412858] SMP: stopping secondary CPUs
> > [    2.412859] Kernel Offset: 0x11e3a00000 from 0xffffffc010000000
> > [    2.412859] PHYS_OFFSET: 0xffffffe780000000
> > [    2.412860] CPU features: 0x0006,2a80aa18
> > [    2.412860] Memory Limit: none
> > 
> 
> Hi Matthias,
> 
> The display device node is not present and we encounter this crash, would it
> be possible to add ALWAYS_ON for the MDSS GDSC and give it a try.

It still crashes when ALWAYS_ON is set for the MDSS GDSC.

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk
  2020-03-04 17:09       ` Matthias Kaehlcke
@ 2020-03-13 18:01         ` Stephen Boyd
       [not found]           ` <c6b19bac-b018-28a0-421f-f40f85245bee@codeaurora.org>
  0 siblings, 1 reply; 10+ messages in thread
From: Stephen Boyd @ 2020-03-13 18:01 UTC (permalink / raw)
  To: Matthias Kaehlcke, Taniya Das
  Cc: Michael Turquette, robh, David Brown, Rajendra Nayak,
	linux-arm-msm, linux-soc, linux-clk, linux-kernel, Andy Gross,
	devicetree, robh+dt, Doug Anderson

Quoting Matthias Kaehlcke (2020-03-04 09:09:39)
> Hi Taniya,
> 
> On Wed, Mar 04, 2020 at 09:05:20AM +0530, Taniya Das wrote:
> > 
> > Hi Matthias,
> > 
> > The display device node is not present and we encounter this crash, would it
> > be possible to add ALWAYS_ON for the MDSS GDSC and give it a try.
> 
> It still crashes when ALWAYS_ON is set for the MDSS GDSC.

Any updates here? I'm about to send this patch off to Linus and I'm
wondering if there will be a resolution besides reverting it.

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk
       [not found]           ` <c6b19bac-b018-28a0-421f-f40f85245bee@codeaurora.org>
@ 2020-03-16 18:33             ` Stephen Boyd
  0 siblings, 0 replies; 10+ messages in thread
From: Stephen Boyd @ 2020-03-16 18:33 UTC (permalink / raw)
  To: Matthias Kaehlcke, Taniya Das
  Cc: Michael Turquette, robh, David Brown, Rajendra Nayak,
	linux-arm-msm, linux-soc, linux-clk, linux-kernel, Andy Gross,
	devicetree, robh+dt, Doug Anderson

Quoting Taniya Das (2020-03-14 04:08:47)
> I would require you suggestion on the below.
> 
> On 3/13/2020 11:31 PM, Stephen Boyd wrote:
> > Quoting Matthias Kaehlcke (2020-03-04 09:09:39)
> >> Hi Taniya,
> >>
> >> On Wed, Mar 04, 2020 at 09:05:20AM +0530, Taniya Das wrote:
> >>>
> >>> Hi Matthias,
> >>>
> >>> The display device node is not present and we encounter this crash, would it
> >>> be possible to add ALWAYS_ON for the MDSS GDSC and give it a try.
> >>
> >> It still crashes when ALWAYS_ON is set for the MDSS GDSC.
> > 
> > Any updates here? I'm about to send this patch off to Linus and I'm
> > wondering if there will be a resolution besides reverting it.
> > 
> 
> Looks like the AHB clock needs to be left enabled till the last clocks 
> get disabled. I need to add a new dependency o n this clock.
> 
> Hi Stephen,
> 
> 
> Any way to keep this dependency using the framework  or I need to split 
> the probes to register them independently?
> 

Sorry, I don't understand the problem fully. This AHB clk is left on in
the bootloader? Why do we need to touch it if it's always left enabled?
Is it actually the case that the bootloader hasn't turned this clk on
and we're getting lucky having it marked here as CLK_IS_CRITICAL? Can we
force it on in driver probe and then ignore it after that?

^ permalink raw reply	[flat|nested] 10+ messages in thread

end of thread, other threads:[~2020-03-16 18:33 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-02-11 12:13 [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk Taniya Das
2020-02-11 12:13 ` [PATCH v1 2/2] clk: qcom: dispcc: Remove support of disp_cc_mdss_rscc_ahb_clk Taniya Das
2020-02-12 23:03   ` Stephen Boyd
2020-03-03 20:16   ` Matthias Kaehlcke
2020-03-04  3:35     ` Taniya Das
2020-03-04 17:09       ` Matthias Kaehlcke
2020-03-13 18:01         ` Stephen Boyd
     [not found]           ` <c6b19bac-b018-28a0-421f-f40f85245bee@codeaurora.org>
2020-03-16 18:33             ` Stephen Boyd
2020-02-12  0:33 ` [PATCH v1 1/2] clk: qcom: videocc: Update the clock flag for video_cc_vcodec0_core_clk Stephen Boyd
2020-02-12 23:03 ` Stephen Boyd

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).