Linux Media Controller development
 help / color / mirror / Atom feed
From: Vishnu Reddy <busanna.reddy@oss.qualcomm.com>
To: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Cc: Vikash Garodia <vikash.garodia@oss.qualcomm.com>,
	Dikshita Agarwal <dikshita.agarwal@oss.qualcomm.com>,
	Abhinav Kumar <abhinav.kumar@linux.dev>,
	Bryan O'Donoghue <bod@kernel.org>,
	Mauro Carvalho Chehab <mchehab@kernel.org>,
	Joerg Roedel <joro@8bytes.org>, Will Deacon <will@kernel.org>,
	Robin Murphy <robin.murphy@arm.com>,
	Hans Verkuil <hverkuil@kernel.org>,
	Stefan Schmidt <stefan.schmidt@linaro.org>,
	Rob Herring <robh@kernel.org>,
	Krzysztof Kozlowski <krzk+dt@kernel.org>,
	Conor Dooley <conor+dt@kernel.org>,
	Stanimir Varbanov <stanimir.varbanov@linaro.org>,
	Jorge Ramirez-Ortiz <jorge.ramirez@oss.qualcomm.com>,
	Del Regno <angelogioacchino.delregno@collabora.com>,
	Bjorn Andersson <andersson@kernel.org>,
	Konrad Dybcio <konradybcio@kernel.org>,
	linux-kernel@vger.kernel.org, linux-media@vger.kernel.org,
	linux-arm-msm@vger.kernel.org, iommu@lists.linux.dev,
	Krzysztof Kozlowski <krzk@kernel.org>,
	devicetree@vger.kernel.org
Subject: Re: [PATCH v5 10/14] media: iris: Add power sequence for Glymur
Date: Mon, 11 May 2026 21:42:01 +0530	[thread overview]
Message-ID: <67157e35-8c39-b06b-eb93-602930168f4b@oss.qualcomm.com> (raw)
In-Reply-To: <jqe3ilzqzajuf67hvp7nw6mak36qyqu6dz2oiwtkf6zwiyclln@ydtyd7vkudm4>


On 5/9/2026 2:24 AM, Dmitry Baryshkov wrote:
> On Sat, May 09, 2026 at 12:29:59AM +0530, Vishnu Reddy wrote:
>> Glymur has a secondary video codec core (vcodec1), equivalent to the
>> primary core (vcodec0), but with independent power domains, clocks,
>> and reset lines. Reuse the existing code wherever possible and add
>> power sequence for vcodec1.
>>
>> Reviewed-by: Vikash Garodia <vikash.garodia@oss.qualcomm.com>
>> Signed-off-by: Vishnu Reddy <busanna.reddy@oss.qualcomm.com>
>> ---
>>  .../platform/qcom/iris/iris_platform_common.h      |   4 +
>>  drivers/media/platform/qcom/iris/iris_vpu3x.c      | 141 ++++++++++++++++++++-
>>  drivers/media/platform/qcom/iris/iris_vpu_common.h |   1 +
>>  .../platform/qcom/iris/iris_vpu_register_defines.h |  10 ++
>>  4 files changed, 154 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/media/platform/qcom/iris/iris_platform_common.h b/drivers/media/platform/qcom/iris/iris_platform_common.h
>> index 7d59e6364e9d..8995136ad29e 100644
>> --- a/drivers/media/platform/qcom/iris/iris_platform_common.h
>> +++ b/drivers/media/platform/qcom/iris/iris_platform_common.h
>> @@ -61,6 +61,9 @@ enum platform_clk_type {
>>  	IRIS_VPP0_HW_CLK,
>>  	IRIS_VPP1_HW_CLK,
>>  	IRIS_APV_HW_CLK,
>> +	IRIS_AXI_VCODEC1_CLK,
>> +	IRIS_VCODEC1_CLK,
>> +	IRIS_VCODEC1_FREERUN_CLK,
> I think I have asked the same question while reviewing some other code.
> When seeing such enums my expectation would be that the set of clocks is
> more or less generic, while the platform specifics should be
> encapsulated in platform-specific code. Instead these lists keep on
> growing to accomodate platform details.
>
> Can we stop that tradition? Adding a peculiarity of the platform should
> not require touching of the generic code.
>
>>  };
>>  
>>  struct platform_clk_data {
>> @@ -210,6 +213,7 @@ enum platform_pm_domain_type {
>>  	IRIS_VPP0_HW_POWER_DOMAIN,
>>  	IRIS_VPP1_HW_POWER_DOMAIN,
>>  	IRIS_APV_HW_POWER_DOMAIN,
>> +	IRIS_VCODEC1_POWER_DOMAIN,
>>  };
>>  
>>  struct platform_pd_data {
>> diff --git a/drivers/media/platform/qcom/iris/iris_vpu3x.c b/drivers/media/platform/qcom/iris/iris_vpu3x.c
>> index 13fbb21c2182..ff90c375e805 100644
>> --- a/drivers/media/platform/qcom/iris/iris_vpu3x.c
>> +++ b/drivers/media/platform/qcom/iris/iris_vpu3x.c
>> @@ -22,9 +22,19 @@ static bool iris_vpu3x_hw_power_collapsed(struct iris_core *core)
>>  	u32 value, pwr_status;
>>  
>>  	value = readl(core->reg_base + WRAPPER_CORE_POWER_STATUS);
>> -	pwr_status = value & BIT(1);
>> +	pwr_status = value & VCODEC0_POWER_STATUS;
>>  
>> -	return pwr_status ? false : true;
>> +	return !pwr_status;
>> +}
>> +
>> +static bool iris_vpu36_hw1_power_collapsed(struct iris_core *core)
>> +{
>> +	u32 value, pwr_status;
>> +
>> +	value = readl(core->reg_base + WRAPPER_CORE_POWER_STATUS);
>> +	pwr_status = value & VCODEC1_POWER_STATUS;
>> +
>> +	return !pwr_status;
> Add core as an argument to the function instead of c&p'ing it.

Ack, will pass the vcodecx_power_status bit to this function.

>>  }
>>  
>>  static void iris_vpu3_power_off_hardware(struct iris_core *core)
>> @@ -254,6 +264,124 @@ static void iris_vpu35_power_off_hw(struct iris_core *core)
>>  	iris_disable_unprepare_clock(core, IRIS_AXI_VCODEC_CLK);
>>  }
>>  
>> +static int iris_vpu36_power_on_hw1(struct iris_core *core)
> Hmmm... And if 3.7 gets 4 cores, will we have 4 copies of the function?

As of now, none of the near upcoming targets introduce a significantly higher
number of cores. If that changes in the future, we can revisit and optimize it
then.

>> +{
>> +	int ret;
>> +
>> +	ret = iris_enable_power_domains(core, IRIS_VCODEC1_POWER_DOMAIN);
>> +	if (ret)
>> +		return ret;
>> +
>> +	ret = iris_prepare_enable_clock(core, IRIS_AXI_VCODEC1_CLK);
>> +	if (ret)
>> +		goto err_disable_hw1_power;
>> +
>> +	ret = iris_prepare_enable_clock(core, IRIS_VCODEC1_FREERUN_CLK);
>> +	if (ret)
>> +		goto err_disable_axi1_clk;
>> +
>> +	ret = iris_prepare_enable_clock(core, IRIS_VCODEC1_CLK);
>> +	if (ret)
>> +		goto err_disable_hw1_free_clk;
>> +
>> +	return 0;
>> +
>> +err_disable_hw1_free_clk:
>> +	iris_disable_unprepare_clock(core, IRIS_VCODEC1_FREERUN_CLK);
>> +err_disable_axi1_clk:
>> +	iris_disable_unprepare_clock(core, IRIS_AXI_VCODEC1_CLK);
>> +err_disable_hw1_power:
>> +	iris_disable_power_domains(core, IRIS_VCODEC1_POWER_DOMAIN);
>> +
>> +	return ret;
>> +}
>> +
>> +static int iris_vpu36_power_on_hw(struct iris_core *core)
>> +{
>> +	int ret;
>> +
>> +	ret = iris_vpu35_power_on_hw(core);
>> +	if (ret)
>> +		return ret;
>> +
>> +	ret = iris_vpu36_power_on_hw1(core);
>> +	if (ret)
>> +		goto err_power_off_hw;
>> +
>> +	return 0;
>> +
>> +err_power_off_hw:
>> +	iris_vpu35_power_off_hw(core);
>> +
>> +	return ret;
>> +}
>> +
>> +static void iris_vpu36_power_off_hw1(struct iris_core *core)
>> +{
>> +	u32 value, i;
>> +	int ret;
>> +
>> +	if (iris_vpu36_hw1_power_collapsed(core))
>> +		goto disable_power;
>> +
>> +	value = readl(core->reg_base + WRAPPER_CORE_CLOCK_CONFIG);
>> +	if (value)
>> +		writel(CORE_CLK_RUN, core->reg_base + WRAPPER_CORE_CLOCK_CONFIG);
>> +
>> +	for (i = 0; i < core->iris_platform_data->num_vpp_pipe; i++) {
>> +		ret = readl_poll_timeout(core->reg_base + VCODEC1_SS_IDLE_STATUSN + 4 * i,
>> +					 value, value & DMA_NOC_IDLE, 2000, 20000);
>> +		if (ret)
>> +			goto disable_power;
>> +	}
>> +
>> +	writel(REQ_VCODEC1_POWER_DOWN_PREP, core->reg_base + AON_WRAPPER_MVP_NOC_LPI_CONTROL);
>> +	ret = readl_poll_timeout(core->reg_base + AON_WRAPPER_MVP_NOC_LPI_STATUS,
>> +				 value, value & NOC_LPI_VCODEC1_STATUS_DONE, 2000, 20000);
>> +	if (ret)
>> +		goto disable_power;
>> +
>> +	writel(0, core->reg_base + AON_WRAPPER_MVP_NOC_LPI_CONTROL);
>> +
>> +	writel(VCODEC1_BRIDGE_SW_RESET | VCODEC1_BRIDGE_HW_RESET_DISABLE, core->reg_base +
>> +	       CPU_CS_AHB_BRIDGE_SYNC_RESET);
>> +	writel(VCODEC1_BRIDGE_HW_RESET_DISABLE, core->reg_base + CPU_CS_AHB_BRIDGE_SYNC_RESET);
>> +	writel(0x0, core->reg_base + CPU_CS_AHB_BRIDGE_SYNC_RESET);
>> +
>> +disable_power:
>> +	iris_genpd_set_hwmode(core, IRIS_VCODEC1_POWER_DOMAIN, false);
>> +	iris_disable_unprepare_clock(core, IRIS_VCODEC1_CLK);
>> +	iris_disable_unprepare_clock(core, IRIS_VCODEC1_FREERUN_CLK);
>> +	iris_disable_unprepare_clock(core, IRIS_AXI_VCODEC1_CLK);
>> +	iris_disable_power_domains(core, IRIS_VCODEC1_POWER_DOMAIN);
>> +}
>> +
>> +static void iris_vpu36_power_off_hw(struct iris_core *core)
>> +{
>> +	iris_vpu35_power_off_hw(core);
>> +	iris_vpu36_power_off_hw1(core);
>> +}
>> +
>> +static int iris_vpu36_set_hwmode(struct iris_core *core)
>> +{
>> +	int ret;
>> +
>> +	ret = iris_genpd_set_hwmode(core, IRIS_VCODEC_POWER_DOMAIN, true);
>> +	if (ret)
>> +		return ret;
>> +
>> +	ret = iris_genpd_set_hwmode(core, IRIS_VCODEC1_POWER_DOMAIN, true);
>> +	if (ret)
>> +		goto error_disable_vcodec_hwmode;
>> +
>> +	return 0;
>> +
>> +error_disable_vcodec_hwmode:
>> +	iris_genpd_set_hwmode(core, IRIS_VCODEC_POWER_DOMAIN, false);
>> +
>> +	return ret;
>> +}
>> +
>>  const struct vpu_ops iris_vpu3_ops = {
>>  	.power_off_hw = iris_vpu3_power_off_hardware,
>>  	.power_on_hw = iris_vpu_power_on_hw,
>> @@ -281,3 +409,12 @@ const struct vpu_ops iris_vpu35_ops = {
>>  	.calc_freq = iris_vpu3x_vpu4x_calculate_frequency,
>>  	.set_hwmode = iris_vpu_set_hwmode,
>>  };
>> +
>> +const struct vpu_ops iris_vpu36_ops = {
>> +	.power_off_hw = iris_vpu36_power_off_hw,
>> +	.power_on_hw = iris_vpu36_power_on_hw,
>> +	.power_off_controller = iris_vpu35_vpu4x_power_off_controller,
>> +	.power_on_controller = iris_vpu35_vpu4x_power_on_controller,
>> +	.calc_freq = iris_vpu3x_vpu4x_calculate_frequency,
>> +	.set_hwmode = iris_vpu36_set_hwmode,
>> +};
>> diff --git a/drivers/media/platform/qcom/iris/iris_vpu_common.h b/drivers/media/platform/qcom/iris/iris_vpu_common.h
>> index dee3b1349c5e..bee8ae9b4308 100644
>> --- a/drivers/media/platform/qcom/iris/iris_vpu_common.h
>> +++ b/drivers/media/platform/qcom/iris/iris_vpu_common.h
>> @@ -12,6 +12,7 @@ extern const struct vpu_ops iris_vpu2_ops;
>>  extern const struct vpu_ops iris_vpu3_ops;
>>  extern const struct vpu_ops iris_vpu33_ops;
>>  extern const struct vpu_ops iris_vpu35_ops;
>> +extern const struct vpu_ops iris_vpu36_ops;
>>  extern const struct vpu_ops iris_vpu4x_ops;
>>  
>>  struct vpu_ops {
>> diff --git a/drivers/media/platform/qcom/iris/iris_vpu_register_defines.h b/drivers/media/platform/qcom/iris/iris_vpu_register_defines.h
>> index 72168b9ffa73..e67d98b8c91e 100644
>> --- a/drivers/media/platform/qcom/iris/iris_vpu_register_defines.h
>> +++ b/drivers/media/platform/qcom/iris/iris_vpu_register_defines.h
>> @@ -7,6 +7,7 @@
>>  #define __IRIS_VPU_REGISTER_DEFINES_H__
>>  
>>  #define VCODEC_BASE_OFFS			0x00000000
>> +#define VCODEC1_BASE_OFFS			0x00040000
>>  #define AON_MVP_NOC_RESET			0x0001F000
>>  #define CPU_BASE_OFFS				0x000A0000
>>  #define WRAPPER_BASE_OFFS			0x000B0000
>> @@ -14,6 +15,8 @@
>>  #define AON_BASE_OFFS				0x000E0000
>>  
>>  #define VCODEC_SS_IDLE_STATUSN			(VCODEC_BASE_OFFS + 0x70)
>> +#define VCODEC1_SS_IDLE_STATUSN			(VCODEC1_BASE_OFFS + 0x70)
>> +#define DMA_NOC_IDLE				BIT(22)
>>  
>>  #define AON_WRAPPER_MVP_NOC_RESET_REQ		(AON_MVP_NOC_RESET + 0x000)
>>  #define VIDEO_NOC_RESET_REQ			(BIT(0) | BIT(1))
>> @@ -35,6 +38,8 @@
>>  #define CPU_CS_AHB_BRIDGE_SYNC_RESET		(CPU_CS_BASE_OFFS + 0x160)
>>  #define CORE_BRIDGE_SW_RESET			BIT(0)
>>  #define CORE_BRIDGE_HW_RESET_DISABLE		BIT(1)
>> +#define VCODEC1_BRIDGE_SW_RESET			BIT(2)
>> +#define VCODEC1_BRIDGE_HW_RESET_DISABLE		BIT(3)
>>  
>>  #define CPU_CS_X2RPMH				(CPU_CS_BASE_OFFS + 0x168)
>>  #define MSK_SIGNAL_FROM_TENSILICA		BIT(0)
>> @@ -52,14 +57,19 @@
>>  #define WRAPPER_DEBUG_BRIDGE_LPI_STATUS		(WRAPPER_BASE_OFFS + 0x58)
>>  #define WRAPPER_IRIS_CPU_NOC_LPI_CONTROL	(WRAPPER_BASE_OFFS + 0x5C)
>>  #define REQ_POWER_DOWN_PREP			BIT(0)
>> +#define REQ_VCODEC1_POWER_DOWN_PREP		BIT(1)
>>  
>>  #define WRAPPER_IRIS_CPU_NOC_LPI_STATUS		(WRAPPER_BASE_OFFS + 0x60)
>>  #define NOC_LPI_STATUS_DONE			BIT(0) /* Indicates the NOC handshake is complete */
>>  #define NOC_LPI_STATUS_DENY			BIT(1) /* Indicates the NOC handshake is denied */
>>  #define NOC_LPI_STATUS_ACTIVE			BIT(2) /* Indicates the NOC is active */
>> +#define NOC_LPI_VCODEC1_STATUS_DONE		BIT(8)
>>  
>>  #define WRAPPER_IRIS_VCODEC_VPU_WRAPPER_SPARE_0	(WRAPPER_BASE_OFFS + 0x78)
>>  #define WRAPPER_CORE_POWER_STATUS		(WRAPPER_BASE_OFFS + 0x80)
>> +#define VCODEC0_POWER_STATUS			BIT(1)
>> +#define VCODEC1_POWER_STATUS			BIT(4)
>> +
>>  #define WRAPPER_CORE_CLOCK_CONFIG		(WRAPPER_BASE_OFFS + 0x88)
>>  #define CORE_CLK_RUN				0x0
>>  
>>
>> -- 
>> 2.34.1
>>

  reply	other threads:[~2026-05-11 16:12 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-08 18:59 [PATCH v5 00/14] media: iris: Add support for glymur platform Vishnu Reddy
2026-05-08 18:59 ` [PATCH v5 01/14] media: iris: Add iris vpu bus support Vishnu Reddy
2026-05-08 19:16   ` Dmitry Baryshkov
2026-05-09 17:05     ` Vishnu Reddy
2026-05-08 18:59 ` [PATCH v5 02/14] iommu: Add iris-vpu-bus to iommu_buses Vishnu Reddy
2026-05-08 19:16   ` Dmitry Baryshkov
2026-05-08 18:59 ` [PATCH v5 03/14] media: iris: Fix VM count passed to firmware Vishnu Reddy
2026-05-08 19:20   ` Dmitry Baryshkov
2026-05-08 18:59 ` [PATCH v5 04/14] dt-bindings: media: qcom,venus: Remove clock, power-domain, and iommus from common schema Vishnu Reddy
2026-05-08 19:22   ` Dmitry Baryshkov
2026-05-09 17:04     ` Vishnu Reddy
2026-05-08 18:59 ` [PATCH v5 05/14] dt-bindings: media: qcom,glymur-iris: Add glymur video codec Vishnu Reddy
2026-05-08 18:59 ` [PATCH v5 06/14] media: iris: Add context bank hooks for platform specific initialization Vishnu Reddy
2026-05-08 18:59 ` [PATCH v5 07/14] media: iris: Enable Secure PAS support with IOMMU managed by Linux Vishnu Reddy
2026-05-08 19:05   ` Trilok Soni
2026-05-11  8:06     ` Mukesh Ojha
2026-05-08 20:20   ` Dmitry Baryshkov
2026-05-11  8:01     ` Mukesh Ojha
2026-05-08 18:59 ` [PATCH v5 08/14] media: iris: Rename clock and power domain macros to use vcodec prefix Vishnu Reddy
2026-05-08 20:22   ` Dmitry Baryshkov
2026-05-09 17:07     ` Vishnu Reddy
2026-05-08 18:59 ` [PATCH v5 09/14] media: iris: Use power domain type to look up pd_devs index Vishnu Reddy
2026-05-08 20:44   ` Dmitry Baryshkov
2026-05-09 17:02     ` Vishnu Reddy
2026-05-08 18:59 ` [PATCH v5 10/14] media: iris: Add power sequence for Glymur Vishnu Reddy
2026-05-08 20:54   ` Dmitry Baryshkov
2026-05-11 16:12     ` Vishnu Reddy [this message]
2026-05-08 19:00 ` [PATCH v5 11/14] media: iris: Add support to select core for dual core platforms Vishnu Reddy
2026-05-08 21:00   ` Dmitry Baryshkov
2026-05-11 16:12     ` Vishnu Reddy
2026-05-08 19:00 ` [PATCH v5 12/14] media: iris: Add platform data for glymur Vishnu Reddy
2026-05-08 21:05   ` Dmitry Baryshkov
2026-05-11 16:15     ` Vishnu Reddy
2026-05-08 19:00 ` [PATCH v5 13/14] arm64: dts: qcom: glymur: Add iris video node Vishnu Reddy
2026-05-08 19:27   ` Dmitry Baryshkov
2026-05-09 16:56     ` Vishnu Reddy
2026-05-08 19:00 ` [PATCH v5 14/14] arm64: dts: qcom: glymur-crd: Enable iris video codec node Vishnu Reddy
2026-05-08 23:54   ` Dmitry Baryshkov

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=67157e35-8c39-b06b-eb93-602930168f4b@oss.qualcomm.com \
    --to=busanna.reddy@oss.qualcomm.com \
    --cc=abhinav.kumar@linux.dev \
    --cc=andersson@kernel.org \
    --cc=angelogioacchino.delregno@collabora.com \
    --cc=bod@kernel.org \
    --cc=conor+dt@kernel.org \
    --cc=devicetree@vger.kernel.org \
    --cc=dikshita.agarwal@oss.qualcomm.com \
    --cc=dmitry.baryshkov@oss.qualcomm.com \
    --cc=hverkuil@kernel.org \
    --cc=iommu@lists.linux.dev \
    --cc=jorge.ramirez@oss.qualcomm.com \
    --cc=joro@8bytes.org \
    --cc=konradybcio@kernel.org \
    --cc=krzk+dt@kernel.org \
    --cc=krzk@kernel.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@kernel.org \
    --cc=robh@kernel.org \
    --cc=robin.murphy@arm.com \
    --cc=stanimir.varbanov@linaro.org \
    --cc=stefan.schmidt@linaro.org \
    --cc=vikash.garodia@oss.qualcomm.com \
    --cc=will@kernel.org \
    /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