Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <7b493d85-0691-8797-367e-1d71ea87c826@linaro.org>
Date:   Sat, 22 Apr 2023 01:16:45 +0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.9.0
Subject: Re: [PATCH v1 5/5] drm/msm/dpu: add DSC 1.2 hw blocks for relevant
 chipsets
Content-Language: en-GB
To:     Kuogee Hsieh <quic_khsieh@quicinc.com>,
        dri-devel@lists.freedesktop.org, robdclark@gmail.com,
        sean@poorly.run, swboyd@chromium.org, dianders@chromium.org,
        vkoul@kernel.org, daniel@ffwll.ch, airlied@gmail.com,
        agross@kernel.org, andersson@kernel.org
Cc:     Abhinav Kumar <quic_abhinavk@quicinc.com>,
        quic_sbillaka@quicinc.com, marijn.suijten@somainline.org,
        freedreno@lists.freedesktop.org, linux-arm-msm@vger.kernel.org,
        linux-kernel@vger.kernel.org
References: <1682033114-28483-1-git-send-email-quic_khsieh@quicinc.com>
 <1682033114-28483-6-git-send-email-quic_khsieh@quicinc.com>
 <b26dfb22-bf97-b65e-ef06-62098c4eafec@linaro.org>
 <3ee67248-c94c-5f3d-527e-914e8c8b4a31@quicinc.com>
From:   Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
In-Reply-To: <3ee67248-c94c-5f3d-527e-914e8c8b4a31@quicinc.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 22/04/2023 01:05, Kuogee Hsieh wrote:
> 
> On 4/20/2023 5:07 PM, Dmitry Baryshkov wrote:
>> On 21/04/2023 02:25, Kuogee Hsieh wrote:
>>> From: Abhinav Kumar <quic_abhinavk@quicinc.com>
>>>
>>> Add DSC 1.2 hardware blocks to the catalog with necessary
>>> sub-block and feature flag information.
>>> Each display compression engine (DCE) contains dual hard
>>> slice DSC encoders so both share same base address but with
>>> its own different sub block address.
>>
>> Please correct line wrapping. 72-75 is usually the preferred width
>>
>>>
>>> Signed-off-by: Abhinav Kumar <quic_abhinavk@quicinc.com>
>>> Signed-off-by: Kuogee Hsieh <quic_khsieh@quicinc.com>
>>> ---
>>>   .../gpu/drm/msm/disp/dpu1/catalog/dpu_7_0_sm8350.h  | 19 
>>> +++++++++++++++++++
>>>   .../gpu/drm/msm/disp/dpu1/catalog/dpu_7_2_sc7280.h  | 11 +++++++++++
>>>   .../drm/msm/disp/dpu1/catalog/dpu_8_0_sc8280xp.h    | 21 
>>> +++++++++++++++++++++
>>>   .../gpu/drm/msm/disp/dpu1/catalog/dpu_8_1_sm8450.h  | 19 
>>> +++++++++++++++++++
>>>   .../gpu/drm/msm/disp/dpu1/catalog/dpu_9_0_sm8550.h  | 19 
>>> +++++++++++++++++++
>>>   drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c      | 12 ++++++++++--
>>>   6 files changed, 99 insertions(+), 2 deletions(-)
>>>
>>
>>
>> [I commented on sm8550, it applies to all the rest of platforms]
>>
>>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/catalog/dpu_9_0_sm8550.h 
>>> b/drivers/gpu/drm/msm/disp/dpu1/catalog/dpu_9_0_sm8550.h
>>> index 9e40303..72a7bcf 100644
>>> --- a/drivers/gpu/drm/msm/disp/dpu1/catalog/dpu_9_0_sm8550.h
>>> +++ b/drivers/gpu/drm/msm/disp/dpu1/catalog/dpu_9_0_sm8550.h
>>> @@ -165,6 +165,23 @@ static const struct dpu_merge_3d_cfg 
>>> sm8550_merge_3d[] = {
>>>       MERGE_3D_BLK("merge_3d_3", MERGE_3D_3, 0x66700),
>>>   };
>>>   +static const struct dpu_dsc_sub_blks sm8550_dsc_sblk_0 = {
>>> +    .enc = {.base = 0x100, .len = 0x100},
>>> +    .ctl = {.base = 0xF00, .len = 0x10},
>>> +};
>>> +
>>> +static const struct dpu_dsc_sub_blks sm8550_dsc_sblk_1 = {
>>> +    .enc = {.base = 0x200, .len = 0x100},
>>> +    .ctl = {.base = 0xF80, .len = 0x10},
>>> +};
>>
>> Please keep sblk in dpu_hw_catalog for now.
>>
>>> +
>>> +static const struct dpu_dsc_cfg sm8550_dsc[] = {
>>> +    DSC_BLK_1_2("dsc_0", DSC_0, 0x80000, 0x100, 0, sm8550_dsc_sblk_0),
>>> +    DSC_BLK_1_2("dsc_0", DSC_1, 0x80000, 0x100, 0, sm8550_dsc_sblk_1),
>>
>> Is there a reason why index in "dsc_N" doesn't match the DSC_n which 
>> comes next to it?
> 
> usually each DCE (display compression engine) contains two hard slice 
> encoders.
> 
> DSC_0 and DSC_1 (index) is belong to dsc_0.
> 
> If there are two DCE, then DSC_2 and DSC_3 belong to dsc_1

Ah, I see now. So, the block register space is the following:
DCEi ->
   common
   dsc0_enc
   dsc1_enc
   dsc0_ctl
   dsc1_ctl

Instead of declaring a single DCE unit with two DSC blocks, we declare 
two distinct DSC blocks. This raises a question, how independent are 
these two parts of a single DCE block? For example, can we use them to 
perform compression with different parameters? Or use one of them for 
the DP DSC and another one for DSI DSC? Can we have the following 
configuration:

DSC_0 => DP DSC
DSC_1, DSC_2 => DSI DSC in DSC_MERGE topology?

> 
>>
>>> +    DSC_BLK_1_2("dsc_1", DSC_2, 0x81000, 0x100, 
>>> BIT(DPU_DSC_NATIVE_422_EN), sm8550_dsc_sblk_0),
>>> +    DSC_BLK_1_2("dsc_1", DSC_3, 0x81000, 0x100, 
>>> BIT(DPU_DSC_NATIVE_422_EN), sm8550_dsc_sblk_1),
>>> +};
>>> +
>>>   static const struct dpu_intf_cfg sm8550_intf[] = {
>>>       INTF_BLK("intf_0", INTF_0, 0x34000, 0x280, INTF_DP, 
>>> MSM_DP_CONTROLLER_0, 24, INTF_SC7280_MASK, MDP_SSPP_TOP0_INTR, 24, 25),
>>>       /* TODO TE sub-blocks for intf1 & intf2 */
>>> @@ -218,6 +235,8 @@ const struct dpu_mdss_cfg dpu_sm8550_cfg = {
>>>       .dspp = sm8550_dspp,
>>>       .pingpong_count = ARRAY_SIZE(sm8550_pp),
>>>       .pingpong = sm8550_pp,
>>> +    .dsc = sm8550_dsc,
>>> +    .dsc_count = ARRAY_SIZE(sm8550_dsc),
>>>       .merge_3d_count = ARRAY_SIZE(sm8550_merge_3d),
>>>       .merge_3d = sm8550_merge_3d,
>>>       .intf_count = ARRAY_SIZE(sm8550_intf),
>>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c 
>>> b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c
>>> index 03f162a..be08158 100644
>>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c
>>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_catalog.c
>>> @@ -1,6 +1,6 @@
>>>   // SPDX-License-Identifier: GPL-2.0-only
>>>   /* Copyright (c) 2015-2018, The Linux Foundation. All rights reserved.
>>> - * Copyright (c) 2022. Qualcomm Innovation Center, Inc. All rights 
>>> reserved.
>>> + * Copyright (c) 2022-2023, Qualcomm Innovation Center, Inc. All 
>>> rights reserved.
>>>    */
>>>     #define pr_fmt(fmt)    "[drm:%s:%d] " fmt, __func__, __LINE__
>>> @@ -540,7 +540,15 @@ static const struct dpu_pingpong_sub_blks 
>>> sc7280_pp_sblk = {
>>>       {\
>>>       .name = _name, .id = _id, \
>>>       .base = _base, .len = 0x140, \
>>> -    .features = _features, \
>>> +    .features = BIT(DPU_DSC_HW_REV_1_1) | _features, \
>>> +    }
>>> +
>>> +#define DSC_BLK_1_2(_name, _id, _base, _len, _features, _sblk) \
>>> +    {\
>>> +    .name = _name, .id = _id, \
>>> +    .base = _base, .len = _len, \
>>> +    .features = BIT(DPU_DSC_HW_REV_1_2) | _features, \
>>> +    .sblk = &_sblk, \
>>>       }
>>> /*************************************************************
>>

-- 
With best wishes
Dmitry