| Message ID | 20230221153740.1620529-1-angelogioacchino.delregno@collabora.com |
|---|---|
| Headers | show |
| Series | Panfrost: Improve and add MediaTek SoCs support | expand |
On Wed, Feb 22, 2023 at 5:13 PM AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> wrote: > > Il 22/02/23 09:37, Chen-Yu Tsai ha scritto: > > On Tue, Feb 21, 2023 at 11:37 PM AngeloGioacchino Del Regno > > <angelogioacchino.delregno@collabora.com> wrote: > >> > >> Get GPU support on MT8186 by adding its compatible. > >> > >> Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> > >> --- > >> Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml | 5 +++++ > >> 1 file changed, 5 insertions(+) > >> > >> diff --git a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml > >> index be18b161959b..43a841d4e94d 100644 > >> --- a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml > >> +++ b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml > >> @@ -15,6 +15,11 @@ properties: > >> > >> compatible: > >> oneOf: > >> + - items: > >> + - enum: > >> + - mediatek,mt8186-mali > >> + - const: mediatek,mt8183b-mali > >> + - const: arm,mali-bifrost # Mali Bifrost GPU model/revision is fully discoverable > > > > The MT8186 has Mali-G52 MC2 2EE, while the MT8183 has Mali-G72 MP3. > > Keeping in mind the obvious - which is that G52 and G72 are both Bifrost.... > > > So we actually need a new entry with two power domains. > > > > ...This is my node for MT8186: > > gpu: gpu@13040000 { > compatible = "mediatek,mt8186-mali", > "mediatek,mt8183b-mali", > "arm,mali-bifrost"; > reg = <0 0x13040000 0 0x4000>; > > clocks = <&mfgsys CLK_MFG_BG3D>; > interrupts = <GIC_SPI 276 IRQ_TYPE_LEVEL_HIGH 0>, > <GIC_SPI 275 IRQ_TYPE_LEVEL_HIGH 0>, > <GIC_SPI 274 IRQ_TYPE_LEVEL_HIGH 0>; > interrupt-names = "job", "mmu", "gpu"; > power-domains = <&spm MT8186_POWER_DOMAIN_MFG1>, > <&spm MT8186_POWER_DOMAIN_MFG2>, > <&spm MT8186_POWER_DOMAIN_MFG3>; > power-domain-names = "core0", "core1", "core2"; > > /* Please ignore speedbin, that's for another time :-) */ > nvmem-cells = <&gpu_volt_bin>; > nvmem-cell-names = "speed-bin"; > #cooling-cells = <2>; > }; > > There are three MFG power domains... MFG2 and MFG3 are parents of MFG1, on that > I agree, but we can avoid adding a new entry just for MT8186 and use the MT8183-b > one while still being technically correct. > > Besides, Mali G52 and Mali G72 are both Bifrost... so I don't think that this > commit is incorrect. For the sake of simplicity, I would push on getting this > one picked. I'm aware. In case it wasn't obvious, Mali-G52 MC2 2EE has 2 cores, while Mali-G72 MP3 has 3 cores. I think that is reason enough to do a new entry. Otherwise you are describing power domains for 3 cores for a GPU that only has two. > Unless there are any real-strong opinions against... Yes. ChenYu
Il 23/02/23 03:43, Chen-Yu Tsai ha scritto: > On Wed, Feb 22, 2023 at 5:13 PM AngeloGioacchino Del Regno > <angelogioacchino.delregno@collabora.com> wrote: >> >> Il 22/02/23 09:37, Chen-Yu Tsai ha scritto: >>> On Tue, Feb 21, 2023 at 11:37 PM AngeloGioacchino Del Regno >>> <angelogioacchino.delregno@collabora.com> wrote: >>>> >>>> Get GPU support on MT8186 by adding its compatible. >>>> >>>> Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> >>>> --- >>>> Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml | 5 +++++ >>>> 1 file changed, 5 insertions(+) >>>> >>>> diff --git a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml >>>> index be18b161959b..43a841d4e94d 100644 >>>> --- a/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml >>>> +++ b/Documentation/devicetree/bindings/gpu/arm,mali-bifrost.yaml >>>> @@ -15,6 +15,11 @@ properties: >>>> >>>> compatible: >>>> oneOf: >>>> + - items: >>>> + - enum: >>>> + - mediatek,mt8186-mali >>>> + - const: mediatek,mt8183b-mali >>>> + - const: arm,mali-bifrost # Mali Bifrost GPU model/revision is fully discoverable >>> >>> The MT8186 has Mali-G52 MC2 2EE, while the MT8183 has Mali-G72 MP3. >> >> Keeping in mind the obvious - which is that G52 and G72 are both Bifrost.... >> >>> So we actually need a new entry with two power domains. >>> >> >> ...This is my node for MT8186: >> >> gpu: gpu@13040000 { >> compatible = "mediatek,mt8186-mali", >> "mediatek,mt8183b-mali", >> "arm,mali-bifrost"; >> reg = <0 0x13040000 0 0x4000>; >> >> clocks = <&mfgsys CLK_MFG_BG3D>; >> interrupts = <GIC_SPI 276 IRQ_TYPE_LEVEL_HIGH 0>, >> <GIC_SPI 275 IRQ_TYPE_LEVEL_HIGH 0>, >> <GIC_SPI 274 IRQ_TYPE_LEVEL_HIGH 0>; >> interrupt-names = "job", "mmu", "gpu"; >> power-domains = <&spm MT8186_POWER_DOMAIN_MFG1>, >> <&spm MT8186_POWER_DOMAIN_MFG2>, >> <&spm MT8186_POWER_DOMAIN_MFG3>; >> power-domain-names = "core0", "core1", "core2"; >> >> /* Please ignore speedbin, that's for another time :-) */ >> nvmem-cells = <&gpu_volt_bin>; >> nvmem-cell-names = "speed-bin"; >> #cooling-cells = <2>; >> }; >> >> There are three MFG power domains... MFG2 and MFG3 are parents of MFG1, on that >> I agree, but we can avoid adding a new entry just for MT8186 and use the MT8183-b >> one while still being technically correct. >> >> Besides, Mali G52 and Mali G72 are both Bifrost... so I don't think that this >> commit is incorrect. For the sake of simplicity, I would push on getting this >> one picked. > > I'm aware. In case it wasn't obvious, Mali-G52 MC2 2EE has 2 cores, while > Mali-G72 MP3 has 3 cores. I think that is reason enough to do a new entry. > Otherwise you are describing power domains for 3 cores for a GPU that only > has two. > >> Unless there are any real-strong opinions against... > > Yes. > Ok, I will do that for v3! Cheers, Angelo > ChenYu
Changes in v2: - Add power-domain-names commit from Chen-Yu to the series - Kept sram-supply in base schema, overridden for non-MediaTek - Added Reviewed-by tags from Steven Price to the driver commits (as released in reply to v1's cover letter - thanks!) This series adds support for new MediaTek SoCs (MT8186/MT8192/MT8195) and improves MT8183 support: since the mtk-regulator-coupler driver was picked, it is now useless for Panfrost to look for, and manage, two regulators (GPU Vcore and GPU SRAM) on MediaTek; The aforementioned driver will take care of keeping the voltage relation (/constraints) of the two regulators on its own when a voltage change request is sent to the Vcore, solving the old time issue with not working DVFS on Panfrost+MediaTek (due to devfreq supporting only single regulator). In the specific case of MT8183, in order to not break the ABI, it was necessary to add a new compatible for enabling DVFS. Alyssa Rosenzweig (3): drm/panfrost: Increase MAX_PM_DOMAINS to 5 drm/panfrost: Add the MT8192 GPU ID drm/panfrost: Add mediatek,mt8192-mali compatible AngeloGioacchino Del Regno (6): dt-bindings: gpu: mali-bifrost: Split out MediaTek power-domains variation dt-bindings: gpu: mali-bifrost: Allow up to 5 power domains for MT8192 dt-bindings: gpu: mali-bifrost: Add compatible for MT8195 SoC dt-bindings: gpu: mali-bifrost: Add new MT8183 compatible dt-bindings: gpu: mali-bifrost: Add a compatible for MediaTek MT8186 drm/panfrost: Add new compatible for Mali on the MT8183 SoC Chen-Yu Tsai (1): dt-bindings: gpu: mali-bifrost: Add power-domain-names to base schema .../bindings/gpu/arm,mali-bifrost.yaml | 67 ++++++++++++++++++- drivers/gpu/drm/panfrost/panfrost_device.h | 2 +- drivers/gpu/drm/panfrost/panfrost_drv.c | 28 ++++++++ drivers/gpu/drm/panfrost/panfrost_gpu.c | 8 +++ 4 files changed, 101 insertions(+), 4 deletions(-)