From: Thomas Turner <tturner@lineageos.org>
To: Alexey Klimov <alexey.klimov@linaro.org>,
semen.protsenko@linaro.org, robh@kernel.org, krzk+dt@kernel.org,
conor+dt@kernel.org, peter.griffin@linaro.org
Cc: alim.akhtar@samsung.com, linux-arm-kernel@lists.infradead.org,
linux-samsung-soc@vger.kernel.org, devicetree@vger.kernel.org,
linux-kernel@vger.kernel.org, thomas_turner@talktalk.net
Subject: Re: [PATCH v3 2/2] arm64: dts: exynos: Add GPU/Mali Bifrost node to Exynos850
Date: Mon, 7 Sep 2026 18:47:23 +0100 [thread overview]
Message-ID: <45014ba7-e213-4624-be45-78e88c1b52a0@lineageos.org> (raw)
In-Reply-To: <DL8NQGNAVSYK.18TYE03LPQSW5@linaro.org>
Hi Alexey,
On 07/09/2026 01:28, Alexey Klimov wrote:
> On Sat Sep 5, 2026 at 6:09 PM BST, Thomas Turner wrote:
>> Add nodes for GPU (Mali Bifrost) to Exynos850. Does not provide support
>> for temperature throttling, DVFS, power domains, governor support,
>
> One of the sensors exposed via ACPM is "linked" to the GPU/G3D as far as
> I know. The driver part was sent already and three sensors work there
> (two cpu clusters and GPU).
>
> I've never tested yet the DVFS of G3D via ACPM. Maybe it is worth a try
> on A21s. But it should work the same way as clusters clocks.
>
I have had a little look at your ACPM stuff currently staged, looks like
some really good work, would be cool to see if we could get dvfs working
via acpm at some point :D.
>> boosting and voltage control. Once PMIC is properly implemented, support
>> will hopefully increase. However, the panfrost driver probes, the
>> Mali-G52 GPU initializes and the core clock is set-up.
>>
>> Testing for this patch was done on a device not currently supported by
>> linux, the Galaxy A21s. After enabling this node for that device
>> locally, I was able to test and confirm basic GPU functionality. As of
>> now, this patch has no users which have tested GPU capabilities and so
>> the node has been set to "disabled".
>>
>> Here are some logs I pulled:
>>
>> panfrost 11500000.gpu: clock rate = 377000000
>> panfrost 11500000.gpu: mali-g52 id 0x7212 major 0x0 minor 0x0 status 0x0
>> panfrost 11500000.gpu: features: 00000000,00000df7, issues: 00000000,00000400
>> panfrost 11500000.gpu: Features: L2:0x07100206 Shader:0x00000000 Tiler:0x00000809 Mem:0x1 MMU:0x00002830 AS:0xff JS:0x7
>> panfrost 11500000.gpu: shader_present=0x1 l2_present=0x1
>> panfrost 11500000.gpu: [drm] Using Transparent Hugepage
>> [drm] Initialized panfrost 1.6.0 for 11500000.gpu on minor 0
>>
>> OpenGL core profile vendor: Mesa
>> OpenGL core profile renderer: Mali-G52 MC1 (Panfrost)
>> OpenGL core profile version: 3.1 (Core Profile) Mesa 26.1.6
>> OpenGL core profile shading language version: 1.40
>> OpenGL compatibility profile vendor: Mesa
>> OpenGL compatibility profile renderer: Mali-G52 MC1 (Panfrost)
>> OpenGL compatibility profile version: 3.1 (Core Profile) Mesa 26.1.6
>> OpenGL compatibility profile shading language version: 1.40
>> OpenGL ES profile vendor: Mesa
>> OpenGL ES profile renderer: Mali-G52 MC1 (Panfrost)
>> OpenGL ES profile version: OpenGL ES 3.1 Mesa 26.1.6
>> OpenGL ES profile shading language version: OpenGL ES GLSL ES 3.10
>>
>> Signed-off-by: Thomas Turner <tturner@lineageos.org>
>
> So, I've got A21s and installed pmOS to test all that.
> My logs are slightly different but graphics is definitely alive:
> [ 0.200787] [drm] Initialized simpledrm 1.0.0 for fa000000.framebuffer on minor 0
> [ 0.215130] Console: switching to colour frame buffer device 90x100
> [ 0.225540] simple-framebuffer fa000000.framebuffer: [drm] fb0: simpledrmdrmfb frame buffer device
> [ 0.226134] panfrost 11500000.gpu: clock rate = 377000000
> [ 0.230960] panfrost 11500000.gpu: mali-g52 id 0x7212 major 0x0 minor 0x0 status 0x0
> [ 0.230981] panfrost 11500000.gpu: features: 00000000,00000df7, issues: 00000000,00000400
> [ 0.230991] panfrost 11500000.gpu: Features: L2:0x07100206 Shader:0x00000000 Tiler:0x00000809 Mem:0x1 MMU:0x00002830 AS:0xff JS:0x7
> [ 0.231002] panfrost 11500000.gpu: shader_present=0x1 l2_present=0x1
> [ 0.232329] panfrost 11500000.gpu: [drm] Using Transparent Hugepage
> [ 0.232726] [drm] Initialized panfrost 1.6.0 for 11500000.gpu on minor 1
>
> Also, when this is applied with status="okay" on Winlink board then
> panfrost itself seems to be alive and drm initialises despite missing
> regulator as indicated in my other email in v2 of this series, hence:
Alright cool, just for the records we have tested this series on the
A21s both with and without proper regulators and GPU seems to work fine
with dummy regs.
> Reviewed-and-tested-by: Alexey Klimov <alexey.klimov@linaro.org>
>
Cheers, will give this version a few days to stew then I will add this
onto v4.
> Thanks for working on this.
> Best regareds,
> Alexey
I really appreciate all of the testing and insight you have provided in
the last few messages, now that we have Samsung E850s booting I hope to
expand support across the whole SoC :D.
Cheers,
Tom
prev parent reply other threads:[~2026-09-07 17:47 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-05 17:09 [PATCH v3 0/2] Exynos850 GPU support Thomas Turner
2026-09-05 17:09 ` [PATCH v3 1/2] dt-bindings: gpu: arm,mali-bifrost: add exynos850-mali compatible Thomas Turner
2026-09-05 17:09 ` [PATCH v3 2/2] arm64: dts: exynos: Add GPU/Mali Bifrost node to Exynos850 Thomas Turner
2026-09-07 0:28 ` Alexey Klimov
2026-09-07 17:47 ` Thomas Turner [this message]
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=45014ba7-e213-4624-be45-78e88c1b52a0@lineageos.org \
--to=tturner@lineageos.org \
--cc=alexey.klimov@linaro.org \
--cc=alim.akhtar@samsung.com \
--cc=conor+dt@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=krzk+dt@kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-samsung-soc@vger.kernel.org \
--cc=peter.griffin@linaro.org \
--cc=robh@kernel.org \
--cc=semen.protsenko@linaro.org \
--cc=thomas_turner@talktalk.net \
/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