From: Venkat Yadlapati via B4 Relay <devnull+venkat.yadlapati.yahoo.com@kernel.org>
To: Rob Clark <robin.clark@oss.qualcomm.com>,
Sean Paul <sean@poorly.run>,
Konrad Dybcio <konradybcio@kernel.org>,
Akhil P Oommen <akhilpo@oss.qualcomm.com>,
Dmitry Baryshkov <lumag@kernel.org>,
Abhinav Kumar <abhinav.kumar@linux.dev>,
Jessica Zhang <jesszhan0024@gmail.com>,
Marijn Suijten <marijn.suijten@somainline.org>,
David Airlie <airlied@gmail.com>,
Simona Vetter <simona@ffwll.ch>
Cc: linux-arm-msm@vger.kernel.org, dri-devel@lists.freedesktop.org,
freedreno@lists.freedesktop.org, linux-kernel@vger.kernel.org,
phone-devel@vger.kernel.org,
Venkat Yadlapati <venkat.yadlapati@yahoo.com>
Subject: [PATCH v2] drm/msm/a6xx: Fall back to MX when GMxC has no levels
Date: Mon, 23 Nov 2026 00:50:07 +0530 [thread overview]
Message-ID: <20261123-msm-gmxc-fallback-v2-1-5dccc6a2ce34@yahoo.com> (raw)
From: Venkat Yadlapati <venkat.yadlapati@yahoo.com>
The command DB of Molokai (SM8845) lists the gmxc.lvl resource without
any level. Since a6xx_gmu_rpmh_read_arc() trims the trailing zero
levels, reading it now fails with -EINVAL, so the GX votes are never
built and the GMU of the Adreno 829 hangs on the first GX vote:
Message HFI_H2F_MSG_GX_BW_PERF_VOTE id 8 timed out waiting for response
Timeout waiting for GMU OOB set GPU_SET: 0x4
gpu hw init failed: -110
Before, the empty list made the GX votes use the lowest GMxC corner and
the GPU worked. Read gmxc.lvl the same way when choosing the secondary
rail, so that a GMxC without levels is treated like a missing one and
the GX votes use MX.
Fixes: b5269bf0e1ad ("drm/msm/a6xx: Factor out RPMh arc-vote helpers")
Signed-off-by: Venkat Yadlapati <venkat.yadlapati@yahoo.com>
---
Found while bringing up the Adreno 829 of the Motorola Signature
(Qualcomm Molokai, SM8845) on next-20261113: the GMU firmware loads but
never answers the first GX vote. Bisecting the arc vote rework points at
commit b5269bf0e1ad: the Molokai command DB lists gmxc.lvl without any
level, which the new helper rejects, so the GX votes are never built.
Tested on the Motorola Signature: the GMU answers, the GPU comes up and
Phosh renders on it.
I used an LLM (Claude) to help write this patch and its commit message.
I reviewed it, bisected the regression on the device and tested the
result there.
---
Changes in v2:
- Drop the Assisted-by trailers and the LLM notes from the commit
messages, the LLM use is stated in this cover letter.
- Link to v1: https://patch.msgid.link/20261116-msm-gmxc-fallback-v1-1-a9daad716a1e@yahoo.com
---
drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 8 ++++++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
index 662f09bec9d5..914434b49eed 100644
--- a/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
+++ b/drivers/gpu/drm/msm/adreno/a6xx_gmu.c
@@ -1889,13 +1889,17 @@ static int a6xx_gmu_rpmh_votes_init(struct a6xx_gmu *gmu)
struct msm_gpu *gpu = &adreno_gpu->base;
const char *sec_id;
const u16 *gmxc;
+ size_t gmxc_count;
int ret;
- gmxc = cmd_db_read_aux_data("gmxc.lvl", NULL);
+ gmxc = a6xx_gmu_rpmh_read_arc("gmxc.lvl", &gmxc_count);
if (gmxc == ERR_PTR(-EPROBE_DEFER))
return -EPROBE_DEFER;
- /* If GMxC is present, prefer that as secondary rail for GX votes */
+ /*
+ * If GMxC is present, prefer that as secondary rail for GX votes. Some
+ * SoCs list it in the command DB without any level, use MX there.
+ */
sec_id = IS_ERR_OR_NULL(gmxc) ? "mx.lvl" : "gmxc.lvl";
/* Build the GX votes */
---
base-commit: d024361c631b3f020a26c07d683e5103404b0fbf
change-id: 20261116-msm-gmxc-fallback-b59be82e6c2d
Best regards,
--
Venkat Yadlapati <venkat.yadlapati@yahoo.com>
reply other threads:[~2026-11-22 19:20 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20261123-msm-gmxc-fallback-v2-1-5dccc6a2ce34@yahoo.com \
--to=devnull+venkat.yadlapati.yahoo.com@kernel.org \
--cc=abhinav.kumar@linux.dev \
--cc=airlied@gmail.com \
--cc=akhilpo@oss.qualcomm.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=freedreno@lists.freedesktop.org \
--cc=jesszhan0024@gmail.com \
--cc=konradybcio@kernel.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lumag@kernel.org \
--cc=marijn.suijten@somainline.org \
--cc=phone-devel@vger.kernel.org \
--cc=robin.clark@oss.qualcomm.com \
--cc=sean@poorly.run \
--cc=simona@ffwll.ch \
--cc=venkat.yadlapati@yahoo.com \
/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
all inboxes | Powered by JetHome®