Re: [PATCH] Revert "clk: qcom: dispcc-sm8250: Enable parents for pixel clocks"

On Thu, 13 Aug 2026 16:21:54 +0300
Dmitry Baryshkov <[email protected]> wrote:

> > Yes, please.  
> 
> And then almost immediately (sorry). No. The flag is correct. To
> unblock SDM845 try adding GENPD_FLAG_NO_STAY_ON to the MDSS_GDSC. The
> MDSS reset is not working / accessible, so it might be the only way
> forward for now.
> 
> 

Like this?

diff --git a/drivers/clk/qcom/dispcc-sdm845.c
b/drivers/clk/qcom/dispcc-sdm845.c index e060097be413..bfa657113ec0
100644 --- a/drivers/clk/qcom/dispcc-sdm845.c
+++ b/drivers/clk/qcom/dispcc-sdm845.c
@@ -764,7 +764,7 @@ static struct gdsc mdss_gdsc = {
                .name = "mdss_gdsc",
        },
        .pwrsts = PWRSTS_OFF_ON,
-       .flags = HW_CTRL | POLL_CFG_GDSCR,
+       .flags = HW_CTRL | POLL_CFG_GDSCR | GENPD_FLAG_NO_STAY_ON,
 };
 
 static struct clk_regmap *disp_cc_sdm845_clocks[] = {

This alone does not seem to solve the problem, I still get crashreports
like:

[    3.493402] /soc@0/display-subsystem@ae00000/dsi@ae94000: Fixed
dependency cycle(s) with
/soc@0/geniqup@8c0000/i2c@88c000/hdmi-bridge@3b [    3.513397]
/soc@0/display-subsystem@ae00000/dsi@ae94000: Fixed dependency cycle(s)
with /soc@0/display-subsystem@ae00000/display-controller@ae01000 [
3.537096] msm_dsi_phy ae94400.phy: [drm:dsi_pll_10nm_vco_prepare]
*ERROR* DSI PLL(0) lock failed, status=0x00000000 [    3.547813]
msm_dsi_phy ae94400.phy: [drm:dsi_pll_10nm_vco_prepare] *ERROR* PLL(0)
lock failed [    3.548807] qcom-pcie 1c00000.pcie: iATU: unroll F, 32
ob, 8 ib, align 4K, limit 4G [    3.556493] ------------[ cut here
]------------ [    3.556495] dsi0_phy_pll_out_dsiclk already disabled [
   3.556512] WARNING: drivers/clk/clk.c:1188 at
clk_core_disable+0xb0/0xd0, CPU#4: kworker/u32:1/60 [    3.556518]
Modules linked in: [    3.556524] CPU: 4 UID: 0 PID: 60 Comm:
kworker/u32:1 Not tainted 7.2.0-rc7-sdm845 #1-postmarketos-qcom-sdm845
PREEMPT [    3.556528] Hardware name: qcom  Inc. SDM845 HDK/ Inc.
SDM845 HDK, BIOS 2026.07-rc1-g2abf33b7c8ea-dirty 07/01/2026 [
3.556530] Workqueue: events_unbound deferred_probe_work_func [
3.572719] qcom-pcie 1c08000.pcie: iATU: unroll F, 8 ob, 8 ib, align 4K,
limit 1024G

[    3.573911] pstate: 604000c5 (nZCv daIF +PAN -UAO -TCO -DIT -SSBS
BTYPE=--) [    3.573914] pc : clk_core_disable+0xb0/0xd0
[    3.573916] lr : clk_core_disable+0xb0/0xd0

Can share the full dump off-list if wanted.

BR,
RObert Eckelmann