On Tue, Sep 16, 2025 at 01:54:54PM +0200, Thorsten Leemhuis wrote:
> On 16.09.25 01:22, ninelore wrote:
> > starting with the bisected commit
> > fae58d0155a979a8c414bbc12db09dd4b2f910d0 ("[v2] drm/mediatek: Fix
> > device/node reference count leaks in mtk_drm_get_all_drm_priv") on
> > 6.16.y my chromebook codenamed google-tomato with the SoC MT8195
> > seems to hang while initializing drm. Kernel messages over serial
> > show CPU stalls some minutes later. I am attaching the full kernel
> > logs below.
> >
> > This issue is present in 6.16.7, however not in v6.17-rc6
>
> A problem caused by the upstream commit of the change you mentioned
> above was fixed for v6.17-rc6, see 4de37a48b6b58f ("drm/mediatek: fix
> potential OF node use-after-free") or
> https://lore.kernel.org/all/[email protected]/
>
> I wonder if that might fix your problem. It's currently scheduled to be
> included in the next 6.16.y release.
This is a regression in 6.17-rc4 and the offending commit has indeed
been backported to the stable trees.
A fix for the regression has been sitting here for a few weeks now:
https://lore.kernel.org/lkml/[email protected]/
Johan
