Re: [PATCH v2 0/2] extcon: synchronize all-connector removal and fix Lenovo fastcharger teardown
From: Rong Zhang
Date: Mon Nov 16 2026 - 08:04:09 EST
Hi Hongyan,
On Sun, 2026-11-15 at 19:47 +0800, Hongyan Xu wrote:
> The Lenovo Yoga Tab 2 fastcharger driver releases managed serdev,
> pinctrl and GPIO resources while its extcon callback can still publish
> work. Ordering the devres actions alone is not enough because the
> all-connector raw notifier does not wait for an in-flight callback.
>
> Patch 1 converts only the all-connector notifier chain to an atomic
> chain so unregister waits for callbacks and is safe from interrupt
> context. The per-ID chains stay raw because some current callbacks
> perform sleeping regulator operations. Patch 2 then installs the
> managed work cancellation before the notifier, so reverse devres order
> stops the producer before releasing the resources it uses.
>
Where's the link to v1?
>
> Hongyan Xu (2):
> extcon: synchronize all-connector notifier removal
> platform/x86: lenovo: cancel fastcharger work on detach
>
> Documentation/driver-api/extcon.rst | 2 +-
> drivers/extcon/extcon.c | 28 ++++++-------------
> drivers/extcon/extcon.h | 2 +-
> .../lenovo/yoga-tab2-pro-1380-fastcharger.c | 6 +++-
> include/linux/extcon.h | 4 +++
> 5 files changed, 20 insertions(+), 22 deletions(-)
>
> --
> 2.50.1.windows.1
Using Windows to submit this is quite weird.
You submitted a lot of patches to different subsystems at the same
minute. Did you send the email yourself, or asked an LLM agent to flood
LKML for you? Did you (or your LLM agent) test it?
Could you at least figure out how to send patches in UTF-8 correctly
before deciding to use git-send-email on Windows? By using Windows, your
configuration couldn't even encode Ilpo's last name correctly (I've fixed
it in my reply) :-/
>
> Discovery: found by the author's in-house static analysis tool.
Wasn't this just LLM generated?
Confused,
Rong