On Tue, Jun 23, 2020 at 07:11:49PM +0200, Julia Lawall wrote:
> It looks like the put_device (line 891) should be one line earlier.
>
Yikes - thanks for that.
I'll be sure to coccinelle my code before submitting from now on.
Cheers,
Kent.
On Tue, Jun 23, 2020 at 08:44:38PM +0300, Dan Carpenter wrote:
> [ The copy_to_user() overflow code is weird. Why do we need to do a
> atomic_read()? That suggests that there is a potential time of check
> time of use bug where we do:
>
It is weird, but you conveniently left out the guard
[ The copy_to_user() overflow code is weird. Why do we need to do a
atomic_read()? That suggests that there is a potential time of check
time of use bug where we do:
if (atomic_read(>watch_abi_version) == 2) // <<-- time of check
event_size = sizeof(struct
It looks like the put_device (line 891) should be one line earlier.
julia
-- Forwarded message --
Date: Tue, 23 Jun 2020 15:26:47 +0800
From: kernel test robot
To: kbu...@lists.01.org
Cc: l...@intel.com, Julia Lawall
Subject: Re: [PATCH 16/22] gpiolib: cdev: add V2 uAPI
Add implementation of the V2 uAPI up to parity with V1.
Signed-off-by: Kent Gibson
---
This patch only covers the V2 functionality that is a direct analogue of
the V1. New V2 functionality is added in subsequent patches.
The core of the implementation is the struct line, which is drawn from
5 matches
Mail list logo