[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <alpine.LFD.2.21.2002102306000.191510@ninjahub.org>
Date: Mon, 10 Feb 2020 23:09:59 +0000 (GMT)
From: Jules Irenge <jbi.octave@...il.com>
To: Boqun Feng <boqun.feng@...il.com>
cc: Jules Irenge <jbi.octave@...il.com>, linux-kernel@...r.kernel.org,
linux-mm@...ck.org, kasan-dev@...glegroups.com,
akpm@...ux-foundation.org, dvyukov@...gle.com, glider@...gle.com,
aryabinin@...tuozzo.com, bsegall@...gle.com, rostedt@...dmis.org,
dietmar.eggemann@....com, vincent.guittot@...aro.org,
juri.lelli@...hat.com, peterz@...radead.org, mingo@...hat.com,
mgorman@...e.de, dvhart@...radead.org, tglx@...utronix.de,
namhyung@...nel.org, jolsa@...hat.com,
alexander.shishkin@...ux.intel.com, mark.rutland@....com,
acme@...nel.org, viro@...iv.linux.org.uk,
linux-fsdevel@...r.kernel.org
Subject: Re: [PATCH 00/11] Lock warning cleanup
On Mon, 10 Feb 2020, Boqun Feng wrote:
> Hi Jules,
>
> On Sun, Feb 09, 2020 at 10:24:42PM +0000, Jules Irenge wrote:
> > This patch series adds missing annotations to functions that register warnings of context imbalance when built with Sparse tool.
> > The adds fix the warnings and give insight on what the functions are actually doing.
> >
> > 1. Within the futex subsystem, a __releases(&pi_state->.pi_mutex.wait_lock) is added because wake_futex_pi() only releases the lock at exit,
> > must_hold(q->lock_ptr) have been added to fixup_pi_state_owner() because the lock is held at entry and exit;
> > a __releases(&hb->lock) added to futex_wait_queue_me() as it only releases the lock.
> >
> > 2. Within fs_pin, a __releases(RCU) is added because the function exit RCU critical section at exit.
> >
> > 3. In kasan, an __acquires(&report_lock) has been added to start_report() and __releases(&report_lock) to end_report()
> >
> > 4. Within ring_buffer subsystem, a __releases(RCU) has been added perf_output_end()
> >
> > 5. schedule subsystem recorded an addition of the __releases(rq->lock) annotation and a __must_hold(this_rq->lock)
> >
> > 6. At hrtimer subsystem, __acquires(timer) is added to lock_hrtimer_base() as the function acquire the lock but never releases it.
> > Jules Irenge (11):
> > hrtimer: Add missing annotation to lock_hrtimer_base()
> > futex: Add missing annotation for wake_futex_pi()
> > futex: Add missing annotation for fixup_pi_state_owner()
>
> Given that those three patches have been sent and reviewed, please do
> increase the version number (this time, for example, using v2) when
> sending the updated ones. Also please add a few sentences after the
> commit log describing what you have changed between versions.
>
> Here is an example:
>
> https://lore.kernel.org/lkml/20200124231834.63628-4-pmalani@chromium.org/
>
> Regards,
> Boqun
>
> > perf/ring_buffer: Add missing annotation to perf_output_end()
> > sched/fair: Add missing annotation for nohz_newidle_balance()
> > sched/deadline: Add missing annotation for dl_task_offline_migration()
> > fs_pin: Add missing annotation for pin_kill() declaration
> > fs_pin: Add missing annotation for pin_kill() definition
> > kasan: add missing annotation for start_report()
> > kasan: add missing annotation for end_report()
> > futex: Add missing annotation for futex_wait_queue_me()
> >
> > fs/fs_pin.c | 2 +-
> > include/linux/fs_pin.h | 2 +-
> > kernel/events/ring_buffer.c | 2 +-
> > kernel/futex.c | 3 +++
> > kernel/sched/deadline.c | 1 +
> > kernel/sched/fair.c | 2 +-
> > kernel/time/hrtimer.c | 1 +
> > mm/kasan/report.c | 4 ++--
> > 8 files changed, 11 insertions(+), 6 deletions(-)
> >
> > --
> > 2.24.1
> >
>
Thanks for the feedback, I take good notes. I am working on the
second version.
Kind regards,
Jules
Powered by blists - more mailing lists