[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <172140359332.12097.1558892281494577679.b4-ty@kernel.dk>
Date: Fri, 19 Jul 2024 09:39:53 -0600
From: Jens Axboe <axboe@...nel.dk>
To: Bart Van Assche <bvanassche@....org>,
Andrew Morton <akpm@...ux-foundation.org>, Ming Lei <ming.lei@...hat.com>,
Omar Sandoval <osandov@...com>, linux-kernel@...r.kernel.org,
linux-block@...r.kernel.org, Yang Yang <yang.yang@...o.com>
Subject: Re: [PATCH v8] sbitmap: fix io hung due to race on
sbitmap_word::cleared
On Tue, 16 Jul 2024 16:26:27 +0800, Yang Yang wrote:
> Configuration for sbq:
> depth=64, wake_batch=6, shift=6, map_nr=1
>
> 1. There are 64 requests in progress:
> map->word = 0xFFFFFFFFFFFFFFFF
> 2. After all the 64 requests complete, and no more requests come:
> map->word = 0xFFFFFFFFFFFFFFFF, map->cleared = 0xFFFFFFFFFFFFFFFF
> 3. Now two tasks try to allocate requests:
> T1: T2:
> __blk_mq_get_tag .
> __sbitmap_queue_get .
> sbitmap_get .
> sbitmap_find_bit .
> sbitmap_find_bit_in_word .
> __sbitmap_get_word -> nr=-1 __blk_mq_get_tag
> sbitmap_deferred_clear __sbitmap_queue_get
> /* map->cleared=0xFFFFFFFFFFFFFFFF */ sbitmap_find_bit
> if (!READ_ONCE(map->cleared)) sbitmap_find_bit_in_word
> return false; __sbitmap_get_word -> nr=-1
> mask = xchg(&map->cleared, 0) sbitmap_deferred_clear
> atomic_long_andnot() /* map->cleared=0 */
> if (!(map->cleared))
> return false;
> /*
> * map->cleared is cleared by T1
> * T2 fail to acquire the tag
> */
>
> [...]
Applied, thanks!
[1/1] sbitmap: fix io hung due to race on sbitmap_word::cleared
commit: 72d04bdcf3f7d7e07d82f9757946f68802a7270a
Best regards,
--
Jens Axboe
Powered by blists - more mailing lists