From: "Yu Kuai" <yukuai@fnnas.com>
To: "Li Nan" <linan666@huaweicloud.com>, <song@kernel.org>,
<linux-raid@vger.kernel.org>
Cc: <linux-kernel@vger.kernel.org>, <filippo@debian.org>,
<colyli@fnnas.com>, <yukuai@fnnas.com>
Subject: Re: [PATCH v2 04/11] md/raid5: use mempool to allocate stripe_request_ctx
Date: Sat, 3 Jan 2026 19:12:33 +0800 [thread overview]
Message-ID: <5460e6b3-9b23-49e9-a52b-57e1e6d2490a@fnnas.com> (raw)
In-Reply-To: <343e3081-7d24-f783-b040-2f61aad4ea4f@huaweicloud.com>
Hi,
在 2025/12/26 16:33, Li Nan 写道:
>
>
> 在 2025/11/24 14:31, Yu Kuai 写道:
>> On the one hand, stripe_request_ctx is 72 bytes, and it's a bit huge for
>> a stack variable.
>>
>> On the other hand, the bitmap sectors_to_do is a fixed size, result in
>> max_hw_sector_kb of raid5 array is at most 256 * 4k = 1Mb, and this will
>> make full stripe IO impossible for the array that chunk_size *
>> data_disks
>> is bigger. Allocate ctx during runtime will make it possible to get rid
>> of this limit.
>>
>> Signed-off-by: Yu Kuai <yukuai@fnnas.com>
>> ---
>> drivers/md/md.h | 4 +++
>> drivers/md/raid1-10.c | 5 ----
>> drivers/md/raid5.c | 61 +++++++++++++++++++++++++++----------------
>> drivers/md/raid5.h | 2 ++
>> 4 files changed, 45 insertions(+), 27 deletions(-)
>>
>
> [...]
>
>> @@ -7374,6 +7380,10 @@ static void free_conf(struct r5conf *conf)
>> bioset_exit(&conf->bio_split);
>> kfree(conf->stripe_hashtbl);
>> kfree(conf->pending_data);
>> +
>> + if (conf->ctx_pool)
>> + mempool_destroy(conf->ctx_pool);
>> +
>> kfree(conf);
>> }
>> @@ -8057,6 +8067,13 @@ static int raid5_run(struct mddev *mddev)
>> goto abort;
>> }
>> + conf->ctx_pool = mempool_create_kmalloc_pool(NR_RAID_BIOS,
>> + sizeof(struct stripe_request_ctx));
>> + if (!conf->ctx_pool) {
>> + ret = -ENOMEM;
>> + goto abort;
>> + }
>> +
>
> What about moving create to setup_conf()? If so, call destroy in
> free_conf() without checks.
No, we can't, this must be done after raid5_set_limits(), which is called
at the end of raid5_run().
>
>> if (log_init(conf, journal_dev, raid5_has_ppl(conf)))
>> goto abort;
>> diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
>> index eafc6e9ed6ee..6e3f07119fa4 100644
>> --- a/drivers/md/raid5.h
>> +++ b/drivers/md/raid5.h
>> @@ -690,6 +690,8 @@ struct r5conf {
>> struct list_head pending_list;
>> int pending_data_cnt;
>> struct r5pending_data *next_pending_data;
>> +
>> + mempool_t *ctx_pool;
>> };
>> #if PAGE_SIZE == DEFAULT_STRIPE_SIZE
>
--
Thansk,
Kuai
next prev parent reply other threads:[~2026-01-03 11:12 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-11-24 6:31 [PATCH v2 00/11] md: align bio to io_opt and fix abnormal io_opt Yu Kuai
2025-11-24 6:31 ` [PATCH v2 01/11] md: merge mddev has_superblock into mddev_flags Yu Kuai
2025-12-26 3:04 ` Li Nan
2025-11-24 6:31 ` [PATCH v2 02/11] md: merge mddev faillast_dev " Yu Kuai
2025-12-26 3:46 ` Li Nan
2025-11-24 6:31 ` [PATCH v2 03/11] md: merge mddev serialize_policy " Yu Kuai
2025-12-26 6:33 ` Li Nan
2025-11-24 6:31 ` [PATCH v2 04/11] md/raid5: use mempool to allocate stripe_request_ctx Yu Kuai
2025-12-26 8:33 ` Li Nan
2026-01-03 11:12 ` Yu Kuai [this message]
2025-12-30 9:38 ` Li Nan
2026-01-03 11:13 ` Yu Kuai
2025-11-24 6:31 ` [PATCH v2 05/11] md/raid5: make sure max_sectors is not less than io_opt Yu Kuai
2025-11-24 6:31 ` [PATCH v2 06/11] md: support to align bio to limits Yu Kuai
2025-11-27 0:51 ` kernel test robot
2025-11-30 2:38 ` Yu Kuai
2025-11-27 7:05 ` kernel test robot
2025-11-24 6:31 ` [PATCH v2 07/11] md: add a helper md_config_align_limits() Yu Kuai
2025-11-24 6:32 ` [PATCH v2 08/11] md/raid5: align bio to io_opt Yu Kuai
2025-11-24 6:32 ` [PATCH v2 09/11] md/raid10: " Yu Kuai
2025-11-24 6:32 ` [PATCH v2 10/11] md/raid0: " Yu Kuai
2025-11-24 6:32 ` [PATCH v2 11/11] md: fix abnormal io_opt from member disks Yu Kuai
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=5460e6b3-9b23-49e9-a52b-57e1e6d2490a@fnnas.com \
--to=yukuai@fnnas.com \
--cc=colyli@fnnas.com \
--cc=filippo@debian.org \
--cc=linan666@huaweicloud.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-raid@vger.kernel.org \
--cc=song@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®