From: Damien Le Moal <Damien.LeMoal@wdc.com>
To: Johannes Thumshirn <Johannes.Thumshirn@wdc.com>,
"Martin K . Petersen" <martin.petersen@oracle.com>
Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com>,
"linux-scsi@vger.kernel.org" <linux-scsi@vger.kernel.org>
Subject: Re: [PATCH] scsi: sd_zbc: update write pointer offset cache
Date: Wed, 10 Mar 2021 22:45:31 +0000 [thread overview]
Message-ID: <BL0PR04MB651472FBB9FBD637C6701991E7919@BL0PR04MB6514.namprd04.prod.outlook.com> (raw)
In-Reply-To: 3cfebe48d09db73041b7849be71ffbcec7ee40b3.1615369586.git.johannes.thumshirn@wdc.com
On 2021/03/10 18:48, Johannes Thumshirn wrote:
> Recent changes changed the completion of SCSI commands from Soft-IRQ
> context to IRQ context. This triggers the following warning, when we're
> completing writes to zoned block devices that go through the zone append
> emulation:
>
> CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.12.0-rc2+ #2
> Hardware name: Supermicro Super Server/X10SRL-F, BIOS 2.0 12/17/2015
> RIP: 0010:__local_bh_disable_ip+0x3f/0x50
> RSP: 0018:ffff8883e1409ba8 EFLAGS: 00010006
> RAX: 0000000080010001 RBX: 0000000000000001 RCX: 0000000000000013
> RDX: ffff888129e4d200 RSI: 0000000000000201 RDI: ffffffff915b9dbd
> RBP: ffff888113e9a540 R08: ffff888113e9a540 R09: 00000000000077f0
> R10: 0000000000080000 R11: 0000000000000001 R12: ffff888129e4d200
> R13: 0000000000001000 R14: 00000000000077f0 R15: ffff888129e4d218
> FS: 0000000000000000(0000) GS:ffff8883e1400000(0000) knlGS:0000000000000000
> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00007f2f8418ebc0 CR3: 000000021202a006 CR4: 00000000001706f0
> Call Trace:
> <IRQ>
> _raw_spin_lock_bh+0x18/0x40
> sd_zbc_complete+0x43d/0x1150
> sd_done+0x631/0x1040
> ? mark_lock+0xe4/0x2fd0
> ? provisioning_mode_store+0x3f0/0x3f0
> scsi_finish_command+0x31b/0x5c0
> _scsih_io_done+0x960/0x29e0 [mpt3sas]
> ? mpt3sas_scsih_scsi_lookup_get+0x1c7/0x340 [mpt3sas]
> ? __lock_acquire+0x166b/0x58b0
> ? _get_st_from_smid+0x4a/0x80 [mpt3sas]
> _base_process_reply_queue+0x23f/0x26e0 [mpt3sas]
> ? lock_is_held_type+0x98/0x110
> ? find_held_lock+0x2c/0x110
> ? mpt3sas_base_sync_reply_irqs+0x360/0x360 [mpt3sas]
> _base_interrupt+0x8d/0xd0 [mpt3sas]
> ? rcu_read_lock_sched_held+0x3f/0x70
> __handle_irq_event_percpu+0x24d/0x600
> handle_irq_event+0xef/0x240
> ? handle_irq_event_percpu+0x110/0x110
> handle_edge_irq+0x1f6/0xb60
> __common_interrupt+0x75/0x160
> common_interrupt+0x7b/0xa0
> </IRQ>
> asm_common_interrupt+0x1e/0x40
>
> Don't use spin_lock_bh() to protect the update of the write pointer offset
> cache, but use spin_lock_irqsave() for it.
>
> Reported-by: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com>
> Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
> ---
> drivers/scsi/sd_zbc.c | 19 +++++++++++--------
> 1 file changed, 11 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c
> index ee558675eab4..994f1b8e3504 100644
> --- a/drivers/scsi/sd_zbc.c
> +++ b/drivers/scsi/sd_zbc.c
> @@ -280,27 +280,28 @@ static int sd_zbc_update_wp_offset_cb(struct blk_zone *zone, unsigned int idx,
> static void sd_zbc_update_wp_offset_workfn(struct work_struct *work)
> {
> struct scsi_disk *sdkp;
> + unsigned long flags;
> unsigned int zno;
> int ret;
>
> sdkp = container_of(work, struct scsi_disk, zone_wp_offset_work);
>
> - spin_lock_bh(&sdkp->zones_wp_offset_lock);
> + spin_lock_irqsave(&sdkp->zones_wp_offset_lock, flags);
> for (zno = 0; zno < sdkp->nr_zones; zno++) {
> if (sdkp->zones_wp_offset[zno] != SD_ZBC_UPDATING_WP_OFST)
> continue;
>
> - spin_unlock_bh(&sdkp->zones_wp_offset_lock);
> + spin_unlock_irqrestore(&sdkp->zones_wp_offset_lock, flags);
> ret = sd_zbc_do_report_zones(sdkp, sdkp->zone_wp_update_buf,
> SD_BUF_SIZE,
> zno * sdkp->zone_blocks, true);
> - spin_lock_bh(&sdkp->zones_wp_offset_lock);
> + spin_lock_irqsave(&sdkp->zones_wp_offset_lock, flags);
> if (!ret)
> sd_zbc_parse_report(sdkp, sdkp->zone_wp_update_buf + 64,
> zno, sd_zbc_update_wp_offset_cb,
> sdkp);
> }
> - spin_unlock_bh(&sdkp->zones_wp_offset_lock);
> + spin_unlock_irqrestore(&sdkp->zones_wp_offset_lock, flags);
>
> scsi_device_put(sdkp->device);
> }
> @@ -324,6 +325,7 @@ blk_status_t sd_zbc_prepare_zone_append(struct scsi_cmnd *cmd, sector_t *lba,
> struct request *rq = cmd->request;
> struct scsi_disk *sdkp = scsi_disk(rq->rq_disk);
> unsigned int wp_offset, zno = blk_rq_zone_no(rq);
> + unsigned long flags;
> blk_status_t ret;
>
> ret = sd_zbc_cmnd_checks(cmd);
> @@ -337,7 +339,7 @@ blk_status_t sd_zbc_prepare_zone_append(struct scsi_cmnd *cmd, sector_t *lba,
> if (!blk_req_zone_write_trylock(rq))
> return BLK_STS_ZONE_RESOURCE;
>
> - spin_lock_bh(&sdkp->zones_wp_offset_lock);
> + spin_lock_irqsave(&sdkp->zones_wp_offset_lock, flags);
> wp_offset = sdkp->zones_wp_offset[zno];
> switch (wp_offset) {
> case SD_ZBC_INVALID_WP_OFST:
> @@ -366,7 +368,7 @@ blk_status_t sd_zbc_prepare_zone_append(struct scsi_cmnd *cmd, sector_t *lba,
>
> *lba += wp_offset;
> }
> - spin_unlock_bh(&sdkp->zones_wp_offset_lock);
> + spin_unlock_irqrestore(&sdkp->zones_wp_offset_lock, flags);
> if (ret)
> blk_req_zone_write_unlock(rq);
> return ret;
> @@ -445,6 +447,7 @@ static unsigned int sd_zbc_zone_wp_update(struct scsi_cmnd *cmd,
> struct scsi_disk *sdkp = scsi_disk(rq->rq_disk);
> unsigned int zno = blk_rq_zone_no(rq);
> enum req_opf op = req_op(rq);
> + unsigned long flags;
>
> /*
> * If we got an error for a command that needs updating the write
> @@ -452,7 +455,7 @@ static unsigned int sd_zbc_zone_wp_update(struct scsi_cmnd *cmd,
> * invalid to force an update from disk the next time a zone append
> * command is issued.
> */
> - spin_lock_bh(&sdkp->zones_wp_offset_lock);
> + spin_lock_irqsave(&sdkp->zones_wp_offset_lock, flags);
>
> if (result && op != REQ_OP_ZONE_RESET_ALL) {
> if (op == REQ_OP_ZONE_APPEND) {
> @@ -496,7 +499,7 @@ static unsigned int sd_zbc_zone_wp_update(struct scsi_cmnd *cmd,
> }
>
> unlock_wp_offset:
> - spin_unlock_bh(&sdkp->zones_wp_offset_lock);
> + spin_unlock_irqrestore(&sdkp->zones_wp_offset_lock, flags);
>
> return good_bytes;
> }
>
Looks good to me.
Reviewed-by: Damien Le Moal <damien.lemoal@wdc.com>
--
Damien Le Moal
Western Digital Research
next prev parent reply other threads:[~2021-03-10 22:46 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-10 9:48 [PATCH] scsi: sd_zbc: update write pointer offset cache Johannes Thumshirn
2021-03-10 22:45 ` Damien Le Moal [this message]
2021-03-11 3:14 ` Shinichiro Kawasaki
2021-03-11 15:13 ` Bart Van Assche
2021-03-11 15:18 ` Johannes Thumshirn
2021-03-11 15:48 ` Bart Van Assche
2021-03-11 15:54 ` Johannes Thumshirn
2021-03-12 2:56 ` Bart Van Assche
2021-03-12 4:38 ` Shinichiro Kawasaki
2021-03-12 5:51 ` Johannes Thumshirn
2021-03-12 7:27 ` Damien Le Moal
2021-03-12 7:59 ` Johannes Thumshirn
2021-03-12 8:20 ` Damien Le Moal
2021-03-12 8:58 ` Johannes Thumshirn
2021-03-12 10:05 ` Shinichiro Kawasaki
2021-03-12 10:28 ` Johannes Thumshirn
2021-03-12 11:48 ` Shinichiro Kawasaki
2021-03-12 12:02 ` Johannes Thumshirn
2021-03-16 7:53 ` Johannes Thumshirn
2021-03-17 1:42 ` Martin K. Petersen
2021-03-17 7:45 ` Johannes Thumshirn
2021-03-19 2:21 ` Martin K. Petersen
2021-03-19 3:45 ` Martin K. Petersen
2021-03-19 8:15 ` Johannes Thumshirn
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=BL0PR04MB651472FBB9FBD637C6701991E7919@BL0PR04MB6514.namprd04.prod.outlook.com \
--to=damien.lemoal@wdc.com \
--cc=Johannes.Thumshirn@wdc.com \
--cc=linux-scsi@vger.kernel.org \
--cc=martin.petersen@oracle.com \
--cc=shinichiro.kawasaki@wdc.com \
/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;
as well as URLs for NNTP newsgroup(s).