From: ZiyangZhang <ZiyangZhang@linux.alibaba.com>
To: ming.lei@redhat.com, axboe@kernel.dk
Cc: xiaoguang.wang@linux.alibaba.com, linux-block@vger.kernel.org,
linux-kernel@vger.kernel.org, joseph.qi@linux.alibaba.com,
ZiyangZhang <ZiyangZhang@linux.alibaba.com>
Subject: [RFC PATCH 7/9] ublk_drv: define macros for recovery feature and check them
Date: Wed, 24 Aug 2022 13:47:42 +0800 [thread overview]
Message-ID: <20220824054744.77812-8-ZiyangZhang@linux.alibaba.com> (raw)
In-Reply-To: <20220824054744.77812-1-ZiyangZhang@linux.alibaba.com>
Define some macros for recovery feature. Especially define a new state:
UBLK_S_DEV_RECOVERING which implies the ublk_device is recovering.
UBLK_F_USER_RECOVERY implies that:
(1) ublk_drv enables recovery feature. It won't schedule monitor_work to
automatically abort rqs and release the device. Instead, it waits
for user's START_USER_RECOVERY ctrl-cmd.
(2) while re-initing a ubq, ublk_drv ends(aborts) rqs issued to
userspace(ublksrv) before crash.
(3) while re-initing a ubq, ublk_drv requeues rqs dispatched after crash.
UBLK_F_USER_RECOVERY_REISSUE implies that:
(1) everything UBLK_F_USER_RECOVERY implies except
(2) ublk_drv requeues rqs issued to userspace(ublksrv) before crash.
UBLK_F_USER_RECOVERY_REISSUE is designed for backends which:
(1) tolerate double-writes because we may issue the same rq twice.
(2) cannot let frontend users get I/O error, such as a RDONLY system.
For now, we do not allow STOP_DEV while we are in UBLK_S_DEV_RECOVERING.
This means that user must assign a new ubq_daemon for each ubq after
sending START_USER_RECOVERY ctrl-cmd.
Also modify checks on state in START_DEV and SET_PARAMS because now we
have three states.
Signed-off-by: ZiyangZhang <ZiyangZhang@linux.alibaba.com>
---
drivers/block/ublk_drv.c | 47 ++++++++++++++++++++++++++++++-----
include/uapi/linux/ublk_cmd.h | 7 ++++++
2 files changed, 48 insertions(+), 6 deletions(-)
diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c
index 4bbd97ccaedf..0ee871fa3f92 100644
--- a/drivers/block/ublk_drv.c
+++ b/drivers/block/ublk_drv.c
@@ -49,7 +49,9 @@
/* All UBLK_F_* have to be included into UBLK_F_ALL */
#define UBLK_F_ALL (UBLK_F_SUPPORT_ZERO_COPY \
| UBLK_F_URING_CMD_COMP_IN_TASK \
- | UBLK_F_NEED_GET_DATA)
+ | UBLK_F_NEED_GET_DATA \
+ | UBLK_F_USER_RECOVERY \
+ | UBLK_F_USER_RECOVERY_REISSUE)
/* All UBLK_PARAM_TYPE_* should be included here */
#define UBLK_PARAM_TYPE_ALL (UBLK_PARAM_TYPE_BASIC | UBLK_PARAM_TYPE_DISCARD)
@@ -322,6 +324,33 @@ static inline int ublk_queue_cmd_buf_size(struct ublk_device *ub, int q_id)
PAGE_SIZE);
}
+/*
+ * TODO: UBLK_F_USER_RECOVERY should be a flag for device, not for queue,
+ * since "some queues are aborted while others are recovered" is really weird.
+ */
+static inline bool ublk_can_use_recovery(struct ublk_device *ub)
+{
+ struct ublk_queue *ubq = ublk_get_queue(ub, 0);
+
+ if (ubq->flags & UBLK_F_USER_RECOVERY)
+ return true;
+ return false;
+}
+
+/*
+ * TODO: UBLK_F_USER_RECOVERY_REISSUE should be a flag for device, not for queue,
+ * since "some queues are aborted while others are recovered" is really weird.
+ */
+static inline bool ublk_can_use_recovery_reissue(struct ublk_device *ub)
+{
+ struct ublk_queue *ubq = ublk_get_queue(ub, 0);
+
+ if ((ubq->flags & UBLK_F_USER_RECOVERY) &&
+ (ubq->flags & UBLK_F_USER_RECOVERY_REISSUE))
+ return true;
+ return false;
+}
+
static void ublk_free_disk(struct gendisk *disk)
{
struct ublk_device *ub = disk->private_data;
@@ -1029,10 +1058,15 @@ static void ublk_stop_dev(struct ublk_device *ub)
{
mutex_lock(&ub->mutex);
/* No gendisk is live. ubq may be ready or not */
- if (ub->dev_info.state == UBLK_S_DEV_DEAD)
+ if (ub->dev_info.state == UBLK_S_DEV_DEAD) {
goto out_cancel_dev;
-
- mod_delayed_work(system_wq, &ub->monitor_work, 0);
+ /* TODO: support stop_dev just after start_recovery */
+ } else if (ub->dev_info.state == UBLK_S_DEV_RECOVERING) {
+ goto out_unlock;
+ /* schedule monitor_work to abort any dying queue */
+ } else {
+ mod_delayed_work(system_wq, &ub->monitor_work, 0);
+ }
pr_devel("%s: Wait for all requests ended...\n", __func__);
blk_mq_freeze_queue(ub->ub_disk->queue);
ub->dev_info.state = UBLK_S_DEV_DEAD;
@@ -1044,6 +1078,7 @@ static void ublk_stop_dev(struct ublk_device *ub)
ub->ub_disk = NULL;
out_cancel_dev:
ublk_cancel_dev(ub);
+ out_unlock:
mutex_unlock(&ub->mutex);
}
@@ -1403,7 +1438,7 @@ static int ublk_ctrl_start_dev(struct io_uring_cmd *cmd)
schedule_delayed_work(&ub->monitor_work, UBLK_DAEMON_MONITOR_PERIOD);
mutex_lock(&ub->mutex);
- if (ub->dev_info.state == UBLK_S_DEV_LIVE ||
+ if (ub->dev_info.state != UBLK_S_DEV_DEAD ||
test_bit(UB_STATE_USED, &ub->state)) {
ret = -EEXIST;
goto out_unlock;
@@ -1746,7 +1781,7 @@ static int ublk_ctrl_set_params(struct io_uring_cmd *cmd)
/* parameters can only be changed when device isn't live */
mutex_lock(&ub->mutex);
- if (ub->dev_info.state == UBLK_S_DEV_LIVE) {
+ if (ub->dev_info.state != UBLK_S_DEV_DEAD) {
ret = -EACCES;
} else if (copy_from_user(&ub->params, argp, ph.len)) {
ret = -EFAULT;
diff --git a/include/uapi/linux/ublk_cmd.h b/include/uapi/linux/ublk_cmd.h
index 677edaab2b66..7f7e6f44cec5 100644
--- a/include/uapi/linux/ublk_cmd.h
+++ b/include/uapi/linux/ublk_cmd.h
@@ -17,6 +17,8 @@
#define UBLK_CMD_STOP_DEV 0x07
#define UBLK_CMD_SET_PARAMS 0x08
#define UBLK_CMD_GET_PARAMS 0x09
+#define UBLK_CMD_START_USER_RECOVERY 0x10
+#define UBLK_CMD_END_USER_RECOVERY 0x11
/*
* IO commands, issued by ublk server, and handled by ublk driver.
@@ -74,9 +76,14 @@
*/
#define UBLK_F_NEED_GET_DATA (1UL << 2)
+#define UBLK_F_USER_RECOVERY (1UL << 3)
+
+#define UBLK_F_USER_RECOVERY_REISSUE (1UL << 4)
+
/* device state */
#define UBLK_S_DEV_DEAD 0
#define UBLK_S_DEV_LIVE 1
+#define UBLK_S_DEV_RECOVERING 2
/* shipped via sqe->cmd of io_uring command */
struct ublksrv_ctrl_cmd {
--
2.27.0
next prev parent reply other threads:[~2022-08-24 5:49 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-24 5:47 [RFC PATCH 0/9] ublk_drv: add USER_RECOVERY support ZiyangZhang
2022-08-24 5:47 ` [RFC PATCH 1/9] ublk_drv: check 'current' instead of 'ubq_daemon' ZiyangZhang
2022-08-29 2:13 ` Ming Lei
2022-08-24 5:47 ` [RFC PATCH 2/9] ublk_drv: refactor ublk_cancel_queue() ZiyangZhang
2022-08-29 3:01 ` Ming Lei
2022-08-29 4:50 ` Ziyang Zhang
2022-08-24 5:47 ` [RFC PATCH 3/9] ublk_drv: add a helper to get ioucmd from pdu ZiyangZhang
2022-08-29 3:06 ` Ming Lei
2022-08-29 4:59 ` Ziyang Zhang
2022-08-24 5:47 ` [RFC PATCH 4/9] ublk_drv: refactor __ublk_rq_task_work() and aborting machenism ZiyangZhang
2022-08-29 5:40 ` Ming Lei
2022-08-29 6:13 ` Ziyang Zhang
2022-08-29 8:11 ` Ming Lei
2022-08-29 9:09 ` Ziyang Zhang
2022-08-29 10:02 ` Ming Lei
2022-08-24 5:47 ` [RFC PATCH 5/9] ublk_drv: refactor ublk_stop_dev() ZiyangZhang
2022-08-24 5:47 ` [RFC PATCH 6/9] ublk_drv: add pr_devel() to prepare for recovery feature ZiyangZhang
2022-08-24 5:47 ` ZiyangZhang [this message]
2022-08-24 5:47 ` [RFC PATCH 8/9] ublk_drv: add START_USER_RECOVERY and END_USER_RECOVERY support ZiyangZhang
2022-08-24 5:47 ` [RFC PATCH 9/9] ublk_drv: do not schedule monitor_work with recovery feature enabled ZiyangZhang
2022-08-29 2:08 ` [RFC PATCH 0/9] ublk_drv: add USER_RECOVERY support Ming Lei
2022-08-29 4:00 ` Ziyang Zhang
2022-08-29 5:56 ` Ming Lei
2022-08-29 7:29 ` Ziyang Zhang
2022-08-29 8:38 ` Ming Lei
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=20220824054744.77812-8-ZiyangZhang@linux.alibaba.com \
--to=ziyangzhang@linux.alibaba.com \
--cc=axboe@kernel.dk \
--cc=joseph.qi@linux.alibaba.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=ming.lei@redhat.com \
--cc=xiaoguang.wang@linux.alibaba.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