FAILED: patch "[PATCH] null_blk: Fix zone reset all tracing" failed to apply to 5.9-stable tree

PATCH In review [email protected] 2 participants 15 messages Nov 3, 2020 – Nov 4, 2020

Subsystems (2)

Stable branch Block layer

lore

Share & tools

lore

[email protected]

author maintainer

2020-11-03 14:24 UTC #0 lore raw

The patch below does not apply to the 5.9-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From f9c9104288da543cd64f186f9e2fba389f415630 Mon Sep 17 00:00:00 2001
From: Damien Le Moal <[email protected]>
Date: Thu, 29 Oct 2020 20:04:59 +0900
Subject: [PATCH] null_blk: Fix zone reset all tracing

In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.

Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.

Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")

Signed-off-by Damien Le Moal <[email protected]>

Cc [email protected]

Signed-off-by Jens Axboe <[email protected]>

drivers/block/null_blk_zoned.c +8 -3

index 98056c88926b..b637b16a5f54 100644
@@ -475,9 +475,14 @@ static blk_status_t null_zone_mgmt(struct nullb_cmd *cmd, enum req_opf op,
 
 	switch (op) {
 	case REQ_OP_ZONE_RESET_ALL:
-		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++)
-			null_reset_zone(dev, &dev->zones[i]);
-		break;
+		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++) {
+			zone = &dev->zones[i];
+			if (zone->cond != BLK_ZONE_COND_EMPTY) {
+				null_reset_zone(dev, zone);
+				trace_nullb_zone_op(cmd, i, zone->cond);
+			}
+		}
+		return BLK_STS_OK;
 	case REQ_OP_ZONE_RESET:
 		ret = null_reset_zone(dev, zone);
 		break;

Damien Le Moal

2020-11-03 23:37 UTC #1 lore raw

Quoted 5 lines · On 2020/11/03 23:23, [email protected] wrote:

The patch below does not apply to the 5.9-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

I will send a backported patch shortly. Thanks.

Quoted 50 lines

thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From f9c9104288da543cd64f186f9e2fba389f415630 Mon Sep 17 00:00:00 2001
From: Damien Le Moal <[email protected]>
Date: Thu, 29 Oct 2020 20:04:59 +0900
Subject: [PATCH] null_blk: Fix zone reset all tracing
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.
Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
drivers/block/null_blk_zoned.c +8 -3
index 98056c88926b..b637b16a5f54 100644
@@ -475,9 +475,14 @@ static blk_status_t null_zone_mgmt(struct nullb_cmd *cmd, enum req_opf op,
 
 	switch (op) {
 	case REQ_OP_ZONE_RESET_ALL:
-		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++)
-			null_reset_zone(dev, &dev->zones[i]);
-		break;
+		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++) {
+			zone = &dev->zones[i];
+			if (zone->cond != BLK_ZONE_COND_EMPTY) {
+				null_reset_zone(dev, zone);
+				trace_nullb_zone_op(cmd, i, zone->cond);
+			}
+		}
+		return BLK_STS_OK;
 	case REQ_OP_ZONE_RESET:
 		ret = null_reset_zone(dev, zone);
 		break;

Signature · 2 lines

Damien Le Moal
Western Digital Research

Damien Le Moal

2020-11-04 05:29 UTC #2 lore raw

PATCH 1/1 null_blk: Fix zone reset all tracing

In reply to [email protected]

commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.

Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.

Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")

Signed-off-by Damien Le Moal <[email protected]>

Cc [email protected]

Signed-off-by Jens Axboe <[email protected]>

---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

drivers/block/null_blk_zoned.c +8 -6

index 3d25c9ad2383..fcbbe2e4ef2d 100644
@@ -226,13 +226,15 @@ static blk_status_t null_zone_mgmt(struct nullb_cmd *cmd, enum req_opf op,
 
 	switch (op) {
 	case REQ_OP_ZONE_RESET_ALL:
-		for (i = 0; i < dev->nr_zones; i++) {
-			if (zone[i].type == BLK_ZONE_TYPE_CONVENTIONAL)
-				continue;
-			zone[i].cond = BLK_ZONE_COND_EMPTY;
-			zone[i].wp = zone[i].start;
+		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++) {
+			zone = &dev->zones[i];
+			if (zone->cond != BLK_ZONE_COND_EMPTY) {
+				zone->cond = BLK_ZONE_COND_EMPTY;
+				zone->wp = zone->start;
+				trace_nullb_zone_op(cmd, i, zone->cond);
+			}
 		}
-		break;
+		return BLK_STS_OK;
 	case REQ_OP_ZONE_RESET:
 		if (zone->type == BLK_ZONE_TYPE_CONVENTIONAL)
 			return BLK_STS_IOERR;

Signature · 1 line

2.26.2

Greg Kroah-Hartman

author maintainer

2020-11-04 09:09 UTC #3 lore raw

Re: [PATCH] null_blk: Fix zone reset all tracing

Quoted 11 lines · Damien Le Moal

commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.

Quoted 7 lines

Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

Now queued up, thanks.

greg k-h

Greg Kroah-Hartman

author maintainer

2020-11-04 09:14 UTC #4 lore raw

Re: [PATCH] null_blk: Fix zone reset all tracing

Quoted 20 lines · Greg Kroah-Hartman

Quoted 11 lines · Damien Le Moal
commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.
Quoted 7 lines
Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

Quoted 1 line

Now queued up, thanks.

Wait, no, I'll delay this one until the next round as it's not fixing something introduced in this -rc series.

thanks,

greg k-h

Damien Le Moal

2020-11-04 09:21 UTC #5 lore raw

Re: [PATCH] null_blk: Fix zone reset all tracing

Quoted 21 lines · Greg Kroah-Hartman

Quoted 20 lines · Greg Kroah-Hartman
Quoted 18 lines · Damien Le Moal
commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.
Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

Quoted 4 lines

Quoted 1 line
Now queued up, thanks.
Wait, no, I'll delay this one until the next round as it's not fixing something introduced in this -rc series.

Yes, that problem is older.
The lock fix I sent goes on top of this one though. I can send the backport for the lock fix without this patch applied. Is that OK ?

Quoted 5 lines

thanks,
greg k-h

Signature · 2 lines

Damien Le Moal
Western Digital Research

Greg Kroah-Hartman

author maintainer

2020-11-04 09:26 UTC #6 lore raw

Re: [PATCH] null_blk: Fix zone reset all tracing

Quoted 22 lines · Damien Le Moal

Quoted 21 lines · Greg Kroah-Hartman
Quoted 20 lines · Greg Kroah-Hartman
Quoted 18 lines · Damien Le Moal
commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.
Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

Quoted 8 lines

Quoted 4 lines
Quoted 1 line
Now queued up, thanks.
Wait, no, I'll delay this one until the next round as it's not fixing something introduced in this -rc series.
Yes, that problem is older.
The lock fix I sent goes on top of this one though. I can send the backport for the lock fix without this patch applied. Is that OK ?

If the order of the patches is needed, then yes, I can take both, please submit them as a patch series so that I know this is needed.

thanks,

greg k-h

Damien Le Moal

2020-11-04 09:31 UTC #7 lore raw

Re: [PATCH] null_blk: Fix zone reset all tracing

Quoted 28 lines · Greg Kroah-Hartman

Quoted 27 lines · Damien Le Moal
Quoted 25 lines · Greg Kroah-Hartman
Quoted 20 lines · Greg Kroah-Hartman
Quoted 11 lines · Damien Le Moal
commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.
Quoted 7 lines
Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
Quoted 1 line
Now queued up, thanks.
Wait, no, I'll delay this one until the next round as it's not fixing something introduced in this -rc series.

Quoted 6 lines

Quoted 3 lines
Yes, that problem is older.
The lock fix I sent goes on top of this one though. I can send the backport for the lock fix without this patch applied. Is that OK ?
If the order of the patches is needed, then yes, I can take both, please submit them as a patch series so that I know this is needed.

OK. Sending that. Note that I still do not see Kanchan patch applied in stable 5.9.y branch, so I will do the backport assuming it is applied. Or I can send all 3 patches as the series. Which do you prefer ?

Quoted 5 lines

thanks,
greg k-h

Signature · 2 lines

Damien Le Moal
Western Digital Research

Greg Kroah-Hartman

author maintainer

2020-11-04 09:50 UTC #8 lore raw

Re: [PATCH] null_blk: Fix zone reset all tracing

Quoted 33 lines · Damien Le Moal

Quoted 32 lines · Greg Kroah-Hartman
Quoted 30 lines · Damien Le Moal
Quoted 21 lines · Greg Kroah-Hartman
Quoted 20 lines · Greg Kroah-Hartman
Quoted 18 lines · Damien Le Moal
commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.
Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
Quoted 4 lines
Quoted 1 line
Now queued up, thanks.
Wait, no, I'll delay this one until the next round as it's not fixing something introduced in this -rc series.
Yes, that problem is older.
The lock fix I sent goes on top of this one though. I can send the backport for the lock fix without this patch applied. Is that OK ?

Quoted 6 lines

Quoted 2 lines
If the order of the patches is needed, then yes, I can take both, please submit them as a patch series so that I know this is needed.
OK. Sending that. Note that I still do not see Kanchan patch applied in stable 5.9.y branch, so I will do the backport assuming it is applied. Or I can send all 3 patches as the series. Which do you prefer ?

All 3 is great, to ensure I have them all as I don't know what you mean by "Kanchan patch".

thanks,

greg k-h

Damien Le Moal

2020-11-04 09:54 UTC #9 lore raw

Re: [PATCH] null_blk: Fix zone reset all tracing

Quoted 37 lines · Greg Kroah-Hartman

Quoted 36 lines · Damien Le Moal
Quoted 34 lines · Greg Kroah-Hartman
Quoted 22 lines · Damien Le Moal
Quoted 21 lines · Greg Kroah-Hartman
Quoted 20 lines · Greg Kroah-Hartman
Quoted 18 lines · Damien Le Moal
commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.
In the cae of the REQ_OP_ZONE_RESET_ALL operation, the command sector is ignored and the operation is applied to all sequential zones. For these commands, tracing the effect of the command using the command sector to determine the target zone is thus incorrect.
Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.
Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")
Signed-off-by Damien Le Moal <[email protected]>
Cc [email protected]
Signed-off-by Jens Axboe <[email protected]>
---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Now queued up, thanks.
Wait, no, I'll delay this one until the next round as it's not fixing something introduced in this -rc series.
Yes, that problem is older.
The lock fix I sent goes on top of this one though. I can send the backport for the lock fix without this patch applied. Is that OK ?
If the order of the patches is needed, then yes, I can take both, please submit them as a patch series so that I know this is needed.

Quoted 6 lines

Quoted 3 lines
OK. Sending that. Note that I still do not see Kanchan patch applied in stable 5.9.y branch, so I will do the backport assuming it is applied. Or I can send all 3 patches as the series. Which do you prefer ?
All 3 is great, to ensure I have them all as I don't know what you mean by "Kanchan patch".

I was talking about "
commit 35bc10b2eafbb701064b94f283b77c54d3304842 upstream." that you already applied to 5.9.y.

Sending all 3 patches backported in a series to be sure. SInce the first patchin the series will be the above mentioned patch, everything should still apply cleanly on your side. Thanks!

Quoted 5 lines

thanks,
greg k-h

Signature · 2 lines

Damien Le Moal
Western Digital Research

Damien Le Moal

2020-11-04 11:27 UTC #10 lore raw

cover null_blk fixes for 5.9 stable

In reply to Greg Kroah-Hartman

Greg,

Here are three backported patches from this cycle for 5.9 stable. The upstream patches are:
1) 35bc10b2eafbb701064b94f283b77c54d3304842
2) f9c9104288da543cd64f186f9e2fba389f415630
3) aa1c09cb65e2ed17cb8e652bc7ec84e0af1229eb

Thanks.

Damien Le Moal (2):
null_blk: Fix zone reset all tracing
null_blk: Fix locking in zoned mode

Kanchan Joshi (1):
null_blk: synchronization fix for zoned device

 drivers/block/null_blk.h       |   1 +
 drivers/block/null_blk_zoned.c | 157 ++++++++++++++++++++++++---------
 2 files changed, 118 insertions(+), 40 deletions(-)

Signature · 1 line

2.26.2

Damien Le Moal

2020-11-04 11:27 UTC #11 lore raw

PATCH 1/3 null_blk: synchronization fix for zoned device

From: Kanchan Joshi <[email protected]>

commit 35bc10b2eafbb701064b94f283b77c54d3304842 upstream.

Parallel write,read,zone-mgmt operations accessing/altering zone state and write-pointer may get into race. Avoid the situation by using a new spinlock for zoned device.
Concurrent zone-appends (on a zone) returning same write-pointer issue is also avoided using this lock.

Cc [email protected]

Fixes e0489ed5daeb ("null_blk: Support REQ_OP_ZONE_APPEND")

Signed-off-by Kanchan Joshi <[email protected]>

Reviewed-by Damien Le Moal <[email protected]>

Signed-off-by Jens Axboe <[email protected]>

---
 drivers/block/null_blk.h       |  1 +
 drivers/block/null_blk_zoned.c | 22 ++++++++++++++++++----
 2 files changed, 19 insertions(+), 4 deletions(-)

drivers/block/null_blk.h +1 -0

index daed4a9c3436..28099be50395 100644
@@ -44,6 +44,7 @@ struct nullb_device {
 	unsigned int nr_zones;
 	struct blk_zone *zones;
 	sector_t zone_size_sects;
+	spinlock_t zone_lock;
 
 	unsigned long size; /* device size in MB */
 	unsigned long completion_nsec; /* time in ns to complete a request */

drivers/block/null_blk_zoned.c +18 -4

index 3d25c9ad2383..e8d8b13aaa5a 100644
@@ -45,6 +45,7 @@ int null_init_zoned_dev(struct nullb_device *dev, struct request_queue *q)
 	if (!dev->zones)
 		return -ENOMEM;
 
+	spin_lock_init(&dev->zone_lock);
 	if (dev->zone_nr_conv >= dev->nr_zones) {
 		dev->zone_nr_conv = dev->nr_zones - 1;
 		pr_info("changed the number of conventional zones to %u",

@@ -131,8 +132,11 @@ int null_report_zones(struct gendisk *disk, sector_t sector,
 		 * So use a local copy to avoid corruption of the device zone
 		 * array.
 		 */
+		spin_lock_irq(&dev->zone_lock);
 		memcpy(&zone, &dev->zones[first_zone + i],
 		       sizeof(struct blk_zone));
+		spin_unlock_irq(&dev->zone_lock);
+
 		error = cb(&zone, i, data);
 		if (error)
 			return error;

@@ -277,18 +281,28 @@ static blk_status_t null_zone_mgmt(struct nullb_cmd *cmd, enum req_opf op,
 blk_status_t null_process_zoned_cmd(struct nullb_cmd *cmd, enum req_opf op,
 				    sector_t sector, sector_t nr_sectors)
 {
+	blk_status_t sts;
+	struct nullb_device *dev = cmd->nq->dev;
+
+	spin_lock_irq(&dev->zone_lock);
 	switch (op) {
 	case REQ_OP_WRITE:
-		return null_zone_write(cmd, sector, nr_sectors, false);
+		sts = null_zone_write(cmd, sector, nr_sectors, false);
+		break;
 	case REQ_OP_ZONE_APPEND:
-		return null_zone_write(cmd, sector, nr_sectors, true);
+		sts = null_zone_write(cmd, sector, nr_sectors, true);
+		break;
 	case REQ_OP_ZONE_RESET:
 	case REQ_OP_ZONE_RESET_ALL:
 	case REQ_OP_ZONE_OPEN:
 	case REQ_OP_ZONE_CLOSE:
 	case REQ_OP_ZONE_FINISH:
-		return null_zone_mgmt(cmd, op, sector);
+		sts = null_zone_mgmt(cmd, op, sector);
+		break;
 	default:
-		return null_process_cmd(cmd, op, sector, nr_sectors);
+		sts = null_process_cmd(cmd, op, sector, nr_sectors);
 	}
+	spin_unlock_irq(&dev->zone_lock);
+
+	return sts;
 }

Signature · 1 line

2.26.2

Damien Le Moal

2020-11-04 11:27 UTC #12 lore raw

PATCH 2/3 null_blk: Fix zone reset all tracing

In reply to Damien Le Moal

commit f9c9104288da543cd64f186f9e2fba389f415630 upstream.

Fix null_zone_mgmt() zone condition tracing in the case of REQ_OP_ZONE_RESET_ALL to apply tracing to all sequential zones that are not already empty.

Fixes 766c3297d7e1 ("null_blk: add trace in null_blk_zoned.c")

Signed-off-by Damien Le Moal <[email protected]>

Cc [email protected]

Signed-off-by Jens Axboe <[email protected]>

---
 drivers/block/null_blk_zoned.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

drivers/block/null_blk_zoned.c +8 -6

index e8d8b13aaa5a..0342696db0ce 100644
@@ -230,13 +230,15 @@ static blk_status_t null_zone_mgmt(struct nullb_cmd *cmd, enum req_opf op,
 
 	switch (op) {
 	case REQ_OP_ZONE_RESET_ALL:
-		for (i = 0; i < dev->nr_zones; i++) {
-			if (zone[i].type == BLK_ZONE_TYPE_CONVENTIONAL)
-				continue;
-			zone[i].cond = BLK_ZONE_COND_EMPTY;
-			zone[i].wp = zone[i].start;
+		for (i = dev->zone_nr_conv; i < dev->nr_zones; i++) {
+			zone = &dev->zones[i];
+			if (zone->cond != BLK_ZONE_COND_EMPTY) {
+				zone->cond = BLK_ZONE_COND_EMPTY;
+				zone->wp = zone->start;
+				trace_nullb_zone_op(cmd, i, zone->cond);
+			}
 		}
-		break;
+		return BLK_STS_OK;
 	case REQ_OP_ZONE_RESET:
 		if (zone->type == BLK_ZONE_TYPE_CONVENTIONAL)
 			return BLK_STS_IOERR;

Signature · 1 line

2.26.2

Damien Le Moal

2020-11-04 11:27 UTC #13 lore raw

PATCH 3/3 null_blk: Fix locking in zoned mode

In reply to Damien Le Moal

commit aa1c09cb65e2ed17cb8e652bc7ec84e0af1229eb upstream.

When the zoned mode is enabled in null_blk, Serializing read, write and zone management operations for each zone is necessary to protect device level information for managing zone resources (zone open and closed counters) as well as each zone condition and write pointer position. Commit 35bc10b2eafb ("null_blk: synchronization fix for zoned device") introduced a spinlock to implement this serialization. However, when memory backing is also enabled, GFP_NOIO memory allocations are executed under the spinlock, resulting in might_sleep() warnings. Furthermore, the zone_lock spinlock is locked/unlocked using spin_lock_irq/spin_unlock_irq, similarly to the memory backing code with the nullb->lock spinlock. This nested use of irq locks wrecks the irq enabled/disabled state.

Fix all this by introducing a bitmap for per-zone lock, with locking implemented using wait_on_bit_lock_io() and clear_and_wake_up_bit(). This locking mechanism allows keeping a zone locked while executing null_process_cmd(), serializing all operations to the zone while allowing to sleep during memory backing allocation with GFP_NOIO. Device level zone resource management information is protected using a spinlock which is not held while executing null_process_cmd();

Fixes 35bc10b2eafb ("null_blk: synchronization fix for zoned device")

Signed-off-by Damien Le Moal <[email protected]>

Signed-off-by Jens Axboe <[email protected]>

---
 drivers/block/null_blk.h       |   2 +-
 drivers/block/null_blk_zoned.c | 133 ++++++++++++++++++++++++---------
 2 files changed, 98 insertions(+), 37 deletions(-)

drivers/block/null_blk.h +1 -1

index 28099be50395..206309ecc7e4 100644
@@ -44,7 +44,7 @@ struct nullb_device {
 	unsigned int nr_zones;
 	struct blk_zone *zones;
 	sector_t zone_size_sects;
-	spinlock_t zone_lock;
+	unsigned long *zone_locks;
 
 	unsigned long size; /* device size in MB */
 	unsigned long completion_nsec; /* time in ns to complete a request */

drivers/block/null_blk_zoned.c +97 ~~-36~~

261 lines not on this page · raw message

Signature · 1 line

2.26.2

Greg Kroah-Hartman

author maintainer

2020-11-04 12:23 UTC #14 lore raw

Re: [PATCH 0/3] null_blk fixes for 5.9 stable

In reply to Damien Le Moal

Quoted 14 lines · Damien Le Moal

Greg,
Here are three backported patches from this cycle for 5.9 stable. The upstream patches are:
1) 35bc10b2eafbb701064b94f283b77c54d3304842
2) f9c9104288da543cd64f186f9e2fba389f415630
3) aa1c09cb65e2ed17cb8e652bc7ec84e0af1229eb
Thanks.
Damien Le Moal (2):
null_blk: Fix zone reset all tracing
null_blk: Fix locking in zoned mode

Quoted 6 lines

Kanchan Joshi (1):
null_blk: synchronization fix for zoned device

 drivers/block/null_blk.h       |   1 +
 drivers/block/null_blk_zoned.c | 157 ++++++++++++++++++++++++---------
 2 files changed, 118 insertions(+), 40 deletions(-)

All now queued up, thanks.

greg k-h