[PATCH] null_blk: Support configuring the maximum segment size

[PATCH] null_blk: Support configuring the maximum segment size

[Bart Van Assche]

Add support for configuring the maximum segment size.

Add support for segments smaller than the page size.

This patch enables testing segments smaller than the page size with a
driver that does not call blk_rq_map_sg().

Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Cc: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
---
 drivers/block/null_blk/main.c     | 19 ++++++++++++++++---
 drivers/block/null_blk/null_blk.h |  1 +
 2 files changed, 17 insertions(+), 3 deletions(-)

diff --git a/drivers/block/null_blk/main.c b/drivers/block/null_blk/main.c
index 4c601ca9552a..06eaa7ff4a86 100644
--- a/drivers/block/null_blk/main.c
+++ b/drivers/block/null_blk/main.c
@@ -157,6 +157,10 @@ static int g_max_sectors;
 module_param_named(max_sectors, g_max_sectors, int, 0444);
 MODULE_PARM_DESC(max_sectors, "Maximum size of a command (in 512B sectors)");
 
+static unsigned int g_max_segment_size = BLK_MAX_SEGMENT_SIZE;
+module_param_named(max_segment_size, g_max_segment_size, int, 0444);
+MODULE_PARM_DESC(max_segment_size, "Maximum size of a segment in bytes");
+
 static unsigned int nr_devices = 1;
 module_param(nr_devices, uint, 0444);
 MODULE_PARM_DESC(nr_devices, "Number of devices to register");
@@ -409,6 +413,7 @@ NULLB_DEVICE_ATTR(home_node, uint, NULL);
 NULLB_DEVICE_ATTR(queue_mode, uint, NULL);
 NULLB_DEVICE_ATTR(blocksize, uint, NULL);
 NULLB_DEVICE_ATTR(max_sectors, uint, NULL);
+NULLB_DEVICE_ATTR(max_segment_size, uint, NULL);
 NULLB_DEVICE_ATTR(irqmode, uint, NULL);
 NULLB_DEVICE_ATTR(hw_queue_depth, uint, NULL);
 NULLB_DEVICE_ATTR(index, uint, NULL);
@@ -550,6 +555,7 @@ static struct configfs_attribute *nullb_device_attrs[] = {
 	&nullb_device_attr_queue_mode,
 	&nullb_device_attr_blocksize,
 	&nullb_device_attr_max_sectors,
+	&nullb_device_attr_max_segment_size,
 	&nullb_device_attr_irqmode,
 	&nullb_device_attr_hw_queue_depth,
 	&nullb_device_attr_index,
@@ -630,7 +636,8 @@ static ssize_t memb_group_features_show(struct config_item *item, char *page)
 	return snprintf(page, PAGE_SIZE,
 			"badblocks,blocking,blocksize,cache_size,"
 			"completion_nsec,discard,home_node,hw_queue_depth,"
-			"irqmode,max_sectors,mbps,memory_backed,no_sched,"
+			"irqmode,max_sectors,max_segment_size,mbps,"
+			"memory_backed,no_sched,"
 			"poll_queues,power,queue_mode,shared_tag_bitmap,size,"
 			"submit_queues,use_per_node_hctx,virt_boundary,zoned,"
 			"zone_capacity,zone_max_active,zone_max_open,"
@@ -693,6 +700,7 @@ static struct nullb_device *null_alloc_dev(void)
 	dev->queue_mode = g_queue_mode;
 	dev->blocksize = g_bs;
 	dev->max_sectors = g_max_sectors;
+	dev->max_segment_size = g_max_segment_size;
 	dev->irqmode = g_irqmode;
 	dev->hw_queue_depth = g_hw_queue_depth;
 	dev->blocking = g_blocking;
@@ -1234,6 +1242,8 @@ static int null_transfer(struct nullb *nullb, struct page *page,
 	unsigned int valid_len = len;
 	int err = 0;
 
+	WARN_ONCE(len > dev->max_segment_size, "%u > %u\n", len,
+		  dev->max_segment_size);
 	if (!is_write) {
 		if (dev->zoned)
 			valid_len = null_zone_valid_read_len(nullb,
@@ -1269,7 +1279,8 @@ static int null_handle_rq(struct nullb_cmd *cmd)
 
 	spin_lock_irq(&nullb->lock);
 	rq_for_each_segment(bvec, rq, iter) {
-		len = bvec.bv_len;
+		len = min(bvec.bv_len, nullb->dev->max_segment_size);
+		bvec.bv_len = len;
 		err = null_transfer(nullb, bvec.bv_page, len, bvec.bv_offset,
 				     op_is_write(req_op(rq)), sector,
 				     rq->cmd_flags & REQ_FUA);
@@ -1296,7 +1307,8 @@ static int null_handle_bio(struct nullb_cmd *cmd)
 
 	spin_lock_irq(&nullb->lock);
 	bio_for_each_segment(bvec, bio, iter) {
-		len = bvec.bv_len;
+		len = min(bvec.bv_len, nullb->dev->max_segment_size);
+		bvec.bv_len = len;
 		err = null_transfer(nullb, bvec.bv_page, len, bvec.bv_offset,
 				     op_is_write(bio_op(bio)), sector,
 				     bio->bi_opf & REQ_FUA);
@@ -2125,6 +2137,7 @@ static int null_add_dev(struct nullb_device *dev)
 		dev->max_sectors = queue_max_hw_sectors(nullb->q);
 	dev->max_sectors = min(dev->max_sectors, BLK_DEF_MAX_SECTORS);
 	blk_queue_max_hw_sectors(nullb->q, dev->max_sectors);
+	blk_queue_max_segment_size(nullb->q, dev->max_segment_size);
 
 	if (dev->virt_boundary)
 		blk_queue_virt_boundary(nullb->q, PAGE_SIZE - 1);
diff --git a/drivers/block/null_blk/null_blk.h b/drivers/block/null_blk/null_blk.h
index eb5972c50be8..8cb73fe05fa3 100644
--- a/drivers/block/null_blk/null_blk.h
+++ b/drivers/block/null_blk/null_blk.h
@@ -102,6 +102,7 @@ struct nullb_device {
 	unsigned int queue_mode; /* block interface */
 	unsigned int blocksize; /* block size */
 	unsigned int max_sectors; /* Max sectors per command */
+	unsigned int max_segment_size; /* Max size of a single DMA segment. */
 	unsigned int irqmode; /* IRQ completion handler */
 	unsigned int hw_queue_depth; /* queue depth */
 	unsigned int index; /* index of the disk, only valid with a disk */

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Damien Le Moal]

On 1/28/23 09:59, Bart Van Assche wrote:
> Add support for configuring the maximum segment size.
> 
> Add support for segments smaller than the page size.
> 
> This patch enables testing segments smaller than the page size with a
> driver that does not call blk_rq_map_sg().
> 
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: Ming Lei <ming.lei@redhat.com>
> Cc: Damien Le Moal <damien.lemoal@opensource.wdc.com>
> Cc: Chaitanya Kulkarni <kch@nvidia.com>
> Signed-off-by: Bart Van Assche <bvanassche@acm.org>
> ---
>  drivers/block/null_blk/main.c     | 19 ++++++++++++++++---
>  drivers/block/null_blk/null_blk.h |  1 +
>  2 files changed, 17 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/block/null_blk/main.c b/drivers/block/null_blk/main.c
> index 4c601ca9552a..06eaa7ff4a86 100644
> --- a/drivers/block/null_blk/main.c
> +++ b/drivers/block/null_blk/main.c
> @@ -157,6 +157,10 @@ static int g_max_sectors;
>  module_param_named(max_sectors, g_max_sectors, int, 0444);
>  MODULE_PARM_DESC(max_sectors, "Maximum size of a command (in 512B sectors)");
>  
> +static unsigned int g_max_segment_size = BLK_MAX_SEGMENT_SIZE;
> +module_param_named(max_segment_size, g_max_segment_size, int, 0444);
> +MODULE_PARM_DESC(max_segment_size, "Maximum size of a segment in bytes");
> +
>  static unsigned int nr_devices = 1;
>  module_param(nr_devices, uint, 0444);
>  MODULE_PARM_DESC(nr_devices, "Number of devices to register");
> @@ -409,6 +413,7 @@ NULLB_DEVICE_ATTR(home_node, uint, NULL);
>  NULLB_DEVICE_ATTR(queue_mode, uint, NULL);
>  NULLB_DEVICE_ATTR(blocksize, uint, NULL);
>  NULLB_DEVICE_ATTR(max_sectors, uint, NULL);
> +NULLB_DEVICE_ATTR(max_segment_size, uint, NULL);
>  NULLB_DEVICE_ATTR(irqmode, uint, NULL);
>  NULLB_DEVICE_ATTR(hw_queue_depth, uint, NULL);
>  NULLB_DEVICE_ATTR(index, uint, NULL);
> @@ -550,6 +555,7 @@ static struct configfs_attribute *nullb_device_attrs[] = {
>  	&nullb_device_attr_queue_mode,
>  	&nullb_device_attr_blocksize,
>  	&nullb_device_attr_max_sectors,
> +	&nullb_device_attr_max_segment_size,
>  	&nullb_device_attr_irqmode,
>  	&nullb_device_attr_hw_queue_depth,
>  	&nullb_device_attr_index,
> @@ -630,7 +636,8 @@ static ssize_t memb_group_features_show(struct config_item *item, char *page)
>  	return snprintf(page, PAGE_SIZE,
>  			"badblocks,blocking,blocksize,cache_size,"
>  			"completion_nsec,discard,home_node,hw_queue_depth,"
> -			"irqmode,max_sectors,mbps,memory_backed,no_sched,"
> +			"irqmode,max_sectors,max_segment_size,mbps,"
> +			"memory_backed,no_sched,"
>  			"poll_queues,power,queue_mode,shared_tag_bitmap,size,"
>  			"submit_queues,use_per_node_hctx,virt_boundary,zoned,"
>  			"zone_capacity,zone_max_active,zone_max_open,"
> @@ -693,6 +700,7 @@ static struct nullb_device *null_alloc_dev(void)
>  	dev->queue_mode = g_queue_mode;
>  	dev->blocksize = g_bs;
>  	dev->max_sectors = g_max_sectors;
> +	dev->max_segment_size = g_max_segment_size;
>  	dev->irqmode = g_irqmode;
>  	dev->hw_queue_depth = g_hw_queue_depth;
>  	dev->blocking = g_blocking;
> @@ -1234,6 +1242,8 @@ static int null_transfer(struct nullb *nullb, struct page *page,
>  	unsigned int valid_len = len;
>  	int err = 0;
>  
> +	WARN_ONCE(len > dev->max_segment_size, "%u > %u\n", len,
> +		  dev->max_segment_size);

Shouldn't this be an EIO return as this is not supposed to happen ?

>  	if (!is_write) {
>  		if (dev->zoned)
>  			valid_len = null_zone_valid_read_len(nullb,
> @@ -1269,7 +1279,8 @@ static int null_handle_rq(struct nullb_cmd *cmd)
>  
>  	spin_lock_irq(&nullb->lock);
>  	rq_for_each_segment(bvec, rq, iter) {
> -		len = bvec.bv_len;
> +		len = min(bvec.bv_len, nullb->dev->max_segment_size);
> +		bvec.bv_len = len;
>  		err = null_transfer(nullb, bvec.bv_page, len, bvec.bv_offset,
>  				     op_is_write(req_op(rq)), sector,
>  				     rq->cmd_flags & REQ_FUA);
> @@ -1296,7 +1307,8 @@ static int null_handle_bio(struct nullb_cmd *cmd)
>  
>  	spin_lock_irq(&nullb->lock);
>  	bio_for_each_segment(bvec, bio, iter) {
> -		len = bvec.bv_len;
> +		len = min(bvec.bv_len, nullb->dev->max_segment_size);
> +		bvec.bv_len = len;
>  		err = null_transfer(nullb, bvec.bv_page, len, bvec.bv_offset,
>  				     op_is_write(bio_op(bio)), sector,
>  				     bio->bi_opf & REQ_FUA);
> @@ -2125,6 +2137,7 @@ static int null_add_dev(struct nullb_device *dev)
>  		dev->max_sectors = queue_max_hw_sectors(nullb->q);
>  	dev->max_sectors = min(dev->max_sectors, BLK_DEF_MAX_SECTORS);
>  	blk_queue_max_hw_sectors(nullb->q, dev->max_sectors);
> +	blk_queue_max_segment_size(nullb->q, dev->max_segment_size);
>  
>  	if (dev->virt_boundary)
>  		blk_queue_virt_boundary(nullb->q, PAGE_SIZE - 1);
> diff --git a/drivers/block/null_blk/null_blk.h b/drivers/block/null_blk/null_blk.h
> index eb5972c50be8..8cb73fe05fa3 100644
> --- a/drivers/block/null_blk/null_blk.h
> +++ b/drivers/block/null_blk/null_blk.h
> @@ -102,6 +102,7 @@ struct nullb_device {
>  	unsigned int queue_mode; /* block interface */
>  	unsigned int blocksize; /* block size */
>  	unsigned int max_sectors; /* Max sectors per command */
> +	unsigned int max_segment_size; /* Max size of a single DMA segment. */
>  	unsigned int irqmode; /* IRQ completion handler */
>  	unsigned int hw_queue_depth; /* queue depth */
>  	unsigned int index; /* index of the disk, only valid with a disk */

-- 
Damien Le Moal
Western Digital Research

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Bart Van Assche]

On 1/27/23 17:18, Damien Le Moal wrote:
> On 1/28/23 09:59, Bart Van Assche wrote:
>> +	WARN_ONCE(len > dev->max_segment_size, "%u > %u\n", len,
>> +		  dev->max_segment_size);
> 
> Shouldn't this be an EIO return as this is not supposed to happen ?

Hmm ... the above WARN_ONCE() statement is intended as a precondition 
check. This statement is intended as a help for developers to check that 
the code below works fine. I'm not sure how returning EIO here would help?

Thanks

Bart.

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Damien Le Moal]

On 1/28/23 11:48, Bart Van Assche wrote:
> On 1/27/23 17:18, Damien Le Moal wrote:
>> On 1/28/23 09:59, Bart Van Assche wrote:
>>> +	WARN_ONCE(len > dev->max_segment_size, "%u > %u\n", len,
>>> +		  dev->max_segment_size);
>>
>> Shouldn't this be an EIO return as this is not supposed to happen ?
> 
> Hmm ... the above WARN_ONCE() statement is intended as a precondition 
> check. This statement is intended as a help for developers to check that 
> the code below works fine. I'm not sure how returning EIO here would help?

My point was that given that null_transfer() is called like this:

	rq_for_each_segment(bvec, rq, iter) {
                len = bvec.bv_len;
                err = null_transfer(nullb, bvec.bv_page, len, bvec.bv_offset,
                                     ...);

len should never be larger than max_segment_size, no ? Unless I am missing
something else...

> 
> Thanks
> 
> Bart.

-- 
Damien Le Moal
Western Digital Research

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Bart Van Assche]

On 1/28/23 18:03, Damien Le Moal wrote:
> On 1/28/23 11:48, Bart Van Assche wrote:
>> On 1/27/23 17:18, Damien Le Moal wrote:
>>> On 1/28/23 09:59, Bart Van Assche wrote:
>>>> +	WARN_ONCE(len > dev->max_segment_size, "%u > %u\n", len,
>>>> +		  dev->max_segment_size);
>>>
>>> Shouldn't this be an EIO return as this is not supposed to happen ?
>>
>> Hmm ... the above WARN_ONCE() statement is intended as a precondition
>> check. This statement is intended as a help for developers to check that
>> the code below works fine. I'm not sure how returning EIO here would help?
> 
> My point was that given that null_transfer() is called like this:
> 
> 	rq_for_each_segment(bvec, rq, iter) {
>                  len = bvec.bv_len;
>                  err = null_transfer(nullb, bvec.bv_page, len, bvec.bv_offset,
>                                       ...);
> 
> len should never be larger than max_segment_size, no ? Unless I am missing
> something else...

Hi Damien,

null_transfer() supports the case len > dev->max_segment_size so I think 
that a warning is sufficient and also that we don't need to return -EIO.

Thanks,

Bart.

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Chaitanya Kulkarni]

On 1/27/23 16:59, Bart Van Assche wrote:
> Add support for configuring the maximum segment size.
> 
> Add support for segments smaller than the page size.
> 
> This patch enables testing segments smaller than the page size with a
> driver that does not call blk_rq_map_sg().
> 
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: Ming Lei <ming.lei@redhat.com>
> Cc: Damien Le Moal <damien.lemoal@opensource.wdc.com>
> Cc: Chaitanya Kulkarni <kch@nvidia.com>
> Signed-off-by: Bart Van Assche <bvanassche@acm.org>
> ---
>   drivers/block/null_blk/main.c     | 19 ++++++++++++++++---

Looks good.

Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>

-ck

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Chaitanya Kulkarni]

On 1/27/23 16:59, Bart Van Assche wrote:
> Add support for configuring the maximum segment size.
> 
> Add support for segments smaller than the page size.
> 
> This patch enables testing segments smaller than the page size with a
> driver that does not call blk_rq_map_sg().
> 
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: Ming Lei <ming.lei@redhat.com>
> Cc: Damien Le Moal <damien.lemoal@opensource.wdc.com>
> Cc: Chaitanya Kulkarni <kch@nvidia.com>
> Signed-off-by: Bart Van Assche <bvanassche@acm.org>

Bart, if you can send a blktest for this it will be great.

-ck

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Bart Van Assche]

On 1/31/23 13:43, Chaitanya Kulkarni wrote:
> On 1/27/23 16:59, Bart Van Assche wrote:
>> Add support for configuring the maximum segment size.
>>
>> Add support for segments smaller than the page size.
>>
>> This patch enables testing segments smaller than the page size with a
>> driver that does not call blk_rq_map_sg().
>>
>> Cc: Christoph Hellwig <hch@lst.de>
>> Cc: Ming Lei <ming.lei@redhat.com>
>> Cc: Damien Le Moal <damien.lemoal@opensource.wdc.com>
>> Cc: Chaitanya Kulkarni <kch@nvidia.com>
>> Signed-off-by: Bart Van Assche <bvanassche@acm.org>
> 
> Bart, if you can send a blktest for this it will be great.

Hi Chaitanya,

This patch has been resent as patch 7/7 of the following series: "[PATCH 
v4 0/7] Add support for limits below the page size" 
(https://lore.kernel.org/linux-block/20230130212656.876311-1-bvanassche@acm.org/T/#t).

blktests for that patch series are available here: 
https://github.com/bvanassche/blktests/commits/master. Do you prefer 
that I send a pull request for these blktests now or do you perhaps 
prefer that I wait until the kernel patch series has been merged?

Thanks,

Bart.

Re: [PATCH] null_blk: Support configuring the maximum segment size

[Chaitanya Kulkarni]

> Hi Chaitanya,
> 
> This patch has been resent as patch 7/7 of the following series: "[PATCH 
> v4 0/7] Add support for limits below the page size" 
> (https://lore.kernel.org/linux-block/20230130212656.876311-1-bvanassche@acm.org/T/#t). 
> 

I see.

> 
> blktests for that patch series are available here: 
> https://github.com/bvanassche/blktests/commits/master. Do you prefer 
> that I send a pull request for these blktests now or do you perhaps 
> prefer that I wait until the kernel patch series has been merged?
> 
> Thanks,
> 
> Bart.

Let's wait till it is merged.

-ck