Skip to contentSkip to conversation

scsi: sd: fix write_same16 and write_same10 for sector size > PAGE_SIZE

RFC Superseded [email protected] 3 participants 7 messages Dec 3, 2025 – Dec 4, 2025 SCSI

v2 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

RFC v2 1/2 scsi: sd: fix write_same16 and write_same10 for sector size > PAGE_SIZE

From: Swarna Prabhu <[email protected]>

The WRITE SAME(16) and WRITE SAME(10) scsi commands uses a page from a dedicated mempool('sd_page_pool') for its payload. This pool was initialized to allocate single pages, which was sufficient as long as the device sector size did not exceed the PAGE_SIZE.

Given that block layer now supports block size upto 64K ie beyond PAGE_SIZE, adapt sd_set_special_bvec() to accommodate that.

With the above fix, enable sector sizes > PAGE_SIZE in scsi sd driver.

---
Note: We are allocating pages of order aligned to BLK_MAX_BLOCK_SIZE for the mempool page allocator 'sd_page_pool' all the time. This is because we only know that a bigger sector device is attached at sd_probe and it might be too late to re allocate mempool with order > 0.

drivers/scsi/sd.c | 19 +++++++++++++------

1 file changed, 13 insertions(+), 6 deletions(-)
drivers/scsi/sd.c +13 -6
index 0252d3f6bed1..f2eac79d7263 100644
@@ -895,11 +895,20 @@ static void sd_config_discard(struct scsi_disk *sdkp, struct queue_limits *lim,
 static void *sd_set_special_bvec(struct request *rq, unsigned int data_len)
 {
 	struct page *page;
+	struct scsi_device *sdp = scsi_disk(rq->q->disk)->device;
+	unsigned sector_size = sdp->sector_size;
+	unsigned int nr_pages = DIV_ROUND_UP(sector_size, PAGE_SIZE);
+	int n = 0;
 
 	page = mempool_alloc(sd_page_pool, GFP_ATOMIC);
 	if (!page)
 		return NULL;
-	clear_highpage(page);
+
+	do {
+		clear_highpage(page + n);
+		n++;
+	} while (n < nr_pages);
+
 	bvec_set_page(&rq->special_vec, page, data_len, 0);
 	rq->rq_flags |= RQF_SPECIAL_PAYLOAD;
 	return bvec_virt(&rq->special_vec);
@@ -2880,10 +2889,8 @@ sd_read_capacity(struct scsi_disk *sdkp, struct queue_limits *lim,
 			  "assuming 512.\n");
 	}
 
-	if (sector_size != 512 &&
-	    sector_size != 1024 &&
-	    sector_size != 2048 &&
-	    sector_size != 4096) {
+	if (sector_size < 512 || sector_size > BLK_MAX_BLOCK_SIZE ||
+	    !is_power_of_2(sector_size)) {
 		sd_printk(KERN_NOTICE, sdkp, "Unsupported sector size %d.\n",
 			  sector_size);
 		/*
@@ -4368,7 +4375,7 @@ static int __init init_sd(void)
 	if (err)
 		goto err_out;
 
-	sd_page_pool = mempool_create_page_pool(SD_MEMPOOL_SIZE, 0);
+	sd_page_pool = mempool_create_page_pool(SD_MEMPOOL_SIZE, get_order(BLK_MAX_BLOCK_SIZE));
 	if (!sd_page_pool) {
 		printk(KERN_ERR "sd: can't init discard page pool\n");
 		err = -ENOMEM;
Signature · 1 line
2.51.0
#1 lore raw

RFC v2 2/2 scsi: scsi_debug: enable sdebug_sector_size > PAGE_SIZE

From: Swarna Prabhu <[email protected]>

Now that block layer can support block size > PAGE_SIZE and the issue with WRITE SAME(16) and WRITE SAME(10) are fixed for sector sizes > PAGE_SIZE, enable sdebug_sector_size

Quoted 1 line

PAGE_SIZE in scsi_debug.

---
 drivers/scsi/scsi_debug.c | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)
drivers/scsi/scsi_debug.c +2 -7
index b2ab97be5db3..b5839e62d3bb 100644
@@ -8459,13 +8459,8 @@ static int __init scsi_debug_init(void)
 	} else if (sdebug_ndelay > 0)
 		sdebug_jdelay = JDELAY_OVERRIDDEN;
 
-	switch (sdebug_sector_size) {
-	case  512:
-	case 1024:
-	case 2048:
-	case 4096:
-		break;
-	default:
+	if (sdebug_sector_size < 512 || sdebug_sector_size > BLK_MAX_BLOCK_SIZE ||
+	    !is_power_of_2(sdebug_sector_size)) {
 		pr_err("invalid sector_size %d\n", sdebug_sector_size);
 		return -EINVAL;
 	}
Signature · 1 line
2.51.0
Reviewed-by on v2 2/2
#2 lore raw

Re: [RFC v2 2/2] scsi: scsi_debug: enable sdebug_sector_size > PAGE_SIZE

Quoted 8 lines · On Wed, Dec 03, 2025 at 11:05:47PM +0000, [email protected] wrote:

From: Swarna Prabhu <[email protected]>

Now that block layer can support block size > PAGE_SIZE and the issue with WRITE SAME(16) and WRITE SAME(10) are fixed for sector sizes > PAGE_SIZE, enable sdebug_sector_size

Quoted 1 line

PAGE_SIZE in scsi_debug.

Luis

#3 lore raw

Re: [RFC v2 2/2] scsi: scsi_debug: enable sdebug_sector_size > PAGE_SIZE

In reply to [email protected]
Quoted 10 lines · On 12/3/25 1:05 PM, [email protected] wrote:
drivers/scsi/scsi_debug.c +0 -2
index b2ab97be5db3..b5839e62d3bb 100644
@@ -8459,13 +8459,8 @@ static int __init scsi_debug_init(void)
  	} else if (sdebug_ndelay > 0)
  		sdebug_jdelay = JDELAY_OVERRIDDEN;
 
-	switch (sdebug_sector_size) {
-	case  512:

drivers/scsi/scsi_debug.c static int __init scsi_debug_init(void) · line 8466

Quoted 10 lines
drivers/scsi/scsi_debug.c +2 -5
-	case 1024:
-	case 2048:
-	case 4096:
-		break;
-	default:
+	if (sdebug_sector_size < 512 || sdebug_sector_size > BLK_MAX_BLOCK_SIZE ||
+	    !is_power_of_2(sdebug_sector_size)) {
  		pr_err("invalid sector_size %d\n", sdebug_sector_size);
  		return -EINVAL;
  	}

Please use blk_validate_block_size() instead of open-coding it.

Thanks,

Bart.

Reviewed-by on v2 1/2
#4 lore raw
In reply to [email protected]
Quoted 12 lines · On Wed, Dec 03, 2025 at 11:05:46PM +0000, [email protected] wrote:

From: Swarna Prabhu <[email protected]>

The WRITE SAME(16) and WRITE SAME(10) scsi commands uses a page from a dedicated mempool('sd_page_pool') for its payload. This pool was initialized to allocate single pages, which was sufficient as long as the device sector size did not exceed the PAGE_SIZE.

Given that block layer now supports block size upto 64K ie beyond PAGE_SIZE, adapt sd_set_special_bvec() to accommodate that.

Quoted 6 lines

With the above fix, enable sector sizes > PAGE_SIZE in scsi sd driver.

Just missing the Cc stable # v6.15 tag. Other than that:

Luis

#5 lore raw
In reply to [email protected]

drivers/scsi/sd.c static void sd_config_discard(struct scsi_disk *sdkp, struct queue_limits *lim, · line 898

Quoted 4 lines · On 12/3/25 1:05 PM, [email protected] wrote:

static void *sd_set_special_bvec(struct request *rq, unsigned int data_len) {
struct page *page;

drivers/scsi/sd.c +1 -0
+	struct scsi_device *sdp = scsi_disk(rq->q->disk)->device;

Instead of using this cumbersome approach to obtain the SCSI device pointer, I recommend to change the 'struct request *rq' argument into 'struct scsi_cmnd *cmd' and to obtain the SCSI device pointer as follows:

	struct scsi_device *sdp = cmd->device;

Otherwise this patch looks good to me.

Thanks,

Bart.

#6 lore raw
In reply to [email protected]

drivers/scsi/sd.c sd_read_capacity(struct scsi_disk *sdkp, struct queue_limits *lim, · line 2896

Quoted 9 lines · On 12/3/25 1:05 PM, [email protected] wrote:
drivers/scsi/sd.c +2 -4
-	if (sector_size != 512 &&
-	    sector_size != 1024 &&
-	    sector_size != 2048 &&
-	    sector_size != 4096) {
+	if (sector_size < 512 || sector_size > BLK_MAX_BLOCK_SIZE ||
+	    !is_power_of_2(sector_size)) {
  		sd_printk(KERN_NOTICE, sdkp, "Unsupported sector size %d.\n",
  			  sector_size);
  		/*

Please use blk_validate_block_size() instead of open-coding it.

Thanks,

Bart.