Skip to contentSkip to conversation

scsi: sd: Set a default optimal IO size if one is not defined

PATCH In review Damien Le Moal 2 participants 3 messages Jun 11, 2025 SCSI

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 scsi: sd: Set a default optimal IO size if one is not defined

Introduce the helper function sd_set_io_opt() to set a disk io_opt limit. This new way of setting this limit falls back to using the max_sectors limit if the host does not define an optimal sector limit and the device did not indicate an optimal transfer size (e.g. as is the case for ATA devices). This fallback io_opt limit avoids the disk to be setup with the rather small 128 KB for as the read_ahead_kb attribute. The larger read_ahead_kb value set with the default io_opt limit significantly improves buffered read performance with file systems without any intervention from the user.

---
Changes from v1:
- Changed message level from wrong WARNING level to INFO level
- Added review tag

 drivers/scsi/sd.c | 37 +++++++++++++++++++++++++++----------
 1 file changed, 27 insertions(+), 10 deletions(-)
drivers/scsi/sd.c +27 -10
index daddef2e9e87..92d5b4900c0e 100644
@@ -3681,6 +3681,32 @@ static void sd_read_block_zero(struct scsi_disk *sdkp)
 	kfree(buffer);
 }
 
+/*
+ * Set the optimal I/O size: limit the default to the SCSI host optimal sector
+ * limit if it is set. There may be an impact on performance when the size of
+ * a request exceeds this host limit. If the host did not set any optimal
+ * sector limit and the device did not indicate an optimal transfer size
+ * (e.g. ATA devices), default to using the device max_sectors limit.
+ */
+static void sd_set_io_opt(struct scsi_disk *sdkp, unsigned int dev_max,
+			  struct queue_limits *lim)
+{
+	struct scsi_device *sdp = sdkp->device;
+	struct Scsi_Host *shost = sdp->host;
+
+	lim->io_opt = shost->opt_sectors << SECTOR_SHIFT;
+	if (sd_validate_opt_xfer_size(sdkp, dev_max))
+		lim->io_opt = min_not_zero(lim->io_opt,
+				logical_to_bytes(sdp, sdkp->opt_xfer_blocks));
+	if (!lim->io_opt) {
+		lim->io_opt = ALIGN_DOWN(lim->max_sectors << SECTOR_SHIFT,
+					 sdkp->physical_block_size - 1);
+		sd_first_printk(KERN_INFO, sdkp,
+			"Using default optimal transfer size of %u bytes\n",
+			lim->io_opt);
+	}
+}
+
 /**
  *	sd_revalidate_disk - called the first time a new disk is seen,
  *	performs disk spin up, read_capacity, etc.
@@ -3777,16 +3803,7 @@ static int sd_revalidate_disk(struct gendisk *disk)
 	else
 		lim.io_min = 0;
 
-	/*
-	 * Limit default to SCSI host optimal sector limit if set. There may be
-	 * an impact on performance for when the size of a request exceeds this
-	 * host limit.
-	 */
-	lim.io_opt = sdp->host->opt_sectors << SECTOR_SHIFT;
-	if (sd_validate_opt_xfer_size(sdkp, dev_max)) {
-		lim.io_opt = min_not_zero(lim.io_opt,
-				logical_to_bytes(sdp, sdkp->opt_xfer_blocks));
-	}
+	sd_set_io_opt(sdkp, dev_max, &lim);
 
 	sdkp->first_scan = 0;
 
Signature · 1 line
2.49.0
#1 lore raw
Quoted 8 lines · Damien Le Moal
drivers/scsi/sd.c +8 -0
+static void sd_set_io_opt(struct scsi_disk *sdkp, unsigned int dev_max,
+			  struct queue_limits *lim)
+{
+	struct scsi_device *sdp = sdkp->device;
+	struct Scsi_Host *shost = sdp->host;
+
+	lim->io_opt = shost->opt_sectors << SECTOR_SHIFT;
+	if (sd_validate_opt_xfer_size(sdkp, dev_max))

drivers/scsi/sd.c static void sd_read_block_zero(struct scsi_disk *sdkp) · line 3708

Quoted 10 lines
drivers/scsi/sd.c +10 -0
+		lim->io_opt = min_not_zero(lim->io_opt,
+				logical_to_bytes(sdp, sdkp->opt_xfer_blocks));
+	if (!lim->io_opt) {
+		lim->io_opt = ALIGN_DOWN(lim->max_sectors << SECTOR_SHIFT,
+					 sdkp->physical_block_size - 1);
+		sd_first_printk(KERN_INFO, sdkp,
+			"Using default optimal transfer size of %u bytes\n",
+			lim->io_opt);
+	}
+}

shost->opt_sectors and lim->max_sectors both have type unsigned int so when shifting these left there is a risk of overflow. As an example, from the scsi_debug driver:

	.max_sectors =		-1U,

Shouldn't integer overflows be prevented instead of letting these happen?

Thanks,

Bart.

#2 lore raw
Quoted 24 lines · Bart Van Assche
Quoted 8 lines · Damien Le Moal
drivers/scsi/sd.c +8 -0
+static void sd_set_io_opt(struct scsi_disk *sdkp, unsigned int dev_max,
+			  struct queue_limits *lim)
+{
+	struct scsi_device *sdp = sdkp->device;
+	struct Scsi_Host *shost = sdp->host;
+
+	lim->io_opt = shost->opt_sectors << SECTOR_SHIFT;
+	if (sd_validate_opt_xfer_size(sdkp, dev_max))

drivers/scsi/sd.c static void sd_read_block_zero(struct scsi_disk *sdkp) · line 3708

Quoted 10 lines
drivers/scsi/sd.c +10 -0
+		lim->io_opt = min_not_zero(lim->io_opt,
+				logical_to_bytes(sdp, sdkp->opt_xfer_blocks));
+	if (!lim->io_opt) {
+		lim->io_opt = ALIGN_DOWN(lim->max_sectors << SECTOR_SHIFT,
+					 sdkp->physical_block_size - 1);
+		sd_first_printk(KERN_INFO, sdkp,
+			"Using default optimal transfer size of %u bytes\n",
+			lim->io_opt);
+	}
+}

shost->opt_sectors and lim->max_sectors both have type unsigned int so when shifting these left there is a risk of overflow. As an example, from the scsi_debug driver:

Quoted 1 line
	.max_sectors =		-1U,

The code already was like that for opt_sectors and no-one complained.

Quoted 2 lines

Shouldn't integer overflows be prevented instead of letting these happen?

Sure, can do that.

Signature · 2 lines
Damien Le Moal
Western Digital Research