Re: [PATCH v3] scsi: sd: Fix build warning in sd_revalidate_disk()

From: Damien Le Moal
Date: Fri Aug 08 2025 - 09:36:11 EST


On 8/8/25 20:30, Abinash Singh wrote:
> A build warning was triggered due to excessive stack usage in
> sd_revalidate_disk():
>
> drivers/scsi/sd.c: In function ‘sd_revalidate_disk.isra’:
> drivers/scsi/sd.c:3824:1: warning: the frame size of 1160 bytes is larger than 1024 bytes [-Wframe-larger-than=]
>
> This is caused by a large local struct queue_limits (~400B) allocated
> on the stack. Replacing it with a heap allocation using kmalloc()
> significantly reduces frame usage. Kernel stack is limited (~8 KB),
> and allocating large structs on the stack is discouraged.
> As the function already performs heap allocations (e.g. for buffer),
> this change fits well.
>
> Signed-off-by: Abinash Singh <abinashsinghlalotra@xxxxxxxxx>
> ---
>
> Hi ,
>
> As I mentioned in v2,.. this is the corresponding
> v3 of that. Only difference is in using __free() attribute
> Please infor me which one is better.

Matter of taste. Personally, I dislike annotations that hide code.
I'll let Martin decide on this.

>
> changelog v2->v3:
> used __free(kfree) attribute.
> no extra goto statements (i.e `free_lim`)
>
> lim was initialized to NULL because there is a return path
> before its allocation. Early exit will pass uninitlized pointer to
> `kfree`.
> We could move the allocation earlier, before this check:
>
> if (!scsi_device_online(sdp))
> goto out;
> However, doing so would result in unnecessary allocation
> if the device is not online, leading to wasted resources.
>
> Thanks,
> ---
> drivers/scsi/sd.c | 42 +++++++++++++++++++++++++-----------------
> 1 file changed, 25 insertions(+), 17 deletions(-)
>
> diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
> index 4a68b2ab2804..50abbab7e27a 100644
> --- a/drivers/scsi/sd.c
> +++ b/drivers/scsi/sd.c
> @@ -34,6 +34,7 @@
> */
>
> #include <linux/bio-integrity.h>
> +#include <linux/cleanup.h>
> #include <linux/module.h>
> #include <linux/fs.h>
> #include <linux/kernel.h>
> @@ -3696,7 +3697,7 @@ static int sd_revalidate_disk(struct gendisk *disk)
> struct scsi_disk *sdkp = scsi_disk(disk);
> struct scsi_device *sdp = sdkp->device;
> sector_t old_capacity = sdkp->capacity;
> - struct queue_limits lim;
> + struct queue_limits *lim __free(kfree) = NULL;
> unsigned char *buffer;
> unsigned int dev_max;
> int err;
> @@ -3711,6 +3712,13 @@ static int sd_revalidate_disk(struct gendisk *disk)
> if (!scsi_device_online(sdp))
> goto out;
>
> + lim = kmalloc(sizeof(*lim), GFP_KERNEL);
> + if (!lim) {
> + sd_printk(KERN_WARNING, sdkp,
> + "sd_revalidate_disk: Disk limit allocation failure.\n");
> + goto out;
> + }
> +
> buffer = kmalloc(SD_BUF_SIZE, GFP_KERNEL);
> if (!buffer) {
> sd_printk(KERN_WARNING, sdkp, "sd_revalidate_disk: Memory "
> @@ -3720,14 +3728,14 @@ static int sd_revalidate_disk(struct gendisk *disk)
>
> sd_spinup_disk(sdkp);
>
> - lim = queue_limits_start_update(sdkp->disk->queue);
> + *lim = queue_limits_start_update(sdkp->disk->queue);
>
> /*
> * Without media there is no reason to ask; moreover, some devices
> * react badly if we do.
> */
> if (sdkp->media_present) {
> - sd_read_capacity(sdkp, &lim, buffer);
> + sd_read_capacity(sdkp, lim, buffer);
> /*
> * Some USB/UAS devices return generic values for mode pages
> * until the media has been accessed. Trigger a READ operation
> @@ -3741,17 +3749,17 @@ static int sd_revalidate_disk(struct gendisk *disk)
> * cause this to be updated correctly and any device which
> * doesn't support it should be treated as rotational.
> */
> - lim.features |= (BLK_FEAT_ROTATIONAL | BLK_FEAT_ADD_RANDOM);
> + lim->features |= (BLK_FEAT_ROTATIONAL | BLK_FEAT_ADD_RANDOM);
>
> if (scsi_device_supports_vpd(sdp)) {
> sd_read_block_provisioning(sdkp);
> - sd_read_block_limits(sdkp, &lim);
> + sd_read_block_limits(sdkp, lim);
> sd_read_block_limits_ext(sdkp);
> - sd_read_block_characteristics(sdkp, &lim);
> - sd_zbc_read_zones(sdkp, &lim, buffer);
> + sd_read_block_characteristics(sdkp, lim);
> + sd_zbc_read_zones(sdkp, lim, buffer);
> }
>
> - sd_config_discard(sdkp, &lim, sd_discard_mode(sdkp));
> + sd_config_discard(sdkp, lim, sd_discard_mode(sdkp));
>
> sd_print_capacity(sdkp, old_capacity);
>
> @@ -3761,45 +3769,45 @@ static int sd_revalidate_disk(struct gendisk *disk)
> sd_read_app_tag_own(sdkp, buffer);
> sd_read_write_same(sdkp, buffer);
> sd_read_security(sdkp, buffer);
> - sd_config_protection(sdkp, &lim);
> + sd_config_protection(sdkp, lim);
> }
>
> /*
> * We now have all cache related info, determine how we deal
> * with flush requests.
> */
> - sd_set_flush_flag(sdkp, &lim);
> + sd_set_flush_flag(sdkp, lim);
>
> /* Initial block count limit based on CDB TRANSFER LENGTH field size. */
> dev_max = sdp->use_16_for_rw ? SD_MAX_XFER_BLOCKS : SD_DEF_XFER_BLOCKS;
>
> /* Some devices report a maximum block count for READ/WRITE requests. */
> dev_max = min_not_zero(dev_max, sdkp->max_xfer_blocks);
> - lim.max_dev_sectors = logical_to_sectors(sdp, dev_max);
> + lim->max_dev_sectors = logical_to_sectors(sdp, dev_max);
>
> if (sd_validate_min_xfer_size(sdkp))
> - lim.io_min = logical_to_bytes(sdp, sdkp->min_xfer_blocks);
> + lim->io_min = logical_to_bytes(sdp, sdkp->min_xfer_blocks);
> else
> - lim.io_min = 0;
> + 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;
> + 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,
> + lim->io_opt = min_not_zero(lim->io_opt,
> logical_to_bytes(sdp, sdkp->opt_xfer_blocks));
> }
>
> sdkp->first_scan = 0;
>
> set_capacity_and_notify(disk, logical_to_sectors(sdp, sdkp->capacity));
> - sd_config_write_same(sdkp, &lim);
> + sd_config_write_same(sdkp, lim);
> kfree(buffer);
>
> - err = queue_limits_commit_update_frozen(sdkp->disk->queue, &lim);
> + err = queue_limits_commit_update_frozen(sdkp->disk->queue, lim);
> if (err)
> return err;
>


--
Damien Le Moal
Western Digital Research