![Coly Li](/assets/img/avatar_default.png)
If create a loop device with a backing NVMe SSD, current loop device driver doesn't correctly set its queue's limits.discard_granularity and leaves it as 0. If a discard request at LBA 0 on this loop device, in __blkdev_issue_discard() the calculated req_sects will be 0, and a zero length discard request will trigger a BUG() panic in generic block layer code at block/blk-mq.c:563. [ 955.565006][ C39] ------------[ cut here ]------------ [ 955.559660][ C39] invalid opcode: 0000 [#1] SMP NOPTI [ 955.622171][ C39] CPU: 39 PID: 248 Comm: ksoftirqd/39 Tainted: G E 5.8.0-default+ #40 [ 955.622171][ C39] Hardware name: Lenovo ThinkSystem SR650 -[7X05CTO1WW]-/-[7X05CTO1WW]-, BIOS -[IVE160M-2.70]- 07/17/2020 [ 955.622175][ C39] RIP: 0010:blk_mq_end_request+0x107/0x110 [ 955.622177][ C39] Code: 48 8b 03 e9 59 ff ff ff 48 89 df 5b 5d 41 5c e9 9f ed ff ff 48 8b 35 98 3c f4 00 48 83 c7 10 48 83 c6 19 e8 cb 56 c9 ff eb cb <0f> 0b 0f 1f 80 00 00 00 00 0f 1f 44 00 00 55 48 89 e5 41 56 41 54 [ 955.622179][ C39] RSP: 0018:ffffb1288701fe28 EFLAGS: 00010202 [ 955.749277][ C39] RAX: 0000000000000001 RBX: ffff956fffba5080 RCX: 0000000000004003 [ 955.749278][ C39] RDX: 0000000000000003 RSI: 0000000000000000 RDI: 0000000000000000 [ 955.749279][ C39] RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 [ 955.749279][ C39] R10: ffffb1288701fd28 R11: 0000000000000001 R12: ffffffffa8e05160 [ 955.749280][ C39] R13: 0000000000000004 R14: 0000000000000004 R15: ffffffffa7ad3a1e [ 955.749281][ C39] FS: 0000000000000000(0000) GS:ffff95bfbda00000(0000) knlGS:0000000000000000 [ 955.749282][ C39] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 955.749282][ C39] CR2: 00007f6f0ef766a8 CR3: 0000005a37012002 CR4: 00000000007606e0 [ 955.749283][ C39] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 955.749284][ C39] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 955.749284][ C39] PKRU: 55555554 [ 955.749285][ C39] Call Trace: [ 955.749290][ C39] blk_done_softirq+0x99/0xc0 [ 957.550669][ C39] __do_softirq+0xd3/0x45f [ 957.550677][ C39] ? smpboot_thread_fn+0x2f/0x1e0 [ 957.550679][ C39] ? smpboot_thread_fn+0x74/0x1e0 [ 957.550680][ C39] ? smpboot_thread_fn+0x14e/0x1e0 [ 957.550684][ C39] run_ksoftirqd+0x30/0x60 [ 957.550687][ C39] smpboot_thread_fn+0x149/0x1e0 [ 957.886225][ C39] ? sort_range+0x20/0x20 [ 957.886226][ C39] kthread+0x137/0x160 [ 957.886228][ C39] ? kthread_park+0x90/0x90 [ 957.886231][ C39] ret_from_fork+0x22/0x30 [ 959.117120][ C39] ---[ end trace 3dacdac97e2ed164 ]--- This is the procedure to reproduce the panic, # modprobe scsi_debug delay=0 dev_size_mb=2048 max_queue=1 # losetup -f /dev/nvme0n1 --direct-io=on # blkdiscard /dev/loop0 -o 0 -l 0x200 This patch fixes the issue by checking q->limits.discard_granularity in __blkdev_issue_discard() before composing the discard bio. If the value is 0, then prints a warning oops information and returns -EOPNOTSUPP to the caller to indicate that this buggy device driver doesn't support discard request. Fixes:9b15d109a6
("block: improve discard bio alignment in __blkdev_issue_discard()") Fixes:c52abf5630
("loop: Better discard support for block devices") Reported-and-suggested-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: Coly Li <colyli@suse.de> Reviewed-by: Ming Lei <ming.lei@redhat.com> Reviewed-by: Hannes Reinecke <hare@suse.de> Reviewed-by: Jack Wang <jinpu.wang@cloud.ionos.com> Cc: Bart Van Assche <bvanassche@acm.org> Cc: Christoph Hellwig <hch@lst.de> Cc: Darrick J. Wong <darrick.wong@oracle.com> Cc: Enzo Matsumiya <ematsumiya@suse.com> Cc: Evan Green <evgreen@chromium.org> Cc: Jens Axboe <axboe@kernel.dk> Cc: Martin K. Petersen <martin.petersen@oracle.com> Cc: Xiao Ni <xni@redhat.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
442 lines
11 KiB
C
442 lines
11 KiB
C
// SPDX-License-Identifier: GPL-2.0
|
|
/*
|
|
* Functions related to generic helpers functions
|
|
*/
|
|
#include <linux/kernel.h>
|
|
#include <linux/module.h>
|
|
#include <linux/bio.h>
|
|
#include <linux/blkdev.h>
|
|
#include <linux/scatterlist.h>
|
|
|
|
#include "blk.h"
|
|
|
|
struct bio *blk_next_bio(struct bio *bio, unsigned int nr_pages, gfp_t gfp)
|
|
{
|
|
struct bio *new = bio_alloc(gfp, nr_pages);
|
|
|
|
if (bio) {
|
|
bio_chain(bio, new);
|
|
submit_bio(bio);
|
|
}
|
|
|
|
return new;
|
|
}
|
|
|
|
int __blkdev_issue_discard(struct block_device *bdev, sector_t sector,
|
|
sector_t nr_sects, gfp_t gfp_mask, int flags,
|
|
struct bio **biop)
|
|
{
|
|
struct request_queue *q = bdev_get_queue(bdev);
|
|
struct bio *bio = *biop;
|
|
unsigned int op;
|
|
sector_t bs_mask, part_offset = 0;
|
|
|
|
if (!q)
|
|
return -ENXIO;
|
|
|
|
if (bdev_read_only(bdev))
|
|
return -EPERM;
|
|
|
|
if (flags & BLKDEV_DISCARD_SECURE) {
|
|
if (!blk_queue_secure_erase(q))
|
|
return -EOPNOTSUPP;
|
|
op = REQ_OP_SECURE_ERASE;
|
|
} else {
|
|
if (!blk_queue_discard(q))
|
|
return -EOPNOTSUPP;
|
|
op = REQ_OP_DISCARD;
|
|
}
|
|
|
|
/* In case the discard granularity isn't set by buggy device driver */
|
|
if (WARN_ON_ONCE(!q->limits.discard_granularity)) {
|
|
char dev_name[BDEVNAME_SIZE];
|
|
|
|
bdevname(bdev, dev_name);
|
|
pr_err_ratelimited("%s: Error: discard_granularity is 0.\n", dev_name);
|
|
return -EOPNOTSUPP;
|
|
}
|
|
|
|
bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
|
|
if ((sector | nr_sects) & bs_mask)
|
|
return -EINVAL;
|
|
|
|
if (!nr_sects)
|
|
return -EINVAL;
|
|
|
|
/* In case the discard request is in a partition */
|
|
if (bdev->bd_partno)
|
|
part_offset = bdev->bd_part->start_sect;
|
|
|
|
while (nr_sects) {
|
|
sector_t granularity_aligned_lba, req_sects;
|
|
sector_t sector_mapped = sector + part_offset;
|
|
|
|
granularity_aligned_lba = round_up(sector_mapped,
|
|
q->limits.discard_granularity >> SECTOR_SHIFT);
|
|
|
|
/*
|
|
* Check whether the discard bio starts at a discard_granularity
|
|
* aligned LBA,
|
|
* - If no: set (granularity_aligned_lba - sector_mapped) to
|
|
* bi_size of the first split bio, then the second bio will
|
|
* start at a discard_granularity aligned LBA on the device.
|
|
* - If yes: use bio_aligned_discard_max_sectors() as the max
|
|
* possible bi_size of the first split bio. Then when this bio
|
|
* is split in device drive, the split ones are very probably
|
|
* to be aligned to discard_granularity of the device's queue.
|
|
*/
|
|
if (granularity_aligned_lba == sector_mapped)
|
|
req_sects = min_t(sector_t, nr_sects,
|
|
bio_aligned_discard_max_sectors(q));
|
|
else
|
|
req_sects = min_t(sector_t, nr_sects,
|
|
granularity_aligned_lba - sector_mapped);
|
|
|
|
WARN_ON_ONCE((req_sects << 9) > UINT_MAX);
|
|
|
|
bio = blk_next_bio(bio, 0, gfp_mask);
|
|
bio->bi_iter.bi_sector = sector;
|
|
bio_set_dev(bio, bdev);
|
|
bio_set_op_attrs(bio, op, 0);
|
|
|
|
bio->bi_iter.bi_size = req_sects << 9;
|
|
sector += req_sects;
|
|
nr_sects -= req_sects;
|
|
|
|
/*
|
|
* We can loop for a long time in here, if someone does
|
|
* full device discards (like mkfs). Be nice and allow
|
|
* us to schedule out to avoid softlocking if preempt
|
|
* is disabled.
|
|
*/
|
|
cond_resched();
|
|
}
|
|
|
|
*biop = bio;
|
|
return 0;
|
|
}
|
|
EXPORT_SYMBOL(__blkdev_issue_discard);
|
|
|
|
/**
|
|
* blkdev_issue_discard - queue a discard
|
|
* @bdev: blockdev to issue discard for
|
|
* @sector: start sector
|
|
* @nr_sects: number of sectors to discard
|
|
* @gfp_mask: memory allocation flags (for bio_alloc)
|
|
* @flags: BLKDEV_DISCARD_* flags to control behaviour
|
|
*
|
|
* Description:
|
|
* Issue a discard request for the sectors in question.
|
|
*/
|
|
int blkdev_issue_discard(struct block_device *bdev, sector_t sector,
|
|
sector_t nr_sects, gfp_t gfp_mask, unsigned long flags)
|
|
{
|
|
struct bio *bio = NULL;
|
|
struct blk_plug plug;
|
|
int ret;
|
|
|
|
blk_start_plug(&plug);
|
|
ret = __blkdev_issue_discard(bdev, sector, nr_sects, gfp_mask, flags,
|
|
&bio);
|
|
if (!ret && bio) {
|
|
ret = submit_bio_wait(bio);
|
|
if (ret == -EOPNOTSUPP)
|
|
ret = 0;
|
|
bio_put(bio);
|
|
}
|
|
blk_finish_plug(&plug);
|
|
|
|
return ret;
|
|
}
|
|
EXPORT_SYMBOL(blkdev_issue_discard);
|
|
|
|
/**
|
|
* __blkdev_issue_write_same - generate number of bios with same page
|
|
* @bdev: target blockdev
|
|
* @sector: start sector
|
|
* @nr_sects: number of sectors to write
|
|
* @gfp_mask: memory allocation flags (for bio_alloc)
|
|
* @page: page containing data to write
|
|
* @biop: pointer to anchor bio
|
|
*
|
|
* Description:
|
|
* Generate and issue number of bios(REQ_OP_WRITE_SAME) with same page.
|
|
*/
|
|
static int __blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
|
|
sector_t nr_sects, gfp_t gfp_mask, struct page *page,
|
|
struct bio **biop)
|
|
{
|
|
struct request_queue *q = bdev_get_queue(bdev);
|
|
unsigned int max_write_same_sectors;
|
|
struct bio *bio = *biop;
|
|
sector_t bs_mask;
|
|
|
|
if (!q)
|
|
return -ENXIO;
|
|
|
|
if (bdev_read_only(bdev))
|
|
return -EPERM;
|
|
|
|
bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
|
|
if ((sector | nr_sects) & bs_mask)
|
|
return -EINVAL;
|
|
|
|
if (!bdev_write_same(bdev))
|
|
return -EOPNOTSUPP;
|
|
|
|
/* Ensure that max_write_same_sectors doesn't overflow bi_size */
|
|
max_write_same_sectors = bio_allowed_max_sectors(q);
|
|
|
|
while (nr_sects) {
|
|
bio = blk_next_bio(bio, 1, gfp_mask);
|
|
bio->bi_iter.bi_sector = sector;
|
|
bio_set_dev(bio, bdev);
|
|
bio->bi_vcnt = 1;
|
|
bio->bi_io_vec->bv_page = page;
|
|
bio->bi_io_vec->bv_offset = 0;
|
|
bio->bi_io_vec->bv_len = bdev_logical_block_size(bdev);
|
|
bio_set_op_attrs(bio, REQ_OP_WRITE_SAME, 0);
|
|
|
|
if (nr_sects > max_write_same_sectors) {
|
|
bio->bi_iter.bi_size = max_write_same_sectors << 9;
|
|
nr_sects -= max_write_same_sectors;
|
|
sector += max_write_same_sectors;
|
|
} else {
|
|
bio->bi_iter.bi_size = nr_sects << 9;
|
|
nr_sects = 0;
|
|
}
|
|
cond_resched();
|
|
}
|
|
|
|
*biop = bio;
|
|
return 0;
|
|
}
|
|
|
|
/**
|
|
* blkdev_issue_write_same - queue a write same operation
|
|
* @bdev: target blockdev
|
|
* @sector: start sector
|
|
* @nr_sects: number of sectors to write
|
|
* @gfp_mask: memory allocation flags (for bio_alloc)
|
|
* @page: page containing data
|
|
*
|
|
* Description:
|
|
* Issue a write same request for the sectors in question.
|
|
*/
|
|
int blkdev_issue_write_same(struct block_device *bdev, sector_t sector,
|
|
sector_t nr_sects, gfp_t gfp_mask,
|
|
struct page *page)
|
|
{
|
|
struct bio *bio = NULL;
|
|
struct blk_plug plug;
|
|
int ret;
|
|
|
|
blk_start_plug(&plug);
|
|
ret = __blkdev_issue_write_same(bdev, sector, nr_sects, gfp_mask, page,
|
|
&bio);
|
|
if (ret == 0 && bio) {
|
|
ret = submit_bio_wait(bio);
|
|
bio_put(bio);
|
|
}
|
|
blk_finish_plug(&plug);
|
|
return ret;
|
|
}
|
|
EXPORT_SYMBOL(blkdev_issue_write_same);
|
|
|
|
static int __blkdev_issue_write_zeroes(struct block_device *bdev,
|
|
sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
|
|
struct bio **biop, unsigned flags)
|
|
{
|
|
struct bio *bio = *biop;
|
|
unsigned int max_write_zeroes_sectors;
|
|
struct request_queue *q = bdev_get_queue(bdev);
|
|
|
|
if (!q)
|
|
return -ENXIO;
|
|
|
|
if (bdev_read_only(bdev))
|
|
return -EPERM;
|
|
|
|
/* Ensure that max_write_zeroes_sectors doesn't overflow bi_size */
|
|
max_write_zeroes_sectors = bdev_write_zeroes_sectors(bdev);
|
|
|
|
if (max_write_zeroes_sectors == 0)
|
|
return -EOPNOTSUPP;
|
|
|
|
while (nr_sects) {
|
|
bio = blk_next_bio(bio, 0, gfp_mask);
|
|
bio->bi_iter.bi_sector = sector;
|
|
bio_set_dev(bio, bdev);
|
|
bio->bi_opf = REQ_OP_WRITE_ZEROES;
|
|
if (flags & BLKDEV_ZERO_NOUNMAP)
|
|
bio->bi_opf |= REQ_NOUNMAP;
|
|
|
|
if (nr_sects > max_write_zeroes_sectors) {
|
|
bio->bi_iter.bi_size = max_write_zeroes_sectors << 9;
|
|
nr_sects -= max_write_zeroes_sectors;
|
|
sector += max_write_zeroes_sectors;
|
|
} else {
|
|
bio->bi_iter.bi_size = nr_sects << 9;
|
|
nr_sects = 0;
|
|
}
|
|
cond_resched();
|
|
}
|
|
|
|
*biop = bio;
|
|
return 0;
|
|
}
|
|
|
|
/*
|
|
* Convert a number of 512B sectors to a number of pages.
|
|
* The result is limited to a number of pages that can fit into a BIO.
|
|
* Also make sure that the result is always at least 1 (page) for the cases
|
|
* where nr_sects is lower than the number of sectors in a page.
|
|
*/
|
|
static unsigned int __blkdev_sectors_to_bio_pages(sector_t nr_sects)
|
|
{
|
|
sector_t pages = DIV_ROUND_UP_SECTOR_T(nr_sects, PAGE_SIZE / 512);
|
|
|
|
return min(pages, (sector_t)BIO_MAX_PAGES);
|
|
}
|
|
|
|
static int __blkdev_issue_zero_pages(struct block_device *bdev,
|
|
sector_t sector, sector_t nr_sects, gfp_t gfp_mask,
|
|
struct bio **biop)
|
|
{
|
|
struct request_queue *q = bdev_get_queue(bdev);
|
|
struct bio *bio = *biop;
|
|
int bi_size = 0;
|
|
unsigned int sz;
|
|
|
|
if (!q)
|
|
return -ENXIO;
|
|
|
|
if (bdev_read_only(bdev))
|
|
return -EPERM;
|
|
|
|
while (nr_sects != 0) {
|
|
bio = blk_next_bio(bio, __blkdev_sectors_to_bio_pages(nr_sects),
|
|
gfp_mask);
|
|
bio->bi_iter.bi_sector = sector;
|
|
bio_set_dev(bio, bdev);
|
|
bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
|
|
|
|
while (nr_sects != 0) {
|
|
sz = min((sector_t) PAGE_SIZE, nr_sects << 9);
|
|
bi_size = bio_add_page(bio, ZERO_PAGE(0), sz, 0);
|
|
nr_sects -= bi_size >> 9;
|
|
sector += bi_size >> 9;
|
|
if (bi_size < sz)
|
|
break;
|
|
}
|
|
cond_resched();
|
|
}
|
|
|
|
*biop = bio;
|
|
return 0;
|
|
}
|
|
|
|
/**
|
|
* __blkdev_issue_zeroout - generate number of zero filed write bios
|
|
* @bdev: blockdev to issue
|
|
* @sector: start sector
|
|
* @nr_sects: number of sectors to write
|
|
* @gfp_mask: memory allocation flags (for bio_alloc)
|
|
* @biop: pointer to anchor bio
|
|
* @flags: controls detailed behavior
|
|
*
|
|
* Description:
|
|
* Zero-fill a block range, either using hardware offload or by explicitly
|
|
* writing zeroes to the device.
|
|
*
|
|
* If a device is using logical block provisioning, the underlying space will
|
|
* not be released if %flags contains BLKDEV_ZERO_NOUNMAP.
|
|
*
|
|
* If %flags contains BLKDEV_ZERO_NOFALLBACK, the function will return
|
|
* -EOPNOTSUPP if no explicit hardware offload for zeroing is provided.
|
|
*/
|
|
int __blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
|
|
sector_t nr_sects, gfp_t gfp_mask, struct bio **biop,
|
|
unsigned flags)
|
|
{
|
|
int ret;
|
|
sector_t bs_mask;
|
|
|
|
bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
|
|
if ((sector | nr_sects) & bs_mask)
|
|
return -EINVAL;
|
|
|
|
ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects, gfp_mask,
|
|
biop, flags);
|
|
if (ret != -EOPNOTSUPP || (flags & BLKDEV_ZERO_NOFALLBACK))
|
|
return ret;
|
|
|
|
return __blkdev_issue_zero_pages(bdev, sector, nr_sects, gfp_mask,
|
|
biop);
|
|
}
|
|
EXPORT_SYMBOL(__blkdev_issue_zeroout);
|
|
|
|
/**
|
|
* blkdev_issue_zeroout - zero-fill a block range
|
|
* @bdev: blockdev to write
|
|
* @sector: start sector
|
|
* @nr_sects: number of sectors to write
|
|
* @gfp_mask: memory allocation flags (for bio_alloc)
|
|
* @flags: controls detailed behavior
|
|
*
|
|
* Description:
|
|
* Zero-fill a block range, either using hardware offload or by explicitly
|
|
* writing zeroes to the device. See __blkdev_issue_zeroout() for the
|
|
* valid values for %flags.
|
|
*/
|
|
int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector,
|
|
sector_t nr_sects, gfp_t gfp_mask, unsigned flags)
|
|
{
|
|
int ret = 0;
|
|
sector_t bs_mask;
|
|
struct bio *bio;
|
|
struct blk_plug plug;
|
|
bool try_write_zeroes = !!bdev_write_zeroes_sectors(bdev);
|
|
|
|
bs_mask = (bdev_logical_block_size(bdev) >> 9) - 1;
|
|
if ((sector | nr_sects) & bs_mask)
|
|
return -EINVAL;
|
|
|
|
retry:
|
|
bio = NULL;
|
|
blk_start_plug(&plug);
|
|
if (try_write_zeroes) {
|
|
ret = __blkdev_issue_write_zeroes(bdev, sector, nr_sects,
|
|
gfp_mask, &bio, flags);
|
|
} else if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
|
|
ret = __blkdev_issue_zero_pages(bdev, sector, nr_sects,
|
|
gfp_mask, &bio);
|
|
} else {
|
|
/* No zeroing offload support */
|
|
ret = -EOPNOTSUPP;
|
|
}
|
|
if (ret == 0 && bio) {
|
|
ret = submit_bio_wait(bio);
|
|
bio_put(bio);
|
|
}
|
|
blk_finish_plug(&plug);
|
|
if (ret && try_write_zeroes) {
|
|
if (!(flags & BLKDEV_ZERO_NOFALLBACK)) {
|
|
try_write_zeroes = false;
|
|
goto retry;
|
|
}
|
|
if (!bdev_write_zeroes_sectors(bdev)) {
|
|
/*
|
|
* Zeroing offload support was indicated, but the
|
|
* device reported ILLEGAL REQUEST (for some devices
|
|
* there is no non-destructive way to verify whether
|
|
* WRITE ZEROES is actually supported).
|
|
*/
|
|
ret = -EOPNOTSUPP;
|
|
}
|
|
}
|
|
|
|
return ret;
|
|
}
|
|
EXPORT_SYMBOL(blkdev_issue_zeroout);
|