Merge: dm: sync with upstream 6.6 fixes and improvements

MR: https://gitlab.com/redhat/centos-stream/src/kernel/centos-stream-9/-/merge_requests/3199

JIRA: https://issues.redhat.com/browse/RHEL-12342
JIRA: https://issues.redhat.com/browse/RHEL-8220
JIRA: https://issues.redhat.com/browse/RHEL-12435
Tested: lvm2-testsuite and dmtest
Upstream Status: kernel/git/torvalds/linux.git

Pull in various dm changes from upstream.

Signed-off-by: Benjamin Marzinski <[email protected]>

Approved-by: Heinz Mauelshagen <[email protected]>
Approved-by: Mike Snitzer <[email protected]>
Approved-by: Mikuláš Patočka <[email protected]>
Approved-by: Nigel Croxon <[email protected]>

Signed-off-by: Scott Weaver <[email protected]>

Author: swvrrh

Documentation/admin-guide/device-mapper/dm-flakey.rst

@@ -67,6 +67,16 @@ Optional feature parameters:
 	Perform the replacement only if bio->bi_opf has all the
 	selected flags set.
 
+  random_read_corrupt <probability>
+	During <down interval>, replace random byte in a read bio
+	with a random value. probability is an integer between
+	0 and 1000000000 meaning 0% to 100% probability of corruption.
+
+  random_write_corrupt <probability>
+	During <down interval>, replace random byte in a write bio
+	with a random value. probability is an integer between
+	0 and 1000000000 meaning 0% to 100% probability of corruption.
+
 Examples:
 
 Replaces the 32nd byte of READ bios with the value 1::

Documentation/admin-guide/device-mapper/dm-integrity.rst

@@ -25,7 +25,7 @@ mode it calculates and verifies the integrity tag internally. In this
 mode, the dm-integrity target can be used to detect silent data
 corruption on the disk or in the I/O path.
 
-There's an alternate mode of operation where dm-integrity uses bitmap
+There's an alternate mode of operation where dm-integrity uses a bitmap
 instead of a journal. If a bit in the bitmap is 1, the corresponding
 region's data and integrity tags are not synchronized - if the machine
 crashes, the unsynchronized regions will be recalculated. The bitmap mode
@@ -38,6 +38,15 @@ the device. But it will only format the device if the superblock contains
 zeroes. If the superblock is neither valid nor zeroed, the dm-integrity
 target can't be loaded.
 
+Accesses to the on-disk metadata area containing checksums (aka tags) are
+buffered using dm-bufio. When an access to any given metadata area
+occurs, each unique metadata area gets its own buffer(s). The buffer size
+is capped at the size of the metadata area, but may be smaller, thereby
+requiring multiple buffers to represent the full metadata area. A smaller
+buffer size will produce a smaller resulting read/write operation to the
+metadata area for small reads/writes. The metadata is still read even in
+a full write to the data covered by a single buffer.
+
 To use the target for the first time:
 
 1. overwrite the superblock with zeroes
@@ -93,7 +102,7 @@ journal_sectors:number
 	device. If the device is already formatted, the value from the
 	superblock is used.
 
-interleave_sectors:number
+interleave_sectors:number (default 32768)
 	The number of interleaved sectors. This values is rounded down to
 	a power of two. If the device is already formatted, the value from
 	the superblock is used.
@@ -102,20 +111,16 @@ meta_device:device
 	Don't interleave the data and metadata on the device. Use a
 	separate device for metadata.
 
-buffer_sectors:number
-	The number of sectors in one buffer. The value is rounded down to
-	a power of two.
-
-	The tag area is accessed using buffers, the buffer size is
-	configurable. The large buffer size means that the I/O size will
-	be larger, but there could be less I/Os issued.
+buffer_sectors:number (default 128)
+	The number of sectors in one metadata buffer. The value is rounded
+	down to a power of two.
 
-journal_watermark:number
+journal_watermark:number (default 50)
 	The journal watermark in percents. When the size of the journal
 	exceeds this watermark, the thread that flushes the journal will
 	be started.
 
-commit_time:number
+commit_time:number (default 10000)
 	Commit time in milliseconds. When this time passes, the journal is
 	written. The journal is also written immediately if the FLUSH
 	request is received.
@@ -163,11 +168,10 @@ journal_mac:algorithm(:key)	(the key is optional)
 	the journal. Thus, modified sector number would be detected at
 	this stage.
 
-block_size:number
-	The size of a data block in bytes.  The larger the block size the
+block_size:number (default 512)
+	The size of a data block in bytes. The larger the block size the
 	less overhead there is for per-block integrity metadata.
-	Supported values are 512, 1024, 2048 and 4096 bytes.  If not
-	specified the default block size is 512 bytes.
+	Supported values are 512, 1024, 2048 and 4096 bytes.
 
 sectors_per_bit:number
 	In the bitmap mode, this parameter specifies the number of
@@ -209,6 +213,12 @@ table and swap the tables with suspend and resume). The other arguments
 should not be changed when reloading the target because the layout of disk
 data depend on them and the reloaded target would be non-functional.
 
+For example, on a device using the default interleave_sectors of 32768, a
+block_size of 512, and an internal_hash of crc32c with a tag size of 4
+bytes, it will take 128 KiB of tags to track a full data area, requiring
+256 sectors of metadata per data area. With the default buffer_sectors of
+128, that means there will be 2 buffers per metadata area, or 2 buffers
+per 16 MiB of data.
 
 Status line:
 
@@ -286,7 +296,8 @@ The layout of the formatted block device:
     Each run contains:
 
 	* tag area - it contains integrity tags. There is one tag for each
-	  sector in the data area
+	  sector in the data area. The size of this area is always 4KiB or
+	  greater.
 	* data area - it contains data sectors. The number of data sectors
 	  in one run must be a power of two. log2 of this value is stored
 	  in the superblock.

MAINTAINERS

@@ -5534,8 +5534,8 @@ F:	include/linux/devm-helpers.h
 DEVICE-MAPPER  (LVM)
 M:	Alasdair Kergon <[email protected]>
 M:	Mike Snitzer <[email protected]>
-M:	dm-devel@redhat.com
-L:	dm-devel@redhat.com
+M:	dm-devel@lists.linux.dev
+L:	dm-devel@lists.linux.dev
 S:	Maintained
 W:	http://sources.redhat.com/dm
 Q:	http://patchwork.kernel.org/project/dm-devel/list/

drivers/md/dm-bufio.c

@@ -1157,23 +1157,6 @@ static void *alloc_buffer_data(struct dm_bufio_client *c, gfp_t gfp_mask,
 
 	*data_mode = DATA_MODE_VMALLOC;
 
-	/*
-	 * __vmalloc allocates the data pages and auxiliary structures with
-	 * gfp_flags that were specified, but pagetables are always allocated
-	 * with GFP_KERNEL, no matter what was specified as gfp_mask.
-	 *
-	 * Consequently, we must set per-process flag PF_MEMALLOC_NOIO so that
-	 * all allocations done by this process (including pagetables) are done
-	 * as if GFP_NOIO was specified.
-	 */
-	if (gfp_mask & __GFP_NORETRY) {
-		unsigned int noio_flag = memalloc_noio_save();
-		void *ptr = __vmalloc(c->block_size, gfp_mask);
-
-		memalloc_noio_restore(noio_flag);
-		return ptr;
-	}
-
 	return __vmalloc(c->block_size, gfp_mask);
 }
 
@@ -2592,6 +2575,13 @@ void dm_bufio_client_destroy(struct dm_bufio_client *c)
 }
 EXPORT_SYMBOL_GPL(dm_bufio_client_destroy);
 
+void dm_bufio_client_reset(struct dm_bufio_client *c)
+{
+	drop_buffers(c);
+	flush_work(&c->shrink_work);
+}
+EXPORT_SYMBOL_GPL(dm_bufio_client_reset);
+
 void dm_bufio_set_sector_offset(struct dm_bufio_client *c, sector_t start)
 {
 	c->start = start;

drivers/md/dm-core.h

@@ -214,6 +214,7 @@ struct dm_table {
 
 	/* a list of devices used by this table */
 	struct list_head devices;
+	struct rw_semaphore devices_lock;
 
 	/* events get handed up using this callback */
 	void (*event_fn)(void *data);
@@ -306,7 +307,8 @@ struct dm_io {
  */
 enum {
 	DM_IO_ACCOUNTED,
-	DM_IO_WAS_SPLIT
+	DM_IO_WAS_SPLIT,
+	DM_IO_BLK_STAT
 };
 
 static inline bool dm_io_flagged(struct dm_io *io, unsigned int bit)

drivers/md/dm-crypt.c

@@ -1661,15 +1661,18 @@ static void crypt_free_buffer_pages(struct crypt_config *cc, struct bio *clone);
  * In order to not degrade performance with excessive locking, we try
  * non-blocking allocations without a mutex first but on failure we fallback
  * to blocking allocations with a mutex.
+ *
+ * In order to reduce allocation overhead, we try to allocate compound pages in
+ * the first pass. If they are not available, we fall back to the mempool.
  */
 static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned int size)
 {
 	struct crypt_config *cc = io->cc;
 	struct bio *clone;
 	unsigned int nr_iovecs = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
 	gfp_t gfp_mask = GFP_NOWAIT | __GFP_HIGHMEM;
-	unsigned int i, len, remaining_size;
-	struct page *page;
+	unsigned int remaining_size;
+	unsigned int order = MAX_ORDER - 1;
 
 retry:
 	if (unlikely(gfp_mask & __GFP_DIRECT_RECLAIM))
@@ -1682,19 +1685,40 @@ static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned int size)
 
 	remaining_size = size;
 
-	for (i = 0; i < nr_iovecs; i++) {
-		page = mempool_alloc(&cc->page_pool, gfp_mask);
-		if (!page) {
+	while (remaining_size) {
+		struct page *pages;
+		unsigned size_to_add;
+		unsigned remaining_order = __fls((remaining_size + PAGE_SIZE - 1) >> PAGE_SHIFT);
+		order = min(order, remaining_order);
+
+		while (order > 0) {
+			if (unlikely(percpu_counter_read_positive(&cc->n_allocated_pages) +
+					(1 << order) > dm_crypt_pages_per_client))
+				goto decrease_order;
+			pages = alloc_pages(gfp_mask
+				| __GFP_NOMEMALLOC | __GFP_NORETRY | __GFP_NOWARN | __GFP_COMP,
+				order);
+			if (likely(pages != NULL)) {
+				percpu_counter_add(&cc->n_allocated_pages, 1 << order);
+				goto have_pages;
+			}
+decrease_order:
+			order--;
+		}
+
+		pages = mempool_alloc(&cc->page_pool, gfp_mask);
+		if (!pages) {
 			crypt_free_buffer_pages(cc, clone);
 			bio_put(clone);
 			gfp_mask |= __GFP_DIRECT_RECLAIM;
+			order = 0;
 			goto retry;
 		}
 
-		len = (remaining_size > PAGE_SIZE) ? PAGE_SIZE : remaining_size;
-
-		__bio_add_page(clone, page, len, 0);
-		remaining_size -= len;
+have_pages:
+		size_to_add = min((unsigned)PAGE_SIZE << order, remaining_size);
+		__bio_add_page(clone, pages, size_to_add, 0);
+		remaining_size -= size_to_add;
 	}
 
 	/* Allocate space for integrity tags */
@@ -1712,12 +1736,18 @@ static struct bio *crypt_alloc_buffer(struct dm_crypt_io *io, unsigned int size)
 
 static void crypt_free_buffer_pages(struct crypt_config *cc, struct bio *clone)
 {
-	struct bio_vec *bv;
-	struct bvec_iter_all iter_all;
+	struct folio_iter fi;
 
-	bio_for_each_segment_all(bv, clone, iter_all) {
-		BUG_ON(!bv->bv_page);
-		mempool_free(bv->bv_page, &cc->page_pool);
+	if (clone->bi_vcnt > 0) { /* bio_for_each_folio_all crashes with an empty bio */
+		bio_for_each_folio_all(fi, clone) {
+			if (folio_test_large(fi.folio)) {
+				percpu_counter_sub(&cc->n_allocated_pages,
+						1 << folio_order(fi.folio));
+				folio_put(fi.folio);
+			} else {
+				mempool_free(&fi.folio->page, &cc->page_pool);
+			}
+		}
 	}
 }
 
@@ -2888,7 +2918,7 @@ static int crypt_ctr_cipher_new(struct dm_target *ti, char *cipher_in, char *key
 		ret = crypt_ctr_auth_cipher(cc, cipher_api);
 		if (ret < 0) {
 			ti->error = "Invalid AEAD cipher spec";
-			return -ENOMEM;
+			return ret;
 		}
 	}