btrfs: prepare scrub to support bs > ps cases

This involves:

- Migrate scrub_stripe::pages[] to folios[]

- Use btrfs_alloc_folio_array() and folio_put() to alloc above array.

- Migrate scrub_stripe_get_kaddr() and scrub_stripe_get_paddr() to use
  folio interfaces

- Migrate raid56_parity_cache_data_pages() to
  raid56_parity_cache_data_folios()
  Since scrub is the only caller still using pages.

  This helper will copy the folio array contents into rbio::stripe_pages,
  with sector uptodate flags updated.

  And a new ASSERT() to make sure bs > ps cases will not hit this path.

Since most scrub code is based on kaddr/paddr, the migration itself is
pretty straightforward.

And since we're here, also move the loop to set the
stripe_sectors[].uptodate out of the copy loop.
As we always mark all the sectors as uptodate for the data stripe, it's
easier to do in one go, other than doing it inside the copy loop.

Signed-off-by: Qu Wenruo <[email protected]>
Signed-off-by: David Sterba <[email protected]>

Author: adam900710

fs/btrfs/raid56.c

@@ -2844,19 +2844,22 @@ void raid56_parity_submit_scrub_rbio(struct btrfs_raid_bio *rbio)
  * This is for scrub call sites where we already have correct data contents.
  * This allows us to avoid reading data stripes again.
  *
- * Unfortunately here we have to do page copy, other than reusing the pages.
+ * Unfortunately here we have to do folio copy, other than reusing the pages.
  * This is due to the fact rbio has its own page management for its cache.
  */
-void raid56_parity_cache_data_pages(struct btrfs_raid_bio *rbio,
-				    struct page **data_pages, u64 data_logical)
+void raid56_parity_cache_data_folios(struct btrfs_raid_bio *rbio,
+				     struct folio **data_folios, u64 data_logical)
 {
+	struct btrfs_fs_info *fs_info = rbio->bioc->fs_info;
 	const u64 offset_in_full_stripe = data_logical -
 					  rbio->bioc->full_stripe_logical;
-	const int page_index = offset_in_full_stripe >> PAGE_SHIFT;
-	const u32 sectorsize = rbio->bioc->fs_info->sectorsize;
-	const u32 sectors_per_page = PAGE_SIZE / sectorsize;
+	unsigned int findex = 0;
+	unsigned int foffset = 0;
 	int ret;
 
+	/* We shouldn't hit RAID56 for bs > ps cases for now. */
+	ASSERT(fs_info->sectorsize <= PAGE_SIZE);
+
 	/*
 	 * If we hit ENOMEM temporarily, but later at
 	 * raid56_parity_submit_scrub_rbio() time it succeeded, we just do
@@ -2873,14 +2876,25 @@ void raid56_parity_cache_data_pages(struct btrfs_raid_bio *rbio,
 	ASSERT(IS_ALIGNED(offset_in_full_stripe, BTRFS_STRIPE_LEN));
 	ASSERT(offset_in_full_stripe < (rbio->nr_data << BTRFS_STRIPE_LEN_SHIFT));
 
-	for (int page_nr = 0; page_nr < (BTRFS_STRIPE_LEN >> PAGE_SHIFT); page_nr++) {
-		struct page *dst = rbio->stripe_pages[page_nr + page_index];
-		struct page *src = data_pages[page_nr];
-
-		memcpy_page(dst, 0, src, 0, PAGE_SIZE);
-		for (int sector_nr = sectors_per_page * page_index;
-		     sector_nr < sectors_per_page * (page_index + 1);
-		     sector_nr++)
-			rbio->stripe_sectors[sector_nr].uptodate = true;
+	for (unsigned int cur_off = offset_in_full_stripe;
+	     cur_off < offset_in_full_stripe + BTRFS_STRIPE_LEN;
+	     cur_off += PAGE_SIZE) {
+		const unsigned int pindex = cur_off >> PAGE_SHIFT;
+		void *kaddr;
+
+		kaddr = kmap_local_page(rbio->stripe_pages[pindex]);
+		memcpy_from_folio(kaddr, data_folios[findex], foffset, PAGE_SIZE);
+		kunmap_local(kaddr);
+
+		foffset += PAGE_SIZE;
+		ASSERT(foffset <= folio_size(data_folios[findex]));
+		if (foffset == folio_size(data_folios[findex])) {
+			findex++;
+			foffset = 0;
+		}
 	}
+	for (unsigned int sector_nr = offset_in_full_stripe >> fs_info->sectorsize_bits;
+	     sector_nr < (offset_in_full_stripe + BTRFS_STRIPE_LEN) >> fs_info->sectorsize_bits;
+	     sector_nr++)
+		rbio->stripe_sectors[sector_nr].uptodate = true;
 }

fs/btrfs/raid56.h

@@ -201,8 +201,8 @@ struct btrfs_raid_bio *raid56_parity_alloc_scrub_rbio(struct bio *bio,
 				unsigned long *dbitmap, int stripe_nsectors);
 void raid56_parity_submit_scrub_rbio(struct btrfs_raid_bio *rbio);
 
-void raid56_parity_cache_data_pages(struct btrfs_raid_bio *rbio,
-				    struct page **data_pages, u64 data_logical);
+void raid56_parity_cache_data_folios(struct btrfs_raid_bio *rbio,
+				     struct folio **data_folios, u64 data_logical);
 
 int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info);
 void btrfs_free_stripe_hash_table(struct btrfs_fs_info *info);

fs/btrfs/scrub.c

@@ -130,7 +130,7 @@ enum {
 	scrub_bitmap_nr_last,
 };
 
-#define SCRUB_STRIPE_PAGES		(BTRFS_STRIPE_LEN / PAGE_SIZE)
+#define SCRUB_STRIPE_MAX_FOLIOS		(BTRFS_STRIPE_LEN / PAGE_SIZE)
 
 /*
  * Represent one contiguous range with a length of BTRFS_STRIPE_LEN.
@@ -139,7 +139,7 @@ struct scrub_stripe {
 	struct scrub_ctx *sctx;
 	struct btrfs_block_group *bg;
 
-	struct page *pages[SCRUB_STRIPE_PAGES];
+	struct folio *folios[SCRUB_STRIPE_MAX_FOLIOS];
 	struct scrub_sector_verification *sectors;
 
 	struct btrfs_device *dev;
@@ -339,10 +339,10 @@ static void release_scrub_stripe(struct scrub_stripe *stripe)
 	if (!stripe)
 		return;
 
-	for (int i = 0; i < SCRUB_STRIPE_PAGES; i++) {
-		if (stripe->pages[i])
-			__free_page(stripe->pages[i]);
-		stripe->pages[i] = NULL;
+	for (int i = 0; i < SCRUB_STRIPE_MAX_FOLIOS; i++) {
+		if (stripe->folios[i])
+			folio_put(stripe->folios[i]);
+		stripe->folios[i] = NULL;
 	}
 	kfree(stripe->sectors);
 	kfree(stripe->csums);
@@ -355,6 +355,7 @@ static void release_scrub_stripe(struct scrub_stripe *stripe)
 static int init_scrub_stripe(struct btrfs_fs_info *fs_info,
 			     struct scrub_stripe *stripe)
 {
+	const u32 min_folio_shift = PAGE_SHIFT + fs_info->block_min_order;
 	int ret;
 
 	memset(stripe, 0, sizeof(*stripe));
@@ -367,7 +368,9 @@ static int init_scrub_stripe(struct btrfs_fs_info *fs_info,
 	atomic_set(&stripe->pending_io, 0);
 	spin_lock_init(&stripe->write_error_lock);
 
-	ret = btrfs_alloc_page_array(SCRUB_STRIPE_PAGES, stripe->pages, false);
+	ASSERT(BTRFS_STRIPE_LEN >> min_folio_shift <= SCRUB_STRIPE_MAX_FOLIOS);
+	ret = btrfs_alloc_folio_array(BTRFS_STRIPE_LEN >> min_folio_shift,
+				      fs_info->block_min_order, stripe->folios);
 	if (ret < 0)
 		goto error;
 
@@ -687,27 +690,30 @@ static int fill_writer_pointer_gap(struct scrub_ctx *sctx, u64 physical)
 
 static void *scrub_stripe_get_kaddr(struct scrub_stripe *stripe, int sector_nr)
 {
-	u32 offset = (sector_nr << stripe->bg->fs_info->sectorsize_bits);
-	const struct page *page = stripe->pages[offset >> PAGE_SHIFT];
+	struct btrfs_fs_info *fs_info = stripe->bg->fs_info;
+	const u32 min_folio_shift = PAGE_SHIFT + fs_info->block_min_order;
+	u32 offset = (sector_nr << fs_info->sectorsize_bits);
+	const struct folio *folio = stripe->folios[offset >> min_folio_shift];
 
-	/* stripe->pages[] is allocated by us and no highmem is allowed. */
-	ASSERT(page);
-	ASSERT(!PageHighMem(page));
-	return page_address(page) + offset_in_page(offset);
+	/* stripe->folios[] is allocated by us and no highmem is allowed. */
+	ASSERT(folio);
+	ASSERT(!folio_test_partial_kmap(folio));
+	return folio_address(folio) + offset_in_folio(folio, offset);
 }
 
 static phys_addr_t scrub_stripe_get_paddr(struct scrub_stripe *stripe, int sector_nr)
 {
 	struct btrfs_fs_info *fs_info = stripe->bg->fs_info;
+	const u32 min_folio_shift = PAGE_SHIFT + fs_info->block_min_order;
 	u32 offset = (sector_nr << fs_info->sectorsize_bits);
-	const struct page *page = stripe->pages[offset >> PAGE_SHIFT];
+	const struct folio *folio = stripe->folios[offset >> min_folio_shift];
 
-	/* stripe->pages[] is allocated by us and no highmem is allowed. */
-	ASSERT(page);
-	ASSERT(!PageHighMem(page));
-	/* And the range must be contained inside the page. */
-	ASSERT(offset_in_page(offset) + fs_info->sectorsize <= PAGE_SIZE);
-	return page_to_phys(page) + offset_in_page(offset);
+	/* stripe->folios[] is allocated by us and no highmem is allowed. */
+	ASSERT(folio);
+	ASSERT(!folio_test_partial_kmap(folio));
+	/* And the range must be contained inside the folio. */
+	ASSERT(offset_in_folio(folio, offset) + fs_info->sectorsize <= folio_size(folio));
+	return page_to_phys(folio_page(folio, 0)) + offset_in_folio(folio, offset);
 }
 
 static void scrub_verify_one_metadata(struct scrub_stripe *stripe, int sector_nr)
@@ -1872,6 +1878,7 @@ static void scrub_submit_initial_read(struct scrub_ctx *sctx,
 {
 	struct btrfs_fs_info *fs_info = sctx->fs_info;
 	struct btrfs_bio *bbio;
+	const u32 min_folio_shift = PAGE_SHIFT + fs_info->block_min_order;
 	unsigned int nr_sectors = stripe_length(stripe) >> fs_info->sectorsize_bits;
 	int mirror = stripe->mirror_num;
 
@@ -1884,7 +1891,7 @@ static void scrub_submit_initial_read(struct scrub_ctx *sctx,
 		return;
 	}
 
-	bbio = btrfs_bio_alloc(SCRUB_STRIPE_PAGES, REQ_OP_READ, fs_info,
+	bbio = btrfs_bio_alloc(BTRFS_STRIPE_LEN >> min_folio_shift, REQ_OP_READ, fs_info,
 			       scrub_read_endio, stripe);
 
 	bbio->bio.bi_iter.bi_sector = stripe->logical >> SECTOR_SHIFT;
@@ -2215,7 +2222,7 @@ static int scrub_raid56_parity_stripe(struct scrub_ctx *sctx,
 	for (int i = 0; i < data_stripes; i++) {
 		stripe = &sctx->raid56_data_stripes[i];
 
-		raid56_parity_cache_data_pages(rbio, stripe->pages,
+		raid56_parity_cache_data_folios(rbio, stripe->folios,
 				full_stripe_start + (i << BTRFS_STRIPE_LEN_SHIFT));
 	}
 	raid56_parity_submit_scrub_rbio(rbio);