1 files changed, 177 insertions, 169 deletions

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index d6fe002a4a71..424f656cd765 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -19,10 +19,7 @@
 typedef void *z_erofs_next_pcluster_t;
 
 struct z_erofs_bvec {
-	union {
-		struct page *page;
-		struct folio *folio;
-	};
+	struct page *page;
 	int offset;
 	unsigned int end;
 };
@@ -449,44 +446,51 @@ static inline int erofs_cpu_hotplug_init(void) { return 0; }
 static inline void erofs_cpu_hotplug_destroy(void) {}
 #endif
 
-void z_erofs_exit_zip_subsystem(void)
+void z_erofs_exit_subsystem(void)
 {
 	erofs_cpu_hotplug_destroy();
 	erofs_destroy_percpu_workers();
 	destroy_workqueue(z_erofs_workqueue);
 	z_erofs_destroy_pcluster_pool();
+	z_erofs_exit_decompressor();
 }
 
-int __init z_erofs_init_zip_subsystem(void)
+int __init z_erofs_init_subsystem(void)
 {
-	int err = z_erofs_create_pcluster_pool();
+	int err = z_erofs_init_decompressor();
 
 	if (err)
-		goto out_error_pcluster_pool;
+		goto err_decompressor;
+
+	err = z_erofs_create_pcluster_pool();
+	if (err)
+		goto err_pcluster_pool;
 
 	z_erofs_workqueue = alloc_workqueue("erofs_worker",
 			WQ_UNBOUND | WQ_HIGHPRI, num_possible_cpus());
 	if (!z_erofs_workqueue) {
 		err = -ENOMEM;
-		goto out_error_workqueue_init;
+		goto err_workqueue_init;
 	}
 
 	err = erofs_init_percpu_workers();
 	if (err)
-		goto out_error_pcpu_worker;
+		goto err_pcpu_worker;
 
 	err = erofs_cpu_hotplug_init();
 	if (err < 0)
-		goto out_error_cpuhp_init;
+		goto err_cpuhp_init;
 	return err;
 
-out_error_cpuhp_init:
+err_cpuhp_init:
 	erofs_destroy_percpu_workers();
-out_error_pcpu_worker:
+err_pcpu_worker:
 	destroy_workqueue(z_erofs_workqueue);
-out_error_workqueue_init:
+err_workqueue_init:
 	z_erofs_destroy_pcluster_pool();
-out_error_pcluster_pool:
+err_pcluster_pool:
+	z_erofs_exit_decompressor();
+err_decompressor:
 	return err;
 }
 
@@ -617,32 +621,31 @@ static void z_erofs_bind_cache(struct z_erofs_decompress_frontend *fe)
 		fe->mode = Z_EROFS_PCLUSTER_FOLLOWED_NOINPLACE;
 }
 
-/* called by erofs_shrinker to get rid of all cached compressed bvecs */
+/* (erofs_shrinker) disconnect cached encoded data with pclusters */
 int erofs_try_to_free_all_cached_folios(struct erofs_sb_info *sbi,
 					struct erofs_workgroup *grp)
 {
 	struct z_erofs_pcluster *const pcl =
 		container_of(grp, struct z_erofs_pcluster, obj);
 	unsigned int pclusterpages = z_erofs_pclusterpages(pcl);
+	struct folio *folio;
 	int i;
 
 	DBG_BUGON(z_erofs_is_inline_pcluster(pcl));
-	/* There is no actice user since the pcluster is now freezed */
+	/* Each cached folio contains one page unless bs > ps is supported */
 	for (i = 0; i < pclusterpages; ++i) {
-		struct folio *folio = pcl->compressed_bvecs[i].folio;
-
-		if (!folio)
-			continue;
+		if (pcl->compressed_bvecs[i].page) {
+			folio = page_folio(pcl->compressed_bvecs[i].page);
+			/* Avoid reclaiming or migrating this folio */
+			if (!folio_trylock(folio))
+				return -EBUSY;
 
-		/* Avoid reclaiming or migrating this folio */
-		if (!folio_trylock(folio))
-			return -EBUSY;
-
-		if (!erofs_folio_is_managed(sbi, folio))
-			continue;
-		pcl->compressed_bvecs[i].folio = NULL;
-		folio_detach_private(folio);
-		folio_unlock(folio);
+			if (!erofs_folio_is_managed(sbi, folio))
+				continue;
+			pcl->compressed_bvecs[i].page = NULL;
+			folio_detach_private(folio);
+			folio_unlock(folio);
+		}
 	}
 	return 0;
 }
@@ -650,9 +653,9 @@ int erofs_try_to_free_all_cached_folios(struct erofs_sb_info *sbi,
 static bool z_erofs_cache_release_folio(struct folio *folio, gfp_t gfp)
 {
 	struct z_erofs_pcluster *pcl = folio_get_private(folio);
-	unsigned int pclusterpages = z_erofs_pclusterpages(pcl);
+	struct z_erofs_bvec *bvec = pcl->compressed_bvecs;
+	struct z_erofs_bvec *end = bvec + z_erofs_pclusterpages(pcl);
 	bool ret;
-	int i;
 
 	if (!folio_test_private(folio))
 		return true;
@@ -661,9 +664,9 @@ static bool z_erofs_cache_release_folio(struct folio *folio, gfp_t gfp)
 	spin_lock(&pcl->obj.lockref.lock);
 	if (pcl->obj.lockref.count <= 0) {
 		DBG_BUGON(z_erofs_is_inline_pcluster(pcl));
-		for (i = 0; i < pclusterpages; ++i) {
-			if (pcl->compressed_bvecs[i].folio == folio) {
-				pcl->compressed_bvecs[i].folio = NULL;
+		for (; bvec < end; ++bvec) {
+			if (bvec->page && page_folio(bvec->page) == folio) {
+				bvec->page = NULL;
 				folio_detach_private(folio);
 				ret = true;
 				break;
@@ -925,7 +928,7 @@ static void z_erofs_pcluster_end(struct z_erofs_decompress_frontend *fe)
 	fe->pcl = NULL;
 }
 
-static int z_erofs_read_fragment(struct super_block *sb, struct page *page,
+static int z_erofs_read_fragment(struct super_block *sb, struct folio *folio,
 			unsigned int cur, unsigned int end, erofs_off_t pos)
 {
 	struct inode *packed_inode = EROFS_SB(sb)->packed_inode;
@@ -938,113 +941,109 @@ static int z_erofs_read_fragment(struct super_block *sb, struct page *page,
 
 	buf.mapping = packed_inode->i_mapping;
 	for (; cur < end; cur += cnt, pos += cnt) {
-		cnt = min_t(unsigned int, end - cur,
-			    sb->s_blocksize - erofs_blkoff(sb, pos));
+		cnt = min(end - cur, sb->s_blocksize - erofs_blkoff(sb, pos));
 		src = erofs_bread(&buf, pos, EROFS_KMAP);
 		if (IS_ERR(src)) {
 			erofs_put_metabuf(&buf);
 			return PTR_ERR(src);
 		}
-		memcpy_to_page(page, cur, src, cnt);
+		memcpy_to_folio(folio, cur, src, cnt);
 	}
 	erofs_put_metabuf(&buf);
 	return 0;
 }
 
-static int z_erofs_scan_folio(struct z_erofs_decompress_frontend *fe,
+static int z_erofs_scan_folio(struct z_erofs_decompress_frontend *f,
 			      struct folio *folio, bool ra)
 {
-	struct inode *const inode = fe->inode;
-	struct erofs_map_blocks *const map = &fe->map;
+	struct inode *const inode = f->inode;
+	struct erofs_map_blocks *const map = &f->map;
 	const loff_t offset = folio_pos(folio);
-	const unsigned int bs = i_blocksize(inode), fs = folio_size(folio);
-	bool tight = true, exclusive;
-	unsigned int cur, end, len, split;
+	const unsigned int bs = i_blocksize(inode);
+	unsigned int end = folio_size(folio), split = 0, cur, pgs;
+	bool tight, excl;
 	int err = 0;
 
+	tight = (bs == PAGE_SIZE);
 	z_erofs_onlinefolio_init(folio);
-	split = 0;
-	end = fs;
-repeat:
-	if (offset + end - 1 < map->m_la ||
-	    offset + end - 1 >= map->m_la + map->m_llen) {
-		z_erofs_pcluster_end(fe);
-		map->m_la = offset + end - 1;
-		map->m_llen = 0;
-		err = z_erofs_map_blocks_iter(inode, map, 0);
-		if (err)
-			goto out;
-	}
-
-	cur = offset > map->m_la ? 0 : map->m_la - offset;
-	/* bump split parts first to avoid several separate cases */
-	++split;
-
-	if (!(map->m_flags & EROFS_MAP_MAPPED)) {
-		folio_zero_segment(folio, cur, end);
-		tight = false;
-		goto next_part;
-	}
-
-	if (map->m_flags & EROFS_MAP_FRAGMENT) {
-		erofs_off_t fpos = offset + cur - map->m_la;
+	do {
+		if (offset + end - 1 < map->m_la ||
+		    offset + end - 1 >= map->m_la + map->m_llen) {
+			z_erofs_pcluster_end(f);
+			map->m_la = offset + end - 1;
+			map->m_llen = 0;
+			err = z_erofs_map_blocks_iter(inode, map, 0);
+			if (err)
+				break;
+		}
 
-		len = min_t(unsigned int, map->m_llen - fpos, end - cur);
-		err = z_erofs_read_fragment(inode->i_sb, &folio->page, cur,
-			cur + len, EROFS_I(inode)->z_fragmentoff + fpos);
-		if (err)
-			goto out;
-		tight = false;
-		goto next_part;
-	}
+		cur = offset > map->m_la ? 0 : map->m_la - offset;
+		pgs = round_down(cur, PAGE_SIZE);
+		/* bump split parts first to avoid several separate cases */
+		++split;
+
+		if (!(map->m_flags & EROFS_MAP_MAPPED)) {
+			folio_zero_segment(folio, cur, end);
+			tight = false;
+		} else if (map->m_flags & EROFS_MAP_FRAGMENT) {
+			erofs_off_t fpos = offset + cur - map->m_la;
+
+			err = z_erofs_read_fragment(inode->i_sb, folio, cur,
+					cur + min(map->m_llen - fpos, end - cur),
+					EROFS_I(inode)->z_fragmentoff + fpos);
+			if (err)
+				break;
+			tight = false;
+		} else {
+			if (!f->pcl) {
+				err = z_erofs_pcluster_begin(f);
+				if (err)
+					break;
+				f->pcl->besteffort |= !ra;
+			}
 
-	if (!fe->pcl) {
-		err = z_erofs_pcluster_begin(fe);
-		if (err)
-			goto out;
-		fe->pcl->besteffort |= !ra;
-	}
+			pgs = round_down(end - 1, PAGE_SIZE);
+			/*
+			 * Ensure this partial page belongs to this submit chain
+			 * rather than other concurrent submit chains or
+			 * noio(bypass) chains since those chains are handled
+			 * asynchronously thus it cannot be used for inplace I/O
+			 * or bvpage (should be processed in the strict order.)
+			 */
+			tight &= (f->mode >= Z_EROFS_PCLUSTER_FOLLOWED);
+			excl = false;
+			if (cur <= pgs) {
+				excl = (split <= 1) || tight;
+				cur = pgs;
+			}
 
-	/*
-	 * Ensure the current partial folio belongs to this submit chain rather
-	 * than other concurrent submit chains or the noio(bypass) chain since
-	 * those chains are handled asynchronously thus the folio cannot be used
-	 * for inplace I/O or bvpage (should be processed in a strict order.)
-	 */
-	tight &= (fe->mode > Z_EROFS_PCLUSTER_FOLLOWED_NOINPLACE);
-	exclusive = (!cur && ((split <= 1) || (tight && bs == fs)));
-	if (cur)
-		tight &= (fe->mode >= Z_EROFS_PCLUSTER_FOLLOWED);
-
-	err = z_erofs_attach_page(fe, &((struct z_erofs_bvec) {
-					.page = &folio->page,
-					.offset = offset - map->m_la,
-					.end = end,
-				  }), exclusive);
-	if (err)
-		goto out;
-
-	z_erofs_onlinefolio_split(folio);
-	if (fe->pcl->pageofs_out != (map->m_la & ~PAGE_MASK))
-		fe->pcl->multibases = true;
-	if (fe->pcl->length < offset + end - map->m_la) {
-		fe->pcl->length = offset + end - map->m_la;
-		fe->pcl->pageofs_out = map->m_la & ~PAGE_MASK;
-	}
-	if ((map->m_flags & EROFS_MAP_FULL_MAPPED) &&
-	    !(map->m_flags & EROFS_MAP_PARTIAL_REF) &&
-	    fe->pcl->length == map->m_llen)
-		fe->pcl->partial = false;
-next_part:
-	/* shorten the remaining extent to update progress */
-	map->m_llen = offset + cur - map->m_la;
-	map->m_flags &= ~EROFS_MAP_FULL_MAPPED;
-
-	end = cur;
-	if (end > 0)
-		goto repeat;
+			err = z_erofs_attach_page(f, &((struct z_erofs_bvec) {
+				.page = folio_page(folio, pgs >> PAGE_SHIFT),
+				.offset = offset + pgs - map->m_la,
+				.end = end - pgs, }), excl);
+			if (err)
+				break;
 
-out:
+			z_erofs_onlinefolio_split(folio);
+			if (f->pcl->pageofs_out != (map->m_la & ~PAGE_MASK))
+				f->pcl->multibases = true;
+			if (f->pcl->length < offset + end - map->m_la) {
+				f->pcl->length = offset + end - map->m_la;
+				f->pcl->pageofs_out = map->m_la & ~PAGE_MASK;
+			}
+			if ((map->m_flags & EROFS_MAP_FULL_MAPPED) &&
+			    !(map->m_flags & EROFS_MAP_PARTIAL_REF) &&
+			    f->pcl->length == map->m_llen)
+				f->pcl->partial = false;
+		}
+		/* shorten the remaining extent to update progress */
+		map->m_llen = offset + cur - map->m_la;
+		map->m_flags &= ~EROFS_MAP_FULL_MAPPED;
+		if (cur <= pgs) {
+			split = cur < pgs;
+			tight = (bs == PAGE_SIZE);
+		}
+	} while ((end = cur) > 0);
 	z_erofs_onlinefolio_end(folio, err);
 	return err;
 }
@@ -1066,7 +1065,7 @@ static bool z_erofs_is_sync_decompress(struct erofs_sb_info *sbi,
 
 static bool z_erofs_page_is_invalidated(struct page *page)
 {
-	return !page->mapping && !z_erofs_is_shortlived_page(page);
+	return !page_folio(page)->mapping && !z_erofs_is_shortlived_page(page);
 }
 
 struct z_erofs_decompress_backend {
@@ -1221,8 +1220,8 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
 	struct z_erofs_pcluster *pcl = be->pcl;
 	unsigned int pclusterpages = z_erofs_pclusterpages(pcl);
 	const struct z_erofs_decompressor *decomp =
-				&erofs_decompressors[pcl->algorithmformat];
-	int i, err2;
+				z_erofs_decomp[pcl->algorithmformat];
+	int i, j, jtop, err2;
 	struct page *page;
 	bool overlapped;
 
@@ -1280,10 +1279,9 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
 		WRITE_ONCE(pcl->compressed_bvecs[0].page, NULL);
 		put_page(page);
 	} else {
+		/* managed folios are still left in compressed_bvecs[] */
 		for (i = 0; i < pclusterpages; ++i) {
-			/* consider shortlived pages added when decompressing */
 			page = be->compressed_pages[i];
-
 			if (!page ||
 			    erofs_folio_is_managed(sbi, page_folio(page)))
 				continue;
@@ -1294,21 +1292,31 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
 	if (be->compressed_pages < be->onstack_pages ||
 	    be->compressed_pages >= be->onstack_pages + Z_EROFS_ONSTACK_PAGES)
 		kvfree(be->compressed_pages);
-	z_erofs_fill_other_copies(be, err);
 
+	jtop = 0;
+	z_erofs_fill_other_copies(be, err);
 	for (i = 0; i < be->nr_pages; ++i) {
 		page = be->decompressed_pages[i];
 		if (!page)
 			continue;
 
 		DBG_BUGON(z_erofs_page_is_invalidated(page));
-
-		/* recycle all individual short-lived pages */
-		if (z_erofs_put_shortlivedpage(be->pagepool, page))
+		if (!z_erofs_is_shortlived_page(page)) {
+			z_erofs_onlinefolio_end(page_folio(page), err);
 			continue;
-		z_erofs_onlinefolio_end(page_folio(page), err);
+		}
+		if (pcl->algorithmformat != Z_EROFS_COMPRESSION_LZ4) {
+			erofs_pagepool_add(be->pagepool, page);
+			continue;
+		}
+		for (j = 0; j < jtop && be->decompressed_pages[j] != page; ++j)
+			;
+		if (j >= jtop)	/* this bounce page is newly detected */
+			be->decompressed_pages[jtop++] = page;
 	}
-
+	while (jtop)
+		erofs_pagepool_add(be->pagepool,
+				   be->decompressed_pages[--jtop]);
 	if (be->decompressed_pages != be->onstack_pages)
 		kvfree(be->decompressed_pages);
 
@@ -1419,7 +1427,7 @@ static void z_erofs_fill_bio_vec(struct bio_vec *bvec,
 	bool tocache = false;
 	struct z_erofs_bvec zbv;
 	struct address_space *mapping;
-	struct page *page;
+	struct folio *folio;
 	int bs = i_blocksize(f->inode);
 
 	/* Except for inplace folios, the entire folio can be used for I/Os */
@@ -1429,23 +1437,25 @@ repeat:
 	spin_lock(&pcl->obj.lockref.lock);
 	zbv = pcl->compressed_bvecs[nr];
 	spin_unlock(&pcl->obj.lockref.lock);
-	if (!zbv.folio)
+	if (!zbv.page)
 		goto out_allocfolio;
 
-	bvec->bv_page = &zbv.folio->page;
+	bvec->bv_page = zbv.page;
 	DBG_BUGON(z_erofs_is_shortlived_page(bvec->bv_page));
+
+	folio = page_folio(zbv.page);
 	/*
 	 * Handle preallocated cached folios.  We tried to allocate such folios
 	 * without triggering direct reclaim.  If allocation failed, inplace
 	 * file-backed folios will be used instead.
 	 */
-	if (zbv.folio->private == (void *)Z_EROFS_PREALLOCATED_PAGE) {
-		zbv.folio->private = 0;
+	if (folio->private == (void *)Z_EROFS_PREALLOCATED_PAGE) {
+		folio->private = 0;
 		tocache = true;
 		goto out_tocache;
 	}
 
-	mapping = READ_ONCE(zbv.folio->mapping);
+	mapping = READ_ONCE(folio->mapping);
 	/*
 	 * File-backed folios for inplace I/Os are all locked steady,
 	 * therefore it is impossible for `mapping` to be NULL.
@@ -1457,21 +1467,21 @@ repeat:
 		return;
 	}
 
-	folio_lock(zbv.folio);
-	if (zbv.folio->mapping == mc) {
+	folio_lock(folio);
+	if (folio->mapping == mc) {
 		/*
 		 * The cached folio is still in managed cache but without
 		 * a valid `->private` pcluster hint.  Let's reconnect them.
 		 */
-		if (!folio_test_private(zbv.folio)) {
-			folio_attach_private(zbv.folio, pcl);
+		if (!folio_test_private(folio)) {
+			folio_attach_private(folio, pcl);
 			/* compressed_bvecs[] already takes a ref before */
-			folio_put(zbv.folio);
+			folio_put(folio);
 		}
 
 		/* no need to submit if it is already up-to-date */
-		if (folio_test_uptodate(zbv.folio)) {
-			folio_unlock(zbv.folio);
+		if (folio_test_uptodate(folio)) {
+			folio_unlock(folio);
 			bvec->bv_page = NULL;
 		}
 		return;
@@ -1481,32 +1491,31 @@ repeat:
 	 * It has been truncated, so it's unsafe to reuse this one. Let's
 	 * allocate a new page for compressed data.
 	 */
-	DBG_BUGON(zbv.folio->mapping);
+	DBG_BUGON(folio->mapping);
 	tocache = true;
-	folio_unlock(zbv.folio);
-	folio_put(zbv.folio);
+	folio_unlock(folio);
+	folio_put(folio);
 out_allocfolio:
-	page = erofs_allocpage(&f->pagepool, gfp | __GFP_NOFAIL);
+	zbv.page = erofs_allocpage(&f->pagepool, gfp | __GFP_NOFAIL);
 	spin_lock(&pcl->obj.lockref.lock);
-	if (pcl->compressed_bvecs[nr].folio) {
-		erofs_pagepool_add(&f->pagepool, page);
+	if (pcl->compressed_bvecs[nr].page) {
+		erofs_pagepool_add(&f->pagepool, zbv.page);
 		spin_unlock(&pcl->obj.lockref.lock);
 		cond_resched();
 		goto repeat;
 	}
-	pcl->compressed_bvecs[nr].folio = zbv.folio = page_folio(page);
+	bvec->bv_page = pcl->compressed_bvecs[nr].page = zbv.page;
+	folio = page_folio(zbv.page);
+	/* first mark it as a temporary shortlived folio (now 1 ref) */
+	folio->private = (void *)Z_EROFS_SHORTLIVED_PAGE;
 	spin_unlock(&pcl->obj.lockref.lock);
-	bvec->bv_page = page;
 out_tocache:
 	if (!tocache || bs != PAGE_SIZE ||
-	    filemap_add_folio(mc, zbv.folio, pcl->obj.index + nr, gfp)) {
-		/* turn into a temporary shortlived folio (1 ref) */
-		zbv.folio->private = (void *)Z_EROFS_SHORTLIVED_PAGE;
+	    filemap_add_folio(mc, folio, pcl->obj.index + nr, gfp))
 		return;
-	}
-	folio_attach_private(zbv.folio, pcl);
+	folio_attach_private(folio, pcl);
 	/* drop a refcount added by allocpage (then 2 refs in total here) */
-	folio_put(zbv.folio);
+	folio_put(folio);
 }
 
 static struct z_erofs_decompressqueue *jobqueue_init(struct super_block *sb,
@@ -1767,7 +1776,6 @@ static void z_erofs_pcluster_readmore(struct z_erofs_decompress_frontend *f,
 		end = round_up(end, PAGE_SIZE);
 	} else {
 		end = round_up(map->m_la, PAGE_SIZE);
-
 		if (!map->m_llen)
 			return;
 	}
@@ -1775,15 +1783,15 @@ static void z_erofs_pcluster_readmore(struct z_erofs_decompress_frontend *f,
 	cur = map->m_la + map->m_llen - 1;
 	while ((cur >= end) && (cur < i_size_read(inode))) {
 		pgoff_t index = cur >> PAGE_SHIFT;
-		struct page *page;
+		struct folio *folio;
 
-		page = erofs_grab_cache_page_nowait(inode->i_mapping, index);
-		if (page) {
-			if (PageUptodate(page))
-				unlock_page(page);
+		folio = erofs_grab_folio_nowait(inode->i_mapping, index);
+		if (!IS_ERR_OR_NULL(folio)) {
+			if (folio_test_uptodate(folio))
+				folio_unlock(folio);
 			else
-				z_erofs_scan_folio(f, page_folio(page), !!rac);
-			put_page(page);
+				z_erofs_scan_folio(f, folio, !!rac);
+			folio_put(folio);
 		}
 
 		if (cur < PAGE_SIZE)