Merge tag 'erofs-for-5.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs

Pull erofs updates from Gao Xiang:
 "In this cycle, we would like to introduce a new feature called big
  pcluster so EROFS can compress file data into more than 1 fs block and
  different pcluster size can be selected for each (sub-)files by
  design.

  The current EROFS test results on my laptop are [1]:

    Testscript: erofs-openbenchmark
    Testdata: enwik9 (1000000000 bytes)
     ________________________________________________________________
    |  file system  |   size    | seq read | rand read | rand9m read |
    |_______________|___________|_ MiB/s __|__ MiB/s __|___ MiB/s ___|
    |___erofs_4k____|_556879872_|_ 781.4 __|__ 55.3 ___|___ 25.3  ___|
    |___erofs_16k___|_452509696_|_ 864.8 __|_ 123.2 ___|___ 20.8  ___|
    |___erofs_32k___|_415223808_|_ 899.8 __|_ 105.8 _*_|___ 16.8 ____|
    |___erofs_64k___|_393814016_|_ 906.6 __|__ 66.6 _*_|___ 11.8 ____|
    |__squashfs_8k__|_556191744_|_  64.9 __|__ 19.3 ___|____ 9.1 ____|
    |__squashfs_16k_|_502661120_|_  98.9 __|__ 38.0 ___|____ 9.8 ____|
    |__squashfs_32k_|_458784768_|_ 115.4 __|__ 71.6 _*_|___ 10.0 ____|
    |_squashfs_128k_|_398204928_|_ 257.2 __|_ 253.8 _*_|___ 10.9 ____|
    |____ext4_4k____|____()_____|_ 786.6 __|__ 28.6 ___|___ 27.8 ____|

  which has been verified but I'd like warn it as experimental for a
  while. This matches erofs-utils dev branch and I'll also release a new
  userspace version for this later.

  Apart from that, several improvements are also included: eg complete a
  missing case for inplace I/O, optimize endio decompression logic for
  non-atomic contexts and support adjustable sliding window size, ... In
  addition to those, there are some cleanups as always.

  Summary:

   - avoid memory failure when applying rolling decompression

   - optimize endio decompression logic for non-atomic contexts

   - complete a missing case which can be safely selected for inplace
     I/O and thus decreasing more memory footprint

   - check unsupported on-disk inode i_format strictly

   - support adjustable lz4 sliding window size to decrease runtime
     memory footprint

   - support on-disk compression configurations

   - support big pcluster decompression

   - several code cleanups / spelling correction"

* tag 'erofs-for-5.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs: (21 commits)
  erofs: enable big pcluster feature
  erofs: support decompress big pcluster for lz4 backend
  erofs: support parsing big pcluster compact indexes
  erofs: support parsing big pcluster compress indexes
  erofs: adjust per-CPU buffers according to max_pclusterblks
  erofs: add big physical cluster definition
  erofs: fix up inplace I/O pointer for big pcluster
  erofs: introduce physical cluster slab pools
  erofs: introduce multipage per-CPU buffers
  erofs: reserve physical_clusterbits[]
  erofs: Clean up spelling mistakes found in fs/erofs
  erofs: add on-disk compression configurations
  erofs: introduce on-disk lz4 fs configurations
  erofs: support adjust lz4 history window size
  erofs: introduce erofs_sb_has_xxx() helpers
  erofs: add unsupported inode i_format check
  erofs: don't use erofs_map_blocks() any more
  erofs: complete a missing case for inplace I/O
  erofs: use sync decompression for atomic contexts only
  erofs: use workqueue decompression for atomic contexts only
  ...

fs/erofs/Kconfig

@@ -76,17 +76,3 @@ config EROFS_FS_ZIP
 
 	  If you don't want to enable compression feature, say N.
 
-config EROFS_FS_CLUSTER_PAGE_LIMIT
-	int "EROFS Cluster Pages Hard Limit"
-	depends on EROFS_FS_ZIP
-	range 1 256
-	default "1"
-	help
-	  Indicates maximum # of pages of a compressed
-	  physical cluster.
-
-	  For example, if files in a image were compressed
-	  into 8k-unit, hard limit should not be configured
-	  less than 2. Otherwise, the image will be refused
-	  to mount on this kernel.
-

fs/erofs/Makefile

 # SPDX-License-Identifier: GPL-2.0-only
 
 obj-$(CONFIG_EROFS_FS) += erofs.o
-erofs-objs := super.o inode.o data.o namei.o dir.o utils.o
+erofs-objs := super.o inode.o data.o namei.o dir.o utils.o pcpubuf.o
 erofs-$(CONFIG_EROFS_FS_XATTR) += xattr.o
 erofs-$(CONFIG_EROFS_FS_ZIP) += decompressor.o zmap.o zdata.o

fs/erofs/data.c

@@ -109,21 +109,6 @@ static int erofs_map_blocks_flatmode(struct inode *inode,
 	return err;
 }
 
-int erofs_map_blocks(struct inode *inode,
-		     struct erofs_map_blocks *map, int flags)
-{
-	if (erofs_inode_is_data_compressed(EROFS_I(inode)->datalayout)) {
-		int err = z_erofs_map_blocks_iter(inode, map, flags);
-
-		if (map->mpage) {
-			put_page(map->mpage);
-			map->mpage = NULL;
-		}
-		return err;
-	}
-	return erofs_map_blocks_flatmode(inode, map, flags);
-}
-
 static inline struct bio *erofs_read_raw_page(struct bio *bio,
 					      struct address_space *mapping,
 					      struct page *page,
@@ -159,7 +144,7 @@ static inline struct bio *erofs_read_raw_page(struct bio *bio,
 		erofs_blk_t blknr;
 		unsigned int blkoff;
 
-		err = erofs_map_blocks(inode, &map, EROFS_GET_BLOCKS_RAW);
+		err = erofs_map_blocks_flatmode(inode, &map, EROFS_GET_BLOCKS_RAW);
 		if (err)
 			goto err_out;
 
@@ -318,7 +303,7 @@ static sector_t erofs_bmap(struct address_space *mapping, sector_t block)
 			return 0;
 	}
 
-	if (!erofs_map_blocks(inode, &map, EROFS_GET_BLOCKS_RAW))
+	if (!erofs_map_blocks_flatmode(inode, &map, EROFS_GET_BLOCKS_RAW))
 		return erofs_blknr(map.m_pa);
 
 	return 0;

fs/erofs/erofs_fs.h

@@ -18,15 +18,22 @@
  * be incompatible with this kernel version.
  */
 #define EROFS_FEATURE_INCOMPAT_LZ4_0PADDING	0x00000001
-#define EROFS_ALL_FEATURE_INCOMPAT		EROFS_FEATURE_INCOMPAT_LZ4_0PADDING
+#define EROFS_FEATURE_INCOMPAT_COMPR_CFGS	0x00000002
+#define EROFS_FEATURE_INCOMPAT_BIG_PCLUSTER	0x00000002
+#define EROFS_ALL_FEATURE_INCOMPAT		\
+	(EROFS_FEATURE_INCOMPAT_LZ4_0PADDING | \
+	 EROFS_FEATURE_INCOMPAT_COMPR_CFGS | \
+	 EROFS_FEATURE_INCOMPAT_BIG_PCLUSTER)
 
-/* 128-byte erofs on-disk super block */
+#define EROFS_SB_EXTSLOT_SIZE	16
+
+/* erofs on-disk super block (currently 128 bytes) */
 struct erofs_super_block {
 	__le32 magic;           /* file system magic number */
 	__le32 checksum;        /* crc32c(super_block) */
 	__le32 feature_compat;
 	__u8 blkszbits;         /* support block_size == PAGE_SIZE only */
-	__u8 reserved;
+	__u8 sb_extslots;	/* superblock size = 128 + sb_extslots * 16 */
 
 	__le16 root_nid;	/* nid of root directory */
 	__le64 inos;            /* total valid ino # (== f_files - f_favail) */
@@ -39,7 +46,13 @@ struct erofs_super_block {
 	__u8 uuid[16];          /* 128-bit uuid for volume */
 	__u8 volume_name[16];   /* volume name */
 	__le32 feature_incompat;
-	__u8 reserved2[44];
+	union {
+		/* bitmap for available compression algorithms */
+		__le16 available_compr_algs;
+		/* customized sliding window size instead of 64k by default */
+		__le16 lz4_max_distance;
+	} __packed u1;
+	__u8 reserved2[42];
 };
 
 /*
@@ -75,6 +88,9 @@ static inline bool erofs_inode_is_data_compressed(unsigned int datamode)
 #define EROFS_I_VERSION_BIT             0
 #define EROFS_I_DATALAYOUT_BIT          1
 
+#define EROFS_I_ALL	\
+	((1 << (EROFS_I_DATALAYOUT_BIT + EROFS_I_DATALAYOUT_BITS)) - 1)
+
 /* 32-byte reduced form of an ondisk inode */
 struct erofs_inode_compact {
 	__le16 i_format;	/* inode format hints */
@@ -189,20 +205,33 @@ static inline unsigned int erofs_xattr_entry_size(struct erofs_xattr_entry *e)
 				 e->e_name_len + le16_to_cpu(e->e_value_size));
 }
 
+/* maximum supported size of a physical compression cluster */
+#define Z_EROFS_PCLUSTER_MAX_SIZE	(1024 * 1024)
+
 /* available compression algorithm types (for h_algorithmtype) */
 enum {
 	Z_EROFS_COMPRESSION_LZ4	= 0,
 	Z_EROFS_COMPRESSION_MAX
 };
+#define Z_EROFS_ALL_COMPR_ALGS		(1 << (Z_EROFS_COMPRESSION_MAX - 1))
+
+/* 14 bytes (+ length field = 16 bytes) */
+struct z_erofs_lz4_cfgs {
+	__le16 max_distance;
+	__le16 max_pclusterblks;
+	u8 reserved[10];
+} __packed;
 
 /*
  * bit 0 : COMPACTED_2B indexes (0 - off; 1 - on)
  *  e.g. for 4k logical cluster size,      4B        if compacted 2B is off;
  *                                  (4B) + 2B + (4B) if compacted 2B is on.
+ * bit 1 : HEAD1 big pcluster (0 - off; 1 - on)
+ * bit 2 : HEAD2 big pcluster (0 - off; 1 - on)
  */
-#define Z_EROFS_ADVISE_COMPACTED_2B_BIT         0
-
-#define Z_EROFS_ADVISE_COMPACTED_2B     (1 << Z_EROFS_ADVISE_COMPACTED_2B_BIT)
+#define Z_EROFS_ADVISE_COMPACTED_2B		0x0001
+#define Z_EROFS_ADVISE_BIG_PCLUSTER_1		0x0002
+#define Z_EROFS_ADVISE_BIG_PCLUSTER_2		0x0004
 
 struct z_erofs_map_header {
 	__le32	h_reserved1;
@@ -214,9 +243,7 @@ struct z_erofs_map_header {
 	__u8	h_algorithmtype;
 	/*
 	 * bit 0-2 : logical cluster bits - 12, e.g. 0 for 4096;
-	 * bit 3-4 : (physical - logical) cluster bits of head 1:
-	 *       For example, if logical clustersize = 4096, 1 for 8192.
-	 * bit 5-7 : (physical - logical) cluster bits of head 2.
+	 * bit 3-7 : reserved.
 	 */
 	__u8	h_clusterbits;
 };
@@ -259,6 +286,13 @@ enum {
 #define Z_EROFS_VLE_DI_CLUSTER_TYPE_BITS        2
 #define Z_EROFS_VLE_DI_CLUSTER_TYPE_BIT         0
 
+/*
+ * D0_CBLKCNT will be marked _only_ at the 1st non-head lcluster to store the
+ * compressed block count of a compressed extent (in logical clusters, aka.
+ * block count of a pcluster).
+ */
+#define Z_EROFS_VLE_DI_D0_CBLKCNT		(1 << 11)
+
 struct z_erofs_vle_decompressed_index {
 	__le16 di_advise;
 	/* where to decompress in the head cluster */

fs/erofs/inode.c

@@ -44,6 +44,13 @@ static struct page *erofs_read_inode(struct inode *inode,
 	dic = page_address(page) + *ofs;
 	ifmt = le16_to_cpu(dic->i_format);
 
+	if (ifmt & ~EROFS_I_ALL) {
+		erofs_err(inode->i_sb, "unsupported i_format %u of nid %llu",
+			  ifmt, vi->nid);
+		err = -EOPNOTSUPP;
+		goto err_out;
+	}
+
 	vi->datalayout = erofs_inode_datalayout(ifmt);
 	if (vi->datalayout >= EROFS_INODE_DATALAYOUT_MAX) {
 		erofs_err(inode->i_sb, "unsupported datalayout %u of nid %llu",

fs/erofs/internal.h

@@ -50,6 +50,8 @@ struct erofs_fs_context {
 #ifdef CONFIG_EROFS_FS_ZIP
 	/* current strategy of how to use managed cache */
 	unsigned char cache_strategy;
+	/* strategy of sync decompression (false - auto, true - force on) */
+	bool readahead_sync_decompress;
 
 	/* threshold for decompression synchronously */
 	unsigned int max_sync_decompress_pages;
@@ -57,6 +59,14 @@ struct erofs_fs_context {
 	unsigned int mount_opt;
 };
 
+/* all filesystem-wide lz4 configurations */
+struct erofs_sb_lz4_info {
+	/* # of pages needed for EROFS lz4 rolling decompression */
+	u16 max_distance_pages;
+	/* maximum possible blocks for pclusters in the filesystem */
+	u16 max_pclusterblks;
+};
+
 struct erofs_sb_info {
 #ifdef CONFIG_EROFS_FS_ZIP
 	/* list for all registered superblocks, mainly for shrinker */
@@ -67,9 +77,12 @@ struct erofs_sb_info {
 	struct xarray managed_pslots;
 
 	unsigned int shrinker_run_no;
+	u16 available_compr_algs;
 
 	/* pseudo inode to manage cached pages */
 	struct inode *managed_cache;
+
+	struct erofs_sb_lz4_info lz4;
 #endif	/* CONFIG_EROFS_FS_ZIP */
 	u32 blocks;
 	u32 meta_blkaddr;
@@ -80,6 +93,7 @@ struct erofs_sb_info {
 	/* inode slot unit size in bit shift */
 	unsigned char islotbits;
 
+	u32 sb_size;			/* total superblock size */
 	u32 build_time_nsec;
 	u64 build_time;
 
@@ -182,12 +196,6 @@ static inline int erofs_wait_on_workgroup_freezed(struct erofs_workgroup *grp)
 	return v;
 }
 #endif	/* !CONFIG_SMP */
-
-/* hard limit of pages per compressed cluster */
-#define Z_EROFS_CLUSTER_MAX_PAGES       (CONFIG_EROFS_FS_CLUSTER_PAGE_LIMIT)
-#define EROFS_PCPUBUF_NR_PAGES          Z_EROFS_CLUSTER_MAX_PAGES
-#else
-#define EROFS_PCPUBUF_NR_PAGES          0
 #endif	/* !CONFIG_EROFS_FS_ZIP */
 
 /* we strictly follow PAGE_SIZE and no buffer head yet */
@@ -216,6 +224,17 @@ static inline erofs_off_t iloc(struct erofs_sb_info *sbi, erofs_nid_t nid)
 	return blknr_to_addr(sbi->meta_blkaddr) + (nid << sbi->islotbits);
 }
 
+#define EROFS_FEATURE_FUNCS(name, compat, feature) \
+static inline bool erofs_sb_has_##name(struct erofs_sb_info *sbi) \
+{ \
+	return sbi->feature_##compat & EROFS_FEATURE_##feature; \
+}
+
+EROFS_FEATURE_FUNCS(lz4_0padding, incompat, INCOMPAT_LZ4_0PADDING)
+EROFS_FEATURE_FUNCS(compr_cfgs, incompat, INCOMPAT_COMPR_CFGS)
+EROFS_FEATURE_FUNCS(big_pcluster, incompat, INCOMPAT_BIG_PCLUSTER)
+EROFS_FEATURE_FUNCS(sb_chksum, compat, COMPAT_SB_CHKSUM)
+
 /* atomic flag definitions */
 #define EROFS_I_EA_INITED_BIT	0
 #define EROFS_I_Z_INITED_BIT	1
@@ -244,7 +263,6 @@ struct erofs_inode {
 			unsigned short z_advise;
 			unsigned char  z_algorithmtype[2];
 			unsigned char  z_logical_clusterbits;
-			unsigned char  z_physical_clusterbits[2];
 		};
 #endif	/* CONFIG_EROFS_FS_ZIP */
 	};
@@ -287,7 +305,7 @@ extern const struct address_space_operations erofs_raw_access_aops;
 extern const struct address_space_operations z_erofs_aops;
 
 /*
- * Logical to physical block mapping, used by erofs_map_blocks()
+ * Logical to physical block mapping
  *
  * Different with other file systems, it is used for 2 access modes:
  *
@@ -334,7 +352,7 @@ struct erofs_map_blocks {
 	struct page *mpage;
 };
 
-/* Flags used by erofs_map_blocks() */
+/* Flags used by erofs_map_blocks_flatmode() */
 #define EROFS_GET_BLOCKS_RAW    0x0001
 
 /* zmap.c */
@@ -356,8 +374,6 @@ static inline int z_erofs_map_blocks_iter(struct inode *inode,
 /* data.c */
 struct page *erofs_get_meta_page(struct super_block *sb, erofs_blk_t blkaddr);
 
-int erofs_map_blocks(struct inode *, struct erofs_map_blocks *, int);
-
 /* inode.c */
 static inline unsigned long erofs_inode_hash(erofs_nid_t nid)
 {
@@ -386,23 +402,30 @@ int erofs_namei(struct inode *dir, struct qstr *name,
 /* dir.c */
 extern const struct file_operations erofs_dir_fops;
 
-/* utils.c / zdata.c */
-struct page *erofs_allocpage(struct list_head *pool, gfp_t gfp);
-
-#if (EROFS_PCPUBUF_NR_PAGES > 0)
-void *erofs_get_pcpubuf(unsigned int pagenr);
-#define erofs_put_pcpubuf(buf) do { \
-	(void)&(buf);	\
-	preempt_enable();	\
-} while (0)
-#else
-static inline void *erofs_get_pcpubuf(unsigned int pagenr)
+static inline void *erofs_vm_map_ram(struct page **pages, unsigned int count)
 {
-	return ERR_PTR(-EOPNOTSUPP);
+	int retried = 0;
+
+	while (1) {
+		void *p = vm_map_ram(pages, count, -1);
+
+		/* retry two more times (totally 3 times) */
+		if (p || ++retried >= 3)
+			return p;
+		vm_unmap_aliases();
+	}
+	return NULL;
 }
 
-#define erofs_put_pcpubuf(buf) do {} while (0)
-#endif
+/* pcpubuf.c */
+void *erofs_get_pcpubuf(unsigned int requiredpages);
+void erofs_put_pcpubuf(void *ptr);
+int erofs_pcpubuf_growsize(unsigned int nrpages);
+void erofs_pcpubuf_init(void);
+void erofs_pcpubuf_exit(void);
+
+/* utils.c / zdata.c */
+struct page *erofs_allocpage(struct list_head *pool, gfp_t gfp);
 
 #ifdef CONFIG_EROFS_FS_ZIP
 int erofs_workgroup_put(struct erofs_workgroup *grp);
@@ -421,6 +444,9 @@ int erofs_try_to_free_all_cached_pages(struct erofs_sb_info *sbi,
 				       struct erofs_workgroup *egrp);
 int erofs_try_to_free_cached_page(struct address_space *mapping,
 				  struct page *page);
+int z_erofs_load_lz4_config(struct super_block *sb,
+			    struct erofs_super_block *dsb,
+			    struct z_erofs_lz4_cfgs *lz4, int len);
 #else
 static inline void erofs_shrinker_register(struct super_block *sb) {}
 static inline void erofs_shrinker_unregister(struct super_block *sb) {}
@@ -428,6 +454,16 @@ static inline int erofs_init_shrinker(void) { return 0; }
 static inline void erofs_exit_shrinker(void) {}
 static inline int z_erofs_init_zip_subsystem(void) { return 0; }
 static inline void z_erofs_exit_zip_subsystem(void) {}
+static inline int z_erofs_load_lz4_config(struct super_block *sb,
+				  struct erofs_super_block *dsb,
+				  struct z_erofs_lz4_cfgs *lz4, int len)
+{
+	if (lz4 || dsb->u1.lz4_max_distance) {
+		erofs_err(sb, "lz4 algorithm isn't enabled");
+		return -EINVAL;
+	}
+	return 0;
+}
 #endif	/* !CONFIG_EROFS_FS_ZIP */
 
 #define EFSCORRUPTED    EUCLEAN         /* Filesystem is corrupted */

fs/erofs/pcpubuf.c

+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) Gao Xiang <xiang@kernel.org>
+ *
+ * For low-latency decompression algorithms (e.g. lz4), reserve consecutive
+ * per-CPU virtual memory (in pages) in advance to store such inplace I/O
+ * data if inplace decompression is failed (due to unmet inplace margin for
+ * example).
+ */
+#include "internal.h"
+
+struct erofs_pcpubuf {
+	raw_spinlock_t lock;
+	void *ptr;
+	struct page **pages;
+	unsigned int nrpages;
+};
+
+static DEFINE_PER_CPU(struct erofs_pcpubuf, erofs_pcb);
+
+void *erofs_get_pcpubuf(unsigned int requiredpages)
+	__acquires(pcb->lock)
+{
+	struct erofs_pcpubuf *pcb = &get_cpu_var(erofs_pcb);
+
+	raw_spin_lock(&pcb->lock);
+	/* check if the per-CPU buffer is too small */
+	if (requiredpages > pcb->nrpages) {
+		raw_spin_unlock(&pcb->lock);
+		put_cpu_var(erofs_pcb);
+		/* (for sparse checker) pretend pcb->lock is still taken */
+		__acquire(pcb->lock);
+		return NULL;
+	}
+	return pcb->ptr;
+}
+
+void erofs_put_pcpubuf(void *ptr) __releases(pcb->lock)
+{
+	struct erofs_pcpubuf *pcb = &per_cpu(erofs_pcb, smp_processor_id());
+
+	DBG_BUGON(pcb->ptr != ptr);
+	raw_spin_unlock(&pcb->lock);
+	put_cpu_var(erofs_pcb);
+}
+
+/* the next step: support per-CPU page buffers hotplug */
+int erofs_pcpubuf_growsize(unsigned int nrpages)
+{
+	static DEFINE_MUTEX(pcb_resize_mutex);
+	static unsigned int pcb_nrpages;
+	LIST_HEAD(pagepool);
+	int delta, cpu, ret, i;
+
+	mutex_lock(&pcb_resize_mutex);
+	delta = nrpages - pcb_nrpages;
+	ret = 0;
+	/* avoid shrinking pcpubuf, since no idea how many fses rely on */
+	if (delta <= 0)
+		goto out;
+
+	for_each_possible_cpu(cpu) {
+		struct erofs_pcpubuf *pcb = &per_cpu(erofs_pcb, cpu);
+		struct page **pages, **oldpages;
+		void *ptr, *old_ptr;
+
+		pages = kmalloc_array(nrpages, sizeof(*pages), GFP_KERNEL);
+		if (!pages) {
+			ret = -ENOMEM;
+			break;
+		}
+
+		for (i = 0; i < nrpages; ++i) {
+			pages[i] = erofs_allocpage(&pagepool, GFP_KERNEL);
+			if (!pages[i]) {
+				ret = -ENOMEM;
+				oldpages = pages;
+				goto free_pagearray;
+			}
+		}
+		ptr = vmap(pages, nrpages, VM_MAP, PAGE_KERNEL);
+		if (!ptr) {
+			ret = -ENOMEM;
+			oldpages = pages;
+			goto free_pagearray;
+		}
+		raw_spin_lock(&pcb->lock);
+		old_ptr = pcb->ptr;
+		pcb->ptr = ptr;
+		oldpages = pcb->pages;
+		pcb->pages = pages;
+		i = pcb->nrpages;
+		pcb->nrpages = nrpages;
+		raw_spin_unlock(&pcb->lock);
+
+		if (!oldpages) {
+			DBG_BUGON(old_ptr);
+			continue;
+		}
+
+		if (old_ptr)
+			vunmap(old_ptr);
+free_pagearray:
+		while (i)
+			list_add(&oldpages[--i]->lru, &pagepool);
+		kfree(oldpages);
+		if (ret)
+			break;
+	}
+	pcb_nrpages = nrpages;
+	put_pages_list(&pagepool);
+out:
+	mutex_unlock(&pcb_resize_mutex);
+	return ret;
+}
+
+void erofs_pcpubuf_init(void)
+{
+	int cpu;
+
+	for_each_possible_cpu(cpu) {
+		struct erofs_pcpubuf *pcb = &per_cpu(erofs_pcb, cpu);
+
+		raw_spin_lock_init(&pcb->lock);
+	}
+}
+
+void erofs_pcpubuf_exit(void)
+{
+	int cpu, i;
+
+	for_each_possible_cpu(cpu) {
+		struct erofs_pcpubuf *pcb = &per_cpu(erofs_pcb, cpu);
+
+		if (pcb->ptr) {
+			vunmap(pcb->ptr);
+			pcb->ptr = NULL;
+		}
+		if (!pcb->pages)
+			continue;
+
+		for (i = 0; i < pcb->nrpages; ++i)
+			if (pcb->pages[i])
+				put_page(pcb->pages[i]);
+		kfree(pcb->pages);
+		pcb->pages = NULL;
+	}
+}

fs/erofs/super.c

@@ -122,6 +122,136 @@ static bool check_layout_compatibility(struct super_block *sb,
 	return true;
 }
 
+#ifdef CONFIG_EROFS_FS_ZIP
+/* read variable-sized metadata, offset will be aligned by 4-byte */
+static void *erofs_read_metadata(struct super_block *sb, struct page **pagep,
+				 erofs_off_t *offset, int *lengthp)
+{
+	struct page *page = *pagep;
+	u8 *buffer, *ptr;
+	int len, i, cnt;
+	erofs_blk_t blk;
+
+	*offset = round_up(*offset, 4);
+	blk = erofs_blknr(*offset);
+
+	if (!page || page->index != blk) {
+		if (page) {
+			unlock_page(page);
+			put_page(page);
+		}
+		page = erofs_get_meta_page(sb, blk);
+		if (IS_ERR(page))
+			goto err_nullpage;
+	}
+
+	ptr = kmap(page);
+	len = le16_to_cpu(*(__le16 *)&ptr[erofs_blkoff(*offset)]);
+	if (!len)
+		len = U16_MAX + 1;
+	buffer = kmalloc(len, GFP_KERNEL);
+	if (!buffer) {
+		buffer = ERR_PTR(-ENOMEM);
+		goto out;
+	}
+	*offset += sizeof(__le16);
+	*lengthp = len;
+
+	for (i = 0; i < len; i += cnt) {
+		cnt = min(EROFS_BLKSIZ - (int)erofs_blkoff(*offset), len - i);
+		blk = erofs_blknr(*offset);
+
+		if (!page || page->index != blk) {
+			if (page) {
+				kunmap(page);
+				unlock_page(page);
+				put_page(page);
+			}
+			page = erofs_get_meta_page(sb, blk);
+			if (IS_ERR(page)) {
+				kfree(buffer);
+				goto err_nullpage;
+			}
+			ptr = kmap(page);
+		}
+		memcpy(buffer + i, ptr + erofs_blkoff(*offset), cnt);
+		*offset += cnt;
+	}
+out:
+	kunmap(page);
+	*pagep = page;
+	return buffer;
+err_nullpage:
+	*pagep = NULL;
+	return page;
+}
+
+static int erofs_load_compr_cfgs(struct super_block *sb,
+				 struct erofs_super_block *dsb)
+{
+	struct erofs_sb_info *sbi;
+	struct page *page;
+	unsigned int algs, alg;
+	erofs_off_t offset;
+	int size, ret;
+
+	sbi = EROFS_SB(sb);
+	sbi->available_compr_algs = le16_to_cpu(dsb->u1.available_compr_algs);
+
+	if (sbi->available_compr_algs & ~Z_EROFS_ALL_COMPR_ALGS) {
+		erofs_err(sb, "try to load compressed fs with unsupported algorithms %x",
+			  sbi->available_compr_algs & ~Z_EROFS_ALL_COMPR_ALGS);
+		return -EINVAL;
+	}
+
+	offset = EROFS_SUPER_OFFSET + sbi->sb_size;
+	page = NULL;
+	alg = 0;
+	ret = 0;
+
+	for (algs = sbi->available_compr_algs; algs; algs >>= 1, ++alg) {
+		void *data;
+
+		if (!(algs & 1))
+			continue;
+
+		data = erofs_read_metadata(sb, &page, &offset, &size);
+		if (IS_ERR(data)) {
+			ret = PTR_ERR(data);
+			goto err;
+		}
+
+		switch (alg) {
+		case Z_EROFS_COMPRESSION_LZ4:
+			ret = z_erofs_load_lz4_config(sb, dsb, data, size);
+			break;
+		default:
+			DBG_BUGON(1);
+			ret = -EFAULT;
+		}
+		kfree(data);
+		if (ret)
+			goto err;
+	}
+err:
+	if (page) {
+		unlock_page(page);
+		put_page(page);
+	}
+	return ret;
+}
+#else
+static int erofs_load_compr_cfgs(struct super_block *sb,
+				 struct erofs_super_block *dsb)
+{
+	if (dsb->u1.available_compr_algs) {
+		erofs_err(sb, "try to load compressed fs when compression is disabled");
+		return -EINVAL;
+	}
+	return 0;
+}
+#endif
+
 static int erofs_read_superblock(struct super_block *sb)
 {
 	struct erofs_sb_info *sbi;
@@ -149,7 +279,7 @@ static int erofs_read_superblock(struct super_block *sb)
 	}
 
 	sbi->feature_compat = le32_to_cpu(dsb->feature_compat);
-	if (sbi->feature_compat & EROFS_FEATURE_COMPAT_SB_CHKSUM) {
+	if (erofs_sb_has_sb_chksum(sbi)) {
 		ret = erofs_superblock_csum_verify(sb, data);
 		if (ret)
 			goto out;
@@ -166,6 +296,12 @@ static int erofs_read_superblock(struct super_block *sb)
 	if (!check_layout_compatibility(sb, dsb))
 		goto out;
 
+	sbi->sb_size = 128 + dsb->sb_extslots * EROFS_SB_EXTSLOT_SIZE;
+	if (sbi->sb_size > EROFS_BLKSIZ) {
+		erofs_err(sb, "invalid sb_extslots %u (more than a fs block)",
+			  sbi->sb_size);
+		goto out;
+	}
 	sbi->blocks = le32_to_cpu(dsb->blocks);
 	sbi->meta_blkaddr = le32_to_cpu(dsb->meta_blkaddr);
 #ifdef CONFIG_EROFS_FS_XATTR
@@ -187,7 +323,12 @@ static int erofs_read_superblock(struct super_block *sb)
 		ret = -EFSCORRUPTED;
 		goto out;
 	}
-	ret = 0;
+
+	/* parse on-disk compression configurations */
+	if (erofs_sb_has_compr_cfgs(sbi))
+		ret = erofs_load_compr_cfgs(sb, dsb);
+	else
+		ret = z_erofs_load_lz4_config(sb, dsb, NULL, 0);
 out:
 	kunmap(page);
 	put_page(page);
@@ -200,6 +341,7 @@ static void erofs_default_options(struct erofs_fs_context *ctx)
 #ifdef CONFIG_EROFS_FS_ZIP
 	ctx->cache_strategy = EROFS_ZIP_CACHE_READAROUND;
 	ctx->max_sync_decompress_pages = 3;
+	ctx->readahead_sync_decompress = false;
 #endif
 #ifdef CONFIG_EROFS_FS_XATTR
 	set_opt(ctx, XATTR_USER);
@@ -513,6 +655,7 @@ static int __init erofs_module_init(void)
 	if (err)
 		goto shrinker_err;
 
+	erofs_pcpubuf_init();
 	err = z_erofs_init_zip_subsystem();
 	if (err)
 		goto zip_err;
@@ -542,6 +685,7 @@ static void __exit erofs_module_exit(void)
 	/* Ensure all RCU free inodes are safe before cache is destroyed. */
 	rcu_barrier();
 	kmem_cache_destroy(erofs_inode_cachep);
+	erofs_pcpubuf_exit();
 }
 
 /* get filesystem statistics */

fs/erofs/utils.c

@@ -21,18 +21,6 @@ struct page *erofs_allocpage(struct list_head *pool, gfp_t gfp)
 	return page;
 }
 
-#if (EROFS_PCPUBUF_NR_PAGES > 0)
-static struct {
-	u8 data[PAGE_SIZE * EROFS_PCPUBUF_NR_PAGES];
-} ____cacheline_aligned_in_smp erofs_pcpubuf[NR_CPUS];
-
-void *erofs_get_pcpubuf(unsigned int pagenr)
-{
-	preempt_disable();
-	return &erofs_pcpubuf[smp_processor_id()].data[pagenr * PAGE_SIZE];
-}
-#endif
-
 #ifdef CONFIG_EROFS_FS_ZIP
 /* global shrink count (for all mounted EROFS instances) */
 static atomic_long_t erofs_global_shrink_cnt;

fs/erofs/zdata.h

@@ -10,6 +10,7 @@
 #include "internal.h"
 #include "zpvec.h"
 
+#define Z_EROFS_PCLUSTER_MAX_PAGES	(Z_EROFS_PCLUSTER_MAX_SIZE / PAGE_SIZE)
 #define Z_EROFS_NR_INLINE_PAGEVECS      3
 
 /*
@@ -59,16 +60,17 @@ struct z_erofs_pcluster {
 	/* A: point to next chained pcluster or TAILs */
 	z_erofs_next_pcluster_t next;
 
-	/* A: compressed pages (including multi-usage pages) */
-	struct page *compressed_pages[Z_EROFS_CLUSTER_MAX_PAGES];
-
 	/* A: lower limit of decompressed length and if full length or not */
 	unsigned int length;
 
+	/* I: physical cluster size in pages */
+	unsigned short pclusterpages;
+
 	/* I: compression algorithm format */
 	unsigned char algorithmformat;
-	/* I: bit shift of physical cluster size */
-	unsigned char clusterbits;
+
+	/* A: compressed pages (can be cached or inplaced pages) */
+	struct page *compressed_pages[];
 };
 
 #define z_erofs_primarycollection(pcluster) (&(pcluster)->primary_collection)
@@ -82,8 +84,6 @@ struct z_erofs_pcluster {
 
 #define Z_EROFS_PCLUSTER_NIL            (NULL)
 
-#define Z_EROFS_WORKGROUP_SIZE  sizeof(struct z_erofs_pcluster)
-
 struct z_erofs_decompressqueue {
 	struct super_block *sb;
 	atomic_t pending_bios;

fs/erofs/zmap.c

@@ -11,17 +11,16 @@
 int z_erofs_fill_inode(struct inode *inode)
 {
 	struct erofs_inode *const vi = EROFS_I(inode);
+	struct erofs_sb_info *sbi = EROFS_SB(inode->i_sb);
 
-	if (vi->datalayout == EROFS_INODE_FLAT_COMPRESSION_LEGACY) {
+	if (!erofs_sb_has_big_pcluster(sbi) &&
+	    vi->datalayout == EROFS_INODE_FLAT_COMPRESSION_LEGACY) {
 		vi->z_advise = 0;
 		vi->z_algorithmtype[0] = 0;
 		vi->z_algorithmtype[1] = 0;
 		vi->z_logical_clusterbits = LOG_BLOCK_SIZE;
-		vi->z_physical_clusterbits[0] = vi->z_logical_clusterbits;
-		vi->z_physical_clusterbits[1] = vi->z_logical_clusterbits;
 		set_bit(EROFS_I_Z_INITED_BIT, &vi->flags);
 	}
-
 	inode->i_mapping->a_ops = &z_erofs_aops;
 	return 0;
 }
@@ -52,7 +51,8 @@ static int z_erofs_fill_inode_lazy(struct inode *inode)
 	if (test_bit(EROFS_I_Z_INITED_BIT, &vi->flags))
 		goto out_unlock;
 
-	DBG_BUGON(vi->datalayout == EROFS_INODE_FLAT_COMPRESSION_LEGACY);
+	DBG_BUGON(!erofs_sb_has_big_pcluster(EROFS_SB(sb)) &&
+		  vi->datalayout == EROFS_INODE_FLAT_COMPRESSION_LEGACY);
 
 	pos = ALIGN(iloc(EROFS_SB(sb), vi->nid) + vi->inode_isize +
 		    vi->xattr_isize, 8);
@@ -77,18 +77,22 @@ static int z_erofs_fill_inode_lazy(struct inode *inode)
 	}
 
 	vi->z_logical_clusterbits = LOG_BLOCK_SIZE + (h->h_clusterbits & 7);
-	vi->z_physical_clusterbits[0] = vi->z_logical_clusterbits +
-					((h->h_clusterbits >> 3) & 3);
-
-	if (vi->z_physical_clusterbits[0] != LOG_BLOCK_SIZE) {
-		erofs_err(sb, "unsupported physical clusterbits %u for nid %llu, please upgrade kernel",
-			  vi->z_physical_clusterbits[0], vi->nid);
-		err = -EOPNOTSUPP;
+	if (!erofs_sb_has_big_pcluster(EROFS_SB(sb)) &&
+	    vi->z_advise & (Z_EROFS_ADVISE_BIG_PCLUSTER_1 |
+			    Z_EROFS_ADVISE_BIG_PCLUSTER_2)) {
+		erofs_err(sb, "per-inode big pcluster without sb feature for nid %llu",
+			  vi->nid);
+		err = -EFSCORRUPTED;
+		goto unmap_done;
+	}
+	if (vi->datalayout == EROFS_INODE_FLAT_COMPRESSION &&
+	    !(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_1) ^
+	    !(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_2)) {
+		erofs_err(sb, "big pcluster head1/2 of compact indexes should be consistent for nid %llu",
+			  vi->nid);
+		err = -EFSCORRUPTED;
 		goto unmap_done;
 	}
-
-	vi->z_physical_clusterbits[1] = vi->z_logical_clusterbits +
-					((h->h_clusterbits >> 5) & 7);
 	/* paired with smp_mb() at the beginning of the function */
 	smp_mb();
 	set_bit(EROFS_I_Z_INITED_BIT, &vi->flags);
@@ -111,7 +115,7 @@ struct z_erofs_maprecorder {
 	u8  type;
 	u16 clusterofs;
 	u16 delta[2];
-	erofs_blk_t pblk;
+	erofs_blk_t pblk, compressedlcs;
 };
 
 static int z_erofs_reload_indexes(struct z_erofs_maprecorder *m,
@@ -174,6 +178,15 @@ static int legacy_load_cluster_from_disk(struct z_erofs_maprecorder *m,
 	case Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD:
 		m->clusterofs = 1 << vi->z_logical_clusterbits;
 		m->delta[0] = le16_to_cpu(di->di_u.delta[0]);
+		if (m->delta[0] & Z_EROFS_VLE_DI_D0_CBLKCNT) {
+			if (!(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_1)) {
+				DBG_BUGON(1);
+				return -EFSCORRUPTED;
+			}
+			m->compressedlcs = m->delta[0] &
+				~Z_EROFS_VLE_DI_D0_CBLKCNT;
+			m->delta[0] = 1;
+		}
 		m->delta[1] = le16_to_cpu(di->di_u.delta[1]);
 		break;
 	case Z_EROFS_VLE_CLUSTER_TYPE_PLAIN:
@@ -210,6 +223,7 @@ static int unpack_compacted_index(struct z_erofs_maprecorder *m,
 	unsigned int vcnt, base, lo, encodebits, nblk;
 	int i;
 	u8 *in, type;
+	bool big_pcluster;
 
 	if (1 << amortizedshift == 4)
 		vcnt = 2;
@@ -218,6 +232,7 @@ static int unpack_compacted_index(struct z_erofs_maprecorder *m,
 	else
 		return -EOPNOTSUPP;
 
+	big_pcluster = vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_1;
 	encodebits = ((vcnt << amortizedshift) - sizeof(__le32)) * 8 / vcnt;
 	base = round_down(eofs, vcnt << amortizedshift);
 	in = m->kaddr + base;
@@ -229,7 +244,15 @@ static int unpack_compacted_index(struct z_erofs_maprecorder *m,
 	m->type = type;
 	if (type == Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD) {
 		m->clusterofs = 1 << lclusterbits;
-		if (i + 1 != vcnt) {
+		if (lo & Z_EROFS_VLE_DI_D0_CBLKCNT) {
+			if (!big_pcluster) {
+				DBG_BUGON(1);
+				return -EFSCORRUPTED;
+			}
+			m->compressedlcs = lo & ~Z_EROFS_VLE_DI_D0_CBLKCNT;
+			m->delta[0] = 1;
+			return 0;
+		} else if (i + 1 != (int)vcnt) {
 			m->delta[0] = lo;
 			return 0;
 		}
@@ -242,22 +265,48 @@ static int unpack_compacted_index(struct z_erofs_maprecorder *m,
 					  in, encodebits * (i - 1), &type);
 		if (type != Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD)
 			lo = 0;
+		else if (lo & Z_EROFS_VLE_DI_D0_CBLKCNT)
+			lo = 1;
 		m->delta[0] = lo + 1;
 		return 0;
 	}
 	m->clusterofs = lo;
 	m->delta[0] = 0;
 	/* figout out blkaddr (pblk) for HEAD lclusters */
-	nblk = 1;
-	while (i > 0) {
-		--i;
-		lo = decode_compactedbits(lclusterbits, lomask,
-					  in, encodebits * i, &type);
-		if (type == Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD)
-			i -= lo;
-
-		if (i >= 0)
+	if (!big_pcluster) {
+		nblk = 1;
+		while (i > 0) {
+			--i;
+			lo = decode_compactedbits(lclusterbits, lomask,
+						  in, encodebits * i, &type);
+			if (type == Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD)
+				i -= lo;
+
+			if (i >= 0)
+				++nblk;
+		}
+	} else {
+		nblk = 0;
+		while (i > 0) {
+			--i;
+			lo = decode_compactedbits(lclusterbits, lomask,
+						  in, encodebits * i, &type);
+			if (type == Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD) {
+				if (lo & Z_EROFS_VLE_DI_D0_CBLKCNT) {
+					--i;
+					nblk += lo & ~Z_EROFS_VLE_DI_D0_CBLKCNT;
+					continue;
+				}
+				/* bigpcluster shouldn't have plain d0 == 1 */
+				if (lo <= 1) {
+					DBG_BUGON(1);
+					return -EFSCORRUPTED;
+				}
+				i -= lo - 2;
+				continue;
+			}
 			++nblk;
+		}
 	}
 	in += (vcnt << amortizedshift) - sizeof(__le32);
 	m->pblk = le32_to_cpu(*(__le32 *)in) + nblk;
@@ -381,6 +430,58 @@ static int z_erofs_extent_lookback(struct z_erofs_maprecorder *m,
 	return 0;
 }
 
+static int z_erofs_get_extent_compressedlen(struct z_erofs_maprecorder *m,
+					    unsigned int initial_lcn)
+{
+	struct erofs_inode *const vi = EROFS_I(m->inode);
+	struct erofs_map_blocks *const map = m->map;
+	const unsigned int lclusterbits = vi->z_logical_clusterbits;
+	unsigned long lcn;
+	int err;
+
+	DBG_BUGON(m->type != Z_EROFS_VLE_CLUSTER_TYPE_PLAIN &&
+		  m->type != Z_EROFS_VLE_CLUSTER_TYPE_HEAD);
+	if (!(map->m_flags & EROFS_MAP_ZIPPED) ||
+	    !(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_1)) {
+		map->m_plen = 1 << lclusterbits;
+		return 0;
+	}
+
+	lcn = m->lcn + 1;
+	if (m->compressedlcs)
+		goto out;
+	if (lcn == initial_lcn)
+		goto err_bonus_cblkcnt;
+
+	err = z_erofs_load_cluster_from_disk(m, lcn);
+	if (err)
+		return err;
+
+	switch (m->type) {
+	case Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD:
+		if (m->delta[0] != 1)
+			goto err_bonus_cblkcnt;
+		if (m->compressedlcs)
+			break;
+		fallthrough;
+	default:
+		erofs_err(m->inode->i_sb,
+			  "cannot found CBLKCNT @ lcn %lu of nid %llu",
+			  lcn, vi->nid);
+		DBG_BUGON(1);
+		return -EFSCORRUPTED;
+	}
+out:
+	map->m_plen = m->compressedlcs << lclusterbits;
+	return 0;
+err_bonus_cblkcnt:
+	erofs_err(m->inode->i_sb,
+		  "bogus CBLKCNT @ lcn %lu of nid %llu",
+		  lcn, vi->nid);
+	DBG_BUGON(1);
+	return -EFSCORRUPTED;
+}
+
 int z_erofs_map_blocks_iter(struct inode *inode,
 			    struct erofs_map_blocks *map,
 			    int flags)
@@ -392,6 +493,7 @@ int z_erofs_map_blocks_iter(struct inode *inode,
 	};
 	int err = 0;
 	unsigned int lclusterbits, endoff;
+	unsigned long initial_lcn;
 	unsigned long long ofs, end;
 
 	trace_z_erofs_map_blocks_iter_enter(inode, map, flags);
@@ -410,10 +512,10 @@ int z_erofs_map_blocks_iter(struct inode *inode,
 
 	lclusterbits = vi->z_logical_clusterbits;
 	ofs = map->m_la;
-	m.lcn = ofs >> lclusterbits;
+	initial_lcn = ofs >> lclusterbits;
 	endoff = ofs & ((1 << lclusterbits) - 1);
 
-	err = z_erofs_load_cluster_from_disk(&m, m.lcn);
+	err = z_erofs_load_cluster_from_disk(&m, initial_lcn);
 	if (err)
 		goto unmap_out;
 
@@ -443,7 +545,7 @@ int z_erofs_map_blocks_iter(struct inode *inode,
 		m.delta[0] = 1;
 		fallthrough;
 	case Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD:
-		/* get the correspoinding first chunk */
+		/* get the corresponding first chunk */
 		err = z_erofs_extent_lookback(&m, m.delta[0]);
 		if (err)
 			goto unmap_out;
@@ -457,10 +559,12 @@ int z_erofs_map_blocks_iter(struct inode *inode,
 	}
 
 	map->m_llen = end - map->m_la;
-	map->m_plen = 1 << lclusterbits;
 	map->m_pa = blknr_to_addr(m.pblk);
 	map->m_flags |= EROFS_MAP_MAPPED;
 
+	err = z_erofs_get_extent_compressedlen(&m, initial_lcn);
+	if (err)
+		goto out;
 unmap_out:
 	if (m.kaddr)
 		kunmap_atomic(m.kaddr);