diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2020-04-07 13:48:26 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2020-04-07 13:48:26 -0700 |
commit | f40f31cadc0ea5dcdd224c8b324add26469c2379 (patch) | |
tree | f3343b39982a438a1753726f2f0927753b7fa770 /fs/f2fs | |
parent | 763dede1b24886d327bfaed7cf59ee3c01c7913e (diff) | |
parent | 531dfae52e8c386f72fc4935b97486409291fdb1 (diff) |
Merge tag 'f2fs-for-5.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs
Pull f2fs updates from Jaegeuk Kim:
"In this round, we've mainly focused on fixing bugs and addressing
issues in recently introduced compression support.
Enhancement:
- add zstd support, and set LZ4 by default
- add ioctl() to show # of compressed blocks
- show mount time in debugfs
- replace rwsem with spinlock
- avoid lock contention in DIO reads
Some major bug fixes wrt compression:
- compressed block count
- memory access and leak
- remove obsolete fields
- flag controls
Other bug fixes and clean ups:
- fix overflow when handling .flags in inode_info
- fix SPO issue during resize FS flow
- fix compression with fsverity enabled
- potential deadlock when writing compressed pages
- show missing mount options"
* tag 'f2fs-for-5.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (66 commits)
f2fs: keep inline_data when compression conversion
f2fs: fix to disable compression on directory
f2fs: add missing CONFIG_F2FS_FS_COMPRESSION
f2fs: switch discard_policy.timeout to bool type
f2fs: fix to verify tpage before releasing in f2fs_free_dic()
f2fs: show compression in statx
f2fs: clean up dic->tpages assignment
f2fs: compress: support zstd compress algorithm
f2fs: compress: add .{init,destroy}_decompress_ctx callback
f2fs: compress: fix to call missing destroy_compress_ctx()
f2fs: change default compression algorithm
f2fs: clean up {cic,dic}.ref handling
f2fs: fix to use f2fs_readpage_limit() in f2fs_read_multi_pages()
f2fs: xattr.h: Make stub helpers inline
f2fs: fix to avoid double unlock
f2fs: fix potential .flags overflow on 32bit architecture
f2fs: fix NULL pointer dereference in f2fs_verity_work()
f2fs: fix to clear PG_error if fsverity failed
f2fs: don't call fscrypt_get_encryption_info() explicitly in f2fs_tmpfile()
f2fs: don't trigger data flush in foreground operation
...
Diffstat (limited to 'fs/f2fs')
-rw-r--r-- | fs/f2fs/Kconfig | 9 | ||||
-rw-r--r-- | fs/f2fs/checkpoint.c | 42 | ||||
-rw-r--r-- | fs/f2fs/compress.c | 317 | ||||
-rw-r--r-- | fs/f2fs/data.c | 141 | ||||
-rw-r--r-- | fs/f2fs/debug.c | 3 | ||||
-rw-r--r-- | fs/f2fs/dir.c | 16 | ||||
-rw-r--r-- | fs/f2fs/f2fs.h | 206 | ||||
-rw-r--r-- | fs/f2fs/file.c | 91 | ||||
-rw-r--r-- | fs/f2fs/gc.c | 51 | ||||
-rw-r--r-- | fs/f2fs/inode.c | 29 | ||||
-rw-r--r-- | fs/f2fs/namei.c | 12 | ||||
-rw-r--r-- | fs/f2fs/node.c | 33 | ||||
-rw-r--r-- | fs/f2fs/recovery.c | 12 | ||||
-rw-r--r-- | fs/f2fs/segment.c | 54 | ||||
-rw-r--r-- | fs/f2fs/segment.h | 2 | ||||
-rw-r--r-- | fs/f2fs/shrinker.c | 2 | ||||
-rw-r--r-- | fs/f2fs/super.c | 89 | ||||
-rw-r--r-- | fs/f2fs/sysfs.c | 50 | ||||
-rw-r--r-- | fs/f2fs/xattr.c | 67 | ||||
-rw-r--r-- | fs/f2fs/xattr.h | 9 |
20 files changed, 810 insertions, 425 deletions
diff --git a/fs/f2fs/Kconfig b/fs/f2fs/Kconfig index f0faada30f30..bb68d21e1f8c 100644 --- a/fs/f2fs/Kconfig +++ b/fs/f2fs/Kconfig @@ -118,3 +118,12 @@ config F2FS_FS_LZ4 default y help Support LZ4 compress algorithm, if unsure, say Y. + +config F2FS_FS_ZSTD + bool "ZSTD compression support" + depends on F2FS_FS_COMPRESSION + select ZSTD_COMPRESS + select ZSTD_DECOMPRESS + default y + help + Support ZSTD compress algorithm, if unsure, say Y. diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 44e84ac5c941..852890b72d6a 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -50,9 +50,6 @@ repeat: return page; } -/* - * We guarantee no failure on the returned page. - */ static struct page *__get_meta_page(struct f2fs_sb_info *sbi, pgoff_t index, bool is_meta) { @@ -206,7 +203,7 @@ bool f2fs_is_valid_blkaddr(struct f2fs_sb_info *sbi, } /* - * Readahead CP/NAT/SIT/SSA pages + * Readahead CP/NAT/SIT/SSA/POR pages */ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages, int type, bool sync) @@ -898,7 +895,7 @@ int f2fs_get_valid_checkpoint(struct f2fs_sb_info *sbi) return -ENOMEM; /* * Finding out valid cp block involves read both - * sets( cp pack1 and cp pack 2) + * sets( cp pack 1 and cp pack 2) */ cp_start_blk_no = le32_to_cpu(fsb->cp_blkaddr); cp1 = validate_checkpoint(sbi, cp_start_blk_no, &cp1_version); @@ -1250,20 +1247,20 @@ static void unblock_operations(struct f2fs_sb_info *sbi) f2fs_unlock_all(sbi); } -void f2fs_wait_on_all_pages_writeback(struct f2fs_sb_info *sbi) +void f2fs_wait_on_all_pages(struct f2fs_sb_info *sbi, int type) { DEFINE_WAIT(wait); for (;;) { prepare_to_wait(&sbi->cp_wait, &wait, TASK_UNINTERRUPTIBLE); - if (!get_pages(sbi, F2FS_WB_CP_DATA)) + if (!get_pages(sbi, type)) break; if (unlikely(f2fs_cp_error(sbi))) break; - io_schedule_timeout(5*HZ); + io_schedule_timeout(DEFAULT_IO_TIMEOUT); } finish_wait(&sbi->cp_wait, &wait); } @@ -1301,10 +1298,14 @@ static void update_ckpt_flags(struct f2fs_sb_info *sbi, struct cp_control *cpc) else __clear_ckpt_flags(ckpt, CP_ORPHAN_PRESENT_FLAG); - if (is_sbi_flag_set(sbi, SBI_NEED_FSCK) || - is_sbi_flag_set(sbi, SBI_IS_RESIZEFS)) + if (is_sbi_flag_set(sbi, SBI_NEED_FSCK)) __set_ckpt_flags(ckpt, CP_FSCK_FLAG); + if (is_sbi_flag_set(sbi, SBI_IS_RESIZEFS)) + __set_ckpt_flags(ckpt, CP_RESIZEFS_FLAG); + else + __clear_ckpt_flags(ckpt, CP_RESIZEFS_FLAG); + if (is_sbi_flag_set(sbi, SBI_CP_DISABLED)) __set_ckpt_flags(ckpt, CP_DISABLED_FLAG); else @@ -1384,13 +1385,8 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) /* Flush all the NAT/SIT pages */ f2fs_sync_meta_pages(sbi, META, LONG_MAX, FS_CP_META_IO); - f2fs_bug_on(sbi, get_pages(sbi, F2FS_DIRTY_META) && - !f2fs_cp_error(sbi)); - /* - * modify checkpoint - * version number is already updated - */ + /* start to update checkpoint, cp ver is already updated previously */ ckpt->elapsed_time = cpu_to_le64(get_mtime(sbi, true)); ckpt->free_segment_count = cpu_to_le32(free_segments(sbi)); for (i = 0; i < NR_CURSEG_NODE_TYPE; i++) { @@ -1493,11 +1489,11 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) /* Here, we have one bio having CP pack except cp pack 2 page */ f2fs_sync_meta_pages(sbi, META, LONG_MAX, FS_CP_META_IO); - f2fs_bug_on(sbi, get_pages(sbi, F2FS_DIRTY_META) && - !f2fs_cp_error(sbi)); + /* Wait for all dirty meta pages to be submitted for IO */ + f2fs_wait_on_all_pages(sbi, F2FS_DIRTY_META); /* wait for previous submitted meta pages writeback */ - f2fs_wait_on_all_pages_writeback(sbi); + f2fs_wait_on_all_pages(sbi, F2FS_WB_CP_DATA); /* flush all device cache */ err = f2fs_flush_device_cache(sbi); @@ -1506,7 +1502,7 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) /* barrier and flush checkpoint cp pack 2 page if it can */ commit_checkpoint(sbi, ckpt, start_blk); - f2fs_wait_on_all_pages_writeback(sbi); + f2fs_wait_on_all_pages(sbi, F2FS_WB_CP_DATA); /* * invalidate intermediate page cache borrowed from meta inode which are @@ -1543,9 +1539,6 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) return unlikely(f2fs_cp_error(sbi)) ? -EIO : 0; } -/* - * We guarantee that this checkpoint procedure will not fail. - */ int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) { struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi); @@ -1613,7 +1606,6 @@ int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) f2fs_flush_sit_entries(sbi, cpc); - /* unlock all the fs_lock[] in do_checkpoint() */ err = do_checkpoint(sbi, cpc); if (err) f2fs_release_discard_addrs(sbi); @@ -1626,7 +1618,7 @@ stop: if (cpc->reason & CP_RECOVERY) f2fs_notice(sbi, "checkpoint: version = %llx", ckpt_ver); - /* do checkpoint periodically */ + /* update CP_TIME to trigger checkpoint periodically */ f2fs_update_time(sbi, CP_TIME); trace_f2fs_write_checkpoint(sbi->sb, cpc->reason, "finish checkpoint"); out: diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index d8a64be90a50..df7b2d15eacd 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -11,6 +11,7 @@ #include <linux/backing-dev.h> #include <linux/lzo.h> #include <linux/lz4.h> +#include <linux/zstd.h> #include "f2fs.h" #include "node.h" @@ -20,6 +21,8 @@ struct f2fs_compress_ops { int (*init_compress_ctx)(struct compress_ctx *cc); void (*destroy_compress_ctx)(struct compress_ctx *cc); int (*compress_pages)(struct compress_ctx *cc); + int (*init_decompress_ctx)(struct decompress_io_ctx *dic); + void (*destroy_decompress_ctx)(struct decompress_io_ctx *dic); int (*decompress_pages)(struct decompress_io_ctx *dic); }; @@ -52,7 +55,7 @@ bool f2fs_is_compressed_page(struct page *page) } static void f2fs_set_compressed_page(struct page *page, - struct inode *inode, pgoff_t index, void *data, refcount_t *r) + struct inode *inode, pgoff_t index, void *data) { SetPagePrivate(page); set_page_private(page, (unsigned long)data); @@ -60,8 +63,6 @@ static void f2fs_set_compressed_page(struct page *page, /* i_crypto_info and iv index */ page->index = index; page->mapping = inode->i_mapping; - if (r) - refcount_inc(r); } static void f2fs_put_compressed_page(struct page *page) @@ -291,6 +292,165 @@ static const struct f2fs_compress_ops f2fs_lz4_ops = { }; #endif +#ifdef CONFIG_F2FS_FS_ZSTD +#define F2FS_ZSTD_DEFAULT_CLEVEL 1 + +static int zstd_init_compress_ctx(struct compress_ctx *cc) +{ + ZSTD_parameters params; + ZSTD_CStream *stream; + void *workspace; + unsigned int workspace_size; + + params = ZSTD_getParams(F2FS_ZSTD_DEFAULT_CLEVEL, cc->rlen, 0); + workspace_size = ZSTD_CStreamWorkspaceBound(params.cParams); + + workspace = f2fs_kvmalloc(F2FS_I_SB(cc->inode), + workspace_size, GFP_NOFS); + if (!workspace) + return -ENOMEM; + + stream = ZSTD_initCStream(params, 0, workspace, workspace_size); + if (!stream) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initCStream failed\n", + KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, + __func__); + kvfree(workspace); + return -EIO; + } + + cc->private = workspace; + cc->private2 = stream; + + cc->clen = cc->rlen - PAGE_SIZE - COMPRESS_HEADER_SIZE; + return 0; +} + +static void zstd_destroy_compress_ctx(struct compress_ctx *cc) +{ + kvfree(cc->private); + cc->private = NULL; + cc->private2 = NULL; +} + +static int zstd_compress_pages(struct compress_ctx *cc) +{ + ZSTD_CStream *stream = cc->private2; + ZSTD_inBuffer inbuf; + ZSTD_outBuffer outbuf; + int src_size = cc->rlen; + int dst_size = src_size - PAGE_SIZE - COMPRESS_HEADER_SIZE; + int ret; + + inbuf.pos = 0; + inbuf.src = cc->rbuf; + inbuf.size = src_size; + + outbuf.pos = 0; + outbuf.dst = cc->cbuf->cdata; + outbuf.size = dst_size; + + ret = ZSTD_compressStream(stream, &outbuf, &inbuf); + if (ZSTD_isError(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n", + KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, + __func__, ZSTD_getErrorCode(ret)); + return -EIO; + } + + ret = ZSTD_endStream(stream, &outbuf); + if (ZSTD_isError(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_endStream returned %d\n", + KERN_ERR, F2FS_I_SB(cc->inode)->sb->s_id, + __func__, ZSTD_getErrorCode(ret)); + return -EIO; + } + + cc->clen = outbuf.pos; + return 0; +} + +static int zstd_init_decompress_ctx(struct decompress_io_ctx *dic) +{ + ZSTD_DStream *stream; + void *workspace; + unsigned int workspace_size; + + workspace_size = ZSTD_DStreamWorkspaceBound(MAX_COMPRESS_WINDOW_SIZE); + + workspace = f2fs_kvmalloc(F2FS_I_SB(dic->inode), + workspace_size, GFP_NOFS); + if (!workspace) + return -ENOMEM; + + stream = ZSTD_initDStream(MAX_COMPRESS_WINDOW_SIZE, + workspace, workspace_size); + if (!stream) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_initDStream failed\n", + KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, + __func__); + kvfree(workspace); + return -EIO; + } + + dic->private = workspace; + dic->private2 = stream; + + return 0; +} + +static void zstd_destroy_decompress_ctx(struct decompress_io_ctx *dic) +{ + kvfree(dic->private); + dic->private = NULL; + dic->private2 = NULL; +} + +static int zstd_decompress_pages(struct decompress_io_ctx *dic) +{ + ZSTD_DStream *stream = dic->private2; + ZSTD_inBuffer inbuf; + ZSTD_outBuffer outbuf; + int ret; + + inbuf.pos = 0; + inbuf.src = dic->cbuf->cdata; + inbuf.size = dic->clen; + + outbuf.pos = 0; + outbuf.dst = dic->rbuf; + outbuf.size = dic->rlen; + + ret = ZSTD_decompressStream(stream, &outbuf, &inbuf); + if (ZSTD_isError(ret)) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD_compressStream failed, ret: %d\n", + KERN_ERR, F2FS_I_SB(dic->inode)->sb->s_id, + __func__, ZSTD_getErrorCode(ret)); + return -EIO; + } + + if (dic->rlen != outbuf.pos) { + printk_ratelimited("%sF2FS-fs (%s): %s ZSTD invalid rlen:%zu, " + "expected:%lu\n", KERN_ERR, + F2FS_I_SB(dic->inode)->sb->s_id, + __func__, dic->rlen, + PAGE_SIZE << dic->log_cluster_size); + return -EIO; + } + + return 0; +} + +static const struct f2fs_compress_ops f2fs_zstd_ops = { + .init_compress_ctx = zstd_init_compress_ctx, + .destroy_compress_ctx = zstd_destroy_compress_ctx, + .compress_pages = zstd_compress_pages, + .init_decompress_ctx = zstd_init_decompress_ctx, + .destroy_decompress_ctx = zstd_destroy_decompress_ctx, + .decompress_pages = zstd_decompress_pages, +}; +#endif + static const struct f2fs_compress_ops *f2fs_cops[COMPRESS_MAX] = { #ifdef CONFIG_F2FS_FS_LZO &f2fs_lzo_ops, @@ -302,6 +462,11 @@ static const struct f2fs_compress_ops *f2fs_cops[COMPRESS_MAX] = { #else NULL, #endif +#ifdef CONFIG_F2FS_FS_ZSTD + &f2fs_zstd_ops, +#else + NULL, +#endif }; bool f2fs_is_compress_backend_ready(struct inode *inode) @@ -334,9 +499,11 @@ static int f2fs_compress_pages(struct compress_ctx *cc) trace_f2fs_compress_pages_start(cc->inode, cc->cluster_idx, cc->cluster_size, fi->i_compress_algorithm); - ret = cops->init_compress_ctx(cc); - if (ret) - goto out; + if (cops->init_compress_ctx) { + ret = cops->init_compress_ctx(cc); + if (ret) + goto out; + } max_len = COMPRESS_HEADER_SIZE + cc->clen; cc->nr_cpages = DIV_ROUND_UP(max_len, PAGE_SIZE); @@ -380,21 +547,27 @@ static int f2fs_compress_pages(struct compress_ctx *cc) } cc->cbuf->clen = cpu_to_le32(cc->clen); - cc->cbuf->chksum = cpu_to_le32(0); for (i = 0; i < COMPRESS_DATA_RESERVED_SIZE; i++) cc->cbuf->reserved[i] = cpu_to_le32(0); + nr_cpages = DIV_ROUND_UP(cc->clen + COMPRESS_HEADER_SIZE, PAGE_SIZE); + + /* zero out any unused part of the last page */ + memset(&cc->cbuf->cdata[cc->clen], 0, + (nr_cpages * PAGE_SIZE) - (cc->clen + COMPRESS_HEADER_SIZE)); + vunmap(cc->cbuf); vunmap(cc->rbuf); - nr_cpages = DIV_ROUND_UP(cc->clen + COMPRESS_HEADER_SIZE, PAGE_SIZE); - for (i = nr_cpages; i < cc->nr_cpages; i++) { f2fs_put_compressed_page(cc->cpages[i]); cc->cpages[i] = NULL; } + if (cops->destroy_compress_ctx) + cops->destroy_compress_ctx(cc); + cc->nr_cpages = nr_cpages; trace_f2fs_compress_pages_end(cc->inode, cc->cluster_idx, @@ -413,7 +586,8 @@ out_free_cpages: kfree(cc->cpages); cc->cpages = NULL; destroy_compress_ctx: - cops->destroy_compress_ctx(cc); + if (cops->destroy_compress_ctx) + cops->destroy_compress_ctx(cc); out: trace_f2fs_compress_pages_end(cc->inode, cc->cluster_idx, cc->clen, ret); @@ -447,10 +621,16 @@ void f2fs_decompress_pages(struct bio *bio, struct page *page, bool verity) goto out_free_dic; } + if (cops->init_decompress_ctx) { + ret = cops->init_decompress_ctx(dic); + if (ret) + goto out_free_dic; + } + dic->rbuf = vmap(dic->tpages, dic->cluster_size, VM_MAP, PAGE_KERNEL); if (!dic->rbuf) { ret = -ENOMEM; - goto out_free_dic; + goto destroy_decompress_ctx; } dic->cbuf = vmap(dic->cpages, dic->nr_cpages, VM_MAP, PAGE_KERNEL_RO); @@ -473,7 +653,12 @@ out_vunmap_cbuf: vunmap(dic->cbuf); out_vunmap_rbuf: vunmap(dic->rbuf); +destroy_decompress_ctx: + if (cops->destroy_decompress_ctx) + cops->destroy_decompress_ctx(dic); out_free_dic: + if (verity) + refcount_set(&dic->ref, dic->nr_cpages); if (!verity) f2fs_decompress_end_io(dic->rpages, dic->cluster_size, ret, false); @@ -532,8 +717,7 @@ static bool __cluster_may_compress(struct compress_ctx *cc) return true; } -/* return # of compressed block addresses */ -static int f2fs_compressed_blocks(struct compress_ctx *cc) +static int __f2fs_cluster_blocks(struct compress_ctx *cc, bool compr) { struct dnode_of_data dn; int ret; @@ -554,10 +738,15 @@ static int f2fs_compressed_blocks(struct compress_ctx *cc) for (i = 1; i < cc->cluster_size; i++) { block_t blkaddr; - blkaddr = datablock_addr(dn.inode, + blkaddr = data_blkaddr(dn.inode, dn.node_page, dn.ofs_in_node + i); - if (blkaddr != NULL_ADDR) - ret++; + if (compr) { + if (__is_valid_data_blkaddr(blkaddr)) + ret++; + } else { + if (blkaddr != NULL_ADDR) + ret++; + } } } fail: @@ -565,6 +754,18 @@ fail: return ret; } +/* return # of compressed blocks in compressed cluster */ +static int f2fs_compressed_blocks(struct compress_ctx *cc) +{ + return __f2fs_cluster_blocks(cc, true); +} + +/* return # of valid blocks in compressed cluster */ +static int f2fs_cluster_blocks(struct compress_ctx *cc, bool compr) +{ + return __f2fs_cluster_blocks(cc, false); +} + int f2fs_is_compressed_cluster(struct inode *inode, pgoff_t index) { struct compress_ctx cc = { @@ -574,7 +775,7 @@ int f2fs_is_compressed_cluster(struct inode *inode, pgoff_t index) .cluster_idx = index >> F2FS_I(inode)->i_log_cluster_size, }; - return f2fs_compressed_blocks(&cc); + return f2fs_cluster_blocks(&cc, false); } static bool cluster_may_compress(struct compress_ctx *cc) @@ -623,7 +824,7 @@ static int prepare_compress_overwrite(struct compress_ctx *cc, bool prealloc; retry: - ret = f2fs_compressed_blocks(cc); + ret = f2fs_cluster_blocks(cc, false); if (ret <= 0) return ret; @@ -653,7 +854,7 @@ retry: struct bio *bio = NULL; ret = f2fs_read_multi_pages(cc, &bio, cc->cluster_size, - &last_block_in_bio, false); + &last_block_in_bio, false, true); f2fs_destroy_compress_ctx(cc); if (ret) goto release_pages; @@ -772,7 +973,6 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, .encrypted_page = NULL, .compressed_page = NULL, .submitted = false, - .need_lock = LOCK_RETRY, .io_type = io_type, .io_wbc = wbc, .encrypted = f2fs_encrypted_file(cc->inode), @@ -785,16 +985,17 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, loff_t psize; int i, err; - set_new_dnode(&dn, cc->inode, NULL, NULL, 0); + if (!f2fs_trylock_op(sbi)) + return -EAGAIN; - f2fs_lock_op(sbi); + set_new_dnode(&dn, cc->inode, NULL, NULL, 0); err = f2fs_get_dnode_of_data(&dn, start_idx, LOOKUP_NODE); if (err) goto out_unlock_op; for (i = 0; i < cc->cluster_size; i++) { - if (datablock_addr(dn.inode, dn.node_page, + if (data_blkaddr(dn.inode, dn.node_page, dn.ofs_in_node + i) == NULL_ADDR) goto out_put_dnode; } @@ -813,7 +1014,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, cic->magic = F2FS_COMPRESSED_PAGE_MAGIC; cic->inode = inode; - refcount_set(&cic->ref, 1); + refcount_set(&cic->ref, cc->nr_cpages); cic->rpages = f2fs_kzalloc(sbi, sizeof(struct page *) << cc->log_cluster_size, GFP_NOFS); if (!cic->rpages) @@ -823,8 +1024,7 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, for (i = 0; i < cc->nr_cpages; i++) { f2fs_set_compressed_page(cc->cpages[i], inode, - cc->rpages[i + 1]->index, - cic, i ? &cic->ref : NULL); + cc->rpages[i + 1]->index, cic); fio.compressed_page = cc->cpages[i]; if (fio.encrypted) { fio.page = cc->rpages[i + 1]; @@ -843,9 +1043,8 @@ static int f2fs_write_compressed_pages(struct compress_ctx *cc, for (i = 0; i < cc->cluster_size; i++, dn.ofs_in_node++) { block_t blkaddr; - blkaddr = datablock_addr(dn.inode, dn.node_page, - dn.ofs_in_node); - fio.page = cic->rpages[i]; + blkaddr = f2fs_data_blkaddr(&dn); + fio.page = cc->rpages[i]; fio.old_blkaddr = blkaddr; /* cluster header */ @@ -895,10 +1094,10 @@ unlock_continue: f2fs_put_dnode(&dn); f2fs_unlock_op(sbi); - down_write(&fi->i_sem); + spin_lock(&fi->i_size_lock); if (fi->last_disk_size < psize) fi->last_disk_size = psize; - up_write(&fi->i_sem); + spin_unlock(&fi->i_size_lock); f2fs_put_rpages(cc); f2fs_destroy_compress_ctx(cc); @@ -984,24 +1183,30 @@ retry_write: unlock_page(cc->rpages[i]); ret = 0; } else if (ret == -EAGAIN) { + /* + * for quota file, just redirty left pages to + * avoid deadlock caused by cluster update race + * from foreground operation. + */ + if (IS_NOQUOTA(cc->inode)) { + err = 0; + goto out_err; + } ret = 0; cond_resched(); - congestion_wait(BLK_RW_ASYNC, HZ/50); + congestion_wait(BLK_RW_ASYNC, + DEFAULT_IO_TIMEOUT); lock_page(cc->rpages[i]); clear_page_dirty_for_io(cc->rpages[i]); goto retry_write; } err = ret; - goto out_fail; + goto out_err; } *submitted += _submitted; } return 0; - -out_fail: - /* TODO: revoke partially updated block addresses */ - BUG_ON(compr_blocks); out_err: for (++i; i < cc->cluster_size; i++) { if (!cc->rpages[i]) @@ -1069,7 +1274,7 @@ struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc) dic->magic = F2FS_COMPRESSED_PAGE_MAGIC; dic->inode = cc->inode; - refcount_set(&dic->ref, 1); + refcount_set(&dic->ref, cc->nr_cpages); dic->cluster_idx = cc->cluster_idx; dic->cluster_size = cc->cluster_size; dic->log_cluster_size = cc->log_cluster_size; @@ -1093,8 +1298,7 @@ struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc) goto out_free; f2fs_set_compressed_page(page, cc->inode, - start_idx + i + 1, - dic, i ? &dic->ref : NULL); + start_idx + i + 1, dic); dic->cpages[i] = page; } @@ -1104,20 +1308,16 @@ struct decompress_io_ctx *f2fs_alloc_dic(struct compress_ctx *cc) goto out_free; for (i = 0; i < dic->cluster_size; i++) { - if (cc->rpages[i]) + if (cc->rpages[i]) { + dic->tpages[i] = cc->rpages[i]; continue; + } dic->tpages[i] = f2fs_grab_page(); if (!dic->tpages[i]) goto out_free; } - for (i = 0; i < dic->cluster_size; i++) { - if (dic->tpages[i]) - continue; - dic->tpages[i] = cc->rpages[i]; - } - return dic; out_free: @@ -1133,7 +1333,10 @@ void f2fs_free_dic(struct decompress_io_ctx *dic) for (i = 0; i < dic->cluster_size; i++) { if (dic->rpages[i]) continue; - f2fs_put_page(dic->tpages[i], 1); + if (!dic->tpages[i]) + continue; + unlock_page(dic->tpages[i]); + put_page(dic->tpages[i]); } kfree(dic->tpages); } @@ -1162,15 +1365,17 @@ void f2fs_decompress_end_io(struct page **rpages, if (!rpage) continue; - if (err || PageError(rpage)) { - ClearPageUptodate(rpage); - ClearPageError(rpage); - } else { - if (!verity || fsverity_verify_page(rpage)) - SetPageUptodate(rpage); - else - SetPageError(rpage); + if (err || PageError(rpage)) + goto clear_uptodate; + + if (!verity || fsverity_verify_page(rpage)) { + SetPageUptodate(rpage); + goto unlock; } +clear_uptodate: + ClearPageUptodate(rpage); + ClearPageError(rpage); +unlock: unlock_page(rpage); } } diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index b27b72107911..cdf2f626bea7 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -54,17 +54,13 @@ static inline struct bio *__f2fs_bio_alloc(gfp_t gfp_mask, return bio_alloc_bioset(gfp_mask, nr_iovecs, &f2fs_bioset); } -struct bio *f2fs_bio_alloc(struct f2fs_sb_info *sbi, int npages, bool no_fail) +struct bio *f2fs_bio_alloc(struct f2fs_sb_info *sbi, int npages, bool noio) { - struct bio *bio; - - if (no_fail) { + if (noio) { /* No failure on bio allocation */ - bio = __f2fs_bio_alloc(GFP_NOIO, npages); - if (!bio) - bio = __f2fs_bio_alloc(GFP_NOIO | __GFP_NOFAIL, npages); - return bio; + return __f2fs_bio_alloc(GFP_NOIO, npages); } + if (time_to_inject(sbi, FAULT_ALLOC_BIO)) { f2fs_show_injection_info(sbi, FAULT_ALLOC_BIO); return NULL; @@ -143,6 +139,8 @@ static void __read_end_io(struct bio *bio, bool compr, bool verity) f2fs_decompress_pages(bio, page, verity); continue; } + if (verity) + continue; #endif /* PG_error was set if any post_read step failed */ @@ -191,12 +189,38 @@ static void f2fs_verify_pages(struct page **rpages, unsigned int cluster_size) static void f2fs_verify_bio(struct bio *bio) { - struct page *page = bio_first_page_all(bio); - struct decompress_io_ctx *dic = - (struct decompress_io_ctx *)page_private(page); + struct bio_vec *bv; + struct bvec_iter_all iter_all; - f2fs_verify_pages(dic->rpages, dic->cluster_size); - f2fs_free_dic(dic); + bio_for_each_segment_all(bv, bio, iter_all) { + struct page *page = bv->bv_page; + struct decompress_io_ctx *dic; + + dic = (struct decompress_io_ctx *)page_private(page); + + if (dic) { + if (refcount_dec_not_one(&dic->ref)) + continue; + f2fs_verify_pages(dic->rpages, + dic->cluster_size); + f2fs_free_dic(dic); + continue; + } + + if (bio->bi_status || PageError(page)) + goto clear_uptodate; + + if (fsverity_verify_page(page)) { + SetPageUptodate(page); + goto unlock; + } +clear_uptodate: + ClearPageUptodate(page); + ClearPageError(page); +unlock: + dec_page_count(F2FS_P_SB(page), __read_io_type(page)); + unlock_page(page); + } } #endif @@ -364,9 +388,6 @@ static void f2fs_write_end_io(struct bio *bio) bio_put(bio); } -/* - * Return true, if pre_bio's bdev is same as its target device. - */ struct block_device *f2fs_target_device(struct f2fs_sb_info *sbi, block_t blk_addr, struct bio *bio) { @@ -403,6 +424,9 @@ int f2fs_target_device_index(struct f2fs_sb_info *sbi, block_t blkaddr) return 0; } +/* + * Return true, if pre_bio's bdev is same as its target device. + */ static bool __same_bdev(struct f2fs_sb_info *sbi, block_t blk_addr, struct bio *bio) { @@ -410,9 +434,6 @@ static bool __same_bdev(struct f2fs_sb_info *sbi, return bio->bi_disk == b->bd_disk && bio->bi_partno == b->bd_partno; } -/* - * Low-level block read/write IO operations. - */ static struct bio *__bio_alloc(struct f2fs_io_info *fio, int npages) { struct f2fs_sb_info *sbi = fio->sbi; @@ -445,7 +466,7 @@ static inline void __submit_bio(struct f2fs_sb_info *sbi, if (type != DATA && type != NODE) goto submit_io; - if (test_opt(sbi, LFS) && current->plug) + if (f2fs_lfs_mode(sbi) && current->plug) blk_finish_plug(current->plug); if (F2FS_IO_ALIGNED(sbi)) @@ -928,14 +949,15 @@ static inline bool f2fs_need_verity(const struct inode *inode, pgoff_t idx) static struct bio *f2fs_grab_read_bio(struct inode *inode, block_t blkaddr, unsigned nr_pages, unsigned op_flag, - pgoff_t first_idx) + pgoff_t first_idx, bool for_write) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct bio *bio; struct bio_post_read_ctx *ctx; unsigned int post_read_steps = 0; - bio = f2fs_bio_alloc(sbi, min_t(int, nr_pages, BIO_MAX_PAGES), false); + bio = f2fs_bio_alloc(sbi, min_t(int, nr_pages, BIO_MAX_PAGES), + for_write); if (!bio) return ERR_PTR(-ENOMEM); f2fs_target_device(sbi, blkaddr, bio); @@ -970,12 +992,12 @@ static void f2fs_release_read_bio(struct bio *bio) /* This can handle encryption stuffs */ static int f2fs_submit_page_read(struct inode *inode, struct page *page, - block_t blkaddr) + block_t blkaddr, bool for_write) { struct f2fs_sb_info *sbi = F2FS_I_SB(inode); struct bio *bio; - bio = f2fs_grab_read_bio(inode, blkaddr, 1, 0, page->index); + bio = f2fs_grab_read_bio(inode, blkaddr, 1, 0, page->index, for_write); if (IS_ERR(bio)) return PTR_ERR(bio); @@ -1047,8 +1069,7 @@ int f2fs_reserve_new_blocks(struct dnode_of_data *dn, blkcnt_t count) f2fs_wait_on_page_writeback(dn->node_page, NODE, true, true); for (; count > 0; dn->ofs_in_node++) { - block_t blkaddr = datablock_addr(dn->inode, - dn->node_page, dn->ofs_in_node); + block_t blkaddr = f2fs_data_blkaddr(dn); if (blkaddr == NULL_ADDR) { dn->data_blkaddr = NEW_ADDR; __set_data_blkaddr(dn); @@ -1162,7 +1183,7 @@ got_it: return page; } - err = f2fs_submit_page_read(inode, page, dn.data_blkaddr); + err = f2fs_submit_page_read(inode, page, dn.data_blkaddr, for_write); if (err) goto put_err; return page; @@ -1300,8 +1321,7 @@ static int __allocate_data_block(struct dnode_of_data *dn, int seg_type) if (err) return err; - dn->data_blkaddr = datablock_addr(dn->inode, - dn->node_page, dn->ofs_in_node); + dn->data_blkaddr = f2fs_data_blkaddr(dn); if (dn->data_blkaddr != NULL_ADDR) goto alloc; @@ -1388,13 +1408,9 @@ void __do_map_lock(struct f2fs_sb_info *sbi, int flag, bool lock) } /* - * f2fs_map_blocks() now supported readahead/bmap/rw direct_IO with - * f2fs_map_blocks structure. - * If original data blocks are allocated, then give them to blockdev. - * Otherwise, - * a. preallocate requested block addresses - * b. do not use extent cache for better performance - * c. give the block addresses to blockdev + * f2fs_map_blocks() tries to find or build mapping relationship which + * maps continuous logical blocks to physical blocks, and return such + * info via f2fs_map_blocks structure. */ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, int create, int flag) @@ -1422,7 +1438,7 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, end = pgofs + maxblocks; if (!create && f2fs_lookup_extent_cache(inode, pgofs, &ei)) { - if (test_opt(sbi, LFS) && flag == F2FS_GET_BLOCK_DIO && + if (f2fs_lfs_mode(sbi) && flag == F2FS_GET_BLOCK_DIO && map->m_may_create) goto next_dnode; @@ -1467,7 +1483,7 @@ next_dnode: end_offset = ADDRS_PER_PAGE(dn.node_page, inode); next_block: - blkaddr = datablock_addr(dn.inode, dn.node_page, dn.ofs_in_node); + blkaddr = f2fs_data_blkaddr(&dn);< |