From 069782a1ee55105220e5ae2db448495dac267cb1 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 21 Oct 2008 12:56:31 +0300 Subject: UBIFS: remove printk Remove the "UBIFS background thread ubifs_bgd0_0 started" message. We kill the background thread when we switch to R/O mode, and start it again whan we switch to R/W mode. OLPC is doing this many times during boot, and we see this message many times as well, which is irritating. So just kill the message. Signed-off-by: Artem Bityutskiy --- fs/ubifs/commit.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/commit.c b/fs/ubifs/commit.c index 0a6aa2cc78f0..b49884c8c10e 100644 --- a/fs/ubifs/commit.c +++ b/fs/ubifs/commit.c @@ -234,8 +234,8 @@ int ubifs_bg_thread(void *info) int err; struct ubifs_info *c = info; - ubifs_msg("background thread \"%s\" started, PID %d", - c->bgt_name, current->pid); + dbg_msg("background thread \"%s\" started, PID %d", + c->bgt_name, current->pid); set_freezable(); while (1) { -- cgit v1.2.3 From 0ecb9529a4d47825778e7b0d226eb36019252a9d Mon Sep 17 00:00:00 2001 From: Harvey Harrison Date: Fri, 24 Oct 2008 10:52:57 -0700 Subject: UBIFS: endian handling fixes and annotations Noticed by sparse: fs/ubifs/file.c:75:2: warning: restricted __le64 degrades to integer fs/ubifs/file.c:629:4: warning: restricted __le64 degrades to integer fs/ubifs/dir.c:431:3: warning: restricted __le64 degrades to integer This should be checked to ensure the ubifs_assert is working as intended, I've done the suggested annotation in this patch. fs/ubifs/sb.c:298:6: warning: incorrect type in assignment (different base types) fs/ubifs/sb.c:298:6: expected int [signed] [assigned] tmp fs/ubifs/sb.c:298:6: got restricted __le64 [usertype] fs/ubifs/sb.c:299:19: warning: incorrect type in assignment (different base types) fs/ubifs/sb.c:299:19: expected restricted __le64 [usertype] atime_sec fs/ubifs/sb.c:299:19: got int [signed] [assigned] tmp fs/ubifs/sb.c:300:19: warning: incorrect type in assignment (different base types) fs/ubifs/sb.c:300:19: expected restricted __le64 [usertype] ctime_sec fs/ubifs/sb.c:300:19: got int [signed] [assigned] tmp fs/ubifs/sb.c:301:19: warning: incorrect type in assignment (different base types) fs/ubifs/sb.c:301:19: expected restricted __le64 [usertype] mtime_sec fs/ubifs/sb.c:301:19: got int [signed] [assigned] tmp This looks like a bugfix as your tmp was a u32 so there was truncation in the atime, mtime, ctime value, probably not intentional, add a tmp_le64 and use it here. fs/ubifs/key.h:348:9: warning: cast to restricted __le32 fs/ubifs/key.h:348:9: warning: cast to restricted __le32 fs/ubifs/key.h:419:9: warning: cast to restricted __le32 Read from the annotated union member instead. fs/ubifs/recovery.c:175:13: warning: incorrect type in assignment (different base types) fs/ubifs/recovery.c:175:13: expected unsigned int [unsigned] [usertype] save_flags fs/ubifs/recovery.c:175:13: got restricted __le32 [usertype] flags fs/ubifs/recovery.c:186:13: warning: incorrect type in assignment (different base types) fs/ubifs/recovery.c:186:13: expected restricted __le32 [usertype] flags fs/ubifs/recovery.c:186:13: got unsigned int [unsigned] [usertype] save_flags Do byteshifting at compile time of the flag value. Annotate the saved_flags as le32. fs/ubifs/debug.c:368:10: warning: cast to restricted __le32 fs/ubifs/debug.c:368:10: warning: cast from restricted __le64 Should be checked if the truncation was intentional, I've changed the printk to print the full width. Signed-off-by: Harvey Harrison Signed-off-by: Artem Bityutskiy --- fs/ubifs/debug.c | 4 ++-- fs/ubifs/dir.c | 3 ++- fs/ubifs/file.c | 4 ++-- fs/ubifs/key.h | 4 ++-- fs/ubifs/recovery.c | 4 ++-- fs/ubifs/sb.c | 9 +++++---- 6 files changed, 15 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index 7186400750e7..f9deccbc90c0 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -364,8 +364,8 @@ void dbg_dump_node(const struct ubifs_info *c, const void *node) le32_to_cpu(mst->ihead_lnum)); printk(KERN_DEBUG "\tihead_offs %u\n", le32_to_cpu(mst->ihead_offs)); - printk(KERN_DEBUG "\tindex_size %u\n", - le32_to_cpu(mst->index_size)); + printk(KERN_DEBUG "\tindex_size %llu\n", + (unsigned long long)le64_to_cpu(mst->index_size)); printk(KERN_DEBUG "\tlpt_lnum %u\n", le32_to_cpu(mst->lpt_lnum)); printk(KERN_DEBUG "\tlpt_offs %u\n", diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 526c01ec8003..37a9e604c3e9 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -428,7 +428,8 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) dbg_gen("feed '%s', ino %llu, new f_pos %#x", dent->name, (unsigned long long)le64_to_cpu(dent->inum), key_hash_flash(c, &dent->key)); - ubifs_assert(dent->ch.sqnum > ubifs_inode(dir)->creat_sqnum); + ubifs_assert(le64_to_cpu(dent->ch.sqnum) > + ubifs_inode(dir)->creat_sqnum); nm.len = le16_to_cpu(dent->nlen); over = filldir(dirent, dent->name, nm.len, file->f_pos, diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 51cf511d44d9..9124eee73aea 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -72,7 +72,7 @@ static int read_block(struct inode *inode, void *addr, unsigned int block, return err; } - ubifs_assert(dn->ch.sqnum > ubifs_inode(inode)->creat_sqnum); + ubifs_assert(le64_to_cpu(dn->ch.sqnum) > ubifs_inode(inode)->creat_sqnum); len = le32_to_cpu(dn->size); if (len <= 0 || len > UBIFS_BLOCK_SIZE) @@ -626,7 +626,7 @@ static int populate_page(struct ubifs_info *c, struct page *page, dn = bu->buf + (bu->zbranch[nn].offs - offs); - ubifs_assert(dn->ch.sqnum > + ubifs_assert(le64_to_cpu(dn->ch.sqnum) > ubifs_inode(inode)->creat_sqnum); len = le32_to_cpu(dn->size); diff --git a/fs/ubifs/key.h b/fs/ubifs/key.h index 9ee65086f627..3f1f16bc25c9 100644 --- a/fs/ubifs/key.h +++ b/fs/ubifs/key.h @@ -345,7 +345,7 @@ static inline int key_type_flash(const struct ubifs_info *c, const void *k) { const union ubifs_key *key = k; - return le32_to_cpu(key->u32[1]) >> UBIFS_S_KEY_BLOCK_BITS; + return le32_to_cpu(key->j32[1]) >> UBIFS_S_KEY_BLOCK_BITS; } /** @@ -416,7 +416,7 @@ static inline unsigned int key_block_flash(const struct ubifs_info *c, { const union ubifs_key *key = k; - return le32_to_cpu(key->u32[1]) & UBIFS_S_KEY_BLOCK_MASK; + return le32_to_cpu(key->j32[1]) & UBIFS_S_KEY_BLOCK_MASK; } /** diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 77d26c141cf6..bed97421b972 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -168,12 +168,12 @@ static int write_rcvrd_mst_node(struct ubifs_info *c, struct ubifs_mst_node *mst) { int err = 0, lnum = UBIFS_MST_LNUM, sz = c->mst_node_alsz; - uint32_t save_flags; + __le32 save_flags; dbg_rcvry("recovery"); save_flags = mst->flags; - mst->flags = cpu_to_le32(le32_to_cpu(mst->flags) | UBIFS_MST_RCVRY); + mst->flags |= cpu_to_le32(UBIFS_MST_RCVRY); ubifs_prepare_node(c, mst, UBIFS_MST_NODE_SZ, 1); err = ubi_leb_change(c->ubi, lnum, mst, sz, UBI_SHORTTERM); diff --git a/fs/ubifs/sb.c b/fs/ubifs/sb.c index 2bf753b38889..0f392351dc5a 100644 --- a/fs/ubifs/sb.c +++ b/fs/ubifs/sb.c @@ -81,6 +81,7 @@ static int create_default_filesystem(struct ubifs_info *c) int lpt_lebs, lpt_first, orph_lebs, big_lpt, ino_waste, sup_flags = 0; int min_leb_cnt = UBIFS_MIN_LEB_CNT; uint64_t tmp64, main_bytes; + __le64 tmp_le64; /* Some functions called from here depend on the @c->key_len filed */ c->key_len = UBIFS_SK_LEN; @@ -295,10 +296,10 @@ static int create_default_filesystem(struct ubifs_info *c) ino->ch.node_type = UBIFS_INO_NODE; ino->creat_sqnum = cpu_to_le64(++c->max_sqnum); ino->nlink = cpu_to_le32(2); - tmp = cpu_to_le64(CURRENT_TIME_SEC.tv_sec); - ino->atime_sec = tmp; - ino->ctime_sec = tmp; - ino->mtime_sec = tmp; + tmp_le64 = cpu_to_le64(CURRENT_TIME_SEC.tv_sec); + ino->atime_sec = tmp_le64; + ino->ctime_sec = tmp_le64; + ino->mtime_sec = tmp_le64; ino->atime_nsec = 0; ino->ctime_nsec = 0; ino->mtime_nsec = 0; -- cgit v1.2.3 From e84461ad9c4f0ff91ab8553596acdb7bf5522df4 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Wed, 29 Oct 2008 12:08:43 +0200 Subject: UBIFS: fix compilation warnings We print 'ino_t' type using '%lu' printk() placeholder, but this results in many warnings when compiling for Alpha platform. Fix this by adding (unsingned long) casts. Fixes these warnings: fs/ubifs/journal.c:693: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/journal.c:1131: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/dir.c:163: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/tnc.c:2680: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/tnc.c:2700: warning: format '%lu' expects type 'long unsigned int', but argument 5 has type 'ino_t' fs/ubifs/replay.c:1066: warning: format '%lu' expects type 'long unsigned int', but argument 7 has type 'ino_t' fs/ubifs/orphan.c:108: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:135: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:142: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:154: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:159: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:451: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:539: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:612: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:843: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/orphan.c:856: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/recovery.c:1438: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/recovery.c:1443: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/recovery.c:1475: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/recovery.c:1495: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:105: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:105: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:110: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:110: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:114: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:114: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:118: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:118: warning: format '%lu' expects type 'long unsigned int', but argument 3 has type 'ino_t' fs/ubifs/debug.c:1591: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1671: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1674: warning: format '%lu' expects type 'long unsigned int', but argument 5 has type 'ino_t' fs/ubifs/debug.c:1680: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1699: warning: format '%lu' expects type 'long unsigned int', but argument 5 has type 'ino_t' fs/ubifs/debug.c:1788: warning: format '%lu' expects type 'long unsigned int', but argument 5 has type 'ino_t' fs/ubifs/debug.c:1821: warning: format '%lu' expects type 'long unsigned int', but argument 5 has type 'ino_t' fs/ubifs/debug.c:1833: warning: format '%lu' expects type 'long unsigned int', but argument 5 has type 'ino_t' fs/ubifs/debug.c:1924: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1932: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1938: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1945: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1953: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1960: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1967: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1973: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1988: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'ino_t' fs/ubifs/debug.c:1991: warning: format '%lu' expects type 'long unsigned int', but argument 5 has type 'ino_t' fs/ubifs/debug.c:2009: warning: format '%lu' expects type 'long unsigned int', but argument 2 has type 'ino_t' Reported-by: Randy Dunlap Signed-off-by: Artem Bityutskiy --- fs/ubifs/debug.c | 62 ++++++++++++++++++++++++++++++++--------------------- fs/ubifs/dir.c | 2 +- fs/ubifs/journal.c | 8 ++++--- fs/ubifs/orphan.c | 28 +++++++++++++----------- fs/ubifs/recovery.c | 13 +++++------ fs/ubifs/replay.c | 2 +- fs/ubifs/tnc.c | 5 +++-- 7 files changed, 70 insertions(+), 50 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index f9deccbc90c0..510ffa0bbda4 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -101,21 +101,24 @@ static void sprintf_key(const struct ubifs_info *c, const union ubifs_key *key, if (c->key_fmt == UBIFS_SIMPLE_KEY_FMT) { switch (type) { case UBIFS_INO_KEY: - sprintf(p, "(%lu, %s)", key_inum(c, key), + sprintf(p, "(%lu, %s)", (unsigned long)key_inum(c, key), get_key_type(type)); break; case UBIFS_DENT_KEY: case UBIFS_XENT_KEY: - sprintf(p, "(%lu, %s, %#08x)", key_inum(c, key), + sprintf(p, "(%lu, %s, %#08x)", + (unsigned long)key_inum(c, key), get_key_type(type), key_hash(c, key)); break; case UBIFS_DATA_KEY: - sprintf(p, "(%lu, %s, %u)", key_inum(c, key), + sprintf(p, "(%lu, %s, %u)", + (unsigned long)key_inum(c, key), get_key_type(type), key_block(c, key)); break; case UBIFS_TRUN_KEY: sprintf(p, "(%lu, %s)", - key_inum(c, key), get_key_type(type)); + (unsigned long)key_inum(c, key), + get_key_type(type)); break; default: sprintf(p, "(bad key type: %#08x, %#08x)", @@ -1589,7 +1592,7 @@ static struct fsck_inode *add_inode(struct ubifs_info *c, if (inum > c->highest_inum) { ubifs_err("too high inode number, max. is %lu", - c->highest_inum); + (unsigned long)c->highest_inum); return ERR_PTR(-EINVAL); } @@ -1668,16 +1671,18 @@ static struct fsck_inode *read_add_inode(struct ubifs_info *c, ino_key_init(c, &key, inum); err = ubifs_lookup_level0(c, &key, &znode, &n); if (!err) { - ubifs_err("inode %lu not found in index", inum); + ubifs_err("inode %lu not found in index", (unsigned long)inum); return ERR_PTR(-ENOENT); } else if (err < 0) { - ubifs_err("error %d while looking up inode %lu", err, inum); + ubifs_err("error %d while looking up inode %lu", + err, (unsigned long)inum); return ERR_PTR(err); } zbr = &znode->zbranch[n]; if (zbr->len < UBIFS_INO_NODE_SZ) { - ubifs_err("bad node %lu node length %d", inum, zbr->len); + ubifs_err("bad node %lu node length %d", + (unsigned long)inum, zbr->len); return ERR_PTR(-EINVAL); } @@ -1697,7 +1702,7 @@ static struct fsck_inode *read_add_inode(struct ubifs_info *c, kfree(ino); if (IS_ERR(fscki)) { ubifs_err("error %ld while adding inode %lu node", - PTR_ERR(fscki), inum); + PTR_ERR(fscki), (unsigned long)inum); return fscki; } @@ -1786,7 +1791,8 @@ static int check_leaf(struct ubifs_info *c, struct ubifs_zbranch *zbr, if (IS_ERR(fscki)) { err = PTR_ERR(fscki); ubifs_err("error %d while processing data node and " - "trying to find inode node %lu", err, inum); + "trying to find inode node %lu", + err, (unsigned long)inum); goto out_dump; } @@ -1819,7 +1825,8 @@ static int check_leaf(struct ubifs_info *c, struct ubifs_zbranch *zbr, if (IS_ERR(fscki)) { err = PTR_ERR(fscki); ubifs_err("error %d while processing entry node and " - "trying to find inode node %lu", err, inum); + "trying to find inode node %lu", + err, (unsigned long)inum); goto out_dump; } @@ -1832,7 +1839,7 @@ static int check_leaf(struct ubifs_info *c, struct ubifs_zbranch *zbr, err = PTR_ERR(fscki); ubifs_err("error %d while processing entry node and " "trying to find parent inode node %lu", - err, inum); + err, (unsigned long)inum); goto out_dump; } @@ -1923,7 +1930,8 @@ static int check_inodes(struct ubifs_info *c, struct fsck_data *fsckd) fscki->references != 1) { ubifs_err("directory inode %lu has %d " "direntries which refer it, but " - "should be 1", fscki->inum, + "should be 1", + (unsigned long)fscki->inum, fscki->references); goto out_dump; } @@ -1931,27 +1939,29 @@ static int check_inodes(struct ubifs_info *c, struct fsck_data *fsckd) fscki->references != 0) { ubifs_err("root inode %lu has non-zero (%d) " "direntries which refer it", - fscki->inum, fscki->references); + (unsigned long)fscki->inum, + fscki->references); goto out_dump; } if (fscki->calc_sz != fscki->size) { ubifs_err("directory inode %lu size is %lld, " "but calculated size is %lld", - fscki->inum, fscki->size, - fscki->calc_sz); + (unsigned long)fscki->inum, + fscki->size, fscki->calc_sz); goto out_dump; } if (fscki->calc_cnt != fscki->nlink) { ubifs_err("directory inode %lu nlink is %d, " "but calculated nlink is %d", - fscki->inum, fscki->nlink, - fscki->calc_cnt); + (unsigned long)fscki->inum, + fscki->nlink, fscki->calc_cnt); goto out_dump; } } else { if (fscki->references != fscki->nlink) { ubifs_err("inode %lu nlink is %d, but " - "calculated nlink is %d", fscki->inum, + "calculated nlink is %d", + (unsigned long)fscki->inum, fscki->nlink, fscki->references); goto out_dump; } @@ -1959,20 +1969,21 @@ static int check_inodes(struct ubifs_info *c, struct fsck_data *fsckd) if (fscki->xattr_sz != fscki->calc_xsz) { ubifs_err("inode %lu has xattr size %u, but " "calculated size is %lld", - fscki->inum, fscki->xattr_sz, + (unsigned long)fscki->inum, fscki->xattr_sz, fscki->calc_xsz); goto out_dump; } if (fscki->xattr_cnt != fscki->calc_xcnt) { ubifs_err("inode %lu has %u xattrs, but " - "calculated count is %lld", fscki->inum, + "calculated count is %lld", + (unsigned long)fscki->inum, fscki->xattr_cnt, fscki->calc_xcnt); goto out_dump; } if (fscki->xattr_nms != fscki->calc_xnms) { ubifs_err("inode %lu has xattr names' size %u, but " "calculated names' size is %lld", - fscki->inum, fscki->xattr_nms, + (unsigned long)fscki->inum, fscki->xattr_nms, fscki->calc_xnms); goto out_dump; } @@ -1985,11 +1996,12 @@ out_dump: ino_key_init(c, &key, fscki->inum); err = ubifs_lookup_level0(c, &key, &znode, &n); if (!err) { - ubifs_err("inode %lu not found in index", fscki->inum); + ubifs_err("inode %lu not found in index", + (unsigned long)fscki->inum); return -ENOENT; } else if (err < 0) { ubifs_err("error %d while looking up inode %lu", - err, fscki->inum); + err, (unsigned long)fscki->inum); return err; } @@ -2007,7 +2019,7 @@ out_dump: } ubifs_msg("dump of the inode %lu sitting in LEB %d:%d", - fscki->inum, zbr->lnum, zbr->offs); + (unsigned long)fscki->inum, zbr->lnum, zbr->offs); dbg_dump_node(c, ino); kfree(ino); return -EINVAL; diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 37a9e604c3e9..0422c98e1793 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -161,7 +161,7 @@ struct inode *ubifs_new_inode(struct ubifs_info *c, const struct inode *dir, return ERR_PTR(-EINVAL); } ubifs_warn("running out of inode numbers (current %lu, max %d)", - c->highest_inum, INUM_WATERMARK); + (unsigned long)c->highest_inum, INUM_WATERMARK); } inode->i_ino = ++c->highest_inum; diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index 22993f867d19..f91b745908ea 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -690,8 +690,9 @@ int ubifs_jnl_write_data(struct ubifs_info *c, const struct inode *inode, int dlen = UBIFS_DATA_NODE_SZ + UBIFS_BLOCK_SIZE * WORST_COMPR_FACTOR; struct ubifs_inode *ui = ubifs_inode(inode); - dbg_jnl("ino %lu, blk %u, len %d, key %s", key_inum(c, key), - key_block(c, key), len, DBGKEY(key)); + dbg_jnl("ino %lu, blk %u, len %d, key %s", + (unsigned long)key_inum(c, key), key_block(c, key), len, + DBGKEY(key)); ubifs_assert(len <= UBIFS_BLOCK_SIZE); data = kmalloc(dlen, GFP_NOFS); @@ -1128,7 +1129,8 @@ int ubifs_jnl_truncate(struct ubifs_info *c, const struct inode *inode, ino_t inum = inode->i_ino; unsigned int blk; - dbg_jnl("ino %lu, size %lld -> %lld", inum, old_size, new_size); + dbg_jnl("ino %lu, size %lld -> %lld", + (unsigned long)inum, old_size, new_size); ubifs_assert(!ui->data_len); ubifs_assert(S_ISREG(inode->i_mode)); ubifs_assert(mutex_is_locked(&ui->ui_mutex)); diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c index 02d3462f4d3e..9bd5a43d4526 100644 --- a/fs/ubifs/orphan.c +++ b/fs/ubifs/orphan.c @@ -105,7 +105,7 @@ int ubifs_add_orphan(struct ubifs_info *c, ino_t inum) list_add_tail(&orphan->list, &c->orph_list); list_add_tail(&orphan->new_list, &c->orph_new); spin_unlock(&c->orphan_lock); - dbg_gen("ino %lu", inum); + dbg_gen("ino %lu", (unsigned long)inum); return 0; } @@ -132,14 +132,16 @@ void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum) else { if (o->dnext) { spin_unlock(&c->orphan_lock); - dbg_gen("deleted twice ino %lu", inum); + dbg_gen("deleted twice ino %lu", + (unsigned long)inum); return; } if (o->cnext) { o->dnext = c->orph_dnext; c->orph_dnext = o; spin_unlock(&c->orphan_lock); - dbg_gen("delete later ino %lu", inum); + dbg_gen("delete later ino %lu", + (unsigned long)inum); return; } rb_erase(p, &c->orph_tree); @@ -151,12 +153,12 @@ void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum) } spin_unlock(&c->orphan_lock); kfree(o); - dbg_gen("inum %lu", inum); + dbg_gen("inum %lu", (unsigned long)inum); return; } } spin_unlock(&c->orphan_lock); - dbg_err("missing orphan ino %lu", inum); + dbg_err("missing orphan ino %lu", (unsigned long)inum); dbg_dump_stack(); } @@ -448,7 +450,7 @@ static void erase_deleted(struct ubifs_info *c) rb_erase(&orphan->rb, &c->orph_tree); list_del(&orphan->list); c->tot_orphans -= 1; - dbg_gen("deleting orphan ino %lu", orphan->inum); + dbg_gen("deleting orphan ino %lu", (unsigned long)orphan->inum); kfree(orphan); } c->orph_dnext = NULL; @@ -536,8 +538,8 @@ static int insert_dead_orphan(struct ubifs_info *c, ino_t inum) list_add_tail(&orphan->list, &c->orph_list); orphan->dnext = c->orph_dnext; c->orph_dnext = orphan; - dbg_mnt("ino %lu, new %d, tot %d", - inum, c->new_orphans, c->tot_orphans); + dbg_mnt("ino %lu, new %d, tot %d", (unsigned long)inum, + c->new_orphans, c->tot_orphans); return 0; } @@ -609,7 +611,8 @@ static int do_kill_orphans(struct ubifs_info *c, struct ubifs_scan_leb *sleb, n = (le32_to_cpu(orph->ch.len) - UBIFS_ORPH_NODE_SZ) >> 3; for (i = 0; i < n; i++) { inum = le64_to_cpu(orph->inos[i]); - dbg_rcvry("deleting orphaned inode %lu", inum); + dbg_rcvry("deleting orphaned inode %lu", + (unsigned long)inum); err = ubifs_tnc_remove_ino(c, inum); if (err) return err; @@ -840,8 +843,8 @@ static int dbg_orphan_check(struct ubifs_info *c, struct ubifs_zbranch *zbr, if (inum != ci->last_ino) { /* Lowest node type is the inode node, so it comes first */ if (key_type(c, &zbr->key) != UBIFS_INO_KEY) - ubifs_err("found orphan node ino %lu, type %d", inum, - key_type(c, &zbr->key)); + ubifs_err("found orphan node ino %lu, type %d", + (unsigned long)inum, key_type(c, &zbr->key)); ci->last_ino = inum; ci->tot_inos += 1; err = ubifs_tnc_read_node(c, zbr, ci->node); @@ -853,7 +856,8 @@ static int dbg_orphan_check(struct ubifs_info *c, struct ubifs_zbranch *zbr, /* Must be recorded as an orphan */ if (!dbg_find_check_orphan(&ci->root, inum) && !dbg_find_orphan(c, inum)) { - ubifs_err("missing orphan, ino %lu", inum); + ubifs_err("missing orphan, ino %lu", + (unsigned long)inum); ci->missing += 1; } } diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index bed97421b972..90acac603e63 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -1435,13 +1435,13 @@ static int fix_size_in_place(struct ubifs_info *c, struct size_entry *e) err = ubi_leb_change(c->ubi, lnum, c->sbuf, len, UBI_UNKNOWN); if (err) goto out; - dbg_rcvry("inode %lu at %d:%d size %lld -> %lld ", e->inum, lnum, offs, - i_size, e->d_size); + dbg_rcvry("inode %lu at %d:%d size %lld -> %lld ", + (unsigned long)e->inum, lnum, offs, i_size, e->d_size); return 0; out: ubifs_warn("inode %lu failed to fix size %lld -> %lld error %d", - e->inum, e->i_size, e->d_size, err); + (unsigned long)e->inum, e->i_size, e->d_size, err); return err; } @@ -1472,7 +1472,8 @@ int ubifs_recover_size(struct ubifs_info *c) return err; if (err == -ENOENT) { /* Remove data nodes that have no inode */ - dbg_rcvry("removing ino %lu", e->inum); + dbg_rcvry("removing ino %lu", + (unsigned long)e->inum); err = ubifs_tnc_remove_ino(c, e->inum); if (err) return err; @@ -1493,8 +1494,8 @@ int ubifs_recover_size(struct ubifs_info *c) return PTR_ERR(inode); if (inode->i_size < e->d_size) { dbg_rcvry("ino %lu size %lld -> %lld", - e->inum, e->d_size, - inode->i_size); + (unsigned long)e->inum, + e->d_size, inode->i_size); inode->i_size = e->d_size; ubifs_inode(inode)->ui_size = e->d_size; e->inode = inode; diff --git a/fs/ubifs/replay.c b/fs/ubifs/replay.c index 7399692af859..21f7d047c306 100644 --- a/fs/ubifs/replay.c +++ b/fs/ubifs/replay.c @@ -1065,7 +1065,7 @@ int ubifs_replay_journal(struct ubifs_info *c) ubifs_assert(c->bud_bytes <= c->max_bud_bytes || c->need_recovery); dbg_mnt("finished, log head LEB %d:%d, max_sqnum %llu, " "highest_inum %lu", c->lhead_lnum, c->lhead_offs, c->max_sqnum, - c->highest_inum); + (unsigned long)c->highest_inum); out: destroy_replay_tree(c); destroy_bud_list(c); diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c index d27fd918b9c9..99e9a744cfd0 100644 --- a/fs/ubifs/tnc.c +++ b/fs/ubifs/tnc.c @@ -2677,7 +2677,7 @@ int ubifs_tnc_remove_ino(struct ubifs_info *c, ino_t inum) struct ubifs_dent_node *xent, *pxent = NULL; struct qstr nm = { .name = NULL }; - dbg_tnc("ino %lu", inum); + dbg_tnc("ino %lu", (unsigned long)inum); /* * Walk all extended attribute entries and remove them together with @@ -2697,7 +2697,8 @@ int ubifs_tnc_remove_ino(struct ubifs_info *c, ino_t inum) } xattr_inum = le64_to_cpu(xent->inum); - dbg_tnc("xent '%s', ino %lu", xent->name, xattr_inum); + dbg_tnc("xent '%s', ino %lu", xent->name, + (unsigned long)xattr_inum); nm.name = xent->name; nm.len = le16_to_cpu(xent->nlen); -- cgit v1.2.3 From 7e2d9bfa4eabee3e1919a40f20d2ef8b569bd07e Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 5 Nov 2008 16:09:04 +0200 Subject: UBIFS: allow for gaps when dirtying the LPT The LPT may have gaps in it because initially empty LEBs are not added by mkfs.ubifs - because it does not know how many there are. Then UBIFS allocates empty LEBs in the reverse order that they are discovered i.e. they are added to, and removed from, the front of a list. That creates a gap in the middle of the LPT. The function dirtying the LPT tree (for the purpose of small model garbage collection) assumed that a gap could only occur at the very end of the LPT and stopped dirtying prematurely, which in turn resulted in the LPT running out of space - something that is designed to be impossible. Signed-off-by: Adrian Hunter --- fs/ubifs/lpt_commit.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c index eed5a0025d63..a41434b42785 100644 --- a/fs/ubifs/lpt_commit.c +++ b/fs/ubifs/lpt_commit.c @@ -571,8 +571,6 @@ static struct ubifs_pnode *next_pnode(struct ubifs_info *c, /* We assume here that LEB zero is never an LPT LEB */ if (nnode->nbranch[iip].lnum) return ubifs_get_pnode(c, nnode, iip); - else - return NULL; } /* Go up while can't go right */ -- cgit v1.2.3 From eb60fa1066622ddb2278732cf61e0c4544e82c6f Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 10 Nov 2008 15:28:59 +0900 Subject: block: fix add_partition() error path Partition stats structure was not freed on devt allocation failure path. Fix it. Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe --- fs/partitions/check.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 633f7a0ebb2c..90bcf136a9de 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -395,7 +395,7 @@ int add_partition(struct gendisk *disk, int partno, err = blk_alloc_devt(p, &devt); if (err) - goto out_free; + goto out_free_stats; pdev->devt = devt; /* delay uevent until 'holders' subdir is created */ @@ -426,6 +426,8 @@ int add_partition(struct gendisk *disk, int partno, return 0; +out_free_stats: + free_part_stats(p); out_free: kfree(p); return err; -- cgit v1.2.3 From ba32929a91fe2c0628f5be62d1597b379c8d3062 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 10 Nov 2008 15:29:58 +0900 Subject: block: make add_partition() return pointer to hd_struct Make add_partition() return pointer to the new hd_struct on success and ERR_PTR() value on failure. This change will be used to fix md autodetection bug. Signed-off-by: Tejun Heo Cc: Neil Brown Signed-off-by: Jens Axboe --- block/ioctl.c | 7 +++---- fs/partitions/check.c | 25 +++++++++++++------------ include/linux/genhd.h | 4 +++- 3 files changed, 19 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/block/ioctl.c b/block/ioctl.c index c832d639b6e2..d03985b04d67 100644 --- a/block/ioctl.c +++ b/block/ioctl.c @@ -18,7 +18,6 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user struct disk_part_iter piter; long long start, length; int partno; - int err; if (!capable(CAP_SYS_ADMIN)) return -EACCES; @@ -61,10 +60,10 @@ static int blkpg_ioctl(struct block_device *bdev, struct blkpg_ioctl_arg __user disk_part_iter_exit(&piter); /* all seems OK */ - err = add_partition(disk, partno, start, length, - ADDPART_FLAG_NONE); + part = add_partition(disk, partno, start, length, + ADDPART_FLAG_NONE); mutex_unlock(&bdev->bd_mutex); - return err; + return IS_ERR(part) ? PTR_ERR(part) : 0; case BLKPG_DEL_PARTITION: part = disk_get_part(disk, partno); if (!part) diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 90bcf136a9de..633025340239 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -348,8 +348,8 @@ static ssize_t whole_disk_show(struct device *dev, static DEVICE_ATTR(whole_disk, S_IRUSR | S_IRGRP | S_IROTH, whole_disk_show, NULL); -int add_partition(struct gendisk *disk, int partno, - sector_t start, sector_t len, int flags) +struct hd_struct *add_partition(struct gendisk *disk, int partno, + sector_t start, sector_t len, int flags) { struct hd_struct *p; dev_t devt = MKDEV(0, 0); @@ -361,15 +361,15 @@ int add_partition(struct gendisk *disk, int partno, err = disk_expand_part_tbl(disk, partno); if (err) - return err; + return ERR_PTR(err); ptbl = disk->part_tbl; if (ptbl->part[partno]) - return -EBUSY; + return ERR_PTR(-EBUSY); p = kzalloc(sizeof(*p), GFP_KERNEL); if (!p) - return -ENOMEM; + return ERR_PTR(-EBUSY); if (!init_part_stats(p)) { err = -ENOMEM; @@ -424,20 +424,20 @@ int add_partition(struct gendisk *disk, int partno, if (!ddev->uevent_suppress) kobject_uevent(&pdev->kobj, KOBJ_ADD); - return 0; + return p; out_free_stats: free_part_stats(p); out_free: kfree(p); - return err; + return ERR_PTR(err); out_del: kobject_put(p->holder_dir); device_del(pdev); out_put: put_device(pdev); blk_free_devt(devt); - return err; + return ERR_PTR(err); } /* Not exported, helper to add_disk(). */ @@ -568,10 +568,11 @@ int rescan_partitions(struct gendisk *disk, struct block_device *bdev) disk->disk_name, p, (unsigned long long) size); size = get_capacity(disk) - from; } - res = add_partition(disk, p, from, size, state->parts[p].flags); - if (res) { - printk(KERN_ERR " %s: p%d could not be added: %d\n", - disk->disk_name, p, -res); + part = add_partition(disk, p, from, size, + state->parts[p].flags); + if (IS_ERR(part)) { + printk(KERN_ERR " %s: p%d could not be added: %ld\n", + disk->disk_name, p, -PTR_ERR(part)); continue; } #ifdef CONFIG_BLK_DEV_MD diff --git a/include/linux/genhd.h b/include/linux/genhd.h index e439e6aed832..3df7742ce246 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -522,7 +522,9 @@ extern char *disk_name (struct gendisk *hd, int partno, char *buf); extern int disk_expand_part_tbl(struct gendisk *disk, int target); extern int rescan_partitions(struct gendisk *disk, struct block_device *bdev); -extern int __must_check add_partition(struct gendisk *, int, sector_t, sector_t, int); +extern struct hd_struct * __must_check add_partition(struct gendisk *disk, + int partno, sector_t start, + sector_t len, int flags); extern void delete_partition(struct gendisk *, int); extern void printk_all_partitions(void); -- cgit v1.2.3 From 55e8e30c382d25c34f8aafcc78efec948571a941 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 10 Nov 2008 15:30:47 +0900 Subject: block/md: fix md autodetection Block ext devt conversion missed md_autodetect_dev() call in rescan_partitions() leaving md autodetect unable to see partitions. Fix it. Signed-off-by: Tejun Heo Cc: Neil Brown Signed-off-by: Jens Axboe --- fs/partitions/check.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 633025340239..6d5b213b8a9b 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -577,7 +577,7 @@ int rescan_partitions(struct gendisk *disk, struct block_device *bdev) } #ifdef CONFIG_BLK_DEV_MD if (state->parts[p].flags & ADDPART_FLAG_RAID) - md_autodetect_dev(bdev->bd_dev+p); + md_autodetect_dev(part_to_dev(part)->devt); #endif } kfree(state); -- cgit v1.2.3 From bfb59820ee46616a7bdb4af6b8f7e109646de6ec Mon Sep 17 00:00:00 2001 From: Steve French Date: Tue, 18 Nov 2008 16:33:48 +0000 Subject: [CIFS] fix check for dead tcon in smb_init This was recently changed to check for need_reconnect, but should actually be a check for a tidStatus of CifsExiting. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/CHANGES | 6 +++++- fs/cifs/cifssmb.c | 2 +- 2 files changed, 6 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index 8855331b2fba..e078b7aea143 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -8,7 +8,11 @@ handling fcntl(F_SETLEASE). Convert cifs to using blocking tcp sends, and also let tcp autotune the socket send and receive buffers. This reduces the number of EAGAIN errors returned by TCP/IP in high stress workloads (and the number of retries on socket writes -when sending large SMBWriteX requests). +when sending large SMBWriteX requests). Fix case in which a portion of +data can in some cases not get written to the file on the server before the +file is closed. Fix DFS parsing to properly handle path consumed field, +and to handle certain codepage conversions better. Fix mount and +umount race that can cause oops in mount or umount or reconnect. Version 1.54 ------------ diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index bdda46dd435a..2af8626ced43 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -295,7 +295,7 @@ smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, check for tcp and smb session status done differently for those three - in the calling routine */ if (tcon) { - if (tcon->need_reconnect) { + if (tcon->tidStatus == CifsExiting) { /* only tree disconnect, open, and write, (and ulogoff which does not have tcon) are allowed as we start force umount */ -- cgit v1.2.3 From ac97b9f9a2d0b83488e0bbcb8517b229d5c9b142 Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Wed, 19 Nov 2008 15:36:28 -0800 Subject: eCryptfs: Allocate up to two scatterlists for crypto ops on keys I have received some reports of out-of-memory errors on some older AMD architectures. These errors are what I would expect to see if crypt_stat->key were split between two separate pages. eCryptfs should not assume that any of the memory sent through virt_to_scatterlist() is all contained in a single page, and so this patch allocates two scatterlist structs instead of one when processing keys. I have received confirmation from one person affected by this bug that this patch resolves the issue for him, and so I am submitting it for inclusion in a future stable release. Note that virt_to_scatterlist() runs sg_init_table() on the scatterlist structs passed to it, so the calls to sg_init_table() in decrypt_passphrase_encrypted_session_key() are redundant. Signed-off-by: Michael Halcrow Reported-by: Paulo J. S. Silva Cc: "Leon Woestenberg" Cc: Tim Gardner Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/keystore.c | 31 ++++++++++++++----------------- 1 file changed, 14 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index e22bc3961345..0d713b691941 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -1037,17 +1037,14 @@ static int decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, struct ecryptfs_crypt_stat *crypt_stat) { - struct scatterlist dst_sg; - struct scatterlist src_sg; + struct scatterlist dst_sg[2]; + struct scatterlist src_sg[2]; struct mutex *tfm_mutex; struct blkcipher_desc desc = { .flags = CRYPTO_TFM_REQ_MAY_SLEEP }; int rc = 0; - sg_init_table(&dst_sg, 1); - sg_init_table(&src_sg, 1); - if (unlikely(ecryptfs_verbosity > 0)) { ecryptfs_printk( KERN_DEBUG, "Session key encryption key (size [%d]):\n", @@ -1066,8 +1063,8 @@ decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, } rc = virt_to_scatterlist(auth_tok->session_key.encrypted_key, auth_tok->session_key.encrypted_key_size, - &src_sg, 1); - if (rc != 1) { + src_sg, 2); + if (rc < 1 || rc > 2) { printk(KERN_ERR "Internal error whilst attempting to convert " "auth_tok->session_key.encrypted_key to scatterlist; " "expected rc = 1; got rc = [%d]. " @@ -1079,8 +1076,8 @@ decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, auth_tok->session_key.encrypted_key_size; rc = virt_to_scatterlist(auth_tok->session_key.decrypted_key, auth_tok->session_key.decrypted_key_size, - &dst_sg, 1); - if (rc != 1) { + dst_sg, 2); + if (rc < 1 || rc > 2) { printk(KERN_ERR "Internal error whilst attempting to convert " "auth_tok->session_key.decrypted_key to scatterlist; " "expected rc = 1; got rc = [%d]\n", rc); @@ -1096,7 +1093,7 @@ decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, rc = -EINVAL; goto out; } - rc = crypto_blkcipher_decrypt(&desc, &dst_sg, &src_sg, + rc = crypto_blkcipher_decrypt(&desc, dst_sg, src_sg, auth_tok->session_key.encrypted_key_size); mutex_unlock(tfm_mutex); if (unlikely(rc)) { @@ -1539,8 +1536,8 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, size_t i; size_t encrypted_session_key_valid = 0; char session_key_encryption_key[ECRYPTFS_MAX_KEY_BYTES]; - struct scatterlist dst_sg; - struct scatterlist src_sg; + struct scatterlist dst_sg[2]; + struct scatterlist src_sg[2]; struct mutex *tfm_mutex = NULL; u8 cipher_code; size_t packet_size_length; @@ -1619,8 +1616,8 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, ecryptfs_dump_hex(session_key_encryption_key, 16); } rc = virt_to_scatterlist(crypt_stat->key, key_rec->enc_key_size, - &src_sg, 1); - if (rc != 1) { + src_sg, 2); + if (rc < 1 || rc > 2) { ecryptfs_printk(KERN_ERR, "Error generating scatterlist " "for crypt_stat session key; expected rc = 1; " "got rc = [%d]. key_rec->enc_key_size = [%d]\n", @@ -1629,8 +1626,8 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, goto out; } rc = virt_to_scatterlist(key_rec->enc_key, key_rec->enc_key_size, - &dst_sg, 1); - if (rc != 1) { + dst_sg, 2); + if (rc < 1 || rc > 2) { ecryptfs_printk(KERN_ERR, "Error generating scatterlist " "for crypt_stat encrypted session key; " "expected rc = 1; got rc = [%d]. " @@ -1651,7 +1648,7 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, rc = 0; ecryptfs_printk(KERN_DEBUG, "Encrypting [%d] bytes of the key\n", crypt_stat->key_size); - rc = crypto_blkcipher_encrypt(&desc, &dst_sg, &src_sg, + rc = crypto_blkcipher_encrypt(&desc, dst_sg, src_sg, (*key_rec).enc_key_size); mutex_unlock(tfm_mutex); if (rc) { -- cgit v1.2.3 From f9454548e17cd56bad081bd7d55a09b001950cbb Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Wed, 19 Nov 2008 15:36:38 -0800 Subject: don't unlink an active swapfile Peter Cordes is sorry that he rm'ed his swapfiles while they were in use, he then had no pathname to swapoff. It's a curious little oversight, but not one worth a lot of hackery. Kudos to Willy Tarreau for turning this around from a discussion of synthetic pathnames to how to prevent unlink. Mimic immutable: prohibit unlinking an active swapfile in may_delete() (and don't worry my little head over the tiny race window). Signed-off-by: Hugh Dickins Cc: Willy Tarreau Acked-by: Christoph Hellwig Cc: Peter Cordes Cc: Bodo Eggert <7eggert@gmx.de> Cc: David Newall Cc: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/namei.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 09ce58e49e72..d34e0f9681c6 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1378,7 +1378,7 @@ static int may_delete(struct inode *dir,struct dentry *victim,int isdir) if (IS_APPEND(dir)) return -EPERM; if (check_sticky(dir, victim->d_inode)||IS_APPEND(victim->d_inode)|| - IS_IMMUTABLE(victim->d_inode)) + IS_IMMUTABLE(victim->d_inode) || IS_SWAPFILE(victim->d_inode)) return -EPERM; if (isdir) { if (!S_ISDIR(victim->d_inode->i_mode)) -- cgit v1.2.3 From ea7e743e49b94749fc739baaf160809ed279aeda Mon Sep 17 00:00:00 2001 From: WANG Cong Date: Wed, 19 Nov 2008 15:36:46 -0800 Subject: hostfs: fix a duplicated global function name fs/hostfs/hostfs_user.c defines do_readlink() as non-static, and so does fs/xfs/linux-2.6/xfs_ioctl.c when CONFIG_XFS_DEBUG=y. So rename do_readlink() in hostfs to hostfs_do_readlink(). I think it's better if XFS guys will also rename their do_readlink(), it's not necessary to use such a general name. Signed-off-by: WANG Cong Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hostfs/hostfs.h | 2 +- fs/hostfs/hostfs_kern.c | 4 ++-- fs/hostfs/hostfs_user.c | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/hostfs/hostfs.h b/fs/hostfs/hostfs.h index 6ae9011b95eb..2f34f8f2134b 100644 --- a/fs/hostfs/hostfs.h +++ b/fs/hostfs/hostfs.h @@ -81,7 +81,7 @@ extern int do_rmdir(const char *file); extern int do_mknod(const char *file, int mode, unsigned int major, unsigned int minor); extern int link_file(const char *from, const char *to); -extern int do_readlink(char *file, char *buf, int size); +extern int hostfs_do_readlink(char *file, char *buf, int size); extern int rename_file(char *from, char *to); extern int do_statfs(char *root, long *bsize_out, long long *blocks_out, long long *bfree_out, long long *bavail_out, diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index 7f34f4385de0..3a31451ac170 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -168,7 +168,7 @@ static char *follow_link(char *link) if (name == NULL) goto out; - n = do_readlink(link, name, len); + n = hostfs_do_readlink(link, name, len); if (n < len) break; len *= 2; @@ -943,7 +943,7 @@ int hostfs_link_readpage(struct file *file, struct page *page) name = inode_name(page->mapping->host, 0); if (name == NULL) return -ENOMEM; - err = do_readlink(name, buffer, PAGE_CACHE_SIZE); + err = hostfs_do_readlink(name, buffer, PAGE_CACHE_SIZE); kfree(name); if (err == PAGE_CACHE_SIZE) err = -E2BIG; diff --git a/fs/hostfs/hostfs_user.c b/fs/hostfs/hostfs_user.c index 53fd0a67c11a..b79424f93282 100644 --- a/fs/hostfs/hostfs_user.c +++ b/fs/hostfs/hostfs_user.c @@ -377,7 +377,7 @@ int link_file(const char *to, const char *from) return 0; } -int do_readlink(char *file, char *buf, int size) +int hostfs_do_readlink(char *file, char *buf, int size) { int n; -- cgit v1.2.3 From ddb4cbfc53aa0913ee8da059fcbf628d14f40f63 Mon Sep 17 00:00:00 2001 From: Steve French Date: Thu, 20 Nov 2008 20:00:44 +0000 Subject: [CIFS] Do not attempt to close invalidated file handles If a connection with open file handles has gone down and come back up and reconnected without reopening the file handle yet, do not attempt to send an SMB close request for this handle in cifs_close. We were checking for the connection being invalid in cifs_close but since the connection may have been reconnected we also need to check whether the file handle was marked invalid (otherwise we could close the wrong file handle by accident). Acked-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsglob.h | 10 +++++++++- fs/cifs/file.c | 21 ++++++++++++++------- fs/cifs/misc.c | 3 +++ fs/cifs/readdir.c | 5 ++++- 4 files changed, 30 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index f1ae1f57c30d..c57c0565547f 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -606,7 +606,15 @@ GLOBAL_EXTERN struct list_head cifs_tcp_ses_list; * changes to the tcon->tidStatus should be done while holding this lock. */ GLOBAL_EXTERN rwlock_t cifs_tcp_ses_lock; -GLOBAL_EXTERN rwlock_t GlobalSMBSeslock; /* protects list inserts on 3 above */ + +/* + * This lock protects the cifs_file->llist and cifs_file->flist + * list operations, and updates to some flags (cifs_file->invalidHandle) + * It will be moved to either use the tcon->stat_lock or equivalent later. + * If cifs_tcp_ses_lock and the lock below are both needed to be held, then + * the cifs_tcp_ses_lock must be grabbed first and released last. + */ +GLOBAL_EXTERN rwlock_t GlobalSMBSeslock; GLOBAL_EXTERN struct list_head GlobalOplock_Q; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 6449e1aae621..b691b893a848 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -488,12 +488,13 @@ int cifs_close(struct inode *inode, struct file *file) pTcon = cifs_sb->tcon; if (pSMBFile) { struct cifsLockInfo *li, *tmp; - + write_lock(&GlobalSMBSeslock); pSMBFile->closePend = true; if (pTcon) { /* no sense reconnecting to close a file that is already closed */ if (!pTcon->need_reconnect) { + write_unlock(&GlobalSMBSeslock); timeout = 2; while ((atomic_read(&pSMBFile->wrtPending) != 0) && (timeout <= 2048)) { @@ -510,12 +511,15 @@ int cifs_close(struct inode *inode, struct file *file) timeout *= 4; } if (atomic_read(&pSMBFile->wrtPending)) - cERROR(1, - ("close with pending writes")); - rc = CIFSSMBClose(xid, pTcon, + cERROR(1, ("close with pending write")); + if (!pTcon->need_reconnect && + !pSMBFile->invalidHandle) + rc = CIFSSMBClose(xid, pTcon, pSMBFile->netfid); - } - } + } else + write_unlock(&GlobalSMBSeslock); + } else + write_unlock(&GlobalSMBSeslock); /* Delete any outstanding lock records. We'll lose them when the file is closed anyway. */ @@ -587,15 +591,18 @@ int cifs_closedir(struct inode *inode, struct file *file) pTcon = cifs_sb->tcon; cFYI(1, ("Freeing private data in close dir")); + write_lock(&GlobalSMBSeslock); if (!pCFileStruct->srch_inf.endOfSearch && !pCFileStruct->invalidHandle) { pCFileStruct->invalidHandle = true; + write_unlock(&GlobalSMBSeslock); rc = CIFSFindClose(xid, pTcon, pCFileStruct->netfid); cFYI(1, ("Closing uncompleted readdir with rc %d", rc)); /* not much we can do if it fails anyway, ignore rc */ rc = 0; - } + } else + write_unlock(&GlobalSMBSeslock); ptmp = pCFileStruct->srch_inf.ntwrk_buf_start; if (ptmp) { cFYI(1, ("closedir free smb buf in srch struct")); diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index addd1dcc2d79..9ee3f689c2b0 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -555,12 +555,14 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) continue; cifs_stats_inc(&tcon->num_oplock_brks); + write_lock(&GlobalSMBSeslock); list_for_each(tmp2, &tcon->openFileList) { netfile = list_entry(tmp2, struct cifsFileInfo, tlist); if (pSMB->Fid != netfile->netfid) continue; + write_unlock(&GlobalSMBSeslock); read_unlock(&cifs_tcp_ses_lock); cFYI(1, ("file id match, oplock break")); pCifsInode = CIFS_I(netfile->pInode); @@ -576,6 +578,7 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) return true; } + write_unlock(&GlobalSMBSeslock); read_unlock(&cifs_tcp_ses_lock); cFYI(1, ("No matching file for oplock break")); return true; diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 58d57299f2a0..9f51f9bf0292 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -741,11 +741,14 @@ static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, (index_to_find < first_entry_in_buffer)) { /* close and restart search */ cFYI(1, ("search backing up - close and restart search")); + write_lock(&GlobalSMBSeslock); if (!cifsFile->srch_inf.endOfSearch && !cifsFile->invalidHandle) { cifsFile->invalidHandle = true; + write_unlock(&GlobalSMBSeslock); CIFSFindClose(xid, pTcon, cifsFile->netfid); - } + } else + write_unlock(&GlobalSMBSeslock); if (cifsFile->srch_inf.ntwrk_buf_start) { cFYI(1, ("freeing SMB ff cache buf on search rewind")); if (cifsFile->srch_inf.smallBuf) -- cgit v1.2.3 From 39ce81ce7168aa7226fb9f182c3a2b57060d0905 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 18 Nov 2008 18:09:49 +0200 Subject: UBIFS: do not print scary memory allocation warnings Bulk-read allocates a lot of memory with 'kmalloc()', and when it is/gets fragmented 'kmalloc()' fails with a scarry warning. But because bulk-read is just an optimization, UBIFS keeps working fine. Supress the warning by passing __GFP_NOWARN option to 'kmalloc()'. This patch also introduces a macro for the magic 128KiB constant. This is just neater. Note, this is not really fixes the problem we had, but just hides the warnings. The further patches fix the problem. Signed-off-by: Artem Bityutskiy --- fs/ubifs/file.c | 4 ++-- fs/ubifs/super.c | 17 ++++++++++++----- fs/ubifs/ubifs.h | 2 +- 3 files changed, 15 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 9124eee73aea..8be827cc7078 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -705,12 +705,12 @@ static int ubifs_do_bulk_read(struct ubifs_info *c, struct page *page1) int err, page_idx, page_cnt, ret = 0, n = 0; loff_t isize; - bu = kmalloc(sizeof(struct bu_info), GFP_NOFS); + bu = kmalloc(sizeof(struct bu_info), GFP_NOFS | __GFP_NOWARN); if (!bu) return 0; bu->buf_len = c->bulk_read_buf_size; - bu->buf = kmalloc(bu->buf_len, GFP_NOFS); + bu->buf = kmalloc(bu->buf_len, GFP_NOFS | __GFP_NOWARN); if (!bu->buf) goto out_free; diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 8780efbf40ac..ea493e6f2652 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -36,6 +36,12 @@ #include #include "ubifs.h" +/* + * Maximum amount of memory we may 'kmalloc()' without worrying that we are + * allocating too much. + */ +#define UBIFS_KMALLOC_OK (128*1024) + /* Slab cache for UBIFS inodes */ struct kmem_cache *ubifs_inode_slab; @@ -561,17 +567,18 @@ static int init_constants_early(struct ubifs_info *c) * calculations when reporting free space. */ c->leb_overhead = c->leb_size % UBIFS_MAX_DATA_NODE_SZ; + /* Buffer size for bulk-reads */ c->bulk_read_buf_size = UBIFS_MAX_BULK_READ * UBIFS_MAX_DATA_NODE_SZ; if (c->bulk_read_buf_size > c->leb_size) c->bulk_read_buf_size = c->leb_size; - if (c->bulk_read_buf_size > 128 * 1024) { - /* Check if we can kmalloc more than 128KiB */ - void *try = kmalloc(c->bulk_read_buf_size, GFP_KERNEL); - + if (c->bulk_read_buf_size > UBIFS_KMALLOC_OK) { + /* Check if we can kmalloc that much */ + void *try = kmalloc(c->bulk_read_buf_size, + GFP_KERNEL | __GFP_NOWARN); kfree(try); if (!try) - c->bulk_read_buf_size = 128 * 1024; + c->bulk_read_buf_size = UBIFS_KMALLOC_OK; } return 0; } diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index a7bd32fa15b9..06ba51efd65d 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -753,7 +753,7 @@ struct ubifs_znode { }; /** - * struct bu_info - bulk-read information + * struct bu_info - bulk-read information. * @key: first data node key * @zbranch: zbranches of data nodes to bulk read * @buf: buffer to read into -- cgit v1.2.3 From 6c0c42cdfd73fb161417403d8d077cb136e10bbf Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 18 Nov 2008 20:20:05 +0200 Subject: UBIFS: do not allocate too much Bulk-read allocates 128KiB or more using kmalloc. The allocation starts failing often when the memory gets fragmented. UBIFS still works fine in this case, because it falls-back to standard (non-optimized) read method, though. This patch teaches bulk-read to allocate exactly the amount of memory it needs, instead of allocating 128KiB every time. This patch is also a preparation to the further fix where we'll have a pre-allocated bulk-read buffer as well. For example, now the @bu object is prepared in 'ubifs_bulk_read()', so we could path either pre-allocated or allocated information to 'ubifs_do_bulk_read()' later. Or teaching 'ubifs_do_bulk_read()' not to allocate 'bu->buf' if it is already there. Signed-off-by: Artem Bityutskiy --- fs/ubifs/file.c | 70 +++++++++++++++++++++++++++++++++++++------------------- fs/ubifs/super.c | 12 +++++----- fs/ubifs/tnc.c | 7 +++++- fs/ubifs/ubifs.h | 4 ++-- 4 files changed, 60 insertions(+), 33 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 8be827cc7078..0c5c27d63f6e 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -691,32 +691,22 @@ out_err: /** * ubifs_do_bulk_read - do bulk-read. * @c: UBIFS file-system description object - * @page1: first page + * @bu: bulk-read information + * @page1: first page to read * * This function returns %1 if the bulk-read is done, otherwise %0 is returned. */ -static int ubifs_do_bulk_read(struct ubifs_info *c, struct page *page1) +static int ubifs_do_bulk_read(struct ubifs_info *c, struct bu_info *bu, + struct page *page1) { pgoff_t offset = page1->index, end_index; struct address_space *mapping = page1->mapping; struct inode *inode = mapping->host; struct ubifs_inode *ui = ubifs_inode(inode); - struct bu_info *bu; int err, page_idx, page_cnt, ret = 0, n = 0; + int allocate = bu->buf ? 0 : 1; loff_t isize; - bu = kmalloc(sizeof(struct bu_info), GFP_NOFS | __GFP_NOWARN); - if (!bu) - return 0; - - bu->buf_len = c->bulk_read_buf_size; - bu->buf = kmalloc(bu->buf_len, GFP_NOFS | __GFP_NOWARN); - if (!bu->buf) - goto out_free; - - data_key_init(c, &bu->key, inode->i_ino, - offset << UBIFS_BLOCKS_PER_PAGE_SHIFT); - err = ubifs_tnc_get_bu_keys(c, bu); if (err) goto out_warn; @@ -735,12 +725,25 @@ static int ubifs_do_bulk_read(struct ubifs_info *c, struct page *page1) * together. If all the pages were like this, bulk-read would * reduce performance, so we turn it off for a while. */ - ui->read_in_a_row = 0; - ui->bulk_read = 0; - goto out_free; + goto out_bu_off; } if (bu->cnt) { + if (allocate) { + /* + * Allocate bulk-read buffer depending on how many data + * nodes we are going to read. + */ + bu->buf_len = bu->zbranch[bu->cnt - 1].offs + + bu->zbranch[bu->cnt - 1].len - + bu->zbranch[0].offs; + ubifs_assert(bu->buf_len > 0); + ubifs_assert(bu->buf_len <= c->leb_size); + bu->buf = kmalloc(bu->buf_len, GFP_NOFS | __GFP_NOWARN); + if (!bu->buf) + goto out_bu_off; + } + err = ubifs_tnc_bulk_read(c, bu); if (err) goto out_warn; @@ -779,13 +782,17 @@ static int ubifs_do_bulk_read(struct ubifs_info *c, struct page *page1) ui->last_page_read = offset + page_idx - 1; out_free: - kfree(bu->buf); - kfree(bu); + if (allocate) + kfree(bu->buf); return ret; out_warn: ubifs_warn("ignoring error %d and skipping bulk-read", err); goto out_free; + +out_bu_off: + ui->read_in_a_row = ui->bulk_read = 0; + goto out_free; } /** @@ -803,18 +810,20 @@ static int ubifs_bulk_read(struct page *page) struct ubifs_info *c = inode->i_sb->s_fs_info; struct ubifs_inode *ui = ubifs_inode(inode); pgoff_t index = page->index, last_page_read = ui->last_page_read; - int ret = 0; + struct bu_info *bu; + int err = 0; ui->last_page_read = index; - if (!c->bulk_read) return 0; + /* * Bulk-read is protected by ui_mutex, but it is an optimization, so * don't bother if we cannot lock the mutex. */ if (!mutex_trylock(&ui->ui_mutex)) return 0; + if (index != last_page_read + 1) { /* Turn off bulk-read if we stop reading sequentially */ ui->read_in_a_row = 1; @@ -822,6 +831,7 @@ static int ubifs_bulk_read(struct page *page) ui->bulk_read = 0; goto out_unlock; } + if (!ui->bulk_read) { ui->read_in_a_row += 1; if (ui->read_in_a_row < 3) @@ -829,10 +839,22 @@ static int ubifs_bulk_read(struct page *page) /* Three reads in a row, so switch on bulk-read */ ui->bulk_read = 1; } - ret = ubifs_do_bulk_read(c, page); + + bu = kmalloc(sizeof(struct bu_info), GFP_NOFS | __GFP_NOWARN); + if (!bu) + return 0; + + bu->buf = NULL; + bu->buf_len = c->max_bu_buf_len; + data_key_init(c, &bu->key, inode->i_ino, + page->index << UBIFS_BLOCKS_PER_PAGE_SHIFT); + + err = ubifs_do_bulk_read(c, bu, page); + kfree(bu); + out_unlock: mutex_unlock(&ui->ui_mutex); - return ret; + return err; } static int ubifs_readpage(struct file *file, struct page *page) diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index ea493e6f2652..1d511569c035 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -569,16 +569,16 @@ static int init_constants_early(struct ubifs_info *c) c->leb_overhead = c->leb_size % UBIFS_MAX_DATA_NODE_SZ; /* Buffer size for bulk-reads */ - c->bulk_read_buf_size = UBIFS_MAX_BULK_READ * UBIFS_MAX_DATA_NODE_SZ; - if (c->bulk_read_buf_size > c->leb_size) - c->bulk_read_buf_size = c->leb_size; - if (c->bulk_read_buf_size > UBIFS_KMALLOC_OK) { + c->max_bu_buf_len = UBIFS_MAX_BULK_READ * UBIFS_MAX_DATA_NODE_SZ; + if (c->max_bu_buf_len > c->leb_size) + c->max_bu_buf_len = c->leb_size; + if (c->max_bu_buf_len > UBIFS_KMALLOC_OK) { /* Check if we can kmalloc that much */ - void *try = kmalloc(c->bulk_read_buf_size, + void *try = kmalloc(c->max_bu_buf_len, GFP_KERNEL | __GFP_NOWARN); kfree(try); if (!try) - c->bulk_read_buf_size = UBIFS_KMALLOC_OK; + c->max_bu_buf_len = UBIFS_KMALLOC_OK; } return 0; } diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c index 99e9a744cfd0..6eef5344a145 100644 --- a/fs/ubifs/tnc.c +++ b/fs/ubifs/tnc.c @@ -1501,7 +1501,12 @@ out: * @bu: bulk-read parameters and results * * Lookup consecutive data node keys for the same inode that reside - * consecutively in the same LEB. + * consecutively in the same LEB. This function returns zero in case of success + * and a negative error code in case of failure. + * + * Note, if the bulk-read buffer length (@bu->buf_len) is known, this function + * makes sure bulk-read nodes fit the buffer. Otherwise, this function prepares + * maxumum possible amount of nodes for bulk-read. */ int ubifs_tnc_get_bu_keys(struct ubifs_info *c, struct bu_info *bu) { diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 06ba51efd65d..870b5c479e95 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -969,7 +969,7 @@ struct ubifs_mount_opts { * @mst_node: master node * @mst_offs: offset of valid master node * @mst_mutex: protects the master node area, @mst_node, and @mst_offs - * @bulk_read_buf_size: buffer size for bulk-reads + * @max_bu_buf_len: maximum bulk-read buffer length * * @log_lebs: number of logical eraseblocks in the log * @log_bytes: log size in bytes @@ -1217,7 +1217,7 @@ struct ubifs_info { struct ubifs_mst_node *mst_node; int mst_offs; struct mutex mst_mutex; - int bulk_read_buf_size; + int max_bu_buf_len; int log_lebs; long long log_bytes; -- cgit v1.2.3 From 3477d204658733aa3a87d3ae03b0327c1e599517 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Wed, 19 Nov 2008 11:53:15 +0200 Subject: UBIFS: pre-allocate bulk-read buffer To avoid memory allocation failure during bulk-read, pre-allocate a bulk-read buffer, so that if there is only one bulk-reader at a time, it would just use the pre-allocated buffer and would not do any memory allocation. However, if there are more than 1 bulk- reader, then only one reader would use the pre-allocated buffer, while the other reader would allocate the buffer for itself. Signed-off-by: Artem Bityutskiy --- fs/ubifs/file.c | 31 +++++++++++++++++++++--------- fs/ubifs/super.c | 57 +++++++++++++++++++++++++++++++++++++++++++++++--------- fs/ubifs/ubifs.h | 6 ++++++ 3 files changed, 76 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 0c5c27d63f6e..2624411d9758 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -811,15 +811,15 @@ static int ubifs_bulk_read(struct page *page) struct ubifs_inode *ui = ubifs_inode(inode); pgoff_t index = page->index, last_page_read = ui->last_page_read; struct bu_info *bu; - int err = 0; + int err = 0, allocated = 0; ui->last_page_read = index; if (!c->bulk_read) return 0; /* - * Bulk-read is protected by ui_mutex, but it is an optimization, so - * don't bother if we cannot lock the mutex. + * Bulk-read is protected by @ui->ui_mutex, but it is an optimization, + * so don't bother if we cannot lock the mutex. */ if (!mutex_trylock(&ui->ui_mutex)) return 0; @@ -840,17 +840,30 @@ static int ubifs_bulk_read(struct page *page) ui->bulk_read = 1; } - bu = kmalloc(sizeof(struct bu_info), GFP_NOFS | __GFP_NOWARN); - if (!bu) - return 0; + /* + * If possible, try to use pre-allocated bulk-read information, which + * is protected by @c->bu_mutex. + */ + if (mutex_trylock(&c->bu_mutex)) + bu = &c->bu; + else { + bu = kmalloc(sizeof(struct bu_info), GFP_NOFS | __GFP_NOWARN); + if (!bu) + goto out_unlock; + + bu->buf = NULL; + allocated = 1; + } - bu->buf = NULL; bu->buf_len = c->max_bu_buf_len; data_key_init(c, &bu->key, inode->i_ino, page->index << UBIFS_BLOCKS_PER_PAGE_SHIFT); - err = ubifs_do_bulk_read(c, bu, page); - kfree(bu); + + if (!allocated) + mutex_unlock(&c->bu_mutex); + else + kfree(bu); out_unlock: mutex_unlock(&ui->ui_mutex); diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 1d511569c035..d80b2aef42b6 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -572,14 +572,6 @@ static int init_constants_early(struct ubifs_info *c) c->max_bu_buf_len = UBIFS_MAX_BULK_READ * UBIFS_MAX_DATA_NODE_SZ; if (c->max_bu_buf_len > c->leb_size) c->max_bu_buf_len = c->leb_size; - if (c->max_bu_buf_len > UBIFS_KMALLOC_OK) { - /* Check if we can kmalloc that much */ - void *try = kmalloc(c->max_bu_buf_len, - GFP_KERNEL | __GFP_NOWARN); - kfree(try); - if (!try) - c->max_bu_buf_len = UBIFS_KMALLOC_OK; - } return 0; } @@ -998,6 +990,34 @@ static void destroy_journal(struct ubifs_info *c) free_buds(c); } +/** + * bu_init - initialize bulk-read information. + * @c: UBIFS file-system description object + */ +static void bu_init(struct ubifs_info *c) +{ + ubifs_assert(c->bulk_read == 1); + + if (c->bu.buf) + return; /* Already initialized */ + +again: + c->bu.buf = kmalloc(c->max_bu_buf_len, GFP_KERNEL | __GFP_NOWARN); + if (!c->bu.buf) { + if (c->max_bu_buf_len > UBIFS_KMALLOC_OK) { + c->max_bu_buf_len = UBIFS_KMALLOC_OK; + goto again; + } + + /* Just disable bulk-read */ + ubifs_warn("Cannot allocate %d bytes of memory for bulk-read, " + "disabling it", c->max_bu_buf_len); + c->mount_opts.bulk_read = 1; + c->bulk_read = 0; + return; + } +} + /** * mount_ubifs - mount UBIFS file-system. * @c: UBIFS file-system description object @@ -1066,6 +1086,13 @@ static int mount_ubifs(struct ubifs_info *c) goto out_free; } + if (c->bulk_read == 1) + bu_init(c); + + /* + * We have to check all CRCs, even for data nodes, when we mount the FS + * (specifically, when we are replaying). + */ c->always_chk_crc = 1; err = ubifs_read_superblock(c); @@ -1296,6 +1323,7 @@ out_cbuf: out_dereg: dbg_failure_mode_deregistration(c); out_free: + kfree(c->bu.buf); vfree(c->ileb_buf); vfree(c->sbuf); kfree(c->bottom_up_buf); @@ -1332,10 +1360,11 @@ static void ubifs_umount(struct ubifs_info *c) kfree(c->cbuf); kfree(c->rcvrd_mst_node); kfree(c->mst_node); + kfree(c->bu.buf); + vfree(c->ileb_buf); vfree(c->sbuf); kfree(c->bottom_up_buf); UBIFS_DBG(vfree(c->dbg_buf)); - vfree(c->ileb_buf); dbg_failure_mode_deregistration(c); } @@ -1633,6 +1662,7 @@ static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data) ubifs_err("invalid or unknown remount parameter"); return err; } + if ((sb->s_flags & MS_RDONLY) && !(*flags & MS_RDONLY)) { err = ubifs_remount_rw(c); if (err) @@ -1640,6 +1670,14 @@ static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data) } else if (!(sb->s_flags & MS_RDONLY) && (*flags & MS_RDONLY)) ubifs_remount_ro(c); + if (c->bulk_read == 1) + bu_init(c); + else { + dbg_gen("disable bulk-read"); + kfree(c->bu.buf); + c->bu.buf = NULL; + } + return 0; } @@ -1730,6 +1768,7 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) mutex_init(&c->log_mutex); mutex_init(&c->mst_mutex); mutex_init(&c->umount_mutex); + mutex_init(&c->bu_mutex); init_waitqueue_head(&c->cmt_wq); c->buds = RB_ROOT; c->old_idx = RB_ROOT; diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 870b5c479e95..46b172560a06 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -969,7 +969,10 @@ struct ubifs_mount_opts { * @mst_node: master node * @mst_offs: offset of valid master node * @mst_mutex: protects the master node area, @mst_node, and @mst_offs + * * @max_bu_buf_len: maximum bulk-read buffer length + * @bu_mutex: protects the pre-allocated bulk-read buffer and @c->bu + * @bu: pre-allocated bulk-read information * * @log_lebs: number of logical eraseblocks in the log * @log_bytes: log size in bytes @@ -1217,7 +1220,10 @@ struct ubifs_info { struct ubifs_mst_node *mst_node; int mst_offs; struct mutex mst_mutex; + int max_bu_buf_len; + struct mutex bu_mutex; + struct bu_info bu; int log_lebs; long long log_bytes; -- cgit v1.2.3 From 2c5e76158fcea6e3b9536a74efa7b5e2e846d374 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 20 Nov 2008 14:36:17 -0600 Subject: nfsd: clean up grace period on early exit If nfsd was shut down before the grace period ended, we could end up with a freed object still on grace_list. Thanks to Jeff Moyer for reporting the resulting list corruption warnings. Signed-off-by: J. Bruce Fields Tested-by: Jeff Moyer --- fs/lockd/svc.c | 1 + fs/nfsd/nfs4state.c | 1 + 2 files changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index c631a83931ce..56b076736b56 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -181,6 +181,7 @@ lockd(void *vrqstp) } flush_signals(current); cancel_delayed_work_sync(&grace_period_end); + locks_end_grace(&lockd_manager); if (nlmsvc_ops) nlmsvc_invalidate_all(); nlm_shutdown_hosts(); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index b0bebc552a11..1a052ac2bde9 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -3261,6 +3261,7 @@ nfs4_state_shutdown(void) { cancel_rearming_delayed_workqueue(laundry_wq, &laundromat_work); destroy_workqueue(laundry_wq); + locks_end_grace(&nfsd4_manager); nfs4_lock_state(); nfs4_release_reclaim(); __nfs4_state_shutdown(); -- cgit v1.2.3 From e4625eb826de4f6774ee602c442ba23b686bdcc7 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 24 Nov 2008 10:32:46 -0600 Subject: nfsd: use of unitialized list head on error exit in nfs4recover.c Thanks to Matthew Dodd for this bug report: A file label issue while running SELinux in MLS mode provoked the following bug, which is a result of use before init on a 'struct list_head'. In nfsd4_list_rec_dir() if the call to dentry_open() fails the 'goto out' skips INIT_LIST_HEAD() which results in the normally improbable case where list_entry() returns NULL. Trace follows. NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory SELinux: Context unconfined_t:object_r:var_lib_nfs_t:s0 is not valid (left unmapped). type=1400 audit(1227298063.609:282): avc: denied { read } for pid=1890 comm="rpc.nfsd" name="v4recovery" dev=dm-0 ino=148726 scontext=system_u:system_r:nfsd_t:s0-s15:c0.c1023 tcontext=system_u:object_r:unlabeled_t:s15:c0.c1023 tclass=dir BUG: unable to handle kernel NULL pointer dereference at 00000004 IP: [] list_del+0x6/0x60 *pde = 0d9ce067 *pte = 00000000 Oops: 0000 [#1] SMP Modules linked in: nfsd lockd nfs_acl auth_rpcgss exportfs autofs4 sunrpc ipv6 dm_multipath scsi_dh ppdev parport_pc sg parport floppy ata_piix pata_acpi ata_generic libata pcnet32 i2c_piix4 mii pcspkr i2c_core dm_snapshot dm_zero dm_mirror dm_log dm_mod BusLogic sd_mod scsi_mod crc_t10dif ext3 jbd mbcache uhci_hcd ohci_hcd ehci_hcd [last unloaded: microcode] Pid: 1890, comm: rpc.nfsd Not tainted (2.6.27.5-37.fc9.i686 #1) EIP: 0060:[] EFLAGS: 00010217 CPU: 0 EIP is at list_del+0x6/0x60 EAX: 00000000 EBX: 00000000 ECX: 00000000 EDX: cd99e480 ESI: cf9caed8 EDI: 00000000 EBP: cf9caebc ESP: cf9caeb8 DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 Process rpc.nfsd (pid: 1890, ti=cf9ca000 task=cf4de580 task.ti=cf9ca000) Stack: 00000000 cf9caef0 d0a9f139 c0496d04 d0a9f217 fffffff3 00000000 00000000 00000000 00000000 cf32b220 00000000 00000008 00000801 cf9caefc d0a9f193 00000000 cf9caf08 d0a9b6ea 00000000 cf9caf1c d0a874f2 cf9c3004 00000008 Call Trace: [] ? nfsd4_list_rec_dir+0xf3/0x13a [nfsd] [] ? do_path_lookup+0x12d/0x175 [] ? load_recdir+0x0/0x26 [nfsd] [] ? nfsd4_recdir_load+0x13/0x34 [nfsd] [] ? nfs4_state_start+0x2a/0xc5 [nfsd] [] ? nfsd_svc+0x51/0xff [nfsd] [] ? write_svc+0x0/0x1e [nfsd] [] ? write_svc+0x1b/0x1e [nfsd] [] ? nfsctl_transaction_write+0x3a/0x61 [nfsd] [] ? sys_nfsservctl+0x116/0x154 [] ? putname+0x24/0x2f [] ? putname+0x24/0x2f [] ? do_sys_open+0xad/0xb7 [] ? filp_close+0x50/0x5a [] ? sys_open+0x1e/0x26 [] ? syscall_call+0x7/0xb [] ? init_cyrix+0x185/0x490 ======================= Code: 75 e1 8b 53 08 8d 4b 04 8d 46 04 e8 75 00 00 00 8b 53 10 8d 4b 0c 8d 46 0c e8 67 00 00 00 5b 5e 5f 5d c3 90 90 55 89 e5 53 89 c3 <8b> 40 04 8b 00 39 d8 74 16 50 53 68 3e d6 6f c0 6a 30 68 78 d6 EIP: [] list_del+0x6/0x60 SS:ESP 0068:cf9caeb8 ---[ end trace a89c4ad091c4ad53 ]--- Cc: Matthew N. Dodd Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4recover.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index bb93946ace22..b79ec930d9f1 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -225,12 +225,12 @@ nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f) return 0; nfs4_save_user(&uid, &gid); + INIT_LIST_HEAD(dentries); filp = dentry_open(dget(dir), mntget(rec_dir.mnt), O_RDONLY); status = PTR_ERR(filp); if (IS_ERR(filp)) goto out; - INIT_LIST_HEAD(dentries); status = vfs_readdir(filp, nfsd4_build_dentrylist, &dla); fput(filp); while (!list_empty(dentries)) { -- cgit v1.2.3 From a8d82d9b950213b66b22c9e7c63a058841de2394 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 24 Nov 2008 12:51:55 -0500 Subject: NLM: client-side nlm_lookup_host() should avoid matching on srcaddr Since commit c98451bd, the loop in nlm_lookup_host() unconditionally compares the host's h_srcaddr field to the incoming source address. For client-side nlm_host entries, both are always AF_UNSPEC, so this check is unnecessary. Since commit 781b61a6, which added support for AF_INET6 addresses to nlm_cmp_addr(), nlm_cmp_addr() now returns FALSE for AF_UNSPEC addresses, which causes nlm_lookup_host() to create a fresh nlm_host entry every time it is called on the client. These extra entries will eventually expire once the server is unmounted, so the impact of this regression, introduced with lockd IPv6 support in 2.6.28, should be minor. We could fix this by adding an arm in nlm_cmp_addr() for AF_UNSPEC addresses, but really, nlm_lookup_host() shouldn't be matching on the srcaddr field for client-side nlm_host lookups. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 9fd8889097b7..70fc63a1727b 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -167,7 +167,8 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) continue; if (host->h_server != ni->server) continue; - if (!nlm_cmp_addr(nlm_srcaddr(host), ni->src_sap)) + if (ni->server && + !nlm_cmp_addr(nlm_srcaddr(host), ni->src_sap)) continue; /* Move to head of hash chain. */ -- cgit v1.2.3 From a98ee8c1c707fe3210b00ef9f806ba8e2bf35504 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Wed, 26 Nov 2008 19:32:33 +0000 Subject: [CIFS] fix regression in cifs_write_begin/cifs_write_end The conversion to write_begin/write_end interfaces had a bug where we were passing a bad parameter to cifs_readpage_worker. Rather than passing the page offset of the start of the write, we needed to pass the offset of the beginning of the page. This was reliably showing up as data corruption in the fsx-linux test from LTP. It also became evident that this code was occasionally doing unnecessary read calls. Optimize those away by using the PG_checked flag to indicate that the unwritten part of the page has been initialized. CC: Nick Piggin Acked-by: Dave Kleikamp Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/file.c | 77 ++++++++++++++++++++++++++++++++++++++++++---------------- 1 file changed, 56 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/cifs/file.c b/fs/cifs/file.c index b691b893a848..f0a81e631ae6 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -1475,7 +1475,11 @@ static int cifs_write_end(struct file *file, struct address_space *mapping, cFYI(1, ("write_end for page %p from pos %lld with %d bytes", page, pos, copied)); - if (!PageUptodate(page) && copied == PAGE_CACHE_SIZE) + if (PageChecked(page)) { + if (copied == len) + SetPageUptodate(page); + ClearPageChecked(page); + } else if (!PageUptodate(page) && copied == PAGE_CACHE_SIZE) SetPageUptodate(page); if (!PageUptodate(page)) { @@ -2062,39 +2066,70 @@ static int cifs_write_begin(struct file *file, struct address_space *mapping, { pgoff_t index = pos >> PAGE_CACHE_SHIFT; loff_t offset = pos & (PAGE_CACHE_SIZE - 1); + loff_t page_start = pos & PAGE_MASK; + loff_t i_size; + struct page *page; + int rc = 0; cFYI(1, ("write_begin from %lld len %d", (long long)pos, len)); - *pagep = __grab_cache_page(mapping, index); - if (!*pagep) - return -ENOMEM; - - if (PageUptodate(*pagep)) - return 0; + page = __grab_cache_page(mapping, index); + if (!page) { + rc = -ENOMEM; + goto out; + } - /* If we are writing a full page it will be up to date, - no need to read from the server */ - if (len == PAGE_CACHE_SIZE && flags & AOP_FLAG_UNINTERRUPTIBLE) - return 0; + if (PageUptodate(page)) + goto out; - if ((file->f_flags & O_ACCMODE) != O_WRONLY) { - int rc; + /* + * If we write a full page it will be up to date, no need to read from + * the server. If the write is short, we'll end up doing a sync write + * instead. + */ + if (len == PAGE_CACHE_SIZE) + goto out; - /* might as well read a page, it is fast enough */ - rc = cifs_readpage_worker(file, *pagep, &offset); + /* + * optimize away the read when we have an oplock, and we're not + * expecting to use any of the data we'd be reading in. That + * is, when the page lies beyond the EOF, or straddles the EOF + * and the write will cover all of the existing data. + */ + if (CIFS_I(mapping->host)->clientCanCacheRead) { + i_size = i_size_read(mapping->host); + if (page_start >= i_size || + (offset == 0 && (pos + len) >= i_size)) { + zero_user_segments(page, 0, offset, + offset + len, + PAGE_CACHE_SIZE); + /* + * PageChecked means that the parts of the page + * to which we're not writing are considered up + * to date. Once the data is copied to the + * page, it can be set uptodate. + */ + SetPageChecked(page); + goto out; + } + } - /* we do not need to pass errors back - e.g. if we do not have read access to the file - because cifs_write_end will attempt synchronous writes - -- shaggy */ + if ((file->f_flags & O_ACCMODE) != O_WRONLY) { + /* + * might as well read a page, it is fast enough. If we get + * an error, we don't need to return it. cifs_write_end will + * do a sync write instead since PG_uptodate isn't set. + */ + cifs_readpage_worker(file, page, &page_start); } else { /* we could try using another file handle if there is one - but how would we lock it to prevent close of that handle racing with this read? In any case this will be written out by write_end so is fine */ } - - return 0; +out: + *pagep = page; + return rc; } const struct address_space_operations cifs_addr_ops = { -- cgit v1.2.3 From 52b19ac993f1aeadbce15b55302be9a35346e235 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 23 Sep 2008 18:24:08 +0200 Subject: udf: Fix BUG_ON() in destroy_inode() udf_clear_inode() can leave behind buffers on mapping's i_private list (when we truncated preallocation). Call invalidate_inode_buffers() so that the list is properly cleaned-up before we return from udf_clear_inode(). This is ugly and suggest that we should cleanup preallocation earlier than in clear_inode() but currently there's no such call available since drop_inode() is called under inode lock and thus is unusable for disk operations. Signed-off-by: Jan Kara --- fs/buffer.c | 1 + fs/udf/inode.c | 1 + 2 files changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index 6569fda5cfed..10179cfa1152 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -878,6 +878,7 @@ void invalidate_inode_buffers(struct inode *inode) spin_unlock(&buffer_mapping->private_lock); } } +EXPORT_SYMBOL(invalidate_inode_buffers); /* * Remove any clean buffers from the inode's buffer list. This is called diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 6e74b117aaf0..30ebde490f7f 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -106,6 +106,7 @@ void udf_clear_inode(struct inode *inode) udf_truncate_tail_extent(inode); unlock_kernel(); write_inode_now(inode, 0); + invalidate_inode_buffers(inode); } iinfo = UDF_I(inode); kfree(iinfo->i_ext.i_data); -- cgit v1.2.3 From 3b5da0189c93160e44b878d2c72e9552d642497c Mon Sep 17 00:00:00 2001 From: Coly Li Date: Wed, 5 Nov 2008 15:16:24 +0800 Subject: ocfs2: comments typo fix This patch fixes two typos in comments of ocfs2. Signed-off-by: Coly Li Signed-off-by: Mark Fasheh --- fs/ocfs2/dlm/userdlm.h | 2 +- fs/ocfs2/ocfs2.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlm/userdlm.h b/fs/ocfs2/dlm/userdlm.h index 39ec27738499..0c3cc03c61fa 100644 --- a/fs/ocfs2/dlm/userdlm.h +++ b/fs/ocfs2/dlm/userdlm.h @@ -33,7 +33,7 @@ #include /* user_lock_res->l_flags flags. */ -#define USER_LOCK_ATTACHED (0x00000001) /* have we initialized +#define USER_LOCK_ATTACHED (0x00000001) /* we have initialized * the lvb */ #define USER_LOCK_BUSY (0x00000002) /* we are currently in * dlm_lock */ diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index fef7ece32376..3fed9e3d8992 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -85,7 +85,7 @@ enum ocfs2_unlock_action { }; /* ocfs2_lock_res->l_flags flags. */ -#define OCFS2_LOCK_ATTACHED (0x00000001) /* have we initialized +#define OCFS2_LOCK_ATTACHED (0x00000001) /* we have initialized * the lvb */ #define OCFS2_LOCK_BUSY (0x00000002) /* we are currently in * dlm_lock */ -- cgit v1.2.3 From 66f502a416f18cd36179290746aa53736c6b2828 Mon Sep 17 00:00:00 2001 From: David Teigland Date: Mon, 10 Nov 2008 16:24:57 -0600 Subject: ocfs2: initialize stack_user lvbptr The locking_state dump, ocfs2_dlm_seq_show, reads the lvb on locks where it has not yet been initialized by a lock call. Signed-off-by: David Teigland Acked-by: Joel Becker Signed-off-by: Mark Fasheh --- fs/ocfs2/stack_user.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/ocfs2/stack_user.c b/fs/ocfs2/stack_user.c index faec2d879357..9b76d41a8ac6 100644 --- a/fs/ocfs2/stack_user.c +++ b/fs/ocfs2/stack_user.c @@ -740,6 +740,9 @@ static int user_dlm_lock_status(union ocfs2_dlm_lksb *lksb) static void *user_dlm_lvb(union ocfs2_dlm_lksb *lksb) { + if (!lksb->lksb_fsdlm.sb_lvbptr) + lksb->lksb_fsdlm.sb_lvbptr = (char *)lksb + + sizeof(struct dlm_lksb); return (void *)(lksb->lksb_fsdlm.sb_lvbptr); } -- cgit v1.2.3 From 07f9eebcdfaeefc8f807fa1bcce1d7c3ae6661b1 Mon Sep 17 00:00:00 2001 From: David Teigland Date: Mon, 17 Nov 2008 12:28:48 -0600 Subject: ocfs2: fix wake_up in unlock_ast In ocfs2_unlock_ast(), call wake_up() on lockres before releasing the spin lock on it. As soon as the spin lock is released, the lockres can be freed. Signed-off-by: David Teigland Signed-off-by: Mark Fasheh --- fs/ocfs2/dlmglue.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index ec684426034b..6e6cc0a2e5f7 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -2841,9 +2841,8 @@ static void ocfs2_unlock_ast(void *opaque, int error) lockres_clear_flags(lockres, OCFS2_LOCK_BUSY); lockres->l_unlock_action = OCFS2_UNLOCK_INVALID; - spin_unlock_irqrestore(&lockres->l_lock, flags); - wake_up(&lockres->l_event); + spin_unlock_irqrestore(&lockres->l_lock, flags); mlog_exit_void(); } -- cgit v1.2.3 From 07d9a3954a68764aefe16855bcd0f86deeb5c825 Mon Sep 17 00:00:00 2001 From: Coly Li Date: Mon, 17 Nov 2008 12:38:22 +0800 Subject: ocfs2: fix return value set in init_dlmfs_fs() In init_dlmfs_fs(), if calling kmem_cache_create() failed, the code will use return value from calling bdi_init(). The correct behavior should be set status as -ENOMEM before going to "bail:". Signed-off-by: Coly Li Acked-by: Sunil Mushran Signed-off-by: Mark Fasheh --- fs/ocfs2/dlm/dlmfs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/dlm/dlmfs.c b/fs/ocfs2/dlm/dlmfs.c index 533a789c3ef8..ba962d71b34d 100644 --- a/fs/ocfs2/dlm/dlmfs.c +++ b/fs/ocfs2/dlm/dlmfs.c @@ -608,8 +608,10 @@ static int __init init_dlmfs_fs(void) 0, (SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT| SLAB_MEM_SPREAD), dlmfs_init_once); - if (!dlmfs_inode_cache) + if (!dlmfs_inode_cache) { + status = -ENOMEM; goto bail; + } cleanup_inode = 1; user_dlm_worker = create_singlethread_workqueue("user_dlm"); -- cgit v1.2.3 From d6b58f89f7257c8099c2260e2bea042a917d6cdf Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Fri, 21 Nov 2008 14:06:55 -0800 Subject: ocfs2: fix regression in ocfs2_read_blocks_sync() We're panicing in ocfs2_read_blocks_sync() if a jbd-managed buffer is seen. At first glance, this seems ok but in reality it can happen. My test case was to just run 'exorcist'. A struct inode is being pushed out of memory but is then re-read at a later time, before the buffer has been checkpointed by jbd. This causes a BUG to be hit in ocfs2_read_blocks_sync(). Reviewed-by: Joel Becker Signed-off-by: Mark Fasheh --- fs/ocfs2/buffer_head_io.c | 15 ++++----------- 1 file changed, 4 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c index 7e947c672469..3a178ec48d7c 100644 --- a/fs/ocfs2/buffer_head_io.c +++ b/fs/ocfs2/buffer_head_io.c @@ -112,7 +112,7 @@ int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block, bh = bhs[i]; if (buffer_jbd(bh)) { - mlog(ML_ERROR, + mlog(ML_BH_IO, "trying to sync read a jbd " "managed bh (blocknr = %llu), skipping\n", (unsigned long long)bh->b_blocknr); @@ -147,15 +147,10 @@ int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block, for (i = nr; i > 0; i--) { bh = bhs[i - 1]; - if (buffer_jbd(bh)) { - mlog(ML_ERROR, - "the journal got the buffer while it was " - "locked for io! (blocknr = %llu)\n", - (unsigned long long)bh->b_blocknr); - BUG(); - } + /* No need to wait on the buffer if it's managed by JBD. */ + if (!buffer_jbd(bh)) + wait_on_buffer(bh); - wait_on_buffer(bh); if (!buffer_uptodate(bh)) { /* Status won't be cleared from here on out, * so we can safely record this and loop back @@ -251,8 +246,6 @@ int ocfs2_read_blocks(struct inode *inode, u64 block, int nr, ignore_cache = 1; } - /* XXX: Can we ever get this and *not* have the cached - * flag set? */ if (buffer_jbd(bh)) { if (ignore_cache) mlog(ML_BH_IO, "trying to sync read a jbd " -- cgit v1.2.3 From 7ef9964e6d1b911b78709f144000aacadd0ebc21 Mon Sep 17 00:00:00 2001 From: Davide Libenzi Date: Mon, 1 Dec 2008 13:13:55 -0800 Subject: epoll: introduce resource usage limits It has been thought that the per-user file descriptors limit would also limit the resources that a normal user can request via the epoll interface. Vegard Nossum reported a very simple program (a modified version attached) that can make a normal user to request a pretty large amount of kernel memory, well within the its maximum number of fds. To solve such problem, default limits are now imposed, and /proc based configuration has been introduced. A new directory has been created, named /proc/sys/fs/epoll/ and inside there, there are two configuration points: max_user_instances = Maximum number of devices - per user max_user_watches = Maximum number of "watched" fds - per user The current default for "max_user_watches" limits the memory used by epoll to store "watches", to 1/32 of the amount of the low RAM. As example, a 256MB 32bit machine, will have "max_user_watches" set to roughly 90000. That should be enough to not break existing heavy epoll users. The default value for "max_user_instances" is set to 128, that should be enough too. This also changes the userspace, because a new error code can now come out from EPOLL_CTL_ADD (-ENOSPC). The EMFILE from epoll_create() was already listed, so that should be ok. [akpm@linux-foundation.org: use get_current_user()] Signed-off-by: Davide Libenzi Cc: Michael Kerrisk Cc: Cc: Cyrill Gorcunov Reported-by: Vegard Nossum Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/proc.txt | 27 ++++++++++++ fs/eventpoll.c | 85 ++++++++++++++++++++++++++++++++++---- include/linux/sched.h | 4 ++ kernel/sysctl.c | 10 +++++ 4 files changed, 118 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt index bcceb99b81dd..bb1b0dd3bfcb 100644 --- a/Documentation/filesystems/proc.txt +++ b/Documentation/filesystems/proc.txt @@ -44,6 +44,7 @@ Table of Contents 2.14 /proc//io - Display the IO accounting fields 2.15 /proc//coredump_filter - Core dump filtering settings 2.16 /proc//mountinfo - Information about mounts + 2.17 /proc/sys/fs/epoll - Configuration options for the epoll interface ------------------------------------------------------------------------------ Preface @@ -2483,4 +2484,30 @@ For more information on mount propagation see: Documentation/filesystems/sharedsubtree.txt +2.17 /proc/sys/fs/epoll - Configuration options for the epoll interface +-------------------------------------------------------- + +This directory contains configuration options for the epoll(7) interface. + +max_user_instances +------------------ + +This is the maximum number of epoll file descriptors that a single user can +have open at a given time. The default value is 128, and should be enough +for normal users. + +max_user_watches +---------------- + +Every epoll file descriptor can store a number of files to be monitored +for event readiness. Each one of these monitored files constitutes a "watch". +This configuration option sets the maximum number of "watches" that are +allowed for each user. +Each "watch" costs roughly 90 bytes on a 32bit kernel, and roughly 160 bytes +on a 64bit one. +The current default value for max_user_watches is the 1/32 of the available +low memory, divided for the "watch" cost in bytes. + + ------------------------------------------------------------------------------ + diff --git a/fs/eventpoll.c b/fs/eventpoll.c index aec5c13f6341..96355d505347 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -102,6 +102,8 @@ #define EP_UNACTIVE_PTR ((void *) -1L) +#define EP_ITEM_COST (sizeof(struct epitem) + sizeof(struct eppoll_entry)) + struct epoll_filefd { struct file *file; int fd; @@ -200,6 +202,9 @@ struct eventpoll { * holding ->lock. */ struct epitem *ovflist; + + /* The user that created the eventpoll descriptor */ + struct user_struct *user; }; /* Wait structure used by the poll hooks */ @@ -226,10 +231,18 @@ struct ep_pqueue { struct epitem *epi; }; +/* + * Configuration options available inside /proc/sys/fs/epoll/ + */ +/* Maximum number of epoll devices, per user */ +static int max_user_instances __read_mostly; +/* Maximum number of epoll watched descriptors, per user */ +static int max_user_watches __read_mostly; + /* * This mutex is used to serialize ep_free() and eventpoll_release_file(). */ -static struct mutex epmutex; +static DEFINE_MUTEX(epmutex); /* Safe wake up implementation */ static struct poll_safewake psw; @@ -240,6 +253,33 @@ static struct kmem_cache *epi_cache __read_mostly; /* Slab cache used to allocate "struct eppoll_entry" */ static struct kmem_cache *pwq_cache __read_mostly; +#ifdef CONFIG_SYSCTL + +#include + +static int zero; + +ctl_table epoll_table[] = { + { + .procname = "max_user_instances", + .data = &max_user_instances, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .extra1 = &zero, + }, + { + .procname = "max_user_watches", + .data = &max_user_watches, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .extra1 = &zero, + }, + { .ctl_name = 0 } +}; +#endif /* CONFIG_SYSCTL */ + /* Setup the structure that is used as key for the RB tree */ static inline void ep_set_ffd(struct epoll_filefd *ffd, @@ -402,6 +442,8 @@ static int ep_remove(struct eventpoll *ep, struct epitem *epi) /* At this point it is safe to free the eventpoll item */ kmem_cache_free(epi_cache, epi); + atomic_dec(&ep->user->epoll_watches); + DNPRINTK(3, (KERN_INFO "[%p] eventpoll: ep_remove(%p, %p)\n", current, ep, file)); @@ -449,6 +491,8 @@ static void ep_free(struct eventpoll *ep) mutex_unlock(&epmutex); mutex_destroy(&ep->mtx); + atomic_dec(&ep->user->epoll_devs); + free_uid(ep->user); kfree(ep); } @@ -532,10 +576,19 @@ void eventpoll_release_file(struct file *file) static int ep_alloc(struct eventpoll **pep) { - struct eventpoll *ep = kzalloc(sizeof(*ep), GFP_KERNEL); + int error; + struct user_struct *user; + struct eventpoll *ep; - if (!ep) - return -ENOMEM; + user = get_current_user(); + error = -EMFILE; + if (unlikely(atomic_read(&user->epoll_devs) >= + max_user_instances)) + goto free_uid; + error = -ENOMEM; + ep = kzalloc(sizeof(*ep), GFP_KERNEL); + if (unlikely(!ep)) + goto free_uid; spin_lock_init(&ep->lock); mutex_init(&ep->mtx); @@ -544,12 +597,17 @@ static int ep_alloc(struct eventpoll **pep) INIT_LIST_HEAD(&ep->rdllist); ep->rbr = RB_ROOT; ep->ovflist = EP_UNACTIVE_PTR; + ep->user = user; *pep = ep; DNPRINTK(3, (KERN_INFO "[%p] eventpoll: ep_alloc() ep=%p\n", current, ep)); return 0; + +free_uid: + free_uid(user); + return error; } /* @@ -703,9 +761,11 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event, struct epitem *epi; struct ep_pqueue epq; - error = -ENOMEM; + if (unlikely(atomic_read(&ep->user->epoll_watches) >= + max_user_watches)) + return -ENOSPC; if (!(epi = kmem_cache_alloc(epi_cache, GFP_KERNEL))) - goto error_return; + return -ENOMEM; /* Item initialization follow here ... */ INIT_LIST_HEAD(&epi->rdllink); @@ -735,6 +795,7 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event, * install process. Namely an allocation for a wait queue failed due * high memory pressure. */ + error = -ENOMEM; if (epi->nwait < 0) goto error_unregister; @@ -765,6 +826,8 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event, spin_unlock_irqrestore(&ep->lock, flags); + atomic_inc(&ep->user->epoll_watches); + /* We have to call this outside the lock */ if (pwake) ep_poll_safewake(&psw, &ep->poll_wait); @@ -789,7 +852,7 @@ error_unregister: spin_unlock_irqrestore(&ep->lock, flags); kmem_cache_free(epi_cache, epi); -error_return: + return error; } @@ -1078,6 +1141,7 @@ asmlinkage long sys_epoll_create1(int flags) flags & O_CLOEXEC); if (fd < 0) ep_free(ep); + atomic_inc(&ep->user->epoll_devs); error_return: DNPRINTK(3, (KERN_INFO "[%p] eventpoll: sys_epoll_create(%d) = %d\n", @@ -1299,7 +1363,12 @@ asmlinkage long sys_epoll_pwait(int epfd, struct epoll_event __user *events, static int __init eventpoll_init(void) { - mutex_init(&epmutex); + struct sysinfo si; + + si_meminfo(&si); + max_user_instances = 128; + max_user_watches = (((si.totalram - si.totalhigh) / 32) << PAGE_SHIFT) / + EP_ITEM_COST; /* Initialize the structure used to perform safe poll wait head wake ups */ ep_poll_safewake_init(&psw); diff --git a/include/linux/sched.h b/include/linux/sched.h index 644ffbda17ca..55e30d114477 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -630,6 +630,10 @@ struct user_struct { atomic_t inotify_watches; /* How many inotify watches does this user have? */ atomic_t inotify_devs; /* How many inotify devs does this user have opened? */ #endif +#ifdef CONFIG_EPOLL + atomic_t epoll_devs; /* The number of epoll descriptors currently open */ + atomic_t epoll_watches; /* The number of file descriptors currently watched */ +#endif #ifdef CONFIG_POSIX_MQUEUE /* protected by mq_lock */ unsigned long mq_bytes; /* How many bytes can be allocated to mqueue? */ diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 9d048fa2d902..3d56fe7570da 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -176,6 +176,9 @@ extern struct ctl_table random_table[]; #ifdef CONFIG_INOTIFY_USER extern struct ctl_table inotify_table[]; #endif +#ifdef CONFIG_EPOLL +extern struct ctl_table epoll_table[]; +#endif #ifdef HAVE_ARCH_PICK_MMAP_LAYOUT int sysctl_legacy_va_layout; @@ -1325,6 +1328,13 @@ static struct ctl_table fs_table[] = { .child = inotify_table, }, #endif +#ifdef CONFIG_EPOLL + { + .procname = "epoll", + .mode = 0555, + .child = epoll_table, + }, +#endif #endif { .ctl_name = KERN_SETUID_DUMPABLE, -- cgit v1.2.3 From 03801553630c4bec6682108800c9b2de64bdbd37 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Mon, 1 Dec 2008 13:14:04 -0800 Subject: ntfs: don't fool kernel-doc kernel-doc handles macros now (it has for quite some time), so change the ntfs_debug() macro's kernel-doc to be just before the macro instead of before a phony function prototype. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Randy Dunlap Cc: Anton Altaparmakov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ntfs/debug.h | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ntfs/debug.h b/fs/ntfs/debug.h index 5e6724c1afd1..2142b1c68b61 100644 --- a/fs/ntfs/debug.h +++ b/fs/ntfs/debug.h @@ -30,7 +30,8 @@ extern int debug_msgs; -#if 0 /* Fool kernel-doc since it doesn't do macros yet */ +extern void __ntfs_debug(const char *file, int line, const char *function, + const char *format, ...) __attribute__ ((format (printf, 4, 5))); /** * ntfs_debug - write a debug level message to syslog * @f: a printf format string containing the message @@ -39,11 +40,6 @@ extern int debug_msgs; * ntfs_debug() writes a DEBUG level message to the syslog but only if the * driver was compiled with -DDEBUG. Otherwise, the call turns into a NOP. */ -static void ntfs_debug(const char *f, ...); -#endif - -extern void __ntfs_debug (const char *file, int line, const char *function, - const char *format, ...) __attribute__ ((format (printf, 4, 5))); #define ntfs_debug(f, a...) \ __ntfs_debug(__FILE__, __LINE__, __func__, f, ##a) -- cgit v1.2.3