whiterose

linux unikernel
Log | Files | Refs | README | LICENSE | git clone https://git.ne02ptzero.me/git/whiterose

commit 7bbbf2c2fc84dd24cae8615b11a0c6cac12cbe94
parent b71acb0e372160167bf6d5500b88b30b52ccef6e
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Thu, 27 Dec 2018 16:42:39 -0800

Merge tag 'gfs2-4.21.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2

Pull gfs2 updates from Bob Peterson:

 - Enhancements and performance improvements to journal replay (Abhi
   Das)

 - Cleanup of gfs2_is_ordered and gfs2_is_writeback (Andreas
   Gruenbacher)

 - Fix a potential double-free in inode creation (Andreas Gruenbacher)

 - Fix the bitmap search loop that was searching too far (Andreas
   Gruenbacher)

 - Various cleanups (Andreas Gruenbacher, Bob Peterson)

 - Implement Steve Whitehouse's patch to dump nrpages for inodes (Bob
   Peterson)

 - Fix a withdraw bug where stuffed journaled data files didn't allocate
   enough journal space to be grown (Bob Peterson)

* tag 'gfs2-4.21.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
  gfs2: take jdata unstuff into account in do_grow
  gfs2: Dump nrpages for inodes and their glocks
  gfs2: Fix loop in gfs2_rbm_find
  gfs2: Get rid of potential double-freeing in gfs2_create_inode
  gfs2: Remove vestigial bd_ops
  gfs2: read journal in large chunks to locate the head
  gfs2: add a helper function to get_log_header that can be used elsewhere
  gfs2: changes to gfs2_log_XXX_bio
  gfs2: add more timing info to journal recovery process
  gfs2: Fix the gfs2_invalidatepage description
  gfs2: Clean up gfs2_is_{ordered,writeback}

Diffstat:
Mfs/gfs2/aops.c | 16++++++++--------
Mfs/gfs2/bmap.c | 10++++++++--
Mfs/gfs2/glock.c | 2+-
Mfs/gfs2/glock.h | 2+-
Mfs/gfs2/glops.c | 17+++++++++++++----
Mfs/gfs2/incore.h | 3+--
Mfs/gfs2/inode.c | 18+++++++++---------
Mfs/gfs2/inode.h | 10++++------
Mfs/gfs2/log.c | 5++---
Mfs/gfs2/log.h | 5++---
Mfs/gfs2/lops.c | 257+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++------------
Mfs/gfs2/lops.h | 4+++-
Mfs/gfs2/ops_fstype.c | 1+
Mfs/gfs2/recovery.c | 178+++++++++++++++----------------------------------------------------------------
Mfs/gfs2/recovery.h | 5+++--
Mfs/gfs2/rgrp.c | 4++--
Mfs/gfs2/rgrp.h | 2+-
Mfs/gfs2/super.c | 1+
Mfs/gfs2/trans.c | 8+++-----
19 files changed, 317 insertions(+), 231 deletions(-)

diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c @@ -820,10 +820,10 @@ out: * @page: the page that's being released * @gfp_mask: passed from Linux VFS, ignored by us * - * Call try_to_free_buffers() if the buffers in this page can be - * released. + * Calls try_to_free_buffers() to free the buffers and put the page if the + * buffers can be released. * - * Returns: 0 + * Returns: 1 if the page was put or else 0 */ int gfs2_releasepage(struct page *page, gfp_t gfp_mask) @@ -930,14 +930,14 @@ static const struct address_space_operations gfs2_jdata_aops = { void gfs2_set_aops(struct inode *inode) { struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_sbd *sdp = GFS2_SB(inode); - if (gfs2_is_writeback(ip)) + if (gfs2_is_jdata(ip)) + inode->i_mapping->a_ops = &gfs2_jdata_aops; + else if (gfs2_is_writeback(sdp)) inode->i_mapping->a_ops = &gfs2_writeback_aops; - else if (gfs2_is_ordered(ip)) + else if (gfs2_is_ordered(sdp)) inode->i_mapping->a_ops = &gfs2_ordered_aops; - else if (gfs2_is_jdata(ip)) - inode->i_mapping->a_ops = &gfs2_jdata_aops; else BUG(); } - diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c @@ -14,6 +14,7 @@ #include <linux/gfs2_ondisk.h> #include <linux/crc32.h> #include <linux/iomap.h> +#include <linux/ktime.h> #include "gfs2.h" #include "incore.h" @@ -2083,6 +2084,8 @@ static int do_grow(struct inode *inode, u64 size) } error = gfs2_trans_begin(sdp, RES_DINODE + RES_STATFS + RES_RG_BIT + + (unstuff && + gfs2_is_jdata(ip) ? RES_JDATA : 0) + (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF ? 0 : RES_QUOTA), 0); if (error) @@ -2248,7 +2251,9 @@ int gfs2_map_journal_extents(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd) unsigned int shift = sdp->sd_sb.sb_bsize_shift; u64 size; int rc; + ktime_t start, end; + start = ktime_get(); lblock_stop = i_size_read(jd->jd_inode) >> shift; size = (lblock_stop - lblock) << shift; jd->nr_extents = 0; @@ -2268,8 +2273,9 @@ int gfs2_map_journal_extents(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd) lblock += (bh.b_size >> ip->i_inode.i_blkbits); } while(size > 0); - fs_info(sdp, "journal %d mapped with %u extents\n", jd->jd_jid, - jd->nr_extents); + end = ktime_get(); + fs_info(sdp, "journal %d mapped with %u extents in %lldms\n", jd->jd_jid, + jd->nr_extents, ktime_ms_delta(end, start)); return 0; fail: diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c @@ -1777,7 +1777,7 @@ static const char *gflags2str(char *buf, const struct gfs2_glock *gl) * */ -void gfs2_dump_glock(struct seq_file *seq, const struct gfs2_glock *gl) +void gfs2_dump_glock(struct seq_file *seq, struct gfs2_glock *gl) { const struct gfs2_glock_operations *glops = gl->gl_ops; unsigned long long dtime; diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h @@ -202,7 +202,7 @@ extern int gfs2_glock_nq_num(struct gfs2_sbd *sdp, u64 number, struct gfs2_holder *gh); extern int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs); extern void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs); -extern void gfs2_dump_glock(struct seq_file *seq, const struct gfs2_glock *gl); +extern void gfs2_dump_glock(struct seq_file *seq, struct gfs2_glock *gl); #define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { gfs2_dump_glock(NULL, gl); BUG(); } } while(0) extern __printf(2, 3) void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...); diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c @@ -28,6 +28,7 @@ #include "util.h" #include "trans.h" #include "dir.h" +#include "lops.h" struct workqueue_struct *gfs2_freeze_wq; @@ -466,17 +467,25 @@ static int inode_go_lock(struct gfs2_holder *gh) * */ -static void inode_go_dump(struct seq_file *seq, const struct gfs2_glock *gl) +static void inode_go_dump(struct seq_file *seq, struct gfs2_glock *gl) { - const struct gfs2_inode *ip = gl->gl_object; + struct gfs2_inode *ip = gl->gl_object; + struct inode *inode = &ip->i_inode; + unsigned long nrpages; + if (ip == NULL) return; - gfs2_print_dbg(seq, " I: n:%llu/%llu t:%u f:0x%02lx d:0x%08x s:%llu\n", + + xa_lock_irq(&inode->i_data.i_pages); + nrpages = inode->i_data.nrpages; + xa_unlock_irq(&inode->i_data.i_pages); + + gfs2_print_dbg(seq, " I: n:%llu/%llu t:%u f:0x%02lx d:0x%08x s:%llu p:%lu\n", (unsigned long long)ip->i_no_formal_ino, (unsigned long long)ip->i_no_addr, IF2DT(ip->i_inode.i_mode), ip->i_flags, (unsigned int)ip->i_diskflags, - (unsigned long long)i_size_read(&ip->i_inode)); + (unsigned long long)i_size_read(inode), nrpages); } /** diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h @@ -165,7 +165,6 @@ struct gfs2_bufdata { u64 bd_blkno; struct list_head bd_list; - const struct gfs2_log_operations *bd_ops; struct gfs2_trans *bd_tr; struct list_head bd_ail_st_list; @@ -244,7 +243,7 @@ struct gfs2_glock_operations { int (*go_demote_ok) (const struct gfs2_glock *gl); int (*go_lock) (struct gfs2_holder *gh); void (*go_unlock) (struct gfs2_holder *gh); - void (*go_dump)(struct seq_file *seq, const struct gfs2_glock *gl); + void (*go_dump)(struct seq_file *seq, struct gfs2_glock *gl); void (*go_callback)(struct gfs2_glock *gl, bool remote); const int go_type; const unsigned long go_flags; diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c @@ -744,17 +744,19 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, the gfs2 structures. */ if (default_acl) { error = __gfs2_set_acl(inode, default_acl, ACL_TYPE_DEFAULT); + if (error) + goto fail_gunlock3; posix_acl_release(default_acl); + default_acl = NULL; } if (acl) { - if (!error) - error = __gfs2_set_acl(inode, acl, ACL_TYPE_ACCESS); + error = __gfs2_set_acl(inode, acl, ACL_TYPE_ACCESS); + if (error) + goto fail_gunlock3; posix_acl_release(acl); + acl = NULL; } - if (error) - goto fail_gunlock3; - error = security_inode_init_security(&ip->i_inode, &dip->i_inode, name, &gfs2_initxattrs, NULL); if (error) @@ -789,10 +791,8 @@ fail_free_inode: } gfs2_rsqa_delete(ip, NULL); fail_free_acls: - if (default_acl) - posix_acl_release(default_acl); - if (acl) - posix_acl_release(acl); + posix_acl_release(default_acl); + posix_acl_release(acl); fail_gunlock: gfs2_dir_no_add(&da); gfs2_glock_dq_uninit(ghs); diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h @@ -30,16 +30,14 @@ static inline int gfs2_is_jdata(const struct gfs2_inode *ip) return ip->i_diskflags & GFS2_DIF_JDATA; } -static inline int gfs2_is_writeback(const struct gfs2_inode *ip) +static inline bool gfs2_is_ordered(const struct gfs2_sbd *sdp) { - const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - return (sdp->sd_args.ar_data == GFS2_DATA_WRITEBACK) && !gfs2_is_jdata(ip); + return sdp->sd_args.ar_data == GFS2_DATA_ORDERED; } -static inline int gfs2_is_ordered(const struct gfs2_inode *ip) +static inline bool gfs2_is_writeback(const struct gfs2_sbd *sdp) { - const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - return (sdp->sd_args.ar_data == GFS2_DATA_ORDERED) && !gfs2_is_jdata(ip); + return sdp->sd_args.ar_data == GFS2_DATA_WRITEBACK; } static inline int gfs2_is_dir(const struct gfs2_inode *ip) diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c @@ -605,7 +605,6 @@ void gfs2_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) bd->bd_blkno = bh->b_blocknr; gfs2_remove_from_ail(bd); /* drops ref on bh */ bd->bd_bh = NULL; - bd->bd_ops = &gfs2_revoke_lops; sdp->sd_log_num_revoke++; atomic_inc(&gl->gl_revokes); set_bit(GLF_LFLUSH, &gl->gl_flags); @@ -734,7 +733,7 @@ void gfs2_write_log_header(struct gfs2_sbd *sdp, struct gfs2_jdesc *jd, lh->lh_crc = cpu_to_be32(crc); gfs2_log_write(sdp, page, sb->s_blocksize, 0, addr); - gfs2_log_flush_bio(sdp, REQ_OP_WRITE, op_flags); + gfs2_log_submit_bio(&sdp->sd_log_bio, REQ_OP_WRITE | op_flags); log_flush_wait(sdp); } @@ -811,7 +810,7 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags) gfs2_ordered_write(sdp); lops_before_commit(sdp, tr); - gfs2_log_flush_bio(sdp, REQ_OP_WRITE, 0); + gfs2_log_submit_bio(&sdp->sd_log_bio, REQ_OP_WRITE); if (sdp->sd_log_head != sdp->sd_log_flush_head) { log_flush_wait(sdp); diff --git a/fs/gfs2/log.h b/fs/gfs2/log.h @@ -51,12 +51,11 @@ static inline void gfs2_log_pointers_init(struct gfs2_sbd *sdp, static inline void gfs2_ordered_add_inode(struct gfs2_inode *ip) { - struct gfs2_sbd *sdp; + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - if (!gfs2_is_ordered(ip)) + if (gfs2_is_jdata(ip) || !gfs2_is_ordered(sdp)) return; - sdp = GFS2_SB(&ip->i_inode); if (!test_bit(GIF_ORDERED, &ip->i_flags)) { spin_lock(&sdp->sd_ordered_lock); if (!test_and_set_bit(GIF_ORDERED, &ip->i_flags)) diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c @@ -17,7 +17,9 @@ #include <linux/bio.h> #include <linux/fs.h> #include <linux/list_sort.h> +#include <linux/blkdev.h> +#include "bmap.h" #include "dir.h" #include "gfs2.h" #include "incore.h" @@ -193,7 +195,6 @@ static void gfs2_end_log_write_bh(struct gfs2_sbd *sdp, struct bio_vec *bvec, /** * gfs2_end_log_write - end of i/o to the log * @bio: The bio - * @error: Status of i/o request * * Each bio_vec contains either data from the pagecache or data * relating to the log itself. Here we iterate over the bio_vec @@ -228,83 +229,86 @@ static void gfs2_end_log_write(struct bio *bio) } /** - * gfs2_log_flush_bio - Submit any pending log bio - * @sdp: The superblock - * @op: REQ_OP - * @op_flags: req_flag_bits + * gfs2_log_submit_bio - Submit any pending log bio + * @biop: Address of the bio pointer + * @opf: REQ_OP | op_flags * * Submit any pending part-built or full bio to the block device. If * there is no pending bio, then this is a no-op. */ -void gfs2_log_flush_bio(struct gfs2_sbd *sdp, int op, int op_flags) +void gfs2_log_submit_bio(struct bio **biop, int opf) { - if (sdp->sd_log_bio) { + struct bio *bio = *biop; + if (bio) { + struct gfs2_sbd *sdp = bio->bi_private; atomic_inc(&sdp->sd_log_in_flight); - bio_set_op_attrs(sdp->sd_log_bio, op, op_flags); - submit_bio(sdp->sd_log_bio); - sdp->sd_log_bio = NULL; + bio->bi_opf = opf; + submit_bio(bio); + *biop = NULL; } } /** - * gfs2_log_alloc_bio - Allocate a new bio for log writing - * @sdp: The superblock - * @blkno: The next device block number we want to write to + * gfs2_log_alloc_bio - Allocate a bio + * @sdp: The super block + * @blkno: The device block number we want to write to + * @end_io: The bi_end_io callback * - * This should never be called when there is a cached bio in the - * super block. When it returns, there will be a cached bio in the - * super block which will have as many bio_vecs as the device is - * happy to handle. + * Allocate a new bio, initialize it with the given parameters and return it. * - * Returns: Newly allocated bio + * Returns: The newly allocated bio */ -static struct bio *gfs2_log_alloc_bio(struct gfs2_sbd *sdp, u64 blkno) +static struct bio *gfs2_log_alloc_bio(struct gfs2_sbd *sdp, u64 blkno, + bio_end_io_t *end_io) { struct super_block *sb = sdp->sd_vfs; - struct bio *bio; - - BUG_ON(sdp->sd_log_bio); + struct bio *bio = bio_alloc(GFP_NOIO, BIO_MAX_PAGES); - bio = bio_alloc(GFP_NOIO, BIO_MAX_PAGES); bio->bi_iter.bi_sector = blkno * (sb->s_blocksize >> 9); bio_set_dev(bio, sb->s_bdev); - bio->bi_end_io = gfs2_end_log_write; + bio->bi_end_io = end_io; bio->bi_private = sdp; - sdp->sd_log_bio = bio; - return bio; } /** * gfs2_log_get_bio - Get cached log bio, or allocate a new one - * @sdp: The superblock + * @sdp: The super block * @blkno: The device block number we want to write to + * @bio: The bio to get or allocate + * @op: REQ_OP + * @end_io: The bi_end_io callback + * @flush: Always flush the current bio and allocate a new one? * * If there is a cached bio, then if the next block number is sequential * with the previous one, return it, otherwise flush the bio to the - * device. If there is not a cached bio, or we just flushed it, then + * device. If there is no cached bio, or we just flushed it, then * allocate a new one. * * Returns: The bio to use for log writes */ -static struct bio *gfs2_log_get_bio(struct gfs2_sbd *sdp, u64 blkno) +static struct bio *gfs2_log_get_bio(struct gfs2_sbd *sdp, u64 blkno, + struct bio **biop, int op, + bio_end_io_t *end_io, bool flush) { - struct bio *bio = sdp->sd_log_bio; - u64 nblk; + struct bio *bio = *biop; if (bio) { + u64 nblk; + nblk = bio_end_sector(bio); nblk >>= sdp->sd_fsb2bb_shift; - if (blkno == nblk) + if (blkno == nblk && !flush) return bio; - gfs2_log_flush_bio(sdp, REQ_OP_WRITE, 0); + gfs2_log_submit_bio(biop, op); } - return gfs2_log_alloc_bio(sdp, blkno); + *biop = gfs2_log_alloc_bio(sdp, blkno, end_io); + return *biop; } /** @@ -326,11 +330,12 @@ void gfs2_log_write(struct gfs2_sbd *sdp, struct page *page, struct bio *bio; int ret; - bio = gfs2_log_get_bio(sdp, blkno); + bio = gfs2_log_get_bio(sdp, blkno, &sdp->sd_log_bio, REQ_OP_WRITE, + gfs2_end_log_write, false); ret = bio_add_page(bio, page, size, offset); if (ret == 0) { - gfs2_log_flush_bio(sdp, REQ_OP_WRITE, 0); - bio = gfs2_log_alloc_bio(sdp, blkno); + bio = gfs2_log_get_bio(sdp, blkno, &sdp->sd_log_bio, + REQ_OP_WRITE, gfs2_end_log_write, true); ret = bio_add_page(bio, page, size, offset); WARN_ON(ret == 0); } @@ -370,6 +375,184 @@ void gfs2_log_write_page(struct gfs2_sbd *sdp, struct page *page) gfs2_log_bmap(sdp)); } +/** + * gfs2_end_log_read - end I/O callback for reads from the log + * @bio: The bio + * + * Simply unlock the pages in the bio. The main thread will wait on them and + * process them in order as necessary. + */ + +static void gfs2_end_log_read(struct bio *bio) +{ + struct page *page; + struct bio_vec *bvec; + int i; + + bio_for_each_segment_all(bvec, bio, i) { + page = bvec->bv_page; + if (bio->bi_status) { + int err = blk_status_to_errno(bio->bi_status); + + SetPageError(page); + mapping_set_error(page->mapping, err); + } + unlock_page(page); + } + + bio_put(bio); +} + +/** + * gfs2_jhead_pg_srch - Look for the journal head in a given page. + * @jd: The journal descriptor + * @page: The page to look in + * + * Returns: 1 if found, 0 otherwise. + */ + +static bool gfs2_jhead_pg_srch(struct gfs2_jdesc *jd, + struct gfs2_log_header_host *head, + struct page *page) +{ + struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); + struct gfs2_log_header_host uninitialized_var(lh); + void *kaddr = kmap_atomic(page); + unsigned int offset; + bool ret = false; + + for (offset = 0; offset < PAGE_SIZE; offset += sdp->sd_sb.sb_bsize) { + if (!__get_log_header(sdp, kaddr + offset, 0, &lh)) { + if (lh.lh_sequence > head->lh_sequence) + *head = lh; + else { + ret = true; + break; + } + } + } + kunmap_atomic(kaddr); + return ret; +} + +/** + * gfs2_jhead_process_page - Search/cleanup a page + * @jd: The journal descriptor + * @index: Index of the page to look into + * @done: If set, perform only cleanup, else search and set if found. + * + * Find the page with 'index' in the journal's mapping. Search the page for + * the journal head if requested (cleanup == false). Release refs on the + * page so the page cache can reclaim it (put_page() twice). We grabbed a + * reference on this page two times, first when we did a find_or_create_page() + * to obtain the page to add it to the bio and second when we do a + * find_get_page() here to get the page to wait on while I/O on it is being + * completed. + * This function is also used to free up a page we might've grabbed but not + * used. Maybe we added it to a bio, but not submitted it for I/O. Or we + * submitted the I/O, but we already found the jhead so we only need to drop + * our references to the page. + */ + +static void gfs2_jhead_process_page(struct gfs2_jdesc *jd, unsigned long index, + struct gfs2_log_header_host *head, + bool *done) +{ + struct page *page; + + page = find_get_page(jd->jd_inode->i_mapping, index); + wait_on_page_locked(page); + + if (PageError(page)) + *done = true; + + if (!*done) + *done = gfs2_jhead_pg_srch(jd, head, page); + + put_page(page); /* Once for find_get_page */ + put_page(page); /* Once more for find_or_create_page */ +} + +/** + * gfs2_find_jhead - find the head of a log + * @jd: The journal descriptor + * @head: The log descriptor for the head of the log is returned here + * + * Do a search of a journal by reading it in large chunks using bios and find + * the valid log entry with the highest sequence number. (i.e. the log head) + * + * Returns: 0 on success, errno otherwise + */ + +int gfs2_find_jhead(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head) +{ + struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); + struct address_space *mapping = jd->jd_inode->i_mapping; + struct gfs2_journal_extent *je; + u32 block, read_idx = 0, submit_idx = 0, index = 0; + int shift = PAGE_SHIFT - sdp->sd_sb.sb_bsize_shift; + int blocks_per_page = 1 << shift, sz, ret = 0; + struct bio *bio = NULL; + struct page *page; + bool done = false; + errseq_t since; + + memset(head, 0, sizeof(*head)); + if (list_empty(&jd->extent_list)) + gfs2_map_journal_extents(sdp, jd); + + since = filemap_sample_wb_err(mapping); + list_for_each_entry(je, &jd->extent_list, list) { + for (block = 0; block < je->blocks; block += blocks_per_page) { + index = (je->lblock + block) >> shift; + + page = find_or_create_page(mapping, index, GFP_NOFS); + if (!page) { + ret = -ENOMEM; + done = true; + goto out; + } + + if (bio) { + sz = bio_add_page(bio, page, PAGE_SIZE, 0); + if (sz == PAGE_SIZE) + goto page_added; + submit_idx = index; + submit_bio(bio); + bio = NULL; + } + + bio = gfs2_log_alloc_bio(sdp, + je->dblock + (index << shift), + gfs2_end_log_read); + bio->bi_opf = REQ_OP_READ; + sz = bio_add_page(bio, page, PAGE_SIZE, 0); + gfs2_assert_warn(sdp, sz == PAGE_SIZE); + +page_added: + if (submit_idx <= read_idx + BIO_MAX_PAGES) { + /* Keep at least one bio in flight */ + continue; + } + + gfs2_jhead_process_page(jd, read_idx++, head, &done); + if (done) + goto out; /* found */ + } + } + +out: + if (bio) + submit_bio(bio); + while (read_idx <= index) + gfs2_jhead_process_page(jd, read_idx++, head, &done); + + if (!ret) + ret = filemap_check_wb_err(mapping, since); + + return ret; +} + static struct page *gfs2_get_log_desc(struct gfs2_sbd *sdp, u32 ld_type, u32 ld_length, u32 ld_data1) { diff --git a/fs/gfs2/lops.h b/fs/gfs2/lops.h @@ -30,8 +30,10 @@ extern u64 gfs2_log_bmap(struct gfs2_sbd *sdp); extern void gfs2_log_write(struct gfs2_sbd *sdp, struct page *page, unsigned size, unsigned offset, u64 blkno); extern void gfs2_log_write_page(struct gfs2_sbd *sdp, struct page *page); -extern void gfs2_log_flush_bio(struct gfs2_sbd *sdp, int op, int op_flags); +extern void gfs2_log_submit_bio(struct bio **biop, int opf); extern void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh); +extern int gfs2_find_jhead(struct gfs2_jdesc *jd, + struct gfs2_log_header_host *head); static inline unsigned int buf_limit(struct gfs2_sbd *sdp) { diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c @@ -41,6 +41,7 @@ #include "dir.h" #include "meta_io.h" #include "trace_gfs2.h" +#include "lops.h" #define DO 0 #define UNDO 1 diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c @@ -120,6 +120,35 @@ void gfs2_revoke_clean(struct gfs2_jdesc *jd) } } +int __get_log_header(struct gfs2_sbd *sdp, const struct gfs2_log_header *lh, + unsigned int blkno, struct gfs2_log_header_host *head) +{ + u32 hash, crc; + + if (lh->lh_header.mh_magic != cpu_to_be32(GFS2_MAGIC) || + lh->lh_header.mh_type != cpu_to_be32(GFS2_METATYPE_LH) || + (blkno && be32_to_cpu(lh->lh_blkno) != blkno)) + return 1; + + hash = crc32(~0, lh, LH_V1_SIZE - 4); + hash = ~crc32_le_shift(hash, 4); /* assume lh_hash is zero */ + + if (be32_to_cpu(lh->lh_hash) != hash) + return 1; + + crc = crc32c(~0, (void *)lh + LH_V1_SIZE + 4, + sdp->sd_sb.sb_bsize - LH_V1_SIZE - 4); + + if ((lh->lh_crc != 0 && be32_to_cpu(lh->lh_crc) != crc)) + return 1; + + head->lh_sequence = be64_to_cpu(lh->lh_sequence); + head->lh_flags = be32_to_cpu(lh->lh_flags); + head->lh_tail = be32_to_cpu(lh->lh_tail); + head->lh_blkno = be32_to_cpu(lh->lh_blkno); + + return 0; +} /** * get_log_header - read the log header for a given segment * @jd: the journal @@ -137,159 +166,18 @@ void gfs2_revoke_clean(struct gfs2_jdesc *jd) static int get_log_header(struct gfs2_jdesc *jd, unsigned int blk, struct gfs2_log_header_host *head) { - struct gfs2_log_header *lh; + struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); struct buffer_head *bh; - u32 hash, crc; int error; error = gfs2_replay_read_block(jd, blk, &bh); if (error) return error; - lh = (void *)bh->b_data; - - hash = crc32(~0, lh, LH_V1_SIZE - 4); - hash = ~crc32_le_shift(hash, 4); /* assume lh_hash is zero */ - - crc = crc32c(~0, (void *)lh + LH_V1_SIZE + 4, - bh->b_size - LH_V1_SIZE - 4); - - error = lh->lh_header.mh_magic != cpu_to_be32(GFS2_MAGIC) || - lh->lh_header.mh_type != cpu_to_be32(GFS2_METATYPE_LH) || - be32_to_cpu(lh->lh_blkno) != blk || - be32_to_cpu(lh->lh_hash) != hash || - (lh->lh_crc != 0 && be32_to_cpu(lh->lh_crc) != crc); + error = __get_log_header(sdp, (const struct gfs2_log_header *)bh->b_data, + blk, head); brelse(bh); - if (!error) { - head->lh_sequence = be64_to_cpu(lh->lh_sequence); - head->lh_flags = be32_to_cpu(lh->lh_flags); - head->lh_tail = be32_to_cpu(lh->lh_tail); - head->lh_blkno = be32_to_cpu(lh->lh_blkno); - } - return error; -} - -/** - * find_good_lh - find a good log header - * @jd: the journal - * @blk: the segment to start searching from - * @lh: the log header to fill in - * @forward: if true search forward in the log, else search backward - * - * Call get_log_header() to get a log header for a segment, but if the - * segment is bad, either scan forward or backward until we find a good one. - * - * Returns: errno - */ - -static int find_good_lh(struct gfs2_jdesc *jd, unsigned int *blk, - struct gfs2_log_header_host *head) -{ - unsigned int orig_blk = *blk; - int error; - - for (;;) { - error = get_log_header(jd, *blk, head); - if (error <= 0) - return error; - - if (++*blk == jd->jd_blocks) - *blk = 0; - - if (*blk == orig_blk) { - gfs2_consist_inode(GFS2_I(jd->jd_inode)); - return -EIO; - } - } -} - -/** - * jhead_scan - make sure we've found the head of the log - * @jd: the journal - * @head: this is filled in with the log descriptor of the head - * - * At this point, seg and lh should be either the head of the log or just - * before. Scan forward until we find the head. - * - * Returns: errno - */ - -static int jhead_scan(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head) -{ - unsigned int blk = head->lh_blkno; - struct gfs2_log_header_host lh; - int error; - - for (;;) { - if (++blk == jd->jd_blocks) - blk = 0; - - error = get_log_header(jd, blk, &lh); - if (error < 0) - return error; - if (error == 1) - continue; - - if (lh.lh_sequence == head->lh_sequence) { - gfs2_consist_inode(GFS2_I(jd->jd_inode)); - return -EIO; - } - if (lh.lh_sequence < head->lh_sequence) - break; - - *head = lh; - } - - return 0; -} - -/** - * gfs2_find_jhead - find the head of a log - * @jd: the journal - * @head: the log descriptor for the head of the log is returned here - * - * Do a binary search of a journal and find the valid log entry with the - * highest sequence number. (i.e. the log head) - * - * Returns: errno - */ - -int gfs2_find_jhead(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head) -{ - struct gfs2_log_header_host lh_1, lh_m; - u32 blk_1, blk_2, blk_m; - int error; - - blk_1 = 0; - blk_2 = jd->jd_blocks - 1; - - for (;;) { - blk_m = (blk_1 + blk_2) / 2; - - error = find_good_lh(jd, &blk_1, &lh_1); - if (error) - return error; - - error = find_good_lh(jd, &blk_m, &lh_m); - if (error) - return error; - - if (blk_1 == blk_m || blk_m == blk_2) - break; - - if (lh_1.lh_sequence <= lh_m.lh_sequence) - blk_1 = blk_m; - else - blk_2 = blk_m; - } - - error = jhead_scan(jd, &lh_1); - if (error) - return error; - - *head = lh_1; - return error; } @@ -460,6 +348,8 @@ void gfs2_recover_func(struct work_struct *work) if (error) goto fail_gunlock_ji; t_jhd = ktime_get(); + fs_info(sdp, "jid=%u: Journal head lookup took %lldms\n", jd->jd_jid, + ktime_ms_delta(t_jhd, t_jlck)); if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT)) { fs_info(sdp, "jid=%u: Acquiring the transaction lock...\n", diff --git a/fs/gfs2/recovery.h b/fs/gfs2/recovery.h @@ -27,10 +27,11 @@ extern int gfs2_revoke_add(struct gfs2_jdesc *jd, u64 blkno, unsigned int where) extern int gfs2_revoke_check(struct gfs2_jdesc *jd, u64 blkno, unsigned int where); extern void gfs2_revoke_clean(struct gfs2_jdesc *jd); -extern int gfs2_find_jhead(struct gfs2_jdesc *jd, - struct gfs2_log_header_host *head); extern int gfs2_recover_journal(struct gfs2_jdesc *gfs2_jd, bool wait); extern void gfs2_recover_func(struct work_struct *work); +extern int __get_log_header(struct gfs2_sbd *sdp, + const struct gfs2_log_header *lh, unsigned int blkno, + struct gfs2_log_header_host *head); #endif /* __RECOVERY_DOT_H__ */ diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c @@ -1780,9 +1780,9 @@ static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext, goto next_iter; } if (ret == -E2BIG) { + n += rbm->bii - initial_bii; rbm->bii = 0; rbm->offset = 0; - n += (rbm->bii - initial_bii); goto res_covered_end_of_rgrp; } return ret; @@ -2256,7 +2256,7 @@ static void rgblk_free(struct gfs2_sbd *sdp, struct gfs2_rgrpd *rgd, * */ -void gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl) +void gfs2_rgrp_dump(struct seq_file *seq, struct gfs2_glock *gl) { struct gfs2_rgrpd *rgd = gl->gl_object; struct gfs2_blkreserv *trs; diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h @@ -72,7 +72,7 @@ extern void gfs2_rlist_add(struct gfs2_inode *ip, struct gfs2_rgrp_list *rlist, extern void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist); extern void gfs2_rlist_free(struct gfs2_rgrp_list *rlist); extern u64 gfs2_ri_total(struct gfs2_sbd *sdp); -extern void gfs2_rgrp_dump(struct seq_file *seq, const struct gfs2_glock *gl); +extern void gfs2_rgrp_dump(struct seq_file *seq, struct gfs2_glock *gl); extern int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset, struct buffer_head *bh, const struct gfs2_bitmap *bi, unsigned minlen, u64 *ptrimmed); diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c @@ -45,6 +45,7 @@ #include "util.h" #include "sys.h" #include "xattr.h" +#include "lops.h" #define args_neq(a1, a2, x) ((a1)->ar_##x != (a2)->ar_##x) diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c @@ -124,15 +124,13 @@ void gfs2_trans_end(struct gfs2_sbd *sdp) } static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl, - struct buffer_head *bh, - const struct gfs2_log_operations *lops) + struct buffer_head *bh) { struct gfs2_bufdata *bd; bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS | __GFP_NOFAIL); bd->bd_bh = bh; bd->bd_gl = gl; - bd->bd_ops = lops; INIT_LIST_HEAD(&bd->bd_list); bh->b_private = bd; return bd; @@ -169,7 +167,7 @@ void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) gfs2_log_unlock(sdp); unlock_buffer(bh); if (bh->b_private == NULL) - bd = gfs2_alloc_bufdata(gl, bh, &gfs2_databuf_lops); + bd = gfs2_alloc_bufdata(gl, bh); else bd = bh->b_private; lock_buffer(bh); @@ -210,7 +208,7 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh) unlock_buffer(bh); lock_page(bh->b_page); if (bh->b_private == NULL) - bd = gfs2_alloc_bufdata(gl, bh, &gfs2_buf_lops); + bd = gfs2_alloc_bufdata(gl, bh); else bd = bh->b_private; unlock_page(bh->b_page);