| Message ID | fdb16638-3d83-c065-2e49-996fb6310643@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] ocfs2: flush truncate log when main bitmap run out of space | expand |
On 18/12/28 16:29, Jia Guo wrote: > Truncate log problem has been described in > commit 50308d813bf2 ("ocfs2: Try to free truncate log when meeting > ENOSPC in write.") and > commit 2070ad1aebff ("ocfs2: retry on ENOSPC if sufficient space in > truncate log"), but the following situations cannot be solved: > > case 1: > Main bitmap has been updated, but the transaction of the deleted blocks has > not been committed completely. In this case, function > ocfs2_reserve_cluster_bitmap_bits() returns success while function > __ocfs2_claim_clusters return ENOSPC because we cannot reuse the deleted > blocks before the transaction committed(test by function > ocfs2_test_bg_bit_allocatable()). One can reproduce this by following > steps: > a, prepare a file which size is 50G, and volume still have 30G free space > b, open the file with O_TRUNC flag > c, sleep 5 seconds > d, fallocate a 50G file, fallocate will fail. > > case 2: > Main bitmap doesn't have enough free bits, so does truncate log. But main > bitmap plus truncate log has enough free bits. We are not gonging to try > flush truncate log in this case, which is not reasonable. > > So force commit the transaction when flushing the truncate log for case 1. > For case 2, the value of osb->truncated_clusters doesn't seem to make > sense, do the flush whenever we run out of space seems to be more > reasonable. > > Signed-off-by: Jia Guo <guojia12@huawei.com> > Reviewed-by: Yiwen Jiang <jiangyiwen@huawei.com> > Reviewed-by: Gang He <ghe@suse.com> Acked-by: Joseph Qi <jiangqi903@gmail.com>
Hi Jia, This patch looks good to me, but I have a suggestion as below: On 2018/12/28 16:29, Jia Guo wrote: > Truncate log problem has been described in > commit 50308d813bf2 ("ocfs2: Try to free truncate log when meeting > ENOSPC in write.") and > commit 2070ad1aebff ("ocfs2: retry on ENOSPC if sufficient space in > truncate log"), but the following situations cannot be solved: > > case 1: > Main bitmap has been updated, but the transaction of the deleted blocks has > not been committed completely. In this case, function > ocfs2_reserve_cluster_bitmap_bits() returns success while function > __ocfs2_claim_clusters return ENOSPC because we cannot reuse the deleted > blocks before the transaction committed(test by function > ocfs2_test_bg_bit_allocatable()). One can reproduce this by following > steps: > a, prepare a file which size is 50G, and volume still have 30G free space > b, open the file with O_TRUNC flag > c, sleep 5 seconds > d, fallocate a 50G file, fallocate will fail. > > case 2: > Main bitmap doesn't have enough free bits, so does truncate log. But main > bitmap plus truncate log has enough free bits. We are not gonging to try > flush truncate log in this case, which is not reasonable. > > So force commit the transaction when flushing the truncate log for case 1. > For case 2, the value of osb->truncated_clusters doesn't seem to make > sense, do the flush whenever we run out of space seems to be more > reasonable. > > Signed-off-by: Jia Guo <guojia12@huawei.com> > Reviewed-by: Yiwen Jiang <jiangyiwen@huawei.com> > Reviewed-by: Gang He <ghe@suse.com> > --- > fs/ocfs2/alloc.c | 43 +++---------------------------------------- > fs/ocfs2/alloc.h | 2 -- > fs/ocfs2/aops.c | 8 +++----- > fs/ocfs2/ocfs2.h | 5 ----- > fs/ocfs2/suballoc.c | 6 +++--- > 5 files changed, 9 insertions(+), 55 deletions(-) > > diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c > index d1cbb27..8b6938d 100644 > --- a/fs/ocfs2/alloc.c > +++ b/fs/ocfs2/alloc.c > @@ -5921,7 +5921,6 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb, > > ocfs2_journal_dirty(handle, tl_bh); > > - osb->truncated_clusters += num_clusters; > bail: > return status; > } > @@ -5940,6 +5939,7 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb, > struct inode *tl_inode = osb->osb_tl_inode; > struct buffer_head *tl_bh = osb->osb_tl_bh; > handle_t *handle; > + tid_t target; > > di = (struct ocfs2_dinode *) tl_bh->b_data; > tl = &di->id2.i_dealloc; > @@ -5989,8 +5989,8 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb, > ocfs2_commit_trans(osb, handle); > i--; > } > - > - osb->truncated_clusters = 0; > + if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) > + jbd2_log_wait_commit(osb->journal->j_journal, target); Catch the errno of jbd2_log_wait_commit here. And this will inform uppper caller no need to retry. Thanks, Jun > > bail: > return status; > @@ -6102,43 +6102,6 @@ void ocfs2_schedule_truncate_log_flush(struct ocfs2_super *osb, > } > } > > -/* > - * Try to flush truncate logs if we can free enough clusters from it. > - * As for return value, "< 0" means error, "0" no space and "1" means > - * we have freed enough spaces and let the caller try to allocate again. > - */ > -int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb, > - unsigned int needed) > -{ > - tid_t target; > - int ret = 0; > - unsigned int truncated_clusters; > - > - inode_lock(osb->osb_tl_inode); > - truncated_clusters = osb->truncated_clusters; > - inode_unlock(osb->osb_tl_inode); > - > - /* > - * Check whether we can succeed in allocating if we free > - * the truncate log. > - */ > - if (truncated_clusters < needed) > - goto out; > - > - ret = ocfs2_flush_truncate_log(osb); > - if (ret) { > - mlog_errno(ret); > - goto out; > - } > - > - if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) { > - jbd2_log_wait_commit(osb->journal->j_journal, target); > - ret = 1; > - } > -out: > - return ret; > -} > - > static int ocfs2_get_truncate_log_info(struct ocfs2_super *osb, > int slot_num, > struct inode **tl_inode, > diff --git a/fs/ocfs2/alloc.h b/fs/ocfs2/alloc.h > index 250bcac..b343a6f 100644 > --- a/fs/ocfs2/alloc.h > +++ b/fs/ocfs2/alloc.h > @@ -188,8 +188,6 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb, > u64 start_blk, > unsigned int num_clusters); > int __ocfs2_flush_truncate_log(struct ocfs2_super *osb); > -int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb, > - unsigned int needed); > > /* > * Process local structure which describes the block unlinks done > diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c > index eb1ce30..1a1af0c 100644 > --- a/fs/ocfs2/aops.c > +++ b/fs/ocfs2/aops.c > @@ -1674,7 +1674,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, > struct buffer_head *di_bh, struct page *mmap_page) > { > int ret, cluster_of_pages, credits = OCFS2_INODE_UPDATE_CREDITS; > - unsigned int clusters_to_alloc, extents_to_split, clusters_need = 0; > + unsigned int clusters_to_alloc, extents_to_split; > struct ocfs2_write_ctxt *wc; > struct inode *inode = mapping->host; > struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); > @@ -1723,7 +1723,6 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, > mlog_errno(ret); > goto out; > } else if (ret == 1) { > - clusters_need = wc->w_clen; > ret = ocfs2_refcount_cow(inode, di_bh, > wc->w_cpos, wc->w_clen, UINT_MAX); > if (ret) { > @@ -1738,7 +1737,6 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, > mlog_errno(ret); > goto out; > } > - clusters_need += clusters_to_alloc; > > di = (struct ocfs2_dinode *)wc->w_di_bh->b_data; > > @@ -1893,8 +1891,8 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, > */ > try_free = 0; > > - ret1 = ocfs2_try_to_free_truncate_log(osb, clusters_need); > - if (ret1 == 1) > + ret1 = ocfs2_flush_truncate_log(osb); > + if (!ret1) > goto try_again; > > if (ret1 < 0) > diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h > index 4f86ac0..f913647 100644 > --- a/fs/ocfs2/ocfs2.h > +++ b/fs/ocfs2/ocfs2.h > @@ -440,11 +440,6 @@ struct ocfs2_super > struct buffer_head *osb_tl_bh; > struct delayed_work osb_truncate_log_wq; > atomic_t osb_tl_disable; > - /* > - * How many clusters in our truncate log. > - * It must be protected by osb_tl_inode->i_mutex. > - */ > - unsigned int truncated_clusters; > > struct ocfs2_node_map osb_recovering_orphan_dirs; > unsigned int *osb_orphan_wipes; > diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c > index f7c972f..d086bb9 100644 > --- a/fs/ocfs2/suballoc.c > +++ b/fs/ocfs2/suballoc.c > @@ -1187,14 +1187,14 @@ static int ocfs2_reserve_clusters_with_limit(struct ocfs2_super *osb, > if (status == -ENOSPC) { > retry: > status = ocfs2_reserve_cluster_bitmap_bits(osb, *ac); > - /* Retry if there is sufficient space cached in truncate log */ > + /* Retry after flush truncate log */ > if (status == -ENOSPC && !retried) { > retried = 1; > ocfs2_inode_unlock((*ac)->ac_inode, 1); > inode_unlock((*ac)->ac_inode); > > - ret = ocfs2_try_to_free_truncate_log(osb, bits_wanted); > - if (ret == 1) { > + ret = ocfs2_flush_truncate_log(osb); > + if (!ret) { > iput((*ac)->ac_inode); > (*ac)->ac_inode = NULL; > goto retry; >
diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index d1cbb27..8b6938d 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -5921,7 +5921,6 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb, ocfs2_journal_dirty(handle, tl_bh); - osb->truncated_clusters += num_clusters; bail: return status; } @@ -5940,6 +5939,7 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb, struct inode *tl_inode = osb->osb_tl_inode; struct buffer_head *tl_bh = osb->osb_tl_bh; handle_t *handle; + tid_t target; di = (struct ocfs2_dinode *) tl_bh->b_data; tl = &di->id2.i_dealloc; @@ -5989,8 +5989,8 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb, ocfs2_commit_trans(osb, handle); i--; } - - osb->truncated_clusters = 0; + if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) + jbd2_log_wait_commit(osb->journal->j_journal, target); bail: return status; @@ -6102,43 +6102,6 @@ void ocfs2_schedule_truncate_log_flush(struct ocfs2_super *osb, } } -/* - * Try to flush truncate logs if we can free enough clusters from it. - * As for return value, "< 0" means error, "0" no space and "1" means - * we have freed enough spaces and let the caller try to allocate again. - */ -int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb, - unsigned int needed) -{ - tid_t target; - int ret = 0; - unsigned int truncated_clusters; - - inode_lock(osb->osb_tl_inode); - truncated_clusters = osb->truncated_clusters; - inode_unlock(osb->osb_tl_inode); - - /* - * Check whether we can succeed in allocating if we free - * the truncate log. - */ - if (truncated_clusters < needed) - goto out; - - ret = ocfs2_flush_truncate_log(osb); - if (ret) { - mlog_errno(ret); - goto out; - } - - if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) { - jbd2_log_wait_commit(osb->journal->j_journal, target); - ret = 1; - } -out: - return ret; -} - static int ocfs2_get_truncate_log_info(struct ocfs2_super *osb, int slot_num, struct inode **tl_inode, diff --git a/fs/ocfs2/alloc.h b/fs/ocfs2/alloc.h index 250bcac..b343a6f 100644 --- a/fs/ocfs2/alloc.h +++ b/fs/ocfs2/alloc.h @@ -188,8 +188,6 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb, u64 start_blk, unsigned int num_clusters); int __ocfs2_flush_truncate_log(struct ocfs2_super *osb); -int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb, - unsigned int needed); /* * Process local structure which describes the block unlinks done diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index eb1ce30..1a1af0c 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -1674,7 +1674,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, struct buffer_head *di_bh, struct page *mmap_page) { int ret, cluster_of_pages, credits = OCFS2_INODE_UPDATE_CREDITS; - unsigned int clusters_to_alloc, extents_to_split, clusters_need = 0; + unsigned int clusters_to_alloc, extents_to_split; struct ocfs2_write_ctxt *wc; struct inode *inode = mapping->host; struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); @@ -1723,7 +1723,6 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, mlog_errno(ret); goto out; } else if (ret == 1) { - clusters_need = wc->w_clen; ret = ocfs2_refcount_cow(inode, di_bh, wc->w_cpos, wc->w_clen, UINT_MAX); if (ret) { @@ -1738,7 +1737,6 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, mlog_errno(ret); goto out; } - clusters_need += clusters_to_alloc; di = (struct ocfs2_dinode *)wc->w_di_bh->b_data; @@ -1893,8 +1891,8 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, */ try_free = 0; - ret1 = ocfs2_try_to_free_truncate_log(osb, clusters_need); - if (ret1 == 1) + ret1 = ocfs2_flush_truncate_log(osb); + if (!ret1) goto try_again; if (ret1 < 0) diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 4f86ac0..f913647 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -440,11 +440,6 @@ struct ocfs2_super struct buffer_head *osb_tl_bh; struct delayed_work osb_truncate_log_wq; atomic_t osb_tl_disable; - /* - * How many clusters in our truncate log. - * It must be protected by osb_tl_inode->i_mutex. - */ - unsigned int truncated_clusters; struct ocfs2_node_map osb_recovering_orphan_dirs; unsigned int *osb_orphan_wipes; diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index f7c972f..d086bb9 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -1187,14 +1187,14 @@ static int ocfs2_reserve_clusters_with_limit(struct ocfs2_super *osb, if (status == -ENOSPC) { retry: status = ocfs2_reserve_cluster_bitmap_bits(osb, *ac); - /* Retry if there is sufficient space cached in truncate log */ + /* Retry after flush truncate log */ if (status == -ENOSPC && !retried) { retried = 1; ocfs2_inode_unlock((*ac)->ac_inode, 1); inode_unlock((*ac)->ac_inode); - ret = ocfs2_try_to_free_truncate_log(osb, bits_wanted); - if (ret == 1) { + ret = ocfs2_flush_truncate_log(osb); + if (!ret) { iput((*ac)->ac_inode); (*ac)->ac_inode = NULL; goto retry;
Truncate log problem has been described in commit 50308d813bf2 ("ocfs2: Try to free truncate log when meeting ENOSPC in write.") and commit 2070ad1aebff ("ocfs2: retry on ENOSPC if sufficient space in truncate log"), but the following situations cannot be solved: case 1: Main bitmap has been updated, but the transaction of the deleted blocks has not been committed completely. In this case, function ocfs2_reserve_cluster_bitmap_bits() returns success while function __ocfs2_claim_clusters return ENOSPC because we cannot reuse the deleted blocks before the transaction committed(test by function ocfs2_test_bg_bit_allocatable()). One can reproduce this by following steps: a, prepare a file which size is 50G, and volume still have 30G free space b, open the file with O_TRUNC flag c, sleep 5 seconds d, fallocate a 50G file, fallocate will fail. case 2: Main bitmap doesn't have enough free bits, so does truncate log. But main bitmap plus truncate log has enough free bits. We are not gonging to try flush truncate log in this case, which is not reasonable. So force commit the transaction when flushing the truncate log for case 1. For case 2, the value of osb->truncated_clusters doesn't seem to make sense, do the flush whenever we run out of space seems to be more reasonable. Signed-off-by: Jia Guo <guojia12@huawei.com> Reviewed-by: Yiwen Jiang <jiangyiwen@huawei.com> Reviewed-by: Gang He <ghe@suse.com> --- fs/ocfs2/alloc.c | 43 +++---------------------------------------- fs/ocfs2/alloc.h | 2 -- fs/ocfs2/aops.c | 8 +++----- fs/ocfs2/ocfs2.h | 5 ----- fs/ocfs2/suballoc.c | 6 +++--- 5 files changed, 9 insertions(+), 55 deletions(-)