Messages in this thread | | | Date | Tue, 18 Sep 2018 10:55:33 -0700 | From | Jaegeuk Kim <> | Subject | Re: [f2fs-dev] [PATCH 2/2] f2fs: avoid f2fs_bug_on if f2fs_get_meta_page_nofail got EIO |
| |
On 09/18, Chao Yu wrote: > On 2018/9/18 10:18, Jaegeuk Kim wrote: > > This patch avoids BUG_ON when f2fs_get_meta_page_nofail got EIO during > > xfstests/generic/475. > > > > Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org> > > --- > > fs/f2fs/checkpoint.c | 2 +- > > fs/f2fs/gc.c | 2 ++ > > fs/f2fs/node.c | 12 ++++++++++-- > > fs/f2fs/recovery.c | 2 ++ > > fs/f2fs/segment.c | 3 +++ > > 5 files changed, 18 insertions(+), 3 deletions(-) > > > > diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c > > index 01e0d8f5bbbe..6ce3cb6502dd 100644 > > --- a/fs/f2fs/checkpoint.c > > +++ b/fs/f2fs/checkpoint.c > > @@ -121,7 +121,7 @@ struct page *f2fs_get_meta_page_nofail(struct f2fs_sb_info *sbi, pgoff_t index) > > goto retry; > > > > f2fs_stop_checkpoint(sbi, false); > > - f2fs_bug_on(sbi, 1); > > + return NULL; > > How about propagate PTR_ERR(page) to caller?
Done.
> > > } > > > > return page; > > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > > index 4bcc8a59fdef..d049865887cf 100644 > > --- a/fs/f2fs/gc.c > > +++ b/fs/f2fs/gc.c > > @@ -1070,6 +1070,8 @@ static int do_garbage_collect(struct f2fs_sb_info *sbi, > > /* reference all summary page */ > > while (segno < end_segno) { > > sum_page = f2fs_get_sum_page(sbi, segno++); > > + if (!sum_page) > > + return -EIO; > > Well, for large section, we need to release all referenced sum page by > f2fs_put_page().
Done.
> > > unlock_page(sum_page); > > } > > > > diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c > > index fa2381c0bc47..b3595522c35b 100644 > > --- a/fs/f2fs/node.c > > +++ b/fs/f2fs/node.c > > @@ -126,6 +126,8 @@ static struct page *get_next_nat_page(struct f2fs_sb_info *sbi, nid_t nid) > > > > /* get current nat block page with lock */ > > src_page = get_current_nat_page(sbi, nid); > > + if (!src_page) > > + return NULL; > > dst_page = f2fs_grab_meta_page(sbi, dst_off); > > f2fs_bug_on(sbi, PageDirty(src_page)); > > > > @@ -2265,8 +2267,12 @@ static int __f2fs_build_free_nids(struct f2fs_sb_info *sbi, > > nm_i->nat_block_bitmap)) { > > struct page *page = get_current_nat_page(sbi, nid); > > > > - ret = scan_nat_page(sbi, page, nid); > > - f2fs_put_page(page, 1); > > + if (page) { > > + ret = scan_nat_page(sbi, page, nid); > > + f2fs_put_page(page, 1); > > + } else { > > + ret = -EIO; > > + } > > > > if (ret) { > > up_read(&nm_i->nat_tree_lock); > > Should propagate the error to f2fs_alloc_nid()?
Done.
> > > @@ -2724,6 +2730,8 @@ static void __flush_nat_entry_set(struct f2fs_sb_info *sbi, > > down_write(&curseg->journal_rwsem); > > } else { > > page = get_next_nat_page(sbi, start_nid); > > + if (!page) > > + return; > > Ditto, propagate such error to write_checkpoint()?
Done.
> > > nat_blk = page_address(page); > > f2fs_bug_on(sbi, !nat_blk); > > } > > diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c > > index 56d34193a74b..a3dce16bfd6c 100644 > > --- a/fs/f2fs/recovery.c > > +++ b/fs/f2fs/recovery.c > > @@ -355,6 +355,8 @@ static int check_index_in_prev_nodes(struct f2fs_sb_info *sbi, > > } > > > > sum_page = f2fs_get_sum_page(sbi, segno); > > + if (!sum_page) > > + return -EIO; > > sum_node = (struct f2fs_summary_block *)page_address(sum_page); > > sum = sum_node->entries[blkoff]; > > f2fs_put_page(sum_page, 1); > > diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c > > index aa96a371aaf8..cfc9eb492da1 100644 > > --- a/fs/f2fs/segment.c > > +++ b/fs/f2fs/segment.c > > @@ -2487,6 +2487,7 @@ static void change_curseg(struct f2fs_sb_info *sbi, int type) > > __next_free_blkoff(sbi, curseg, 0); > > > > sum_page = f2fs_get_sum_page(sbi, new_segno); > > + f2fs_bug_on(sbi, !sum_page); > > Well, next time we may panic here...
This is the same as before, and it's almost impossible to hit anyway in production.
Thanks,
> > In product, for EIO case, usually we just reboot cell phone directly to avoid > potential data loss later. > > So I just set DEFAULT_RETRY_IO_COUNT to 32 temporarily to pass xfstest IO error > injection cases. > > Thanks, > > > sum_node = (struct f2fs_summary_block *)page_address(sum_page); > > memcpy(curseg->sum_blk, sum_node, SUM_ENTRY_SIZE); > > f2fs_put_page(sum_page, 1); > > @@ -3971,6 +3972,8 @@ static int build_sit_entries(struct f2fs_sb_info *sbi) > > > > se = &sit_i->sentries[start]; > > page = get_current_sit_page(sbi, start); > > + if (!page) > > + return err; > > sit_blk = (struct f2fs_sit_block *)page_address(page); > > sit = sit_blk->entries[SIT_ENTRY_OFFSET(sit_i, start)]; > > f2fs_put_page(page, 1); > >
| |