On 12/15, heyunlei wrote: > > > >-----Original Message----- > >From: Jaegeuk Kim [mailto:[email protected]] > >Sent: Friday, December 15, 2017 12:21 PM > >To: heyunlei > >Cc: Yuchao (T); [email protected]; Wangbintian; Jianing > >(Euler) > >Subject: Re: [f2fs-dev][PATCH RFC] f2fs: fix an error case of missing update > >inode page > > > >On 12/15, heyunlei wrote: > >> > >> > >> >-----Original Message----- > >> >From: Jaegeuk Kim [mailto:[email protected]] > >> >Sent: Friday, December 15, 2017 3:46 AM > >> >To: heyunlei > >> >Cc: Yuchao (T); [email protected]; Wangbintian; > >> >Jianing (Euler) > >> >Subject: Re: [f2fs-dev][PATCH RFC] f2fs: fix an error case of missing > >> >update inode page > >> > > >> >On 12/05, Yunlei He wrote: > >> >> -Thread A Thread B > >> >> > >> >> -write_checkpoint > >> >> -block_operations > >> >> -f2fs_unlock_all -f2fs_sync_file > >> >> > >> >> -f2fs_write_inode > >> >> > >> >> -f2fs_inode_synced > >> >> > >> >> -f2fs_sync_inode_meta > >> >> -sync_node_pages > >> >> > >> >> -set_page_drity > >> >> > >> >> In this case, if sudden power off without next new checkpoint, > >> >> the last inode page update will lost. wb_writeback is same with > >> >> fsync. > >> > > >> >Does this patch fix the problem? > >> > > >> I modify code as below: > >> @@ -366,7 +366,7 @@ int update_inode(struct inode *inode, struct page > >> *node_page) > >> struct extent_tree *et = F2FS_I(inode)->extent_tree; > >> > >> f2fs_inode_synced(inode); > >> - > >> + msleep(10000); > >> f2fs_wait_on_page_writeback(node_page, NODE, true); > >> > >> shell 1: shell2: > >> dd if=/dev/zero of=./test bs=1M count=10 > >> sync > >> echo "hello" >> ./test > >> fsync test // sleep 10s > >> sync //return quickly > >> > >> echo c > /proc/sysrq-trigger //before fsync return > >> fsck will find iblocks mismatch, and with this patch it 's ok. > > > >So, do you mean it was fixed by adding msleep() on top of this patch as well? > >Like: > > - f2fs_write_inode() > > - set_page_drity > > - f2fs_inode_synced > > - msleep(10000); > > > Yes, write_checkpoint will be blocked in sync dirty node pages, until fsync > update inode > page and unlock node page.
Got it. I've concerned about inversion of PageDirty() and f2fs_inode_synced(), but for now, I can't find any corner case simply. Let me do stress test with this. Thanks, > > Thanks. > >> > >> Besides,I came across another error like this: > >> > >> f2fs_inode-> i_xattr_nid is non-zero, but node blkaddr in raw nat entry is > >> NULL. > >> > >> This case makes sense to this problem also. > >> > >> > -write_checkpoint > >> > -block_operations > >> > -f2fs_unlock_all -f2fs_sync_file > >> > -f2fs_write_inode > >> > -set_page_drity > >> > > >> > -f2fs_inode_synced > >> > -f2fs_sync_inode_meta > >> > -sync_node_pages > >> > > >> >I think, we need to cover f2fs_lock_op() for f2fs_write_inode() again. > >> > >> It can also work well with f2fs_lock_op(), I am afraid other cases will > >> update inode > >> like f2fs_write_inode without f2fs_lock_op > >> > >> Thanks > >> > > >> >Thanks, > >> > > >> >> > >> >> Signed-off-by: Yunlei He <[email protected]> > >> >> --- > >> >> fs/f2fs/f2fs.h | 4 ++-- > >> >> fs/f2fs/inode.c | 15 +++++++-------- > >> >> 2 files changed, 9 insertions(+), 10 deletions(-) > >> >> > >> >> diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h > >> >> index 82f1dc3..38f9324 100644 > >> >> --- a/fs/f2fs/f2fs.h > >> >> +++ b/fs/f2fs/f2fs.h > >> >> @@ -2513,8 +2513,8 @@ int f2fs_getattr(const struct path *path, struct > >> >> kstat *stat, > >> >> struct inode *f2fs_iget(struct super_block *sb, unsigned long ino); > >> >> struct inode *f2fs_iget_retry(struct super_block *sb, unsigned long > >> >> ino); > >> >> int try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink); > >> >> -int update_inode(struct inode *inode, struct page *node_page); > >> >> -int update_inode_page(struct inode *inode); > >> >> +void update_inode(struct inode *inode, struct page *node_page); > >> >> +void update_inode_page(struct inode *inode); > >> >> int f2fs_write_inode(struct inode *inode, struct writeback_control > >> >> *wbc); > >> >> void f2fs_evict_inode(struct inode *inode); > >> >> void handle_failed_inode(struct inode *inode); > >> >> diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c > >> >> index b4c4f2b..10d3c7c 100644 > >> >> --- a/fs/f2fs/inode.c > >> >> +++ b/fs/f2fs/inode.c > >> >> @@ -360,14 +360,15 @@ struct inode *f2fs_iget_retry(struct super_block > >> >> *sb, unsigned long ino) > >> >> return inode; > >> >> } > >> >> > >> >> -int update_inode(struct inode *inode, struct page *node_page) > >> >> +void update_inode(struct inode *inode, struct page *node_page) > >> >> { > >> >> struct f2fs_inode *ri; > >> >> struct extent_tree *et = F2FS_I(inode)->extent_tree; > >> >> > >> >> - f2fs_inode_synced(inode); > >> >> - > >> >> f2fs_wait_on_page_writeback(node_page, NODE, true); > >> >> + set_page_dirty(node_page); > >> >> + > >> >> + f2fs_inode_synced(inode); > >> >> > >> >> ri = F2FS_INODE(node_page); > >> >> > >> >> @@ -426,10 +427,9 @@ int update_inode(struct inode *inode, struct page > >> >> *node_page) > >> >> if (inode->i_nlink == 0) > >> >> clear_inline_node(node_page); > >> >> > >> >> - return set_page_dirty(node_page); > >> >> } > >> >> > >> >> -int update_inode_page(struct inode *inode) > >> >> +void update_inode_page(struct inode *inode) > >> >> { > >> >> struct f2fs_sb_info *sbi = F2FS_I_SB(inode); > >> >> struct page *node_page; > >> >> @@ -444,11 +444,10 @@ int update_inode_page(struct inode *inode) > >> >> } else if (err != -ENOENT) { > >> >> f2fs_stop_checkpoint(sbi, false); > >> >> } > >> >> - return 0; > >> >> + return; > >> >> } > >> >> - ret = update_inode(inode, node_page); > >> >> + update_inode(inode, node_page); > >> >> f2fs_put_page(node_page, 1); > >> >> - return ret; > >> >> } > >> >> > >> >> int f2fs_write_inode(struct inode *inode, struct writeback_control > >> >> *wbc) > >> >> -- > >> >> 1.9.1 ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Linux-f2fs-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
