Re: [f2fs-dev] [PATCH v4 1/2] f2fs: separate buffer and direct io in block allocation statistics

2021-09-28 Thread Chao Yu
On 2021/9/22 14:25, Fengnan Chang wrote: separate buffer and direct io in block allocation statistics. New output will like this: buffer direct segments IPU:0 0N/A SSR:0 0 0 LFS:0 0 0 Sign

[f2fs-dev] [PATCH] f2fs: skip f2fs_preallocate_blocks() for overwrite case

2021-09-28 Thread Chao Yu
In f2fs_file_write_iter(), let's use f2fs_overwrite_io() to check whethere it is overwrite case, for such case, we can skip f2fs_preallocate_blocks() in order to avoid f2fs_do_map_lock(), which may be blocked by checkpoint() potentially. Signed-off-by: Chao Yu --- fs/f2fs/file.c | 4 1 file

[f2fs-dev] [Bug 214551] Roll-Forward recovery recover deleted file

2021-09-28 Thread bugzilla-daemon
https://bugzilla.kernel.org/show_bug.cgi?id=214551 Chao Yu (c...@kernel.org) changed: What|Removed |Added Status|NEW |ASSIGNED CC|

Re: [f2fs-dev] [PATCH] f2fs: fix whint_mode mount option error

2021-09-28 Thread Chao Yu
On 2021/9/27 14:09, Fengnan Chang wrote: When mount with whint_mode option, it doesn't work, Fix it. Fixes: d0b9e42ab615 (f2fs: introduce inmem curseg) Thanks for the patch, it looks Keoseong Park has submitted the same fix patch a little bit earlier I guess we need to select one of your p

Re: [f2fs-dev] [PATCH] f2fs: fix whint_mode mount option error

2021-09-28 Thread Jaegeuk Kim
On 09/28, Chao Yu wrote: > On 2021/9/27 14:09, Fengnan Chang wrote: > > When mount with whint_mode option, it doesn't work, Fix it. > > Fixes: d0b9e42ab615 (f2fs: introduce inmem curseg) > > Thanks for the patch, it looks Keoseong Park has submitted the same > fix patch a little bit earlier >

Re: [f2fs-dev] [PATCH] f2fs: skip f2fs_preallocate_blocks() for overwrite case

2021-09-28 Thread Jaegeuk Kim
On 09/28, Chao Yu wrote: > In f2fs_file_write_iter(), let's use f2fs_overwrite_io() to > check whethere it is overwrite case, for such case, we can skip > f2fs_preallocate_blocks() in order to avoid f2fs_do_map_lock(), > which may be blocked by checkpoint() potentially. > > Signed-off-by: Chao Yu

[f2fs-dev] [PATCH] f2fs-tools: fall back to the original version check when clock_gettime is not supported

2021-09-28 Thread Daeho Jeong
From: Daeho Jeong In lower versions than macOS 10.12, they don't support clock_gettime function. It breaks the build, so we need to fall back to the original kernel version check algorithm, in that case. Signed-off-by: Daeho Jeong --- fsck/mount.c | 11 +++ 1 file changed, 11 insertion

Re: [f2fs-dev] [PATCH] f2fs-tools: fall back to the original version check when clock_gettime is not supported

2021-09-28 Thread Jaegeuk Kim
Merged to address android build breakage. On 09/28, Daeho Jeong wrote: > From: Daeho Jeong > > In lower versions than macOS 10.12, they don't support clock_gettime > function. It breaks the build, so we need to fall back to the original > kernel version check algorithm, in that case. > > Signed

Re: [f2fs-dev] [PATCH] f2fs: skip f2fs_preallocate_blocks() for overwrite case

2021-09-28 Thread Chao Yu
On 2021/9/29 3:08, Jaegeuk Kim wrote: On 09/28, Chao Yu wrote: In f2fs_file_write_iter(), let's use f2fs_overwrite_io() to check whethere it is overwrite case, for such case, we can skip f2fs_preallocate_blocks() in order to avoid f2fs_do_map_lock(), which may be blocked by checkpoint() potentia

[f2fs-dev] [PATCH] f2fs: fix wrong condition to trigger background checkpoint correctly

2021-09-28 Thread Chao Yu
In f2fs_balance_fs_bg(), it needs to check both NAT_ENTRIES and INO_ENTRIES memory usage to decide whether we should skip background checkpoint, otherwise we may always skip checking INO_ENTRIES memory usage, so that INO_ENTRIES may potentially cause high memory footprint. Fixes: 493720a48543 ("f2