Re: [f2fs-dev] [PATCH v2] f2fs: add sysfs nodes to get GC info for each GC mode

From: Daeho Jeong
Date: Sat Jul 10 2021 - 01:19:10 EST


On Fri, Jul 9, 2021 at 5:17 PM Chao Yu <chao@xxxxxxxxxx> wrote:
>
> On 2021/7/10 4:44, Daeho Jeong wrote:
> > From: Daeho Jeong <daehojeong@xxxxxxxxxx>
> >
> > Added gc_dirty_segs and gc_dirty_segs_mode sysfs nodes.
> > 1) "gc_dirty_segs" shows how many dirty segments have been
> > reclaimed by GC during a specific GC mode.
> > 2) "gc_dirty_segs_mode" is used to control for which gc mode
> > the "gc_dirty_segs" node shows.
> >
> > Signed-off-by: Daeho Jeong <daehojeong@xxxxxxxxxx>
> > ---
> > Documentation/ABI/testing/sysfs-fs-f2fs | 14 +++++++++++++
> > fs/f2fs/debug.c | 9 ++++++++
> > fs/f2fs/f2fs.h | 5 +++++
> > fs/f2fs/gc.c | 1 +
> > fs/f2fs/sysfs.c | 28 +++++++++++++++++++++++++
> > 5 files changed, 57 insertions(+)
> >
> > diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs
> > index 95155e4ec7fe..0d863eb185f6 100644
> > --- a/Documentation/ABI/testing/sysfs-fs-f2fs
> > +++ b/Documentation/ABI/testing/sysfs-fs-f2fs
> > @@ -493,3 +493,17 @@ Contact: "Chao Yu" <yuchao0@xxxxxxxxxx>
> > Description: When ATGC is on, it controls age threshold to bypass GCing young
> > candidates whose age is not beyond the threshold, by default it was
> > initialized as 604800 seconds (equals to 7 days).
> > +
> > +What: /sys/fs/f2fs/<disk>/gc_dirty_segs
>
> How about renaming gc_dirty_segs to gc_reclaimed_segments?

Makes sense.

>
>
> > +Date: July 2021
> > +Contact: "Daeho Jeong" <daehojeong@xxxxxxxxxx>
> > +Description: Show how many dirty segments have been reclaimed by GC during
> > + a specific GC mode (0: GC normal, 1: GC idle CB, 2: GC idle greedy,
> > + 3: gc idle AT, 4: GC urgent high, 5: GC urgent low)
>
> 3. GC idle AT

Got it.

>
> > + You can re-initialize this value to "0".
> > +
> > +What: /sys/fs/f2fs/<disk>/gc_dirty_segs_mode
>
> You know, F2FS_IOC_FLUSH_DEVICE/F2FS_IOC_RESIZE_FS needs to migrate all
> blocks of target segment to other place, no matter the segment has partially
> or fully valid blocks, so it looks 'gc_segment_mode' will be more appropriate?
>
> Thanks,

Reasonable~ Got it.