Re: [f2fs-dev] [PATCH] f2fs: change the way of handling range.len in F2FS_IOC_SEC_TRIM_FILE

[Daeho Jeong]

To handle that case, I think we need to handle range.len(-1) differently.
When range.len is -1, we need to find out every block belongs to the
inode regardless of i_size and discard it.

2020ë 7ì 10ì (ê) ìí 12:52, Jaegeuk Kim <jaegeuk@xxxxxxxxxx>ëì ìì:
>
> On 07/10, Chao Yu wrote:
> > On 2020/7/10 11:31, Jaegeuk Kim wrote:
> > > On 07/10, Chao Yu wrote:
> > >> On 2020/7/10 11:02, Jaegeuk Kim wrote:
> > >>> On 07/10, Daeho Jeong wrote:
> > >>>> From: Daeho Jeong <daehojeong@xxxxxxxxxx>
> > >>>>
> > >>>> Changed the way of handling range.len of F2FS_IOC_SEC_TRIM_FILE.
> > >>>>  1. Added -1 value support for range.len to signify the end of file.
> > >>>>  2. If the end of the range passes over the end of file, it means until
> > >>>>     the end of file.
> > >>>>  3. ignored the case of that range.len is zero to prevent the function
> > >>>>     from making end_addr zero and triggering different behaviour of
> > >>>>     the function.
> > >>>>
> > >>>> Signed-off-by: Daeho Jeong <daehojeong@xxxxxxxxxx>
> > >>>> ---
> > >>>>  fs/f2fs/file.c | 16 +++++++---------
> > >>>>  1 file changed, 7 insertions(+), 9 deletions(-)
> > >>>>
> > >>>> diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
> > >>>> index 368c80f8e2a1..1c4601f99326 100644
> > >>>> --- a/fs/f2fs/file.c
> > >>>> +++ b/fs/f2fs/file.c
> > >>>> @@ -3813,21 +3813,19 @@ static int f2fs_sec_trim_file(struct file *filp, unsigned long arg)
> > >>>>          file_start_write(filp);
> > >>>>          inode_lock(inode);
> > >>>>
> > >>>> -        if (f2fs_is_atomic_file(inode) || f2fs_compressed_file(inode)) {
> > >>>> +        if (f2fs_is_atomic_file(inode) || f2fs_compressed_file(inode) ||
> > >>>> +                        range.start >= inode->i_size) {
> > >>>>                  ret = -EINVAL;
> > >>>>                  goto err;
> > >>>>          }
> > >>>>
> > >>>> -        if (range.start >= inode->i_size) {
> > >>>> -                ret = -EINVAL;
> > >>>> +        if (range.len == 0)
> > >>>>                  goto err;
> > >>>> -        }
> > >>>>
> > >>>> -        if (inode->i_size - range.start < range.len) {
> > >>>> -                ret = -E2BIG;
> > >>>> -                goto err;
> > >>>> -        }
> > >>>> -        end_addr = range.start + range.len;
> > >>>> +        if (range.len == (u64)-1 || inode->i_size - range.start < range.len)
> > >>>> +                end_addr = inode->i_size;
> > >>
> > >> We can remove 'range.len == (u64)-1' condition since later condition can cover
> > >> this?
> > >>
> > >>>
> > >>> Hmm, what if there are blocks beyond i_size? Do we need to check i_blocks for
> > >>
> > >> The blocks beyond i_size will never be written, there won't be any valid message
> > >> there, so we don't need to worry about that.
> > >
> > > I don't think we have a way to guarantee the order of i_size and block
> > > allocation in f2fs. See f2fs_write_begin and f2fs_write_end.
> >
> > However, write_begin & write_end are covered by inode_lock, it could not be
> > racy with inode size check in f2fs_sec_trim_file() as it hold inode_lock as
> > well?
>
> Like Daeho said, write_begin -> checkpoint -> power-cut can give bigger i_blocks
> than i_size.
>
> >
> > >
> > >>
> > >> Thanks,
> > >>
> > >>> ending criteria?
> > >>>
> > >>>> +        else
> > >>>> +                end_addr = range.start + range.len;
> > >>>>
> > >>>>          to_end = (end_addr == inode->i_size);
> > >>>>          if (!IS_ALIGNED(range.start, F2FS_BLKSIZE) ||
> > >>>> --
> > >>>> 2.27.0.383.g050319c2ae-goog
> > >>>>
> > >>>>
> > >>>>
> > >>>> _______________________________________________
> > >>>> Linux-f2fs-devel mailing list
> > >>>> Linux-f2fs-devel@xxxxxxxxxxxxxxxxxxxxx
> > >>>> https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
> > >>>
> > >>>
> > >>> _______________________________________________
> > >>> Linux-f2fs-devel mailing list
> > >>> Linux-f2fs-devel@xxxxxxxxxxxxxxxxxxxxx
> > >>> https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
> > >>> .
> > >>>
> > > .
> > >