RE: [PATCH] f2fs: don't break allocation when crossing contiguous sections

[Daejun Park]

Hi Chao Yu,

> +        lfs_dio_write = (flag == F2FS_GET_BLOCK_DIO && f2fs_lfs_mode(sbi) &&
> +                                map->m_may_create);
> +

Consider moving the lfs_dio_write assignment to just after the if (!map->m_may_create …) check
so it isn’t evaluated when creation isn’t allowed.

> @@ -4171,6 +4172,10 @@ static int f2fs_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
>          map.m_next_pgofs = &next_pgofs;
>          map.m_seg_type = f2fs_rw_hint_to_seg_type(F2FS_I_SB(inode),
>                                                  inode->i_write_hint);
> +        if (flags & IOMAP_WRITE && iomap->private) {

Since iomap->private is only set on the LFS DIO path, you can drop the flags & IOMAP_WRITE and
test the pointer directly.

> +                map.m_last_pblk = (unsigned long)iomap->private;
> +                iomap->private = NULL;
> +        }
> 
>          /*
>           * If the blocks being overwritten are already allocated,
> @@ -4209,6 +4214,9 @@ static int f2fs_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
>                  iomap->flags = IOMAP_F_MERGED;
>                  iomap->bdev = map.m_bdev;
>                  iomap->addr = F2FS_BLK_TO_BYTES(map.m_pblk);
> +
> +                if (flags & IOMAP_WRITE && map.m_last_pblk)
> +                        iomap->private = (void *)map.m_last_pblk;

Likewise, checking only map.m_last_pblk is sufficient for restoring.

Reviewed-by: Daejun Park <daejun7.park@xxxxxxxxxxx>