Messages in this thread | | | From | OGAWA Hirofumi <> | Subject | Re: [PATCH RESEND v5] fat: editions to support fat_fallocate | Date | Mon, 29 Apr 2013 23:31:33 +0900 |
| |
Namjae Jeon <linkinjeon@gmail.com> writes:
I couldn't review fully though.
> + if (mmu_private_ideal < MSDOS_I(inode)->mmu_private && > + filp->f_dentry->d_count == 1) > + fat_truncate_blocks(inode, inode->i_size);
Hm, why d_count == 1 check is needed? Feel strange and racy.
> + /* Start the allocation.We are not zeroing out the clusters */ > + while (nr_cluster-- > 0) { > + err = fat_alloc_clusters(inode, &cluster, 1);
Why doesn't allocate clusters at once by fat_alloc_clusters()?
> + size = i_size_read(inode); > + mmu_private_actual = MSDOS_I(inode)->mmu_private; > + mmu_private_ideal = round_up(size, sb->s_blocksize); > + if ((mmu_private_actual > mmu_private_ideal) && (pos > size)) { > + err = fat_zero_falloc_area(file, mapping, pos); > + if (err) { > + fat_msg(sb, KERN_ERR, > + "Error (%d) zeroing fallocated area", err); > + return err; > + } > + }
This way probably inefficient. This would write data twice times (one is zeroed, one is actual data). So, cpu time would be twice higher if user uses fallocated, right?
Difference of fallocated area would be whether get_block() set buffer_new() or not? If true, we should change get_block(), not write_begin()?
Thanks. -- OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
| |