>On Wed, 12 May 1999, Andrea Arcangeli wrote:
>
>> (vma->vm_offset & (inode->i_sb->s_blocksize - 1)))
>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ but here?
[snip]
>
>Look at set_blocksize() and foo_read_super().
I looked them some time ago. If I remeber well s_blocksize should be the
soft-blocksize. It's the blocksize of the buffers relative to such device.
And according to me it's _far_ from be equal to PAGE_SIZE (it has to me
minor but not equal). If I remember well in my HD (as I think in most of
linux installations) s_blocksize is 1k, while the hardblocksize is 512
(while PAGE_SIZE is 4k).
Sure if you create an ext2 fs with -b 4096, then you will allowed to mmap
_only_ at PAGE_SIZE aligned file-offsets (on i386) but that's not the
point.
It make some sense to avoid overlapping page caches but I am not sure it's
a big performance issue (also considering that pre-2_2_8-x broke only
ZMAGIC binaries... ;).
Andrea Arcangeli
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/