>>>>> "daniel" == Daniel Phillips <[EMAIL PROTECTED]>
>writes:
Hi
daniel> I tried it and you are right. The functionality I had to reproduce
daniel> was just:
daniel> if (err = <my prepare write>) goto err_out;
daniel> memset(page_address(page) + offset, 0, PAGE_CACHE_SIZE - offset);
daniel> flush_dcache_page(page);
daniel> <my commit write>
daniel> kunmap(page);
daniel> And the only mysterious are flush_dcache_page and kunmap, both of
daniel> which are macros defined for i386 as:
daniel> do { } while (0)
kmap is definied if you are using HighMemory, and yes,
flush_dcache_page does nothing in i386, see its meaning in
(linux/Documentation/cachetlb.txt).
daniel> A comment on create_empty_buffers: it takes inode as a parameter and
daniel> only uses that to set the b_dev. Shouldn't it just take dev as a
daniel> parameter instead of inode?
talking about that, all the calls to create_emty_buffers call it with
blocksize = inode->i_sb->s_blocksize, if we are going to do the
cleanup, it is better to do it well :))))
Later, Juan.
--
In theory, practice and theory are the same, but in practice they
are different -- Larry McVoy
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [EMAIL PROTECTED]
