block_truncate_page() change in 2.4.0-prerelease

Marcelo Tosatti (marcelo@conectiva.com.br)
Tue, 2 Jan 2001 21:13:52 -0200 (BRST)


Linus,

In 2.4.0-prerelease, you've changed block_truncate_page() to call
__mark_buffer_dirty instead of mark_buffer_dirty before unlocking the
page, so there is no more possibility of it blocking in bdflush while
holding the page locked, which is good.

But now we don't do balancing anymore there.

The patch I sent you to change mark_buffer_dirty() return the old value of
the dirty bit to know if we need to balance could fix that and with it we
can the ext2 superblock locking issue, which is horribly bad IMHO.

Do you want such a change now in 2.4.0-prerelease ?

If so, I can send a patch with all the mark_buffer_dirty and ext2 changes.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/