>code was simply ignoring inuse-buffers. Since I couldn't figure how the
>kernel could recall invalidate_buffers with a mounted fs I added the
I had a fast review in the last minutes and it seems to me that it can via
blckdev_open -> ide_open -> idedisk_open -> check if removable media
changed -> invalidate buffers
The problem/feature is that idedisk_media_change is lazy and return 1 if
drive->removable is set to 1 so all buffer gets invalidated.
drive->removable seems set to one because at probe time ide-probe does:
/*
* Not an ATAPI device: looks like a "regular" hard disk
*/
if (id->config & (1<<7))
drive->removable = 1;
It's not obvious that an regular hard disk has to be considered removable,
but I have no problem in not being "strict" :-).
It seems simply a safe feature but I write this just to make sure it's
well known :).
printk removed from my buffer.c. A new buffer patch will be out in some
minute in the kernel-patches directory.
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/