Messages in this thread |  | | Date | Tue, 2 Jan 2001 15:31:02 +0100 | From | Jens Axboe <> | Subject | Re: Chipsets, DVD-RAM, and timeouts.... |
| |
On Mon, Jan 01 2001, Alan Cox wrote: > > for FAT etc when reading. Writing is a bit more difficult, as that > > then turns out to generate a read before we can commit a dirty > > block. IMO, this type of thing does not belong in the drivers -- > > we should _never_ receive request for < hard block size. > > Unfortunately someone ripped the support out from 2.2 to do this, then didnt > fix it. So right now 2.4 is useless to anyone with an M/O drive.
It wasn't deliberately ripped out, it just vanished when Eric started his SCSI rewrite in 2.3 :-). For 2.5 I would really like the change mentioned earlier -- make sure that no block drivers should have to deal with any request for smaller sector sizes than their hw sector size. For 2.4, I guess I can apply much the same fix for MO as I did for CD-ROM.
-- * Jens Axboe <axboe@suse.de> * SuSE Labs - 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/
|  |