Messages in this thread | | | Date | Wed, 15 Dec 1999 19:35:51 +0100 (CET) | From | Mike Galbraith <> | Subject | Re: raid problem |
| |
On Wed, 15 Dec 1999, Ketil Malde wrote:
> > Hi > > I have a RAID "array" consisting of two IDE disks, which I've been > running a striped partition across. When I upgraded from 2.3.26 to > 2.3.33 and added the necessary partitions (the definitions in the > mdtab had been there for a while) for another striped /dev/md1 and > a mirrored /dev/md2, the system refused to boot. It all ended with > the somewhat cryptic "Got md request, not good..." message from > block/drivers/md.c (I suppose), whereupon the system hangs.
Hi,
I ran into the same thing. I got it working, but :)) have no idea why it works. Don't do this unless you like to tinker and don't mind the occasional mushroom cloud.
What I did <cringe> is first, comment out the message. That is only important after the box actually mounts an fs, but then it gets VERY noisy. Secondly, I copied make_request() to a new function I called make_request_redirect() [for some pretty foggy reasons]. Thirdly, I changed md_make_request() to use make_request_redirect().. same in raid1.c. This works just fine, (tested raid0 and 1over0) but looks dainbramaged as all hell.
The only thing different in make_request_redirect() is the 'r'.. q = get_queue(bh->b_rdev);
I shouldn't post this, but I'm curious enough to take a toasting :)
-Mike
- 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/
| |