Re: 2.6.10-rc2-mm2 usb storage still oopses

[Pete Zaitcev <zaitcev@redhat.com>]

On Thu, 18 Nov 2004 13:35:57 -0800, Andrew Morton <akpm@osdl.org> wrote:

> Fabio Coatti <cova@ferrara.linux.it> wrote:
> >
> > Just a reminder: it's possible to cause a kernel oops simply inserting and 
> > removing a usb storage (flash pen); using ub driver doesn't improve the 
> > situation; noticed in 2.6.9-rc4-mm1 and present in 2.6.10-rc2-mm2.
> > The same device works just fine with 2.6.8.1 (mdk cooker)
> 
> OK, that's something we'd like to get fixed prior to 2.6.10.

Actually Fabio told me that his oops was fixed by the patch present in
2.6.10-rc2. The problem is that his device needs special handling which
I do not know how to provide, so it does not work in the end. I hope it
will resolve itself eventually, as I get testers.

There was one last oops from Martin Schleminger ("Sahara") which I think
I fixed but I need a confirmation before pushing to Greg. Apparently, it
only happens on kernels with preempt enabled. If anyone knows of any other
problems, I'm all ears.

-- Pete

P.S. Current updates:

--- linux-2.6.10-rc2-usb/drivers/block/ub.c	2004-11-16 17:03:02.000000000 -0800
+++ linux-2.6.10-rc1-ub/drivers/block/ub.c	2004-11-07 19:01:03.000000000 -0800
@@ -36,7 +36,7 @@
 #define DRV_NAME "ub"
 #define DEVFS_NAME DRV_NAME
 
-#define UB_MAJOR 125	/* Stolen from Experimental range for a week - XXX */
+#define UB_MAJOR 180
 
 /*
  * Definitions which have to be scattered once we understand the layout better.
@@ -1535,8 +1535,11 @@
 
 	ub_revalidate(sc);
 	/* This is pretty much a long term P3 */
-	printk(KERN_INFO "%s: device %u capacity nsec %ld bsize %u\n",
-	    sc->name, sc->dev->devnum, sc->capacity.nsec, sc->capacity.bsize);
+	if (!atomic_read(&sc->poison)) {		/* Cover sc->dev */
+		printk(KERN_INFO "%s: device %u capacity nsec %ld bsize %u\n",
+		    sc->name, sc->dev->devnum,
+		    sc->capacity.nsec, sc->capacity.bsize);
+	}
 
 	/* XXX Support sector size switching like in sr.c */
 	blk_queue_hardsect_size(disk->queue, sc->capacity.bsize);