[PATCH] initrd fails again in 2.5.18

James Bottomley (James.Bottomley@steeleye.com)
Mon, 27 May 2002 18:56:27 -0400


This is a multipart MIME message.

--==_Exmh_18786864270
Content-Type: text/plain; charset=us-ascii

The initial ramdisk fails to work with a slew of errors like:

generic_make_request: Trying to access nonexistent block-device 01:00 (2)

The problem was caused by change

kdev_t -> bdev cleanups [2/2]

which took out the ability of bdev_get_queue() to create a queue if one didn't
already exist for the device. The functionality was moved to
block_dev.c:do_open() where it exists within the if(!bdev->bd_openers) which
the ramdisk never gets to.

The (tested) fix, I think, is to set bd_queue as part of ramdisk
initialisation, which is what the attached patch does.

James Bottomley

--==_Exmh_18786864270
Content-Type: text/plain ; name="rd-2.5.18.diff"; charset=us-ascii
Content-Description: rd-2.5.18.diff
Content-Disposition: attachment; filename="rd-2.5.18.diff"

===== drivers/block/rd.c 1.38 vs edited =====
--- 1.38/drivers/block/rd.c Thu May 23 08:18:38 2002
+++ edited/drivers/block/rd.c Mon May 27 17:11:40 2002
@@ -379,6 +379,7 @@
rd_bdev[unit]->bd_openers++;
rd_bdev[unit]->bd_block_size = rd_blocksize;
rd_bdev[unit]->bd_inode->i_mapping->a_ops = &ramdisk_aops;
+ rd_bdev[unit]->bd_queue = BLK_DEFAULT_QUEUE(MAJOR_NR);
}

return 0;

--==_Exmh_18786864270--

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