James Bottomley wrote:
> Actually, the patch is wrong. It will wreak havoc with SCSI's use of sysfs.
> The device_register has to be done in scsi_add_host, which is called after all
> the driver specific sysfs setup has been done. The correct fix is to move the
> corresponding device_unregister into scsi_remove_host so that they match.
>
> I've attached it below. I'll also commit it to the scsi-misc-2.5 BK tree.
>
> James
>
>
>
> ------------------------------------------------------------------------
>
> ===== hosts.c 1.31 vs edited =====
> --- 1.31/drivers/scsi/hosts.c Sun Nov 17 15:47:02 2002
> +++ edited/hosts.c Sat Nov 23 17:25:57 2002
> @@ -295,6 +295,8 @@
> kfree(sdev);
> }
>
> + device_unregister(&shost->host_driverfs_dev);
> +
> return 0;
> }
>
> @@ -348,7 +350,6 @@
>
> /* Cleanup proc and driverfs */
> scsi_proc_host_rm(shost);
> - device_unregister(&shost->host_driverfs_dev);
>
> kfree(shost);
> }
James,
The above patch is a bit noisy when applied to lk 2.5.50 .
Attached is the same patch (I hope) which applies cleanly.
Doug Gilbert
--------------090009020605010109080903
Content-Type: text/plain;
name="dev_reg_2550.diff"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
filename="dev_reg_2550.diff"
--- linux/drivers/scsi/hosts.c 2002-11-29 09:27:35.000000000 +1100
+++ linux/drivers/scsi/hosts.c2550jb 2002-11-29 09:45:28.000000000 +1100
@@ -297,6 +297,8 @@
scsi_free_sdev(list_entry(le, Scsi_Device, siblings));
}
+ device_unregister(&shost->host_driverfs_dev);
+
return 0;
}
@@ -348,7 +350,6 @@
/* Cleanup proc and driverfs */
scsi_proc_host_rm(shost);
- device_unregister(&shost->host_driverfs_dev);
kfree(shost);
}
--------------090009020605010109080903--
-
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/