This patch is originally from Alan Stern (as557). It has been re-diffed against a current tree, and I also corrected a minor merging error.
Some time ago we introduced a delay before device scanning, because many
devices do not like to receive SCSI commands right after enumeration.
Now it turns out there's a device that doesn't like to receive Get-Max-LUN
right after enumeration either. Accordingly this patch delays the
Get-Max-LUN request until the beginning of the scanning procedure. This
fixes Bugzilla entry #5010.
Three things are worth noting. First, I removed the locking code from
usb_stor_acquire_resources. It's not needed, because the locking is to
protect against disconnect events and acquire_resources is only called
during probe (so the disconnect routine can't be called). Second, I
initialized to 0 the buffer used for the Get-Max-LUN response. It's not
really necessary, but it will prevent random values from showing up in the
debugging log when the request fails. Third, I added a test against the
SINGLE_LUN flag. This will allow us to use the flag to indicate Bulk-only
devices that can't handle Get-Max-LUN.
Signed-off-by: Alan Stern <[EMAIL PROTECTED]>
Signed-off-by: Matthew Dharm <[EMAIL PROTECTED]>
Index: usb-2.6/drivers/usb/storage/transport.c
===================================================================
--- usb-2.6.orig/drivers/usb/storage/transport.c
+++ usb-2.6/drivers/usb/storage/transport.c
@@ -923,6 +923,7 @@ int usb_stor_Bulk_max_lun(struct us_data
int result;
/* issue the command */
+ us->iobuf[0] = 0;
result = usb_stor_control_msg(us, us->recv_ctrl_pipe,
US_BULK_GET_MAX_LUN,
USB_DIR_IN | USB_TYPE_CLASS |
Index: usb-2.6/drivers/usb/storage/usb.c
===================================================================
--- usb-2.6.orig/drivers/usb/storage/usb.c
+++ usb-2.6/drivers/usb/storage/usb.c
@@ -747,25 +747,13 @@
return -ENOMEM;
}
- /* Lock the device while we carry out the next two operations */
- down(&us->dev_semaphore);
-
- /* For bulk-only devices, determine the max LUN value */
- if (us->protocol == US_PR_BULK) {
- p = usb_stor_Bulk_max_lun(us);
- if (p < 0) {
- up(&us->dev_semaphore);
- return p;
- }
- us->max_lun = p;
- }
-
/* Just before we start our control thread, initialize
* the device if it needs initialization */
- if (us->unusual_dev->initFunction)
- us->unusual_dev->initFunction(us);
-
- up(&us->dev_semaphore);
+ if (us->unusual_dev->initFunction) {
+ p = us->unusual_dev->initFunction(us);
+ if (p)
+ return p;
+ }
/* Start up our control thread */
p = kernel_thread(usb_stor_control_thread, us, CLONE_VM);
@@ -909,6 +894,14 @@ retry:
/* If the device is still connected, perform the scanning */
if (!test_bit(US_FLIDX_DISCONNECTING, &us->flags)) {
+
+ /* For bulk-only devices, determine the max LUN value */
+ if (us->protocol == US_PR_BULK &&
+ !(us->flags & US_FL_SINGLE_LUN)) {
+ down(&us->dev_semaphore);
+ us->max_lun = usb_stor_Bulk_max_lun(us);
+ up(&us->dev_semaphore);
+ }
scsi_scan_host(us_to_host(us));
printk(KERN_DEBUG "usb-storage: device scan complete\n");
--
Matthew Dharm Home: [EMAIL PROTECTED]
Maintainer, Linux USB Mass Storage Driver
Stef, you just got beaten by a ball of DIRT.
-- Greg
User Friendly, 12/7/1997
pgp8RLFRmLDTQ.pgp
Description: PGP signature
