[PATCH] usb-serial: fix regression in Visor/Palm OS module for kernels >= 2.6.24

[Brad Sawatzky <brad+debian@swatter.net>]

Fixes a bug/inconsistency revealed by the additional sanity checking in
   commit 063a2da8f01806906f7d7b1a1424b9afddebc443
introduced in the original 2.6.24 branch.

The Handspring Visor / PalmOS 4 device structure defines .num_bulk_out=2
but the usb-serial probe returns num_bulk_out=3, triggering the check in
the above commit and forcing a bail out when the device (a Garmin iQue in
my case) attempts to connect.  The patch bumps the expected number of
endpoints to 3.

I suppose it's possible that the kernel is identifying 3 bulk endpoints
when there should only be 2 and there is some issue with the lower level
endpoint probe?

FWIW, this patch will probably solve the following kernel bug report for
Treo users (identical symptoms, different model PalmOS units):
  <http://bugzilla.kernel.org/show_bug.cgi?id=10118>


Signed-off-by: Brad Sawatzky <brad+kernel@swatter.net>
---

--- drivers/usb/serial/visor.c.orig	2008-03-25 21:32:40.000000000 -0400
+++ drivers/usb/serial/visor.c	2008-03-25 21:29:57.000000000 -0400
@@ -191,7 +191,7 @@ static struct usb_serial_driver handspri
 	.id_table =		id_table,
 	.num_interrupt_in =	NUM_DONT_CARE,
 	.num_bulk_in =		2,
-	.num_bulk_out =		2,
+	.num_bulk_out =		3,
 	.num_ports =		2,
 	.open =			visor_open,
 	.close =		visor_close,