summaryrefslogtreecommitdiff
path: root/target/linux/brcm63xx/patches-3.13/204-USB-OHCI-allow-other-arches-to-use-the-BE-frame-numb.patch
blob: 1f175f4797d1b05e02ced8233dae59fb526795ac (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
From b2f399dcd674a692a64bb3b300b77b78ae57b530 Mon Sep 17 00:00:00 2001
From: Jonas Gorski <jogo@openwrt.org>
Date: Sun, 12 Jan 2014 16:47:35 +0100
Subject: [PATCH] USB: OHCI: allow other arches to use the BE frame number
 quirk

Intead of guarding it with a certain PPC SoC and expanding the list
for each SoC requiring it, just guard it with USB_OHCI_BIG_ENDIAN_DESC.

This makes it less suprising that passing no_big_frame_no = 1 for the
platform data does not do what expected (or

Checking it for all big endian descriptor setups should not impact
performance much as USB1.1 is rather slow anyway.

Signed-off-by: Jonas Gorski <jogo@openwrt.org>
---
 drivers/usb/host/ohci.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/drivers/usb/host/ohci.h
+++ b/drivers/usb/host/ohci.h
@@ -650,7 +650,7 @@ static inline u32 hc32_to_cpup (const st
  * some big-endian SOC implementations.  Same thing happens with PSW access.
  */
 
-#ifdef CONFIG_PPC_MPC52xx
+#ifdef CONFIG_USB_OHCI_BIG_ENDIAN_DESC
 #define big_endian_frame_no_quirk(ohci)	(ohci->flags & OHCI_QUIRK_FRAME_NO)
 #else
 #define big_endian_frame_no_quirk(ohci)	0