summaryrefslogtreecommitdiffstats
path: root/debian/patches-rt/0035-usb-do-no-disable-interrupts-in-giveback.patch
blob: 799437d213bacac232650853fbb604de284a1311 (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
32
33
34
35
36
37
38
39
40
41
42
From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Date: Fri, 8 Nov 2013 17:34:54 +0100
Subject: [PATCH 035/353] usb: do no disable interrupts in giveback
Origin: https://git.kernel.org/cgit/linux/kernel/git/rt/linux-stable-rt.git/commit?id=f847ddd27f34eff4e942a9f35a4b254bb16abed8

Since commit 94dfd7ed ("USB: HCD: support giveback of URB in tasklet
context") the USB code disables interrupts before invoking the complete
callback.
This should not be required the HCD completes the URBs either in hard-irq
context or in BH context. Lockdep may report false positives if one has two
HCDs (one completes in IRQ and the other in BH context) and is using the same
USB driver (device) with both HCDs. This is safe since the same URBs are never
mixed with those two HCDs.
Longeterm we should force all HCDs to complete in the same context.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
 drivers/usb/core/hcd.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c
index 59d5d506d73c..79133ba4b9e0 100644
--- a/drivers/usb/core/hcd.c
+++ b/drivers/usb/core/hcd.c
@@ -1752,7 +1752,6 @@ static void __usb_hcd_giveback_urb(struct urb *urb)
 	struct usb_hcd *hcd = bus_to_hcd(urb->dev->bus);
 	struct usb_anchor *anchor = urb->anchor;
 	int status = urb->unlinked;
-	unsigned long flags;
 
 	urb->hcpriv = NULL;
 	if (unlikely((urb->transfer_flags & URB_SHORT_NOT_OK) &&
@@ -1780,9 +1779,7 @@ static void __usb_hcd_giveback_urb(struct urb *urb)
 	 * and no one may trigger the above deadlock situation when
 	 * running complete() in tasklet.
 	 */
-	local_irq_save(flags);
 	urb->complete(urb);
-	local_irq_restore(flags);
 
 	usb_anchor_resume_wakeups(anchor);
 	atomic_dec(&urb->use_count);