This patch (as1653) fixes a bug in ehci-hcd. Unlike iTD entries, an
siTD entry in the periodic schedule may not complete until the frame
after the one it belongs to. Consequently, when scanning the periodic
schedule it is necessary to start with the frame _preceding_ the one
where the previous scan ended.
Not doing this properly can result in memory leaks and failures to
complete isochronous URBs.
Signed-off-by: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>
Reported-by: Andy Leiserson <andy@xxxxxxxxxxxxx>
---
Greg:
This patch is for 3.8. The same problem exists in 3.6 and 3.7, but the
code has changed in the meantime so they need a different patch.
Should I submit that one after this has been merged?
Alan
drivers/usb/host/ehci-sched.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
Index: usb-3.8/drivers/usb/host/ehci-sched.c
===================================================================
--- usb-3.8.orig/drivers/usb/host/ehci-sched.c
+++ usb-3.8/drivers/usb/host/ehci-sched.c
@@ -2212,11 +2212,11 @@ static void scan_isoc(struct ehci_hcd *e
}
ehci->now_frame = now_frame;
+ frame = ehci->last_iso_frame;
for (;;) {
union ehci_shadow q, *q_p;
__hc32 type, *hw_p;
- frame = ehci->last_iso_frame;
restart:
/* scan each element in frame's queue for completions */
q_p = &ehci->pshadow [frame];
@@ -2321,6 +2321,9 @@ restart:
/* Stop when we have reached the current frame */
if (frame == now_frame)
break;
- ehci->last_iso_frame = (frame + 1) & fmask;
+
+ /* The last frame may still have active siTDs */
+ ehci->last_iso_frame = frame;
+ frame = (frame + 1) & fmask;
}
}
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
