This patch (as1660) works around a hardware problem present in some (if not all) Intel EHCI controllers. After a QH has been unlinked from the async schedule and the corresponding IAA interrupt has occurred, the controller is not supposed access the QH and its qTDs. There certainly shouldn't be any more DMA writes to those structures. Nevertheless, Intel's controllers have been observed to perform a final writeback to the QH's overlay region and to the most recent qTD. For more information and a test program to determine whether this problem is present in a particular controller, see http://marc.info/?l=linux-usb&m=135492071812265&w=2 http://marc.info/?l=linux-usb&m=136182570800963&w=2 This patch works around the problem by always waiting for two IAA cycles when unlinking an async QH. The extra IAA delay gives the controller time to perform its final writeback. Surprisingly enough, the effects of this silicon bug have gone undetected until quite recently. More through luck than anything else, it hasn't caused any apparent problems. However, it does interact badly with the path that follows this one, so it needs to be addressed. This is the first part of a fix for the regression reported at: https://bugs.launchpad.net/bugs/1088733 Signed-off-by: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> Tested-by: Stephen Thirlwall <sdt@xxxxxx> CC: <stable@xxxxxxxxxxxxxxx> --- For all I know, the same bug may be present in controllers from other vendors too. Later on, if I learn that it's not, a quirk flag can be added to prevent the extra delay from being used where it isn't needed. drivers/usb/host/ehci-hcd.c | 6 ++---- drivers/usb/host/ehci-q.c | 18 ++++++++++++++---- 2 files changed, 16 insertions(+), 8 deletions(-) Index: usb-3.8/drivers/usb/host/ehci-q.c =================================================================== --- usb-3.8.orig/drivers/usb/host/ehci-q.c +++ usb-3.8/drivers/usb/host/ehci-q.c @@ -1170,7 +1170,7 @@ static void single_unlink_async(struct e struct ehci_qh *prev; /* Add to the end of the list of QHs waiting for the next IAAD */ - qh->qh_state = QH_STATE_UNLINK; + qh->qh_state = QH_STATE_UNLINK_WAIT; if (ehci->async_unlink) ehci->async_unlink_last->unlink_next = qh; else @@ -1213,9 +1213,19 @@ static void start_iaa_cycle(struct ehci_ /* Do only the first waiting QH (nVidia bug?) */ qh = ehci->async_unlink; - ehci->async_iaa = qh; - ehci->async_unlink = qh->unlink_next; - qh->unlink_next = NULL; + + /* + * Intel (?) bug: The HC can write back the overlay region + * even after the IAA interrupt occurs. In self-defense, + * always go through two IAA cycles for each QH. + */ + if (qh->qh_state == QH_STATE_UNLINK_WAIT) { + qh->qh_state = QH_STATE_UNLINK; + } else { + ehci->async_iaa = qh; + ehci->async_unlink = qh->unlink_next; + qh->unlink_next = NULL; + } /* Make sure the unlinks are all visible to the hardware */ wmb(); Index: usb-3.8/drivers/usb/host/ehci-hcd.c =================================================================== --- usb-3.8.orig/drivers/usb/host/ehci-hcd.c +++ usb-3.8/drivers/usb/host/ehci-hcd.c @@ -748,11 +748,9 @@ static irqreturn_t ehci_irq (struct usb_ /* guard against (alleged) silicon errata */ if (cmd & CMD_IAAD) ehci_dbg(ehci, "IAA with IAAD still set?\n"); - if (ehci->async_iaa) { + if (ehci->async_iaa) COUNT(ehci->stats.iaa); - end_unlink_async(ehci); - } else - ehci_dbg(ehci, "IAA with nothing unlinked?\n"); + end_unlink_async(ehci); } /* remote wakeup [4.3.1] */ -- 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
