aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorMing Lei <ming.lei@canonical.com>2011-08-31 00:03:13 +0800
committerAndrey Konovalov <andrey.konovalov@linaro.org>2012-02-14 23:03:46 +0400
commit7f2be4a1c8181de24da244673b16ee98926b4a7e (patch)
treeecbacdafd369e8e1869aa8fdf876c70e8d0812a4
parentd65b4e98d7ea3038b767b70fe8be959b2913f16d (diff)
downloadvexpress-lsk-7f2be4a1c8181de24da244673b16ee98926b4a7e.tar.gz
usb: ehci: make HC see up-to-date qh/qtd descriptor ASAP
This patch introduces the helper of ehci_sync_mem to flush qtd/qh into memory immediately on some ARM, so that HC can see the up-to-date qtd/qh descriptor asap. This patch fixs one performance bug on ARM Cortex A9 dual core platform, which has been reported on quite a few ARM machines (OMAP4, Tegra 2, snowball...), see details from link of https://bugs.launchpad.net/bugs/709245. The patch has been tested ok on OMAP4 panda A1 board, and the performance of 'dd' over usb mass storage can be increased from 4~5MB/sec to 14~16MB/sec after applying this patch. Cc: Russell King <linux@arm.linux.org.uk> Signed-off-by: Ming Lei <ming.lei@canonical.com> Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
-rw-r--r--drivers/usb/host/ehci-q.c18
-rw-r--r--drivers/usb/host/ehci.h17
2 files changed, 35 insertions, 0 deletions
diff --git a/drivers/usb/host/ehci-q.c b/drivers/usb/host/ehci-q.c
index 36ca5077cdf..7151535e887 100644
--- a/drivers/usb/host/ehci-q.c
+++ b/drivers/usb/host/ehci-q.c
@@ -1004,6 +1004,12 @@ static void qh_link_async (struct ehci_hcd *ehci, struct ehci_qh *qh)
head->qh_next.qh = qh;
head->hw->hw_next = dma;
+ /*
+ * flush qh descriptor into memory immediately,
+ * see comments in qh_append_tds.
+ */
+ ehci_sync_mem();
+
qh_get(qh);
qh->xacterrs = 0;
qh->qh_state = QH_STATE_LINKED;
@@ -1091,6 +1097,18 @@ static struct ehci_qh *qh_append_tds (
wmb ();
dummy->hw_token = token;
+ /*
+ * Writing to dma coherent buffer on ARM may
+ * be delayed to reach memory, so HC may not see
+ * hw_token of dummy qtd in time, which can cause
+ * the qtd transaction to be executed very late,
+ * and degrade performance a lot. ehci_sync_mem
+ * is added to flush 'token' immediatelly into
+ * memory, so that ehci can execute the transaction
+ * ASAP.
+ */
+ ehci_sync_mem();
+
urb->hcpriv = qh_get (qh);
}
}
diff --git a/drivers/usb/host/ehci.h b/drivers/usb/host/ehci.h
index 0a5fda73b3f..01cefde1d49 100644
--- a/drivers/usb/host/ehci.h
+++ b/drivers/usb/host/ehci.h
@@ -746,6 +746,23 @@ static inline u32 hc32_to_cpup (const struct ehci_hcd *ehci, const __hc32 *x)
#endif
+/*
+ * Writing to dma coherent memory on ARM may be delayed via L2
+ * writing buffer, so introduce the helper which can flush L2 writing
+ * buffer into memory immediately, especially used to flush ehci
+ * descriptor to memory.
+ */
+#ifdef CONFIG_ARM_DMA_MEM_BUFFERABLE
+static inline void ehci_sync_mem()
+{
+ mb();
+}
+#else
+static inline void ehci_sync_mem()
+{
+}
+#endif
+
/*-------------------------------------------------------------------------*/
#ifdef CONFIG_PCI