USB: EHCI support for big-endian descriptors

This patch implements supports for EHCI controllers whose in-memory
data structures are represented in big-endian format. This is needed
(unfortunately) for the AMCC PPC440EPx SoC EHCI controller; the EHCI
spec doesn't specify little-endian format, although that's what most
other implementations use.

The guts of the patch are to introduce the hc32 type and change all
references from le32 to hc32.  All access routines are converted from
cpu_to_le32(...) to cpu_to_hc32(ehci, ...) and similar for the other
"direction".  (This is the same approach used with OHCI.)

David fixed:
	Whitespace fixes; refresh against ehci cpufreq patch; move glue
	for that PPC driver to the patch adding it; fix free symbol
	capture bugs in modified "constant" macros; and make "hc32" etc
	be "le32" unless we really need the BE options, so "sparse" can
	do some real good.

Signed-off-by: Stefan Roese <sr@denx.de>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

1 files changed, 47 insertions, 45 deletions

diff --git a/drivers/usb/host/ehci-q.c b/drivers/usb/host/ehci-q.c
index 903510b..2284028 100644
--- a/drivers/usb/host/ehci-q.c
+++ b/drivers/usb/host/ehci-q.c
@@ -43,15 +43,15 @@
 /* fill a qtd, returning how much of the buffer we were able to queue up */
 
 static int
-qtd_fill (struct ehci_qtd *qtd, dma_addr_t buf, size_t len,
-		int token, int maxpacket)
+qtd_fill(struct ehci_hcd *ehci, struct ehci_qtd *qtd, dma_addr_t buf,
+		  size_t len, int token, int maxpacket)
 {
 	int	i, count;
 	u64	addr = buf;
 
 	/* one buffer entry per 4K ... first might be short or unaligned */
-	qtd->hw_buf [0] = cpu_to_le32 ((u32)addr);
-	qtd->hw_buf_hi [0] = cpu_to_le32 ((u32)(addr >> 32));
+	qtd->hw_buf[0] = cpu_to_hc32(ehci, (u32)addr);
+	qtd->hw_buf_hi[0] = cpu_to_hc32(ehci, (u32)(addr >> 32));
 	count = 0x1000 - (buf & 0x0fff);	/* rest of that page */
 	if (likely (len < count))		/* ... iff needed */
 		count = len;
@@ -62,8 +62,9 @@ qtd_fill (struct ehci_qtd *qtd, dma_addr_t buf, size_t len,
 		/* per-qtd limit: from 16K to 20K (best alignment) */
 		for (i = 1; count < len && i < 5; i++) {
 			addr = buf;
-			qtd->hw_buf [i] = cpu_to_le32 ((u32)addr);
-			qtd->hw_buf_hi [i] = cpu_to_le32 ((u32)(addr >> 32));
+			qtd->hw_buf[i] = cpu_to_hc32(ehci, (u32)addr);
+			qtd->hw_buf_hi[i] = cpu_to_hc32(ehci,
+					(u32)(addr >> 32));
 			buf += 0x1000;
 			if ((count + 0x1000) < len)
 				count += 0x1000;
@@ -75,7 +76,7 @@ qtd_fill (struct ehci_qtd *qtd, dma_addr_t buf, size_t len,
 		if (count != len)
 			count -= (count % maxpacket);
 	}
-	qtd->hw_token = cpu_to_le32 ((count << 16) | token);
+	qtd->hw_token = cpu_to_hc32(ehci, (count << 16) | token);
 	qtd->length = count;
 
 	return count;
@@ -89,28 +90,28 @@ qh_update (struct ehci_hcd *ehci, struct ehci_qh *qh, struct ehci_qtd *qtd)
 	/* writes to an active overlay are unsafe */
 	BUG_ON(qh->qh_state != QH_STATE_IDLE);
 
-	qh->hw_qtd_next = QTD_NEXT (qtd->qtd_dma);
-	qh->hw_alt_next = EHCI_LIST_END;
+	qh->hw_qtd_next = QTD_NEXT(ehci, qtd->qtd_dma);
+	qh->hw_alt_next = EHCI_LIST_END(ehci);
 
 	/* Except for control endpoints, we make hardware maintain data
 	 * toggle (like OHCI) ... here (re)initialize the toggle in the QH,
 	 * and set the pseudo-toggle in udev. Only usb_clear_halt() will
 	 * ever clear it.
 	 */
-	if (!(qh->hw_info1 & cpu_to_le32(1 << 14))) {
+	if (!(qh->hw_info1 & cpu_to_hc32(ehci, 1 << 14))) {
 		unsigned	is_out, epnum;
 
-		is_out = !(qtd->hw_token & cpu_to_le32(1 << 8));
-		epnum = (le32_to_cpup(&qh->hw_info1) >> 8) & 0x0f;
+		is_out = !(qtd->hw_token & cpu_to_hc32(ehci, 1 << 8));
+		epnum = (hc32_to_cpup(ehci, &qh->hw_info1) >> 8) & 0x0f;
 		if (unlikely (!usb_gettoggle (qh->dev, epnum, is_out))) {
-			qh->hw_token &= ~__constant_cpu_to_le32 (QTD_TOGGLE);
+			qh->hw_token &= ~cpu_to_hc32(ehci, QTD_TOGGLE);
 			usb_settoggle (qh->dev, epnum, is_out, 1);
 		}
 	}
 
 	/* HC must see latest qtd and qh data before we clear ACTIVE+HALT */
 	wmb ();
-	qh->hw_token &= __constant_cpu_to_le32 (QTD_TOGGLE | QTD_STS_PING);
+	qh->hw_token &= cpu_to_hc32(ehci, QTD_TOGGLE | QTD_STS_PING);
 }
 
 /* if it weren't for a common silicon quirk (writing the dummy into the qh
@@ -128,7 +129,7 @@ qh_refresh (struct ehci_hcd *ehci, struct ehci_qh *qh)
 		qtd = list_entry (qh->qtd_list.next,
 				struct ehci_qtd, qtd_list);
 		/* first qtd may already be partially processed */
-		if (cpu_to_le32 (qtd->qtd_dma) == qh->hw_current)
+		if (cpu_to_hc32(ehci, qtd->qtd_dma) == qh->hw_current)
 			qtd = NULL;
 	}
 
@@ -222,7 +223,7 @@ __acquires(ehci->lock)
 		struct ehci_qh	*qh = (struct ehci_qh *) urb->hcpriv;
 
 		/* S-mask in a QH means it's an interrupt urb */
-		if ((qh->hw_info2 & __constant_cpu_to_le32 (QH_SMASK)) != 0) {
+		if ((qh->hw_info2 & cpu_to_hc32(ehci, QH_SMASK)) != 0) {
 
 			/* ... update hc-wide periodic stats (for usbfs) */
 			ehci_to_hcd(ehci)->self.bandwidth_int_reqs--;
@@ -277,7 +278,6 @@ static int qh_schedule (struct ehci_hcd *ehci, struct ehci_qh *qh);
  * Chases up to qh->hw_current.  Returns number of completions called,
  * indicating how much "real" work we did.
  */
-#define HALT_BIT __constant_cpu_to_le32(QTD_STS_HALT)
 static unsigned
 qh_completions (struct ehci_hcd *ehci, struct ehci_qh *qh)
 {
@@ -287,6 +287,7 @@ qh_completions (struct ehci_hcd *ehci, struct ehci_qh *qh)
 	unsigned		count = 0;
 	int			do_status = 0;
 	u8			state;
+	u32			halt = HALT_BIT(ehci);
 
 	if (unlikely (list_empty (&qh->qtd_list)))
 		return count;
@@ -334,7 +335,7 @@ qh_completions (struct ehci_hcd *ehci, struct ehci_qh *qh)
 
 		/* hardware copies qtd out of qh overlay */
 		rmb ();
-		token = le32_to_cpu (qtd->hw_token);
+		token = hc32_to_cpu(ehci, qtd->hw_token);
 
 		/* always clean up qtds the hc de-activated */
 		if ((token & QTD_STS_ACTIVE) == 0) {
@@ -346,7 +347,8 @@ qh_completions (struct ehci_hcd *ehci, struct ehci_qh *qh)
 			 * that silicon quirk can kick in with this dummy too.
 			 */
 			} else if (IS_SHORT_READ (token)
-					&& !(qtd->hw_alt_next & EHCI_LIST_END)) {
+					&& !(qtd->hw_alt_next
+						& EHCI_LIST_END(ehci))) {
 				stopped = 1;
 				goto halt;
 			}
@@ -378,17 +380,17 @@ qh_completions (struct ehci_hcd *ehci, struct ehci_qh *qh)
 
 			/* token in overlay may be most current */
 			if (state == QH_STATE_IDLE
-					&& cpu_to_le32 (qtd->qtd_dma)
+					&& cpu_to_hc32(ehci, qtd->qtd_dma)
 						== qh->hw_current)
-				token = le32_to_cpu (qh->hw_token);
+				token = hc32_to_cpu(ehci, qh->hw_token);
 
 			/* force halt for unlinked or blocked qh, so we'll
 			 * patch the qh later and so that completions can't
 			 * activate it while we "know" it's stopped.
 			 */
-			if ((HALT_BIT & qh->hw_token) == 0) {
+			if ((halt & qh->hw_token) == 0) {
 halt:
-				qh->hw_token |= HALT_BIT;
+				qh->hw_token |= halt;
 				wmb ();
 			}
 		}
@@ -423,7 +425,7 @@ halt:
 	 * it after fault cleanup, or recovering from silicon wrongly
 	 * overlaying the dummy qtd (which reduces DMA chatter).
 	 */
-	if (stopped != 0 || qh->hw_qtd_next == EHCI_LIST_END) {
+	if (stopped != 0 || qh->hw_qtd_next == EHCI_LIST_END(ehci)) {
 		switch (state) {
 		case QH_STATE_IDLE:
 			qh_refresh(ehci, qh);
@@ -432,7 +434,7 @@ halt:
 			/* should be rare for periodic transfers,
 			 * except maybe high bandwidth ...
 			 */
-			if ((__constant_cpu_to_le32 (QH_SMASK)
+			if ((cpu_to_hc32(ehci, QH_SMASK)
 					& qh->hw_info2) != 0) {
 				intr_deschedule (ehci, qh);
 				(void) qh_schedule (ehci, qh);
@@ -506,8 +508,9 @@ qh_urb_transaction (
 	is_input = usb_pipein (urb->pipe);
 	if (usb_pipecontrol (urb->pipe)) {
 		/* SETUP pid */
-		qtd_fill (qtd, urb->setup_dma, sizeof (struct usb_ctrlrequest),
-			token | (2 /* "setup" */ << 8), 8);
+		qtd_fill(ehci, qtd, urb->setup_dma,
+				sizeof (struct usb_ctrlrequest),
+				token | (2 /* "setup" */ << 8), 8);
 
 		/* ... and always at least one more pid */
 		token ^= QTD_TOGGLE;
@@ -516,7 +519,7 @@ qh_urb_transaction (
 		if (unlikely (!qtd))
 			goto cleanup;
 		qtd->urb = urb;
-		qtd_prev->hw_next = QTD_NEXT (qtd->qtd_dma);
+		qtd_prev->hw_next = QTD_NEXT(ehci, qtd->qtd_dma);
 		list_add_tail (&qtd->qtd_list, head);
 
 		/* for zero length DATA stages, STATUS is always IN */
@@ -543,7 +546,7 @@ qh_urb_transaction (
 	for (;;) {
 		int this_qtd_len;
 
-		this_qtd_len = qtd_fill (qtd, buf, len, token, maxpacket);
+		this_qtd_len = qtd_fill(ehci, qtd, buf, len, token, maxpacket);
 		len -= this_qtd_len;
 		buf += this_qtd_len;
 		if (is_input)
@@ -561,7 +564,7 @@ qh_urb_transaction (
 		if (unlikely (!qtd))
 			goto cleanup;
 		qtd->urb = urb;
-		qtd_prev->hw_next = QTD_NEXT (qtd->qtd_dma);
+		qtd_prev->hw_next = QTD_NEXT(ehci, qtd->qtd_dma);
 		list_add_tail (&qtd->qtd_list, head);
 	}
 
@@ -570,7 +573,7 @@ qh_urb_transaction (
 	 */
 	if (likely ((urb->transfer_flags & URB_SHORT_NOT_OK) == 0
 				|| usb_pipecontrol (urb->pipe)))
-		qtd->hw_alt_next = EHCI_LIST_END;
+		qtd->hw_alt_next = EHCI_LIST_END(ehci);
 
 	/*
 	 * control requests may need a terminating data "status" ack;
@@ -594,17 +597,17 @@ qh_urb_transaction (
 			if (unlikely (!qtd))
 				goto cleanup;
 			qtd->urb = urb;
-			qtd_prev->hw_next = QTD_NEXT (qtd->qtd_dma);
+			qtd_prev->hw_next = QTD_NEXT(ehci, qtd->qtd_dma);
 			list_add_tail (&qtd->qtd_list, head);
 
 			/* never any data in such packets */
-			qtd_fill (qtd, 0, 0, token, 0);
+			qtd_fill(ehci, qtd, 0, 0, token, 0);
 		}
 	}
 
 	/* by default, enable interrupt on urb completion */
 	if (likely (!(urb->transfer_flags & URB_NO_INTERRUPT)))
-		qtd->hw_token |= __constant_cpu_to_le32 (QTD_IOC);
+		qtd->hw_token |= cpu_to_hc32(ehci, QTD_IOC);
 	return head;
 
 cleanup:
@@ -773,8 +776,8 @@ done:
 
 	/* init as live, toggle clear, advance to dummy */
 	qh->qh_state = QH_STATE_IDLE;
-	qh->hw_info1 = cpu_to_le32 (info1);
-	qh->hw_info2 = cpu_to_le32 (info2);
+	qh->hw_info1 = cpu_to_hc32(ehci, info1);
+	qh->hw_info2 = cpu_to_hc32(ehci, info2);
 	usb_settoggle (urb->dev, usb_pipeendpoint (urb->pipe), !is_input, 1);
 	qh_refresh (ehci, qh);
 	return qh;
@@ -786,7 +789,7 @@ done:
 
 static void qh_link_async (struct ehci_hcd *ehci, struct ehci_qh *qh)
 {
-	__le32		dma = QH_NEXT (qh->qh_dma);
+	__hc32		dma = QH_NEXT(ehci, qh->qh_dma);
 	struct ehci_qh	*head;
 
 	/* (re)start the async schedule? */
@@ -824,8 +827,6 @@ static void qh_link_async (struct ehci_hcd *ehci, struct ehci_qh *qh)
 
 /*-------------------------------------------------------------------------*/
 
-#define	QH_ADDR_MASK	__constant_cpu_to_le32(0x7f)
-
 /*
  * For control/bulk/interrupt, return QH with these TDs appended.
  * Allocates and initializes the QH if necessary.
@@ -841,6 +842,7 @@ static struct ehci_qh *qh_append_tds (
 )
 {
 	struct ehci_qh		*qh = NULL;
+	u32			qh_addr_mask = cpu_to_hc32(ehci, 0x7f);
 
 	qh = (struct ehci_qh *) *ptr;
 	if (unlikely (qh == NULL)) {
@@ -862,7 +864,7 @@ static struct ehci_qh *qh_append_tds (
 
                         /* usb_reset_device() briefly reverts to address 0 */
                         if (usb_pipedevice (urb->pipe) == 0)
-                                qh->hw_info1 &= ~QH_ADDR_MASK;
+                                qh->hw_info1 &= ~qh_addr_mask;
 		}
 
 		/* just one way to queue requests: swap with the dummy qtd.
@@ -871,7 +873,7 @@ static struct ehci_qh *qh_append_tds (
 		if (likely (qtd != NULL)) {
 			struct ehci_qtd		*dummy;
 			dma_addr_t		dma;
-			__le32			token;
+			__hc32			token;
 
 			/* to avoid racing the HC, use the dummy td instead of
 			 * the first td of our list (becomes new dummy).  both
@@ -879,7 +881,7 @@ static struct ehci_qh *qh_append_tds (
 			 * HC is allowed to fetch the old dummy (4.10.2).
 			 */
 			token = qtd->hw_token;
-			qtd->hw_token = HALT_BIT;
+			qtd->hw_token = HALT_BIT(ehci);
 			wmb ();
 			dummy = qh->dummy;
 
@@ -891,14 +893,14 @@ static struct ehci_qh *qh_append_tds (
 			list_add (&dummy->qtd_list, qtd_list);
 			__list_splice (qtd_list, qh->qtd_list.prev);
 
-			ehci_qtd_init (qtd, qtd->qtd_dma);
+			ehci_qtd_init(ehci, qtd, qtd->qtd_dma);
 			qh->dummy = qtd;
 
 			/* hc must see the new dummy at list end */
 			dma = qtd->qtd_dma;
 			qtd = list_entry (qh->qtd_list.prev,
 					struct ehci_qtd, qtd_list);
-			qtd->hw_next = QTD_NEXT (dma);
+			qtd->hw_next = QTD_NEXT(ehci, dma);
 
 			/* let the hc process these next qtds */
 			wmb ();
@@ -974,7 +976,7 @@ static void end_unlink_async (struct ehci_hcd *ehci)
 
 	timer_action_done (ehci, TIMER_IAA_WATCHDOG);
 
-	// qh->hw_next = cpu_to_le32 (qh->qh_dma);
+	// qh->hw_next = cpu_to_hc32(qh->qh_dma);
 	qh->qh_state = QH_STATE_IDLE;
 	qh->qh_next.qh = NULL;
 	qh_put (qh);			// refcount from reclaim