From: Bryan Wu <cooloney@kernel.org>
To: jeff@garzik.org
Cc: netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
Bryan Wu <cooloney@kernel.org>
Subject: [PATCH 2/3] Blackfin EMAC Driver: enable TXDWA new feature for new silicon (rev > 0.2)
Date: Sun, 27 Jul 2008 22:45:04 +0800 [thread overview]
Message-ID: <1217169905-7055-3-git-send-email-cooloney@kernel.org> (raw)
In-Reply-To: <1217169905-7055-1-git-send-email-cooloney@kernel.org>
Signed-off-by: Bryan Wu <cooloney@kernel.org>
---
drivers/net/bfin_mac.c | 103 ++++++++++++++++++++++++++++++++++++------------
1 files changed, 77 insertions(+), 26 deletions(-)
diff --git a/drivers/net/bfin_mac.c b/drivers/net/bfin_mac.c
index 17951cb..57ebe6d 100644
--- a/drivers/net/bfin_mac.c
+++ b/drivers/net/bfin_mac.c
@@ -606,36 +606,87 @@ adjust_head:
static int bfin_mac_hard_start_xmit(struct sk_buff *skb,
struct net_device *dev)
{
- unsigned int data;
+ u16 *data;
current_tx_ptr->skb = skb;
- /*
- * Is skb->data always 16-bit aligned?
- * Do we need to memcpy((char *)(tail->packet + 2), skb->data, len)?
- */
- if ((((unsigned int)(skb->data)) & 0x02) == 2) {
- /* move skb->data to current_tx_ptr payload */
- data = (unsigned int)(skb->data) - 2;
- *((unsigned short *)data) = (unsigned short)(skb->len);
- current_tx_ptr->desc_a.start_addr = (unsigned long)data;
- /* this is important! */
- blackfin_dcache_flush_range(data, (data + (skb->len)) + 2);
-
+ if (ANOMALY_05000285) {
+ /*
+ * TXDWA feature is not avaible to older revision < 0.3 silicon
+ * of BF537
+ *
+ * Only if data buffer is ODD WORD alignment, we do not
+ * need to memcpy
+ */
+ u32 data_align = (u32)(skb->data) & 0x3;
+ if (data_align == 0x2) {
+ /* move skb->data to current_tx_ptr payload */
+ data = (u16 *)(skb->data) - 1;
+ *data = (u16)(skb->len);
+ current_tx_ptr->desc_a.start_addr = (u32)data;
+ /* this is important! */
+ blackfin_dcache_flush_range((u32)data,
+ (u32)((u8 *)data + skb->len + 4));
+ } else {
+ *((u16 *)(current_tx_ptr->packet)) = (u16)(skb->len);
+ memcpy((u8 *)(current_tx_ptr->packet + 2), skb->data,
+ skb->len);
+ current_tx_ptr->desc_a.start_addr =
+ (u32)current_tx_ptr->packet;
+ if (current_tx_ptr->status.status_word != 0)
+ current_tx_ptr->status.status_word = 0;
+ blackfin_dcache_flush_range(
+ (u32)current_tx_ptr->packet,
+ (u32)(current_tx_ptr->packet + skb->len + 2));
+ }
} else {
- *((unsigned short *)(current_tx_ptr->packet)) =
- (unsigned short)(skb->len);
- memcpy((char *)(current_tx_ptr->packet + 2), skb->data,
- (skb->len));
- current_tx_ptr->desc_a.start_addr =
- (unsigned long)current_tx_ptr->packet;
- if (current_tx_ptr->status.status_word != 0)
- current_tx_ptr->status.status_word = 0;
- blackfin_dcache_flush_range((unsigned int)current_tx_ptr->
- packet,
- (unsigned int)(current_tx_ptr->
- packet + skb->len) +
- 2);
+ /*
+ * TXDWA feature is avaible to revision < 0.3 silicon of
+ * BF537 and always avaible to BF52x
+ */
+ u32 data_align = (u32)(skb->data) & 0x3;
+ if (data_align == 0x0) {
+ u16 sysctl = bfin_read_EMAC_SYSCTL();
+ sysctl |= TXDWA;
+ bfin_write_EMAC_SYSCTL(sysctl);
+
+ /* move skb->data to current_tx_ptr payload */
+ data = (u16 *)(skb->data) - 2;
+ *data = (u16)(skb->len);
+ current_tx_ptr->desc_a.start_addr = (u32)data;
+ /* this is important! */
+ blackfin_dcache_flush_range(
+ (u32)data,
+ (u32)((u8 *)data + skb->len + 4));
+ } else if (data_align == 0x2) {
+ u16 sysctl = bfin_read_EMAC_SYSCTL();
+ sysctl &= ~TXDWA;
+ bfin_write_EMAC_SYSCTL(sysctl);
+
+ /* move skb->data to current_tx_ptr payload */
+ data = (u16 *)(skb->data) - 1;
+ *data = (u16)(skb->len);
+ current_tx_ptr->desc_a.start_addr = (u32)data;
+ /* this is important! */
+ blackfin_dcache_flush_range(
+ (u32)data,
+ (u32)((u8 *)data + skb->len + 4));
+ } else {
+ u16 sysctl = bfin_read_EMAC_SYSCTL();
+ sysctl &= ~TXDWA;
+ bfin_write_EMAC_SYSCTL(sysctl);
+
+ *((u16 *)(current_tx_ptr->packet)) = (u16)(skb->len);
+ memcpy((u8 *)(current_tx_ptr->packet + 2), skb->data,
+ skb->len);
+ current_tx_ptr->desc_a.start_addr =
+ (u32)current_tx_ptr->packet;
+ if (current_tx_ptr->status.status_word != 0)
+ current_tx_ptr->status.status_word = 0;
+ blackfin_dcache_flush_range(
+ (u32)current_tx_ptr->packet,
+ (u32)(current_tx_ptr->packet + skb->len + 2));
+ }
}
/* enable this packet's dma */
--
1.5.6
next prev parent reply other threads:[~2008-07-27 14:46 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-07-27 14:45 [PATCH 0/3] Blackfin EMAC Driver updates for 2.6.27 (v2) Bryan Wu
2008-07-27 14:45 ` [PATCH 1/3] Blackfin EMAC Driver: add proper __devinit/__devexit markings Bryan Wu
2008-07-28 15:21 ` Paulius Zaleckas
2008-07-28 15:21 ` Paulius Zaleckas
2008-07-29 3:34 ` Bryan Wu
2008-07-29 21:49 ` Jeff Garzik
2008-07-27 14:45 ` Bryan Wu [this message]
2008-07-27 14:45 ` [PATCH 3/3] Blackfin EMAC Driver: Functional power management support Bryan Wu
-- strict thread matches above, loose matches on Subject: below --
2008-07-27 6:13 [PATCH 0/3] Blackfin EMAC Driver updates for 2.6.27 Bryan Wu
2008-07-27 6:13 ` [PATCH 2/3] Blackfin EMAC Driver: enable TXDWA new feature for new silicon (rev > 0.2) Bryan Wu
2008-07-28 10:05 ` Ingo Oeser
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1217169905-7055-3-git-send-email-cooloney@kernel.org \
--to=cooloney@kernel.org \
--cc=jeff@garzik.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.