netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: David Woodhouse <dwmw2@infradead.org>
To: netdev@vger.kernel.org
Subject: [PATCH 16/30] solos: Tidy up tx_mask handling for ports which need TX
Date: Tue, 17 Mar 2009 21:29:28 +0000	[thread overview]
Message-ID: <1237325368.27681.345.camel@macbook.infradead.org> (raw)
In-Reply-To: <1237310370.27681.314.camel@macbook.infradead.org>

Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
---
 drivers/atm/solos-pci.c |   20 +++++++++++++-------
 1 files changed, 13 insertions(+), 7 deletions(-)

diff --git a/drivers/atm/solos-pci.c b/drivers/atm/solos-pci.c
index acba08d..bf59c40 100644
--- a/drivers/atm/solos-pci.c
+++ b/drivers/atm/solos-pci.c
@@ -100,6 +100,7 @@ struct solos_card {
 	void __iomem *config_regs;
 	void __iomem *buffers;
 	int nr_ports;
+	int tx_mask;
 	struct pci_dev *dev;
 	struct atm_dev *atmdev[4];
 	struct tasklet_struct tlet;
@@ -590,15 +591,13 @@ void solos_bh(unsigned long card_arg)
 	struct solos_card *card = (void *)card_arg;
 	int port;
 	uint32_t card_flags;
-	uint32_t tx_mask;
 	uint32_t rx_done = 0;
 
 	card_flags = ioread32(card->config_regs + FLAGS_ADDR);
 
 	/* The TX bits are set if the channel is busy; clear if not. We want to
 	   invoke fpga_tx() unless _all_ the bits for active channels are set */
-	tx_mask = (1 << card->nr_ports) - 1;
-	if ((card_flags & tx_mask) != tx_mask)
+	if ((card_flags & card->tx_mask) != card->tx_mask)
 		fpga_tx(card);
 
 	for (port = 0; port < card->nr_ports; port++) {
@@ -887,15 +886,20 @@ static void fpga_queue(struct solos_card *card, int port, struct sk_buff *skb,
 		       struct atm_vcc *vcc)
 {
 	int old_len;
+	unsigned long flags;
 
 	SKB_CB(skb)->vcc = vcc;
 
-	spin_lock(&card->tx_queue_lock);
+	spin_lock_irqsave(&card->tx_queue_lock, flags);
 	old_len = skb_queue_len(&card->tx_queue[port]);
 	skb_queue_tail(&card->tx_queue[port], skb);
-	spin_unlock(&card->tx_queue_lock);
+	if (!old_len) {
+		card->tx_mask |= (1 << port);
+	}
+	spin_unlock_irqrestore(&card->tx_queue_lock, flags);
 
-	/* If TX might need to be started, do so */
+	/* Theoretically we could just schedule the tasklet here, but
+	   that introduces latency we don't want -- it's noticeable */
 	if (!old_len)
 		fpga_tx(card);
 }
@@ -911,7 +915,7 @@ static int fpga_tx(struct solos_card *card)
 
 	spin_lock_irqsave(&card->tx_lock, flags);
 
-	tx_pending = ioread32(card->config_regs + FLAGS_ADDR);
+	tx_pending = ioread32(card->config_regs + FLAGS_ADDR) & card->tx_mask;
 
 	dev_vdbg(&card->dev->dev, "TX Flags are %X\n", tx_pending);
 
@@ -925,6 +929,8 @@ static int fpga_tx(struct solos_card *card)
 			
 			spin_lock(&card->tx_queue_lock);
 			skb = skb_dequeue(&card->tx_queue[port]);
+			if (!skb)
+				card->tx_mask &= ~(1 << port);
 			spin_unlock(&card->tx_queue_lock);
 
 			if (skb && !card->using_dma) {
-- 
1.6.0.6



  parent reply	other threads:[~2009-03-17 21:29 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-17 17:19 [GIT *] Solos PCI ADSL card update David Woodhouse
2009-03-17 19:23 ` David Miller
2009-03-17 20:41   ` Stephen Hemminger
2009-03-17 21:36     ` David Woodhouse
2009-03-17 21:33   ` David Woodhouse
2009-03-17 21:29 ` [PATCH 01/30] solos: Fix length header in FPGA transfers David Woodhouse
2009-03-17 21:29 ` [PATCH 02/30] solos: Slight debugging improvements David Woodhouse
2009-03-17 21:29 ` [PATCH 03/30] solos: FPGA and firmware update support David Woodhouse
2009-03-17 21:29 ` [PATCH 04/30] solos: Clean up firmware loading code David Woodhouse
2009-03-17 21:29 ` [PATCH 05/30] solos: Kill global 'opens' count David Woodhouse
2009-03-17 21:29 ` [PATCH 06/30] solos: Handle attribute show/store in kernel more sanely David Woodhouse
2009-03-17 22:44   ` Stephen Hemminger
2009-03-17 22:49     ` David Woodhouse
2009-03-21 20:22       ` David Miller
2009-03-17 21:29 ` [PATCH 08/30] solos: Handle new line status change packets, hook up to ATM layer info David Woodhouse
2009-03-17 21:29 ` [PATCH 07/30] solos: Add initial list of parameters David Woodhouse
2009-03-17 21:29 ` [PATCH 09/30] solos: Kill existing connections on link down event David Woodhouse
2009-03-17 21:29 ` [PATCH 10/30] solos: Reject non-AAL5 connections.... for now David Woodhouse
2009-03-17 21:29 ` [PATCH 11/30] solos: Add SNR and Attn to status packet, fix oops on load David Woodhouse
2009-03-17 21:29 ` [PATCH 12/30] solos: Fix under-allocation of skb size for get/set parameters David Woodhouse
2009-03-17 21:29 ` [PATCH 13/30] solos: Remove parameter group from sysfs on ATM dev deregister David Woodhouse
2009-03-17 21:29 ` [PATCH 14/30] solos: First attempt at DMA support David Woodhouse
2009-03-17 21:29 ` [PATCH 15/30] solos: Tidy up DMA handling a little. Still untested David Woodhouse
2009-03-17 21:29 ` David Woodhouse [this message]
2009-03-17 21:29 ` [PATCH 17/30] solos: Remove unused loopback debug stuff David Woodhouse
2009-03-17 21:29 ` [PATCH 18/30] solos: Remove IRQF_DISABLED, don't frob IRQ enable on the FPGA in solos_irq() David Woodhouse
2009-03-17 21:29 ` [PATCH 19/30] solos: Remove superfluous wait_queue_head_t from struct solos_param David Woodhouse
2009-03-17 21:29 ` [PATCH 20/30] solos: Fix various bugs in status packet handling David Woodhouse
2009-03-17 21:29 ` [PATCH 21/30] solos: Clean up handling of card->tx_mask a little David Woodhouse
2009-03-17 21:29 ` [PATCH 22/30] solos: Remove debugging, commented-out test code David Woodhouse
2009-03-17 21:29 ` [PATCH 23/30] solos: Add 'reset' module parameter to reset the DSL chips on load David Woodhouse
2009-03-17 21:29 ` [PATCH 24/30] solos: Tidy up status interrupt handling, cope with 'ERROR' status David Woodhouse
2009-03-17 21:29 ` [PATCH 25/30] solos: Don't clear config registers at startup David Woodhouse
2009-03-17 21:29 ` [PATCH 26/30] solos: Set RX empty flag at startup only for !dma mode David Woodhouse
2009-03-17 21:29 ` [PATCH 27/30] solos: Swap upstream/downstream rates in status packet, clean up some more David Woodhouse
2009-03-17 21:29 ` [PATCH 28/30] solos: Reset device on unload, free pending skbs David Woodhouse
2009-03-17 21:29 ` [PATCH 30/30] solos: Disable DMA until we have an FPGA update with it actually implemented David Woodhouse
2009-03-17 21:29 ` [PATCH 29/30] solos: Automatically determine number of ports David Woodhouse

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=1237325368.27681.345.camel@macbook.infradead.org \
    --to=dwmw2@infradead.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).