netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Kyle McMartin <kyle@parisc-linux.org>
To: netdev@vger.kernel.org
Cc: grundler@parisc-linux.org, val_henson@linux.intel.com, akpm@osdl.org
Subject: [PATCH 6/9] [TULIP] Fix IRQ/DMA race
Date: Mon, 07 Aug 2006 20:53:29 +0000	[thread overview]
Message-ID: <11549840343640-git-send-email-kyle@parisc-linux.org> (raw)
In-Reply-To: <11549840341728-git-send-email-kyle@parisc-linux.org>

From: Grant Grundler <grundler@parisc-linux.org>

IRQs are racing with tulip_down().
DMA can be restarted by tulip_interrupt() _after_ we call
tulip_stop_rxtx() and the DMA buffers are unmapped.  The result
is an MCA (hard crash on ia64) because of an IO TLB miss.

Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
---
 drivers/net/tulip/interrupt.c  |    4 ++++
 drivers/net/tulip/tulip_core.c |   17 +++++++----------
 2 files changed, 11 insertions(+), 10 deletions(-)

diff --git a/drivers/net/tulip/interrupt.c b/drivers/net/tulip/interrupt.c
index 99ccf2e..19faa0e 100644
--- a/drivers/net/tulip/interrupt.c
+++ b/drivers/net/tulip/interrupt.c
@@ -87,6 +87,10 @@ int tulip_refill_rx(struct net_device *d
 		}
 		tp->rx_ring[entry].status = cpu_to_le32(DescOwned);
 	}
+
+/* FIXME: restarting DMA breaks tulip_down() code path.
+   tulip_down() will unmap the RX and TX descriptors.
+ */
 	if(tp->chip_id == LC82C168) {
 		if(((ioread32(tp->base_addr + CSR5)>>17)&0x07) == 4) {
 			/* Rx stopped due to out of buffers,
diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c
index 81905f4..363e5f6 100644
--- a/drivers/net/tulip/tulip_core.c
+++ b/drivers/net/tulip/tulip_core.c
@@ -742,21 +742,20 @@ #endif
 
 	/* Disable interrupts by clearing the interrupt mask. */
 	iowrite32 (0x00000000, ioaddr + CSR7);
+	ioread32 (ioaddr + CSR7);	/* flush posted write */
 
-	/* Stop the Tx and Rx processes. */
-	tulip_stop_rxtx(tp);
+	spin_unlock_irqrestore (&tp->lock, flags);
 
-	/* prepare receive buffers */
-	tulip_refill_rx(dev);
+	free_irq (dev->irq, dev);	/* no more races after this */
+	tulip_stop_rxtx(tp);		/* Stop DMA */
 
-	/* release any unconsumed transmit buffers */
-	tulip_clean_tx_ring(tp);
+	/* Put driver back into the state we start with */
+	tulip_refill_rx(dev);		/* prepare RX buffers */
+	tulip_clean_tx_ring(tp);	/* clean up unsent TX buffers */
 
 	if (ioread32 (ioaddr + CSR6) != 0xffffffff)
 		tp->stats.rx_missed_errors += ioread32 (ioaddr + CSR8) & 0xffff;
 
-	spin_unlock_irqrestore (&tp->lock, flags);
-
 	init_timer(&tp->timer);
 	tp->timer.data = (unsigned long)dev;
 	tp->timer.function = tulip_tbl[tp->chip_id].media_timer;
@@ -782,7 +781,6 @@ static int tulip_close (struct net_devic
 		printk (KERN_DEBUG "%s: Shutting down ethercard, status was %2.2x.\n",
 			dev->name, ioread32 (ioaddr + CSR5));
 
-	free_irq (dev->irq, dev);
 
 	/* Free all the skbuffs in the Rx queue. */
 	for (i = 0; i < RX_RING_SIZE; i++) {
@@ -1752,7 +1750,6 @@ static int tulip_suspend (struct pci_dev
 		tulip_down(dev);
 
 	netif_device_detach(dev);
-	free_irq(dev->irq, dev);
 
 	pci_save_state(pdev);
 	pci_disable_device(pdev);
-- 
1.4.1.1


  reply	other threads:[~2006-08-07 20:53 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-08-07 20:44 [PATCH 0/9] tulip patches from parisc-linux Kyle McMartin
2006-08-07 20:53 ` [PATCH 1/9] [TULIP] Fix PHY init and reset Kyle McMartin
2006-08-07 20:53   ` [PATCH 2/9] [TULIP] Print physical address in tulip_init_one Kyle McMartin
2006-08-07 20:53     ` [PATCH 3/9] [TULIP] Make tulip_stop_rxtx() failure case slightly more informative Kyle McMartin
2006-08-07 20:53       ` [PATCH 4/9] [TULIP] Clean tulip.h so it can be used by winbond-840.c Kyle McMartin
2006-08-07 20:53         ` [PATCH 5/9] [TULIP] Flush MMIO writes in reset sequence Kyle McMartin
2006-08-07 20:53           ` Kyle McMartin [this message]
2006-08-07 20:53             ` [PATCH 7/9] [TULIP] Defer tulip_select_media() to process context Kyle McMartin
2006-08-07 20:53               ` [PATCH 8/9] [TULIP] Make DS21143 printout match lspci output Kyle McMartin
2006-08-07 20:53                 ` [PATCH 9/9] [TULIP] Fix section mismatch in de2104x.c Kyle McMartin
2006-08-09  5:37                   ` Jeff Garzik
2006-08-09  5:36                 ` [PATCH 8/9] [TULIP] Make DS21143 printout match lspci output Jeff Garzik
2006-08-09  5:35             ` [PATCH 6/9] [TULIP] Fix IRQ/DMA race Jeff Garzik
2006-08-09  6:44               ` Valerie Henson
2006-08-09  5:33           ` [PATCH 5/9] [TULIP] Flush MMIO writes in reset sequence Jeff Garzik
2006-08-09  5:33         ` [PATCH 4/9] [TULIP] Clean tulip.h so it can be used by winbond-840.c Jeff Garzik
2006-08-09 15:00           ` Grant Grundler
2006-08-10 12:26             ` Jeff Garzik
2006-08-09  5:30       ` [PATCH 3/9] [TULIP] Make tulip_stop_rxtx() failure case slightly more informative Jeff Garzik
2006-08-09  5:29   ` [PATCH 1/9] [TULIP] Fix PHY init and reset Jeff Garzik

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=11549840343640-git-send-email-kyle@parisc-linux.org \
    --to=kyle@parisc-linux.org \
    --cc=akpm@osdl.org \
    --cc=grundler@parisc-linux.org \
    --cc=netdev@vger.kernel.org \
    --cc=val_henson@linux.intel.com \
    /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).