* [PATCH 4/5] forcedeth: recover error support
@ 2009-01-30 18:56 Ayaz Abdulla
0 siblings, 0 replies; 3+ messages in thread
From: Ayaz Abdulla @ 2009-01-30 18:56 UTC (permalink / raw)
To: Jeff Garzik, Manfred Spraul, David S. Miller, Andrew Morton,
nedev
[-- Attachment #1: Type: text/plain, Size: 213 bytes --]
This patch adds another type of recoverable error to the driver. It also
modifies the sequence for recovery to include a mac reset and clearing
of interrupts.
Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
[-- Attachment #2: patch-forcedeth-recover-error --]
[-- Type: text/plain, Size: 1674 bytes --]
--- old/drivers/net/forcedeth.c 2009-01-28 12:42:58.000000000 -0800
+++ new/drivers/net/forcedeth.c 2009-01-28 12:44:18.000000000 -0800
@@ -102,7 +102,7 @@
enum {
NvRegIrqStatus = 0x000,
#define NVREG_IRQSTAT_MIIEVENT 0x040
-#define NVREG_IRQSTAT_MASK 0x81ff
+#define NVREG_IRQSTAT_MASK 0x83ff
NvRegIrqMask = 0x004,
#define NVREG_IRQ_RX_ERROR 0x0001
#define NVREG_IRQ_RX 0x0002
@@ -113,7 +113,7 @@
#define NVREG_IRQ_LINK 0x0040
#define NVREG_IRQ_RX_FORCED 0x0080
#define NVREG_IRQ_TX_FORCED 0x0100
-#define NVREG_IRQ_RECOVER_ERROR 0x8000
+#define NVREG_IRQ_RECOVER_ERROR 0x8200
#define NVREG_IRQMASK_THROUGHPUT 0x00df
#define NVREG_IRQMASK_CPU 0x0060
#define NVREG_IRQ_TX_ALL (NVREG_IRQ_TX_ERR|NVREG_IRQ_TX_OK|NVREG_IRQ_TX_FORCED)
@@ -4064,13 +4064,15 @@
if (np->recover_error) {
np->recover_error = 0;
- printk(KERN_INFO "forcedeth: MAC in recoverable error state\n");
+ printk(KERN_INFO "%s: MAC in recoverable error state\n", dev->name);
if (netif_running(dev)) {
netif_tx_lock_bh(dev);
netif_addr_lock(dev);
spin_lock(&np->lock);
/* stop engines */
nv_stop_rxtx(dev);
+ if (np->driver_data & DEV_HAS_POWER_CNTRL)
+ nv_mac_reset(dev);
nv_txrx_reset(dev);
/* drain rx queue */
nv_drain_rxtx(dev);
@@ -4088,6 +4090,11 @@
pci_push(base);
writel(NVREG_TXRXCTL_KICK|np->txrxctl_bits, get_hwbase(dev) + NvRegTxRxControl);
pci_push(base);
+ /* clear interrupts */
+ if (!(np->msi_flags & NV_MSI_X_ENABLED))
+ writel(NVREG_IRQSTAT_MASK, base + NvRegIrqStatus);
+ else
+ writel(NVREG_IRQSTAT_MASK, base + NvRegMSIXIrqStatus);
/* restart rx engine */
nv_start_rxtx(dev);
^ permalink raw reply [flat|nested] 3+ messages in thread
* [PATCH 4/5] forcedeth: recover error support
@ 2009-02-06 19:27 Ayaz Abdulla
2009-02-07 8:26 ` David Miller
0 siblings, 1 reply; 3+ messages in thread
From: Ayaz Abdulla @ 2009-02-06 19:27 UTC (permalink / raw)
To: Jeff Garzik, Manfred Spraul, David S. Miller, Andrew Morton,
nedev
[-- Attachment #1: Type: text/plain, Size: 212 bytes --]
This patch adds another type of recoverable error to the driver. It also
modifies the sequence for recovery to include a mac reset and clearing
of interrupts.
Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
[-- Attachment #2: patch-forcedeth-recover-error --]
[-- Type: text/plain, Size: 1674 bytes --]
--- old/drivers/net/forcedeth.c 2009-02-06 14:16:30.000000000 -0500
+++ new/drivers/net/forcedeth.c 2009-02-06 14:16:47.000000000 -0500
@@ -102,7 +102,7 @@
enum {
NvRegIrqStatus = 0x000,
#define NVREG_IRQSTAT_MIIEVENT 0x040
-#define NVREG_IRQSTAT_MASK 0x81ff
+#define NVREG_IRQSTAT_MASK 0x83ff
NvRegIrqMask = 0x004,
#define NVREG_IRQ_RX_ERROR 0x0001
#define NVREG_IRQ_RX 0x0002
@@ -113,7 +113,7 @@
#define NVREG_IRQ_LINK 0x0040
#define NVREG_IRQ_RX_FORCED 0x0080
#define NVREG_IRQ_TX_FORCED 0x0100
-#define NVREG_IRQ_RECOVER_ERROR 0x8000
+#define NVREG_IRQ_RECOVER_ERROR 0x8200
#define NVREG_IRQMASK_THROUGHPUT 0x00df
#define NVREG_IRQMASK_CPU 0x0060
#define NVREG_IRQ_TX_ALL (NVREG_IRQ_TX_ERR|NVREG_IRQ_TX_OK|NVREG_IRQ_TX_FORCED)
@@ -4073,13 +4073,15 @@
if (np->recover_error) {
np->recover_error = 0;
- printk(KERN_INFO "forcedeth: MAC in recoverable error state\n");
+ printk(KERN_INFO "%s: MAC in recoverable error state\n", dev->name);
if (netif_running(dev)) {
netif_tx_lock_bh(dev);
netif_addr_lock(dev);
spin_lock(&np->lock);
/* stop engines */
nv_stop_rxtx(dev);
+ if (np->driver_data & DEV_HAS_POWER_CNTRL)
+ nv_mac_reset(dev);
nv_txrx_reset(dev);
/* drain rx queue */
nv_drain_rxtx(dev);
@@ -4097,6 +4099,11 @@
pci_push(base);
writel(NVREG_TXRXCTL_KICK|np->txrxctl_bits, get_hwbase(dev) + NvRegTxRxControl);
pci_push(base);
+ /* clear interrupts */
+ if (!(np->msi_flags & NV_MSI_X_ENABLED))
+ writel(NVREG_IRQSTAT_MASK, base + NvRegIrqStatus);
+ else
+ writel(NVREG_IRQSTAT_MASK, base + NvRegMSIXIrqStatus);
/* restart rx engine */
nv_start_rxtx(dev);
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH 4/5] forcedeth: recover error support
2009-02-06 19:27 Ayaz Abdulla
@ 2009-02-07 8:26 ` David Miller
0 siblings, 0 replies; 3+ messages in thread
From: David Miller @ 2009-02-07 8:26 UTC (permalink / raw)
To: aabdulla; +Cc: jgarzik, manfred, akpm, netdev
From: Ayaz Abdulla <aabdulla@nvidia.com>
Date: Fri, 06 Feb 2009 14:27:27 -0500
> This patch adds another type of recoverable error to the driver. It also
> modifies the sequence for recovery to include a mac reset and clearing
> of interrupts.
>
> Signed-off-by: Ayaz Abdulla <aabdulla@nvidia.com>
Applied to net-next-2.6
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2009-02-07 8:26 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-01-30 18:56 [PATCH 4/5] forcedeth: recover error support Ayaz Abdulla
-- strict thread matches above, loose matches on Subject: below --
2009-02-06 19:27 Ayaz Abdulla
2009-02-07 8:26 ` David Miller
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).