Skip to contentSkip to conversation

sh_eth: Change handling of IRQ

PATCH New Nobuhiro Iwamatsu 2 participants 5 messages Mar 17, 2009 – Mar 31, 2009

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 sh_eth: Change handling of IRQ

Handling of IRQ of the SH7763/SH7764 CPU which sh_eth supported was changed.
This revises it for this change.

---
 drivers/net/sh_eth.c |   20 +++++++++++++++++---
 1 files changed, 17 insertions(+), 3 deletions(-)
drivers/net/sh_eth.c +17 -3
index 7f8e514..7b18827 100644
@@ -687,6 +687,7 @@ static irqreturn_t sh_eth_interrupt(int irq, void *netdev)
 {
 	struct net_device *ndev = netdev;
 	struct sh_eth_private *mdp = netdev_priv(ndev);
+	irqreturn_t ret = IRQ_NONE;
 	u32 ioaddr, boguscnt = RX_RING_SIZE;
 	u32 intr_status = 0;
 
@@ -696,7 +697,13 @@ static irqreturn_t sh_eth_interrupt(int irq, void *netdev)
 	/* Get interrpt stat */
 	intr_status = ctrl_inl(ioaddr + EESR);
 	/* Clear interrupt */
-	ctrl_outl(intr_status, ioaddr + EESR);
+	if (intr_status & (EESR_FRC | EESR_RMAF | EESR_RRF |
+			EESR_RTLF | EESR_RTSF | EESR_PRE | EESR_CERF |
+			TX_CHECK | EESR_ERR_CHECK)) {
+		ctrl_outl(intr_status, ioaddr + EESR);
+		ret = IRQ_HANDLED;
+	} else
+		goto other_irq;
 
 	if (intr_status & (EESR_FRC | /* Frame recv*/
 			EESR_RMAF | /* Multi cast address recv*/
@@ -723,9 +730,10 @@ static irqreturn_t sh_eth_interrupt(int irq, void *netdev)
 		       ndev->name, intr_status);
 	}
 
+other_irq:
 	spin_unlock(&mdp->lock);
 
-	return IRQ_HANDLED;
+	return ret;
 }
 
 static void sh_eth_timer(unsigned long data)
@@ -844,7 +852,13 @@ static int sh_eth_open(struct net_device *ndev)
 	int ret = 0;
 	struct sh_eth_private *mdp = netdev_priv(ndev);
 
-	ret = request_irq(ndev->irq, &sh_eth_interrupt, 0, ndev->name, ndev);
+	ret = request_irq(ndev->irq, &sh_eth_interrupt,
+#if defined(CONFIG_CPU_SUBTYPE_SH7763) || defined(CONFIG_CPU_SUBTYPE_SH7764)
+				IRQF_SHARED,
+#else
+				0,
+#endif
+				ndev->name, ndev);
 	if (ret) {
 		printk(KERN_ERR "Can not assign IRQ number to %s\n", CARDNAME);
 		return ret;
Signature · 1 line
1.6.2
#1 lore raw

PATCH 1/1 sh_eth: Fix mistake of the address of SH7763

Address of SH_TSU_ADDR and ARSTR of SH7763 was wrong. This revise it.

---
 drivers/net/sh_eth.h |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)
drivers/net/sh_eth.h +2 -2
index 73bc718..1537e13 100644
@@ -43,8 +43,8 @@
 
 #define SH7763_SKB_ALIGN 32
 /* Chip Base Address */
-# define SH_TSU_ADDR  0xFFE01800
-# define ARSTR 		  0xFFE01800
+# define SH_TSU_ADDR	0xFEE01800
+# define ARSTR			SH_TSU_ADDR
 
 /* Chip Registers */
 /* E-DMAC */
Signature · 1 line
1.6.2
#2 lore raw
Quoted 5 lines · Nobuhiro Iwamatsu

Handling of IRQ of the SH7763/SH7764 CPU which sh_eth supported was changed.
This revises it for this change.

Quoted 4 lines · Nobuhiro Iwamatsu

Address of SH_TSU_ADDR and ARSTR of SH7763 was wrong. This revise it.

The sh changes this depends on for the CPU subtypes are merged upstream now, so this can go through either tree. As there is nothing really netdev specific about this, I suppose the easiest is to just roll both of these in to the sh tree.

#3 lore raw
Quoted 11 lines · Paul Mundt
Quoted 5 lines · Nobuhiro Iwamatsu

Handling of IRQ of the SH7763/SH7764 CPU which sh_eth supported was changed.
This revises it for this change.

Quoted 3 lines · Nobuhiro Iwamatsu

Address of SH_TSU_ADDR and ARSTR of SH7763 was wrong. This revise it.

Quoted 6 lines
Quoted 1 line

The sh changes this depends on for the CPU subtypes are merged upstream now, so this can go through either tree. As there is nothing really netdev specific about this, I suppose the easiest is to just roll both of these in to the sh tree.

Oh, nevermind, I see these were merged already through the net tree..

#4 lore raw
Quoted 11 lines · Paul Mundt
Quoted 10 lines · Paul Mundt
Quoted 5 lines · Nobuhiro Iwamatsu

Handling of IRQ of the SH7763/SH7764 CPU which sh_eth supported was changed.
This revises it for this change.

Quoted 3 lines · Nobuhiro Iwamatsu

Address of SH_TSU_ADDR and ARSTR of SH7763 was wrong. This revise it.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

The sh changes this depends on for the CPU subtypes are merged upstream now, so this can go through either tree. As there is nothing really netdev specific about this, I suppose the easiest is to just roll both of these in to the sh tree.

Oh, nevermind, I see these were merged already through the net tree..

I should have reported that these were taken in.
I will report it from the next time.

Best regards,
Nobuhiro