Re: [PATCH] switch iph5526_probe to initcalls

Re: [PATCH] switch iph5526_probe to initcalls

[David S. Miller]

   From: Christoph Hellwig <hch@lst.de>
   Date: Sun, 15 Jun 2003 20:00:11 +0200

   It's a pci driver so it has no business in Space.c
   
   p.s. it doesn't compile with or without the patch due to scsi issues..
   
Can we put this patch on hold until that's resolved?
Thanks.

[PATCH] switch iph5526_probe to initcalls

[Christoph Hellwig]

It's a pci driver so it has no business in Space.c

p.s. it doesn't compile with or without the patch due to scsi issues..

--- 1.21/drivers/net/Space.c	Mon Jun  9 18:52:43 2003
+++ edited/drivers/net/Space.c	Sat Jun 14 21:32:48 2003
@@ -398,24 +398,6 @@
 	return -ENODEV;
 }
 
-#ifdef CONFIG_NET_FC
-static int fcif_probe(struct net_device *dev)
-{
-	if (dev->base_addr == -1)
-		return 1;
-
-	if (1
-#ifdef CONFIG_IPHASE5526
-	    && iph5526_probe(dev)
-#endif
-	    && 1 ) {
-		return 1; /* -ENODEV or -EAGAIN would be more accurate. */
-	}
-	return 0;
-}
-#endif  /* CONFIG_NET_FC */
-
-
 #ifdef CONFIG_ETHERTAP
 static struct net_device tap0_dev = {
 	.name		= "tap0",
@@ -588,22 +570,6 @@
 #define      NEXT_DEV        (&tr0_dev)
 
 #endif 
-
-#ifdef CONFIG_NET_FC
-static struct net_device fc1_dev = {
-	.name		= "fc1",
-	.next		= NEXT_DEV,
-	.init		= fcif_probe
-};
-static struct net_device fc0_dev = {
-	.name		= "fc0",
-	.next		=  &fc1_dev,
-	.init		= fcif_probe
-};
-#undef       NEXT_DEV
-#define      NEXT_DEV        (&fc0_dev)
-#endif
-
 
 #ifdef CONFIG_SBNI
 static struct net_device sbni7_dev = {
--- 1.25/drivers/net/fc/iph5526.c	Tue Jun 10 00:49:03 2003
+++ edited/drivers/net/fc/iph5526.c	Sat Jun 14 21:32:48 2003
@@ -239,19 +239,7 @@
 
 static int __init iph5526_probe_pci(struct net_device *dev)
 {
-#ifdef MODULE
 	struct fc_info *fi = (struct fc_info *)dev->priv;
-#else
-	struct fc_info *fi = fc[count];
-	static int count;
-	int err;
- 
-	if (!fi)
-		return -ENODEV;
-
-	fc_setup(dev);
-	count++;
-#endif
 	fi->dev = dev;
 	dev->base_addr = fi->base_addr;
 	dev->irq = fi->irq;
@@ -4479,8 +4467,6 @@
 	return buf;
 }
 
-#ifdef MODULE
-
 #define NAMELEN		8	/* # of chars for storing dev->name */
 
 static struct net_device *dev_fc[MAX_FC_CARDS];
@@ -4491,7 +4477,7 @@
 static int scsi_registered;
 
 
-int init_module(void)
+static int __init iph5526_init(void)
 {
 	int i = 0;
 
@@ -4530,7 +4516,7 @@
 	return 0;
 }
 
-void cleanup_module(void)
+static void __exit iph5526_exit(void)
 {
 	int i = 0;
 	while(fc[i] != NULL) {
@@ -4549,7 +4535,9 @@
 	if (scsi_registered == TRUE)
 		scsi_unregister_host(&driver_template); 
 }
-#endif /* MODULE */
+
+module_init(iph5526_init);
+module_exit(iph5526_exit);
 
 void clean_up_memory(struct fc_info *fi)
 {

Re: [PATCH] switch iph5526_probe to initcalls

[David S. Miller]

   From: Christoph Hellwig <hch@lst.de>
   Date: Sun, 15 Jun 2003 20:05:14 +0200

   On Sun, Jun 15, 2003 at 10:58:11AM -0700, David S. Miller wrote:
   >    p.s. it doesn't compile with or without the patch due to scsi issues..
   >    
   > Can we put this patch on hold until that's resolved?
   
   Well, we could.  It's just that I doubt anyone will fix it anytime soon,
   the hardware doesn't seem to be widely used and it's hard to touch the
   driver without a barfbag nearby.

Ugh, ok.

At least it's impossible for you to break the tree
with this change. :)

Re: [PATCH] switch iph5526_probe to initcalls

[Christoph Hellwig]

On Sun, Jun 15, 2003 at 10:58:11AM -0700, David S. Miller wrote:
>    From: Christoph Hellwig <hch@lst.de>
>    Date: Sun, 15 Jun 2003 20:00:11 +0200
> 
>    It's a pci driver so it has no business in Space.c
>    
>    p.s. it doesn't compile with or without the patch due to scsi issues..
>    
> Can we put this patch on hold until that's resolved?

Well, we could.  It's just that I doubt anyone will fix it anytime soon,
the hardware doesn't seem to be widely used and it's hard to touch the
driver without a barfbag nearby.  (and you also applied the alloc_fcdev
changes for it :))

Re: [PATCH] switch iph5526_probe to initcalls

[Christoph Hellwig]

On Sun, Jun 15, 2003 at 11:03:58AM -0700, David S. Miller wrote:
> At least it's impossible for you to break the tree
> with this change. :)

Take care, I could have smuggled something evil into Space.c..