2.6.0-test3 current - firewire compile error

2.6.0-test3 current - firewire compile error

[Ivan Gyurdiev]

Hopefully, this is not a duplicate post:
===========================================

drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
drivers/ieee1394/nodemgr.c:471: error: structure has no member named `name'
drivers/ieee1394/nodemgr.c: In function `nodemgr_create_node':
drivers/ieee1394/nodemgr.c:723: error: structure has no member named `name'
drivers/ieee1394/nodemgr.c: In function `nodemgr_update_node':
drivers/ieee1394/nodemgr.c:1318: error: structure has no member named `name'
drivers/ieee1394/nodemgr.c: In function `nodemgr_add_host':
drivers/ieee1394/nodemgr.c:1775: error: structure has no member named `name'
make[2]: *** [drivers/ieee1394/nodemgr.o] Error 1
make[1]: *** [drivers/ieee1394] Error 2
make: *** [drivers] Error 2

Re: 2.6.0-test3 current - firewire compile error

[walt]

Ivan Gyurdiev wrote:
> Hopefully, this is not a duplicate post:
> ===========================================
> 
> drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
> drivers/ieee1394/nodemgr.c:471: error: structure has no member named `name'

I got a similar error starting with last night's bk pull:

drivers/pnp/core.c: In function `pnp_register_protocol':
drivers/pnp/core.c:72: structure has no member named `name'

Re: 2.6.0-test3 current - compile error - no member named 'name'

[Ivan Gyurdiev]

walt wrote:
> Ivan Gyurdiev wrote:
> 
>> Hopefully, this is not a duplicate post:
>> ===========================================
>>
>> drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
>> drivers/ieee1394/nodemgr.c:471: error: structure has no member named 
>> `name'
> 
> 
> I got a similar error starting with last night's bk pull:
> 
> drivers/pnp/core.c: In function `pnp_register_protocol':
> drivers/pnp/core.c:72: structure has no member named `name'
> 
> -

And I just got another one of those trying to compile the nvidia driver 
for this kernel. So apparently this is not firewire or pnp specific.

Re: 2.6.0-test3 current - compile error - no member named 'name'

[Greg KH]

On Sat, Aug 16, 2003 at 09:59:10AM -0400, Ivan Gyurdiev wrote:
> walt wrote:
> >Ivan Gyurdiev wrote:
> >
> >>Hopefully, this is not a duplicate post:
> >>===========================================
> >>
> >>drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
> >>drivers/ieee1394/nodemgr.c:471: error: structure has no member named 
> >>`name'
> >
> >
> >I got a similar error starting with last night's bk pull:
> >
> >drivers/pnp/core.c: In function `pnp_register_protocol':
> >drivers/pnp/core.c:72: structure has no member named `name'
> >
> >-
> 
> And I just got another one of those trying to compile the nvidia driver 
> for this kernel. So apparently this is not firewire or pnp specific.

Well go bug nvidia about that, nothing we can do for closed source
modules...

thanks,

greg k-h

Re: 2.6.0-test3 current - firewire compile error

[Greg KH]

On Sat, Aug 16, 2003 at 12:11:58AM -0700, walt wrote:
> Ivan Gyurdiev wrote:
> >Hopefully, this is not a duplicate post:
> >===========================================
> >
> >drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
> >drivers/ieee1394/nodemgr.c:471: error: structure has no member named `name'
> 
> I got a similar error starting with last night's bk pull:
> 
> drivers/pnp/core.c: In function `pnp_register_protocol':
> drivers/pnp/core.c:72: structure has no member named `name'

Hm, I thought the pnp layer had already been fixed up.

Adam?

thanks,

greg k-h

Re: 2.6.0-test3 current - firewire compile error

[Adam Belay]

On Sat, Aug 16, 2003 at 09:36:44AM -0700, Greg KH wrote:
> On Sat, Aug 16, 2003 at 12:11:58AM -0700, walt wrote:
> > Ivan Gyurdiev wrote:
> > >Hopefully, this is not a duplicate post:
> > >===========================================
> > >
> > >drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
> > >drivers/ieee1394/nodemgr.c:471: error: structure has no member named `name'
> > 
> > I got a similar error starting with last night's bk pull:
> > 
> > drivers/pnp/core.c: In function `pnp_register_protocol':
> > drivers/pnp/core.c:72: structure has no member named `name'
> 
> Hm, I thought the pnp layer had already been fixed up.
> 
> Adam?
> 

I apoligize for forgetting to correct this.  Here is a fix.

Thanks,
Adam


diff -urN a/drivers/pnp/core.c b/drivers/pnp/core.c
--- a/drivers/pnp/core.c	2003-08-09 04:33:18.000000000 +0000
+++ b/drivers/pnp/core.c	2003-08-16 16:26:59.000000000 +0000
@@ -69,7 +69,6 @@
 
 	protocol->number = nodenum;
 	sprintf(protocol->dev.bus_id, "pnp%d", nodenum);
-	strlcpy(protocol->dev.name,protocol->name,DEVICE_NAME_SIZE);
 	return device_register(&protocol->dev);
 }
 
diff -urN a/drivers/pnp/isapnp/core.c b/drivers/pnp/isapnp/core.c
--- a/drivers/pnp/isapnp/core.c	2003-08-09 04:37:19.000000000 +0000
+++ b/drivers/pnp/isapnp/core.c	2003-08-16 16:45:16.000000000 +0000
@@ -743,7 +743,7 @@
 			size = 0;
 			break;
 		case _LTAG_ANSISTR:
-			isapnp_parse_name(dev->dev.name, sizeof(dev->dev.name), &size);
+			isapnp_parse_name(dev->name, sizeof(dev->name), &size);
 			break;
 		case _LTAG_UNICODESTR:
 			/* silently ignore */
@@ -808,7 +808,7 @@
 		case _STAG_VENDOR:
 			break;
 		case _LTAG_ANSISTR:
-			isapnp_parse_name(card->dev.name, sizeof(card->dev.name), &size);
+			isapnp_parse_name(card->name, sizeof(card->name), &size);
 			break;
 		case _LTAG_UNICODESTR:
 			/* silently ignore */
@@ -1144,11 +1144,11 @@
 	protocol_for_each_card(&isapnp_protocol,card) {
 		cards++;
 		if (isapnp_verbose) {
-			printk(KERN_INFO "isapnp: Card '%s'\n", card->dev.name[0]?card->dev.name:"Unknown");
+			printk(KERN_INFO "isapnp: Card '%s'\n", card->name[0]?card->name:"Unknown");
 			if (isapnp_verbose < 2)
 				continue;
 			card_for_each_dev(card,dev) {
-				printk(KERN_INFO "isapnp:   Device '%s'\n", dev->dev.name[0]?dev->dev.name:"Unknown");
+				printk(KERN_INFO "isapnp:   Device '%s'\n", dev->name[0]?dev->name:"Unknown");
 			}
 		}
 	}
--- a/drivers/serial/8250_pnp.c	2003-08-09 04:31:45.000000000 +0000
+++ b/drivers/serial/8250_pnp.c	2003-08-16 16:47:21.000000000 +0000
@@ -369,7 +369,7 @@
  */
 static int __devinit serial_pnp_guess_board(struct pnp_dev *dev, int *flags)
 {
-	if (!(check_name(dev->dev.name) || (dev->card && check_name(dev->card->dev.name))))
+	if (!(check_name(pnp_dev_name(dev)) || (dev->card && check_name(dev->card->name))))
 		return -ENODEV;
 
 	if (check_resources(dev->independent))
--- a/include/linux/pnp.h	2003-08-09 04:42:16.000000000 +0000
+++ b/include/linux/pnp.h	2003-08-16 16:43:32.000000000 +0000
@@ -19,6 +19,7 @@
 #define PNP_MAX_DMA		2
 #define PNP_MAX_DEVICES		8
 #define PNP_ID_LEN		8
+#define PNP_NAME_LEN		50
 
 struct pnp_protocol;
 struct pnp_dev;
@@ -133,6 +134,7 @@
 	struct pnp_protocol * protocol;
 	struct pnp_id * id;		/* contains supported EISA IDs*/
 
+	char name[PNP_NAME_LEN];	/* contains a human-readable name */
 	unsigned char	pnpver;		/* Plug & Play version */
 	unsigned char	productver;	/* product version */
 	unsigned int	serial;		/* serial number */
@@ -187,6 +189,7 @@
 	struct pnp_option * dependent;
 	struct pnp_resource_table res;
 
+	char name[PNP_NAME_LEN];	/* contains a human-readable name */
 	unsigned short	regs;		/* ISAPnP: supported registers */
 	int 		flags;		/* used by protocols */
 	struct proc_dir_entry *procent;	/* device entry in /proc/bus/isapnp */
@@ -204,7 +207,7 @@
 	for((dev) = card_to_pnp_dev((card)->devices.next); \
 	(dev) != card_to_pnp_dev(&(card)->devices); \
 	(dev) = card_to_pnp_dev((dev)->card_list.next))
-#define pnp_dev_name(dev) (dev)->dev.name
+#define pnp_dev_name(dev) (dev)->name
 
 static inline void *pnp_get_drvdata (struct pnp_dev *pdev)
 {

Re: 2.6.0-test3 current - firewire compile error

[walt]

Adam Belay wrote:

>>On Sat, Aug 16, 2003 at 12:11:58AM -0700, walt wrote:

>>>drivers/pnp/core.c: In function `pnp_register_protocol':
>>>drivers/pnp/core.c:72: structure has no member named `name'

> I apoligize for forgetting to correct this.  Here is a fix.

> diff -urN a/drivers/pnp/core.c b/drivers/pnp/core.c
> --- a/drivers/pnp/core.c	2003-08-09 04:33:18.000000000 +0000
> +++ b/drivers/pnp/core.c	2003-08-16 16:26:59.000000000 +0000
> @@ -69,7 +69,6 @@
>  
>  	protocol->number = nodenum;
>  	sprintf(protocol->dev.bus_id, "pnp%d", nodenum);
> -	strlcpy(protocol->dev.name,protocol->name,DEVICE_NAME_SIZE);
>  	return device_register(&protocol->dev);
>  }
>  
> diff -urN a/drivers/pnp/isapnp/core.c b/drivers/pnp/isapnp/core.c
> --- a/drivers/pnp/isapnp/core.c	2003-08-09 04:37:19.000000000 +0000
> +++ b/drivers/pnp/isapnp/core.c	2003-08-16 16:45:16.000000000 +0000
> @@ -743,7 +743,7 @@
>  			size = 0;
>  			break;
>  		case _LTAG_ANSISTR:
> -			isapnp_parse_name(dev->dev.name, sizeof(dev->dev.name), &size);
> +			isapnp_parse_name(dev->name, sizeof(dev->name), &size);
>  			break;
>  		case _LTAG_UNICODESTR:
>  			/* silently ignore */
> @@ -808,7 +808,7 @@
>  		case _STAG_VENDOR:
>  			break;
>  		case _LTAG_ANSISTR:
> -			isapnp_parse_name(card->dev.name, sizeof(card->dev.name), &size);
> +			isapnp_parse_name(card->name, sizeof(card->name), &size);
>  			break;
>  		case _LTAG_UNICODESTR:
>  			/* silently ignore */
> @@ -1144,11 +1144,11 @@
>  	protocol_for_each_card(&isapnp_protocol,card) {
>  		cards++;
>  		if (isapnp_verbose) {
> -			printk(KERN_INFO "isapnp: Card '%s'\n", card->dev.name[0]?card->dev.name:"Unknown");
> +			printk(KERN_INFO "isapnp: Card '%s'\n", card->name[0]?card->name:"Unknown");
>  			if (isapnp_verbose < 2)
>  				continue;
>  			card_for_each_dev(card,dev) {
> -				printk(KERN_INFO "isapnp:   Device '%s'\n", dev->dev.name[0]?dev->dev.name:"Unknown");
> +				printk(KERN_INFO "isapnp:   Device '%s'\n", dev->name[0]?dev->name:"Unknown");
>  			}
>  		}
>  	}
> --- a/drivers/serial/8250_pnp.c	2003-08-09 04:31:45.000000000 +0000
> +++ b/drivers/serial/8250_pnp.c	2003-08-16 16:47:21.000000000 +0000
> @@ -369,7 +369,7 @@
>   */
>  static int __devinit serial_pnp_guess_board(struct pnp_dev *dev, int *flags)
>  {
> -	if (!(check_name(dev->dev.name) || (dev->card && check_name(dev->card->dev.name))))
> +	if (!(check_name(pnp_dev_name(dev)) || (dev->card && check_name(dev->card->name))))
>  		return -ENODEV;
>  
>  	if (check_resources(dev->independent))
> --- a/include/linux/pnp.h	2003-08-09 04:42:16.000000000 +0000
> +++ b/include/linux/pnp.h	2003-08-16 16:43:32.000000000 +0000
> @@ -19,6 +19,7 @@
>  #define PNP_MAX_DMA		2
>  #define PNP_MAX_DEVICES		8
>  #define PNP_ID_LEN		8
> +#define PNP_NAME_LEN		50
>  
>  struct pnp_protocol;
>  struct pnp_dev;
> @@ -133,6 +134,7 @@
>  	struct pnp_protocol * protocol;
>  	struct pnp_id * id;		/* contains supported EISA IDs*/
>  
> +	char name[PNP_NAME_LEN];	/* contains a human-readable name */
>  	unsigned char	pnpver;		/* Plug & Play version */
>  	unsigned char	productver;	/* product version */
>  	unsigned int	serial;		/* serial number */
> @@ -187,6 +189,7 @@
>  	struct pnp_option * dependent;
>  	struct pnp_resource_table res;
>  
> +	char name[PNP_NAME_LEN];	/* contains a human-readable name */
>  	unsigned short	regs;		/* ISAPnP: supported registers */
>  	int 		flags;		/* used by protocols */
>  	struct proc_dir_entry *procent;	/* device entry in /proc/bus/isapnp */
> @@ -204,7 +207,7 @@
>  	for((dev) = card_to_pnp_dev((card)->devices.next); \
>  	(dev) != card_to_pnp_dev(&(card)->devices); \
>  	(dev) = card_to_pnp_dev((dev)->card_list.next))
> -#define pnp_dev_name(dev) (dev)->dev.name
> +#define pnp_dev_name(dev) (dev)->name
>  
>  static inline void *pnp_get_drvdata (struct pnp_dev *pdev)
>  {

Your patch works for me, thanks!

Re: 2.6.0-test3 current - firewire compile error

[Greg KH]

On Sat, Aug 16, 2003 at 08:50:19AM -0400, Ivan Gyurdiev wrote:
> Hopefully, this is not a duplicate post:
> ===========================================
> 
> drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
> drivers/ieee1394/nodemgr.c:471: error: structure has no member named `name'
> drivers/ieee1394/nodemgr.c: In function `nodemgr_create_node':
> drivers/ieee1394/nodemgr.c:723: error: structure has no member named `name'
> drivers/ieee1394/nodemgr.c: In function `nodemgr_update_node':
> drivers/ieee1394/nodemgr.c:1318: error: structure has no member named `name'
> drivers/ieee1394/nodemgr.c: In function `nodemgr_add_host':
> drivers/ieee1394/nodemgr.c:1775: error: structure has no member named `name'
> make[2]: *** [drivers/ieee1394/nodemgr.o] Error 1
> make[1]: *** [drivers/ieee1394] Error 2
> make: *** [drivers] Error 2

I removed struct device.name and forgot to change the firewire code :(

I'll work on a patch for this later this evening, unless someone beats
me to it.

Sorry,

greg k-h

Re: 2.6.0-test3 current - firewire compile error

[OGAWA Hirofumi]

Greg KH <greg@kroah.com> writes:

> I removed struct device.name and forgot to change the firewire code :(
> 
> I'll work on a patch for this later this evening, unless someone beats
> me to it.

Why wasn't DEVICE_NAME_SIZE/_HALF killed? Looks like these also should
define by drivers.
-- 
OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>

Re: 2.6.0-test3 current - firewire compile error

[Greg KH]

On Sun, Aug 17, 2003 at 02:38:16AM +0900, OGAWA Hirofumi wrote:
> Greg KH <greg@kroah.com> writes:
> 
> > I removed struct device.name and forgot to change the firewire code :(
> > 
> > I'll work on a patch for this later this evening, unless someone beats
> > me to it.
> 
> Why wasn't DEVICE_NAME_SIZE/_HALF killed? Looks like these also should
> define by drivers.

Some subsystems use that #define for their device name size.  Yeah, it
probably could be gotten rid of too, just didn't really see the need.

Patches are always welcome if you want to fix this up.  :)

thanks,

greg k-h

Re: 2.6.0-test3 current - firewire compile error

[Greg KH]

On Sat, Aug 16, 2003 at 09:35:53AM -0700, Greg KH wrote:
> On Sat, Aug 16, 2003 at 08:50:19AM -0400, Ivan Gyurdiev wrote:
> > Hopefully, this is not a duplicate post:
> > ===========================================
> > 
> > drivers/ieee1394/nodemgr.c: In function `nodemgr_update_ud_names':
> > drivers/ieee1394/nodemgr.c:471: error: structure has no member named `name'
> > drivers/ieee1394/nodemgr.c: In function `nodemgr_create_node':
> > drivers/ieee1394/nodemgr.c:723: error: structure has no member named `name'
> > drivers/ieee1394/nodemgr.c: In function `nodemgr_update_node':
> > drivers/ieee1394/nodemgr.c:1318: error: structure has no member named `name'
> > drivers/ieee1394/nodemgr.c: In function `nodemgr_add_host':
> > drivers/ieee1394/nodemgr.c:1775: error: structure has no member named `name'
> > make[2]: *** [drivers/ieee1394/nodemgr.o] Error 1
> > make[1]: *** [drivers/ieee1394] Error 2
> > make: *** [drivers] Error 2
> 
> I removed struct device.name and forgot to change the firewire code :(
> 
> I'll work on a patch for this later this evening, unless someone beats
> me to it.

Ok, here's the patch that I just sent to Linus and Ben for this, and
some other problems in the ieee1394 code (due to the i2c changes in
2.6.0-test3.)  Please let me know if you have any problems after
applying them.

Again, sorry for missing this.

greg k-h


--- a/drivers/ieee1394/nodemgr.c	Tue Aug  5 12:13:20 2003
+++ b/drivers/ieee1394/nodemgr.c	Sat Aug 16 12:08:09 2003
@@ -460,21 +460,6 @@
 }
 
 
-static void nodemgr_update_ud_names(struct host_info *hi, struct node_entry *ne)
-{
-	struct list_head *lh;
-
-	list_for_each(lh, &ne->device.children) {
-		struct unit_directory *ud;
-		ud = container_of(list_to_dev(lh), struct unit_directory, device);
-
-		snprintf(ud->device.name, DEVICE_NAME_SIZE,
-			 "IEEE-1394 unit directory " NODE_BUS_FMT "-%u",
-			 NODE_BUS_ARGS(hi->host, ne->nodeid), ud->id);
-	}
-}
-
-
 static void nodemgr_remove_ne(struct node_entry *ne)
 {
 	struct device *dev = &ne->device;
@@ -720,9 +705,6 @@
 	ne->device.parent = &host->device;
 	snprintf(ne->device.bus_id, BUS_ID_SIZE, "%016Lx",
 		 (unsigned long long)(ne->guid));
-	snprintf(ne->device.name, DEVICE_NAME_SIZE,
-		 "IEEE-1394 device " NODE_BUS_FMT,
-		 NODE_BUS_ARGS(host, ne->nodeid));
 
 	device_register(&ne->device);
 
@@ -732,8 +714,6 @@
 
 	nodemgr_process_config_rom (hi, ne, busoptions);
 
-	nodemgr_update_ud_names(hi, ne);
-
 	HPSB_DEBUG("%s added: ID:BUS[" NODE_BUS_FMT "]  GUID[%016Lx]",
 		   (host->node_id == nodeid) ? "Host" : "Node",
 		   NODE_BUS_ARGS(host, nodeid), (unsigned long long)guid);
@@ -1312,18 +1292,11 @@
 				struct host_info *hi, nodeid_t nodeid,
 				unsigned int generation)
 {
-	int update_ud_names = 0;
-
 	if (ne->nodeid != nodeid) {
-		snprintf(ne->device.name, DEVICE_NAME_SIZE,
-			 "IEEE-1394 device " NODE_BUS_FMT,
-			 NODE_BUS_ARGS(hi->host, ne->nodeid));
 		HPSB_DEBUG("Node changed: " NODE_BUS_FMT " -> " NODE_BUS_FMT,
 			   NODE_BUS_ARGS(ne->host, ne->nodeid),
 			   NODE_BUS_ARGS(ne->host, nodeid));
 		ne->nodeid = nodeid;
-
-		update_ud_names++;
 	}
 
 	if (ne->busopt.generation != ((busoptions >> 4) & 0xf)) {
@@ -1333,13 +1306,8 @@
 
 		/* This will re-register our unitdir's */
 		nodemgr_process_config_rom (hi, ne, busoptions);
-
-		update_ud_names++;
 	}
 
-	if (update_ud_names)
-		nodemgr_update_ud_names(hi, ne);
-
 	/* Since that's done, we can declare this record current */
 	ne->generation = generation;
 
@@ -1772,8 +1740,6 @@
 	       sizeof(host->device));
 	host->device.parent = &host->pdev->dev;
 	snprintf(host->device.bus_id, BUS_ID_SIZE, "fw-host%d", host->id);
-	snprintf(host->device.name, DEVICE_NAME_SIZE, "IEEE-1394 Host %s-%d",
-		 host->driver->name, host->id);
 
 	sprintf(hi->daemon_name, "knodemgrd_%d", host->id);
 
--- a/drivers/ieee1394/pcilynx.c	Mon Aug 11 09:21:39 2003
+++ b/drivers/ieee1394/pcilynx.c	Sat Aug 16 12:08:53 2003
@@ -144,9 +144,7 @@
 	.id 			= 0xAA, //FIXME: probably we should get an id in i2c-id.h
 	.client_register	= bit_reg,
 	.client_unregister	= bit_unreg,
-	.dev			= {
-		.name		= "PCILynx I2C",
-	},
+	.name			= "PCILynx I2C",
 };