ipw2100 wireless driver

Hi!

What is the status of ipw2100? Is there chance that it would be pushed
into mainline?

I have few problems with that:

* it will not compile with gcc-2.95. Attached patch fixes one problem
but more remain.

--- ipw2100-ofic/ieee80211.h	2004-07-09 06:32:17.000000000 +0200
+++ ipw2100-0.49/ieee80211.h	2004-07-14 13:18:50.000000000 +0200

@@ -440,7 +440,7 @@
 	u16 reserved;
 	u16 frag_size;
 	u16 payload_size;
-	struct sk_buff *fragments[];
+	struct sk_buff *fragments[1];
 };
 
 extern struct ieee80211_txb *ieee80211_skb_to_txb(struct ieee80211_device *ieee, 

* it requires CONFIG_CRYPTO, but fails to force it in Kconfig.

									Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!

On Wed, 14 Jul 2004, Pavel Machek wrote:

Hi!

What is the status of ipw2100? Is there chance that it would be pushed
into mainline?

I have few problems with that:

* it will not compile with gcc-2.95. Attached patch fixes one problem
but more remain.

I've given up hope on that. Don't think it'll ever compile on 2.95. I'm 
using ndiswrapper and it works nicely.

Jeff.

Hi!

quoted

What is the status of ipw2100? Is there chance that it would be pushed
into mainline?

I have few problems with that:

* it will not compile with gcc-2.95. Attached patch fixes one problem
but more remain.

I've given up hope on that. Don't think it'll ever compile on 2.95. I'm 
using ndiswrapper and it works nicely.

No, I think that can be fixed... I'll rather fix ipw2100 than use
ndiswrapper.

								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!

Hi!

What is the status of ipw2100? Is there chance that it would be pushed
into mainline?

Hmm, it has very "nice" interface in proc, where it tells you (in
english *sentences*) if radio kill switch is enabled or not. Ouch.

iwconfig eth1 mode ad-hoc

followed by

iwconfig eth1

makes it crash in ipw2100_wx_get_power called from
wireless_process_ioctl.

								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!

On Jul-14 2004, Wed, 19:48 +0800
Jeff Chua [off-list ref] wrote:

quoted

I have few problems with that:

* it will not compile with gcc-2.95. Attached patch fixes one problem
but more remain.

I've given up hope on that. Don't think it'll ever compile on 2.95. I'm 
using ndiswrapper and it works nicely.

I'd rather fiddle around with gcc versions than risk stack overflows
from running Windows drivers in Linux kernel space.

-- 
Tomas Szepe [off-list ref]

On Wed, Jul 14, 2004 at 01:41:35PM +0200, Pavel Machek wrote:

Hi!

What is the status of ipw2100? Is there chance that it would be pushed
into mainline?

I have few problems with that:

* it will not compile with gcc-2.95. Attached patch fixes one problem
but more remain.

Wouldn't "struct sk_buff **fragments" be a more correct fix?

-- 
Vojtech Pavlik
SuSE Labs, SuSE CR

Hi!

quoted

What is the status of ipw2100? Is there chance that it would be pushed
into mainline?

I have few problems with that:

* it will not compile with gcc-2.95. Attached patch fixes one problem
but more remain.

Wouldn't "struct sk_buff **fragments" be a more correct fix?

Yep, that would certainly be better. I'll wait if I get some  reply
from ipw2100 people. If so I'll update the patch.
								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!

On Jul-14 2004, Wed, 13:55 +0200
Pavel Machek [off-list ref] wrote:

quoted

quoted

What is the status of ipw2100? Is there chance that it would be pushed
into mainline?

I have few problems with that:

* it will not compile with gcc-2.95. Attached patch fixes one problem
but more remain.

I've given up hope on that. Don't think it'll ever compile on 2.95. I'm 
using ndiswrapper and it works nicely.

No, I think that can be fixed... I'll rather fix ipw2100 than use
ndiswrapper.

ipw2100 0.51 from ipw2100.sf.net builds using gcc-2.95.3 "out of the box."
Also make sure to use the attached patch for 2.6.8pre.

-- 
Tomas Szepe [off-list ref]


diff -urN linux-2.6.7/drivers/net/wireless/ipw2100/ipw2100_fw.c linux-2.6.7.x/drivers/net/wireless/ipw2100/ipw2100_fw.c

--- linux-2.6.7/drivers/net/wireless/ipw2100/ipw2100_fw.c	2004-08-09 21:37:11.000000000 +0200
+++ linux-2.6.7.x/drivers/net/wireless/ipw2100/ipw2100_fw.c	2004-08-09 21:36:15.000000000 +0200

@@ -200,7 +200,7 @@
 			goto fail;
 			
 		}
-		if (read(fd, c->buf, c->len) != c->len) {
+		if (sys_read(fd, c->buf, c->len) != c->len) {
 			printk(KERN_INFO "Failed to read chunk firmware "
 			       "chunk %d.\n", i);
 			goto fail;

@@ -231,17 +231,17 @@
 	INIT_LIST_HEAD(&fw->fw.chunk_list);
 	INIT_LIST_HEAD(&fw->uc.chunk_list);
 	
-	fd = open(fn, 0, 0);
+	fd = sys_open(fn, 0, 0);
 	if (fd == -1) {
 		printk(KERN_INFO "Unable to load '%s'.\n", fn);
 		return 1;
 	}
-	l = lseek(fd, 0L, 2);
-	lseek(fd, 0L, 0);
+	l = sys_lseek(fd, 0L, 2);
+	sys_lseek(fd, 0L, 0);
 	
 	IPW2100_DEBUG_FW("Loading %ld bytes for firmware '%s'\n", l, fn);
 	
-	if (read(fd, (char *)&h, sizeof(h)) != sizeof(h)) {
+	if (sys_read(fd, (char *)&h, sizeof(h)) != sizeof(h)) {
 		printk(KERN_INFO "Failed to read '%s'.\n", fn);
 		goto fail;
 	}

@@ -262,12 +262,12 @@
 	if (ipw2100_fw_load(fd, &fw->uc, h.uc_size))
 		goto fail;
 
-	close(fd);
+	sys_close(fd);
 	return 0;
 
  fail:
 	ipw2100_fw_free(fw);
-	close(fd);
+	sys_close(fd);
 	return 1;
 }
 

On Mon, 9 Aug 2004, Tomas Szepe wrote:

ipw2100 0.51 from ipw2100.sf.net builds using gcc-2.95.3 "out of the box."

Well, this is really good news!

I just downloaded 0.51 compiled with gcc-2.95.3 and got it working on my 
IBM X31 with WEP. Even better, 0.51 doesn't need hostap-driver.

Looks good.

Thanks,
Jeff