2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Alessandro Suardi]

[-- Attachment #1: Type: text/plain, Size: 599 bytes --]

Dell Latitude D610, FC5-latest, ipw2200 configured to associate
 with a D-Link DSL-G604T (combo of router/ADSL modem/802.11g AP).

2.6.18-git8 (plus semaphore.h) is ok
-git9, -git10, -git11 fail to associate
-git11 with reverted wireless changes is ok

Attaching diff of what I reverted in -git11 to make it work again.

wpa_supplicant log of failing session available upon request.

Thanks, ciao,

--alessandro

"Well a man has two reasons for things that he does
  the first one is pride and the second one is love
  all understandings must come by this way"

     (Husker Du, 'She Floated Away')

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: wifi.diff --]
[-- Type: text/x-patch; name="wifi.diff", Size: 8930 bytes --]

diff -urN linux-2.6.18-git11/drivers/net/wireless/ipw2200.c linux-2.6.18-git11-oldwifi/drivers/net/wireless/ipw2200.c
--- linux-2.6.18-git11/drivers/net/wireless/ipw2200.c	2006-09-29 21:11:19.000000000 +0200
+++ linux-2.6.18-git11-oldwifi/drivers/net/wireless/ipw2200.c	2006-09-29 20:24:33.000000000 +0200
@@ -8875,6 +8875,8 @@
         }
 
 	length = min((int)wrqu->essid.length, IW_ESSID_MAX_SIZE);
+	if (!extra[length - 1])
+		length--;
 
 	priv->config |= CFG_STATIC_ESSID;
 
@@ -8951,7 +8953,7 @@
 	struct ipw_priv *priv = ieee80211_priv(dev);
 	IPW_DEBUG_WX("Getting nick\n");
 	mutex_lock(&priv->mutex);
-	wrqu->data.length = strlen(priv->nick);
+	wrqu->data.length = strlen(priv->nick) + 1;
 	memcpy(extra, priv->nick, wrqu->data.length);
 	wrqu->data.flags = 1;	/* active */
 	mutex_unlock(&priv->mutex);
@@ -9274,9 +9276,9 @@
 		return -EINVAL;
 
 	mutex_lock(&priv->mutex);
-	if (wrqu->retry.flags & IW_RETRY_SHORT)
+	if (wrqu->retry.flags & IW_RETRY_MIN)
 		priv->short_retry_limit = (u8) wrqu->retry.value;
-	else if (wrqu->retry.flags & IW_RETRY_LONG)
+	else if (wrqu->retry.flags & IW_RETRY_MAX)
 		priv->long_retry_limit = (u8) wrqu->retry.value;
 	else {
 		priv->short_retry_limit = (u8) wrqu->retry.value;
@@ -9305,11 +9307,11 @@
 		return -EINVAL;
 	}
 
-	if (wrqu->retry.flags & IW_RETRY_LONG) {
-		wrqu->retry.flags = IW_RETRY_LIMIT | IW_RETRY_LONG;
+	if (wrqu->retry.flags & IW_RETRY_MAX) {
+		wrqu->retry.flags = IW_RETRY_LIMIT | IW_RETRY_MAX;
 		wrqu->retry.value = priv->long_retry_limit;
-	} else if (wrqu->retry.flags & IW_RETRY_SHORT) {
-		wrqu->retry.flags = IW_RETRY_LIMIT | IW_RETRY_SHORT;
+	} else if (wrqu->retry.flags & IW_RETRY_MIN) {
+		wrqu->retry.flags = IW_RETRY_LIMIT | IW_RETRY_MIN;
 		wrqu->retry.value = priv->short_retry_limit;
 	} else {
 		wrqu->retry.flags = IW_RETRY_LIMIT;
diff -urN linux-2.6.18-git11/include/linux/netdevice.h linux-2.6.18-git11-oldwifi/include/linux/netdevice.h
--- linux-2.6.18-git11/include/linux/netdevice.h	2006-09-29 21:11:30.000000000 +0200
+++ linux-2.6.18-git11-oldwifi/include/linux/netdevice.h	2006-09-29 20:27:49.000000000 +0200
@@ -334,6 +334,7 @@
 
 
 	struct net_device_stats* (*get_stats)(struct net_device *dev);
+	struct iw_statistics*   (*get_wireless_stats)(struct net_device *dev);
 
 	/* List of functions to handle Wireless Extensions (instead of ioctl).
 	 * See <net/iw_handler.h> for details. Jean II */
diff -urN linux-2.6.18-git11/net/core/net-sysfs.c linux-2.6.18-git11-oldwifi/net/core/net-sysfs.c
--- linux-2.6.18-git11/net/core/net-sysfs.c	2006-09-29 21:11:31.000000000 +0200
+++ linux-2.6.18-git11-oldwifi/net/core/net-sysfs.c	2006-09-29 20:24:34.000000000 +0200
@@ -344,6 +344,8 @@
 		if(dev->wireless_handlers &&
 		   dev->wireless_handlers->get_wireless_stats)
 			iw = dev->wireless_handlers->get_wireless_stats(dev);
+		else if (dev->get_wireless_stats)
+			iw = dev->get_wireless_stats(dev);
 		if (iw != NULL)
 			ret = (*format)(iw, buf);
 	}
@@ -463,7 +465,8 @@
 		*groups++ = &netstat_group;
 
 #ifdef WIRELESS_EXT
-	if (net->wireless_handlers && net->wireless_handlers->get_wireless_stats)
+	if (net->get_wireless_stats
+	    || (net->wireless_handlers && net->wireless_handlers->get_wireless_stats))
 		*groups++ = &wireless_group;
 #endif
 
diff -urN linux-2.6.18-git11/net/core/wireless.c linux-2.6.18-git11-oldwifi/net/core/wireless.c
--- linux-2.6.18-git11/net/core/wireless.c	2006-09-29 21:11:31.000000000 +0200
+++ linux-2.6.18-git11-oldwifi/net/core/wireless.c	2006-09-29 20:24:35.000000000 +0200
@@ -68,14 +68,6 @@
  *
  * v8 - 17.02.06 - Jean II
  *	o RtNetlink requests support (SET/GET)
- *
- * v8b - 03.08.06 - Herbert Xu
- *	o Fix Wireless Event locking issues.
- *
- * v9 - 14.3.06 - Jean II
- *	o Change length in ESSID and NICK to strlen() instead of strlen()+1
- *	o Make standard_ioctl_num and standard_event_num unsigned
- *	o Remove (struct net_device *)->get_wireless_stats()
  */
 
 /***************************** INCLUDES *****************************/
@@ -242,24 +234,24 @@
 	[SIOCSIWESSID	- SIOCIWFIRST] = {
 		.header_type	= IW_HEADER_TYPE_POINT,
 		.token_size	= 1,
-		.max_tokens	= IW_ESSID_MAX_SIZE,
+		.max_tokens	= IW_ESSID_MAX_SIZE + 1,
 		.flags		= IW_DESCR_FLAG_EVENT,
 	},
 	[SIOCGIWESSID	- SIOCIWFIRST] = {
 		.header_type	= IW_HEADER_TYPE_POINT,
 		.token_size	= 1,
-		.max_tokens	= IW_ESSID_MAX_SIZE,
+		.max_tokens	= IW_ESSID_MAX_SIZE + 1,
 		.flags		= IW_DESCR_FLAG_DUMP,
 	},
 	[SIOCSIWNICKN	- SIOCIWFIRST] = {
 		.header_type	= IW_HEADER_TYPE_POINT,
 		.token_size	= 1,
-		.max_tokens	= IW_ESSID_MAX_SIZE,
+		.max_tokens	= IW_ESSID_MAX_SIZE + 1,
 	},
 	[SIOCGIWNICKN	- SIOCIWFIRST] = {
 		.header_type	= IW_HEADER_TYPE_POINT,
 		.token_size	= 1,
-		.max_tokens	= IW_ESSID_MAX_SIZE,
+		.max_tokens	= IW_ESSID_MAX_SIZE + 1,
 	},
 	[SIOCSIWRATE	- SIOCIWFIRST] = {
 		.header_type	= IW_HEADER_TYPE_PARAM,
@@ -346,8 +338,8 @@
 		.max_tokens	= sizeof(struct iw_pmksa),
 	},
 };
-static const unsigned standard_ioctl_num = (sizeof(standard_ioctl) /
-					    sizeof(struct iw_ioctl_description));
+static const int standard_ioctl_num = (sizeof(standard_ioctl) /
+				       sizeof(struct iw_ioctl_description));
 
 /*
  * Meta-data about all the additional standard Wireless Extension events
@@ -397,8 +389,8 @@
 		.max_tokens	= sizeof(struct iw_pmkid_cand),
 	},
 };
-static const unsigned standard_event_num = (sizeof(standard_event) /
-					    sizeof(struct iw_ioctl_description));
+static const int standard_event_num = (sizeof(standard_event) /
+				       sizeof(struct iw_ioctl_description));
 
 /* Size (in bytes) of the various private data types */
 static const char iw_priv_type_size[] = {
@@ -473,6 +465,17 @@
 	   (dev->wireless_handlers->get_wireless_stats != NULL))
 		return dev->wireless_handlers->get_wireless_stats(dev);
 
+	/* Old location, field to be removed in next WE */
+	if(dev->get_wireless_stats) {
+		static int printed_message;
+
+		if (!printed_message++)
+			printk(KERN_DEBUG "%s (WE) : Driver using old /proc/net/wireless support, please fix driver !\n",
+				dev->name);
+
+		return dev->get_wireless_stats(dev);
+	}
+
 	/* Not found */
 	return (struct iw_statistics *) NULL;
 }
@@ -1840,33 +1843,8 @@
  */
 
 #ifdef WE_EVENT_RTNETLINK
-/* ---------------------------------------------------------------- */
-/*
- * Locking...
- * ----------
- *
- * Thanks to Herbert Xu <herbert@gondor.apana.org.au> for fixing
- * the locking issue in here and implementing this code !
- *
- * The issue : wireless_send_event() is often called in interrupt context,
- * while the Netlink layer can never be called in interrupt context.
- * The fully formed RtNetlink events are queued, and then a tasklet is run
- * to feed those to Netlink.
- * The skb_queue is interrupt safe, and its lock is not held while calling
- * Netlink, so there is no possibility of dealock.
- * Jean II
- */
-
 static struct sk_buff_head wireless_nlevent_queue;
 
-static int __init wireless_nlevent_init(void)
-{
-	skb_queue_head_init(&wireless_nlevent_queue);
-	return 0;
-}
-
-subsys_initcall(wireless_nlevent_init);
-
 static void wireless_nlevent_process(unsigned long data)
 {
 	struct sk_buff *skb;
@@ -1943,6 +1921,13 @@
 	tasklet_schedule(&wireless_nlevent_tasklet);
 }
 
+static int __init wireless_nlevent_init(void)
+{
+	skb_queue_head_init(&wireless_nlevent_queue);
+	return 0;
+}
+
+subsys_initcall(wireless_nlevent_init);
 #endif	/* WE_EVENT_RTNETLINK */
 
 /* ---------------------------------------------------------------- */
diff -urN linux-2.6.18-git11/net/ieee80211/softmac/ieee80211softmac_wx.c linux-2.6.18-git11-oldwifi/net/ieee80211/softmac/ieee80211softmac_wx.c
--- linux-2.6.18-git11/net/ieee80211/softmac/ieee80211softmac_wx.c	2006-09-29 21:11:32.000000000 +0200
+++ linux-2.6.18-git11-oldwifi/net/ieee80211/softmac/ieee80211softmac_wx.c	2006-09-29 20:24:37.000000000 +0200
@@ -80,10 +80,10 @@
 	 * If it's our network, ignore the change, we're already doing it!
 	 */
 	if((sm->associnfo.associating || sm->associated) &&
-	   (data->essid.flags && data->essid.length)) {
+	   (data->essid.flags && data->essid.length && extra)) {
 		/* Get the associating network */
 		n = ieee80211softmac_get_network_by_bssid(sm, sm->associnfo.bssid);
-		if(n && n->essid.len == data->essid.length &&
+		if(n && n->essid.len == (data->essid.length - 1) &&
 		   !memcmp(n->essid.data, extra, n->essid.len)) {
 			dprintk(KERN_INFO PFX "Already associating or associated to "MAC_FMT"\n",
 				MAC_ARG(sm->associnfo.bssid));
@@ -109,8 +109,8 @@
 	sm->associnfo.static_essid = 0;
 	sm->associnfo.assoc_wait = 0;
 
-	if (data->essid.flags && data->essid.length) {
-		length = min((int)data->essid.length, IW_ESSID_MAX_SIZE);
+	if (data->essid.flags && data->essid.length && extra /*required?*/) {
+		length = min(data->essid.length - 1, IW_ESSID_MAX_SIZE);
 		if (length) {
 			memcpy(sm->associnfo.req_essid.data, extra, length);
 			sm->associnfo.static_essid = 1;

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[John W. Linville]

On Fri, Sep 29, 2006 at 09:25:53PM +0200, Alessandro Suardi wrote:
> Dell Latitude D610, FC5-latest, ipw2200 configured to associate
> with a D-Link DSL-G604T (combo of router/ADSL modem/802.11g AP).
> 
> 2.6.18-git8 (plus semaphore.h) is ok
> -git9, -git10, -git11 fail to associate
> -git11 with reverted wireless changes is ok
> 
> Attaching diff of what I reverted in -git11 to make it work again.
> 
> wpa_supplicant log of failing session available upon request.

It looks like you reverted the WE-21 stuff.  Is your wireless-tools
package up to date?

John
-- 
John W. Linville
linville@tuxdriver.com

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Alessandro Suardi]

On 9/29/06, John W. Linville <linville@tuxdriver.com> wrote:
> On Fri, Sep 29, 2006 at 09:25:53PM +0200, Alessandro Suardi wrote:
> > Dell Latitude D610, FC5-latest, ipw2200 configured to associate
> > with a D-Link DSL-G604T (combo of router/ADSL modem/802.11g AP).
> >
> > 2.6.18-git8 (plus semaphore.h) is ok
> > -git9, -git10, -git11 fail to associate
> > -git11 with reverted wireless changes is ok
> >
> > Attaching diff of what I reverted in -git11 to make it work again.
> >
> > wpa_supplicant log of failing session available upon request.
>
> It looks like you reverted the WE-21 stuff.  Is your wireless-tools
> package up to date?

Well, that's the latest I get under FC5:

[asuardi@sandman ~]$ rpm -q wireless-tools
wireless-tools-28-0.pre13.5.1

 but indeed (-git11 minus the reverts) iwconfig says

[asuardi@sandman ~]$ iwconfig eth1
Warning: Driver for device eth1 has been compiled with version 21
of Wireless Extension, while this program supports up to version 19.
Some things may be broken...

The criteria for the revert was actually "take out anything that
 might remotely hit ipw2200 or wireless" - it was basically to
 first rule out that either my WPA configuration or anything in
 my environment could be at fault.

If you have suggestions about either upgrading wireless-tools
 from a non-FC5 repository or narrowing down the reverts, I'm
 up for giving them a go :)


Thanks, ciao,

--alessandro

"Well a man has two reasons for things that he does
  the first one is pride and the second one is love
  all understandings must come by this way"

     (Husker Du, 'She Floated Away')

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Jean Tourrilhes]

On Fri, Sep 29, 2006 at 10:40:29PM +0200, Alessandro Suardi wrote:
> On 9/29/06, John W. Linville <linville@tuxdriver.com> wrote:
> >On Fri, Sep 29, 2006 at 09:25:53PM +0200, Alessandro Suardi wrote:
> >> Dell Latitude D610, FC5-latest, ipw2200 configured to associate
> >> with a D-Link DSL-G604T (combo of router/ADSL modem/802.11g AP).
> >>
> >> 2.6.18-git8 (plus semaphore.h) is ok
> >> -git9, -git10, -git11 fail to associate
> >> -git11 with reverted wireless changes is ok
> >>
> >> Attaching diff of what I reverted in -git11 to make it work again.
> >>
> >> wpa_supplicant log of failing session available upon request.
> >
> >It looks like you reverted the WE-21 stuff.  Is your wireless-tools
> >package up to date?
> 
> Well, that's the latest I get under FC5:
> 
> [asuardi@sandman ~]$ rpm -q wireless-tools
> wireless-tools-28-0.pre13.5.1

	That's too old, the cutoff is 27-pre15.

> but indeed (-git11 minus the reverts) iwconfig says
> 
> [asuardi@sandman ~]$ iwconfig eth1
> Warning: Driver for device eth1 has been compiled with version 21
> of Wireless Extension, while this program supports up to version 19.
> Some things may be broken...

	That's exactly the point of this warning (some distro like to
kill it), I think it spells pretty clearly what's wrong. Don't say I
did not warn you...

> If you have suggestions about either upgrading wireless-tools
> from a non-FC5 repository or narrowing down the reverts, I'm
> up for giving them a go :)

	If you run a custom kernel, I think you won't see any problems
running a custom version of Wireless Tools. They are available on my
web site, pretty easy to install, and have minimal
implications. Usually distro do no customisation of my tools.
	On the other hand, FC6, which is in beta, contains already the
proper version of the tools. I have been monitoring the various distro
in the last few months before sending those WE-21 patches, and all
major distro have WT-28 in the pipeline.
	Actually, you might be able to install the wireless-tools RPM
of FC6 of FC-dev onto your FC5.

> Thanks, ciao,
> 
> --alessandro

	Have fun...

	Jean

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Alessandro Suardi]

On 9/29/06, Jean Tourrilhes <jt@hpl.hp.com> wrote:
> On Fri, Sep 29, 2006 at 10:40:29PM +0200, Alessandro Suardi wrote:
> > On 9/29/06, John W. Linville <linville@tuxdriver.com> wrote:
> > >On Fri, Sep 29, 2006 at 09:25:53PM +0200, Alessandro Suardi wrote:
> > >> Dell Latitude D610, FC5-latest, ipw2200 configured to associate
> > >> with a D-Link DSL-G604T (combo of router/ADSL modem/802.11g AP).
> > >>
> > >> 2.6.18-git8 (plus semaphore.h) is ok
> > >> -git9, -git10, -git11 fail to associate
> > >> -git11 with reverted wireless changes is ok
> > >>
> > >> Attaching diff of what I reverted in -git11 to make it work again.
> > >>
> > >> wpa_supplicant log of failing session available upon request.
> > >
> > >It looks like you reverted the WE-21 stuff.  Is your wireless-tools
> > >package up to date?
> >
> > Well, that's the latest I get under FC5:
> >
> > [asuardi@sandman ~]$ rpm -q wireless-tools
> > wireless-tools-28-0.pre13.5.1
>
>         That's too old, the cutoff is 27-pre15.

Are you sure ? For how I read it, 28-0.pre13.5.1 is more recent
 than 27-pre15, not older.

> > but indeed (-git11 minus the reverts) iwconfig says
> >
> > [asuardi@sandman ~]$ iwconfig eth1
> > Warning: Driver for device eth1 has been compiled with version 21
> > of Wireless Extension, while this program supports up to version 19.
> > Some things may be broken...
>
>         That's exactly the point of this warning (some distro like to
> kill it), I think it spells pretty clearly what's wrong. Don't say I
> did not warn you...

Oh, but I think I've seen these warnings forever in the last years ;)

> > If you have suggestions about either upgrading wireless-tools
> > from a non-FC5 repository or narrowing down the reverts, I'm
> > up for giving them a go :)
>
>         If you run a custom kernel, I think you won't see any problems
> running a custom version of Wireless Tools. They are available on my
> web site, pretty easy to install, and have minimal
> implications. Usually distro do no customisation of my tools.

Indeed - it's not a problem to me :)

However, I downloaded and built the wireless_tools.29.pre10.tar.gz
 sources from your pages and the problem is still there - minus the
 iwconfig warning... more precisely,

 -git8 works
 -git9 fails
 -git11 fails
 -git11 minus the reverts works, and says

[asuardi@sandman ~]$ iwconfig -v
iwconfig  Wireless-Tools version 29
          Compatible with Wireless Extension v11 to v21.

Kernel    Currently compiled with Wireless Extension v21.

eth1      Recommend Wireless Extension v18 or later,
          Currently compiled with Wireless Extension v21.

[asuardi@sandman ~]$ ldd /sbin/iwconfig
        linux-gate.so.1 =>  (0xb7f87000)
        libiw.so.29 => /lib/libiw.so.29 (0xb7f72000)
        libm.so.6 => /lib/libm.so.6 (0x46b0d000)
        libc.so.6 => /lib/libc.so.6 (0x469d8000)
        /lib/ld-linux.so.2 (0x469bb000)

So I guess there's an actual bug that doesn't depend on the
 wireless-tools. Or maybe it's wpa_supplicant that has to be
 upgraded ?

[asuardi@sandman ~]$ rpm -q wpa_supplicant
wpa_supplicant-0.4.8-10.fc5

>         On the other hand, FC6, which is in beta, contains already the
> proper version of the tools. I have been monitoring the various distro
> in the last few months before sending those WE-21 patches, and all
> major distro have WT-28 in the pipeline.

Even if so, wireless-tools would be the only package I have to
 build out of the FC5 distribution to keep up with the latest -git
 snapshot of the Torvalds kernel... I'm not especially troubled
 with this anyway. Perhaps you could push the Fedora folks to
 be a bit more up-to-date with wireless-tools in their current
 main version ?

>         Actually, you might be able to install the wireless-tools RPM
> of FC6 of FC-dev onto your FC5.

Still listening on how to further research the issue... many thanks
 in the meantime, ciao,

--alessandro

"Well a man has two reasons for things that he does
  the first one is pride and the second one is love
  all understandings must come by this way"

     (Husker Du, 'She Floated Away')

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Dave Jones]

On Sat, Sep 30, 2006 at 12:04:31AM +0200, Alessandro Suardi wrote:

 > Even if so, wireless-tools would be the only package I have to
 >  build out of the FC5 distribution to keep up with the latest -git
 >  snapshot of the Torvalds kernel... I'm not especially troubled
 >  with this anyway. Perhaps you could push the Fedora folks to
 >  be a bit more up-to-date with wireless-tools in their current
 >  main version ?

We do. Frequently, because it tends to break in some manner every
single time we push a rebased kernel update.
wireless-tools v28 is in the updates repo right now, which afaik, is
already the newest available.

It's seriously unfunny to have to update fundamental userspace packages
like this when moving to a newer kernel. Especially when say for eg,
the newer kernel then has a broken sound driver, so the user goes
back to their old 'working' kernel. Oops, now they have a choice
of a kernel with broken sound, or a kernel with broken wireless.


	Dave

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Jean Tourrilhes]

On Fri, Sep 29, 2006 at 06:24:10PM -0400, Dave Jones wrote:
> On Sat, Sep 30, 2006 at 12:04:31AM +0200, Alessandro Suardi wrote:
> 
>  > Even if so, wireless-tools would be the only package I have to
>  >  build out of the FC5 distribution to keep up with the latest -git
>  >  snapshot of the Torvalds kernel... I'm not especially troubled
>  >  with this anyway. Perhaps you could push the Fedora folks to
>  >  be a bit more up-to-date with wireless-tools in their current
>  >  main version ?
> 
> We do. Frequently, because it tends to break in some manner every
> single time we push a rebased kernel update.
> wireless-tools v28 is in the updates repo right now, which afaik, is
> already the newest available.

	Correct.

> It's seriously unfunny to have to update fundamental userspace packages
> like this when moving to a newer kernel. Especially when say for eg,
> the newer kernel then has a broken sound driver, so the user goes
> back to their old 'working' kernel. Oops, now they have a choice
> of a kernel with broken sound, or a kernel with broken wireless.

	I'm 100% with you Dave.
	The ESSID change was pretty much forced on me. You can check
the thread on this mailing list starting around 27th of January. Jouni
and Dan were ones of the main advocate of this change. The main reason
is that half of the driver were doing the old way, and the other half
the new way, so we needed to bring back consistency.
	I pushed WT-28 just after the wireless summit, when this
change was reconfirmed, so that it had time to percolate in the
distro. WT-28 works with both kernel before and after the changeover,
so it's transparent. As I say, most distro have it in their dev
version for already some time (even Slackware).
	At the same time, I also looked at other userspace apps. I
looked personally at the source code, and most apps were not affected,
because they wrap around iwconfig or never SET the ESSID.
	The main apps that were affected are wpa_supplicant and
xsupplicant. I told Jouni about required change in wpa_supplicant as
early as 15 March, and reminded him later.
	In other words, I believe I did everything I could do to make
the transition smooth and transparent on userspace. If you think I
could have done better, please tell me. Of course, there are always
people using bleeding edge kernel in old distro, but in theory they
are able to interpret the warning given to them by the Wireless Tools.

> 	Dave

	Have fun...

	Jean

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Jean Tourrilhes]

On Sat, Sep 30, 2006 at 12:04:31AM +0200, Alessandro Suardi wrote:
> On 9/29/06, Jean Tourrilhes <jt@hpl.hp.com> wrote:
> >>
> >> [asuardi@sandman ~]$ rpm -q wireless-tools
> >> wireless-tools-28-0.pre13.5.1
> >
> >        That's too old, the cutoff is 27-pre15.
> 
> Are you sure ? For how I read it, 28-0.pre13.5.1 is more recent
> than 27-pre15, not older.

	Sorry, I'm mixing up my numbers.
	The cutoff for the ESSID fix is 28-pre15, so your version is
just a little bit older. I'm mixing up with the iwpoint cutoff which
was 27-pre25.

> So I guess there's an actual bug that doesn't depend on the
> wireless-tools. Or maybe it's wpa_supplicant that has to be
> upgraded ?

	I don't have the start of the thread, so I don't know the
exact failure mode. If you are using wpa_supplicant, it bypasses the
wireless tools so it would have to be updated.
	Note that I've been pestering Jouni about the fact that he had
to update wpa_supplicant for that since last May, when Jouni himself
asked me to change the ESSID API. Ironic, isn't it ?
	The epitest.fi site seems unfortunately down...

> >        On the other hand, FC6, which is in beta, contains already the
> >proper version of the tools. I have been monitoring the various distro
> >in the last few months before sending those WE-21 patches, and all
> >major distro have WT-28 in the pipeline.
> 
> Even if so, wireless-tools would be the only package I have to
> build out of the FC5 distribution to keep up with the latest -git
> snapshot of the Torvalds kernel... I'm not especially troubled
> with this anyway. Perhaps you could push the Fedora folks to
> be a bit more up-to-date with wireless-tools in their current
> main version ?

	The FC people are busy.

> Still listening on how to further research the issue... many thanks
> in the meantime, ciao,
> 
> --alessandro

	Jean

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Alessandro Suardi]

On 9/30/06, Jean Tourrilhes <jt@hpl.hp.com> wrote:
> On Sat, Sep 30, 2006 at 12:04:31AM +0200, Alessandro Suardi wrote:
> > On 9/29/06, Jean Tourrilhes <jt@hpl.hp.com> wrote:
> > >>
> > >> [asuardi@sandman ~]$ rpm -q wireless-tools
> > >> wireless-tools-28-0.pre13.5.1
> > >
> > >        That's too old, the cutoff is 27-pre15.
> >
> > Are you sure ? For how I read it, 28-0.pre13.5.1 is more recent
> > than 27-pre15, not older.
>
>         Sorry, I'm mixing up my numbers.
>         The cutoff for the ESSID fix is 28-pre15, so your version is
> just a little bit older. I'm mixing up with the iwpoint cutoff which
> was 27-pre25.

OK.

> > So I guess there's an actual bug that doesn't depend on the
> > wireless-tools. Or maybe it's wpa_supplicant that has to be
> > upgraded ?
>
>         I don't have the start of the thread, so I don't know the
> exact failure mode. If you are using wpa_supplicant, it bypasses the
> wireless tools so it would have to be updated.
>         Note that I've been pestering Jouni about the fact that he had
> to update wpa_supplicant for that since last May, when Jouni himself
> asked me to change the ESSID API. Ironic, isn't it ?
>         The epitest.fi site seems unfortunately down...

Yup, same from here. I was about to go downloading and rebuilding
 wpa_supplicant from the 0.4.9 (stable) and failing that from the
 0.5.5 (dev) tarball, but epitest.fi isn't reachable.

> > >        On the other hand, FC6, which is in beta, contains already the
> > >proper version of the tools. I have been monitoring the various distro
> > >in the last few months before sending those WE-21 patches, and all
> > >major distro have WT-28 in the pipeline.
> >
> > Even if so, wireless-tools would be the only package I have to
> > build out of the FC5 distribution to keep up with the latest -git
> > snapshot of the Torvalds kernel... I'm not especially troubled
> > with this anyway. Perhaps you could push the Fedora folks to
> > be a bit more up-to-date with wireless-tools in their current
> > main version ?
>
>         The FC people are busy.

In any case, cc'ing Dave who chipped in earlier in the thread -
 if wpa_supplicant needs to be rebuilt, then it's very likely that
 even FC6 which has 0.4.8-something (just checked ;) will not
 work with the current kernel changes.

I will post an update when I can get hold of the newer sources
 for wpa_supplicant...

Thanks, ciao,

--alessandro

"Well a man has two reasons for things that he does
  the first one is pride and the second one is love
  all understandings must come by this way"

     (Husker Du, 'She Floated Away')

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Alessandro Suardi]

On 9/30/06, Alessandro Suardi <alessandro.suardi@gmail.com> wrote:
> On 9/30/06, Jean Tourrilhes <jt@hpl.hp.com> wrote:
> > On Sat, Sep 30, 2006 at 12:04:31AM +0200, Alessandro Suardi wrote:
> > > On 9/29/06, Jean Tourrilhes <jt@hpl.hp.com> wrote:
> > > >>
> > > >> [asuardi@sandman ~]$ rpm -q wireless-tools
> > > >> wireless-tools-28-0.pre13.5.1
> > > >
> > > >        That's too old, the cutoff is 27-pre15.
> > >
> > > Are you sure ? For how I read it, 28-0.pre13.5.1 is more recent
> > > than 27-pre15, not older.
> >
> >         Sorry, I'm mixing up my numbers.
> >         The cutoff for the ESSID fix is 28-pre15, so your version is
> > just a little bit older. I'm mixing up with the iwpoint cutoff which
> > was 27-pre25.
>
> OK.
>
> > > So I guess there's an actual bug that doesn't depend on the
> > > wireless-tools. Or maybe it's wpa_supplicant that has to be
> > > upgraded ?
> >
> >         I don't have the start of the thread, so I don't know the
> > exact failure mode. If you are using wpa_supplicant, it bypasses the
> > wireless tools so it would have to be updated.
> >         Note that I've been pestering Jouni about the fact that he had
> > to update wpa_supplicant for that since last May, when Jouni himself
> > asked me to change the ESSID API. Ironic, isn't it ?
> >         The epitest.fi site seems unfortunately down...
>
> Yup, same from here. I was about to go downloading and rebuilding
>  wpa_supplicant from the 0.4.9 (stable) and failing that from the
>  0.5.5 (dev) tarball, but epitest.fi isn't reachable.
>
> > > >        On the other hand, FC6, which is in beta, contains already the
> > > >proper version of the tools. I have been monitoring the various distro
> > > >in the last few months before sending those WE-21 patches, and all
> > > >major distro have WT-28 in the pipeline.
> > >
> > > Even if so, wireless-tools would be the only package I have to
> > > build out of the FC5 distribution to keep up with the latest -git
> > > snapshot of the Torvalds kernel... I'm not especially troubled
> > > with this anyway. Perhaps you could push the Fedora folks to
> > > be a bit more up-to-date with wireless-tools in their current
> > > main version ?
> >
> >         The FC people are busy.
>
> In any case, cc'ing Dave who chipped in earlier in the thread -
>  if wpa_supplicant needs to be rebuilt, then it's very likely that
>  even FC6 which has 0.4.8-something (just checked ;) will not
>  work with the current kernel changes.
>
> I will post an update when I can get hold of the newer sources
>  for wpa_supplicant...

Good news, WPA association is back to work for me using
 wireless_tools.29.pre10 and wpa_supplicant-0.4.9 with

 2.6.18-git11 vanilla
 2.6.18-git11 with reverted wireless fixes
 2.6.18-git13

 which appears to mean that backward compatibility of the
 new tools with older kernel features has also been tested :)

Dave, do you want me to file a request for updated FC5 RPMs
 for wireless-tools and wpa_supplicant in bugzilla or is it
  - already happening
  - never going to happen
 ?


Thanks, ciao,

--alessandro

"Well a man has two reasons for things that he does
  the first one is pride and the second one is love
  all understandings must come by this way"

     (Husker Du, 'She Floated Away')

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Dave Jones]

On Sat, Sep 30, 2006 at 02:27:01PM +0200, Alessandro Suardi wrote:
 > 
 > Good news, WPA association is back to work for me using
 >  wireless_tools.29.pre10 and wpa_supplicant-0.4.9 with
 > 
 >  2.6.18-git11 vanilla
 >  2.6.18-git11 with reverted wireless fixes
 >  2.6.18-git13
 > 
 >  which appears to mean that backward compatibility of the
 >  new tools with older kernel features has also been tested :)
 > 
 > Dave, do you want me to file a request for updated FC5 RPMs
 >  for wireless-tools and wpa_supplicant in bugzilla or is it
 >   - already happening
 >   - never going to happen
 >  ?

I'm not sure if we usually wait for a 'real' release, or if there's
precedent for us shipping pre's before, but filing a bug-report
is a good idea so that it doesn't go unnoticed when I eventually
push a 2.6.19 update to FC5-updates in a few months.

	Dave

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Alessandro Suardi]

On 9/30/06, Dave Jones <davej@redhat.com> wrote:
> On Sat, Sep 30, 2006 at 02:27:01PM +0200, Alessandro Suardi wrote:
>  >
>  > Good news, WPA association is back to work for me using
>  >  wireless_tools.29.pre10 and wpa_supplicant-0.4.9 with
>  >
>  >  2.6.18-git11 vanilla
>  >  2.6.18-git11 with reverted wireless fixes
>  >  2.6.18-git13
>  >
>  >  which appears to mean that backward compatibility of the
>  >  new tools with older kernel features has also been tested :)
>  >
>  > Dave, do you want me to file a request for updated FC5 RPMs
>  >  for wireless-tools and wpa_supplicant in bugzilla or is it
>  >   - already happening
>  >   - never going to happen
>  >  ?
>
> I'm not sure if we usually wait for a 'real' release, or if there's
> precedent for us shipping pre's before, but filing a bug-report
> is a good idea so that it doesn't go unnoticed when I eventually
> push a 2.6.19 update to FC5-updates in a few months.
>
>         Dave
>

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=208719
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=208720

respectively filed for wpa_supplicant and wireless-tools.

Thanks, ciao,

--alessandro

"Well a man has two reasons for things that he does
  the first one is pride and the second one is love
  all understandings must come by this way"

     (Husker Du, 'She Floated Away')

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Jean Tourrilhes]

On Sat, Sep 30, 2006 at 02:27:01PM +0200, Alessandro Suardi wrote:
> 
> Good news, WPA association is back to work for me using
> wireless_tools.29.pre10 and wpa_supplicant-0.4.9 with
> 
> 2.6.18-git11 vanilla
> 2.6.18-git11 with reverted wireless fixes
> 2.6.18-git13
> 
> which appears to mean that backward compatibility of the
> new tools with older kernel features has also been tested :)

	We can't really test everything, but at least we try to do the
basics. The strong versioning of the API does help with regards to
this kind of change.

> Dave, do you want me to file a request for updated FC5 RPMs
> for wireless-tools and wpa_supplicant in bugzilla or is it
>  - already happening
>  - never going to happen
> ?

	I believe we are actually pretty close to FC6, which should
have all the right bits...

> Thanks, ciao,
> 
> --alessandro

	Ciao...

	Jean

Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA

[Jouni Malinen]

On Sat, Sep 30, 2006 at 12:04:31AM +0200, Alessandro Suardi wrote:

> So I guess there's an actual bug that doesn't depend on the
> wireless-tools. Or maybe it's wpa_supplicant that has to be
> upgraded ?
> 
> [asuardi@sandman ~]$ rpm -q wpa_supplicant
> wpa_supplicant-0.4.8-10.fc5

2006-05-06 - v0.4.9
 * driver_wext: added support for WE-21 change to SSID configuration

I.e., not really a "bug", but an unfortunate result of making WLAN
drivers use one part of WE consistently..

-- 
Jouni Malinen                                            PGP id EFC895FA

wireless abi breakage (was Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA)

[Pavel Machek]

Hi!

> > >> wpa_supplicant log of failing session available upon request.
> > >
> > >It looks like you reverted the WE-21 stuff.  Is your wireless-tools
> > >package up to date?
> > 
> > Well, that's the latest I get under FC5:
> > 
> > [asuardi@sandman ~]$ rpm -q wireless-tools
> > wireless-tools-28-0.pre13.5.1
> 
> 	That's too old, the cutoff is 27-pre15.
> 
> > but indeed (-git11 minus the reverts) iwconfig says
> > 
> > [asuardi@sandman ~]$ iwconfig eth1
> > Warning: Driver for device eth1 has been compiled with version 21
> > of Wireless Extension, while this program supports up to version 19.
> > Some things may be broken...
> 
> 	That's exactly the point of this warning (some distro like to
> kill it), I think it spells pretty clearly what's wrong. Don't say I
> did not warn you...

Well... we are trying to have stable abi here. Breaking older wireless
tools randomly is *not* okay in the middle of stable series.

> > If you have suggestions about either upgrading wireless-tools
> > from a non-FC5 repository or narrowing down the reverts, I'm
> > up for giving them a go :)
> 
> 	If you run a custom kernel, I think you won't see any problems
> running a custom version of Wireless Tools. They are available on my
> web site, pretty easy to install, and have minimal

No. Kernel abi is stable in 2.6.x.
-- 
Thanks for all the (sleeping) penguins.

Re: wireless abi breakage (was Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA)

[Pekka Enberg]

On 9/30/06, Pavel Machek <pavel@ucw.cz> wrote:
> Well... we are trying to have stable abi here. Breaking older wireless
> tools randomly is *not* okay in the middle of stable series.

Seconded. This is totally unacceptable so please fix up the ABI. You
must follow the proper procedure and add incompatible changes to
Documentation/feature-removal-schedule.txt so people will have enough
to migrate.

Re: wireless abi breakage (was Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA)

[Jean Tourrilhes]

On Sat, Sep 30, 2006 at 07:38:53PM +0000, Pavel Machek wrote:
> Hi!
> 
> > 	That's exactly the point of this warning (some distro like to
> > kill it), I think it spells pretty clearly what's wrong. Don't say I
> > did not warn you...
> 
> Well... we are trying to have stable abi here. Breaking older wireless
> tools randomly is *not* okay in the middle of stable series.

	I'm sorry, but as there is no longer any "devel" serie, to me
there is no longer any "stable" serie. Do you mean that we are going
to get frozen with the same APIs until then end of time ? I don't
think so...
	You can see the glass half-full or half-empty. Maybe you can
see that both Wireless Tools and wpa_supplicant compatible with those
changes were released last May, which means by the time this kernel
change hit the distro, people won't notice...

> > 	If you run a custom kernel, I think you won't see any problems
> > running a custom version of Wireless Tools. They are available on my
> > web site, pretty easy to install, and have minimal
> 
> No. Kernel abi is stable in 2.6.x.

	Sure, I saw for examples a few warning about udev API changes.

	Jean

Re: wireless abi breakage (was Re: 2.6.18-git9 wireless fixes break ipw2200 association to AP with WPA)

[Pavel Machek]

Hi!

> > > 	That's exactly the point of this warning (some distro like to
> > > kill it), I think it spells pretty clearly what's wrong. Don't say I
> > > did not warn you...
> > 
> > Well... we are trying to have stable abi here. Breaking older wireless
> > tools randomly is *not* okay in the middle of stable series.
> 
> 	I'm sorry, but as there is no longer any "devel" serie, to me
> there is no longer any "stable" serie. Do you mean that we are going
> to get frozen with the same APIs until then end of time ? I don't
> think so...

I mean that proprt procedure for removing APIs needs to be followed,
and that is deprecating them in Doc*/feature-removal-schedule, along
with date, waiting a year, then removing them.

> 	You can see the glass half-full or half-empty. Maybe you can

No, not in this case.
-- 
Thanks for all the (sleeping) penguins.