pull request: wireless-2.6 2008-06-03

pull request: wireless-2.6 2008-06-03

[John W. Linville]

Dave,

A smattering of fixes for 2.6.26...nothing major...

John

---

Individual patches are available here:

	http://www.kernel.org/pub/linux/kernel/people/linville/wireless-2.6/

---

The following changes since commit 78b8c91d226f2bc1fae8cbdab7de8b8367ad758b:
  Adam Langley (1):
        TCP: Fix corrupt packets when options space overflows with MD5SIG enabled

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-2.6.git master

Andrew Morton (1):
      airo warning fix

Dan Williams (1):
      ipw2200: expire and use oldest BSS on adhoc create

Felix Homann (1):
      USB ID for Philips CPWUA054/00 Wireless USB Adapter 11g

Holger Schurig (1):
      libertas: fix command size for CMD_802_11_SUBSCRIBE_EVENT

Kirill A. Shutemov (1):
      wireless.h: improve userland include-ability

Michael Buesch (2):
      b43legacy: Fix controller restart crash
      ssb: Fix context assertion in ssb_pcicore_dev_irqvecs_enable

 drivers/net/wireless/airo.c             |    2 +-
 drivers/net/wireless/b43legacy/main.c   |   17 +++++++++--------
 drivers/net/wireless/ipw2200.c          |   27 +++++++++++++++++++++++++--
 drivers/net/wireless/libertas/debugfs.c |    4 ++--
 drivers/net/wireless/p54/p54usb.c       |    1 +
 drivers/ssb/driver_pcicore.c            |    4 ++--
 include/linux/wireless.h                |    6 ++++++
 7 files changed, 46 insertions(+), 15 deletions(-)

diff --git a/drivers/net/wireless/airo.c b/drivers/net/wireless/airo.c
index 4e1c690..32019fb 100644
--- a/drivers/net/wireless/airo.c
+++ b/drivers/net/wireless/airo.c
@@ -2905,7 +2905,7 @@ EXPORT_SYMBOL(init_airo_card);
 
 static int waitbusy (struct airo_info *ai) {
 	int delay = 0;
-	while ((IN4500 (ai, COMMAND) & COMMAND_BUSY) && (delay < 10000)) {
+	while ((IN4500(ai, COMMAND) & COMMAND_BUSY) && (delay < 10000)) {
 		udelay (10);
 		if ((++delay % 20) == 0)
 			OUT4500(ai, EVACK, EV_CLEARCOMMANDBUSY);
diff --git a/drivers/net/wireless/b43legacy/main.c b/drivers/net/wireless/b43legacy/main.c
index 14a5eea..204077c 100644
--- a/drivers/net/wireless/b43legacy/main.c
+++ b/drivers/net/wireless/b43legacy/main.c
@@ -3039,7 +3039,6 @@ static void b43legacy_set_pretbtt(struct b43legacy_wldev *dev)
 /* Locking: wl->mutex */
 static void b43legacy_wireless_core_exit(struct b43legacy_wldev *dev)
 {
-	struct b43legacy_wl *wl = dev->wl;
 	struct b43legacy_phy *phy = &dev->phy;
 	u32 macctl;
 
@@ -3054,12 +3053,6 @@ static void b43legacy_wireless_core_exit(struct b43legacy_wldev *dev)
 	macctl |= B43legacy_MACCTL_PSM_JMP0;
 	b43legacy_write32(dev, B43legacy_MMIO_MACCTL, macctl);
 
-	mutex_unlock(&wl->mutex);
-	/* Must unlock as it would otherwise deadlock. No races here.
-	 * Cancel possibly pending workqueues. */
-	cancel_work_sync(&dev->restart_work);
-	mutex_lock(&wl->mutex);
-
 	b43legacy_leds_exit(dev);
 	b43legacy_rng_exit(dev->wl);
 	b43legacy_pio_free(dev);
@@ -3486,6 +3479,8 @@ static void b43legacy_chip_reset(struct work_struct *work)
 		}
 	}
 out:
+	if (err)
+		wl->current_dev = NULL; /* Failed to init the dev. */
 	mutex_unlock(&wl->mutex);
 	if (err)
 		b43legacyerr(wl, "Controller restart FAILED\n");
@@ -3618,9 +3613,11 @@ static void b43legacy_one_core_detach(struct ssb_device *dev)
 	struct b43legacy_wldev *wldev;
 	struct b43legacy_wl *wl;
 
+	/* Do not cancel ieee80211-workqueue based work here.
+	 * See comment in b43legacy_remove(). */
+
 	wldev = ssb_get_drvdata(dev);
 	wl = wldev->wl;
-	cancel_work_sync(&wldev->restart_work);
 	b43legacy_debugfs_remove_device(wldev);
 	b43legacy_wireless_core_detach(wldev);
 	list_del(&wldev->list);
@@ -3789,6 +3786,10 @@ static void b43legacy_remove(struct ssb_device *dev)
 	struct b43legacy_wl *wl = ssb_get_devtypedata(dev);
 	struct b43legacy_wldev *wldev = ssb_get_drvdata(dev);
 
+	/* We must cancel any work here before unregistering from ieee80211,
+	 * as the ieee80211 unreg will destroy the workqueue. */
+	cancel_work_sync(&wldev->restart_work);
+
 	B43legacy_WARN_ON(!wl);
 	if (wl->current_dev == wldev)
 		ieee80211_unregister_hw(wl->hw);
diff --git a/drivers/net/wireless/ipw2200.c b/drivers/net/wireless/ipw2200.c
index d74c061..7293367 100644
--- a/drivers/net/wireless/ipw2200.c
+++ b/drivers/net/wireless/ipw2200.c
@@ -7558,8 +7558,31 @@ static int ipw_associate(void *data)
 	    priv->ieee->iw_mode == IW_MODE_ADHOC &&
 	    priv->config & CFG_ADHOC_CREATE &&
 	    priv->config & CFG_STATIC_ESSID &&
-	    priv->config & CFG_STATIC_CHANNEL &&
-	    !list_empty(&priv->ieee->network_free_list)) {
+	    priv->config & CFG_STATIC_CHANNEL) {
+		/* Use oldest network if the free list is empty */
+		if (list_empty(&priv->ieee->network_free_list)) {
+			struct ieee80211_network *oldest = NULL;
+			struct ieee80211_network *target;
+			DECLARE_MAC_BUF(mac);
+
+			list_for_each_entry(target, &priv->ieee->network_list, list) {
+				if ((oldest == NULL) ||
+				    (target->last_scanned < oldest->last_scanned))
+					oldest = target;
+			}
+
+			/* If there are no more slots, expire the oldest */
+			list_del(&oldest->list);
+			target = oldest;
+			IPW_DEBUG_ASSOC("Expired '%s' (%s) from "
+					"network list.\n",
+					escape_essid(target->ssid,
+						     target->ssid_len),
+					print_mac(mac, target->bssid));
+			list_add_tail(&target->list,
+				      &priv->ieee->network_free_list);
+		}
+
 		element = priv->ieee->network_free_list.next;
 		network = list_entry(element, struct ieee80211_network, list);
 		ipw_adhoc_create(priv, network);
diff --git a/drivers/net/wireless/libertas/debugfs.c b/drivers/net/wireless/libertas/debugfs.c
index ad2fabc..0aa0ce3 100644
--- a/drivers/net/wireless/libertas/debugfs.c
+++ b/drivers/net/wireless/libertas/debugfs.c
@@ -312,8 +312,8 @@ static ssize_t lbs_threshold_write(uint16_t tlv_type, uint16_t event_mask,
 	if (tlv_type != TLV_TYPE_BCNMISS)
 		tlv->freq = freq;
 
-	/* The command header, the event mask, and the one TLV */
-	events->hdr.size = cpu_to_le16(sizeof(events->hdr) + 2 + sizeof(*tlv));
+	/* The command header, the action, the event mask, and one TLV */
+	events->hdr.size = cpu_to_le16(sizeof(events->hdr) + 4 + sizeof(*tlv));
 
 	ret = lbs_cmd_with_response(priv, CMD_802_11_SUBSCRIBE_EVENT, events);
 
diff --git a/drivers/net/wireless/p54/p54usb.c b/drivers/net/wireless/p54/p54usb.c
index 98ddbb3..1610a73 100644
--- a/drivers/net/wireless/p54/p54usb.c
+++ b/drivers/net/wireless/p54/p54usb.c
@@ -49,6 +49,7 @@ static struct usb_device_id p54u_table[] __devinitdata = {
 	{USB_DEVICE(0x5041, 0x2235)},	/* Linksys WUSB54G Portable */
 
 	/* Version 2 devices (3887) */
+	{USB_DEVICE(0x0471, 0x1230)},   /* Philips CPWUA054/00 */
 	{USB_DEVICE(0x050d, 0x7050)},	/* Belkin F5D7050 ver 1000 */
 	{USB_DEVICE(0x0572, 0x2000)},	/* Cohiba Proto board */
 	{USB_DEVICE(0x0572, 0x2002)},	/* Cohiba Proto board */
diff --git a/drivers/ssb/driver_pcicore.c b/drivers/ssb/driver_pcicore.c
index 75def13..d28c538 100644
--- a/drivers/ssb/driver_pcicore.c
+++ b/drivers/ssb/driver_pcicore.c
@@ -537,12 +537,12 @@ int ssb_pcicore_dev_irqvecs_enable(struct ssb_pcicore *pc,
 	int err = 0;
 	u32 tmp;
 
-	might_sleep();
-
 	if (!pdev)
 		goto out;
 	bus = pdev->bus;
 
+	might_sleep_if(pdev->id.coreid != SSB_DEV_PCI);
+
 	/* Enable interrupts for this device. */
 	if (bus->host_pci &&
 	    ((pdev->id.revision >= 6) || (pdev->id.coreid == SSB_DEV_PCIE))) {
diff --git a/include/linux/wireless.h b/include/linux/wireless.h
index 0a9b5b4..9c0b89c 100644
--- a/include/linux/wireless.h
+++ b/include/linux/wireless.h
@@ -70,8 +70,14 @@
 /***************************** INCLUDES *****************************/
 
 #include <linux/types.h>		/* for __u* and __s* typedefs */
+
+/* This header is used in user-space, therefore need to be sanitised
+ * for that purpose. Those includes are usually not compatible with glibc.
+ * To know which includes to use in user-space, check iwlib.h. */
+#ifdef __KERNEL__
 #include <linux/socket.h>		/* for "struct sockaddr" et al	*/
 #include <linux/if.h>			/* for IFNAMSIZ and co... */
+#endif	/* __KERNEL__ */
 
 /***************************** VERSION *****************************/
 /*
-- 
John W. Linville
linville-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: pull request: wireless-2.6 2008-06-03

[David Miller]

From: "John W. Linville" <linville-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org>
Date: Tue, 3 Jun 2008 15:09:43 -0400

> Kirill A. Shutemov (1):
>       wireless.h: improve userland include-ability

I'm not pulling this into my tree.  I've already replied
to that patch posting saying why I don't like this.

Even if I found this change as the only way to go and therefore
had to take it in, the commit log is way too terse about what
the problem is.  And it must describe the horrifics that are
now necessary to include this file in userspace after the change.

It needs to describe in detail what the conflicting types and
definitions in the kernel headers are that cause problems with
userspace, and the exact sequence of includes userland has to make in
order to get this right.

And I require such enormously scary details in order to make it clear
how bogus this change is.

There is zero point in us sanitizing kernel headers for userspace
usage if this kind of crap is still necessary.  I mean tell me, why
should we even bother in that case?  And this is why I'm not taking in
this changeset.
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: pull request: wireless-2.6 2008-06-03

[John W. Linville]

On Tue, Jun 03, 2008 at 02:16:52PM -0700, David Miller wrote:
> From: "John W. Linville" <linville@tuxdriver.com>
> Date: Tue, 3 Jun 2008 15:09:43 -0400
> 
> > Kirill A. Shutemov (1):
> >       wireless.h: improve userland include-ability
> 
> I'm not pulling this into my tree.  I've already replied
> to that patch posting saying why I don't like this.

For the record, I didn't get that posting until just _after_ I got this
reply from you -- a mail server somewhere must have had indigestion!

> Even if I found this change as the only way to go and therefore
> had to take it in, the commit log is way too terse about what
> the problem is.  And it must describe the horrifics that are
> now necessary to include this file in userspace after the change.

FWIW, I'll take the blame for the overly terse commit log...

> There is zero point in us sanitizing kernel headers for userspace
> usage if this kind of crap is still necessary.  I mean tell me, why
> should we even bother in that case?  And this is why I'm not taking in
> this changeset.

Please see the discussion on the other thread, in case you have
something to add or something there persuades you.  In the meantime,
I retract this pull request and will soon follow with a rebased pull
request that expunges the wireless.h commit.

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