[PATCH 0/2] *** SUBJECT HERE ***

[PATCH 0/2] *** SUBJECT HERE ***

[sukeshs]

From: Sukesh Srikakula <sukeshs@xl-sj1-20.sj.broadcom.com>

*** BLURB HERE ***

Sukesh Srikakula (2):
  brcm80211: FIX for TKIP GTK bug
  brcm80211: Fix for suspend/resume bug

 drivers/staging/brcm80211/brcmfmac/bcmsdh_sdmmc.c |    4 ----
 drivers/staging/brcm80211/brcmfmac/dhd.h          |    3 +++
 drivers/staging/brcm80211/brcmfmac/dhd_linux.c    |    6 ++++--
 drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c  |   12 ++++++++++++
 4 files changed, 19 insertions(+), 6 deletions(-)

-- 
1.7.3.2

[PATCH 1/2] brcm80211: FIX for TKIP GTK bug

[sukeshs]

From: Sukesh Srikakula <sukeshs@xl-sj1-20.sj.broadcom.com>

The last 16 bytes of the key sent by cfg80211 needs to be swapped before installing it to FW for TKIP encryption.
This is not done for the group key in current code, which is corrected with this fix.

Signed-off-by: Sukesh Srikakula <sukeshs@broadcom.com>
Signed-off-by: Sukesh Srikakula <sukeshs@xl-sj1-20.sj.broadcom.com>
---
 drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c b/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c
index b44daf9..c60fc7c 100644
--- a/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c
+++ b/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c
@@ -1616,6 +1616,7 @@ wl_cfg80211_add_key(struct wiphy *wiphy, struct net_device *dev,
 	s32 val;
 	s32 wsec;
 	s32 err = 0;
+	u8 keybuf[8];
 
 	WL_DBG("key index (%d)\n", key_idx);
 	CHECK_SYS_UP();
@@ -1644,6 +1645,9 @@ wl_cfg80211_add_key(struct wiphy *wiphy, struct net_device *dev,
 		WL_DBG("WLAN_CIPHER_SUITE_WEP104\n");
 		break;
 	case WLAN_CIPHER_SUITE_TKIP:
+		memcpy(keybuf, &key.data[24], sizeof(keybuf));
+		memcpy(&key.data[24], &key.data[16], sizeof(keybuf));
+		memcpy(&key.data[16], keybuf, sizeof(keybuf));
 		key.algo = CRYPTO_ALGO_TKIP;
 		WL_DBG("WLAN_CIPHER_SUITE_TKIP\n");
 		break;
-- 
1.7.3.2

[PATCH 2/2] brcm80211: Fix for suspend/resume bug

[sukeshs]

From: Sukesh Srikakula <sukeshs@xl-sj1-20.sj.broadcom.com>

Currently, there are 2 callbacks registered with OS for getting notifications when system goes to suspend/resume.
Racing between these 2 callbacks leads to failure in the suspend/resume path.
With this fix, we avoid registering dhd callback for suspend/resume notification when cfg80211 is used. Relevant functionality in dhd suspend/resume callback function is moved to cfg80211 suspend/resume functions.

Signed-off-by: Sukesh Srikakula <sukeshs@broadcom.com>
Signed-off-by: Sukesh Srikakula <sukeshs@xl-sj1-20.sj.broadcom.com>
---
 drivers/staging/brcm80211/brcmfmac/bcmsdh_sdmmc.c |    4 ----
 drivers/staging/brcm80211/brcmfmac/dhd.h          |    3 +++
 drivers/staging/brcm80211/brcmfmac/dhd_linux.c    |    6 ++++--
 drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c  |    8 ++++++++
 4 files changed, 15 insertions(+), 6 deletions(-)

diff --git a/drivers/staging/brcm80211/brcmfmac/bcmsdh_sdmmc.c b/drivers/staging/brcm80211/brcmfmac/bcmsdh_sdmmc.c
index 05f50d3..54ba30d 100644
--- a/drivers/staging/brcm80211/brcmfmac/bcmsdh_sdmmc.c
+++ b/drivers/staging/brcm80211/brcmfmac/bcmsdh_sdmmc.c
@@ -30,10 +30,6 @@
 #include <dngl_stats.h>
 #include <dhd.h>
 
-#if defined(CONFIG_PM_SLEEP)
-#include <linux/suspend.h>
-extern volatile bool dhd_mmc_suspend;
-#endif
 #include "bcmsdh_sdmmc.h"
 
 extern int sdio_function_init(void);
diff --git a/drivers/staging/brcm80211/brcmfmac/dhd.h b/drivers/staging/brcm80211/brcmfmac/dhd.h
index 99c38dd..41c55a6 100644
--- a/drivers/staging/brcm80211/brcmfmac/dhd.h
+++ b/drivers/staging/brcm80211/brcmfmac/dhd.h
@@ -31,6 +31,7 @@
 #include <linux/random.h>
 #include <linux/spinlock.h>
 #include <linux/ethtool.h>
+#include <linux/suspend.h>
 #include <asm/uaccess.h>
 #include <asm/unaligned.h>
 /* The kernel threading is sdio-specific */
@@ -123,6 +124,8 @@ typedef struct dhd_pub {
 
 #if defined(CONFIG_PM_SLEEP)
 
+extern volatile bool dhd_mmc_suspend;
+
 #define DHD_PM_RESUME_WAIT_INIT(a) DECLARE_WAIT_QUEUE_HEAD(a);
 #define _DHD_PM_RESUME_WAIT(a, b) do {\
 			int retry = 0; \
diff --git a/drivers/staging/brcm80211/brcmfmac/dhd_linux.c b/drivers/staging/brcm80211/brcmfmac/dhd_linux.c
index b34c4ea..37b2c0c 100644
--- a/drivers/staging/brcm80211/brcmfmac/dhd_linux.c
+++ b/drivers/staging/brcm80211/brcmfmac/dhd_linux.c
@@ -2015,7 +2015,8 @@ dhd_pub_t *dhd_attach(struct dhd_bus *bus, uint bus_hdrlen)
 	g_bus = bus;
 #endif
 #if defined(CONFIG_PM_SLEEP)
-	register_pm_notifier(&dhd_sleep_pm_notifier);
+	if (!IS_CFG80211_FAVORITE())
+		register_pm_notifier(&dhd_sleep_pm_notifier);
 #endif	/* defined(CONFIG_PM_SLEEP) */
 	/* && defined(DHD_GPL) */
 	/* Init lock suspend to prevent kernel going to suspend */
@@ -2325,7 +2326,8 @@ void dhd_detach(dhd_pub_t *dhdp)
 				wl_cfg80211_detach();
 
 #if defined(CONFIG_PM_SLEEP)
-			unregister_pm_notifier(&dhd_sleep_pm_notifier);
+			if (!IS_CFG80211_FAVORITE())
+				unregister_pm_notifier(&dhd_sleep_pm_notifier);
 #endif	/* defined(CONFIG_PM_SLEEP) */
 			/* && defined(DHD_GPL) */
 			free_netdev(ifp->net);
diff --git a/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c b/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c
index c60fc7c..db77e60 100644
--- a/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c
+++ b/drivers/staging/brcm80211/brcmfmac/wl_cfg80211.c
@@ -1971,6 +1971,10 @@ static s32 wl_cfg80211_resume(struct wiphy *wiphy)
 {
 	s32 err = 0;
 
+#if defined(CONFIG_PM_SLEEP)
+	dhd_mmc_suspend = false;
+#endif /*  defined(CONFIG_PM_SLEEP) */
+
 	CHECK_SYS_UP();
 	wl_invoke_iscan(wiphy_to_wl(wiphy));
 
@@ -1996,6 +2000,10 @@ static s32 wl_cfg80211_suspend(struct wiphy *wiphy)
 
 	sdioh_sdio_set_host_pm_flags(MMC_PM_KEEP_POWER);
 
+#if defined(CONFIG_PM_SLEEP)
+	dhd_mmc_suspend = true;
+#endif /*  defined(CONFIG_PM_SLEEP) */
+
 	return err;
 }
 
-- 
1.7.3.2

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Arend van Spriel]

On 05/03/2011 09:00 AM, sukeshs@broadcom.com wrote:
> From: Sukesh Srikakula<sukeshs@xl-sj1-20.sj.broadcom.com>
>
> Currently, there are 2 callbacks registered with OS for getting notifications when system goes to suspend/resume.
> Racing between these 2 callbacks leads to failure in the suspend/resume path.
> With this fix, we avoid registering dhd callback for suspend/resume notification when cfg80211 is used. Relevant functionality in dhd suspend/resume callback function is moved to cfg80211 suspend/resume functions.
>
> Signed-off-by: Sukesh Srikakula<sukeshs@broadcom.com>
> Signed-off-by: Sukesh Srikakula<sukeshs@xl-sj1-20.sj.broadcom.com>
> ---
> diff --git a/drivers/staging/brcm80211/brcmfmac/dhd.h b/drivers/staging/brcm80211/brcmfmac/dhd.h
> index 99c38dd..41c55a6 100644
> --- a/drivers/staging/brcm80211/brcmfmac/dhd.h
> +++ b/drivers/staging/brcm80211/brcmfmac/dhd.h
> @@ -123,6 +124,8 @@ typedef struct dhd_pub {
>
>   #if defined(CONFIG_PM_SLEEP)
>
> +extern volatile bool dhd_mmc_suspend;
> +

You probably should run checkpatch.pl on this patch. The volatile 
keyword triggers a warning here. I grepped the fullmac code and this 
variable is only set so what is its purpose (debugging?)? See 
Documentation/volatile-considered-harmful.txt on this topic.

Gr. AvS

Re: [PATCH 0/2] *** SUBJECT HERE ***

[Greg KH]

On Tue, May 03, 2011 at 12:00:55AM -0700, sukeshs@broadcom.com wrote:
> From: Sukesh Srikakula <sukeshs@xl-sj1-20.sj.broadcom.com>
> 
> *** BLURB HERE ***

I think you forgot the subject and the BLURB :(

care to retry?

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Greg KH]

On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
>  #if defined(CONFIG_PM_SLEEP)
>  
> +extern volatile bool dhd_mmc_suspend;

This does not do what you think it does.

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Arend van Spriel]

On 05/03/2011 02:15 PM, Greg KH wrote:
> On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
>>   #if defined(CONFIG_PM_SLEEP)
>>
>> +extern volatile bool dhd_mmc_suspend;
> This does not do what you think it does.

Hi Greg,

I posted the 'catchup' patch set and I rebased it on '[PATCH 1/2] 
brcm80211: FIX for TKIP GTK bug' from Sukesh. The only other two patches 
I saw on the linuxdriver list are:

staging: brcm80211: brcmfmac: Add and use dhd_dbg [joe@perches.com] => 
he rejected it himself.
STAGING: brcm80211 SDIO/MMC cleanups [grundler@chromium.org] => does not 
apply.

So I discarded those two as well.

Gr. AvS

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Greg KH]

On Tue, May 03, 2011 at 02:33:44PM +0200, Arend van Spriel wrote:
> On 05/03/2011 02:15 PM, Greg KH wrote:
> >On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
> >>  #if defined(CONFIG_PM_SLEEP)
> >>
> >>+extern volatile bool dhd_mmc_suspend;
> >This does not do what you think it does.
> 
> Hi Greg,
> 
> I posted the 'catchup' patch set and I rebased it on '[PATCH 1/2]
> brcm80211: FIX for TKIP GTK bug' from Sukesh. The only other two
> patches I saw on the linuxdriver list are:

You rebased on a 1/2 patch in a series with an invalid 0/2 email?

{sigh}

I dropped that patch as the whole series needed to be resent, do you
want me to apply that one patch before yours?

thanks,

greg k-h

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Arend van Spriel]

On 05/03/2011 03:00 PM, Greg KH wrote:
> On Tue, May 03, 2011 at 02:33:44PM +0200, Arend van Spriel wrote:
>> On 05/03/2011 02:15 PM, Greg KH wrote:
>>> On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
>>>>   #if defined(CONFIG_PM_SLEEP)
>>>>
>>>> +extern volatile bool dhd_mmc_suspend;
>>> This does not do what you think it does.
>> Hi Greg,
>>
>> I posted the 'catchup' patch set and I rebased it on '[PATCH 1/2]
>> brcm80211: FIX for TKIP GTK bug' from Sukesh. The only other two
>> patches I saw on the linuxdriver list are:
> You rebased on a 1/2 patch in a series with an invalid 0/2 email?
>
> {sigh}
Sorry.
> I dropped that patch as the whole series needed to be resent, do you
> want me to apply that one patch before yours?
I can resend my series if that is more convenient for you.

Gr. AvS

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Greg KH]

On Tue, May 03, 2011 at 03:09:20PM +0200, Arend van Spriel wrote:
> On 05/03/2011 03:00 PM, Greg KH wrote:
> >On Tue, May 03, 2011 at 02:33:44PM +0200, Arend van Spriel wrote:
> >>On 05/03/2011 02:15 PM, Greg KH wrote:
> >>>On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
> >>>>  #if defined(CONFIG_PM_SLEEP)
> >>>>
> >>>>+extern volatile bool dhd_mmc_suspend;
> >>>This does not do what you think it does.
> >>Hi Greg,
> >>
> >>I posted the 'catchup' patch set and I rebased it on '[PATCH 1/2]
> >>brcm80211: FIX for TKIP GTK bug' from Sukesh. The only other two
> >>patches I saw on the linuxdriver list are:
> >You rebased on a 1/2 patch in a series with an invalid 0/2 email?
> >
> >{sigh}
> Sorry.
> >I dropped that patch as the whole series needed to be resent, do you
> >want me to apply that one patch before yours?
> I can resend my series if that is more convenient for you.

Please, but just send it to me, I doubt everyone wants to see 60+
patches again on the list.

thanks,

greg k-h

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Arend van Spriel]

On 05/03/2011 03:11 PM, Greg KH wrote:
> On Tue, May 03, 2011 at 03:09:20PM +0200, Arend van Spriel wrote:
>> On 05/03/2011 03:00 PM, Greg KH wrote:
>>> On Tue, May 03, 2011 at 02:33:44PM +0200, Arend van Spriel wrote:
>>>> On 05/03/2011 02:15 PM, Greg KH wrote:
>>>>> On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
>>>>>>   #if defined(CONFIG_PM_SLEEP)
>>>>>>
>>>>>> +extern volatile bool dhd_mmc_suspend;
>>>>> This does not do what you think it does.
>>>> Hi Greg,
>>>>
>>>> I posted the 'catchup' patch set and I rebased it on '[PATCH 1/2]
>>>> brcm80211: FIX for TKIP GTK bug' from Sukesh. The only other two
>>>> patches I saw on the linuxdriver list are:
>>> You rebased on a 1/2 patch in a series with an invalid 0/2 email?
>>>
>>> {sigh}
>> Sorry.
>>> I dropped that patch as the whole series needed to be resent, do you
>>> want me to apply that one patch before yours?
>> I can resend my series if that is more convenient for you.
> Please, but just send it to me, I doubt everyone wants to see 60+
> patches again on the list.
Ok. Just for this time I will agree to break the rules ;-)

Gr. AvS

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Arend van Spriel]

On 05/03/2011 03:17 PM, Arend van Spriel wrote:
> On 05/03/2011 03:11 PM, Greg KH wrote:
>> On Tue, May 03, 2011 at 03:09:20PM +0200, Arend van Spriel wrote:
>>> On 05/03/2011 03:00 PM, Greg KH wrote:
>>>> On Tue, May 03, 2011 at 02:33:44PM +0200, Arend van Spriel wrote:
>>>>> On 05/03/2011 02:15 PM, Greg KH wrote:
>>>>>> On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
>>>>>>>    #if defined(CONFIG_PM_SLEEP)
>>>>>>>
>>>>>>> +extern volatile bool dhd_mmc_suspend;
>>>>>> This does not do what you think it does.
>>>>> Hi Greg,
>>>>>
>>>>> I posted the 'catchup' patch set and I rebased it on '[PATCH 1/2]
>>>>> brcm80211: FIX for TKIP GTK bug' from Sukesh. The only other two
>>>>> patches I saw on the linuxdriver list are:
>>>> You rebased on a 1/2 patch in a series with an invalid 0/2 email?
>>>>
>>>> {sigh}
>>> Sorry.
>>>> I dropped that patch as the whole series needed to be resent, do you
>>>> want me to apply that one patch before yours?
>>> I can resend my series if that is more convenient for you.
>> Please, but just send it to me, I doubt everyone wants to see 60+
>> patches again on the list.
> Ok. Just for this time I will agree to break the rules ;-)

Hi Greg,

I rebase and regenerated the patch files. Did a diff before resending 
them with the previous set and there are no relevant differences so the 
patches I sent this morning are ok.

Gr. AvS

Re: [PATCH 2/2] brcm80211: Fix for suspend/resume bug

[Greg KH]

On Tue, May 03, 2011 at 03:41:50PM +0200, Arend van Spriel wrote:
> On 05/03/2011 03:17 PM, Arend van Spriel wrote:
> >On 05/03/2011 03:11 PM, Greg KH wrote:
> >>On Tue, May 03, 2011 at 03:09:20PM +0200, Arend van Spriel wrote:
> >>>On 05/03/2011 03:00 PM, Greg KH wrote:
> >>>>On Tue, May 03, 2011 at 02:33:44PM +0200, Arend van Spriel wrote:
> >>>>>On 05/03/2011 02:15 PM, Greg KH wrote:
> >>>>>>On Tue, May 03, 2011 at 12:00:57AM -0700, sukeshs@broadcom.com wrote:
> >>>>>>>   #if defined(CONFIG_PM_SLEEP)
> >>>>>>>
> >>>>>>>+extern volatile bool dhd_mmc_suspend;
> >>>>>>This does not do what you think it does.
> >>>>>Hi Greg,
> >>>>>
> >>>>>I posted the 'catchup' patch set and I rebased it on '[PATCH 1/2]
> >>>>>brcm80211: FIX for TKIP GTK bug' from Sukesh. The only other two
> >>>>>patches I saw on the linuxdriver list are:
> >>>>You rebased on a 1/2 patch in a series with an invalid 0/2 email?
> >>>>
> >>>>{sigh}
> >>>Sorry.
> >>>>I dropped that patch as the whole series needed to be resent, do you
> >>>>want me to apply that one patch before yours?
> >>>I can resend my series if that is more convenient for you.
> >>Please, but just send it to me, I doubt everyone wants to see 60+
> >>patches again on the list.
> >Ok. Just for this time I will agree to break the rules ;-)
> 
> Hi Greg,
> 
> I rebase and regenerated the patch files. Did a diff before
> resending them with the previous set and there are no relevant
> differences so the patches I sent this morning are ok.

Great, I've applied them all now.

greg k-h