The driver was failing to clear the BSSID when a disconnect happened. That
prevented a reconnection. This problem is reported at
https://bugzilla.redhat.com/show_bug.cgi?id=789605,
https://bugzilla.redhat.com/show_bug.cgi?id=866786,
https://bugzilla.redhat.com/show_bug.cgi?id=906734, and
https://bugzilla.kernel.org/show_bug.cgi?id=46171.
Thanks to Jussi Kivilinna for making the critical observation
that led to the solution.
Reported-by: Jussi Kivilinna <[email protected]>
Tested-by: Jussi Kivilinna <[email protected]>
Tested-by: Alessandro Lannocca <[email protected]>
Signed-off-by: Larry Finger <[email protected]>
Cc: Stable <[email protected]>
---
John,
As you can see by the number of bug reports, this patch should be
pushed as soon as possible.
Thanks,
Larry
---
base.h | 3 +
pci.c | 2 -
rtl8192cu/hw.c | 87 ++++++++++++++++++++++-----------------------------------
3 files changed, 39 insertions(+), 53 deletions(-)
Index: linux-2.6/drivers/net/wireless/rtlwifi/rtl8192cu/hw.c
===================================================================
--- linux-2.6.orig/drivers/net/wireless/rtlwifi/rtl8192cu/hw.c
+++ linux-2.6/drivers/net/wireless/rtlwifi/rtl8192cu/hw.c
@@ -1377,74 +1377,57 @@ void rtl92cu_card_disable(struct ieee802
void rtl92cu_set_check_bssid(struct ieee80211_hw *hw, bool check_bssid)
{
- /* dummy routine needed for callback from rtl_op_configure_filter() */
-}
-
-/*========================================================================== */
-
-static void _rtl92cu_set_check_bssid(struct ieee80211_hw *hw,
- enum nl80211_iftype type)
-{
struct rtl_priv *rtlpriv = rtl_priv(hw);
- u32 reg_rcr = rtl_read_dword(rtlpriv, REG_RCR);
struct rtl_hal *rtlhal = rtl_hal(rtlpriv);
- struct rtl_phy *rtlphy = &(rtlpriv->phy);
- u8 filterout_non_associated_bssid = false;
+ u32 reg_rcr = rtl_read_dword(rtlpriv, REG_RCR);
- switch (type) {
- case NL80211_IFTYPE_ADHOC:
- case NL80211_IFTYPE_STATION:
- filterout_non_associated_bssid = true;
- break;
- case NL80211_IFTYPE_UNSPECIFIED:
- case NL80211_IFTYPE_AP:
- default:
- break;
- }
- if (filterout_non_associated_bssid) {
+ if (rtlpriv->psc.rfpwr_state != ERFON)
+ return;
+
+ if (check_bssid) {
+ u8 tmp;
if (IS_NORMAL_CHIP(rtlhal->version)) {
- switch (rtlphy->current_io_type) {
- case IO_CMD_RESUME_DM_BY_SCAN:
- reg_rcr |= (RCR_CBSSID_DATA | RCR_CBSSID_BCN);
- rtlpriv->cfg->ops->set_hw_reg(hw,
- HW_VAR_RCR, (u8 *)(®_rcr));
- /* enable update TSF */
- _rtl92cu_set_bcn_ctrl_reg(hw, 0, BIT(4));
- break;
- case IO_CMD_PAUSE_DM_BY_SCAN:
- reg_rcr &= ~(RCR_CBSSID_DATA | RCR_CBSSID_BCN);
- rtlpriv->cfg->ops->set_hw_reg(hw,
- HW_VAR_RCR, (u8 *)(®_rcr));
- /* disable update TSF */
- _rtl92cu_set_bcn_ctrl_reg(hw, BIT(4), 0);
- break;
- }
+ reg_rcr |= (RCR_CBSSID_DATA | RCR_CBSSID_BCN);
+ tmp = BIT(4);
} else {
- reg_rcr |= (RCR_CBSSID);
- rtlpriv->cfg->ops->set_hw_reg(hw, HW_VAR_RCR,
- (u8 *)(®_rcr));
- _rtl92cu_set_bcn_ctrl_reg(hw, 0, (BIT(4)|BIT(5)));
+ reg_rcr |= RCR_CBSSID;
+ tmp = BIT(4) | BIT(5);
}
- } else if (filterout_non_associated_bssid == false) {
+ rtlpriv->cfg->ops->set_hw_reg(hw, HW_VAR_RCR,
+ (u8 *) (®_rcr));
+ _rtl92cu_set_bcn_ctrl_reg(hw, 0, tmp);
+ } else {
+ u8 tmp;
if (IS_NORMAL_CHIP(rtlhal->version)) {
- reg_rcr &= (~(RCR_CBSSID_DATA | RCR_CBSSID_BCN));
- rtlpriv->cfg->ops->set_hw_reg(hw, HW_VAR_RCR,
- (u8 *)(®_rcr));
- _rtl92cu_set_bcn_ctrl_reg(hw, BIT(4), 0);
+ reg_rcr &= ~(RCR_CBSSID_DATA | RCR_CBSSID_BCN);
+ tmp = BIT(4);
} else {
- reg_rcr &= (~RCR_CBSSID);
- rtlpriv->cfg->ops->set_hw_reg(hw, HW_VAR_RCR,
- (u8 *)(®_rcr));
- _rtl92cu_set_bcn_ctrl_reg(hw, (BIT(4)|BIT(5)), 0);
+ reg_rcr &= ~RCR_CBSSID;
+ tmp = BIT(4) | BIT(5);
}
+ reg_rcr &= (~(RCR_CBSSID_DATA | RCR_CBSSID_BCN));
+ rtlpriv->cfg->ops->set_hw_reg(hw,
+ HW_VAR_RCR, (u8 *) (®_rcr));
+ _rtl92cu_set_bcn_ctrl_reg(hw, tmp, 0);
}
}
+/*========================================================================== */
+
int rtl92cu_set_network_type(struct ieee80211_hw *hw, enum nl80211_iftype type)
{
+ struct rtl_priv *rtlpriv = rtl_priv(hw);
+
if (_rtl92cu_set_media_status(hw, type))
return -EOPNOTSUPP;
- _rtl92cu_set_check_bssid(hw, type);
+
+ if (rtlpriv->mac80211.link_state == MAC80211_LINKED) {
+ if (type != NL80211_IFTYPE_AP)
+ rtl92cu_set_check_bssid(hw, true);
+ } else {
+ rtl92cu_set_check_bssid(hw, false);
+ }
+
return 0;
}
Index: linux-2.6/drivers/net/wireless/rtlwifi/base.h
===================================================================
--- linux-2.6.orig/drivers/net/wireless/rtlwifi/base.h
+++ linux-2.6/drivers/net/wireless/rtlwifi/base.h
@@ -143,5 +143,8 @@ extern struct attribute_group rtl_attrib
int rtlwifi_rate_mapping(struct ieee80211_hw *hw,
bool isht, u8 desc_rate, bool first_ampdu);
bool rtl_tx_mgmt_proc(struct ieee80211_hw *hw, struct sk_buff *skb);
+struct sk_buff *rtl_make_del_ba(struct ieee80211_hw *hw,
+ u8 *sa, u8 *bssid, u16 tid);
+void rtl_lps_change_work_callback(struct work_struct *work);
#endif
Index: linux-2.6/drivers/net/wireless/rtlwifi/pci.c
===================================================================
--- linux-2.6.orig/drivers/net/wireless/rtlwifi/pci.c
+++ linux-2.6/drivers/net/wireless/rtlwifi/pci.c
@@ -939,7 +939,7 @@ static void _rtl_pci_prepare_bcn_tasklet
return;
}
-static void rtl_lps_leave_work_callback(struct work_struct *work)
+void rtl_lps_leave_work_callback(struct work_struct *work)
{
struct rtl_works *rtlworks =
container_of(work, struct rtl_works, lps_leave_work);
On 03/13/2013 11:43 AM, Jussi Kivilinna wrote:
> On 13.03.2013 17:28, Larry Finger wrote:
>> The driver was failing to clear the BSSID when a disconnect happened. That
>> prevented a reconnection. This problem is reported at
>> https://bugzilla.redhat.com/show_bug.cgi?id=789605,
>> https://bugzilla.redhat.com/show_bug.cgi?id=866786,
>> https://bugzilla.redhat.com/show_bug.cgi?id=906734, and
>> https://bugzilla.kernel.org/show_bug.cgi?id=46171.
>>
>> Thanks to Jussi Kivilinna for making the critical observation
>> that led to the solution.
>>
>> Reported-by: Jussi Kivilinna <[email protected]>
>> Tested-by: Jussi Kivilinna <[email protected]>
>> Tested-by: Alessandro Lannocca <[email protected]>
>> Signed-off-by: Larry Finger <[email protected]>
>> Cc: Stable <[email protected]>
>> ---
>>
>> John,
>>
>> As you can see by the number of bug reports, this patch should be
>> pushed as soon as possible.
>>
>> Thanks,
>>
>> Larry
>> ---
>>
>> base.h | 3 +
>> pci.c | 2 -
>
> Seems like leak from another patch/change?
>
>> rtl8192cu/hw.c | 87 ++++++++++++++++++++++-----------------------------------
>> 3 files changed, 39 insertions(+), 53 deletions(-)
Yes, there were some differences between having the patch at the end of the
line. I'll fix and send V2.
Larry
On 03/13/2013 01:46 PM, John W. Linville wrote:
> On Wed, Mar 13, 2013 at 06:43:55PM +0200, Jussi Kivilinna wrote:
>> On 13.03.2013 17:28, Larry Finger wrote:
>>> The driver was failing to clear the BSSID when a disconnect happened. That
>>> prevented a reconnection. This problem is reported at
>>> https://bugzilla.redhat.com/show_bug.cgi?id=789605,
>>> https://bugzilla.redhat.com/show_bug.cgi?id=866786,
>>> https://bugzilla.redhat.com/show_bug.cgi?id=906734, and
>>> https://bugzilla.kernel.org/show_bug.cgi?id=46171.
>>>
>>> Thanks to Jussi Kivilinna for making the critical observation
>>> that led to the solution.
>>>
>>> Reported-by: Jussi Kivilinna <[email protected]>
>>> Tested-by: Jussi Kivilinna <[email protected]>
>>> Tested-by: Alessandro Lannocca <[email protected]>
>>> Signed-off-by: Larry Finger <[email protected]>
>>> Cc: Stable <[email protected]>
>>> ---
>>>
>>> John,
>>>
>>> As you can see by the number of bug reports, this patch should be
>>> pushed as soon as possible.
>>>
>>> Thanks,
>>>
>>> Larry
>>> ---
>>>
>>> base.h | 3 +
>>> pci.c | 2 -
>>
>> Seems like leak from another patch/change?
>
> I edited-out those bits...
Thanks,
Larry
On Wed, Mar 13, 2013 at 06:43:55PM +0200, Jussi Kivilinna wrote:
> On 13.03.2013 17:28, Larry Finger wrote:
> > The driver was failing to clear the BSSID when a disconnect happened. That
> > prevented a reconnection. This problem is reported at
> > https://bugzilla.redhat.com/show_bug.cgi?id=789605,
> > https://bugzilla.redhat.com/show_bug.cgi?id=866786,
> > https://bugzilla.redhat.com/show_bug.cgi?id=906734, and
> > https://bugzilla.kernel.org/show_bug.cgi?id=46171.
> >
> > Thanks to Jussi Kivilinna for making the critical observation
> > that led to the solution.
> >
> > Reported-by: Jussi Kivilinna <[email protected]>
> > Tested-by: Jussi Kivilinna <[email protected]>
> > Tested-by: Alessandro Lannocca <[email protected]>
> > Signed-off-by: Larry Finger <[email protected]>
> > Cc: Stable <[email protected]>
> > ---
> >
> > John,
> >
> > As you can see by the number of bug reports, this patch should be
> > pushed as soon as possible.
> >
> > Thanks,
> >
> > Larry
> > ---
> >
> > base.h | 3 +
> > pci.c | 2 -
>
> Seems like leak from another patch/change?
I edited-out those bits...
--
John W. Linville Someday the world will need a hero, and you
[email protected] might be all we have. Be ready.
On 13.03.2013 17:28, Larry Finger wrote:
> The driver was failing to clear the BSSID when a disconnect happened. That
> prevented a reconnection. This problem is reported at
> https://bugzilla.redhat.com/show_bug.cgi?id=789605,
> https://bugzilla.redhat.com/show_bug.cgi?id=866786,
> https://bugzilla.redhat.com/show_bug.cgi?id=906734, and
> https://bugzilla.kernel.org/show_bug.cgi?id=46171.
>
> Thanks to Jussi Kivilinna for making the critical observation
> that led to the solution.
>
> Reported-by: Jussi Kivilinna <[email protected]>
> Tested-by: Jussi Kivilinna <[email protected]>
> Tested-by: Alessandro Lannocca <[email protected]>
> Signed-off-by: Larry Finger <[email protected]>
> Cc: Stable <[email protected]>
> ---
>
> John,
>
> As you can see by the number of bug reports, this patch should be
> pushed as soon as possible.
>
> Thanks,
>
> Larry
> ---
>
> base.h | 3 +
> pci.c | 2 -
Seems like leak from another patch/change?
> rtl8192cu/hw.c | 87 ++++++++++++++++++++++-----------------------------------
> 3 files changed, 39 insertions(+), 53 deletions(-)
>
<snip>
>
> Index: linux-2.6/drivers/net/wireless/rtlwifi/base.h
> ===================================================================
> --- linux-2.6.orig/drivers/net/wireless/rtlwifi/base.h
> +++ linux-2.6/drivers/net/wireless/rtlwifi/base.h
> @@ -143,5 +143,8 @@ extern struct attribute_group rtl_attrib
> int rtlwifi_rate_mapping(struct ieee80211_hw *hw,
> bool isht, u8 desc_rate, bool first_ampdu);
> bool rtl_tx_mgmt_proc(struct ieee80211_hw *hw, struct sk_buff *skb);
> +struct sk_buff *rtl_make_del_ba(struct ieee80211_hw *hw,
> + u8 *sa, u8 *bssid, u16 tid);
> +void rtl_lps_change_work_callback(struct work_struct *work);
>
> #endif
> Index: linux-2.6/drivers/net/wireless/rtlwifi/pci.c
> ===================================================================
> --- linux-2.6.orig/drivers/net/wireless/rtlwifi/pci.c
> +++ linux-2.6/drivers/net/wireless/rtlwifi/pci.c
> @@ -939,7 +939,7 @@ static void _rtl_pci_prepare_bcn_tasklet
> return;
> }
>
> -static void rtl_lps_leave_work_callback(struct work_struct *work)
> +void rtl_lps_leave_work_callback(struct work_struct *work)
> {
> struct rtl_works *rtlworks =
> container_of(work, struct rtl_works, lps_leave_work);
>
On Wed, Mar 13, 2013 at 10:28:13AM -0500, Larry Finger wrote:
> The driver was failing to clear the BSSID when a disconnect happened. That
> prevented a reconnection. This problem is reported at
> https://bugzilla.redhat.com/show_bug.cgi?id=789605,
> https://bugzilla.redhat.com/show_bug.cgi?id=866786,
> https://bugzilla.redhat.com/show_bug.cgi?id=906734, and
> https://bugzilla.kernel.org/show_bug.cgi?id=46171.
>
> Thanks to Jussi Kivilinna for making the critical observation
> that led to the solution.
>
> Reported-by: Jussi Kivilinna <[email protected]>
> Tested-by: Jussi Kivilinna <[email protected]>
> Tested-by: Alessandro Lannocca <[email protected]>
> Signed-off-by: Larry Finger <[email protected]>
> Cc: Stable <[email protected]>
> ---
>
> John,
>
> As you can see by the number of bug reports, this patch should be
> pushed as soon as possible.
>
> Thanks,
>
> Larry
Got it -- disregard my earlier query... :-)
--
John W. Linville Someday the world will need a hero, and you
[email protected] might be all we have. Be ready.