* RE: problems setting the channel
From: Albert Lo @ 2013-09-13 17:30 UTC (permalink / raw)
To: Ben Greear; +Cc: linux-wireless@vger.kernel.org
In-Reply-To: <523349B1.8010306@candelatech.com>
Hi Ben,
yes, for normal use, I shouldn't have to specify the channel.
But I have a not normal use case.
Thx.
cheers,
Albert
________________________________________
From: Ben Greear <greearb@candelatech.com>
Sent: Friday, September 13, 2013 10:21 AM
To: Albert Lo
Cc: linux-wireless@vger.kernel.org
Subject: Re: problems setting the channel
On 09/13/2013 10:11 AM, Albert Lo wrote:
>
> Hello,
>
> Forgive the newbie question, but I'm having a hard time setting the channel. Here is my output:
>
> ======================================
> root@OpenWrt:~# iw dev wlan0 info
> Interface wlan0
> ifindex 7
> wdev 0x2
> addr 04:f0:21:01:1b:91
> type AP
> wiphy 0
> channel 11 (2462 MHz) NO HT
> root@OpenWrt:~# iw reg get
> country US:
> (2402 - 2472 @ 40), (3, 27)
> (5170 - 5250 @ 40), (3, 17)
> (5250 - 5330 @ 40), (3, 20), DFS
> (5490 - 5600 @ 40), (3, 20), DFS
> (5650 - 5710 @ 40), (3, 20), DFS
> (5735 - 5835 @ 40), (3, 30)
> root@OpenWrt:/# iw dev wlan0 set channel 10
> root@OpenWrt:/# iw dev wlan0 set channel 19
> command failed: Invalid argument (-22)
> ==================================================
>
>
> Any ideas why setting channel to 10 is ok, but setting channel to 19 fails?
> I do notice that my phy is split into two parts, Band1 and Band2.
There is no channel 19. Band2 starts at channel 36.
> Frequencies:
> * 5180 MHz [36] (17.0 dBm)
> * 5200 MHz [40] (17.0 dBm)
> * 5220 MHz [44] (17.0 dBm)
> * 5240 MHz [48] (17.0 dBm)
> * 5260 MHz [52] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5280 MHz [56] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5300 MHz [60] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5320 MHz [64] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5500 MHz [100] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5520 MHz [104] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5540 MHz [108] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5560 MHz [112] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5580 MHz [116] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5600 MHz [120] (disabled)
> * 5620 MHz [124] (disabled)
> * 5640 MHz [128] (disabled)
> * 5660 MHz [132] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5680 MHz [136] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5700 MHz [140] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5745 MHz [149] (30.0 dBm)
> * 5765 MHz [153] (30.0 dBm)
> * 5785 MHz [157] (30.0 dBm)
> * 5805 MHz [161] (30.0 dBm)
> * 5825 MHz [165] (30.0 dBm)
For normal use, you usually don't need to specify the channel anyway, by the way.
Thanks,
Ben
--
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc http://www.candelatech.com
^ permalink raw reply
* RE: problems setting the channel
From: Albert Lo @ 2013-09-13 17:29 UTC (permalink / raw)
To: Larry Finger; +Cc: linux-wireless@vger.kernel.org
In-Reply-To: <523349F2.2000702@lwfinger.net>
Hi Larry,
Thanks for the info. I tried setting the channel to one that is listed in the link you provided, and even to one that is listed in phy info. ch=36 is nice and safe.
however when I type:
> iw dev wlan0 info
It is still reporting it's on the old channel (11).
Any ideas why it is reporting the old channel?
Thx.
Cheers,
Albert
________________________________________
From: Larry Finger <larry.finger@gmail.com> on behalf of Larry Finger <Larry.Finger@lwfinger.net>
Sent: Friday, September 13, 2013 10:22 AM
To: Albert Lo
Cc: linux-wireless@vger.kernel.org
Subject: Re: problems setting the channel
On 09/13/2013 12:11 PM, Albert Lo wrote:
>
> Hello,
>
> Forgive the newbie question, but I'm having a hard time setting the channel. Here is my output:
>
> ======================================
> root@OpenWrt:~# iw dev wlan0 info
> Interface wlan0
> ifindex 7
> wdev 0x2
> addr 04:f0:21:01:1b:91
> type AP
> wiphy 0
> channel 11 (2462 MHz) NO HT
> root@OpenWrt:~# iw reg get
> country US:
> (2402 - 2472 @ 40), (3, 27)
> (5170 - 5250 @ 40), (3, 17)
> (5250 - 5330 @ 40), (3, 20), DFS
> (5490 - 5600 @ 40), (3, 20), DFS
> (5650 - 5710 @ 40), (3, 20), DFS
> (5735 - 5835 @ 40), (3, 30)
> root@OpenWrt:/# iw dev wlan0 set channel 10
> root@OpenWrt:/# iw dev wlan0 set channel 19
> command failed: Invalid argument (-22)
--snip--
> Any ideas as to how I can set wlan0 to the channels in Band2?
There is no channel 19 in wifi! That is the reason you cannot select it.
To select Band 2, you would need a dual-band Access Point. Do you have one?
Clearly, you have little knowledge of wifi. I suggest you read
http://en.wikipedia.org/wiki/List_of_WLAN_channels.
Larry
^ permalink raw reply
* Re: problems setting the channel
From: Larry Finger @ 2013-09-13 17:22 UTC (permalink / raw)
To: Albert Lo; +Cc: linux-wireless@vger.kernel.org
In-Reply-To: <bd4dc3b2edbb45789121bddb33960aa2@CO1PR05MB345.namprd05.prod.outlook.com>
On 09/13/2013 12:11 PM, Albert Lo wrote:
>
> Hello,
>
> Forgive the newbie question, but I'm having a hard time setting the channel. Here is my output:
>
> ======================================
> root@OpenWrt:~# iw dev wlan0 info
> Interface wlan0
> ifindex 7
> wdev 0x2
> addr 04:f0:21:01:1b:91
> type AP
> wiphy 0
> channel 11 (2462 MHz) NO HT
> root@OpenWrt:~# iw reg get
> country US:
> (2402 - 2472 @ 40), (3, 27)
> (5170 - 5250 @ 40), (3, 17)
> (5250 - 5330 @ 40), (3, 20), DFS
> (5490 - 5600 @ 40), (3, 20), DFS
> (5650 - 5710 @ 40), (3, 20), DFS
> (5735 - 5835 @ 40), (3, 30)
> root@OpenWrt:/# iw dev wlan0 set channel 10
> root@OpenWrt:/# iw dev wlan0 set channel 19
> command failed: Invalid argument (-22)
--snip--
> Any ideas as to how I can set wlan0 to the channels in Band2?
There is no channel 19 in wifi! That is the reason you cannot select it.
To select Band 2, you would need a dual-band Access Point. Do you have one?
Clearly, you have little knowledge of wifi. I suggest you read
http://en.wikipedia.org/wiki/List_of_WLAN_channels.
Larry
^ permalink raw reply
* Re: problems setting the channel
From: Ben Greear @ 2013-09-13 17:21 UTC (permalink / raw)
To: Albert Lo; +Cc: linux-wireless@vger.kernel.org
In-Reply-To: <bd4dc3b2edbb45789121bddb33960aa2@CO1PR05MB345.namprd05.prod.outlook.com>
On 09/13/2013 10:11 AM, Albert Lo wrote:
>
> Hello,
>
> Forgive the newbie question, but I'm having a hard time setting the channel. Here is my output:
>
> ======================================
> root@OpenWrt:~# iw dev wlan0 info
> Interface wlan0
> ifindex 7
> wdev 0x2
> addr 04:f0:21:01:1b:91
> type AP
> wiphy 0
> channel 11 (2462 MHz) NO HT
> root@OpenWrt:~# iw reg get
> country US:
> (2402 - 2472 @ 40), (3, 27)
> (5170 - 5250 @ 40), (3, 17)
> (5250 - 5330 @ 40), (3, 20), DFS
> (5490 - 5600 @ 40), (3, 20), DFS
> (5650 - 5710 @ 40), (3, 20), DFS
> (5735 - 5835 @ 40), (3, 30)
> root@OpenWrt:/# iw dev wlan0 set channel 10
> root@OpenWrt:/# iw dev wlan0 set channel 19
> command failed: Invalid argument (-22)
> ==================================================
>
>
> Any ideas why setting channel to 10 is ok, but setting channel to 19 fails?
> I do notice that my phy is split into two parts, Band1 and Band2.
There is no channel 19. Band2 starts at channel 36.
> Frequencies:
> * 5180 MHz [36] (17.0 dBm)
> * 5200 MHz [40] (17.0 dBm)
> * 5220 MHz [44] (17.0 dBm)
> * 5240 MHz [48] (17.0 dBm)
> * 5260 MHz [52] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5280 MHz [56] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5300 MHz [60] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5320 MHz [64] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5500 MHz [100] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5520 MHz [104] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5540 MHz [108] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5560 MHz [112] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5580 MHz [116] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5600 MHz [120] (disabled)
> * 5620 MHz [124] (disabled)
> * 5640 MHz [128] (disabled)
> * 5660 MHz [132] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5680 MHz [136] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5700 MHz [140] (20.0 dBm) (passive scanning, no IBSS, radar detection)
> * 5745 MHz [149] (30.0 dBm)
> * 5765 MHz [153] (30.0 dBm)
> * 5785 MHz [157] (30.0 dBm)
> * 5805 MHz [161] (30.0 dBm)
> * 5825 MHz [165] (30.0 dBm)
For normal use, you usually don't need to specify the channel anyway, by the way.
Thanks,
Ben
--
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc http://www.candelatech.com
^ permalink raw reply
* Re: brcmfmac, BCM43241
From: Franky Lin @ 2013-09-13 17:16 UTC (permalink / raw)
To: mail; +Cc: linux-wireless
In-Reply-To: <5230D40B.20704@janno-kaiser.de>
On 09/11/2013 01:35 PM, Jan-Oliver Kaiser wrote:
> I copied the firmware to /lib/firmware/brcm/brcmfmac-sdio.bin and copied
> the content of /dev/nvram to /lib/firmware/brcm/brcmfmac-sdio.txt.
Hi Jan-Oliver,
Could you share me the nvram file you retrieve through /dev/nvram? Which
distro are you using?
Thanks,
Franky
^ permalink raw reply
* problems setting the channel
From: Albert Lo @ 2013-09-13 17:11 UTC (permalink / raw)
To: linux-wireless@vger.kernel.org
Hello,
Forgive the newbie question, but I'm having a hard time setting the channel. Here is my output:
======================================
root@OpenWrt:~# iw dev wlan0 info
Interface wlan0
ifindex 7
wdev 0x2
addr 04:f0:21:01:1b:91
type AP
wiphy 0
channel 11 (2462 MHz) NO HT
root@OpenWrt:~# iw reg get
country US:
(2402 - 2472 @ 40), (3, 27)
(5170 - 5250 @ 40), (3, 17)
(5250 - 5330 @ 40), (3, 20), DFS
(5490 - 5600 @ 40), (3, 20), DFS
(5650 - 5710 @ 40), (3, 20), DFS
(5735 - 5835 @ 40), (3, 30)
root@OpenWrt:/# iw dev wlan0 set channel 10
root@OpenWrt:/# iw dev wlan0 set channel 19
command failed: Invalid argument (-22)
==================================================
Any ideas why setting channel to 10 is ok, but setting channel to 19 fails?
I do notice that my phy is split into two parts, Band1 and Band2.
==================================================
root@OpenWrt:~# iw phy phy0 info
Wiphy phy0
Band 1:
Capabilities: 0x11ce
HT20/HT40
SM Power Save disabled
RX HT40 SGI
TX STBC
RX STBC 1-stream
Max AMSDU length: 3839 bytes
DSSS/CCK HT40
Maximum RX AMPDU length 65535 bytes (exponent: 0x003)
Minimum RX AMPDU time spacing: 8 usec (0x06)
HT TX/RX MCS rate indexes supported: 0-15
Frequencies:
* 2412 MHz [1] (27.0 dBm)
* 2417 MHz [2] (27.0 dBm)
* 2422 MHz [3] (27.0 dBm)
* 2427 MHz [4] (27.0 dBm)
* 2432 MHz [5] (27.0 dBm)
* 2437 MHz [6] (27.0 dBm)
* 2442 MHz [7] (27.0 dBm)
* 2447 MHz [8] (27.0 dBm)
* 2452 MHz [9] (27.0 dBm)
* 2457 MHz [10] (27.0 dBm)
* 2462 MHz [11] (27.0 dBm)
* 2467 MHz [12] (disabled)
* 2472 MHz [13] (disabled)
* 2484 MHz [14] (disabled)
Bitrates (non-HT):
* 1.0 Mbps
* 2.0 Mbps (short preamble supported)
* 5.5 Mbps (short preamble supported)
* 11.0 Mbps (short preamble supported)
* 6.0 Mbps
* 9.0 Mbps
* 12.0 Mbps
* 18.0 Mbps
* 24.0 Mbps
* 36.0 Mbps
* 48.0 Mbps
* 54.0 Mbps
Band 2:
Capabilities: 0x11ce
HT20/HT40
SM Power Save disabled
RX HT40 SGI
TX STBC
RX STBC 1-stream
Max AMSDU length: 3839 bytes
DSSS/CCK HT40
Maximum RX AMPDU length 65535 bytes (exponent: 0x003)
Minimum RX AMPDU time spacing: 8 usec (0x06)
HT TX/RX MCS rate indexes supported: 0-15
Frequencies:
* 5180 MHz [36] (17.0 dBm)
* 5200 MHz [40] (17.0 dBm)
* 5220 MHz [44] (17.0 dBm)
* 5240 MHz [48] (17.0 dBm)
* 5260 MHz [52] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5280 MHz [56] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5300 MHz [60] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5320 MHz [64] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5500 MHz [100] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5520 MHz [104] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5540 MHz [108] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5560 MHz [112] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5580 MHz [116] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5600 MHz [120] (disabled)
* 5620 MHz [124] (disabled)
* 5640 MHz [128] (disabled)
* 5660 MHz [132] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5680 MHz [136] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5700 MHz [140] (20.0 dBm) (passive scanning, no IBSS, radar detection)
* 5745 MHz [149] (30.0 dBm)
* 5765 MHz [153] (30.0 dBm)
* 5785 MHz [157] (30.0 dBm)
* 5805 MHz [161] (30.0 dBm)
* 5825 MHz [165] (30.0 dBm)
Bitrates (non-HT):
* 6.0 Mbps
* 9.0 Mbps
* 12.0 Mbps
* 18.0 Mbps
* 24.0 Mbps
* 36.0 Mbps
* 48.0 Mbps
* 54.0 Mbps
.......
Any ideas as to how I can set wlan0 to the channels in Band2?
Best Regards,
Albert
^ permalink raw reply
* [PATCH] ath9k: mark wmi_event_swba as __packed
From: Chris Metcalf @ 2013-09-13 16:14 UTC (permalink / raw)
To: Luis R. Rodriguez, Jouni Malinen, Vasanthakumar Thiagarajan,
Senthil Balasubramanian, John W. Linville, linux-wireless,
ath9k-devel, netdev, linux-kernel
The other structures in wmi.h are already marked this way.
Without this marking, we get an unaliged access panic in the tilegx kernel:
Starting stack dump of tid 0, pid 0 (swapper) on cpu 35 at cycle 198675113844
frame 0: 0xfffffff7103ada90 ath9k_htc_swba+0x120/0x618 [ath9k_htc]
frame 1: 0xfffffff7103a4b10 ath9k_wmi_event_tasklet+0x1b0/0x270 [ath9k_htc]
frame 2: 0xfffffff700326570 tasklet_action+0x148/0x298
[...]
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
---
drivers/net/wireless/ath/ath9k/wmi.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/ath9k/wmi.h b/drivers/net/wireless/ath/ath9k/wmi.h
index fde6da6..0db37f2 100644
--- a/drivers/net/wireless/ath/ath9k/wmi.h
+++ b/drivers/net/wireless/ath/ath9k/wmi.h
@@ -39,7 +39,7 @@ struct wmi_fw_version {
struct wmi_event_swba {
__be64 tsf;
u8 beacon_pending;
-};
+} __packed;
/*
* 64 - HTC header - WMI header - 1 / txstatus
--
1.8.3.1
^ permalink raw reply related
* Re: WEP/WPA/WPA2 implementation in software
From: Oleksij Rempel @ 2013-09-13 14:24 UTC (permalink / raw)
To: Prabhunath G; +Cc: linux-wireless, kernelnewbies
In-Reply-To: <CACPKTAM-mtTS7QVSq817TQ11dS-cLA+nKgXF6yDGYvxw9rhucw@mail.gmail.com>
Am 13.09.2013 13:53, schrieb Prabhunath G:
> Dear All,
>
> Suppose there is no hardware support for
> encryption/decryption for Wi-Fi devices. Will the data be
> encrypted/decrypted by software implementation. If yes, is it done at
> the application level or at the kernel level (mac80211).
> Kindly clarify.
> If the question has to be rephrased, please do it
Hi,
see hostapd on AP site and wpa_supplicant on STA:
http://wireless.kernel.org/en/users/Documentation/wpa_supplicant
http://wireless.kernel.org/en/users/Documentation/hostapd
check documentation here on:
http://wireless.kernel.org/en/users/Documentation
may be it will answer some of your questions.
--
Regards,
Oleksij
^ permalink raw reply
* Re: regulatory will stuck when cfg80211 is built-in
From: Eugene Krasnikov @ 2013-09-13 14:12 UTC (permalink / raw)
To: Kalle Valo; +Cc: linux-wireless
In-Reply-To: <87d2ogtv57.fsf@purkki.adurom.net>
Let me try to describe the problem in more details:
1) When user space is calling "iw reg set <contry_code>" iw utility
will send NL80211_CMD_REQ_SET_REG through nl80211 to cfg80211.
2) cfg80211 will process this call in reg_todo work that is defined
here https://git.kernel.org/cgit/linux/kernel/git/linville/wireless.git/tree/net/wireless/reg.c#n1621
3) reg_todo work is calling reg_process_pending_hints to process the
call from user space where it checks if last request was
processed(https://git.kernel.org/cgit/linux/kernel/git/linville/wireless.git/tree/net/wireless/reg.c#n1576).
But last request is still there! This last request is
"NL80211_REGDOM_SET_BY_CORE" request that cfg80211 sent on init for
first regulatory domain initialization.
So the problem is that the first request that is sent by inbuilt
cfg80211 will be never process and will always stay in last_request
variable(https://git.kernel.org/cgit/linux/kernel/git/linville/wireless.git/tree/net/wireless/reg.c#n91)
Kalle, did you ever tried inbuilt into kernel cfg80211? This problem
is 100% reproducible because inbuilt modules will be always initialize
before user space.
2013/9/10 Kalle Valo <kvalo@adurom.com>:
> Eugene Krasnikov <k.eugene.e@gmail.com> writes:
>
>> No, the flag CONFIG_CFG80211_INTERNAL_REGDB does not help and it should not.
>>
>> The problem is that cfg80211 will stuck in situation when cfg80211 is
>> initialized before user space. One way to get cfg80211 out of this
>> state is to run crda once right after user space is up and running.
>> But again is this intentional? Or it is more like a workaround?
>
> With the internal regdb you do not need crda anymore, so why does it still
> get stuck? Where does it exactly get stuck?
>
> --
> Kalle Valo
--
Best regards,
Eugene
^ permalink raw reply
* Re: WEP/WPA/WPA2 implementation in software
From: Valdis.Kletnieks @ 2013-09-13 13:32 UTC (permalink / raw)
To: Prabhunath G; +Cc: linux-wireless, kernelnewbies
In-Reply-To: <CACPKTAM-mtTS7QVSq817TQ11dS-cLA+nKgXF6yDGYvxw9rhucw@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 800 bytes --]
On Fri, 13 Sep 2013 17:23:22 +0530, Prabhunath G said:
> Suppose there is no hardware support for
> encryption/decryption for Wi-Fi devices
1) Do not use WEP. It's busted. It's *known* busted. You transmit
packets for more than a minute or so using WEP, you may as well not
bother with crypto at all. Yes, it's really that busted.
2) You're much better off spending the $15 to get a wireless card
that actually has hardware WPA2 support on it. The devices are
cheap, there's no reason to inflict pain on yourself.
I'd not be at all surprised if a card that doesn't have crypto
support isn't *also* lacking other hardware support for WPA/WPA2
So while it may well be *possible* to do it in software, I'd have
to rank it as one of the more masochistic ideas I've heard in a
long time.
[-- Attachment #2: Type: application/pgp-signature, Size: 865 bytes --]
^ permalink raw reply
* Re: rtl8192cu: testing with EdiMax USB
From: Mark Cave-Ayland @ 2013-09-13 13:27 UTC (permalink / raw)
To: Larry Finger; +Cc: linux-wireless
In-Reply-To: <52330CB4.7020902@ilande.co.uk>
On 13/09/13 14:01, Mark Cave-Ayland wrote:
>> I spent a bit more time tinkering further with debug=0x5, forgetting
>> that I had left your last diagnostic patch applied. Based upon when the
>> beacon output disappears in the logs (after updating the power
>> registers), it does seem likely that is a power-related problem.
>
> FWIW I just tried a quick test where I commented out the entire
> rtl92c_dm_txpower_tracking_callback_thermalmeter() function to make it a
> nop, and that didn't seem to make any difference...
Aha! The following diff to remove the call to rtl92c_dm_diginit() keeps
me associated to the AP:
diff --git a/drivers/net/wireless/rtlwifi/rtl8192c/dm_common.c
b/drivers/net/wireless/rtlwifi/rtl8192c/dm_common.c
index d2d57a2..c18362d 100644
--- a/drivers/net/wireless/rtlwifi/rtl8192c/dm_common.c
+++ b/drivers/net/wireless/rtlwifi/rtl8192c/dm_common.c
@@ -1275,7 +1275,7 @@ void rtl92c_dm_init(struct ieee80211_hw *hw)
struct rtl_priv *rtlpriv = rtl_priv(hw);
rtlpriv->dm.dm_type = DM_TYPE_BYDRIVER;
- rtl92c_dm_diginit(hw);
+ //rtl92c_dm_diginit(hw);
rtl92c_dm_init_dynamic_txpower(hw);
rtl92c_dm_init_edca_turbo(hw);
rtl92c_dm_init_rate_adaptive_mask(hw);
However, dhclient still takes a very long time get an IP address and the
connection seems extremely lossy, much like it was when I could get a
connection before. Perhaps there are two different bugs here, one for
the association and one for the data loss?
ATB,
Mark.
^ permalink raw reply related
* Re: rtl8192cu: testing with EdiMax USB
From: Mark Cave-Ayland @ 2013-09-13 13:01 UTC (permalink / raw)
To: Larry Finger; +Cc: linux-wireless
In-Reply-To: <5232DAF0.6040907@ilande.co.uk>
On 13/09/13 10:29, Mark Cave-Ayland wrote:
> I spent a bit more time tinkering further with debug=0x5, forgetting
> that I had left your last diagnostic patch applied. Based upon when the
> beacon output disappears in the logs (after updating the power
> registers), it does seem likely that is a power-related problem.
FWIW I just tried a quick test where I commented out the entire
rtl92c_dm_txpower_tracking_callback_thermalmeter() function to make it a
nop, and that didn't seem to make any difference...
ATB,
Mark.
^ permalink raw reply
* [PATCHv2 7/7] ath10k: remove wmi event worker thread
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1379074618-30534-1-git-send-email-michal.kazior@tieto.com>
It's not really necessary to have this processed
in a worker. There are no sleepable calls (and
actually shouldn't be).
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
---
drivers/net/wireless/ath/ath10k/core.h | 3 ---
drivers/net/wireless/ath/ath10k/wmi.c | 44 +-------------------------------
2 files changed, 1 insertion(+), 46 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/core.h b/drivers/net/wireless/ath/ath10k/core.h
index c2b6a76..fcf94ee 100644
--- a/drivers/net/wireless/ath/ath10k/core.h
+++ b/drivers/net/wireless/ath/ath10k/core.h
@@ -113,9 +113,6 @@ struct ath10k_wmi {
struct completion service_ready;
struct completion unified_ready;
wait_queue_head_t tx_credits_wq;
-
- struct sk_buff_head wmi_event_list;
- struct work_struct wmi_event_work;
};
struct ath10k_peer_stat {
diff --git a/drivers/net/wireless/ath/ath10k/wmi.c b/drivers/net/wireless/ath/ath10k/wmi.c
index b29d2b9..6803ead 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.c
+++ b/drivers/net/wireless/ath/ath10k/wmi.c
@@ -1055,7 +1055,7 @@ static int ath10k_wmi_ready_event_rx(struct ath10k *ar, struct sk_buff *skb)
return 0;
}
-static void ath10k_wmi_event_process(struct ath10k *ar, struct sk_buff *skb)
+static void ath10k_wmi_process_rx(struct ath10k *ar, struct sk_buff *skb)
{
struct wmi_cmd_hdr *cmd_hdr;
enum wmi_event_id id;
@@ -1174,43 +1174,6 @@ static void ath10k_wmi_event_process(struct ath10k *ar, struct sk_buff *skb)
dev_kfree_skb(skb);
}
-static void ath10k_wmi_event_work(struct work_struct *work)
-{
- struct ath10k *ar = container_of(work, struct ath10k,
- wmi.wmi_event_work);
- struct sk_buff *skb;
-
- for (;;) {
- skb = skb_dequeue(&ar->wmi.wmi_event_list);
- if (!skb)
- break;
-
- ath10k_wmi_event_process(ar, skb);
- }
-}
-
-static void ath10k_wmi_process_rx(struct ath10k *ar, struct sk_buff *skb)
-{
- struct wmi_cmd_hdr *cmd_hdr = (struct wmi_cmd_hdr *)skb->data;
- enum wmi_event_id event_id;
-
- event_id = MS(__le32_to_cpu(cmd_hdr->cmd_id), WMI_CMD_HDR_CMD_ID);
-
- /* some events require to be handled ASAP
- * thus can't be defered to a worker thread */
- switch (event_id) {
- case WMI_MGMT_RX_EVENTID:
- case WMI_HOST_SWBA_EVENTID:
- ath10k_wmi_event_process(ar, skb);
- return;
- default:
- break;
- }
-
- skb_queue_tail(&ar->wmi.wmi_event_list, skb);
- queue_work(ar->workqueue, &ar->wmi.wmi_event_work);
-}
-
/* WMI Initialization functions */
int ath10k_wmi_attach(struct ath10k *ar)
{
@@ -1218,16 +1181,11 @@ int ath10k_wmi_attach(struct ath10k *ar)
init_completion(&ar->wmi.unified_ready);
init_waitqueue_head(&ar->wmi.tx_credits_wq);
- skb_queue_head_init(&ar->wmi.wmi_event_list);
- INIT_WORK(&ar->wmi.wmi_event_work, ath10k_wmi_event_work);
-
return 0;
}
void ath10k_wmi_detach(struct ath10k *ar)
{
- cancel_work_sync(&ar->wmi.wmi_event_work);
- skb_queue_purge(&ar->wmi.wmi_event_list);
}
int ath10k_wmi_connect_htc_service(struct ath10k *ar)
--
1.7.9.5
^ permalink raw reply related
* [PATCHv2 6/7] ath10k: remove wmi pending count limit
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1379074618-30534-1-git-send-email-michal.kazior@tieto.com>
It is no longer used nor necessary since WMI
commands can block.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
---
drivers/net/wireless/ath/ath10k/core.h | 2 --
drivers/net/wireless/ath/ath10k/mac.c | 9 ---------
drivers/net/wireless/ath/ath10k/wmi.c | 32 --------------------------------
drivers/net/wireless/ath/ath10k/wmi.h | 2 --
4 files changed, 45 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/core.h b/drivers/net/wireless/ath/ath10k/core.h
index 14b7d3d..c2b6a76 100644
--- a/drivers/net/wireless/ath/ath10k/core.h
+++ b/drivers/net/wireless/ath/ath10k/core.h
@@ -112,8 +112,6 @@ struct ath10k_wmi {
enum ath10k_htc_ep_id eid;
struct completion service_ready;
struct completion unified_ready;
- atomic_t pending_tx_count;
- wait_queue_head_t wq;
wait_queue_head_t tx_credits_wq;
struct sk_buff_head wmi_event_list;
diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c
index 6c3e9d1..11aa13e 100644
--- a/drivers/net/wireless/ath/ath10k/mac.c
+++ b/drivers/net/wireless/ath/ath10k/mac.c
@@ -1224,8 +1224,6 @@ static void ath10k_bss_disassoc(struct ieee80211_hw *hw,
/* FIXME: why don't we print error if wmi call fails? */
ret = ath10k_wmi_vdev_down(ar, arvif->vdev_id);
- ath10k_wmi_flush_tx(ar);
-
arvif->def_wep_key_index = 0;
}
@@ -1664,8 +1662,6 @@ static int ath10k_abort_scan(struct ath10k *ar)
return -EIO;
}
- ath10k_wmi_flush_tx(ar);
-
ret = wait_for_completion_timeout(&ar->scan.completed, 3*HZ);
if (ret == 0)
ath10k_warn("timed out while waiting for scan to stop\n");
@@ -1699,10 +1695,6 @@ static int ath10k_start_scan(struct ath10k *ar,
if (ret)
return ret;
- /* make sure we submit the command so the completion
- * timeout makes sense */
- ath10k_wmi_flush_tx(ar);
-
ret = wait_for_completion_timeout(&ar->scan.started, 1*HZ);
if (ret == 0) {
ath10k_abort_scan(ar);
@@ -1924,7 +1916,6 @@ static int ath10k_config(struct ieee80211_hw *hw, u32 changed)
ret = ath10k_monitor_destroy(ar);
}
- ath10k_wmi_flush_tx(ar);
mutex_unlock(&ar->conf_mutex);
return ret;
}
diff --git a/drivers/net/wireless/ath/ath10k/wmi.c b/drivers/net/wireless/ath/ath10k/wmi.c
index 9152dae..b29d2b9 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.c
+++ b/drivers/net/wireless/ath/ath10k/wmi.c
@@ -23,30 +23,6 @@
#include "wmi.h"
#include "mac.h"
-void ath10k_wmi_flush_tx(struct ath10k *ar)
-{
- int ret;
-
- lockdep_assert_held(&ar->conf_mutex);
-
- if (ar->state == ATH10K_STATE_WEDGED) {
- ath10k_warn("wmi flush skipped - device is wedged anyway\n");
- return;
- }
-
- ret = wait_event_timeout(ar->wmi.wq,
- atomic_read(&ar->wmi.pending_tx_count) == 0,
- 5*HZ);
- if (atomic_read(&ar->wmi.pending_tx_count) == 0)
- return;
-
- if (ret == 0)
- ret = -ETIMEDOUT;
-
- if (ret < 0)
- ath10k_warn("wmi flush failed (%d)\n", ret);
-}
-
int ath10k_wmi_wait_for_service_ready(struct ath10k *ar)
{
int ret;
@@ -85,9 +61,6 @@ static struct sk_buff *ath10k_wmi_alloc_skb(u32 len)
static void ath10k_wmi_htc_tx_complete(struct ath10k *ar, struct sk_buff *skb)
{
dev_kfree_skb(skb);
-
- if (atomic_sub_return(1, &ar->wmi.pending_tx_count) == 0)
- wake_up(&ar->wmi.wq);
}
static int ath10k_wmi_cmd_send_nowait(struct ath10k *ar, struct sk_buff *skb,
@@ -1243,7 +1216,6 @@ int ath10k_wmi_attach(struct ath10k *ar)
{
init_completion(&ar->wmi.service_ready);
init_completion(&ar->wmi.unified_ready);
- init_waitqueue_head(&ar->wmi.wq);
init_waitqueue_head(&ar->wmi.tx_credits_wq);
skb_queue_head_init(&ar->wmi.wmi_event_list);
@@ -1254,10 +1226,6 @@ int ath10k_wmi_attach(struct ath10k *ar)
void ath10k_wmi_detach(struct ath10k *ar)
{
- /* HTC should've drained the packets already */
- if (WARN_ON(atomic_read(&ar->wmi.pending_tx_count) > 0))
- ath10k_warn("there are still pending packets\n");
-
cancel_work_sync(&ar->wmi.wmi_event_work);
skb_queue_purge(&ar->wmi.wmi_event_list);
}
diff --git a/drivers/net/wireless/ath/ath10k/wmi.h b/drivers/net/wireless/ath/ath10k/wmi.h
index b100431..2c52c23 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.h
+++ b/drivers/net/wireless/ath/ath10k/wmi.h
@@ -3044,7 +3044,6 @@ struct wmi_force_fw_hang_cmd {
#define WMI_MAX_EVENT 0x1000
/* Maximum number of pending TXed WMI packets */
-#define WMI_MAX_PENDING_TX_COUNT 128
#define WMI_SKB_HEADROOM sizeof(struct wmi_cmd_hdr)
/* By default disable power save for IBSS */
@@ -3057,7 +3056,6 @@ int ath10k_wmi_attach(struct ath10k *ar);
void ath10k_wmi_detach(struct ath10k *ar);
int ath10k_wmi_wait_for_service_ready(struct ath10k *ar);
int ath10k_wmi_wait_for_unified_ready(struct ath10k *ar);
-void ath10k_wmi_flush_tx(struct ath10k *ar);
int ath10k_wmi_connect_htc_service(struct ath10k *ar);
int ath10k_wmi_pdev_set_channel(struct ath10k *ar,
--
1.7.9.5
^ permalink raw reply related
* [PATCHv2 5/7] ath10k: improve beacon submission latency
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1379074618-30534-1-git-send-email-michal.kazior@tieto.com>
The patch prevents beacon misses in some case of
heavy load on a system.
If a beacon can't be transmitted directly from an
SWBA event it will be left in arvif->beacon and
transmission will be retried once TX credits
become available.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
---
v2:
* add comments on ath10k_wmi_tx_beacons_nowait() usage
drivers/net/wireless/ath/ath10k/core.h | 1 +
drivers/net/wireless/ath/ath10k/mac.c | 7 +++
drivers/net/wireless/ath/ath10k/wmi.c | 74 ++++++++++++++++++++++++++------
drivers/net/wireless/ath/ath10k/wmi.h | 3 +-
4 files changed, 71 insertions(+), 14 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/core.h b/drivers/net/wireless/ath/ath10k/core.h
index c953a33..14b7d3d 100644
--- a/drivers/net/wireless/ath/ath10k/core.h
+++ b/drivers/net/wireless/ath/ath10k/core.h
@@ -204,6 +204,7 @@ struct ath10k_vif {
enum wmi_vdev_subtype vdev_subtype;
u32 beacon_interval;
u32 dtim_period;
+ struct sk_buff *beacon;
struct ath10k *ar;
struct ieee80211_vif *vif;
diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c
index 8b9fb66..6c3e9d1 100644
--- a/drivers/net/wireless/ath/ath10k/mac.c
+++ b/drivers/net/wireless/ath/ath10k/mac.c
@@ -2075,6 +2075,13 @@ static void ath10k_remove_interface(struct ieee80211_hw *hw,
mutex_lock(&ar->conf_mutex);
+ spin_lock_bh(&ar->data_lock);
+ if (arvif->beacon) {
+ dev_kfree_skb_any(arvif->beacon);
+ arvif->beacon = NULL;
+ }
+ spin_unlock_bh(&ar->data_lock);
+
ar->free_vdev_map |= 1 << (arvif->vdev_id);
if (arvif->vdev_type == WMI_VDEV_TYPE_AP) {
diff --git a/drivers/net/wireless/ath/ath10k/wmi.c b/drivers/net/wireless/ath/ath10k/wmi.c
index ff407c2..9152dae 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.c
+++ b/drivers/net/wireless/ath/ath10k/wmi.c
@@ -120,8 +120,53 @@ err_pull:
return ret;
}
+static void ath10k_wmi_tx_beacon_nowait(struct ath10k_vif *arvif)
+{
+ struct wmi_bcn_tx_arg arg = {0};
+ int ret;
+
+ lockdep_assert_held(&arvif->ar->data_lock);
+
+ if (arvif->beacon == NULL)
+ return;
+
+ arg.vdev_id = arvif->vdev_id;
+ arg.tx_rate = 0;
+ arg.tx_power = 0;
+ arg.bcn = arvif->beacon->data;
+ arg.bcn_len = arvif->beacon->len;
+
+ ret = ath10k_wmi_beacon_send_nowait(arvif->ar, &arg);
+ if (ret)
+ return;
+
+ dev_kfree_skb_any(arvif->beacon);
+ arvif->beacon = NULL;
+}
+
+static void ath10k_wmi_tx_beacons_iter(void *data, u8 *mac,
+ struct ieee80211_vif *vif)
+{
+ struct ath10k_vif *arvif = ath10k_vif_to_arvif(vif);
+
+ ath10k_wmi_tx_beacon_nowait(arvif);
+}
+
+static void ath10k_wmi_tx_beacons_nowait(struct ath10k *ar)
+{
+ spin_lock_bh(&ar->data_lock);
+ ieee80211_iterate_active_interfaces_atomic(ar->hw,
+ IEEE80211_IFACE_ITER_NORMAL,
+ ath10k_wmi_tx_beacons_iter,
+ NULL);
+ spin_unlock_bh(&ar->data_lock);
+}
+
static void ath10k_wmi_op_ep_tx_credits(struct ath10k *ar)
{
+ /* try to send pending beacons first. they take priority */
+ ath10k_wmi_tx_beacons_nowait(ar);
+
wake_up(&ar->wmi.tx_credits_wq);
}
@@ -131,6 +176,9 @@ static int ath10k_wmi_cmd_send(struct ath10k *ar, struct sk_buff *skb,
int ret = -EINVAL;
wait_event_timeout(ar->wmi.tx_credits_wq, ({
+ /* try to send pending beacons first. they take priority */
+ ath10k_wmi_tx_beacons_nowait(ar);
+
ret = ath10k_wmi_cmd_send_nowait(ar, skb, cmd_id);
(ret != -EAGAIN);
}), 3*HZ);
@@ -760,10 +808,8 @@ static void ath10k_wmi_event_host_swba(struct ath10k *ar, struct sk_buff *skb)
int i = -1;
struct wmi_bcn_info *bcn_info;
struct ath10k_vif *arvif;
- struct wmi_bcn_tx_arg arg;
struct sk_buff *bcn;
int vdev_id = 0;
- int ret;
ath10k_dbg(ATH10K_DBG_MGMT, "WMI_HOST_SWBA_EVENTID\n");
@@ -820,17 +866,17 @@ static void ath10k_wmi_event_host_swba(struct ath10k *ar, struct sk_buff *skb)
ath10k_wmi_update_tim(ar, arvif, bcn, bcn_info);
ath10k_wmi_update_noa(ar, arvif, bcn, bcn_info);
- arg.vdev_id = arvif->vdev_id;
- arg.tx_rate = 0;
- arg.tx_power = 0;
- arg.bcn = bcn->data;
- arg.bcn_len = bcn->len;
+ spin_lock_bh(&ar->data_lock);
+ if (arvif->beacon) {
+ ath10k_warn("SWBA overrun on vdev %d\n",
+ arvif->vdev_id);
+ dev_kfree_skb_any(arvif->beacon);
+ }
- ret = ath10k_wmi_beacon_send(ar, &arg);
- if (ret)
- ath10k_warn("could not send beacon (%d)\n", ret);
+ arvif->beacon = bcn;
- dev_kfree_skb_any(bcn);
+ ath10k_wmi_tx_beacon_nowait(arvif);
+ spin_unlock_bh(&ar->data_lock);
}
}
@@ -1181,6 +1227,7 @@ static void ath10k_wmi_process_rx(struct ath10k *ar, struct sk_buff *skb)
* thus can't be defered to a worker thread */
switch (event_id) {
case WMI_MGMT_RX_EVENTID:
+ case WMI_HOST_SWBA_EVENTID:
ath10k_wmi_event_process(ar, skb);
return;
default:
@@ -2138,7 +2185,8 @@ int ath10k_wmi_peer_assoc(struct ath10k *ar,
return ath10k_wmi_cmd_send(ar, skb, WMI_PEER_ASSOC_CMDID);
}
-int ath10k_wmi_beacon_send(struct ath10k *ar, const struct wmi_bcn_tx_arg *arg)
+int ath10k_wmi_beacon_send_nowait(struct ath10k *ar,
+ const struct wmi_bcn_tx_arg *arg)
{
struct wmi_bcn_tx_cmd *cmd;
struct sk_buff *skb;
@@ -2154,7 +2202,7 @@ int ath10k_wmi_beacon_send(struct ath10k *ar, const struct wmi_bcn_tx_arg *arg)
cmd->hdr.bcn_len = __cpu_to_le32(arg->bcn_len);
memcpy(cmd->bcn, arg->bcn, arg->bcn_len);
- return ath10k_wmi_cmd_send(ar, skb, WMI_BCN_TX_CMDID);
+ return ath10k_wmi_cmd_send_nowait(ar, skb, WMI_BCN_TX_CMDID);
}
static void ath10k_wmi_pdev_set_wmm_param(struct wmi_wmm_params *params,
diff --git a/drivers/net/wireless/ath/ath10k/wmi.h b/drivers/net/wireless/ath/ath10k/wmi.h
index ab46582..b100431 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.h
+++ b/drivers/net/wireless/ath/ath10k/wmi.h
@@ -3110,7 +3110,8 @@ int ath10k_wmi_set_ap_ps_param(struct ath10k *ar, u32 vdev_id, const u8 *mac,
enum wmi_ap_ps_peer_param param_id, u32 value);
int ath10k_wmi_scan_chan_list(struct ath10k *ar,
const struct wmi_scan_chan_list_arg *arg);
-int ath10k_wmi_beacon_send(struct ath10k *ar, const struct wmi_bcn_tx_arg *arg);
+int ath10k_wmi_beacon_send_nowait(struct ath10k *ar,
+ const struct wmi_bcn_tx_arg *arg);
int ath10k_wmi_pdev_set_wmm_params(struct ath10k *ar,
const struct wmi_pdev_set_wmm_params_arg *arg);
int ath10k_wmi_request_stats(struct ath10k *ar, enum wmi_stats_id stats_id);
--
1.7.9.5
^ permalink raw reply related
* [PATCHv2 4/7] ath10k: simplify HTC command submitting
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1379074618-30534-1-git-send-email-michal.kazior@tieto.com>
The patch removes HTC endpoint tx workers in
favour of direct command submission. This makes a
lot more sense for data path.
mac80211 queues are effectively stopped/woken up
in a more timely fashion preventing build up of
frames. It's possible to push more traffic than
the device/system is able to handle and have no
hiccups or performance degradation with UDP
traffic.
WMI commands will now report errors properly and
possibly block as they actively can wait for tx
credits to become available.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
---
v2:
* use DIV_ROUND_UP
drivers/net/wireless/ath/ath10k/htc.c | 176 ++++++++-------------------------
drivers/net/wireless/ath/ath10k/htc.h | 4 -
drivers/net/wireless/ath/ath10k/wmi.c | 3 +-
3 files changed, 43 insertions(+), 140 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/htc.c b/drivers/net/wireless/ath/ath10k/htc.c
index 49da4e5..d0d7212 100644
--- a/drivers/net/wireless/ath/ath10k/htc.c
+++ b/drivers/net/wireless/ath/ath10k/htc.c
@@ -117,99 +117,13 @@ static void ath10k_htc_prepare_tx_skb(struct ath10k_htc_ep *ep,
spin_unlock_bh(&ep->htc->tx_lock);
}
-static int ath10k_htc_issue_skb(struct ath10k_htc *htc,
- struct ath10k_htc_ep *ep,
- struct sk_buff *skb,
- u8 credits)
-{
- struct ath10k_skb_cb *skb_cb = ATH10K_SKB_CB(skb);
- int ret;
-
- ath10k_dbg(ATH10K_DBG_HTC, "%s: ep %d skb %p\n", __func__,
- ep->eid, skb);
-
- ath10k_htc_prepare_tx_skb(ep, skb);
-
- ret = ath10k_skb_map(htc->ar->dev, skb);
- if (ret)
- goto err;
-
- ret = ath10k_hif_send_head(htc->ar,
- ep->ul_pipe_id,
- ep->eid,
- skb->len,
- skb);
- if (unlikely(ret))
- goto err;
-
- return 0;
-err:
- ath10k_warn("HTC issue failed: %d\n", ret);
-
- spin_lock_bh(&htc->tx_lock);
- ep->tx_credits += credits;
- spin_unlock_bh(&htc->tx_lock);
-
- if (ep->ep_ops.ep_tx_credits)
- ep->ep_ops.ep_tx_credits(htc->ar);
-
- /* this is the simplest way to handle out-of-resources for non-credit
- * based endpoints. credit based endpoints can still get -ENOSR, but
- * this is highly unlikely as credit reservation should prevent that */
- if (ret == -ENOSR) {
- spin_lock_bh(&htc->tx_lock);
- __skb_queue_head(&ep->tx_queue, skb);
- spin_unlock_bh(&htc->tx_lock);
-
- return ret;
- }
-
- skb_cb->is_aborted = true;
- ath10k_htc_notify_tx_completion(ep, skb);
-
- return ret;
-}
-
-static void ath10k_htc_send_work(struct work_struct *work)
-{
- struct ath10k_htc_ep *ep = container_of(work,
- struct ath10k_htc_ep, send_work);
- struct ath10k_htc *htc = ep->htc;
- struct sk_buff *skb;
- u8 credits = 0;
- int ret;
-
- while (true) {
- if (ep->ul_is_polled)
- ath10k_htc_send_complete_check(ep, 0);
-
- spin_lock_bh(&htc->tx_lock);
- skb = __skb_dequeue(&ep->tx_queue);
-
- if (ep->tx_credit_flow_enabled) {
- credits = DIV_ROUND_UP(skb->len,
- htc->target_credit_size);
- if (ep->tx_credits < credits) {
- __skb_queue_head(&ep->tx_queue, skb);
- skb = NULL;
- }
- }
- spin_unlock_bh(&htc->tx_lock);
-
- if (!skb)
- break;
-
- ret = ath10k_htc_issue_skb(htc, ep, skb, credits);
- if (ret == -ENOSR)
- break;
- }
-}
-
int ath10k_htc_send(struct ath10k_htc *htc,
enum ath10k_htc_ep_id eid,
struct sk_buff *skb)
{
struct ath10k_htc_ep *ep = &htc->endpoint[eid];
+ int credits = 0;
+ int ret;
if (htc->ar->state == ATH10K_STATE_WEDGED)
return -ECOMM;
@@ -219,18 +133,55 @@ int ath10k_htc_send(struct ath10k_htc *htc,
return -ENOENT;
}
+ /* FIXME: This looks ugly, can we fix it? */
spin_lock_bh(&htc->tx_lock);
if (htc->stopped) {
spin_unlock_bh(&htc->tx_lock);
return -ESHUTDOWN;
}
+ spin_unlock_bh(&htc->tx_lock);
- __skb_queue_tail(&ep->tx_queue, skb);
skb_push(skb, sizeof(struct ath10k_htc_hdr));
- spin_unlock_bh(&htc->tx_lock);
- queue_work(htc->ar->workqueue, &ep->send_work);
+ if (ep->tx_credit_flow_enabled) {
+ credits = DIV_ROUND_UP(skb->len, htc->target_credit_size);
+ spin_lock_bh(&htc->tx_lock);
+ if (ep->tx_credits < credits) {
+ spin_unlock_bh(&htc->tx_lock);
+ ret = -EAGAIN;
+ goto err_pull;
+ }
+ ep->tx_credits -= credits;
+ spin_unlock_bh(&htc->tx_lock);
+ }
+
+ ath10k_htc_prepare_tx_skb(ep, skb);
+
+ ret = ath10k_skb_map(htc->ar->dev, skb);
+ if (ret)
+ goto err_credits;
+
+ ret = ath10k_hif_send_head(htc->ar, ep->ul_pipe_id, ep->eid,
+ skb->len, skb);
+ if (ret)
+ goto err_unmap;
+
return 0;
+
+err_unmap:
+ ath10k_skb_unmap(htc->ar->dev, skb);
+err_credits:
+ if (ep->tx_credit_flow_enabled) {
+ spin_lock_bh(&htc->tx_lock);
+ ep->tx_credits += credits;
+ spin_unlock_bh(&htc->tx_lock);
+
+ if (ep->ep_ops.ep_tx_credits)
+ ep->ep_ops.ep_tx_credits(htc->ar);
+ }
+err_pull:
+ skb_pull(skb, sizeof(struct ath10k_htc_hdr));
+ return ret;
}
static int ath10k_htc_tx_completion_handler(struct ath10k *ar,
@@ -243,39 +194,9 @@ static int ath10k_htc_tx_completion_handler(struct ath10k *ar,
ath10k_htc_notify_tx_completion(ep, skb);
/* the skb now belongs to the completion handler */
- /* note: when using TX credit flow, the re-checking of queues happens
- * when credits flow back from the target. in the non-TX credit case,
- * we recheck after the packet completes */
- spin_lock_bh(&htc->tx_lock);
- if (!ep->tx_credit_flow_enabled && !htc->stopped)
- queue_work(ar->workqueue, &ep->send_work);
- spin_unlock_bh(&htc->tx_lock);
-
return 0;
}
-/* flush endpoint TX queue */
-static void ath10k_htc_flush_endpoint_tx(struct ath10k_htc *htc,
- struct ath10k_htc_ep *ep)
-{
- struct sk_buff *skb;
- struct ath10k_skb_cb *skb_cb;
-
- spin_lock_bh(&htc->tx_lock);
- for (;;) {
- skb = __skb_dequeue(&ep->tx_queue);
- if (!skb)
- break;
-
- skb_cb = ATH10K_SKB_CB(skb);
- skb_cb->is_aborted = true;
- ath10k_htc_notify_tx_completion(ep, skb);
- }
- spin_unlock_bh(&htc->tx_lock);
-
- cancel_work_sync(&ep->send_work);
-}
-
/***********/
/* Receive */
/***********/
@@ -310,9 +231,6 @@ ath10k_htc_process_credit_report(struct ath10k_htc *htc,
ep->ep_ops.ep_tx_credits(htc->ar);
spin_lock_bh(&htc->tx_lock);
}
-
- if (ep->tx_credits && !skb_queue_empty(&ep->tx_queue))
- queue_work(htc->ar->workqueue, &ep->send_work);
}
spin_unlock_bh(&htc->tx_lock);
}
@@ -570,10 +488,8 @@ static void ath10k_htc_reset_endpoint_states(struct ath10k_htc *htc)
ep->max_ep_message_len = 0;
ep->max_tx_queue_depth = 0;
ep->eid = i;
- skb_queue_head_init(&ep->tx_queue);
ep->htc = htc;
ep->tx_credit_flow_enabled = true;
- INIT_WORK(&ep->send_work, ath10k_htc_send_work);
}
}
@@ -916,18 +832,10 @@ int ath10k_htc_start(struct ath10k_htc *htc)
*/
void ath10k_htc_stop(struct ath10k_htc *htc)
{
- int i;
- struct ath10k_htc_ep *ep;
-
spin_lock_bh(&htc->tx_lock);
htc->stopped = true;
spin_unlock_bh(&htc->tx_lock);
- for (i = ATH10K_HTC_EP_0; i < ATH10K_HTC_EP_COUNT; i++) {
- ep = &htc->endpoint[i];
- ath10k_htc_flush_endpoint_tx(htc, ep);
- }
-
ath10k_hif_stop(htc->ar);
}
diff --git a/drivers/net/wireless/ath/ath10k/htc.h b/drivers/net/wireless/ath/ath10k/htc.h
index 92ca29b..4716d33 100644
--- a/drivers/net/wireless/ath/ath10k/htc.h
+++ b/drivers/net/wireless/ath/ath10k/htc.h
@@ -316,15 +316,11 @@ struct ath10k_htc_ep {
int ul_is_polled; /* call HIF to get tx completions */
int dl_is_polled; /* call HIF to fetch rx (not implemented) */
- struct sk_buff_head tx_queue;
-
u8 seq_no; /* for debugging */
int tx_credits;
int tx_credit_size;
int tx_credits_per_max_message;
bool tx_credit_flow_enabled;
-
- struct work_struct send_work;
};
struct ath10k_htc_svc_tx_credits {
diff --git a/drivers/net/wireless/ath/ath10k/wmi.c b/drivers/net/wireless/ath/ath10k/wmi.c
index 66cd892..ff407c2 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.c
+++ b/drivers/net/wireless/ath/ath10k/wmi.c
@@ -120,8 +120,7 @@ err_pull:
return ret;
}
-static void ath10k_wmi_op_ep_tx_credits(struct ath10k *ar,
- enum ath10k_htc_ep_id eid)
+static void ath10k_wmi_op_ep_tx_credits(struct ath10k *ar)
{
wake_up(&ar->wmi.tx_credits_wq);
}
--
1.7.9.5
^ permalink raw reply related
* [PATCHv2 3/7] ath10k: make WMI commands block by design
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1379074618-30534-1-git-send-email-michal.kazior@tieto.com>
This will be necessary for further changes in
command submission scheme.
Once HTC is cleaned up WMI commands will finally
block.
This requires for SWBA to be processed in a
non-atomic context for now. Once other necessary
changes are in this will be reverted.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
---
drivers/net/wireless/ath/ath10k/core.h | 1 +
drivers/net/wireless/ath/ath10k/trace.h | 11 ++++---
drivers/net/wireless/ath/ath10k/wmi.c | 54 +++++++++++++++++++------------
3 files changed, 42 insertions(+), 24 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/core.h b/drivers/net/wireless/ath/ath10k/core.h
index 22b17d6..c953a33 100644
--- a/drivers/net/wireless/ath/ath10k/core.h
+++ b/drivers/net/wireless/ath/ath10k/core.h
@@ -114,6 +114,7 @@ struct ath10k_wmi {
struct completion unified_ready;
atomic_t pending_tx_count;
wait_queue_head_t wq;
+ wait_queue_head_t tx_credits_wq;
struct sk_buff_head wmi_event_list;
struct work_struct wmi_event_work;
diff --git a/drivers/net/wireless/ath/ath10k/trace.h b/drivers/net/wireless/ath/ath10k/trace.h
index bf1ceb8..fd53130 100644
--- a/drivers/net/wireless/ath/ath10k/trace.h
+++ b/drivers/net/wireless/ath/ath10k/trace.h
@@ -111,26 +111,29 @@ TRACE_EVENT(ath10k_log_dbg_dump,
);
TRACE_EVENT(ath10k_wmi_cmd,
- TP_PROTO(int id, void *buf, size_t buf_len),
+ TP_PROTO(int id, void *buf, size_t buf_len, int ret),
- TP_ARGS(id, buf, buf_len),
+ TP_ARGS(id, buf, buf_len, ret),
TP_STRUCT__entry(
__field(unsigned int, id)
__field(size_t, buf_len)
__dynamic_array(u8, buf, buf_len)
+ __field(int ret)
),
TP_fast_assign(
__entry->id = id;
__entry->buf_len = buf_len;
+ __entry->ret = ret;
memcpy(__get_dynamic_array(buf), buf, buf_len);
),
TP_printk(
- "id %d len %zu",
+ "id %d len %zu ret %d",
__entry->id,
- __entry->buf_len
+ __entry->buf_len,
+ __entry->ret
)
);
diff --git a/drivers/net/wireless/ath/ath10k/wmi.c b/drivers/net/wireless/ath/ath10k/wmi.c
index 32fd5e7..66cd892 100644
--- a/drivers/net/wireless/ath/ath10k/wmi.c
+++ b/drivers/net/wireless/ath/ath10k/wmi.c
@@ -90,13 +90,12 @@ static void ath10k_wmi_htc_tx_complete(struct ath10k *ar, struct sk_buff *skb)
wake_up(&ar->wmi.wq);
}
-/* WMI command API */
-static int ath10k_wmi_cmd_send(struct ath10k *ar, struct sk_buff *skb,
- enum wmi_cmd_id cmd_id)
+static int ath10k_wmi_cmd_send_nowait(struct ath10k *ar, struct sk_buff *skb,
+ enum wmi_cmd_id cmd_id)
{
struct ath10k_skb_cb *skb_cb = ATH10K_SKB_CB(skb);
struct wmi_cmd_hdr *cmd_hdr;
- int status;
+ int ret;
u32 cmd = 0;
if (skb_push(skb, sizeof(struct wmi_cmd_hdr)) == NULL)
@@ -107,26 +106,40 @@ static int ath10k_wmi_cmd_send(struct ath10k *ar, struct sk_buff *skb,
cmd_hdr = (struct wmi_cmd_hdr *)skb->data;
cmd_hdr->cmd_id = __cpu_to_le32(cmd);
- if (atomic_add_return(1, &ar->wmi.pending_tx_count) >
- WMI_MAX_PENDING_TX_COUNT) {
- /* avoid using up memory when FW hangs */
- dev_kfree_skb(skb);
- atomic_dec(&ar->wmi.pending_tx_count);
- return -EBUSY;
- }
-
memset(skb_cb, 0, sizeof(*skb_cb));
+ ret = ath10k_htc_send(&ar->htc, ar->wmi.eid, skb);
+ trace_ath10k_wmi_cmd(cmd_id, skb->data, skb->len, ret);
- trace_ath10k_wmi_cmd(cmd_id, skb->data, skb->len);
+ if (ret)
+ goto err_pull;
- status = ath10k_htc_send(&ar->htc, ar->wmi.eid, skb);
- if (status) {
+ return 0;
+
+err_pull:
+ skb_pull(skb, sizeof(struct wmi_cmd_hdr));
+ return ret;
+}
+
+static void ath10k_wmi_op_ep_tx_credits(struct ath10k *ar,
+ enum ath10k_htc_ep_id eid)
+{
+ wake_up(&ar->wmi.tx_credits_wq);
+}
+
+static int ath10k_wmi_cmd_send(struct ath10k *ar, struct sk_buff *skb,
+ enum wmi_cmd_id cmd_id)
+{
+ int ret = -EINVAL;
+
+ wait_event_timeout(ar->wmi.tx_credits_wq, ({
+ ret = ath10k_wmi_cmd_send_nowait(ar, skb, cmd_id);
+ (ret != -EAGAIN);
+ }), 3*HZ);
+
+ if (ret)
dev_kfree_skb_any(skb);
- atomic_dec(&ar->wmi.pending_tx_count);
- return status;
- }
- return 0;
+ return ret;
}
static int ath10k_wmi_event_scan(struct ath10k *ar, struct sk_buff *skb)
@@ -1168,7 +1181,6 @@ static void ath10k_wmi_process_rx(struct ath10k *ar, struct sk_buff *skb)
/* some events require to be handled ASAP
* thus can't be defered to a worker thread */
switch (event_id) {
- case WMI_HOST_SWBA_EVENTID:
case WMI_MGMT_RX_EVENTID:
ath10k_wmi_event_process(ar, skb);
return;
@@ -1186,6 +1198,7 @@ int ath10k_wmi_attach(struct ath10k *ar)
init_completion(&ar->wmi.service_ready);
init_completion(&ar->wmi.unified_ready);
init_waitqueue_head(&ar->wmi.wq);
+ init_waitqueue_head(&ar->wmi.tx_credits_wq);
skb_queue_head_init(&ar->wmi.wmi_event_list);
INIT_WORK(&ar->wmi.wmi_event_work, ath10k_wmi_event_work);
@@ -1215,6 +1228,7 @@ int ath10k_wmi_connect_htc_service(struct ath10k *ar)
/* these fields are the same for all service endpoints */
conn_req.ep_ops.ep_tx_complete = ath10k_wmi_htc_tx_complete;
conn_req.ep_ops.ep_rx_complete = ath10k_wmi_process_rx;
+ conn_req.ep_ops.ep_tx_credits = ath10k_wmi_op_ep_tx_credits;
/* connect to control service */
conn_req.service_id = ATH10K_HTC_SVC_ID_WMI_CONTROL;
--
1.7.9.5
^ permalink raw reply related
* [PATCHv2 2/7] ath10k: add HTC TX credits replenishing notification
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1379074618-30534-1-git-send-email-michal.kazior@tieto.com>
This will allow higher layers to anticipate and
act upon TX credits renewal. This will be
important for some future rework of WMI command
submission.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
---
drivers/net/wireless/ath/ath10k/htc.c | 9 +++++++++
drivers/net/wireless/ath/ath10k/htc.h | 1 +
2 files changed, 10 insertions(+)
diff --git a/drivers/net/wireless/ath/ath10k/htc.c b/drivers/net/wireless/ath/ath10k/htc.c
index 1da0f82..49da4e5 100644
--- a/drivers/net/wireless/ath/ath10k/htc.c
+++ b/drivers/net/wireless/ath/ath10k/htc.c
@@ -150,6 +150,9 @@ err:
ep->tx_credits += credits;
spin_unlock_bh(&htc->tx_lock);
+ if (ep->ep_ops.ep_tx_credits)
+ ep->ep_ops.ep_tx_credits(htc->ar);
+
/* this is the simplest way to handle out-of-resources for non-credit
* based endpoints. credit based endpoints can still get -ENOSR, but
* this is highly unlikely as credit reservation should prevent that */
@@ -302,6 +305,12 @@ ath10k_htc_process_credit_report(struct ath10k_htc *htc,
ep = &htc->endpoint[report->eid];
ep->tx_credits += report->credits;
+ if (ep->ep_ops.ep_tx_credits) {
+ spin_unlock_bh(&htc->tx_lock);
+ ep->ep_ops.ep_tx_credits(htc->ar);
+ spin_lock_bh(&htc->tx_lock);
+ }
+
if (ep->tx_credits && !skb_queue_empty(&ep->tx_queue))
queue_work(htc->ar->workqueue, &ep->send_work);
}
diff --git a/drivers/net/wireless/ath/ath10k/htc.h b/drivers/net/wireless/ath/ath10k/htc.h
index e1dd8c7..92ca29b 100644
--- a/drivers/net/wireless/ath/ath10k/htc.h
+++ b/drivers/net/wireless/ath/ath10k/htc.h
@@ -276,6 +276,7 @@ struct ath10k_htc_ops {
struct ath10k_htc_ep_ops {
void (*ep_tx_complete)(struct ath10k *, struct sk_buff *);
void (*ep_rx_complete)(struct ath10k *, struct sk_buff *);
+ void (*ep_tx_credits)(struct ath10k *);
};
/* service connection information */
--
1.7.9.5
^ permalink raw reply related
* [PATCHv2 1/7] ath10k: simplify HTC credits calculation
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1379074618-30534-1-git-send-email-michal.kazior@tieto.com>
Credit calculation was overly complex
unnecessarily. Now skb dequeing is more unified.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
---
v2:
* use DIV_ROUND_UP
drivers/net/wireless/ath/ath10k/htc.c | 58 ++++++---------------------------
1 file changed, 10 insertions(+), 48 deletions(-)
diff --git a/drivers/net/wireless/ath/ath10k/htc.c b/drivers/net/wireless/ath/ath10k/htc.c
index f03fd8d..1da0f82 100644
--- a/drivers/net/wireless/ath/ath10k/htc.c
+++ b/drivers/net/wireless/ath/ath10k/htc.c
@@ -167,49 +167,6 @@ err:
return ret;
}
-static struct sk_buff *ath10k_htc_get_skb_credit_based(struct ath10k_htc *htc,
- struct ath10k_htc_ep *ep,
- u8 *credits)
-{
- struct sk_buff *skb;
- struct ath10k_skb_cb *skb_cb;
- int credits_required;
- int remainder;
- unsigned int transfer_len;
-
- lockdep_assert_held(&htc->tx_lock);
-
- skb = __skb_dequeue(&ep->tx_queue);
- if (!skb)
- return NULL;
-
- skb_cb = ATH10K_SKB_CB(skb);
- transfer_len = skb->len;
-
- if (likely(transfer_len <= htc->target_credit_size)) {
- credits_required = 1;
- } else {
- /* figure out how many credits this message requires */
- credits_required = transfer_len / htc->target_credit_size;
- remainder = transfer_len % htc->target_credit_size;
-
- if (remainder)
- credits_required++;
- }
-
- ath10k_dbg(ATH10K_DBG_HTC, "Credits required %d got %d\n",
- credits_required, ep->tx_credits);
-
- if (ep->tx_credits < credits_required) {
- __skb_queue_head(&ep->tx_queue, skb);
- return NULL;
- }
-
- ep->tx_credits -= credits_required;
- *credits = credits_required;
- return skb;
-}
-
static void ath10k_htc_send_work(struct work_struct *work)
{
struct ath10k_htc_ep *ep = container_of(work,
@@ -224,11 +181,16 @@ static void ath10k_htc_send_work(struct work_struct *work)
ath10k_htc_send_complete_check(ep, 0);
spin_lock_bh(&htc->tx_lock);
- if (ep->tx_credit_flow_enabled)
- skb = ath10k_htc_get_skb_credit_based(htc, ep,
- &credits);
- else
- skb = __skb_dequeue(&ep->tx_queue);
+ skb = __skb_dequeue(&ep->tx_queue);
+
+ if (ep->tx_credit_flow_enabled) {
+ credits = DIV_ROUND_UP(skb->len,
+ htc->target_credit_size);
+ if (ep->tx_credits < credits) {
+ __skb_queue_head(&ep->tx_queue, skb);
+ skb = NULL;
+ }
+ }
spin_unlock_bh(&htc->tx_lock);
if (!skb)
--
1.7.9.5
^ permalink raw reply related
* [PATCHv2 0/7] ath10k: improve TX path
From: Michal Kazior @ 2013-09-13 12:16 UTC (permalink / raw)
To: ath10k; +Cc: linux-wireless, Michal Kazior
In-Reply-To: <1378821003-22925-1-git-send-email-michal.kazior@tieto.com>
Hi,
This patchset addresses two issues:
* system/userspace starvation on heavy briding
UDP TX
* unstable/inconsistent UDP TX throughput
In short the patchset simplifies TX path by
removing HTC TX workers, makes WMI commands block
and makes ath10k more responsive to queues
becoming full. This contributes to both improved
throughput and makes the system more responsive
under heavy UDP TX load.
Max stable briding TX (ath10k as TX):
UDP w/o patchset: 520mbps
UDP w/ patchset: 570mbps
TCP w/o patchset: 350mbps
TCP w/ patchset: 400mbps
Measured with two 2x2 cards, one acting as an AP
on AP135 board, the other as a STA on a core i5
laptop. The AP135 was passing traffic from a
station on LAN/eth1 to the STA/wlan0.
v2:
* use DIV_ROUND_UP
* comment usage of ath10k_wmi_tx_beacons_nowait
* additional check for tx_credits_flow_enabled
Michal Kazior (7):
ath10k: simplify HTC credits calculation
ath10k: add HTC TX credits replenishing notification
ath10k: make WMI commands block by design
ath10k: simplify HTC command submitting
ath10k: improve beacon submission latency
ath10k: remove wmi pending count limit
ath10k: remove wmi event worker thread
drivers/net/wireless/ath/ath10k/core.h | 7 +-
drivers/net/wireless/ath/ath10k/htc.c | 215 +++++++------------------------
drivers/net/wireless/ath/ath10k/htc.h | 5 +-
drivers/net/wireless/ath/ath10k/mac.c | 16 +--
drivers/net/wireless/ath/ath10k/trace.h | 11 +-
drivers/net/wireless/ath/ath10k/wmi.c | 201 ++++++++++++++---------------
drivers/net/wireless/ath/ath10k/wmi.h | 5 +-
7 files changed, 160 insertions(+), 300 deletions(-)
--
1.7.9.5
^ permalink raw reply
* WEP/WPA/WPA2 implementation in software
From: Prabhunath G @ 2013-09-13 11:53 UTC (permalink / raw)
To: linux-wireless, kernelnewbies
Dear All,
Suppose there is no hardware support for
encryption/decryption for Wi-Fi devices. Will the data be
encrypted/decrypted by software implementation. If yes, is it done at
the application level or at the kernel level (mac80211).
Kindly clarify.
If the question has to be rephrased, please do it
Regards,
Prabhunath G
^ permalink raw reply
* Re: [RFC 4/5] mac80211: add generic quiesce procedure
From: Stanislaw Gruszka @ 2013-09-13 11:26 UTC (permalink / raw)
To: Johannes Berg; +Cc: ilw, linux-wireless
In-Reply-To: <1379070837.14883.11.camel@jlt4.sipsolutions.net>
On Fri, Sep 13, 2013 at 01:13:57PM +0200, Johannes Berg wrote:
> On Fri, 2013-09-13 at 12:36 +0200, Stanislaw Gruszka wrote:
> > Add function ieee80211_quiesce() end export it to allow be used by
> > drivers. It is intended to be used before restart_hw to stop mac80211
> > timers/works . Reverse quiesce on ieee80211_reconfig().
>
> Why can't this be done inside restart_hw()?
iwlwifi stop device before call restart_hw() and we need to
quiesce before iwlwifi stop device (like done in patch 5).
Stanislaw
^ permalink raw reply
* Re: [RFC 4/5] mac80211: add generic quiesce procedure
From: Johannes Berg @ 2013-09-13 11:13 UTC (permalink / raw)
To: Stanislaw Gruszka; +Cc: ilw, linux-wireless
In-Reply-To: <1379068568-27552-5-git-send-email-sgruszka@redhat.com>
On Fri, 2013-09-13 at 12:36 +0200, Stanislaw Gruszka wrote:
> Add function ieee80211_quiesce() end export it to allow be used by
> drivers. It is intended to be used before restart_hw to stop mac80211
> timers/works . Reverse quiesce on ieee80211_reconfig().
Why can't this be done inside restart_hw()?
johannes
^ permalink raw reply
* [RFC 5/5] iwlwifi: quiesce mac80211 before fw restart
From: Stanislaw Gruszka @ 2013-09-13 10:36 UTC (permalink / raw)
To: ilw; +Cc: linux-wireless
In-Reply-To: <1379068568-27552-1-git-send-email-sgruszka@redhat.com>
Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com>
---
drivers/net/wireless/iwlwifi/dvm/mac80211.c | 1 +
drivers/net/wireless/iwlwifi/dvm/main.c | 4 +++-
2 files changed, 4 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/iwlwifi/dvm/mac80211.c b/drivers/net/wireless/iwlwifi/dvm/mac80211.c
index cae4d31..e819e5b 100644
--- a/drivers/net/wireless/iwlwifi/dvm/mac80211.c
+++ b/drivers/net/wireless/iwlwifi/dvm/mac80211.c
@@ -393,6 +393,7 @@ static int iwlagn_mac_suspend(struct ieee80211_hw *hw,
error:
priv->wowlan = false;
+ ieee80211_quiesce(priv->hw);
iwlagn_prepare_restart(priv);
ieee80211_restart_hw(priv->hw);
out:
diff --git a/drivers/net/wireless/iwlwifi/dvm/main.c b/drivers/net/wireless/iwlwifi/dvm/main.c
index 7aad766..2e0fbea 100644
--- a/drivers/net/wireless/iwlwifi/dvm/main.c
+++ b/drivers/net/wireless/iwlwifi/dvm/main.c
@@ -975,10 +975,12 @@ static void iwl_bg_restart(struct work_struct *data)
return;
if (test_and_clear_bit(STATUS_FW_ERROR, &priv->status)) {
+ iwl_cancel_deferred_work(priv);
+ if (priv->mac80211_registered)
+ ieee80211_quiesce(priv->hw);
mutex_lock(&priv->mutex);
iwlagn_prepare_restart(priv);
mutex_unlock(&priv->mutex);
- iwl_cancel_deferred_work(priv);
if (priv->mac80211_registered)
ieee80211_restart_hw(priv->hw);
else
--
1.8.3.1
^ permalink raw reply related
* [RFC 2/5] Revert "mac80211: cleanup suspend/resume on ibss mode"
From: Stanislaw Gruszka @ 2013-09-13 10:36 UTC (permalink / raw)
To: ilw; +Cc: linux-wireless
In-Reply-To: <1379068568-27552-1-git-send-email-sgruszka@redhat.com>
This reverts commit a61829437e68c8b2036cf5005ed0e875451c9120.
---
net/mac80211/ibss.c | 29 ++++++++++++++++++++++++++++-
net/mac80211/ieee80211_i.h | 4 ++++
2 files changed, 32 insertions(+), 1 deletion(-)
diff --git a/net/mac80211/ibss.c b/net/mac80211/ibss.c
index a12afe7..91d1b29 100644
--- a/net/mac80211/ibss.c
+++ b/net/mac80211/ibss.c
@@ -1155,9 +1155,36 @@ static void ieee80211_ibss_timer(unsigned long data)
{
struct ieee80211_sub_if_data *sdata =
(struct ieee80211_sub_if_data *) data;
+ struct ieee80211_if_ibss *ifibss = &sdata->u.ibss;
+ struct ieee80211_local *local = sdata->local;
- ieee80211_queue_work(&sdata->local->hw, &sdata->work);
+ if (local->quiescing) {
+ ifibss->timer_running = true;
+ return;
+ }
+
+ ieee80211_queue_work(&local->hw, &sdata->work);
+}
+
+#ifdef CONFIG_PM
+void ieee80211_ibss_quiesce(struct ieee80211_sub_if_data *sdata)
+{
+ struct ieee80211_if_ibss *ifibss = &sdata->u.ibss;
+
+ if (del_timer_sync(&ifibss->timer))
+ ifibss->timer_running = true;
+}
+
+void ieee80211_ibss_restart(struct ieee80211_sub_if_data *sdata)
+{
+ struct ieee80211_if_ibss *ifibss = &sdata->u.ibss;
+
+ if (ifibss->timer_running) {
+ add_timer(&ifibss->timer);
+ ifibss->timer_running = false;
+ }
}
+#endif
void ieee80211_ibss_setup_sdata(struct ieee80211_sub_if_data *sdata)
{
diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h
index 4b7ef89..96ac8f8 100644
--- a/net/mac80211/ieee80211_i.h
+++ b/net/mac80211/ieee80211_i.h
@@ -492,6 +492,8 @@ struct ieee80211_if_ibss {
u32 basic_rates;
+ bool timer_running;
+
bool fixed_bssid;
bool fixed_channel;
bool privacy;
@@ -1329,6 +1331,8 @@ void ieee80211_ibss_rx_no_sta(struct ieee80211_sub_if_data *sdata,
int ieee80211_ibss_join(struct ieee80211_sub_if_data *sdata,
struct cfg80211_ibss_params *params);
int ieee80211_ibss_leave(struct ieee80211_sub_if_data *sdata);
+void ieee80211_ibss_quiesce(struct ieee80211_sub_if_data *sdata);
+void ieee80211_ibss_restart(struct ieee80211_sub_if_data *sdata);
void ieee80211_ibss_work(struct ieee80211_sub_if_data *sdata);
void ieee80211_ibss_rx_queued_mgmt(struct ieee80211_sub_if_data *sdata,
struct sk_buff *skb);
--
1.8.3.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox