* Re: [PATCH] dt-bindings: net: wireless: Add sg parameters dts parsing
From: Arend van Spriel @ 2017-12-11 9:55 UTC (permalink / raw)
To: Chi-Hsien Lin, devicetree
Cc: linux-wireless, brcm80211-dev-list, brcm80211-dev-list.pdl,
Kalle Valo
In-Reply-To: <20171211081144.GA140567@aremote07.aus.cypress.com>
On 12/11/2017 9:11 AM, Chi-Hsien Lin wrote:
> broken_sg_support, sd_head_align, and sd_sgentry_align are used in
> brcmfmac code but not configurable in dts file. Add the parsing logic.
> Now they can be configured like below in dts:
> brcm,broken_sg_support;
> brcm,sd_head_align = <4>;
> brcm,sd_sgentry_align = <4>;
>
> Signed-off-by: Chi-hsien Lin <chi-hsien.lin@cypress.com>
> ---
> .../devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt b/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt
> index 86602f2..4d42f0d 100644
> --- a/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt
> +++ b/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt
> @@ -17,6 +17,11 @@ Optional properties:
> When not specified the device will use in-band SDIO interrupts.
> - interrupt-names : name of the out-of-band interrupt, which must be set
> to "host-wake".
> + - broken_sg_support : flag for broken sg list support of SDIO host controller.
> + Set this to true if the SDIO host controller has higher align
> + requirement than 32 bytes for each scatterlist item.
In DT context, this does not characterize the device but as described it
is about the SDIO host controller. I think the same applies to the
properties below, but not entirely sure about that.
> + - sg_head_align : alignment requirement for start of data buffer.
> + - sg_sgentry_align : length alignment requirement for each sg entry.
Regards,
Arend
^ permalink raw reply
* pull-request: mac80211 2017-12-11
From: Johannes Berg @ 2017-12-11 9:52 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-wireless
Hi Dave,
Three fixes, two related to build issues with the new regdb stuff,
and one for some patch overlap problem that caused locking to be
missing which in turn caused lots of warnings.
Please pull and let me know if there's any problem.
Thanks,
johannes
The following changes since commit ccab371f746abca05a599d074cb3b95a549ef590:
Merge branch 'sfp-phylink-fixes' (2017-12-01 15:18:42 -0500)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211.git tags/mac80211-for-davem-2017-12-11
for you to fetch changes up to 0afe9d4ab9d40c281bdcdd118661fe8e4bdcef18:
mac80211: fix locking in ieee80211_sta_tear_down_BA_sessions (2017-12-11 10:50:00 +0100)
----------------------------------------------------------------
Three fixes:
* for certificate C file generation, don't use hexdump as it's
not always installed by default, use pure posix instead (od/sed)
* for certificate C file generation, don't write the file if
anything fails, so the build abort will not cause a bad build
upon a second attempt
* fix locking in ieee80211_sta_tear_down_BA_sessions() which had
been causing lots of locking warnings
----------------------------------------------------------------
Johannes Berg (3):
wireless: replace usage of hexdump with od/sed
wireless: don't write C files on failures
mac80211: fix locking in ieee80211_sta_tear_down_BA_sessions
net/mac80211/ht.c | 5 ++---
net/wireless/Makefile | 48 ++++++++++++++++++++++++++++++++++++++----------
2 files changed, 40 insertions(+), 13 deletions(-)
^ permalink raw reply
* Re: [PATCH] brcmfmac: enlarge buffer size of caps to 512 bytes
From: Arend van Spriel @ 2017-12-11 9:10 UTC (permalink / raw)
To: Wright Feng, franky.lin, hante.meuleman, kvalo, chi-hsien.lin
Cc: linux-wireless, brcm80211-dev-list.pdl
In-Reply-To: <1512977901-12726-1-git-send-email-wright.feng@cypress.com>
On 12/11/2017 8:38 AM, Wright Feng wrote:
> The buffer size of return of cap iovar is greater than 256 bytes in some
> firmwares. For instance, the return size of cap iovar is 271 bytes in 4373
> 13.10.246.79 firmare. It makes feature capability parsing failed because
> caps buffer is default value.
> So we enlarge caps buffer size to 512 bytes and add the error print for
> cap iovar error.
Looks fine to me. However, firmware side is also broken for some chips
where released firmware uses 256 bytes (= WLC_IOCTL_SMLEN) but the
capability tokens do no longer fit that size and we get a cut-off
capability string.
Anyway...
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com>
> Signed-off-by: Wright Feng <wright.feng@cypress.com>
> ---
> drivers/net/wireless/broadcom/brcm80211/brcmfmac/feature.c | 12 +++++++++---
> 1 file changed, 9 insertions(+), 3 deletions(-)
^ permalink raw reply
* Re: [PATCH] wcn36xx: Add hardware scan offload support
From: Loic Poulain @ 2017-12-11 8:57 UTC (permalink / raw)
To: Bjorn Andersson
Cc: kvalo, linux-arm-msm, linux-wireless, wcn36xx, Eugene Krasnikov
In-Reply-To: <20171209003452.GA3441@builder>
Hi Bjorn,
On 9 December 2017 at 01:34, Bjorn Andersson <bjorn.andersson@linaro.org> wrote:
>
>> diff --git a/drivers/net/wireless/ath/wcn36xx/smd.c b/drivers/net/wireless/ath/wcn36xx/smd.c
> [..]
>> +static int wcn36xx_smd_hw_scan_ind(struct wcn36xx *wcn, void *buf, size_t len)
>> +{
>> + struct wcn36xx_hal_scan_offload_ind *rsp = buf;
>> + struct cfg80211_scan_info scan_info = {};
>> +
>> + if (len != sizeof(*rsp)) {
>> + wcn36xx_warn("Corrupted delete scan indication\n");
>> + return -EIO;
>> + }
>> +
>> + wcn36xx_dbg(WCN36XX_DBG_HAL, "scan indication (type %x)", rsp->type);
>> +
>> + switch (rsp->type) {
>> + case WCN36XX_HAL_SCAN_IND_FAILED:
>> + scan_info.aborted = true;
>> + case WCN36XX_HAL_SCAN_IND_COMPLETED:
>> + mutex_lock(&wcn->scan_lock);
>
> Grabbing this mutex with DEBUG_ATOMIC_SLEEP causes issues, but that's
> because the locking in ind_smd_work() is to excessive. Will reply with a
> fix for this.
Oops you're right, thanks for the patch.
Regards,
Loic
^ permalink raw reply
* [PATCH v2] wcn36xx: Fix dynamic power saving
From: Loic Poulain @ 2017-12-11 8:52 UTC (permalink / raw)
To: kvalo
Cc: linux-wireless, wcn36xx, linux-arm-msm, k.eugene.e,
bjorn.andersson, Loic Poulain
Since driver does not report hardware dynamic power saving cap,
this is up to the mac80211 to manage power saving timeout and
state machine, using the ieee80211 config callback to report
PS changes. This patch enables/disables PS mode according to
the new configuration.
Remove old behaviour enabling PS mode in a static way, this make
the device unusable when power save is enabled since device is
forced to PS regardless RX/TX traffic.
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
---
v2: remove error msg on unbalanced bmps exit
return -EALREADY if not in bmps mode
drivers/net/wireless/ath/wcn36xx/main.c | 23 ++++++++++++-----------
drivers/net/wireless/ath/wcn36xx/pmc.c | 6 ++++--
2 files changed, 16 insertions(+), 13 deletions(-)
diff --git a/drivers/net/wireless/ath/wcn36xx/main.c b/drivers/net/wireless/ath/wcn36xx/main.c
index f0b4d43..436b8ea 100644
--- a/drivers/net/wireless/ath/wcn36xx/main.c
+++ b/drivers/net/wireless/ath/wcn36xx/main.c
@@ -384,6 +384,18 @@ static int wcn36xx_config(struct ieee80211_hw *hw, u32 changed)
}
}
+ if (changed & IEEE80211_CONF_CHANGE_PS) {
+ list_for_each_entry(tmp, &wcn->vif_list, list) {
+ vif = wcn36xx_priv_to_vif(tmp);
+ if (hw->conf.flags & IEEE80211_CONF_PS) {
+ if (vif->bss_conf.ps) /* ps allowed ? */
+ wcn36xx_pmc_enter_bmps_state(wcn, vif);
+ } else {
+ wcn36xx_pmc_exit_bmps_state(wcn, vif);
+ }
+ }
+ }
+
mutex_unlock(&wcn->conf_mutex);
return 0;
@@ -747,17 +759,6 @@ static void wcn36xx_bss_info_changed(struct ieee80211_hw *hw,
vif_priv->dtim_period = bss_conf->dtim_period;
}
- if (changed & BSS_CHANGED_PS) {
- wcn36xx_dbg(WCN36XX_DBG_MAC,
- "mac bss PS set %d\n",
- bss_conf->ps);
- if (bss_conf->ps) {
- wcn36xx_pmc_enter_bmps_state(wcn, vif);
- } else {
- wcn36xx_pmc_exit_bmps_state(wcn, vif);
- }
- }
-
if (changed & BSS_CHANGED_BSSID) {
wcn36xx_dbg(WCN36XX_DBG_MAC, "mac bss changed_bssid %pM\n",
bss_conf->bssid);
diff --git a/drivers/net/wireless/ath/wcn36xx/pmc.c b/drivers/net/wireless/ath/wcn36xx/pmc.c
index 589fe5f..1976b80 100644
--- a/drivers/net/wireless/ath/wcn36xx/pmc.c
+++ b/drivers/net/wireless/ath/wcn36xx/pmc.c
@@ -45,8 +45,10 @@ int wcn36xx_pmc_exit_bmps_state(struct wcn36xx *wcn,
struct wcn36xx_vif *vif_priv = wcn36xx_vif_to_priv(vif);
if (WCN36XX_BMPS != vif_priv->pw_state) {
- wcn36xx_err("Not in BMPS mode, no need to exit from BMPS mode!\n");
- return -EINVAL;
+ /* Unbalanced call or last BMPS enter failed */
+ wcn36xx_dbg(WCN36XX_DBG_PMC,
+ "Not in BMPS mode, no need to exit\n");
+ return -EALREADY;
}
wcn36xx_smd_exit_bmps(wcn, vif);
vif_priv->pw_state = WCN36XX_FULL_POWER;
--
2.7.4
^ permalink raw reply related
* Re: [PATCH] wcn36xx: Fix dynamic power saving
From: Loic Poulain @ 2017-12-11 8:39 UTC (permalink / raw)
To: Bjorn Andersson
Cc: kvalo, wcn36xx, linux-wireless, linux-arm-msm, Nicolas Dechesne,
Eugene Krasnikov
In-Reply-To: <20171209004534.GB3441@builder>
Hi Bjorn,
On 9 December 2017 at 01:45, Bjorn Andersson <bjorn.andersson@linaro.org> wrote:
>> }
>> }
>>
>> + if (changed & IEEE80211_CONF_CHANGE_PS) {
>> + list_for_each_entry(tmp, &wcn->vif_list, list) {
>> + vif = wcn36xx_priv_to_vif(tmp);
>> + if (hw->conf.flags & IEEE80211_CONF_PS) {
>> + if (vif->bss_conf.ps) /* ps allowed ? */
>> + wcn36xx_pmc_enter_bmps_state(wcn, vif);
>> + } else {
>> + wcn36xx_pmc_exit_bmps_state(wcn, vif);
>
> During startup I get the error print from wcn36xx_pmc_exit_bmps_state()
> that we're not in BMPS state. There's no harm in this, but the error
> might concern people.
>
> How about we in addition to this, change the wcn36xx_err() to a
> wcn36xx_dbg(PMC...) ?
Thanks for reporting this, I assume mac80211 disable PS mode before enabling it
on timeout... It makes sense to avoid this error msg since balancing
is well managed
by this driver.
Regards,
Loic
^ permalink raw reply
* [PATCH] dt-bindings: net: wireless: Add sg parameters dts parsing
From: Chi-Hsien Lin @ 2017-12-11 8:11 UTC (permalink / raw)
To: devicetree
Cc: linux-wireless, brcm80211-dev-list, brcm80211-dev-list.pdl,
Arend van Spriel, Kalle Valo
broken_sg_support, sd_head_align, and sd_sgentry_align are used in
brcmfmac code but not configurable in dts file. Add the parsing logic.
Now they can be configured like below in dts:
brcm,broken_sg_support;
brcm,sd_head_align = <4>;
brcm,sd_sgentry_align = <4>;
Signed-off-by: Chi-hsien Lin <chi-hsien.lin@cypress.com>
---
.../devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt b/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt
index 86602f2..4d42f0d 100644
--- a/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt
+++ b/Documentation/devicetree/bindings/net/wireless/brcm,bcm43xx-fmac.txt
@@ -17,6 +17,11 @@ Optional properties:
When not specified the device will use in-band SDIO interrupts.
- interrupt-names : name of the out-of-band interrupt, which must be set
to "host-wake".
+ - broken_sg_support : flag for broken sg list support of SDIO host controller.
+ Set this to true if the SDIO host controller has higher align
+ requirement than 32 bytes for each scatterlist item.
+ - sg_head_align : alignment requirement for start of data buffer.
+ - sg_sgentry_align : length alignment requirement for each sg entry.
Example:
@@ -36,5 +41,8 @@ mmc3: mmc@1c12000 {
interrupt-parent = <&pio>;
interrupts = <10 8>; /* PH10 / EINT10 */
interrupt-names = "host-wake";
+ brcm,broken_sg_support;
+ brcm,sd_head_align = <4>;
+ brcm,sd_sgentry_align = <4>;
};
};
--
2.1.0
^ permalink raw reply related
* [PATCH V2] brcmfmac: Add sg parameters dts parsing
From: Chi-Hsien Lin @ 2017-12-11 8:06 UTC (permalink / raw)
To: linux-wireless
Cc: brcm80211-dev-list, brcm80211-dev-list.pdl, franky.lin,
hante.meuleman, Wright Feng, Stanley Hsu, devicetree
broken_sg_support, sd_head_align, and sd_sgentry_align are used in
brcmfmac code but not configurable in dts file. Add the parsing logic.
Now they can be configured like below in dts:
brcm,broken_sg_support;
brcm,sd_head_align = <4>;
brcm,sd_sgentry_align = <4>;
Signed-off-by: Chi-hsien Lin <chi-hsien.lin@cypress.com>
---
drivers/net/wireless/broadcom/brcm80211/brcmfmac/of.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/of.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/of.c
index aee6e59..fd028b1 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/of.c
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/of.c
@@ -39,6 +39,13 @@ void brcmf_of_probe(struct device *dev, enum brcmf_bus_type bus_type,
if (of_property_read_u32(np, "brcm,drive-strength", &val) == 0)
sdio->drive_strength = val;
+ sdio->broken_sg_support = of_property_read_bool(np,
+ "brcm,broken_sg_support");
+ if (of_property_read_u32(np, "brcm,sd_head_align", &val) == 0)
+ sdio->sd_head_align = (u16)val;
+ if (of_property_read_u32(np, "brcm,sd_sgentry_align", &val) == 0)
+ sdio->sd_sgentry_align = (u16)val;
+
/* make sure there are interrupts defined in the node */
if (!of_find_property(np, "interrupts", NULL))
return;
--
2.1.0
^ permalink raw reply related
* [PATCH] brcmfmac: enlarge buffer size of caps to 512 bytes
From: Wright Feng @ 2017-12-11 7:38 UTC (permalink / raw)
To: arend.vanspriel, franky.lin, hante.meuleman, kvalo, chi-hsien.lin
Cc: wright.feng, linux-wireless, brcm80211-dev-list.pdl
The buffer size of return of cap iovar is greater than 256 bytes in some
firmwares. For instance, the return size of cap iovar is 271 bytes in 4373
13.10.246.79 firmare. It makes feature capability parsing failed because
caps buffer is default value.
So we enlarge caps buffer size to 512 bytes and add the error print for
cap iovar error.
Signed-off-by: Wright Feng <wright.feng@cypress.com>
---
drivers/net/wireless/broadcom/brcm80211/brcmfmac/feature.c | 12 +++++++++---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/feature.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/feature.c
index 53ae302..47de35a 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/feature.c
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/feature.c
@@ -130,13 +130,19 @@ static void brcmf_feat_iovar_data_set(struct brcmf_if *ifp,
}
}
+#define MAX_CAPS_BUFFER_SIZE 512
static void brcmf_feat_firmware_capabilities(struct brcmf_if *ifp)
{
- char caps[256];
+ char caps[MAX_CAPS_BUFFER_SIZE];
enum brcmf_feat_id id;
- int i;
+ int i, err;
+
+ err = brcmf_fil_iovar_data_get(ifp, "cap", caps, sizeof(caps));
+ if (err) {
+ brcmf_err("could not get firmware cap (%d)\n", err);
+ return;
+ }
- brcmf_fil_iovar_data_get(ifp, "cap", caps, sizeof(caps));
brcmf_dbg(INFO, "[ %s]\n", caps);
for (i = 0; i < ARRAY_SIZE(brcmf_fwcap_map); i++) {
--
1.9.1
^ permalink raw reply related
* [PATCH 1/1] rtlwifi: always initialize variables given to RT_TRACE()
From: Nicolas Iooss @ 2017-12-10 19:51 UTC (permalink / raw)
To: Larry Finger, Chaoming Li, Kalle Valo, linux-wireless
Cc: netdev, linux-kernel, Nicolas Iooss
In rtl_rx_ampdu_apply(), when rtlpriv->cfg->ops->get_btc_status()
returns false, RT_TRACE() is called with the values of variables
reject_agg and agg_size, which have not been initialized.
Always initialize these variables in order to prevent using
uninitialized values.
This issue has been found with clang. The compiler reported:
drivers/net/wireless/realtek/rtlwifi/base.c:1665:6: error: variable
'agg_size' is used uninitialized whenever 'if' condition is false
[-Werror,-Wsometimes-uninitialized]
if (rtlpriv->cfg->ops->get_btc_status())
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/net/wireless/realtek/rtlwifi/base.c:1671:31: note:
uninitialized use occurs here
reject_agg, ctrl_agg_size, agg_size);
^~~~~~~~
drivers/net/wireless/realtek/rtlwifi/base.c:1665:6: error: variable
'reject_agg' is used uninitialized whenever 'if' condition
is false [-Werror,-Wsometimes-uninitialized]
if (rtlpriv->cfg->ops->get_btc_status())
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/net/wireless/realtek/rtlwifi/base.c:1671:4: note:
uninitialized use occurs here
reject_agg, ctrl_agg_size, agg_size);
^~~~~~~~~~
Fixes: 2635664e6e4a ("rtlwifi: Add rx ampdu cfg for btcoexist.")
Signed-off-by: Nicolas Iooss <nicolas.iooss_linux@m4x.org>
---
drivers/net/wireless/realtek/rtlwifi/base.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/realtek/rtlwifi/base.c b/drivers/net/wireless/realtek/rtlwifi/base.c
index cad2272ae21b..704741d6f495 100644
--- a/drivers/net/wireless/realtek/rtlwifi/base.c
+++ b/drivers/net/wireless/realtek/rtlwifi/base.c
@@ -1726,7 +1726,7 @@ int rtl_tx_agg_oper(struct ieee80211_hw *hw,
void rtl_rx_ampdu_apply(struct rtl_priv *rtlpriv)
{
struct rtl_btc_ops *btc_ops = rtlpriv->btcoexist.btc_ops;
- u8 reject_agg, ctrl_agg_size = 0, agg_size;
+ u8 reject_agg = 0, ctrl_agg_size = 0, agg_size = 0;
if (rtlpriv->cfg->ops->get_btc_status())
btc_ops->btc_get_ampdu_cfg(rtlpriv, &reject_agg,
--
2.15.0
^ permalink raw reply related
* Re: [PATCH] mac80211: fix locking in ieee80211_sta_tear_down_BA_sessions
From: Pavel Machek @ 2017-12-10 8:36 UTC (permalink / raw)
To: Johannes Berg; +Cc: linux-wireless, Kalle Valo, Johannes Berg
In-Reply-To: <20171209201338.24931-1-johannes@sipsolutions.net>
[-- Attachment #1: Type: text/plain, Size: 1244 bytes --]
On Sat 2017-12-09 21:13:38, Johannes Berg wrote:
> From: Johannes Berg <johannes.berg@intel.com>
>
> Due to overlap between
> commit 1281103770e9 ("mac80211: Simplify locking in ieee80211_sta_tear_down_BA_sessions()")
> and the way that Luca modified
> commit 72e2c3438ba3 ("mac80211: tear down RX aggregations first")
> when sending it upstream from Intel's internal tree, we get
> the following warning:
>
> WARNING: CPU: 0 PID: 5472 at net/mac80211/agg-tx.c:315 ___ieee80211_stop_tx_ba_session+0x158/0x1f0
>
> since there's no appropriate locking around the call to
> ___ieee80211_stop_tx_ba_session; Sara's original just had
> a call to the locked __ieee80211_stop_tx_ba_session (one
> less underscore) but it looks like Luca modified both of
> the calls when fixing it up for upstream, leading to the
> problem at hand.
>
> Move the locking appropriately to fix this problem.
>
> Reported-by: Kalle Valo <kvalo@codeaurora.org>
> Reported-by: Pavel Machek <pavel@ucw.cz>
> Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Tested-by: Pavel Machek <pavel@ucw.cz>
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]
^ permalink raw reply
* Re: Wireless regressions in v4.15-rc1
From: Johannes Berg @ 2017-12-09 20:15 UTC (permalink / raw)
To: Kalle Valo, linux-wireless, Thorsten Leemhuis
Cc: netdev, linux-kernel, ath10k
In-Reply-To: <87po7xwajg.fsf@kamboji.qca.qualcomm.com>
On Sat, 2017-12-02 at 14:59 +0200, Kalle Valo wrote:
>
> net: netlink: Update attr validation to require exact length for some types
> https://git.kernel.org/linus/28033ae4e0f5
This was reverted, more or less, to print only a warning:
https://git.kernel.org/pub/scm/linux/kernel/git/davem/net.git/commit/?id=6e237d099fac1f73a7b6d7287bb9191f29585a4e
> Jouni fixed this already in hostapd but we also need a fix for kernel so
> that old hostapd versions continue to work:
>
> https://w1.fi/cgit/hostap/commit/?id=a2426829ce426de82d2fa47071ca41ea81c43307
>
> Jouni also found a similar problem with mesh:
>
> https://w1.fi/cgit/hostap/commit/?id=963d3149abfcbab5b83f9023bc50321f777360d1
See above for the kernel revert to let old hostapd versions continue to
work (properly only on little endian platforms) when they don't have
these fixes.
> And Johannes already submitted a revert related to wpa_supplicant:
>
> [net] Revert "net: core: maybe return -EEXIST in __dev_alloc_name" diffmboxseries
> https://patchwork.ozlabs.org/patch/843863/
That was applied.
> And with ath10k I'm now seeing this:
>
> [ 133.175508] WARNING: CPU: 2 PID: 1743 at net/mac80211/agg-tx.c:315 ___ieee80211_stop_tx_ba_session+0x1ab/0x280 [mac80211]
Just sent a fix for this, kinda a merge/patch failure.
> johannes
^ permalink raw reply
* [PATCH] mac80211: fix locking in ieee80211_sta_tear_down_BA_sessions
From: Johannes Berg @ 2017-12-09 20:13 UTC (permalink / raw)
To: linux-wireless; +Cc: Kalle Valo, Pavel Machek, Johannes Berg
From: Johannes Berg <johannes.berg@intel.com>
Due to overlap between
commit 1281103770e9 ("mac80211: Simplify locking in ieee80211_sta_tear_down_BA_sessions()")
and the way that Luca modified
commit 72e2c3438ba3 ("mac80211: tear down RX aggregations first")
when sending it upstream from Intel's internal tree, we get
the following warning:
WARNING: CPU: 0 PID: 5472 at net/mac80211/agg-tx.c:315 ___ieee80211_stop_tx_ba_session+0x158/0x1f0
since there's no appropriate locking around the call to
___ieee80211_stop_tx_ba_session; Sara's original just had
a call to the locked __ieee80211_stop_tx_ba_session (one
less underscore) but it looks like Luca modified both of
the calls when fixing it up for upstream, leading to the
problem at hand.
Move the locking appropriately to fix this problem.
Reported-by: Kalle Valo <kvalo@codeaurora.org>
Reported-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
---
net/mac80211/ht.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c
index 167f83b853e6..1621b6ab17ba 100644
--- a/net/mac80211/ht.c
+++ b/net/mac80211/ht.c
@@ -291,16 +291,15 @@ void ieee80211_sta_tear_down_BA_sessions(struct sta_info *sta,
int i;
mutex_lock(&sta->ampdu_mlme.mtx);
- for (i = 0; i < IEEE80211_NUM_TIDS; i++) {
+ for (i = 0; i < IEEE80211_NUM_TIDS; i++)
___ieee80211_stop_rx_ba_session(sta, i, WLAN_BACK_RECIPIENT,
WLAN_REASON_QSTA_LEAVE_QBSS,
reason != AGG_STOP_DESTROY_STA &&
reason != AGG_STOP_PEER_REQUEST);
- }
- mutex_unlock(&sta->ampdu_mlme.mtx);
for (i = 0; i < IEEE80211_NUM_TIDS; i++)
___ieee80211_stop_tx_ba_session(sta, i, reason);
+ mutex_unlock(&sta->ampdu_mlme.mtx);
/* stopping might queue the work again - so cancel only afterwards */
cancel_work_sync(&sta->ampdu_mlme.work);
--
2.14.2
^ permalink raw reply related
* Re: [PATCH 1/2] mac80211: Populate RSC counter in set_key method
From: Johannes Berg @ 2017-12-09 19:58 UTC (permalink / raw)
To: Govind Singh, ath10k; +Cc: linux-wireless
In-Reply-To: <1512739922-1529-1-git-send-email-govinds@qti.qualcomm.com>
On Fri, 2017-12-08 at 19:02 +0530, Govind Singh wrote:
> Send RSC counter to driver in set_key method, so that FW/driver
> can drop the packet in PN check if received packet sequence
> no is less than current RSC counter during group keys(GTK)
> exchange.
Good idea, but ... does it really work this way? :-)
> /**
> * struct ieee80211_key_conf - key information
> *
> @@ -1586,6 +1588,7 @@ enum ieee80211_key_flags {
> * - Temporal Authenticator Rx MIC Key (64 bits)
> * @icv_len: The ICV length for this key type
> * @iv_len: The IV length for this key type
> + * @rx_pn: Last received packet number, must be in little endian.
> */
> struct ieee80211_key_conf {
> atomic64_t tx_pn;
> @@ -1596,11 +1599,10 @@ struct ieee80211_key_conf {
> u8 flags;
> s8 keyidx;
> u8 keylen;
> + u8 rx_pn[IEEE80211_MAX_PN_LEN];
> u8 key[0];
> };
I don't think you should put that here, we keep this around in the key
structure forever. We can pass it out of band, I'd say?
> diff --git a/net/mac80211/key.c b/net/mac80211/key.c
> index 9380493..15e1822 100644
> --- a/net/mac80211/key.c
> +++ b/net/mac80211/key.c
> @@ -538,6 +538,12 @@ struct ieee80211_key *
> }
> memcpy(key->conf.key, key_data, key_len);
> INIT_LIST_HEAD(&key->list);
> + /* Assign receive packet sequence no, rx_pn remains in
> + * little endian format as seq is guaranteed to be in little
> + * endian format.
> + */
> + if (seq)
> + memcpy(&key->conf.rx_pn, seq, seq_len);
It seems that perhaps we should make this more robust - there are
sometimes cases where the key can enable HW acceleration only after
some time. Not the case with ath10k though, but we should probably
handle that.
And once we do, we probably need this per TID, because the RX PN might
have moved around, or we look up the max over all TIDs. All the more
reason to pass it out of band.
Another reason to pass it out of band would be to tell the driver that
indeed it's *not* maintained over the lifetime of the key, unlike the
other fields which are maintained (because constant) over the lifetime
of the key.
johannes
^ permalink raw reply
* Re: Setting TX power on a monitoring interface
From: Johannes Berg @ 2017-12-09 19:53 UTC (permalink / raw)
To: Peter Große, linux-wireless
In-Reply-To: <20171201183446.2c5f8504@fem-net.de>
Hi,
(side note - I might reply faster if you don't drop me from Cc, since
then it goes to my inbox in addition to my list folder)
> > I meant to pass the &monitor_sdata.vif pointer instead of the real
> > monitor interface that's coming through cfg80211. The former is virtual
> > and has no netdev, but the diver is aware of it.
>
> I'm not sure I get where to pass this to what. Do you mean in
> drv_bss_info_changed or ieee80211_set_tx_power?
Should be in ieee80211_set_tx_power, if the interface is you're getting
in (dev/sdata) is MONITOR. You need to do the right locking (or may
already have rtnl, check in nl80211.c) to get local->monitor_sdata, and
then reject the call if that is NULL or pass it through with that sdata
(or the vif from it)
> Assuming you meant ieee80211_set_tx_power, then I'd have to check whether wdev
> is a monitor interface and reject the configuration, if local->monitor_sdata
> doesn't exist?
Right.
> But in ieee80211_set_tx_power no vif pointers get handed around, so I'm
> confused. Sorry.
You have a wdev, that's equivalent. See the "if (wdev)" clause that
gets an sdata from the wdev, and then you have &sdata->vif as the vif.
johannes
^ permalink raw reply
* [PATCH 3/3] rtlwifi: rtl_pci: Fix the bug when inactiveps is enabled.
From: Larry Finger @ 2017-12-09 17:37 UTC (permalink / raw)
To: kvalo
Cc: linux-wireless, Tsang-Shian Lin, Ping-Ke Shih, Larry Finger,
Yan-Hsuan Chuang, Birming Chiu, Shaofu, Steven Ting
In-Reply-To: <20171209173710.9879-1-Larry.Finger@lwfinger.net>
From: Tsang-Shian Lin <thlin@realtek.com>
Reset the driver current tx read/write index to zero when inactiveps
nic out of sync with HW state. Wrong driver tx read/write index will
cause Tx fail.
Signed-off-by: Tsang-Shian Lin <thlin@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Yan-Hsuan Chuang <yhchuang@realtek.com>
Cc: Birming Chiu <birming@realtek.com>
Cc: Shaofu <shaofu@realtek.com>
Cc: Steven Ting <steventing@realtek.com>
---
drivers/net/wireless/realtek/rtlwifi/pci.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/net/wireless/realtek/rtlwifi/pci.c b/drivers/net/wireless/realtek/rtlwifi/pci.c
index 0bafcefacad0..1f2b04bb7ae4 100644
--- a/drivers/net/wireless/realtek/rtlwifi/pci.c
+++ b/drivers/net/wireless/realtek/rtlwifi/pci.c
@@ -1557,7 +1557,14 @@ int rtl_pci_reset_trx_ring(struct ieee80211_hw *hw)
dev_kfree_skb_irq(skb);
ring->idx = (ring->idx + 1) % ring->entries;
}
+
+ if (rtlpriv->use_new_trx_flow) {
+ rtlpci->tx_ring[i].cur_tx_rp = 0;
+ rtlpci->tx_ring[i].cur_tx_wp = 0;
+ }
+
ring->idx = 0;
+ ring->entries = rtlpci->txringcount[i];
}
}
spin_unlock_irqrestore(&rtlpriv->locks.irq_th_lock, flags);
--
2.15.1
^ permalink raw reply related
* [PATCH 2/3] rtlwifi: Add beacon check mechanism to check if AP settings changed.
From: Larry Finger @ 2017-12-09 17:37 UTC (permalink / raw)
To: kvalo
Cc: linux-wireless, Tsang-Shian Lin, Ping-Ke Shih, Larry Finger,
Yan-Hsuan Chuang, Birming Chiu, Shaofu, Steven Ting
In-Reply-To: <20171209173710.9879-1-Larry.Finger@lwfinger.net>
From: Tsang-Shian Lin <thlin@realtek.com>
AP WiFi settings are changed(channel, bandwidth), but deauth may not
received by STA. For these cases, we need to detect and handle beacon
changes.
Signed-off-by: Tsang-Shian Lin <thlin@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Yan-Hsuan Chuang <yhchuang@realtek.com>
Cc: Birming Chiu <birming@realtek.com>
Cc: Shaofu <shaofu@realtek.com>
Cc: Steven Ting <steventing@realtek.com>
---
drivers/net/wireless/realtek/rtlwifi/base.c | 179 ++++++++++++++++++++++++++++
drivers/net/wireless/realtek/rtlwifi/base.h | 3 +-
drivers/net/wireless/realtek/rtlwifi/core.c | 48 ++++++++
drivers/net/wireless/realtek/rtlwifi/pci.c | 2 +
drivers/net/wireless/realtek/rtlwifi/wifi.h | 13 ++
5 files changed, 244 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/realtek/rtlwifi/base.c b/drivers/net/wireless/realtek/rtlwifi/base.c
index cad2272ae21b..a04afdd55569 100644
--- a/drivers/net/wireless/realtek/rtlwifi/base.c
+++ b/drivers/net/wireless/realtek/rtlwifi/base.c
@@ -2360,6 +2360,185 @@ struct sk_buff *rtl_make_del_ba(struct ieee80211_hw *hw,
return skb;
}
+bool rtl_check_beacon_key(struct ieee80211_hw *hw, void *data, unsigned int len)
+{
+ struct rtl_priv *rtlpriv = rtl_priv(hw);
+ struct rtl_mac *mac = rtl_mac(rtl_priv(hw));
+ struct rtl_phy *rtlphy = &rtlpriv->phy;
+ struct ieee80211_hdr *hdr = data;
+ struct ieee80211_ht_cap *ht_cap_ie;
+ struct ieee80211_ht_operation *ht_oper_ie = NULL;
+ struct rtl_beacon_keys bcn_key = {0};
+ struct rtl_beacon_keys *cur_bcn_key;
+ u8 *ht_cap;
+ u8 ht_cap_len;
+ u8 *ht_oper;
+ u8 ht_oper_len;
+ u8 *ds_param;
+ u8 ds_param_len;
+
+ if (mac->opmode != NL80211_IFTYPE_STATION)
+ return false;
+
+ /* check if this really is a beacon*/
+ if (!ieee80211_is_beacon(hdr->frame_control))
+ return false;
+
+ /* min. beacon length + FCS_LEN */
+ if (len <= 40 + FCS_LEN)
+ return false;
+
+ cur_bcn_key = &mac->cur_beacon_keys;
+
+ if (rtlpriv->mac80211.link_state == MAC80211_NOLINK) {
+ if (cur_bcn_key->valid) {
+ cur_bcn_key->valid = false;
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_LOUD,
+ "Reset cur_beacon_keys.valid to false!\n");
+ }
+ return false;
+ }
+
+ /* and only beacons from the associated BSSID, please */
+ if (!ether_addr_equal(hdr->addr3, rtlpriv->mac80211.bssid))
+ return false;
+
+ /***** Parsing DS Param IE ******/
+ ds_param = rtl_find_ie(data, len - FCS_LEN, WLAN_EID_DS_PARAMS);
+
+ if (ds_param && !(ds_param[1] < sizeof(*ds_param)))
+ ds_param_len = ds_param[1];
+ else
+ ds_param = NULL;
+
+ /***** Parsing HT Cap. IE ******/
+ ht_cap = rtl_find_ie(data, len - FCS_LEN, WLAN_EID_HT_CAPABILITY);
+
+ if (ht_cap && !(ht_cap[1] < sizeof(*ht_cap))) {
+ ht_cap_len = ht_cap[1];
+ ht_cap_ie = (struct ieee80211_ht_cap *)&ht_cap[2];
+ } else {
+ ht_cap = NULL;
+ ht_cap_ie = NULL;
+ }
+
+ /***** Parsing HT Info. IE ******/
+ ht_oper = rtl_find_ie(data, len - FCS_LEN, WLAN_EID_HT_OPERATION);
+
+ if (ht_oper && !(ht_oper[1] < sizeof(*ht_oper))) {
+ ht_oper_len = ht_oper[1];
+ ht_oper_ie = (struct ieee80211_ht_operation *)&ht_oper[2];
+ } else {
+ ht_oper = NULL;
+ }
+
+ /* update bcn_key */
+ memset(&bcn_key, 0, sizeof(bcn_key));
+
+ if (ds_param)
+ bcn_key.bcn_channel = ds_param[2];
+ else if (ht_oper && ht_oper_ie)
+ bcn_key.bcn_channel = ht_oper_ie->primary_chan;
+
+ if (ht_cap_ie)
+ bcn_key.ht_cap_info = ht_cap_ie->cap_info;
+
+ if (ht_oper && ht_oper_ie)
+ bcn_key.ht_info_infos_0_sco = ht_oper_ie->ht_param & 0x03;
+
+ bcn_key.valid = true;
+
+ /* update cur_beacon_keys or compare beacon key */
+ if (rtlpriv->mac80211.link_state != MAC80211_LINKED &&
+ rtlpriv->mac80211.link_state != MAC80211_LINKED_SCANNING)
+ return true;
+
+ if (!cur_bcn_key->valid) {
+ /* update cur_beacon_keys */
+ memset(cur_bcn_key, 0, sizeof(bcn_key));
+ memcpy(cur_bcn_key, &bcn_key, sizeof(bcn_key));
+ cur_bcn_key->valid = true;
+
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_LOUD,
+ "Beacon key update!ch=%d, ht_cap_info=0x%x, sco=0x%x\n",
+ cur_bcn_key->bcn_channel,
+ cur_bcn_key->ht_cap_info,
+ cur_bcn_key->ht_info_infos_0_sco);
+
+ return true;
+ }
+
+ /* compare beacon key */
+ if (!memcmp(cur_bcn_key, &bcn_key, sizeof(bcn_key))) {
+ /* same beacon key */
+ mac->new_beacon_cnt = 0;
+ goto chk_exit;
+ }
+
+ if (cur_bcn_key->bcn_channel == bcn_key.bcn_channel &&
+ cur_bcn_key->ht_cap_info == bcn_key.ht_cap_info) {
+ /* Beacon HT info IE, secondary channel offset check */
+ /* 40M -> 20M */
+ if (cur_bcn_key->ht_info_infos_0_sco >
+ bcn_key.ht_info_infos_0_sco) {
+ /* Not a new beacon */
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_DMESG,
+ "Beacon BW change! sco:0x%x -> 0x%x\n",
+ cur_bcn_key->ht_info_infos_0_sco,
+ bcn_key.ht_info_infos_0_sco);
+
+ cur_bcn_key->ht_info_infos_0_sco =
+ bcn_key.ht_info_infos_0_sco;
+ } else {
+ /* 20M -> 40M */
+ if (rtlphy->max_ht_chan_bw >= HT_CHANNEL_WIDTH_20_40) {
+ /* Not a new beacon */
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_DMESG,
+ "Beacon BW change! sco:0x%x -> 0x%x\n",
+ cur_bcn_key->ht_info_infos_0_sco,
+ bcn_key.ht_info_infos_0_sco);
+
+ cur_bcn_key->ht_info_infos_0_sco =
+ bcn_key.ht_info_infos_0_sco;
+ } else {
+ mac->new_beacon_cnt++;
+ }
+ }
+ } else {
+ mac->new_beacon_cnt++;
+ }
+
+ if (mac->new_beacon_cnt == 1) {
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_DMESG,
+ "Get new beacon.\n");
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_DMESG,
+ "Cur : ch=%d, ht_cap=0x%x, sco=0x%x\n",
+ cur_bcn_key->bcn_channel,
+ cur_bcn_key->ht_cap_info,
+ cur_bcn_key->ht_info_infos_0_sco);
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_DMESG,
+ "New RX : ch=%d, ht_cap=0x%x, sco=0x%x\n",
+ bcn_key.bcn_channel,
+ bcn_key.ht_cap_info,
+ bcn_key.ht_info_infos_0_sco);
+
+ } else if (mac->new_beacon_cnt > 1) {
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_DMESG,
+ "new beacon cnt: %d\n",
+ mac->new_beacon_cnt);
+ }
+
+ if (mac->new_beacon_cnt > 3) {
+ ieee80211_connection_loss(rtlpriv->mac80211.vif);
+ RT_TRACE(rtlpriv, COMP_BEACON, DBG_DMESG,
+ "new beacon cnt >3, disconnect !\n");
+ }
+
+chk_exit:
+
+ return true;
+}
+EXPORT_SYMBOL_GPL(rtl_check_beacon_key);
/*********************************************************
*
* IOT functions
diff --git a/drivers/net/wireless/realtek/rtlwifi/base.h b/drivers/net/wireless/realtek/rtlwifi/base.h
index 26735319b38f..3ad8e8107209 100644
--- a/drivers/net/wireless/realtek/rtlwifi/base.h
+++ b/drivers/net/wireless/realtek/rtlwifi/base.h
@@ -176,5 +176,6 @@ u8 rtl_tid_to_ac(u8 tid);
void rtl_easy_concurrent_retrytimer_callback(struct timer_list *t);
extern struct rtl_global_var rtl_global_var;
void rtl_phy_scan_operation_backup(struct ieee80211_hw *hw, u8 operation);
-
+bool rtl_check_beacon_key(struct ieee80211_hw *hw, void *data,
+ unsigned int len);
#endif
diff --git a/drivers/net/wireless/realtek/rtlwifi/core.c b/drivers/net/wireless/realtek/rtlwifi/core.c
index 3cb88825473e..195f7c41b4aa 100644
--- a/drivers/net/wireless/realtek/rtlwifi/core.c
+++ b/drivers/net/wireless/realtek/rtlwifi/core.c
@@ -909,6 +909,7 @@ static int rtl_op_sta_add(struct ieee80211_hw *hw,
struct ieee80211_sta *sta)
{
struct rtl_priv *rtlpriv = rtl_priv(hw);
+ struct rtl_phy *rtlphy = &rtlpriv->phy;
struct rtl_hal *rtlhal = rtl_hal(rtl_priv(hw));
struct rtl_mac *mac = rtl_mac(rtl_priv(hw));
struct rtl_sta_info *sta_entry;
@@ -941,6 +942,16 @@ static int rtl_op_sta_add(struct ieee80211_hw *hw,
if (mac->p2p)
sta->supp_rates[0] &= 0xfffffff0;
+ if (sta->ht_cap.ht_supported) {
+ if (sta->ht_cap.cap & IEEE80211_HT_CAP_SUP_WIDTH_20_40)
+ rtlphy->max_ht_chan_bw = HT_CHANNEL_WIDTH_20_40;
+ else
+ rtlphy->max_ht_chan_bw = HT_CHANNEL_WIDTH_20;
+ }
+
+ if (sta->vht_cap.vht_supported)
+ rtlphy->max_vht_chan_bw = HT_CHANNEL_WIDTH_80;
+
memcpy(sta_entry->mac_addr, sta->addr, ETH_ALEN);
RT_TRACE(rtlpriv, COMP_MAC80211, DBG_DMESG,
"Add sta addr is %pM\n", sta->addr);
@@ -1039,6 +1050,7 @@ static void rtl_op_bss_info_changed(struct ieee80211_hw *hw,
u32 changed)
{
struct rtl_priv *rtlpriv = rtl_priv(hw);
+ struct rtl_phy *rtlphy = &rtlpriv->phy;
struct rtl_hal *rtlhal = rtl_hal(rtlpriv);
struct rtl_mac *mac = rtl_mac(rtl_priv(hw));
struct rtl_ps_ctl *ppsc = rtl_psc(rtl_priv(hw));
@@ -1241,6 +1253,16 @@ static void rtl_op_bss_info_changed(struct ieee80211_hw *hw,
mac->current_ampdu_factor)
mac->current_ampdu_factor =
sta->ht_cap.ampdu_factor;
+
+ if (sta->ht_cap.ht_supported) {
+ if (sta->ht_cap.cap &
+ IEEE80211_HT_CAP_SUP_WIDTH_20_40)
+ rtlphy->max_ht_chan_bw =
+ HT_CHANNEL_WIDTH_20_40;
+ else
+ rtlphy->max_ht_chan_bw =
+ HT_CHANNEL_WIDTH_20;
+ }
}
rcu_read_unlock();
@@ -1252,6 +1274,32 @@ static void rtl_op_bss_info_changed(struct ieee80211_hw *hw,
&mac->current_ampdu_density);
}
+ if (changed & BSS_CHANGED_BANDWIDTH) {
+ struct ieee80211_sta *sta = NULL;
+
+ RT_TRACE(rtlpriv, COMP_MAC80211, DBG_TRACE,
+ "BSS_CHANGED_BANDWIDTH\n");
+
+ rcu_read_lock();
+ sta = ieee80211_find_sta(vif, (u8 *)bss_conf->bssid);
+
+ if (sta) {
+ if (sta->ht_cap.ht_supported) {
+ if (sta->ht_cap.cap &
+ IEEE80211_HT_CAP_SUP_WIDTH_20_40)
+ rtlphy->max_ht_chan_bw =
+ HT_CHANNEL_WIDTH_20_40;
+ else
+ rtlphy->max_ht_chan_bw =
+ HT_CHANNEL_WIDTH_20;
+ }
+
+ if (sta->vht_cap.vht_supported)
+ rtlphy->max_vht_chan_bw = HT_CHANNEL_WIDTH_80;
+ }
+ rcu_read_unlock();
+ }
+
if (changed & BSS_CHANGED_BSSID) {
u32 basic_rates;
struct ieee80211_sta *sta = NULL;
diff --git a/drivers/net/wireless/realtek/rtlwifi/pci.c b/drivers/net/wireless/realtek/rtlwifi/pci.c
index 8ae36a263426..0bafcefacad0 100644
--- a/drivers/net/wireless/realtek/rtlwifi/pci.c
+++ b/drivers/net/wireless/realtek/rtlwifi/pci.c
@@ -877,6 +877,8 @@ static void _rtl_pci_rx_interrupt(struct ieee80211_hw *hw)
ieee80211_is_probe_resp(fc))) {
dev_kfree_skb_any(skb);
} else {
+ rtl_check_beacon_key(hw, (void *)skb->data,
+ skb->len);
_rtl_pci_rx_to_mac80211(hw, skb, rx_status);
}
} else {
diff --git a/drivers/net/wireless/realtek/rtlwifi/wifi.h b/drivers/net/wireless/realtek/rtlwifi/wifi.h
index e2b14793b705..d3ea6260f01c 100644
--- a/drivers/net/wireless/realtek/rtlwifi/wifi.h
+++ b/drivers/net/wireless/realtek/rtlwifi/wifi.h
@@ -962,6 +962,15 @@ struct rtl_probe_rsp {
struct rtl_info_element info_element[0];
} __packed;
+struct rtl_beacon_keys {
+ /*u8 ssid[32];*/
+ /*u32 ssid_len;*/
+ u8 bcn_channel;
+ __le16 ht_cap_info;
+ u8 ht_info_infos_0_sco; /* bit0 & bit1 in infos[0] is 2nd ch offset */
+ bool valid;
+};
+
/*LED related.*/
/*ledpin Identify how to implement this SW led.*/
struct rtl_led {
@@ -1209,6 +1218,8 @@ struct rtl_phy {
u8 rf_mode;
u8 rf_type;
u8 current_chan_bw;
+ u8 max_ht_chan_bw;
+ u8 max_vht_chan_bw;
u8 set_bwmode_inprogress;
u8 sw_chnl_inprogress;
u8 sw_chnl_stage;
@@ -1380,6 +1391,8 @@ struct rtl_mac {
/*Probe Beacon management */
struct rtl_tid_data tids[MAX_TID_COUNT];
enum rtl_link_state link_state;
+ struct rtl_beacon_keys cur_beacon_keys;
+ u8 new_beacon_cnt;
int n_channels;
int n_bitrates;
--
2.15.1
^ permalink raw reply related
* [PATCH 1/3] rtlwifi: rtl_pci: 8822BE puts broadcast and multicast packet to HIQ
From: Larry Finger @ 2017-12-09 17:37 UTC (permalink / raw)
To: kvalo
Cc: linux-wireless, Ping-Ke Shih, Larry Finger, Yan-Hsuan Chuang,
Birming Chiu, Shaofu, Steven Ting
In-Reply-To: <20171209173710.9879-1-Larry.Finger@lwfinger.net>
From: Ping-Ke Shih <pkshih@realtek.com>
Making this change to HIQ, which has high priority, improves the response
time for transmission after TBTT or beacon.
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Yan-Hsuan Chuang <yhchuang@realtek.com>
Cc: Birming Chiu <birming@realtek.com>
Cc: Shaofu <shaofu@realtek.com>
Cc: Steven Ting <steventing@realtek.com>
---
drivers/net/wireless/realtek/rtlwifi/pci.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/drivers/net/wireless/realtek/rtlwifi/pci.c b/drivers/net/wireless/realtek/rtlwifi/pci.c
index 3e9eb11a38f9..8ae36a263426 100644
--- a/drivers/net/wireless/realtek/rtlwifi/pci.c
+++ b/drivers/net/wireless/realtek/rtlwifi/pci.c
@@ -59,6 +59,7 @@ static u8 _rtl_mac_to_hwqueue(struct ieee80211_hw *hw, struct sk_buff *skb)
struct rtl_hal *rtlhal = rtl_hal(rtl_priv(hw));
__le16 fc = rtl_get_fc(skb);
u8 queue_index = skb_get_queue_mapping(skb);
+ struct ieee80211_hdr *hdr;
if (unlikely(ieee80211_is_beacon(fc)))
return BEACON_QUEUE;
@@ -67,6 +68,13 @@ static u8 _rtl_mac_to_hwqueue(struct ieee80211_hw *hw, struct sk_buff *skb)
if (rtlhal->hw_type == HARDWARE_TYPE_RTL8192SE)
if (ieee80211_is_nullfunc(fc))
return HIGH_QUEUE;
+ if (rtlhal->hw_type == HARDWARE_TYPE_RTL8822BE) {
+ hdr = rtl_get_hdr(skb);
+
+ if (is_multicast_ether_addr(hdr->addr1) ||
+ is_broadcast_ether_addr(hdr->addr1))
+ return HIGH_QUEUE;
+ }
return ac_to_hwq[queue_index];
}
--
2.15.1
^ permalink raw reply related
* [PATCH 0/3] rtlwifi: rtl_pci: More updates to mini driver for RTL8822BE
From: Larry Finger @ 2017-12-09 17:37 UTC (permalink / raw)
To: kvalo
Cc: linux-wireless, Larry Finger, Ping-Ke Shih, Yan-Hsuan Chuang,
Birming Chiu, Shaofu, Steven Ting
These changes are for the 4.16 stream.
The modifications implement nearly all the changes needed to allow the
RTL8822BE in staging to use rtl_pci in the wireless tree. The only
exceptions are related to BT coexistence, which will be added once btcoex
is updated.
Once these patches reach mainline, I will submit a patch to staging that
changes r8822be to use the standard rtl_pci, and deletes the duplicated
code.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Ping-Ke Shih <pkshih@realtek.com>
Cc: Yan-Hsuan Chuang <yhchuang@realtek.com>
Cc: Birming Chiu <birming@realtek.com>
Cc: Shaofu <shaofu@realtek.com>
Cc: Steven Ting <steventing@realtek.com>
Ping-Ke Shih (1):
rtlwifi: rtl_pci: 8822BE puts broadcast and multicast packet to HIQ
Tsang-Shian Lin (2):
rtlwifi: Add beacon check mechanism to check if AP settings changed.
rtlwifi: rtl_pci: Fix the bug when inactiveps is enabled.
drivers/net/wireless/realtek/rtlwifi/base.c | 179 ++++++++++++++++++++++++++++
drivers/net/wireless/realtek/rtlwifi/base.h | 3 +-
drivers/net/wireless/realtek/rtlwifi/core.c | 48 ++++++++
drivers/net/wireless/realtek/rtlwifi/pci.c | 17 +++
drivers/net/wireless/realtek/rtlwifi/wifi.h | 13 ++
5 files changed, 259 insertions(+), 1 deletion(-)
--
2.15.1
^ permalink raw reply
* Re: [PATCH] wcn36xx: Fix dynamic power saving
From: Bjorn Andersson @ 2017-12-09 0:45 UTC (permalink / raw)
To: Loic Poulain
Cc: kvalo, wcn36xx, linux-wireless, linux-arm-msm, nicolas.dechesne,
k.eugene.e
In-Reply-To: <1512750846-13591-1-git-send-email-loic.poulain@linaro.org>
On Fri 08 Dec 08:34 PST 2017, Loic Poulain wrote:
> Since driver does not report hardware dynamic power saving cap,
> this is up to the mac80211 to manage power saving timeout and
> state machine, using the ieee80211 config callback to report
> PS changes. This patch enables/disables PS mode according to
> the new configuration.
>
> Remove old behaviour enabling PS mode in a static way, this make
> the device unusable when power save is enabled since device is
> forced to PS regardless RX/TX traffic.
>
> Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
With below nit.
> ---
> drivers/net/wireless/ath/wcn36xx/main.c | 23 ++++++++++++-----------
> 1 file changed, 12 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/wcn36xx/main.c b/drivers/net/wireless/ath/wcn36xx/main.c
> index f0b4d43..436b8ea 100644
> --- a/drivers/net/wireless/ath/wcn36xx/main.c
> +++ b/drivers/net/wireless/ath/wcn36xx/main.c
> @@ -384,6 +384,18 @@ static int wcn36xx_config(struct ieee80211_hw *hw, u32 changed)
> }
> }
>
> + if (changed & IEEE80211_CONF_CHANGE_PS) {
> + list_for_each_entry(tmp, &wcn->vif_list, list) {
> + vif = wcn36xx_priv_to_vif(tmp);
> + if (hw->conf.flags & IEEE80211_CONF_PS) {
> + if (vif->bss_conf.ps) /* ps allowed ? */
> + wcn36xx_pmc_enter_bmps_state(wcn, vif);
> + } else {
> + wcn36xx_pmc_exit_bmps_state(wcn, vif);
During startup I get the error print from wcn36xx_pmc_exit_bmps_state()
that we're not in BMPS state. There's no harm in this, but the error
might concern people.
How about we in addition to this, change the wcn36xx_err() to a
wcn36xx_dbg(PMC...) ?
Regards,
Bjorn
^ permalink raw reply
* [PATCH] wcn36xx: Reduce spinlock in indication handler
From: Bjorn Andersson @ 2017-12-09 0:35 UTC (permalink / raw)
To: Eugene Krasnikov, Kalle Valo, Loic Poulain
Cc: wcn36xx, linux-wireless, netdev, linux-kernel
In-Reply-To: <1512725730-19188-1-git-send-email-loic.poulain@linaro.org>
The purpose of pushing indication on a list and handle these in a
separate worker is to allow the handlers to sleep. It does therefor not
make much sense to hold the queue spinlock through the entire indication
worker function.
By removing items from the queue early we don't need to hold the lock
throughout the indication worker, allowing the individual handlers to
sleep.
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
---
drivers/net/wireless/ath/wcn36xx/smd.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ath/wcn36xx/smd.c b/drivers/net/wireless/ath/wcn36xx/smd.c
index fa88a2a460aa..52daae863aed 100644
--- a/drivers/net/wireless/ath/wcn36xx/smd.c
+++ b/drivers/net/wireless/ath/wcn36xx/smd.c
@@ -2414,6 +2414,8 @@ static void wcn36xx_ind_smd_work(struct work_struct *work)
hal_ind_msg = list_first_entry(&wcn->hal_ind_queue,
struct wcn36xx_hal_ind_msg,
list);
+ list_del(wcn->hal_ind_queue.next);
+ spin_unlock_irqrestore(&wcn->hal_ind_lock, flags);
msg_header = (struct wcn36xx_hal_msg_header *)hal_ind_msg->msg;
@@ -2450,8 +2452,6 @@ static void wcn36xx_ind_smd_work(struct work_struct *work)
wcn36xx_err("SMD_EVENT (%d) not supported\n",
msg_header->msg_type);
}
- list_del(wcn->hal_ind_queue.next);
- spin_unlock_irqrestore(&wcn->hal_ind_lock, flags);
kfree(hal_ind_msg);
}
int wcn36xx_smd_open(struct wcn36xx *wcn)
--
2.15.0
^ permalink raw reply related
* Re: [PATCH] wcn36xx: Add hardware scan offload support
From: Bjorn Andersson @ 2017-12-09 0:34 UTC (permalink / raw)
To: Loic Poulain; +Cc: kvalo, linux-arm-msm, linux-wireless, wcn36xx, k.eugene.e
In-Reply-To: <1512725730-19188-1-git-send-email-loic.poulain@linaro.org>
On Fri 08 Dec 01:35 PST 2017, Loic Poulain wrote:
> Current hw_scan implementation does not trigger offloaded
> hardware scan and seems to only put the device in a kind of
> listening mode (beacon/probe-response) for software scan.
> Since no probe request are generated by the software, current
> scanning method is similar to a passive scan.
>
> This patch introduces support for 'true' hardware offloaded scan.
> Hardware scan is configured and started via the start-scan-offload
> firmware message. Once scan has been completed a scan indicator
> message is received from firmware.
>
> Moreover, this patch includes support for directed probe-request,
> allowing connection with hidden APs. It also fixes scan issues with
> band-steering AP which are not 'visible' with passive scan (due to
> hidden ssid in beacons).
>
I'm puzzled to why I didn't see this when I dumped the control channel
that lead up to the initial hw_scan patch. But this looks good and with
a related locking issue fixed this seems to work well.
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
> diff --git a/drivers/net/wireless/ath/wcn36xx/smd.c b/drivers/net/wireless/ath/wcn36xx/smd.c
[..]
> +static int wcn36xx_smd_hw_scan_ind(struct wcn36xx *wcn, void *buf, size_t len)
> +{
> + struct wcn36xx_hal_scan_offload_ind *rsp = buf;
> + struct cfg80211_scan_info scan_info = {};
> +
> + if (len != sizeof(*rsp)) {
> + wcn36xx_warn("Corrupted delete scan indication\n");
> + return -EIO;
> + }
> +
> + wcn36xx_dbg(WCN36XX_DBG_HAL, "scan indication (type %x)", rsp->type);
> +
> + switch (rsp->type) {
> + case WCN36XX_HAL_SCAN_IND_FAILED:
> + scan_info.aborted = true;
> + case WCN36XX_HAL_SCAN_IND_COMPLETED:
> + mutex_lock(&wcn->scan_lock);
Grabbing this mutex with DEBUG_ATOMIC_SLEEP causes issues, but that's
because the locking in ind_smd_work() is to excessive. Will reply with a
fix for this.
> + wcn->scan_req = NULL;
> + mutex_unlock(&wcn->scan_lock);
> + ieee80211_scan_completed(wcn->hw, &scan_info);
> + break;
Regards,
Bjorn
^ permalink raw reply
* Re: pull-request: wireless-drivers 2017-12-08
From: David Miller @ 2017-12-08 19:50 UTC (permalink / raw)
To: kvalo; +Cc: linux-wireless, netdev, linux-kernel
In-Reply-To: <87k1xxqoi7.fsf@kamboji.qca.qualcomm.com>
From: Kalle Valo <kvalo@codeaurora.org>
Date: Fri, 08 Dec 2017 16:32:16 +0200
> this is a pull request to net tree for 4.15, more info in the signed tag
> below. All small fixes and not really expecting any problems, but please
> let me know if you have any.
Pulled, thanks Kalle.
^ permalink raw reply
* [PATCH] wcn36xx: Fix dynamic power saving
From: Loic Poulain @ 2017-12-08 16:34 UTC (permalink / raw)
To: kvalo
Cc: wcn36xx, linux-wireless, linux-arm-msm, nicolas.dechesne,
k.eugene.e, Loic Poulain
Since driver does not report hardware dynamic power saving cap,
this is up to the mac80211 to manage power saving timeout and
state machine, using the ieee80211 config callback to report
PS changes. This patch enables/disables PS mode according to
the new configuration.
Remove old behaviour enabling PS mode in a static way, this make
the device unusable when power save is enabled since device is
forced to PS regardless RX/TX traffic.
Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
---
drivers/net/wireless/ath/wcn36xx/main.c | 23 ++++++++++++-----------
1 file changed, 12 insertions(+), 11 deletions(-)
diff --git a/drivers/net/wireless/ath/wcn36xx/main.c b/drivers/net/wireless/ath/wcn36xx/main.c
index f0b4d43..436b8ea 100644
--- a/drivers/net/wireless/ath/wcn36xx/main.c
+++ b/drivers/net/wireless/ath/wcn36xx/main.c
@@ -384,6 +384,18 @@ static int wcn36xx_config(struct ieee80211_hw *hw, u32 changed)
}
}
+ if (changed & IEEE80211_CONF_CHANGE_PS) {
+ list_for_each_entry(tmp, &wcn->vif_list, list) {
+ vif = wcn36xx_priv_to_vif(tmp);
+ if (hw->conf.flags & IEEE80211_CONF_PS) {
+ if (vif->bss_conf.ps) /* ps allowed ? */
+ wcn36xx_pmc_enter_bmps_state(wcn, vif);
+ } else {
+ wcn36xx_pmc_exit_bmps_state(wcn, vif);
+ }
+ }
+ }
+
mutex_unlock(&wcn->conf_mutex);
return 0;
@@ -747,17 +759,6 @@ static void wcn36xx_bss_info_changed(struct ieee80211_hw *hw,
vif_priv->dtim_period = bss_conf->dtim_period;
}
- if (changed & BSS_CHANGED_PS) {
- wcn36xx_dbg(WCN36XX_DBG_MAC,
- "mac bss PS set %d\n",
- bss_conf->ps);
- if (bss_conf->ps) {
- wcn36xx_pmc_enter_bmps_state(wcn, vif);
- } else {
- wcn36xx_pmc_exit_bmps_state(wcn, vif);
- }
- }
-
if (changed & BSS_CHANGED_BSSID) {
wcn36xx_dbg(WCN36XX_DBG_MAC, "mac bss changed_bssid %pM\n",
bss_conf->bssid);
--
2.7.4
^ permalink raw reply related
* Re: [08/11] ath10k_sdio: common read write
From: Gary Bisson @ 2017-12-08 14:42 UTC (permalink / raw)
To: Alagu Sankar; +Cc: silexcommon, linux-wireless, ath10k
In-Reply-To: <e9eb78a9e0ed5e1a3acf48af88d02119@silex-india.com>
Hi Alagu,
On Thu, Oct 05, 2017 at 11:03:12PM +0530, Alagu Sankar wrote:
> Hi Gary,
>
>
> On 2017-10-05 15:39, Gary Bisson wrote:
> > Hi Alagu,
> >
> > On Sat, Sep 30, 2017 at 11:07:45PM +0530, silexcommon@gmail.com wrote:
> > > From: Alagu Sankar <alagusankar@silex-india.com>
> > >
> > > convert different read write functions in sdio hif to bring it under a
> > > single read-write path. This helps in having a common dma bounce
> > > buffer
> > > implementation. Also helps in address modification that is required
> > > specific to change in certain mbox addresses of sdio_write.
> > >
> > > Signed-off-by: Alagu Sankar <alagusankar@silex-india.com>
> > > ---
> > > drivers/net/wireless/ath/ath10k/sdio.c | 131
> > > ++++++++++++++++-----------------
> > > 1 file changed, 64 insertions(+), 67 deletions(-)
> > >
> > > diff --git a/drivers/net/wireless/ath/ath10k/sdio.c
> > > b/drivers/net/wireless/ath/ath10k/sdio.c
> > > index 77d4fa4..bb6fa67 100644
> > > --- a/drivers/net/wireless/ath/ath10k/sdio.c
> > > +++ b/drivers/net/wireless/ath/ath10k/sdio.c
> > > @@ -36,6 +36,11 @@
> > >
> > > #define ATH10K_SDIO_DMA_BUF_SIZE (32 * 1024)
> > >
> > > +static int ath10k_sdio_read(struct ath10k *ar, u32 addr, void *buf,
> > > + u32 len, bool incr);
> > > +static int ath10k_sdio_write(struct ath10k *ar, u32 addr, const
> > > void *buf,
> > > + u32 len, bool incr);
> > > +
> >
> > As mentioned by Kalle, u32 needs to be size_t.
> Yes, the compiler I used is probably a step older and did not catch this.
> >
> > > /* inlined helper functions */
> > >
> > > /* Macro to check if DMA buffer is WORD-aligned and DMA-able.
> > > @@ -152,6 +157,7 @@ static int ath10k_sdio_config(struct ath10k *ar)
> > > struct sdio_func *func = ar_sdio->func;
> > > unsigned char byte, asyncintdelay = 2;
> > > int ret;
> > > + u32 addr;
> > >
> > > ath10k_dbg(ar, ATH10K_DBG_BOOT, "sdio configuration\n");
> > >
> > > @@ -180,9 +186,8 @@ static int ath10k_sdio_config(struct ath10k *ar)
> > > CCCR_SDIO_DRIVER_STRENGTH_ENABLE_C |
> > > CCCR_SDIO_DRIVER_STRENGTH_ENABLE_D);
> > >
> > > - ret = ath10k_sdio_func0_cmd52_wr_byte(func->card,
> > > - CCCR_SDIO_DRIVER_STRENGTH_ENABLE_ADDR,
> > > - byte);
> > > + addr = CCCR_SDIO_DRIVER_STRENGTH_ENABLE_ADDR,
> > > + ret = ath10k_sdio_func0_cmd52_wr_byte(func->card, addr, byte);
> >
> > Not sure this part is needed.
> This is to overcome checkpatch warning. Too big a names for the function and
> macro
> not helping in there. Will have to move it as a separate patch.
> >
> > > if (ret) {
> > > ath10k_warn(ar, "failed to enable driver strength: %d\n", ret);
> > > goto out;
> > > @@ -233,13 +238,16 @@ static int ath10k_sdio_config(struct ath10k *ar)
> > >
> > > static int ath10k_sdio_write32(struct ath10k *ar, u32 addr, u32 val)
> > > {
> > > - struct ath10k_sdio *ar_sdio = ath10k_sdio_priv(ar);
> > > - struct sdio_func *func = ar_sdio->func;
> > > + __le32 *buf;
> > > int ret;
> > >
> > > - sdio_claim_host(func);
> > > + buf = kzalloc(sizeof(*buf), GFP_KERNEL);
> > > + if (!buf)
> > > + return -ENOMEM;
> > >
> > > - sdio_writel(func, val, addr, &ret);
> > > + *buf = cpu_to_le32(val);
> > > +
> > > + ret = ath10k_sdio_write(ar, addr, &val, sizeof(val), true);
> >
> > Shouldn't we use buf instead of val? buf seems pretty useless otherwise.
> Yes, thanks for pointing this out. will be corrected in v2.
Do you have any timeframe on the v2?
Regards,
Gary
^ permalink raw reply
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