Skip to contentSkip to conversation

wifi: mt76: mt7915: remove VHT160 capability on MT7915

PATCH In review Felix Fietkau 5 participants 11 messages Jul 26, 2023 – Oct 13, 2023

Subsystems (4)
lore
Share & tools
lore

Felix Fietkau

author maintainer
#0 lore raw

PATCH 1/3 wifi: mt76: mt7915: remove VHT160 capability on MT7915

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
 drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
index ee976657bfc3..78552f10b377 100644
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
 			if (!dev->dbdc_support)
 				vht_cap->cap |=
 					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
 					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
 		} else {
 			vht_cap->cap |=
Signature · 1 line
2.41.0
author maintainer
#1 lore raw

PATCH 2/3 wifi: mt76: mt7603: fix beacon interval after disabling a single vif

When disabling beacons on a vif, intval is 0. Ensure that dev->mt76.beacon_int is not overwritten in this case, so that beacons continue to work for other interfaces.

---
 drivers/net/wireless/mediatek/mt76/mt7603/beacon.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7603/beacon.c +2 -1
index b65b0a88c1de..888678732f29 100644
@@ -161,7 +161,8 @@ void mt7603_beacon_set_timer(struct mt7603_dev *dev, int idx, int intval)
 		return;
 	}
 
-	dev->mt76.beacon_int = intval;
+	if (intval)
+		dev->mt76.beacon_int = intval;
 	mt76_wr(dev, MT_TBTT,
 		FIELD_PREP(MT_TBTT_PERIOD, intval) | MT_TBTT_CAL_ENABLE);
 
Signature · 1 line
2.41.0
author maintainer
#2 lore raw

PATCH 3/3 wifi: mt76: mt7603: fix tx filter/flush function

In reply to Felix Fietkau

Setting MT_TX_ABORT does not abort any transmission for a wtbl index on its own. Instead, it modifies the behavior of a queue flush to make it selectively flush packets for a particular wtbl index.
Adjust powersave filtering to make use of this in order to avoid running into unnecessary timeouts while flushing

---
 .../net/wireless/mediatek/mt76/mt7603/mac.c   | 21 ++++++++++++++-----
 .../net/wireless/mediatek/mt76/mt7603/main.c  |  5 ++++-
 .../wireless/mediatek/mt76/mt7603/mt7603.h    |  2 +-
 .../net/wireless/mediatek/mt76/mt7603/regs.h  |  7 +++++++
 4 files changed, 28 insertions(+), 7 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7603/mac.c +16 -5
index de11557eb04c..99ae080502d8 100644
@@ -178,8 +178,9 @@ mt7603_wtbl_set_skip_tx(struct mt7603_dev *dev, int idx, bool enabled)
 	mt76_wr(dev, addr + 3 * 4, val);
 }
 
-void mt7603_filter_tx(struct mt7603_dev *dev, int idx, bool abort)
+void mt7603_filter_tx(struct mt7603_dev *dev, int mac_idx, int idx, bool abort)
 {
+	u32 flush_mask;
 	int i, port, queue;
 
 	if (abort) {
@@ -195,6 +196,18 @@ void mt7603_filter_tx(struct mt7603_dev *dev, int idx, bool abort)
 	mt76_wr(dev, MT_TX_ABORT, MT_TX_ABORT_EN |
 			FIELD_PREP(MT_TX_ABORT_WCID, idx));
 
+	flush_mask = MT_WF_ARB_TX_FLUSH_AC0 |
+		     MT_WF_ARB_TX_FLUSH_AC1 |
+		     MT_WF_ARB_TX_FLUSH_AC2 |
+		     MT_WF_ARB_TX_FLUSH_AC3;
+	flush_mask <<= mac_idx;
+
+	mt76_wr(dev, MT_WF_ARB_TX_FLUSH_0, flush_mask);
+	mt76_poll(dev, MT_WF_ARB_TX_FLUSH_0, flush_mask, 0, 20000);
+	mt76_wr(dev, MT_WF_ARB_TX_START_0, flush_mask);
+
+	mt76_wr(dev, MT_TX_ABORT, 0);
+
 	for (i = 0; i < 4; i++) {
 		mt76_wr(dev, MT_DMA_FQCR0, MT_DMA_FQCR0_BUSY |
 			FIELD_PREP(MT_DMA_FQCR0_TARGET_WCID, idx) |
@@ -202,13 +215,11 @@ void mt7603_filter_tx(struct mt7603_dev *dev, int idx, bool abort)
 			FIELD_PREP(MT_DMA_FQCR0_DEST_PORT_ID, port) |
 			FIELD_PREP(MT_DMA_FQCR0_DEST_QUEUE_ID, queue));
 
-		mt76_poll(dev, MT_DMA_FQCR0, MT_DMA_FQCR0_BUSY, 0, 15000);
+		mt76_poll(dev, MT_DMA_FQCR0, MT_DMA_FQCR0_BUSY, 0, 5000);
 	}
 
 	WARN_ON_ONCE(mt76_rr(dev, MT_DMA_FQCR0) & MT_DMA_FQCR0_BUSY);
 
-	mt76_wr(dev, MT_TX_ABORT, 0);
-
 	mt7603_wtbl_set_skip_tx(dev, idx, false);
 }
 
@@ -245,7 +256,7 @@ void mt7603_wtbl_set_ps(struct mt7603_dev *dev, struct mt7603_sta *sta,
 	mt76_poll(dev, MT_PSE_RTA, MT_PSE_RTA_BUSY, 0, 5000);
 
 	if (enabled)
-		mt7603_filter_tx(dev, idx, false);
+		mt7603_filter_tx(dev, sta->vif->idx, idx, false);
 
 	addr = mt7603_wtbl1_addr(idx);
 	mt76_set(dev, MT_WTBL1_OR, MT_WTBL1_OR_PSM_WRITE);
drivers/net/wireless/mediatek/mt76/mt7603/main.c +4 -1
index 1d4893410ca5..c213fd2a5216 100644
@@ -69,6 +69,7 @@ mt7603_add_interface(struct ieee80211_hw *hw, struct ieee80211_vif *vif)
 	INIT_LIST_HEAD(&mvif->sta.wcid.poll_list);
 	mvif->sta.wcid.idx = idx;
 	mvif->sta.wcid.hw_key_idx = -1;
+	mvif->sta.vif = mvif;
 	mt76_packet_id_init(&mvif->sta.wcid);
 
 	eth_broadcast_addr(bc_addr);
@@ -357,6 +358,7 @@ mt7603_sta_add(struct mt76_dev *mdev, struct ieee80211_vif *vif,
 	msta->smps = ~0;
 	msta->wcid.sta = 1;
 	msta->wcid.idx = idx;
+	msta->vif = mvif;
 	mt7603_wtbl_init(dev, idx, mvif->idx, sta->addr);
 	mt7603_wtbl_set_ps(dev, msta, false);
 
@@ -380,12 +382,13 @@ mt7603_sta_remove(struct mt76_dev *mdev, struct ieee80211_vif *vif,
 		  struct ieee80211_sta *sta)
 {
 	struct mt7603_dev *dev = container_of(mdev, struct mt7603_dev, mt76);
+	struct mt7603_vif *mvif = (struct mt7603_vif *)vif->drv_priv;
 	struct mt7603_sta *msta = (struct mt7603_sta *)sta->drv_priv;
 	struct mt76_wcid *wcid = (struct mt76_wcid *)sta->drv_priv;
 
 	spin_lock_bh(&dev->ps_lock);
 	__skb_queue_purge(&msta->psq);
-	mt7603_filter_tx(dev, wcid->idx, true);
+	mt7603_filter_tx(dev, mvif->idx, wcid->idx, true);
 	spin_unlock_bh(&dev->ps_lock);
 
 	spin_lock_bh(&mdev->sta_poll_lock);
drivers/net/wireless/mediatek/mt76/mt7603/mt7603.h +1 -1
index 354b189862f7..9e58df7042ad 100644
@@ -230,7 +230,7 @@ void mt7603_wtbl_set_ps(struct mt7603_dev *dev, struct mt7603_sta *sta,
 			bool enabled);
 void mt7603_wtbl_set_smps(struct mt7603_dev *dev, struct mt7603_sta *sta,
 			  bool enabled);
-void mt7603_filter_tx(struct mt7603_dev *dev, int idx, bool abort);
+void mt7603_filter_tx(struct mt7603_dev *dev, int mac_idx, int idx, bool abort);
 
 int mt7603_tx_prepare_skb(struct mt76_dev *mdev, void *txwi_ptr,
 			  enum mt76_txq_id qid, struct mt76_wcid *wcid,
drivers/net/wireless/mediatek/mt76/mt7603/regs.h +7 -0
index 3b901090b29c..a39c9a0fcb1c 100644
@@ -309,6 +309,13 @@ enum {
 #define MT_WF_ARB_TX_STOP_0		MT_WF_ARB(0x110)
 #define MT_WF_ARB_TX_STOP_1		MT_WF_ARB(0x114)
 
+#define MT_WF_ARB_TX_FLUSH_AC0		BIT(0)
+#define MT_WF_ARB_TX_FLUSH_AC1		BIT(5)
+#define MT_WF_ARB_TX_FLUSH_AC2		BIT(10)
+#define MT_WF_ARB_TX_FLUSH_AC3		BIT(16)
+#define MT_WF_ARB_TX_FLUSH_AC4		BIT(21)
+#define MT_WF_ARB_TX_FLUSH_AC5		BIT(26)
+
 #define MT_WF_ARB_BCN_START		MT_WF_ARB(0x118)
 #define MT_WF_ARB_BCN_START_BSSn(n)	BIT(0 + (n))
 #define MT_WF_ARB_BCN_START_T_PRE_TTTT	BIT(10)
Signature · 1 line
2.41.0
#3 lore raw
In reply to Felix Fietkau

Hello Felix.

Quoted 15 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
 drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -0
index ee976657bfc3..78552f10b377 100644

drivers/net/wireless/mediatek/mt76/mt7915/init.c mt7915_init_wiphy(struct mt7915_phy *phy) · line 419

Quoted 7 lines
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
 			if (!dev->dbdc_support)
 				vht_cap->cap |=
 					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
 					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
 		} else {
 			vht_cap->cap |=

For some reason this got backported into the stable kernel:

```
$ git log --oneline v6.5.2..v6.5.4 -- drivers/net/wireless/mediatek/mt76/mt7915/
c43017fbebcc3 wifi: mt76: mt7915: fix power-limits while chan_switch
edb1afe042c74 wifi: mt76: mt7915: fix tlv length of mt7915_mcu_get_chan_mib_info
9ec0dec0baea3 wifi: mt76: mt7915: remove VHT160 capability on MT7915
0e61f73e6ebc0 wifi: mt76: mt7915: fix capabilities in non-AP mode
6bce28ce28390 wifi: mt76: mt7915: fix command timeout in AP stop period
7af917d4864c6 wifi: mt76: mt7915: rework tx bytes counting when WED is active
feae00c6468ce wifi: mt76: mt7915: rework tx packets counting when WED is active
70bbcc4ad6544 wifi: mt76: mt7915: fix background radar event being blocked
```

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

Please check.

Thanks.

Signature · 1 line
Oleksandr Natalenko (post-factum)
#4 lore raw
Quoted 40 lines · Oleksandr Natalenko

Hello Felix.

Quoted 15 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
 drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -0
index ee976657bfc3..78552f10b377 100644

drivers/net/wireless/mediatek/mt76/mt7915/init.c mt7915_init_wiphy(struct mt7915_phy *phy) · line 419

Quoted 7 lines
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
 			if (!dev->dbdc_support)
 				vht_cap->cap |=
 					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
 					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
 		} else {
 			vht_cap->cap |=

For some reason this got backported into the stable kernel:

```
$ git log --oneline v6.5.2..v6.5.4 -- drivers/net/wireless/mediatek/mt76/mt7915/
c43017fbebcc3 wifi: mt76: mt7915: fix power-limits while chan_switch
edb1afe042c74 wifi: mt76: mt7915: fix tlv length of mt7915_mcu_get_chan_mib_info
9ec0dec0baea3 wifi: mt76: mt7915: remove VHT160 capability on MT7915
0e61f73e6ebc0 wifi: mt76: mt7915: fix capabilities in non-AP mode
6bce28ce28390 wifi: mt76: mt7915: fix command timeout in AP stop period
7af917d4864c6 wifi: mt76: mt7915: rework tx bytes counting when WED is active
feae00c6468ce wifi: mt76: mt7915: rework tx packets counting when WED is active
70bbcc4ad6544 wifi: mt76: mt7915: fix background radar event being blocked
```
Quoted 5 lines

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

Is your device also broken in 6.6-rc2?

thanks,

greg k-h

#5 lore raw

Hello.

Quoted 41 lines · On čtvrtek 21. září 2023 9:19:58 CEST Greg Kroah-Hartman wrote:
Quoted 40 lines · Oleksandr Natalenko

Hello Felix.

Quoted 15 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
 drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -0
index ee976657bfc3..78552f10b377 100644

drivers/net/wireless/mediatek/mt76/mt7915/init.c mt7915_init_wiphy(struct mt7915_phy *phy) · line 419

Quoted 7 lines
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
 			if (!dev->dbdc_support)
 				vht_cap->cap |=
 					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
 					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
 		} else {
 			vht_cap->cap |=

For some reason this got backported into the stable kernel:

```
$ git log --oneline v6.5.2..v6.5.4 -- drivers/net/wireless/mediatek/mt76/mt7915/
c43017fbebcc3 wifi: mt76: mt7915: fix power-limits while chan_switch
edb1afe042c74 wifi: mt76: mt7915: fix tlv length of mt7915_mcu_get_chan_mib_info
9ec0dec0baea3 wifi: mt76: mt7915: remove VHT160 capability on MT7915
0e61f73e6ebc0 wifi: mt76: mt7915: fix capabilities in non-AP mode
6bce28ce28390 wifi: mt76: mt7915: fix command timeout in AP stop period
7af917d4864c6 wifi: mt76: mt7915: rework tx bytes counting when WED is active
feae00c6468ce wifi: mt76: mt7915: rework tx packets counting when WED is active
70bbcc4ad6544 wifi: mt76: mt7915: fix background radar event being blocked
```
Quoted 7 lines
Quoted 5 lines

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

Is your device also broken in 6.6-rc2?

Yes, the same behaviour is observed with v6.6-rc2:

```
hostapd[1316]: Configured VHT capability [VHT_CAP_SUPP_CHAN_WIDTH_MASK] exceeds max value supported by the driver (1 > 0)
```

while having [VT160] in vht_capab=.

Thanks.

Quoted 4 lines

thanks,

greg k-h

Signature · 1 line
Oleksandr Natalenko (post-factum)
#6 lore raw

Seems nothing happened since this regression was reported more that a week ago. From a quick search on lore it seems Felix is not around currently; thus bringing the other mt76 maintainers in, maybe they can help out here to get this fixed rather sooner than later, as the culprit unfortunately made it to various stable trees. Ciao, Thorsten.

Quoted 49 lines · Oleksandr Natalenko
Quoted 40 lines · On čtvrtek 21. září 2023 9:19:58 CEST Greg Kroah-Hartman wrote:
Quoted 39 lines · Oleksandr Natalenko
Quoted 15 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
 drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -0
index ee976657bfc3..78552f10b377 100644

drivers/net/wireless/mediatek/mt76/mt7915/init.c mt7915_init_wiphy(struct mt7915_phy *phy) · line 419

Quoted 7 lines
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
 			if (!dev->dbdc_support)
 				vht_cap->cap |=
 					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
 					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
 		} else {
 			vht_cap->cap |=

For some reason this got backported into the stable kernel:

```
$ git log --oneline v6.5.2..v6.5.4 -- drivers/net/wireless/mediatek/mt76/mt7915/
c43017fbebcc3 wifi: mt76: mt7915: fix power-limits while chan_switch
edb1afe042c74 wifi: mt76: mt7915: fix tlv length of mt7915_mcu_get_chan_mib_info
9ec0dec0baea3 wifi: mt76: mt7915: remove VHT160 capability on MT7915
0e61f73e6ebc0 wifi: mt76: mt7915: fix capabilities in non-AP mode
6bce28ce28390 wifi: mt76: mt7915: fix command timeout in AP stop period
7af917d4864c6 wifi: mt76: mt7915: rework tx bytes counting when WED is active
feae00c6468ce wifi: mt76: mt7915: rework tx packets counting when WED is active
70bbcc4ad6544 wifi: mt76: mt7915: fix background radar event being blocked
```
Quoted 7 lines
Quoted 5 lines

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

Is your device also broken in 6.6-rc2?

Quoted 7 lines

Yes, the same behaviour is observed with v6.6-rc2:

```
hostapd[1316]: Configured VHT capability [VHT_CAP_SUPP_CHAN_WIDTH_MASK] exceeds max value supported by the driver (1 > 0)
```

while having [VT160] in vht_capab=.

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat) --
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
If I did something stupid, please tell me, as explained on that page.

#regzbot introduced 3ec5ac12ac8a4e
#regzbot ignore-activity

#7 lore raw
In reply to Oleksandr Natalenko

[TLDR: I'm adding this report to the list of tracked Linux kernel regressions; the text you find below is based on a few templates paragraphs you might have encountered already in similar form. See link in footer if these mails annoy you.]

Quoted 39 lines · Oleksandr Natalenko
Quoted 15 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
 drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -0
index ee976657bfc3..78552f10b377 100644

drivers/net/wireless/mediatek/mt76/mt7915/init.c mt7915_init_wiphy(struct mt7915_phy *phy) · line 419

Quoted 7 lines
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
 			if (!dev->dbdc_support)
 				vht_cap->cap |=
 					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
 					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
 		} else {
 			vht_cap->cap |=

For some reason this got backported into the stable kernel:

```
$ git log --oneline v6.5.2..v6.5.4 -- drivers/net/wireless/mediatek/mt76/mt7915/
c43017fbebcc3 wifi: mt76: mt7915: fix power-limits while chan_switch
edb1afe042c74 wifi: mt76: mt7915: fix tlv length of mt7915_mcu_get_chan_mib_info
9ec0dec0baea3 wifi: mt76: mt7915: remove VHT160 capability on MT7915
0e61f73e6ebc0 wifi: mt76: mt7915: fix capabilities in non-AP mode
6bce28ce28390 wifi: mt76: mt7915: fix command timeout in AP stop period
7af917d4864c6 wifi: mt76: mt7915: rework tx bytes counting when WED is active
feae00c6468ce wifi: mt76: mt7915: rework tx packets counting when WED is active
70bbcc4ad6544 wifi: mt76: mt7915: fix background radar event being blocked
```
Quoted 5 lines

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

Thanks for the report. To be sure the issue doesn't fall through the cracks unnoticed, I'm adding it to regzbot, the Linux kernel regression tracking bot:

#regzbot ^introduced 3ec5ac12ac8a4e..fe0ea395f0a351
#regzbot title wifi: mt76: mt7915: removal of VHT160 capability broke hostap #regzbot ignore-activity

This isn't a regression? This issue or a fix for it are already discussed somewhere else? It was fixed already? You want to clarify when the regression started to happen? Or point out I got the title or something else totally wrong? Then just reply and tell me -- ideally while also telling regzbot about it, as explained by the page listed in the footer of this mail.

Developers: When fixing the issue, remember to add 'Link:' tags pointing to the report (the parent of this mail). See page linked in footer for details.

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat) --
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
That page also explains what to do if mails like this annoy you.

#8 lore raw

[TLDR: This mail in primarily relevant for Linux kernel regression tracking. See link in footer if these mails annoy you.]

Quoted 5 lines · Linux regression tracking #adding (Thorsten Leemhuis)
Quoted 3 lines · Oleksandr Natalenko
Quoted 2 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Quoted 8 lines

[...]

Quoted 5 lines

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

Thanks for the report.

Removing this from the regression tracking after mentioning the intent to do so due to the tricky striation in my last report to Linus.

#regzbot inconclusive: tricky situation, no simple way out afaics #regzbot ignore-activity

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat) --
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
That page also explains what to do if mails like this annoy you.

#9 lore raw
In reply to Oleksandr Natalenko
Quoted 40 lines · Oleksandr Natalenko

Hello Felix.

Quoted 15 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
  drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
  1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -0
index ee976657bfc3..78552f10b377 100644

drivers/net/wireless/mediatek/mt76/mt7915/init.c mt7915_init_wiphy(struct mt7915_phy *phy) · line 419

Quoted 7 lines
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
  			if (!dev->dbdc_support)
  				vht_cap->cap |=
  					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
  					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
  		} else {
  			vht_cap->cap |=

For some reason this got backported into the stable kernel:

```
$ git log --oneline v6.5.2..v6.5.4 -- drivers/net/wireless/mediatek/mt76/mt7915/
c43017fbebcc3 wifi: mt76: mt7915: fix power-limits while chan_switch
edb1afe042c74 wifi: mt76: mt7915: fix tlv length of mt7915_mcu_get_chan_mib_info
9ec0dec0baea3 wifi: mt76: mt7915: remove VHT160 capability on MT7915
0e61f73e6ebc0 wifi: mt76: mt7915: fix capabilities in non-AP mode
6bce28ce28390 wifi: mt76: mt7915: fix command timeout in AP stop period
7af917d4864c6 wifi: mt76: mt7915: rework tx bytes counting when WED is active
feae00c6468ce wifi: mt76: mt7915: rework tx packets counting when WED is active
70bbcc4ad6544 wifi: mt76: mt7915: fix background radar event being blocked
```
Quoted 5 lines

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

I would say it is expected.

hostapd seems to solely rely on the VHT Supported Channel Width and does not seem to support the VHT Extended NSS BW stuff. So it only knows about full VHT 160 MHz support and not about half NSS VHT 160 MHz.

The hardware does not actually support full 160 MHz (despite the driver erroneously claiming support for it before this patch) so it make sense that hostapd fails to start the AP if the config file requests (full) VHT 160 MHz.

However, hostapd knows about half NSS HE 160 MHz and I suspect your configuration also requests HE 160 MHz, so 160 MHz works fine in HE but not in VHT.

In any case, it would help to know the hostapd version and your configuration file.

#10 lore raw

Hello.

Quoted 53 lines · On pátek 29. září 2023 14:59:26 CEST Nicolas Cavallari wrote:
Quoted 40 lines · Oleksandr Natalenko

Hello Felix.

Quoted 15 lines · On středa 26. července 2023 11:17:02 CEST Felix Fietkau wrote:

The IEEE80211_VHT_CAP_EXT_NSS_BW value already indicates support for half-NSS 160 MHz support, so it is wrong to also advertise full 160 MHz support.

Fixes c2f73eacee3b ("wifi: mt76: mt7915: add back 160MHz channel width support for MT7915")
---
  drivers/net/wireless/mediatek/mt76/mt7915/init.c | 1 -
  1 file changed, 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -0
index ee976657bfc3..78552f10b377 100644

drivers/net/wireless/mediatek/mt76/mt7915/init.c mt7915_init_wiphy(struct mt7915_phy *phy) · line 419

Quoted 7 lines
drivers/net/wireless/mediatek/mt76/mt7915/init.c +0 -1
@@ -414,7 +414,6 @@ mt7915_init_wiphy(struct mt7915_phy *phy)
  			if (!dev->dbdc_support)
  				vht_cap->cap |=
  					IEEE80211_VHT_CAP_SHORT_GI_160 |
-					IEEE80211_VHT_CAP_SUPP_CHAN_WIDTH_160MHZ |
  					FIELD_PREP(IEEE80211_VHT_CAP_EXT_NSS_BW_MASK, 1);
  		} else {
  			vht_cap->cap |=

For some reason this got backported into the stable kernel:

```
$ git log --oneline v6.5.2..v6.5.4 -- drivers/net/wireless/mediatek/mt76/mt7915/
c43017fbebcc3 wifi: mt76: mt7915: fix power-limits while chan_switch
edb1afe042c74 wifi: mt76: mt7915: fix tlv length of mt7915_mcu_get_chan_mib_info
9ec0dec0baea3 wifi: mt76: mt7915: remove VHT160 capability on MT7915
0e61f73e6ebc0 wifi: mt76: mt7915: fix capabilities in non-AP mode
6bce28ce28390 wifi: mt76: mt7915: fix command timeout in AP stop period
7af917d4864c6 wifi: mt76: mt7915: rework tx bytes counting when WED is active
feae00c6468ce wifi: mt76: mt7915: rework tx packets counting when WED is active
70bbcc4ad6544 wifi: mt76: mt7915: fix background radar event being blocked
```
Quoted 5 lines

and this broke my mt7915-based AP.

However, if I remove [VT160] capability from the hostapd config, things go back to normal. It does seem that 160 MHz still works even.

Is this expected?

I would say it is expected.

hostapd seems to solely rely on the VHT Supported Channel Width and does not seem to support the VHT Extended NSS BW stuff. So it only knows about full VHT 160 MHz support and not about half NSS VHT 160 MHz.

Quoted 6 lines

The hardware does not actually support full 160 MHz (despite the driver erroneously claiming support for it before this patch) so it make sense that hostapd fails to start the AP if the config file requests (full) VHT 160 MHz.

However, hostapd knows about half NSS HE 160 MHz and I suspect your configuration also requests HE 160 MHz, so 160 MHz works fine in HE but not in VHT.

I see, thanks for the explanation.

Quoted 1 line

In any case, it would help to know the hostapd version and your configuration file.

It's 2.10 with the following config at the very moment:

```
interface=wlp1s0
ctrl_interface=/run/hostapd/wlp1s0
driver=nl80211
ssid=xxx
utf8_ssid=1
interworking=1
ipaddr_type_availability=12
country_code=XX
auth_algs=1
wpa=2
wpa_key_mgmt=WPA-PSK FT-PSK
wpa_pairwise=CCMP
rsn_pairwise=CCMP
wpa_psk_file=/run/credentials/[email protected]/hostapd.wpa_psk
macaddr_acl=1
accept_mac_file=/etc/hostapd/hostapd.accept
noscan=1
hw_mode=a
ieee80211n=1
channel=100
vht_oper_chwidth=2
vht_oper_centr_freq_seg0_idx=114
ieee80211ax=1
he_oper_chwidth=2
he_oper_centr_freq_seg0_idx=114
ieee80211w=1
ocv=1
ieee80211d=1
ieee80211h=1
wmm_enabled=1
ht_capab=[RXLDPC][HT40+][GF][SHORT-GI-20][SHORT-GI-40][TX-STBC][RX-STBC1][MAX-AMSDU-7935][DSSS_CCK-40]
mobility_domain=abcd
ft_psk_generate_local=1
pmk_r1_push=1
rsn_preauth=1
rsn_preauth_interfaces=wifi-br
nas_identifier=yyy
r1_key_holder=yyy
r0kh=ff:ff:ff:ff:ff:ff * zzz
r1kh=00:00:00:00:00:00 00:00:00:00:00:00 zzz
disassoc_low_ack=1
fst_group_id=wifi-br
fst_priority=100
mbo=1
mbo_cell_data_conn_pref=1
require_ht=1
ieee80211ac=1
require_vht=1
vht_capab=[MAX-MPDU-7991][RXLDPC][SHORT-GI-80][SHORT-GI-160][TX-STBC-2BY1][RX-STBC-1][SU-BEAMFORMER][SU-BEAMFORMEE][MU-BEAMFORMER][MU-BEAMFORMEE][RX-ANTENNA-PATTERN][TX-ANTENNA-PATTERN][MAX-A-MPDU-LEN-EXP7]
```

(I wonder if I did other obviously stupid things here)

Signature · 1 line
Oleksandr Natalenko (post-factum)