[PATCH V2 3.2] b43: fill ctl1 word on all newer PHYs, fix PHY errors

[PATCH V2 3.2] b43: fill ctl1 word on all newer PHYs, fix PHY errors

[Rafał Miłecki]

This fixes PHY transmission errors reported on some LP-PHY and HT-PHY
cards. For LP-PHY they were quite rare and not really noticable. On
HT-PHY they were critical, OFDM rates were not available at all.

Signed-off-by: Rafa? Mi?ecki <zajec5@gmail.com>
---
V2: don't use >= for checking "old" vs. "new" PHY type, that was a
little hacky.
---
 drivers/net/wireless/b43/xmit.c |    8 ++++++--
 1 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/net/wireless/b43/xmit.c b/drivers/net/wireless/b43/xmit.c
index c73e860..8d69f7f 100644
--- a/drivers/net/wireless/b43/xmit.c
+++ b/drivers/net/wireless/b43/xmit.c
@@ -175,6 +175,7 @@ void b43_generate_plcp_hdr(struct b43_plcp_hdr4 *plcp,
 	}
 }
 
+/* TODO: verify if needed for SSLPN or LCN  */
 static u16 b43_generate_tx_phy_ctl1(struct b43_wldev *dev, u8 bitrate)
 {
 	const struct b43_phy *phy = &dev->phy;
@@ -256,6 +257,9 @@ int b43_generate_txhdr(struct b43_wldev *dev,
 	unsigned int plcp_fragment_len;
 	u32 mac_ctl = 0;
 	u16 phy_ctl = 0;
+	bool fill_phy_ctl1 = (phy->type == B43_PHYTYPE_LP ||
+			      phy->type == B43_PHYTYPE_N ||
+			      phy->type == B43_PHYTYPE_HT);
 	u8 extra_ft = 0;
 	struct ieee80211_rate *txrate;
 	struct ieee80211_tx_rate *rates;
@@ -531,7 +535,7 @@ int b43_generate_txhdr(struct b43_wldev *dev,
 			extra_ft |= B43_TXH_EFT_RTSFB_CCK;
 
 		if (rates[0].flags & IEEE80211_TX_RC_USE_RTS_CTS &&
-		    phy->type == B43_PHYTYPE_N) {
+		    fill_phy_ctl1) {
 			txhdr->phy_ctl1_rts = cpu_to_le16(
 				b43_generate_tx_phy_ctl1(dev, rts_rate));
 			txhdr->phy_ctl1_rts_fb = cpu_to_le16(
@@ -552,7 +556,7 @@ int b43_generate_txhdr(struct b43_wldev *dev,
 		break;
 	}
 
-	if (phy->type == B43_PHYTYPE_N) {
+	if (fill_phy_ctl1) {
 		txhdr->phy_ctl1 =
 			cpu_to_le16(b43_generate_tx_phy_ctl1(dev, rate));
 		txhdr->phy_ctl1_fb =
-- 
1.7.3.4

[PATCH V2 3.2] b43: fill ctl1 word on all newer PHYs, fix PHY errors

[Chris Vine]

On Mon, 17 Oct 2011 16:59:25 +0200
Rafa? Mi?ecki <zajec5@gmail.com> wrote:
> This fixes PHY transmission errors reported on some LP-PHY and HT-PHY
> cards. For LP-PHY they were quite rare and not really noticable. On
> HT-PHY they were critical, OFDM rates were not available at all.
[snip]

Does this fix the "b43-phy0 debug: Out of order TX status report on DMA
ring 1" stuff?

Chris

[PATCH V2 3.2] b43: fill ctl1 word on all newer PHYs, fix PHY errors

[Rafał Miłecki]

W dniu 18 pa?dziernika 2011 02:04 u?ytkownik Chris Vine
<chris@cvine.freeserve.co.uk> napisa?:
> On Mon, 17 Oct 2011 16:59:25 +0200
> Rafa? Mi?ecki <zajec5@gmail.com> wrote:
>> This fixes PHY transmission errors reported on some LP-PHY and HT-PHY
>> cards. For LP-PHY they were quite rare and not really noticable. On
>> HT-PHY they were critical, OFDM rates were not available at all.
> [snip]
>
> Does this fix the "b43-phy0 debug: Out of order TX status report on DMA
> ring 1" stuff?

I don't think so :( It's about
b43-phy0 ERROR: PHY transmission error

-- 
Rafa?