From: Francesco Dolcini <francesco.dolcini@toradex.com>

This reverts commit 34253084291cb210b251d64657958b8041ce4ab1.

That commit introduces a regression, when HT40 mode is enabled,
received packets are lost, this was experience with W8997 with both
SDIO-UART and SDIO-SDIO variants. From an initial investigation the
issue solves on its own after some time, but it's not clear what is
the reason. Given that this was just a performance optimization, let's
revert it till we have a better understanding of the issue and a proper
fix.

Cc: Jeff Chen <jeff.chen_1@nxp.com>
Cc: stable@vger.kernel.org
Fixes: 34253084291c ("wifi: mwifiex: Fix HT40 bandwidth issue.")
Closes: https://lore.kernel.org/all/20250603203337.GA109929@francesco-nb/
Signed-off-by: Francesco Dolcini <francesco.dolcini@toradex.com>
---
 drivers/net/wireless/marvell/mwifiex/11n.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/net/wireless/marvell/mwifiex/11n.c b/drivers/net/wireless/marvell/mwifiex/11n.c
index 738bafc3749b..66f0f5377ac1 100644
--- a/drivers/net/wireless/marvell/mwifiex/11n.c
+++ b/drivers/net/wireless/marvell/mwifiex/11n.c
@@ -403,14 +403,12 @@ mwifiex_cmd_append_11n_tlv(struct mwifiex_private *priv,
 
 		if (sband->ht_cap.cap & IEEE80211_HT_CAP_SUP_WIDTH_20_40 &&
 		    bss_desc->bcn_ht_oper->ht_param &
-		    IEEE80211_HT_PARAM_CHAN_WIDTH_ANY) {
-			chan_list->chan_scan_param[0].radio_type |=
-				CHAN_BW_40MHZ << 2;
+		    IEEE80211_HT_PARAM_CHAN_WIDTH_ANY)
 			SET_SECONDARYCHAN(chan_list->chan_scan_param[0].
 					  radio_type,
 					  (bss_desc->bcn_ht_oper->ht_param &
 					  IEEE80211_HT_PARAM_CHA_SEC_OFFSET));
-		}
+
 		*buffer += struct_size(chan_list, chan_scan_param, 1);
 		ret_len += struct_size(chan_list, chan_scan_param, 1);
 	}
-- 
2.39.5

On Thu, 2025-06-05 at 12:03 +0200, Francesco Dolcini wrote:
> From: Francesco Dolcini <francesco.dolcini@toradex.com>
> 
> This reverts commit 34253084291cb210b251d64657958b8041ce4ab1.

I'm confused. If you want it reverted in wireless, this is the wrong
sha1? If you want bit only reverted in stable, why are you tagging it
for wireless?

johannes

On Thu, Jun 05, 2025 at 01:40:41PM +0200, Johannes Berg wrote:
> On Thu, 2025-06-05 at 12:03 +0200, Francesco Dolcini wrote:
> > From: Francesco Dolcini <francesco.dolcini@toradex.com>
> > 
> > This reverts commit 34253084291cb210b251d64657958b8041ce4ab1.
> 
> I'm confused. If you want it reverted in wireless, this is the wrong
> sha1? If you want bit only reverted in stable, why are you tagging it
> for wireless?

Because I did a mistake :-( This is supposed to be reverted on Linus
tree, not just in stable, I'll send a v2 with this fixed, sorry.

Francesco