Received-SPF: pass (google.com: domain of linux-wireless-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Sender: kvalo=codeaurora.org@mg.codeaurora.org
DMARC-Filter: OpenDMARC Filter v1.4.1 smtp.codeaurora.org D89DCC4338F
From:   Kalle Valo <kvalo@codeaurora.org>
To:     Benjamin Li <benl@squareup.com>
Cc:     Bryan O'Donoghue <bryan.odonoghue@linaro.org>,
        Loic Poulain <loic.poulain@linaro.org>,
        linux-arm-msm@vger.kernel.org,
        "David S. Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>, wcn36xx@lists.infradead.org,
        linux-wireless@vger.kernel.org, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/2] wcn36xx: fix RX BD rate mapping for 5GHz legacy rates
References: <20211028223131.897548-1-benl@squareup.com>
        <20211028223131.897548-2-benl@squareup.com>
        <b3473977-5bb6-06df-55c3-85f08a29a964@linaro.org>
        <631a3ab4-56d9-5c1d-be53-c885747e3f7b@squareup.com>
Date:   Mon, 01 Nov 2021 15:00:30 +0200
In-Reply-To: <631a3ab4-56d9-5c1d-be53-c885747e3f7b@squareup.com> (Benjamin
        Li's message of "Thu, 28 Oct 2021 17:39:58 -0700")
Message-ID: <8735og2fpt.fsf@codeaurora.org>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Benjamin Li <benl@squareup.com> writes:

> On 10/28/21 5:30 PM, Bryan O'Donoghue wrote:
>> On 28/10/2021 23:31, Benjamin Li wrote:
>>> -=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 sta=
tus.rate_idx >=3D sband->n_bitrates) {
>> This fix was applied because we were getting a negative index
>>=20
>> If you want to remove that, you'll need to do something about this
>>=20
>> status.rate_idx -=3D 4;
>
> Hmm... so you're saying there's a FW bug where sometimes we get
> bd->rate_id =3D 0-7 (leading to status.rate_idx =3D 0-3) on a 5GHz
> channel?
>
> static const struct wcn36xx_rate wcn36xx_rate_table[] =3D {
>     /* 11b rates */
>     {  10, 0, RX_ENC_LEGACY, 0, RATE_INFO_BW_20 },
>     {  20, 1, RX_ENC_LEGACY, 0, RATE_INFO_BW_20 },
>     {  55, 2, RX_ENC_LEGACY, 0, RATE_INFO_BW_20 },
>     { 110, 3, RX_ENC_LEGACY, 0, RATE_INFO_BW_20 },
>
>     /* 11b SP (short preamble) */
>     {  10, 0, RX_ENC_LEGACY, RX_ENC_FLAG_SHORTPRE, RATE_INFO_BW_20 },
>     {  20, 1, RX_ENC_LEGACY, RX_ENC_FLAG_SHORTPRE, RATE_INFO_BW_20 },
>     {  55, 2, RX_ENC_LEGACY, RX_ENC_FLAG_SHORTPRE, RATE_INFO_BW_20 },
>     { 110, 3, RX_ENC_LEGACY, RX_ENC_FLAG_SHORTPRE, RATE_INFO_BW_20 },
>
> It sounds like we should WARN and drop the frame in that case. If
> you agree I'll send a v2.

BTW, please avoid using WARN() family of functions in the data path as
that can cause host crashes due to too much spamming in the logs. A some
kind of ratelimited version of an error message is much safer. For
example ath11k_warn() is ratelimited, maybe wcn36xx_warn() should be as
well?

--=20
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatc=
hes