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;
MIME-Version: 1.0
References: <20210818150943.1630199-1-festevam@denx.de> <17b5a29d098.279b.9696ff82abe5fb6502268bdc3b0467d4@gmail.com>
 <369c1f89-df33-455b-773e-a2e0dbf582ff@eero.com>
In-Reply-To: <369c1f89-df33-455b-773e-a2e0dbf582ff@eero.com>
From:   Fabio Estevam <festevam@gmail.com>
Date:   Wed, 18 Aug 2021 17:43:53 -0300
Message-ID: <CAOMZO5DbZ11HE4un20tBkLbjGWcyUu8KpTV+gEtVf=rT2JyT+A@mail.gmail.com>
Subject: Re: [PATCH] ath10k: Do not call dma_alloc_coherent() for SDIO and USB
To:     Peter Oh <peter.oh@eero.com>
Cc:     Arend van Spriel <aspriel@gmail.com>,
        Fabio Estevam <festevam@denx.de>,
        Kalle Valo <kvalo@codeaurora.org>, ath10k@lists.infradead.org,
        linux-wireless <linux-wireless@vger.kernel.org>,
        Christoph Hellwig <hch@lst.de>, erik.stromdahl@gmail.com
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi Peter,

On Wed, Aug 18, 2021 at 2:35 PM Peter Oh <peter.oh@eero.com> wrote:
>
>
> >> Fix the problem by not calling dma_alloc_coherent() when the device
> >> is not DMA capable, such as SDIO and USB.
> >>
> ath10k calls dma_alloc_coherent multiple places including
> ath10k_htt_rx_alloc.
>
> Do SDIO and USB not use such data path function at all?

Good point! Maybe I am not just hitting these paths on my simple tests.

Kalle,

How should we handle this?

Thanks