linux-kernel - Re: [PATCH] dmaengine: ti: edma: support sw triggered chans in of_edma

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <55d56d32-64b5-4cef-8c1c-fb29b7b93b87@gmail.com>
Date: Wed, 19 Feb 2025 17:06:54 +0200
From: Péter Ujfalusi <peter.ujfalusi@...il.com>
To: Matthew Majewski <mattwmajewski@...il.com>, Vinod Koul <vkoul@...nel.org>
Cc: dmaengine@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] dmaengine: ti: edma: support sw triggered chans in
 of_edma_xlate()



On 16/02/2025 23:47, Matthew Majewski wrote:
> The .of_edma_xlate() function always sets the hw_triggered flag to
> true. This causes sw triggered channels consumed via the device-tree
> to not function properly, as the driver incorrectly assumes they are
> hw triggered. Modify the xlate() function to correctly set the
> hw_triggered flag to false for channels reserved for memcpy
> operation (ie, sw triggered).
> 
> Signed-off-by: Matthew Majewski <mattwmajewski@...il.com>

Good catch,
Acked-by: Peter Ujfalusi <peter.ujfalusi@...il.com>

> ---
>  drivers/dma/ti/edma.c | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/dma/ti/edma.c b/drivers/dma/ti/edma.c
> index 4ece125b2ae7..0554a18d84ba 100644
> --- a/drivers/dma/ti/edma.c
> +++ b/drivers/dma/ti/edma.c
> @@ -2258,8 +2258,12 @@ static struct dma_chan *of_edma_xlate(struct of_phandle_args *dma_spec,
>  
>  	return NULL;
>  out:
> -	/* The channel is going to be used as HW synchronized */
> -	echan->hw_triggered = true;
> +	/*
> +	 * The channel is going to be HW synchronized, unless it was
> +	 * reserved as a memcpy channel
> +	 */
> +	echan->hw_triggered =
> +		!edma_is_memcpy_channel(i, ecc->info->memcpy_channels);
>  	return dma_get_slave_channel(chan);
>  }
>  #else

-- 
Péter