linux-kernel - Re: [PATCH net] wifi: ath10k: fix dma_free

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

Message-ID: <ccba2a0b-3312-4a49-8de3-1601fa95d10f@oss.qualcomm.com>
Date: Tue, 6 Jan 2026 10:55:20 +0800
From: Baochen Qiang <baochen.qiang@....qualcomm.com>
To: Thomas Fourier <fourier.thomas@...il.com>
Cc: stable@...r.kernel.org, Jeff Johnson <jjohnson@...nel.org>,
        Kalle Valo <kvalo@....qualcomm.com>,
        Govind Singh
 <govinds@....qualcomm.com>,
        linux-wireless@...r.kernel.org, ath10k@...ts.infradead.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH net] wifi: ath10k: fix dma_free_coherent() pointer



On 1/6/2026 5:04 AM, Thomas Fourier wrote:
> dma_alloc_coherent() allocates a DMA mapped buffer and stores the
> addresses in XXX_unaligned fields.  Those should be reused when freeing
> the buffer rather than the aligned addresses.
> 
> Fixes: 2a1e1ad3fd37 ("ath10k: Add support for 64 bit ce descriptor")
> Cc: <stable@...r.kernel.org>
> Signed-off-by: Thomas Fourier <fourier.thomas@...il.com>

Reviewed-by: Baochen Qiang <baochen.qiang@....qualcomm.com>

> ---
>  drivers/net/wireless/ath/ath10k/ce.c | 16 ++++++++--------
>  1 file changed, 8 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/net/wireless/ath/ath10k/ce.c b/drivers/net/wireless/ath/ath10k/ce.c
> index 7bbda46cfd93..82f120ee1c66 100644
> --- a/drivers/net/wireless/ath/ath10k/ce.c
> +++ b/drivers/net/wireless/ath/ath10k/ce.c
> @@ -1727,8 +1727,8 @@ static void _ath10k_ce_free_pipe(struct ath10k *ar, int ce_id)
>  				  (ce_state->src_ring->nentries *
>  				   sizeof(struct ce_desc) +
>  				   CE_DESC_RING_ALIGN),
> -				  ce_state->src_ring->base_addr_owner_space,
> -				  ce_state->src_ring->base_addr_ce_space);
> +				  ce_state->src_ring->base_addr_owner_space_unaligned,
> +				  ce_state->src_ring->base_addr_ce_space_unaligned);
>  		kfree(ce_state->src_ring);
>  	}
>  
> @@ -1737,8 +1737,8 @@ static void _ath10k_ce_free_pipe(struct ath10k *ar, int ce_id)
>  				  (ce_state->dest_ring->nentries *
>  				   sizeof(struct ce_desc) +
>  				   CE_DESC_RING_ALIGN),
> -				  ce_state->dest_ring->base_addr_owner_space,
> -				  ce_state->dest_ring->base_addr_ce_space);
> +				  ce_state->dest_ring->base_addr_owner_space_unaligned,
> +				  ce_state->dest_ring->base_addr_ce_space_unaligned);
>  		kfree(ce_state->dest_ring);
>  	}
>  
> @@ -1758,8 +1758,8 @@ static void _ath10k_ce_free_pipe_64(struct ath10k *ar, int ce_id)
>  				  (ce_state->src_ring->nentries *
>  				   sizeof(struct ce_desc_64) +
>  				   CE_DESC_RING_ALIGN),
> -				  ce_state->src_ring->base_addr_owner_space,
> -				  ce_state->src_ring->base_addr_ce_space);
> +				  ce_state->src_ring->base_addr_owner_space_unaligned,
> +				  ce_state->src_ring->base_addr_ce_space_unaligned);
>  		kfree(ce_state->src_ring);
>  	}
>  
> @@ -1768,8 +1768,8 @@ static void _ath10k_ce_free_pipe_64(struct ath10k *ar, int ce_id)
>  				  (ce_state->dest_ring->nentries *
>  				   sizeof(struct ce_desc_64) +
>  				   CE_DESC_RING_ALIGN),
> -				  ce_state->dest_ring->base_addr_owner_space,
> -				  ce_state->dest_ring->base_addr_ce_space);
> +				  ce_state->dest_ring->base_addr_owner_space_unaligned,
> +				  ce_state->dest_ring->base_addr_ce_space_unaligned);
>  		kfree(ce_state->dest_ring);
>  	}
>  

seems ath12k has the same issue, would you like to fix it as well?