linux-kernel - Re: [PATCH v1 02/19] infiniband/mthca: Fix dma_map

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 for Android: free password hash cracker in your pocket

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

Date:   Sun, 21 Aug 2022 14:46:43 +0300
From:   Leon Romanovsky <leon@...nel.org>
To:     Jack Wang <jinpu.wang@...os.com>
Cc:     linux-kernel@...r.kernel.org, Jason Gunthorpe <jgg@...pe.ca>,
        Christophe JAILLET <christophe.jaillet@...adoo.fr>,
        Kees Cook <keescook@...omium.org>,
        Håkon Bugge <haakon.bugge@...cle.com>,
        linux-rdma@...r.kernel.org
Subject: Re: [PATCH v1 02/19] infiniband/mthca: Fix dma_map_sg error check

On Fri, Aug 19, 2022 at 08:07:44AM +0200, Jack Wang wrote:
> dma_map_sg return 0 on error, in case of error set
> EIO as return code.
> 
> Cc: Jason Gunthorpe <jgg@...pe.ca>
> Cc: Leon Romanovsky <leon@...nel.org>
> Cc: Christophe JAILLET <christophe.jaillet@...adoo.fr>
> Cc: Kees Cook <keescook@...omium.org>
> Cc: "Håkon Bugge" <haakon.bugge@...cle.com>
> Cc: linux-rdma@...r.kernel.org
> Cc: linux-kernel@...r.kernel.org
> Signed-off-by: Jack Wang <jinpu.wang@...os.com>
> ---
>  drivers/infiniband/hw/mthca/mthca_memfree.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/infiniband/hw/mthca/mthca_memfree.c b/drivers/infiniband/hw/mthca/mthca_memfree.c
> index f2734a5c5f26..44fd5fdf64d5 100644
> --- a/drivers/infiniband/hw/mthca/mthca_memfree.c
> +++ b/drivers/infiniband/hw/mthca/mthca_memfree.c
> @@ -189,7 +189,7 @@ struct mthca_icm *mthca_alloc_icm(struct mthca_dev *dev, int npages,
>  						   chunk->npages,
>  						   DMA_BIDIRECTIONAL);
>  
> -				if (chunk->nsg <= 0)
> +				if (!chunk->nsg)
>  					goto fail;
>  			}
>  
> @@ -208,7 +208,7 @@ struct mthca_icm *mthca_alloc_icm(struct mthca_dev *dev, int npages,
>  		chunk->nsg = dma_map_sg(&dev->pdev->dev, chunk->mem,
>  					chunk->npages, DMA_BIDIRECTIONAL);
>  
> -		if (chunk->nsg <= 0)
> +		if (!chunk->nsg)
>  			goto fail;
>  	}
>  
> @@ -482,8 +482,9 @@ int mthca_map_user_db(struct mthca_dev *dev, struct mthca_uar *uar,
>  
>  	ret = dma_map_sg(&dev->pdev->dev, &db_tab->page[i].mem, 1,
>  			 DMA_TO_DEVICE);
> -	if (ret < 0) {
> +	if (!ret) {

This code is not equivalent to original code. mthca didn't count ret == 0
as an error. Most likely, it is a bug, but I don't want to change old and
unmaintained driver without any real need.

Thanks

>  		unpin_user_page(pages[0]);
> +		ret = -EIO;
>  		goto out;
>  	}
>  
> -- 
> 2.34.1
>