linux-kernel - Re: [PATCH] crypto: ccp - fix memleak in ccp_init_dm

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]

Date:   Mon, 27 Nov 2023 09:45:15 +0800 (GMT+08:00)
From:   dinghao.liu@....edu.cn
To:     "Tom Lendacky" <thomas.lendacky@....com>
Cc:     "John Allen" <john.allen@....com>,
        "Herbert Xu" <herbert@...dor.apana.org.au>,
        "David S. Miller" <davem@...emloft.net>,
        linux-crypto@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] crypto: ccp - fix memleak in ccp_init_dm_workarea

> On 11/23/23 01:03, Dinghao Liu wrote:
> > When dma_map_single() fails, we should free wa->address to
> > prevent potential memleak.
> 
> You should expand on this a bit more. The ccp_dm_free() function is 
> normally called to free that memory, but many of the call spots don't 
> expect to have to call ccp_dm_free() on an error return from 
> ccp_init_dm_workarea(). Because of that, the memory is freed in 
> ccp_init_dm_workarea().
> 
> So a more detailed commit message about why this change is needed should 
> be provided.

Thanks for your suggestion! I will resend a new patch soon.

> 
> > 
> > Fixes: 63b945091a07 ("crypto: ccp - CCP device driver and interface support")
> > Signed-off-by: Dinghao Liu <dinghao.liu@....edu.cn>
> > ---
> >   drivers/crypto/ccp/ccp-ops.c | 4 +++-
> >   1 file changed, 3 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/crypto/ccp/ccp-ops.c b/drivers/crypto/ccp/ccp-ops.c
> > index aa4e1a500691..45552ae6347e 100644
> > --- a/drivers/crypto/ccp/ccp-ops.c
> > +++ b/drivers/crypto/ccp/ccp-ops.c
> > @@ -179,8 +179,10 @@ static int ccp_init_dm_workarea(struct ccp_dm_workarea *wa,
> >   
> >   		wa->dma.address = dma_map_single(wa->dev, wa->address, len,
> >   						 dir);
> > -		if (dma_mapping_error(wa->dev, wa->dma.address))
> > +		if (dma_mapping_error(wa->dev, wa->dma.address)) {
> > +			kfree(wa->address);
> 
> If future changes should result in ccp_dm_free() being called on an error 
> returned from ccp_init_dm_workarea(), you should add:
> 
> 			wa->address = NULL;
> 

I will fix this in the new patch, thanks!

Regards,
Dinghao