linux-kernel - Re: [PATCH 12/14] dma-direct: handle the memory encryption bit in common code

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]

Date:   Mon, 19 Mar 2018 16:55:51 +0000
From:   Will Deacon <will.deacon@....com>
To:     Christoph Hellwig <hch@....de>
Cc:     Robin Murphy <robin.murphy@....com>, x86@...nel.org,
        Tom Lendacky <thomas.lendacky@....com>,
        Konrad Rzeszutek Wilk <konrad.wilk@...cle.com>,
        linux-kernel@...r.kernel.org, Muli Ben-Yehuda <mulix@...ix.org>,
        iommu@...ts.linux-foundation.org,
        David Woodhouse <dwmw2@...radead.org>,
        Catalin Marinas <catalin.marinas@....com>
Subject: Re: [PATCH 12/14] dma-direct: handle the memory encryption bit in
 common code

On Mon, Mar 19, 2018 at 05:03:43PM +0100, Christoph Hellwig wrote:
> On Mon, Mar 19, 2018 at 03:48:33PM +0000, Will Deacon wrote:
> > Why can't we just resolve the conflict by adding the underscores?
> 
> We can solve the conflict easily that way.  But that's not the point.
> 
> The point is that I've been fighting hard to consolidate dma code
> given that the behavior really is common and not arch specific.  And
> this one is another case like that:  the fact that the non-coherent
> dma boundary is bigger than the exposed size is something that can
> easily happen elsewhere, so there is no need to duplicate a lot
> of code for that.

Fair enough, although I wouldn't say it's a *lot* of code being duplicated.
Are there other architectures working around this issue too? I couldn't
see anything in the other dma-direct.h headers.

> Nevermind that the commit should at least be three different patches:
> 
>  (1) revert the broken original commit
>  (2) increase the dma min alignment
>  (3) put the swiotlb workaround in place

I'd agree with you if this wasn't already queued and sitting in -next.
Reverting what we currently have seems a bit OTT now. Catalin?

Will