linux-kernel - Re: [PATCH v4 7/7] mtrr, mm, x86: Enhance MTRR checks for KVA huge page mapping

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:	Tue, 12 May 2015 10:57:09 -0600
From:	Toshi Kani <toshi.kani@...com>
To:	Borislav Petkov <bp@...en8.de>
Cc:	akpm@...ux-foundation.org, hpa@...or.com, tglx@...utronix.de,
	mingo@...hat.com, linux-mm@...ck.org, x86@...nel.org,
	linux-kernel@...r.kernel.org, dave.hansen@...el.com,
	Elliott@...com, pebolle@...cali.nl
Subject: Re: [PATCH v4 7/7] mtrr, mm, x86: Enhance MTRR checks for KVA huge
 page mapping

On Tue, 2015-05-12 at 18:31 +0200, Borislav Petkov wrote:
> On Tue, May 12, 2015 at 08:30:30AM -0600, Toshi Kani wrote:
> > MTRR_TYPE_INVALID means MTRRs disabled.  So, the caller checking with
> > this value is the same as checking with mtrr_enabled() you suggested.
> 
> So then you don't have to set *uniform = 1 on entry to
> mtrr_type_lookup(). And change the retval test
> 
> 	if ((!uniform) && (mtrr != MTRR_TYPE_WRBACK))
> 
> to
> 	if ((mtrr != MTRR_TYPE_INVALID) && (!uniform) && (mtrr != MTRR_TYPE_WRBACK))

Yes, that's what I was thinking as well.  Will do.

> You can put the MTRR_TYPE_INVALID first so that it shortcuts.
> 
> You need the distinction between MTRRs *disabled* and an MTRR region
> being {non-,}uniform.
> 
> If MTRRs are disabled, uniform doesn't *mean* *anything* because it is
> undefined. When MTRRs are disabled, the range is *not* covered by MTRRs
> because, well, them MTRRs are disabled.
> 
> And it might be fine for *your* use case to set *uniform even when MTRRs
> are disabled but it might matter in the future. So we better design it
> correct from the beginning.

I think it is a matter of how "uniform" is defined, but your point is
taken and I will change it accordingly.

Thanks,
-Toshi


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/