linux-kernel - Re: [PATCH v5] KVM: x86: avoid large stack allocations in em

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:   Wed, 31 May 2017 18:05:05 -0700
From:   Nick Desaulniers <nick.desaulniers@...il.com>
To:     Paolo Bonzini <pbonzini@...hat.com>
Cc:     Radim Krčmář <rkrcmar@...hat.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>,
        "H. Peter Anvin" <hpa@...or.com>, x86@...nel.org,
        kvm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v5] KVM: x86: avoid large stack allocations in em_fxrstor

On Wed, May 31, 2017 at 07:01:29AM -0400, Paolo Bonzini wrote:
> > +		size = offsetof(struct fxregs_state, xmm_space[16]);
> This still has the same issue (it should be multiplied by 4).

I'm still misunderstanding the math here.

Why multiplied by four, in this case? 8 * 16 / 4 is used in other cases.

Also, previously Radim wrote:

>> +      size = offsetof(struct fxregs_state, xmm_space[8]);
> This should be the size of first 8 XMM registers, but xmm_space is of
> type u32, so the correct size is
>   xmm_space[8 * 16/sizeof(*fx_state.xmm_space)].

So I think my calculation is off in xmm_offset still?  Can we make use
of well-named variables, in place of these constants? Otherwise the math
is hard to follow.

> Thanks Nick for the patches and Radim for the reviews!
> Paolo

Thanks for the code review!