linux-kernel - Re: [PATCH v9 06/27] x86/fpu/xstate: Create guest fpstate with guest specific config

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: Tue, 30 Jan 2024 01:38:45 +0000
From: "Edgecombe, Rick P" <rick.p.edgecombe@...el.com>
To: "Yang, Weijiang" <weijiang.yang@...el.com>, "Hansen, Dave"
	<dave.hansen@...el.com>, "seanjc@...gle.com" <seanjc@...gle.com>,
	"x86@...nel.org" <x86@...nel.org>, "linux-kernel@...r.kernel.org"
	<linux-kernel@...r.kernel.org>, "yuan.yao@...ux.intel.com"
	<yuan.yao@...ux.intel.com>, "kvm@...r.kernel.org" <kvm@...r.kernel.org>,
	"pbonzini@...hat.com" <pbonzini@...hat.com>
CC: "john.allen@....com" <john.allen@....com>, "peterz@...radead.org"
	<peterz@...radead.org>, "Gao, Chao" <chao.gao@...el.com>,
	"mlevitsk@...hat.com" <mlevitsk@...hat.com>
Subject: Re: [PATCH v9 06/27] x86/fpu/xstate: Create guest fpstate with guest
 specific config

On Tue, 2024-01-23 at 18:41 -0800, Yang Weijiang wrote:
> -bool fpu_alloc_guest_fpstate(struct fpu_guest *gfpu)
> +static struct fpstate *__fpu_alloc_init_guest_fpstate(struct
> fpu_guest *gfpu)
>  {
>         struct fpstate *fpstate;
>         unsigned int size;
>  
> -       size = fpu_user_cfg.default_size + ALIGN(offsetof(struct
> fpstate, regs), 64);
> +       /*
> +        * fpu_guest_cfg.default_size is initialized to hold all
> enabled
> +        * xfeatures except the user dynamic xfeatures. If the user
> dynamic
> +        * xfeatures are enabled, the guest fpstate will be re-
> allocated to
> +        * hold all guest enabled xfeatures, so omit user dynamic
> xfeatures
> +        * here.
> +        */
> +       size = fpu_guest_cfg.default_size +
> +              ALIGN(offsetof(struct fpstate, regs), 64);

Minor, I'm not sure the extra char warrants changing it to a wrapped
line, but that's just my personal opinion.

> +
>         fpstate = vzalloc(size);
>         if (!fpstate)
> -               return false;
> +               return NULL;
> +       /*
> +        * Initialize sizes and feature masks, use fpu_user_cfg.*
> +        * for user_* settings for compatibility of exiting uAPIs.
> +        */
> +       fpstate->size           = fpu_guest_cfg.default_size;
> +       fpstate->xfeatures      = fpu_guest_cfg.default_features;
> +       fpstate->user_size      = fpu_user_cfg.default_size;
> +       fpstate->user_xfeatures = fpu_user_cfg.default_features;
> +       fpstate->xfd            = 0;
>  
> -       /* Leave xfd to 0 (the reset value defined by spec) */
> -       __fpstate_reset(fpstate, 0);
>         fpstate_init_user(fpstate);
>         fpstate->is_valloc      = true;
>         fpstate->is_guest       = true;
>  
>         gfpu->fpstate           = fpstate;
> -       gfpu->xfeatures         = fpu_user_cfg.default_features;
> -       gfpu->perm              = fpu_user_cfg.default_features;
> +       gfpu->xfeatures         = fpu_guest_cfg.default_features;
> +       gfpu->perm              = fpu_guest_cfg.default_features;
> +
> +       return fpstate;
> +}
> +
> +bool fpu_alloc_guest_fpstate(struct fpu_guest *gfpu)
> +{
> +       struct fpstate *fpstate;
> +
> +       fpstate = __fpu_alloc_init_guest_fpstate(gfpu);

The above two statements could be just one line and still even fit
under 80 chars.

All the same,

Reviewed-by: Rick Edgecombe <rick.p.edgecombe@...el.com>

> +
> +       if (!fpstate)
> +               return false;
>