linux-kernel - Re: [RFC PATCH] x86/cpufeatures: Enumerate new AVX512 bfloat16 instructions

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, 11 Jun 2019 11:19:20 -0700
From:   Fenghua Yu <fenghua.yu@...el.com>
To:     Borislav Petkov <bp@...en8.de>
Cc:     Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, H Peter Anvin <hpa@...or.com>,
        Ravi V Shankar <ravi.v.shankar@...el.com>,
        linux-kernel <linux-kernel@...r.kernel.org>, x86 <x86@...nel.org>
Subject: Re: [RFC PATCH] x86/cpufeatures: Enumerate new AVX512 bfloat16
 instructions

On Mon, Jun 10, 2019 at 09:20:26PM +0200, Borislav Petkov wrote:
> On Mon, Jun 10, 2019 at 10:02:38AM -0700, Fenghua Yu wrote:
> > AVX512 Vector Neural Network Instructions (VNNI) in Intel Deep Learning
> > Boost support bfloat16 format (BF16). BF16 is a short version of FP32 and
> > has several advantages over FP16. BF16 offers more than enough range for
> > deep learning training tasks and doesn't need to handle hardware exception
> > as this is a performance optimization. FP32 accumulation after the
> > multiply is essential to achieve sufficient numerical behavior on an
> > application level. 
> > 
> > AVX512 bfloat16 instructions can be enumerated by:
> > 	CPUID.(EAX=7,ECX=1):EAX[bit 5] AVX512_BF16
> >     
> > Detailed information of the CPUID bit and AVX512 bfloat16 instructions
> > can be found in the latest Intel Architecture Instruction Set Extensions
> > and Future Features Programming Reference.
> > 
> > Signed-off-by: Fenghua Yu <fenghua.yu@...el.com>
> > ---
> > 
> > Since split lock feature (to-be-upstreamed) occupies the last bit 
> > of word 7, need to create a new word 19 to host AVX512_BF16 and other
> > future features.
> 
> Is CPUID.(EAX=7,ECX=1):EAX going to contain only feature bits? If so,
> just make it a proper feature word instead of a linux-specific one.

AFAICT, there will be more features in the EAX register in the future.
Ok. I will use a feature word for CPUID.(EAX=7,ECX=1):EAX.

> 
> Also, while on the subject, you can recycle word 11
> 
> /* Intel-defined CPU QoS Sub-leaf, CPUID level 0x0000000F:0 (EDX), word 11 */
> #define X86_FEATURE_CQM_LLC             (11*32+ 1) /* LLC QoS if 1 */
> 
> and move it to scattered as it is a complete waste. Word 12 too, for
> that matter. But do that in separate patches.

So can I re-organize word 11 and 12 as follows?

1. Change word 11 to host scattered features.
2. Move the previos features in word 11 and word 12 to word 11:
/*
 * Extended auxiliary flags: Linux defined - For features scattered in various
 * CPUID levels and sub-leaves like CPUID level 7 and sub-leaf 1, etc, word 19.
 */
#define X86_FEATURE_CQM_LLC             (11*32+ 0) /* LLC QoS if 1 */
#define X86_FEATURE_CQM_OCCUP_LLC       (11*32+ 1) /* LLC occupancy monitoring */
#define X86_FEATURE_CQM_MBM_TOTAL       (11*32+ 2) /* LLC Total MBM monitoring */
#define X86_FEATURE_CQM_MBM_LOCAL       (11*32+ 3) /* LLC Local MBM monitoring */

3. Change word 12 to host CPUID.(EAX=7,ECX=1):EAX:
/* Intel-defined CPU features, CPUID level 0x7:1 (EAX), word 12 */
#define X86_FEATURE_AVX512_BF16         (12*32+ 0) /* BFLOAT16 instructions */

4. Do other necessary changes to match the new word 11 and word 12.

Thanks.

-Fenghua