From: "Shankaran, Akash" <akash.shankaran@intel.com>
To: Nathan Bossart <nathandbossart@gmail.com>, "Amonson, Paul D"
	<paul.d.amonson@intel.com>
CC: Tom Lane <tgl@sss.pgh.pa.us>, David Rowley <dgrowleyml@gmail.com>, "Andres
 Freund" <andres@anarazel.de>, Alvaro Herrera <alvherre@alvh.no-ip.org>, "Noah
 Misch" <noah@leadboat.com>, Matthias van de Meent
	<boekewurm+postgres@gmail.com>, "pgsql-hackers@lists.postgresql.org"
	<pgsql-hackers@lists.postgresql.org>
Subject: RE: Popcount optimization using AVX512
Thread-Topic: Popcount optimization using AVX512
Thread-Index: AdoNliAXaie8RHrzTICrAiLVKKiYCgAsQKyAALaJTAAAAQ7nAAAAxSWAAAGNHgAAA/mZgAAeFdGAAZKQbBAAAw6egA3NuLcAAAvVM4ACbQuxsAAhlXkAAHPWMbAAAkOrAACYCUjgAAMc3wABvefW4AD8TV6wADf+AoAAmWkVgACOpocAAAl+RwAAJXNxYAAA1VoAAABadxAAZi2hAAAIeekAAMiJepAACPYvAAAbpMKQAJeTLYAAcNGUAAAm2oYAAAFGgCAAAIPdAAACb8MAAAdpgwAAADR0AAAAsQcAAAAMEoAAAWCPgAAC3eIAACrnWQAABeBEkAA1zxYAACb9mpAAwL24cAAATJYAAAAxGQAABx2FwAAC9E+AAGiTRwAAASAwAAAwbEYAAADYEgAAJL9AgAABFXaAAABdIgAAAC9GEA==
Date: Fri, 29 Mar 2024 16:31:09 +0000
Message-ID: <PH0PR11MB50001026ADCC5D9C83C8F9C9F23A2@PH0PR11MB5000.namprd11.prod.outlook.com>
References: <2490633.1711379522@sss.pgh.pa.us>
 <BL1PR11MB53043014C10F2D30B4231F9ADC362@BL1PR11MB5304.namprd11.prod.outlook.com>
 <BL1PR11MB53049F1796E834DBE0E05864DC362@BL1PR11MB5304.namprd11.prod.outlook.com>
 <20240325200551.GB2924132@nathanxps13> <20240327220010.GA231058@nathanxps13>
 <BL1PR11MB53047289F146DB43C9B2FC71DC342@BL1PR11MB5304.namprd11.prod.outlook.com>
 <20240328213854.GA916732@nathanxps13>
 <BL1PR11MB5304E51336123CE6F041A920DC3B2@BL1PR11MB5304.namprd11.prod.outlook.com>
 <20240329153515.GA1046039@nathanxps13>
 <BL1PR11MB5304D039E04286E9DF048473DC3A2@BL1PR11MB5304.namprd11.prod.outlook.com>
 <20240329161642.GA1106154@nathanxps13>
In-Reply-To: <20240329161642.GA1106154@nathanxps13>
Accept-Language: en-US
Content-Language: en-US
x-ms-exchange-antispam-messagedata-chunkcount: 1
x-ms-exchange-antispam-messagedata-0: =?us-ascii?Q?Kjx88rWLIohAToTnlwL5HlHQSXxTTR6PqeJ2lf167bNeI0Pih9+zMuA05k20?=
 =?us-ascii?Q?BByY0PiJ/XxlxasSn7/8hdbHRbOfEbIfuVNuq6Ix2lelqkHX+MiWhFV5ChPL?=
 =?us-ascii?Q?aLH8gyoafge8cDWaDt6xY/+sNeQN1tQCwNtBMGahjVl22ijnWA6zS5CjThv9?=
 =?us-ascii?Q?Tv/i4bQnQjpltSEniI+0bTCZmQ2J3OxB8xdD0qu1LLnZ6xZwrSdnzZmckv1S?=
 =?us-ascii?Q?PsnIwi1PbJmj+VFi9kE7gVKXxu3JDAGG/oyiqFM8F6s7SV7b9DCejcQs+KIv?=
 =?us-ascii?Q?MU8N5GpjEr/z6YNYJGuyf5mZ7crNNE5txLGNNfhfic4kfRiy0LDX0BKEtS6L?=
 =?us-ascii?Q?IKv0+LBOhRRtNuhUMDVTebadZKiOIo5EIm/VAnnB9MaN/Mc7aLRsAygqQ7dt?=
 =?us-ascii?Q?DL3xMhmx9Vm4Cf2LHNJPNuOSp1O4X0cx+wFsuFov+T+WgLJhadlun7B3tk2d?=
 =?us-ascii?Q?bNGhN7ATQhRKfPmEcAPIf3DXC3z4hI1bMcT8ulbRf1VLkCwoWjJoqufp4xxb?=
 =?us-ascii?Q?oC4Kg4N2mIR729ckcHmjo1as5KcXN9nRUvxD6yZ4LsM0iSyO1+x7Pewvul0X?=
 =?us-ascii?Q?3MIt+mnYjU3uKrIrRi1Z2ah+xNObeBqe4aX/ljGVO+1IuRFYrOnoDRZrqhsx?=
 =?us-ascii?Q?6SxSLltY6P1Z7GNfVO4kLCgCRlmLuHcqQfV0E43/maLTS4axUbLlQzkRHBVS?=
 =?us-ascii?Q?i9699xytu4T6gifFUceY3KDdR+ZyKt+L/ix7spW1EvuIWfp5+utiRU5FOsIi?=
 =?us-ascii?Q?zPfwAYwgGqA7s7rl9/oBrAM2EB/gV7xj3WxZODr5lHzlMV05YDP1hpLu+EAm?=
 =?us-ascii?Q?mp3ikvvgJHC4k9+FW8ahBWj3XbFOgQO7f4Vvj5aKm7rZPPzxge+B8Um2dFtI?=
 =?us-ascii?Q?v3BHLgY7jSRfjgXRVE4xnz9RpwRTO0HbIhTZSklPWFJ5Eikp+mrurhAnpzsZ?=
 =?us-ascii?Q?TeWzxvmTx1WxRTUrxnVTrmo8ktWmnjqzyhQCJK71p69h1Q5fvJzzaDYUb/3d?=
 =?us-ascii?Q?7JfZahT53CofDI9IxbF5j8rEp8SgjpgtcsqFYPgFFpc6aYTVXfEsL5hvVMdq?=
 =?us-ascii?Q?+u5Hc5PvhvGeAwfis2UnhJQGqZuQcQc20y19TfW6KBNFaO36Dtj8GO2SMb1X?=
 =?us-ascii?Q?gwu5KYVTGN46gsC1MxqaFFbzZ//bAkS1WXhNCoJrCzyATW/Pd28JjPrdcbt/?=
 =?us-ascii?Q?2Pow1elH0vQrPg7Tl8FrafaI6xH2XlFi+sC8h/DScmFSSvDmqddA7NPtoue4?=
 =?us-ascii?Q?ytTitK5U4TRp+RCzGD8LpVos8su0MxYJPdvB6jTPQEMWbQ+/40viJWd4s8Bt?=
 =?us-ascii?Q?vsBGRRwCu35u87bFowTEs1GQcbl17vkgoAo7TSHA9v9XeWbuXFZncbnY6+e5?=
 =?us-ascii?Q?Lcm5rcukKYC53i37PAFpSC8rlDdizJYBlmJfppfA1PdpcpTztDfjbh20QPe+?=
 =?us-ascii?Q?MjRd8zKI136SwDGITAujVv4mDPkUpWCErOre6O1wk+yvax71R9ydTNSdv4hZ?=
 =?us-ascii?Q?GF1ofP5OEpqFJzt/ymzaZutQM25cwEaUXjO8ILyTFUty2qNUDGhhOxGrpU5c?=
 =?us-ascii?Q?aZ0noIxx/2h76l8L6KcFsG17Ob16uEdG95WkGVmD?=
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-AuthSource: PH0PR11MB5000.namprd11.prod.outlook.com
X-MS-Exchange-CrossTenant-Network-Message-Id: fda9a691-2e36-4d0c-a6ca-08dc500da3b2
X-MS-Exchange-CrossTenant-originalarrivaltime: 29 Mar 2024 16:31:09.0576
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 46c98d88-e344-4ed4-8496-4ed7712e255d
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: NpX3VPRxzlTECe2We7zr3McT0IGP0YbuFv42O+CJQ1W6wfDGYHTnAR/KJgV0dJyNaoRlCQxueGJJSqS5snyC+7GxbAZPLjzphM0mtXMoByo=
X-MS-Exchange-Transport-CrossTenantHeadersStamped: IA0PR11MB8377
X-OriginatorOrg: intel.com
List-Id: <pgsql-hackers.lists.postgresql.org>
List-Help: <https://lists.postgresql.org/manage/>
List-Subscribe: <https://lists.postgresql.org/manage/>
List-Post: <mailto:pgsql-hackers@lists.postgresql.org>
List-Owner: <mailto:pgsql-hackers-owner@lists.postgresql.org>
List-Archive: <https://www.postgresql.org/list/pgsql-hackers>
Archived-At: <https://www.postgresql.org/message-id/PH0PR11MB50001026ADCC5D9C83C8F9C9F23A2%40PH0PR11MB5000.namprd11.prod.outlook.com>
Precedence: bulk

> From: Nathan Bossart <nathandbossart@gmail.com>=20
> Sent: Friday, March 29, 2024 9:17 AM
> To: Amonson, Paul D <paul.d.amonson@intel.com>

> On Fri, Mar 29, 2024 at 04:06:17PM +0000, Amonson, Paul D wrote:
>> Yeah, I understand that much, but I want to know how portable the=20
>> XGETBV instruction is.  Unless I can assume that all x86_64 systems=20
>> and compilers support that instruction, we might need an additional=20
>> configure check and/or CPUID check.  It looks like MSVC has had=20
>> support for the _xgetbv intrinsic for quite a while, but I'm still resea=
rching the other cases.
>=20
> I see google web references to the xgetbv instruction as far back as=20
> 2009 for Intel 64 bit HW and 2010 for AMD 64bit HW, maybe you could=20
> test for
> _xgetbv() MSVC built-in. How far back do you need to go?

> Hm.  It seems unlikely that a compiler would understand AVX512 intrinsics=
 and not XGETBV then.  I guess the other question is whether CPUID indicati=
ng AVX512 is enabled implies the availability of XGETBV on the CPU.
> If that's not safe, we might need to add another CPUID test.

> It would probably be easy enough to add a couple of tests for this, but i=
f we don't have reason to believe there's any practical case to do so, I do=
n't know why we would.  I'm curious what others think about this.

This seems unlikely. Machines supporting XGETBV would support AVX512 intrin=
sics. Xgetbv instruction seems to be part of xsave feature set as per intel=
 developer manual [2]. XGETBV/XSAVE came first, and seems to be available i=
n all x86 systems available since 2011, since Intel SandyBridge architectur=
e and AMD the Opteron Gen4 [0].
AVX512 first came into a product in 2016 [1]
[0]: https://kb.vmware.com/s/article/1005764
[1]: https://en.wikipedia.org/wiki/AVX-512
[2]: https://cdrdv2-public.intel.com/774475/252046-sdm-change-document.pdf

- Akash Shankaran