Date: Mon, 31 May 2010 11:34:50 +0300 From: Kostik Belousov <kostikbel@gmail.com> To: d@delphij.net Cc: FreeBSD Current <freebsd-current@freebsd.org> Subject: Re: AES NI vs BIOS settings Message-ID: <20100531083450.GQ83316@deviant.kiev.zoral.com.ua> In-Reply-To: <4C034F5B.10806@delphij.net> References: <4C034F5B.10806@delphij.net>
next in thread | previous in thread | raw e-mail | index | archive | help
--mhFomgJQ1VTZ9Fr0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Sun, May 30, 2010 at 10:55:39PM -0700, Xin LI wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA256 >=20 > Hi, >=20 > I just found that if I disable "AES NI" in BIOS setting, FreeBSD would > be able to detect it on boot with: >=20 > Features2=3D0x29ee3ff<SSE3,PCLMULQDQ,DTES64,MON,DS_CPL,VMX,SMX,EST,TM2,SS= SE3,CX16,xTPR,PDCM,<b17>,DCA,SSE4.1,SSE4.2,POPCNT,AESNI> >=20 > However if it's set to "enabled" I got: >=20 > Features2=3D0x9ee3fd<SSE3,DTES64,MON,DS_CPL,VMX,SMX,EST,TM2,SSSE3,CX16,xT= PR,PDCM,<b17>,DCA,SSE4.1,SSE4.2,POPCNT> >=20 > The CPU was Xeon L5630 and motherboard is Supermicro X8STi with BIOS > 1.00c. Should I consider this a BIOS issue with known workaround that I > consider "Disable" as "Enabled"? :) On the only machine with AESNI-capable Core i5 I have access to thanks to Sentex Communications, AESNI bit is reported as 1 and aesni instructions do work. You could try to actually use aesni and see what is broken. I failed to find a magic bit to enable/disable AESNI in the MSRs. --mhFomgJQ1VTZ9Fr0 Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (FreeBSD) iEYEARECAAYFAkwDdKkACgkQC3+MBN1Mb4jsaQCgjk7FlSDrVmiEL2G1G0VS0pd7 bWcAoIyj/7Y2qIuMZfrx+6KBNfsLoBP0 =2ukd -----END PGP SIGNATURE----- --mhFomgJQ1VTZ9Fr0--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100531083450.GQ83316>