Date: Wed, 23 Dec 2015 13:29:29 -0800 From: John Baldwin <jhb@freebsd.org> To: cem@freebsd.org Cc: Ravi Pokala <rpokala@mac.com>, src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: Re: svn commit: r292528 - head/sys/x86/include Message-ID: <19132903.CqPZ5E7ceZ@ralph.baldwin.cx> In-Reply-To: <CAG6CVpVrJGTZrTZi%2By=FU32ne5uzB-z_LhhTLj=mfhbsLETG6Q@mail.gmail.com> References: <201512210442.tBL4gw6T007629@repo.freebsd.org> <9D7D1132-ED42-4F37-BEFD-AD62FB6321C7@panasas.com> <CAG6CVpVrJGTZrTZi%2By=FU32ne5uzB-z_LhhTLj=mfhbsLETG6Q@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Monday, December 21, 2015 01:41:35 PM Conrad Meyer wrote:
> No, I didn't forget. AVX512VBMI is a flag in ECX
> ("cpu_stdext_feature2") rather than EBX ("cpu_stdext_feature"), and
> there aren't any macros for any of the ECX flags yet (PREFETCHWT1,
> AVX512VBMI, PKU, and OSPKE).
Might as well start on them? I would be tempted to just use CPU_STDEXT_foo for those
as well. The standard ext leaf allows for multiple "banks" of feature registers, so
I'm not sure CPU_STDEXT2_*, etc. will scale all that well.
--
John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19132903.CqPZ5E7ceZ>