Date: Thu, 30 Mar 2017 18:56:57 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-bugs@FreeBSD.org Subject: [Bug 218203] Implement AVX2 accelerated Fletcher algorithms Message-ID: <bug-218203-8-jNBMHyAXQG@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-218203-8@https.bugs.freebsd.org/bugzilla/> References: <bug-218203-8@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D218203 --- Comment #3 from kungfujesus06@gmail.com --- Created attachment 181322 --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=3D181322&action= =3Dedit Quick benchmark Please don't mind the C++, the actual function is clean C. Be sure to comp= ile with -mavx2. It's worth noting that the code that's integrated into ZFS On Linux appears= to do benchmark code to select the fastest function and dynamically reassigns a function pointer to that function (at maybe module load time?). Adding the extra layer of function pointers will probably be somewhat necessary for th= is kind of modularity, though this obviously has some security implications (in that corrupting this pointer value can lead to bad things). I hardly think it'd be the first function pointer dispatched in the FreeBSD kernel, though. Given that IXSystems cares about ZFS performance and distributes on hardware with some lower powered Atom CPUs, I figured this would be of interest. --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-218203-8-jNBMHyAXQG>