[PATCH] Add optimized addmul_1 and submul_1 for IBM z13

Marius Hillenbrand mhillen at linux.ibm.com
Fri Mar 5 14:46:28 UTC 2021


Hi Torbjörn,

A few pointers of how PoP, available instructions, and actual Linux
systems fit together, before I dive into your other comments:

The PoP lists ISA extensions, called "facilities", in Chapter 1. They
are called "facilities". Linux uses the term "features". Practically,

$ grep features /proc/cpuinfo

will provide the list of "features" that your combination of hardware
and Linux kernel actually supports.

Programmatically, getauxval(AT_HWCAP) provides a bitmap of supported
features (e.g., as we did in
https://github.ibm.com/ztoolchain/OpenBLAS/blob/develop/driver/others/dynamic_zarch.c).

The translation between features and facilities is not obvious,
unfortunately, so here's a (incomplete) table:

z13: introduced the vector extensions
  PoP: Vector Facility for z/Architecture
  Linux: vx / HWCAP_S390_VX

z14:
  Vector-Enhancements Facility 1
  vxe / HWCAP_S390_VXE

z15:
  Vector-Enhancements Facility 2 (adds VLERG and VSTERG, among others)
  vxe2 / HWCAP_S390_VXRS_EXT2

Marius
-- 
Marius Hillenbrand
Linux on Z development
IBM Deutschland Research & Development GmbH
Vors. des Aufsichtsrats: Gregor Pillen / Geschäftsführung: Dirk Wittkopp
Sitz der Gesellschaft: Böblingen / Registergericht: Amtsgericht
Stuttgart, HRB 243294


More information about the gmp-devel mailing list