Hi,

Marius is not working anymore on gmp, thus I've picked up his v2
patches, rebased them and also integrated some changes which he has not
yet posted to the mailing-list:
[PATCH v2] Add addmul_1, addmul_2, and mul_basecase for IBM z13 and later
(https://gmplib.org/list-archives/gmp-devel/2021-August/006005.html)

Changes from v2:
- Add vectorized addmul_1 / mul_1 for IBM z13:
  - Fixing assert.
  - Rename common-vec.h macros from my_* to s390_*.
- Add addmul_2 / mul_2 for IBM z13 and later:
  - Adjust comment-style.
  - Rename common-vec.h macros from my_* to s390_*.
- s390_64: Add mul_basecase for IBM z13 and later:
  - Adjust comment.

The code passes "make check" and now also when configured with --enable-assert.

Of course, before posting this v3, Marius allowed me to pickup his v2 patches
as well as his not yet posted fixes. Therefore I've added the
"Co-authored-by Stefan Liebler <s...@linux.ibm.com>".

Bye,
Stefan

Marius Hillenbrand (4):
  Add vectorized addmul_1 / mul_1 for IBM z13
  Add addmul_2 / mul_2 for IBM z13 and later
  s390_64: Add mul_basecase for IBM z13 and later
  s390_64: Add tuneup results for z13

 mpn/s390_64/z13/addmul_1.c     | 358 +++++++++++++++++++++++++
 mpn/s390_64/z13/aormul_2.c     | 476 +++++++++++++++++++++++++++++++++
 mpn/s390_64/z13/common-vec.h   | 175 ++++++++++++
 mpn/s390_64/z13/gmp-mparam.h   | 162 +++++++++++
 mpn/s390_64/z13/mul_1.c        |  31 +++
 mpn/s390_64/z13/mul_basecase.c | 124 +++++++++
 6 files changed, 1326 insertions(+)
 create mode 100644 mpn/s390_64/z13/addmul_1.c
 create mode 100644 mpn/s390_64/z13/aormul_2.c
 create mode 100644 mpn/s390_64/z13/common-vec.h
 create mode 100644 mpn/s390_64/z13/gmp-mparam.h
 create mode 100644 mpn/s390_64/z13/mul_1.c
 create mode 100644 mpn/s390_64/z13/mul_basecase.c

-- 
2.40.1

_______________________________________________
gmp-devel mailing list
gmp-devel@gmplib.org
https://gmplib.org/mailman/listinfo/gmp-devel

Reply via email to