[Bug target/89213] New: Optimize V2DI shifts by a constant on power8 & above systems.

2019-02-05 Thread g...@the-meissners.org 
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89213

Bug ID: 89213
   Summary: Optimize V2DI shifts by a constant on power8 & above
systems.
   Product: gcc
   Version: 9.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: target
  Assignee: unassigned at gcc dot gnu.org
  Reporter: g...@the-meissners.org
  Target Milestone: ---

ISA 2.07 (i.e. -mcpu=power8) and above added support for doing various
operations on V2DI (i.e. vector long long) data types, including shifts.

If you generate code to shift a V2DI type by a constant, the compiler generates
sub-optimal code:

For example:

#include 

typedef vector long long vi64_t;
typedef vector unsigned long long vui64_t;

vi64_t
shiftra_test64 (vi64_t a)
{
  vui64_t x = {4, 4};
  return (vi64_t) vec_vsrad (a, x);
}

Generates:

shiftra_test64:
xxspltib 0,4
xxlor 32,0,0
vextsb2d 0,0
vsrad 2,2,0
blr

when it could generate:

shiftra_test64:
vspltisw 0,4
vsrad 2,2,0
blr

This is true of all 3 shift operations (shift left, logical shift right, and
arithmetic shift right).

[Bug target/89319] New: PowerPC generates poor code when using attribute((vector_size(32))

2019-02-12 Thread g...@the-meissners.org 
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89319

Bug ID: 89319
   Summary: PowerPC generates poor code when using
attribute((vector_size(32))
   Product: gcc
   Version: 9.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: target
  Assignee: unassigned at gcc dot gnu.org
  Reporter: g...@the-meissners.org
  Target Milestone: ---

Created attachment 45677
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=45677&action=edit
Source code to show the problem

I tried out some code using the vector_size(32) attribute on PowerPC, and it
generates poor code.  I expected that it would just generate two loads, two
adds, and two stores, but instead it pushed things to the stack.

See the attached code and asm file.

[Bug target/89319] PowerPC generates poor code when using attribute((vector_size(32))

2019-02-12 Thread g...@the-meissners.org 
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89319

--- Comment #1 from Michael Meissner  ---
Created attachment 45678
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=45678&action=edit
Power9 assembly code to show the poor code generation

This was the assembly code using -O3 -mcpu=power9