[Bug middle-end/78468] [7 regression] libgomp.c/reduction-10.c and many more FAIL

["ebotcazou at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78468

--- Comment #24 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
The root cause of this mess is actually init_emit:

  REGNO_POINTER_ALIGN (VIRTUAL_INCOMING_ARGS_REGNUM) = STACK_BOUNDARY;
  REGNO_POINTER_ALIGN (VIRTUAL_STACK_VARS_REGNUM) = STACK_BOUNDARY;
  REGNO_POINTER_ALIGN (VIRTUAL_STACK_DYNAMIC_REGNUM) = STACK_BOUNDARY;
  REGNO_POINTER_ALIGN (VIRTUAL_OUTGOING_ARGS_REGNUM) = STACK_BOUNDARY;

3 out of 4 are wrong for 32-bit SPARC.  VIRTUAL_STACK_DYNAMIC_REGNUM is fixable
but not VIRTUAL_INCOMING_ARGS_REGNUM & VIRTUAL_OUTGOING_ARGS_REGNUM, whose
offset is defined by the ABI.

The code in init_emit looks wrong to me, as STACK_BOUNDARY is documented as:

 -- Macro: STACK_BOUNDARY
     Define this macro to the minimum alignment enforced by hardware
     for the stack pointer on this machine.  The definition is a C
     expression for the desired alignment (measured in bits).  This
     value is used as a default if `PREFERRED_STACK_BOUNDARY' is not
     defined.  On most machines, this should be the same as
     `PARM_BOUNDARY'.

but it goes back to 1995; I guess nobody cared about it until Dominik's patch.