Re: RFC: implement typed DWARF stack

["Ulrich Weigand" <uweigand@de.ibm.com>]

Tom Tromey wrote:
> >>>>> "Ulrich" == Ulrich Weigand <uweigand@de.ibm.com> writes:
> 
> Tom> Yes.  My understanding is that for new-style typed values, DW_OP_shr and
> Tom> DW_OP_shra are actually the same -- the type indicates the operation to
> Tom> perform.  But, for old-style values, we must cast to unsigned for
> Tom> DW_OP_shr.
> 
> Ulrich> I see.  However, the code as implemented casts *all* signed values to
> Ulrich> unsigned for DW_OP_shr, not just old-style values.  That's what got
> Ulrich> me confused ...
> 
> I mentioned this on the GCC list and Jakub said that GCC actually emits
> code assuming that the operation will always be unsigned.
> 
> So, I am reverting my change here.  Patch appended.

So just to clarify: in the discussion a while back, you said:

> Ulrich> B.t.w. your patch always performs an unsigned shift for
> Ulrich> DW_OP_shr, even on signed operands.  However, for DW_OP_shra,
> Ulrich> your patch respects the sign of the operands and might actually
> Ulrich> perform an unsigned shift (even though the opcode explicitly
> Ulrich> says "arithmetic right shift" ...)  This looks like another of
> Ulrich> those signed/unsigned inconsistencies with the proposal to me.
> 
> Yes.  My understanding is that for new-style typed values, DW_OP_shr and
> DW_OP_shra are actually the same -- the type indicates the operation to
> perform.  But, for old-style values, we must cast to unsigned for
> DW_OP_shr.

With this latest patch, it is now definitely *not* the case that DW_OP_shr
and DW_OP_shra behave the same on new-style typed values.  Instead, as I
pointed out originally, DW_OP_shr now always performs an unsigned operation,
while DW_OP_shra respects the value's type ...

Is that really what was intended?  Or should rather DW_OP_shra now also
be changed (to always perform a signed operation as its name suggests)?

Bye,
Ulrich

-- 
  Dr. Ulrich Weigand
  GNU Toolchain for Linux on System z and Cell BE
  Ulrich.Weigand@de.ibm.com