Matching and testing against smulhsm3

[m]

On 2020-02-07 16:44, Segher Boessenkool wrote:
> Hi!
>
> On Fri, Feb 07, 2020 at 03:41:25PM +0100, m wrote:
>> ...so I tried to write a corresponding matching pattern, like so:
>>
>>  Â  (define_insn "smulhshi3"
>>  Â Â Â  [(set (match_operand:HI 0 "register_operand" "=r")
>>  Â Â Â Â Â  (truncate:HI
>>  Â Â Â Â Â Â Â  (ashiftrt:SI
>>  Â Â Â Â Â Â Â Â Â  (mult:SI
>>  Â Â Â Â Â Â Â Â Â Â Â  (sign_extend:SI (match_operand:HI 1
>> "register_operand" "r"))
>>  Â Â Â Â Â Â Â Â Â Â Â  (sign_extend:SI (match_operand:HI 2
>> "register_operand" "r")))
>>  Â Â Â Â Â Â Â Â Â  (const_int 15))))]
>>  Â  "TARGET_PACKED_OPS"
>>  Â  "mulq.h\\t%0, %1, %2")
>>
>> However, I am unable to trigger this code path. I have tried with the
>> following C code:
>>
>> short mulq(short op1, short op2) {
>>  Â  return (short) (((int) op1 * (int) op2) >> (32 / 2 - 1));
>> }
>>
>> But I just get the regular 4-instruction result (2x sign extend, 1x mul,
>> 1x shift).
> What does -fdump-rtl-combine-all show it tried?  *Did* it try anything?
>
>
> Segher


Cool option. I'm not really sure how to read the output tough. The 
closest it seems to try to match is this:


Failed to match this instruction:
(set (reg:SI 85)
 Â Â Â  (ashiftrt:SI (mult:SI (sign_extend:SI (subreg:HI (reg:SI 86) 0))
 Â Â Â Â Â Â Â Â Â Â Â  (reg:SI 83 [ op2D.1381 ]))
 Â Â Â Â Â Â Â  (const_int 15 [0xf])))


It seems that it has already decided to split the instruction into 
several operations (the truncate operation is not there, and the second 
sign_extend:SI (subreg:HI ...) is also missing).


/Marcus