On 10/19/22 13:41, Jakub Jelinek wrote:
> On Wed, Oct 19, 2022 at 07:39:05PM +0200, Jakub Jelinek via Gcc-patches wrote:
>> On Wed, Oct 19, 2022 at 12:55:12PM -0400, Andrew MacLeod via Gcc-patches wrote:
>>>> Not sure I understand this part.  op is whatever we pass as the ith
>>>> argument to IFN_ASSUME.  I'd expect that at this point one needs to
>>>> remap that to the (i-1)th PARM_DECL of assume_id (so e.g. when you
>>>> have the above loop you could as well start with DECL_ARGUMENTS and move
>>>> that to DECL_CHAIN at the end of every iteration.  And then
>>>> query ssa_default_def (DECL_STRUCT_FUNCTION (assume_id), parm)
>>>> in each case and get global range of what that returns.
>>> OK, this is the bit of code I dont know how to write :-)
>>>
>>> yes, op is the name of the value within this current function, and yes, that
>>> needs to be mapped to the argument decl in the assume function.   Then we
>>> need to query what range was given to that name during the assume pass.
>>> when that is returned, the add_range (op, range) will inject it as a side
>>> effect.
>>>
>>> Can you write that loop?
>> I meant something like (untested code):
>>         && gimple_call_internal_fn (s) == IFN_ASSUME)
>>       {
>>         tree assume_id = gimple_call_arg (s, 0);
>> -      for (unsigned i = 1; i < gimple_call_num_args (s); i++)
>> +      tree parm = DECL_ARGUMENTS (assume_id);
>> +      struct function *fun = DECL_STRUCT_FUNCTION (assume_id);
>> +      for (unsigned i = 1;
>> +	   i < gimple_call_num_args (s) && parm;
>> +	   i++, parm = DECL_CHAIN (parm))
>>   	{
>>   	  tree op = gimple_call_arg (s, i);
>>   	  tree type = TREE_TYPE (op);
>> +	  tree arg = ssa_default_def (fun, parm);
>> +	  if (arg == NULL_TREE)
>> +	    continue;
>>   	  if (gimple_range_ssa_p (op) && Value_Range::supports_type_p (type))
>>   	    {
>>   	      Value_Range assume_range (type);
>> and querying SSA_NAME_RANGE_INFO of arg rather than op.

Thanks, I had actually come to most of those conclusions already, except 
for the DECL_STRUCT_FUNCTION bit.. I was stuck on that :-)


So the SSA_NAMEs for default_defs are never disposed of?  so I can query 
them even though they are not in the current function decl?  huh. I did 
not know that.


OK, attached is the latest set of patches.  not bootstrapped or 
anything, but very interesting.


from.assumption  pass:

;; Function _Z3bari._assume.0 (_Z3bari._assume.0, funcdef_no=5, 
decl_uid=2184, cgraph_uid=7, symbol_order=7)

Assumptions :
--------------
x_1(D) -> [irange] int [42, 42] NONZERO 0x2a

__attribute__((no_icf, noclone, noinline, noipa))
bool _Z3bari._assume.0 (int x)
{
   bool _2;

;;   basic block 2, loop depth 0
;;    pred:       ENTRY
   _2 = x_1(D) == 42;
   return _2;
;;    succ:       EXIT

}


and in the VRP2 pass, I see:

_Z3bari._assume.0 assume inferred range of x_1(D) (param x) = [irange] 
int [42, 42] NONZERO 0x2a
int bar (int x)
{
   <bb 2> [local count: 1073741824]:
   .ASSUME (_Z3bari._assume.0, x_1(D));
   return 42;

}


Huh.  lookit that.....


Anyway, let me clean this up a bit more, but so far so good.

I'll try bootstrapping and  such

Andrew