Re: [PATCH PR67921]Use sizetype for CHREC_RIGHT when building pointer type CHREC

[Richard Biener <richard.guenther@gmail.com>]

On Wed, Oct 21, 2015 at 11:39 AM, Bin.Cheng <amker.cheng@gmail.com> wrote:
> On Wed, Oct 21, 2015 at 5:15 PM, Richard Biener
> <richard.guenther@gmail.com> wrote:
>> On Wed, Oct 21, 2015 at 6:46 AM, Bin Cheng <bin.cheng@arm.com> wrote:
>>> Hi,
>>> As analyzed in PR67921, I think the issue is caused by fold_binary_loc which
>>> folds:
>>>   4 - (sizetype) &c - (sizetype) ((int *) p1_8(D) + ((sizetype) a_23 * 24 +
>>> 4))
>>> into below form:
>>>   ((sizetype) -((int *) p1_8(D) + ((sizetype) a_23 * 24 + 4)) - (sizetype)
>>> &c) + 4
>>>
>>> Look the minus sizetype expression is folded as negative pointer expression,
>>> which seems incorrect.  Apart from this, The direct reason of this ICE is in
>>> CHREC because of an overlook.  In general CHREC supports NEGATE_EXPR for
>>> CHREC, the only problem is it uses pointer type for CHREC_RIGHT, rather than
>>> sizetype, when building pointer type CHREC.
>>>
>>> This simple patch fixes the ICE issue.  Bootstrap and test on x86 & x86_64.
>>>
>>> Is it OK?
>>
>> Hmm, I think not - we shouldn't ever get pointer typed
>> multiplications.  Did you track
>> down which is the bogus fold transform (I agree the result above is
>> bogus)?  It's
>> probably related to STRIP_NOPS stripping sizetype conversions from pointers
>> so we might get split_tree to build such negate.  Note that split_tree strips
>> (sign!) nops itself and thus should probably simply receive op0 and op1 instead
>> of arg0 and arg1.
> Yes, I was going to send similar patch for fold stuff.  Just thought
> it might be useful to support POINTER chrec in *_multiply.  I will
> drop this and let you test yours.

The patch regresses

Running target unix/
FAIL: gcc.dg/pr52578.c scan-tree-dump-times original "return 2;" 2
FAIL: gcc.dg/strict-overflow-5.c scan-tree-dump optimized "return 3"
FAIL: gcc.dg/tree-ssa/loop-15.c scan-tree-dump-times optimized "\\\\+" 0
FAIL: gcc.dg/tree-ssa/loop-15.c scan-tree-dump-times optimized "n_. \\\\* n_." 1
FAIL: gcc.dg/vect/vect-align-3.c -flto -ffat-lto-objects  scan-tree-dump-not vec
t "vect_do_peeling_for_loop_bound"
FAIL: gcc.dg/vect/vect-align-3.c scan-tree-dump-not vect "vect_do_peeling_for_lo
op_bound"

and for -m32

FAIL: gcc.target/i386/pr67985-3.c scan-assembler movd[ \\t]%xmm[0-7], %eax
FAIL: gcc.target/i386/pr67985-3.c scan-assembler mulss

as well.  So it requires some more complicated handling.  It might need to
take the negation code in split-tree to be performed only implicitely during
associate_trees (then on the correct type).  But that requires (quite) some
refactoring.  I suggest a struct tree_with_negate { tree t; bool negate_p; }
to pass this along.

Not for me at this point (looks I have quite a few patches to review ;))

Richard.

> Thanks,
> bin
>>
>> I'm testing
>>
>> @@ -9505,8 +9523,8 @@ fold_binary_loc (location_t loc,
>>              then the result with variables.  This increases the chances of
>>              literals being recombined later and of generating relocatable
>>              expressions for the sum of a constant and literal.  */
>> -         var0 = split_tree (arg0, code, &con0, &lit0, &minus_lit0, 0);
>> -         var1 = split_tree (arg1, code, &con1, &lit1, &minus_lit1,
>> +         var0 = split_tree (op0, code, &con0, &lit0, &minus_lit0, 0);
>> +         var1 = split_tree (op1, code, &con1, &lit1, &minus_lit1,
>>                              code == MINUS_EXPR);
>>
>>           /* Recombine MINUS_EXPR operands by using PLUS_EXPR.  */
>>
>> which fixes the testcase for me.
>>
>> Richard.
>>
>>> Note, I do think the associate logic in fold_binary_loc needs fix, but that
>>> should be another patch.
>>>
>>>
>>> 2015-10-20  Bin Cheng  <bin.cheng@arm.com>
>>>
>>>         PR tree-optimization/67921
>>>         * tree-chrec.c (chrec_fold_multiply): Use sizetype for CHREC_RIGHT
>>> if
>>>         type is pointer type.
>>>
>>> 2015-10-20  Bin Cheng  <bin.cheng@arm.com>
>>>
>>>         PR tree-optimization/67921
>>>         * gcc.dg/ubsan/pr67921.c: New test.