> -----Original Message-----
> From: Richard Biener <richard.guent...@gmail.com>
> Sent: Friday, November 29, 2024 8:57 AM
> To: Tamar Christina <tamar.christ...@arm.com>
> Cc: gcc-patches@gcc.gnu.org; nd <n...@arm.com>; rguent...@suse.de;
> j...@ventanamicro.com
> Subject: Re: [PATCH 1/2]middle-end: refactor type to be explicit in
> operand_equal_p [PR114932]
>
> On Tue, Aug 20, 2024 at 3:07 PM Tamar Christina <tamar.christ...@arm.com>
> wrote:
> >
> > Hi All,
> >
> > This is a refactoring with no expected behavioral change.
> > The goal with this is to make the type of the expressions being used
> > explicit.
> >
> > I did not change all the recursive calls to operand_equal_p () to recurse
> > directly to the new function but instead this goes through the top level
> > call
> > which re-extracts the types.
> >
> > This was done because in most of the cases where we recurse type == arg.
> > The second patch makes use of this new flexibility to implement an overload
> > of operand_equal_p which checks for equality under two's complement.
> >
> > Bootstrapped Regtested on aarch64-none-linux-gnu,
> > arm-none-linux-gnueabihf, x86_64-pc-linux-gnu
> > -m32, -m64 and no issues.
> >
> > Ok for master?
> >
> > Thanks,
> > Tamar
> >
> > gcc/ChangeLog:
> >
> > PR tree-optimization/114932
> > * fold-const.cc (operand_compare::operand_equal_p): Split into one
> > that
> > takes explicit type parameters and use that in public one.
> > * fold-const.h (class operand_compare): Add operand_equal_p private
> > overload.
> >
> > ---
> > diff --git a/gcc/fold-const.h b/gcc/fold-const.h
> > index
> b82ef137e2f2096f86c20df3c7749747e604177e..878545b1148b839e8a8e866f
> 38e31161f0d116c8 100644
> > --- a/gcc/fold-const.h
> > +++ b/gcc/fold-const.h
> > @@ -273,6 +273,12 @@ protected:
> > true is returned. Then RET is set to corresponding comparsion
> > result. */
> > bool verify_hash_value (const_tree arg0, const_tree arg1, unsigned int
> > flags,
> > bool *ret);
> > +
> > +private:
> > + /* Return true if two operands are equal. The flags fields can be used
> > + to specify OEP flags described in tree-core.h. */
> > + bool operand_equal_p (tree, const_tree, tree, const_tree,
> > + unsigned int flags);
> > };
> >
> > #endif // GCC_FOLD_CONST_H
> > diff --git a/gcc/fold-const.cc b/gcc/fold-const.cc
> > index
> 8908e7381e72cbbf4a8fd96f18cbf4436aba8441..71e82b1d76d4106c7c23c54af
> 8b35905a1af9f1c 100644
> > --- a/gcc/fold-const.cc
> > +++ b/gcc/fold-const.cc
> > @@ -3156,6 +3156,17 @@ combine_comparisons (location_t loc,
> > bool
> > operand_compare::operand_equal_p (const_tree arg0, const_tree arg1,
> > unsigned int flags)
> > +{
> > + return operand_equal_p (TREE_TYPE (arg0), arg0, TREE_TYPE (arg1), arg1,
> flags);
> > +}
> > +
> > +/* The same as operand_equal_p however the type of ARG0 and ARG1 are
> assumed to be
> > + the TYPE0 and TYPE1 respectively. */
> > +
> > +bool
> > +operand_compare::operand_equal_p (tree type0, const_tree arg0,
> > + tree type1, const_tree arg1,
>
> did you try using const_tree for type0/type1?
>
I did, but types_compatible_p is non-const and it calls
useless_type_conversion_p
which is also non-const. Having a look I don't either of those function changes
type so I could change them all to const_tree if you'd like and see what shakes
out.
It looks like all the calls done in useless_type_conversion_p are already
const_tree.
Do you want me to propagate the const_tree down?
Thanks,
Tamar
> > + unsigned int flags)
> > {
> > bool r;
> > if (verify_hash_value (arg0, arg1, flags, &r))
> > @@ -3166,25 +3177,25 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> >
> > /* If either is ERROR_MARK, they aren't equal. */
> > if (TREE_CODE (arg0) == ERROR_MARK || TREE_CODE (arg1) == ERROR_MARK
> > - || TREE_TYPE (arg0) == error_mark_node
> > - || TREE_TYPE (arg1) == error_mark_node)
> > + || type0 == error_mark_node
> > + || type1 == error_mark_node)
> > return false;
> >
> > /* Similar, if either does not have a type (like a template id),
> > they aren't equal. */
> > - if (!TREE_TYPE (arg0) || !TREE_TYPE (arg1))
> > + if (!type0 || !type1)
> > return false;
> >
> > /* Bitwise identity makes no sense if the values have different layouts.
> > */
> > if ((flags & OEP_BITWISE)
> > - && !tree_nop_conversion_p (TREE_TYPE (arg0), TREE_TYPE (arg1)))
> > + && !tree_nop_conversion_p (type0, type1))
> > return false;
> >
> > /* We cannot consider pointers to different address space equal. */
> > - if (POINTER_TYPE_P (TREE_TYPE (arg0))
> > - && POINTER_TYPE_P (TREE_TYPE (arg1))
> > - && (TYPE_ADDR_SPACE (TREE_TYPE (TREE_TYPE (arg0)))
> > - != TYPE_ADDR_SPACE (TREE_TYPE (TREE_TYPE (arg1)))))
> > + if (POINTER_TYPE_P (type0)
> > + && POINTER_TYPE_P (type1)
> > + && (TYPE_ADDR_SPACE (TREE_TYPE (type0))
> > + != TYPE_ADDR_SPACE (TREE_TYPE (type1))))
> > return false;
> >
> > /* Check equality of integer constants before bailing out due to
> > @@ -3211,12 +3222,15 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> >
> > /* If both types don't have the same precision, then it is not safe
> > to strip NOPs. */
> > - if (element_precision (TREE_TYPE (arg0))
> > - != element_precision (TREE_TYPE (arg1)))
> > + if (element_precision (type0)
> > + != element_precision (type1))
> > return false;
> >
> > STRIP_NOPS (arg0);
> > STRIP_NOPS (arg1);
> > +
> > + type0 = TREE_TYPE (arg0);
> > + type1 = TREE_TYPE (arg1);
> > }
> > #if 0
> > /* FIXME: Fortran FE currently produce ADDR_EXPR of NOP_EXPR. Enable the
> > @@ -3275,9 +3289,9 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> >
> > /* When not checking adddresses, this is needed for conversions and for
> > COMPONENT_REF. Might as well play it safe and always test this. */
> > - if (TREE_CODE (TREE_TYPE (arg0)) == ERROR_MARK
> > - || TREE_CODE (TREE_TYPE (arg1)) == ERROR_MARK
> > - || (TYPE_MODE (TREE_TYPE (arg0)) != TYPE_MODE (TREE_TYPE (arg1))
> > + if (TREE_CODE (type0) == ERROR_MARK
> > + || TREE_CODE (type1) == ERROR_MARK
> > + || (TYPE_MODE (type0) != TYPE_MODE (type1)
> > && !(flags & OEP_ADDRESS_OF)))
> > return false;
> >
> > @@ -3364,8 +3378,8 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> > return true;
> >
> > /* See sem_variable::equals in ipa-icf for a similar approach. */
> > - tree typ0 = TREE_TYPE (arg0);
> > - tree typ1 = TREE_TYPE (arg1);
> > + tree typ0 = type0;
> > + tree typ1 = type1;
>
> I suppose you should instead eliminate typ0 and typ1.
>
> >
> > if (TREE_CODE (typ0) != TREE_CODE (typ1))
> > return false;
> > @@ -3444,8 +3458,8 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> > {
> > CASE_CONVERT:
> > case FIX_TRUNC_EXPR:
> > - if (TYPE_UNSIGNED (TREE_TYPE (arg0))
> > - != TYPE_UNSIGNED (TREE_TYPE (arg1)))
> > + if (TYPE_UNSIGNED (type0)
> > + != TYPE_UNSIGNED (type1))
>
> it seems this now fits one line
>
> > return false;
> > break;
> > default:
> > @@ -3481,12 +3495,12 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> > case INDIRECT_REF:
> > if (!(flags & OEP_ADDRESS_OF))
> > {
> > - if (TYPE_ALIGN (TREE_TYPE (arg0))
> > - != TYPE_ALIGN (TREE_TYPE (arg1)))
> > + if (TYPE_ALIGN (type0)
> > + != TYPE_ALIGN (type1))
>
> likewise (and maybe elsewhere).
>
> Otherwise this looks good. It's OK when 2/2 is approved or OK when doing this
> without adding a new parameter but adding locals initializing them
> from arg0/arg1
> as an intermediate step.
>
> Thanks,
> Richard.
>
> > return false;
> > /* Verify that the access types are compatible. */
> > - if (TYPE_MAIN_VARIANT (TREE_TYPE (arg0))
> > - != TYPE_MAIN_VARIANT (TREE_TYPE (arg1)))
> > + if (TYPE_MAIN_VARIANT (type0)
> > + != TYPE_MAIN_VARIANT (type1))
> > return false;
> > }
> > flags &= ~OEP_ADDRESS_OF;
> > @@ -3494,8 +3508,8 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> >
> > case IMAGPART_EXPR:
> > /* Require the same offset. */
> > - if (!operand_equal_p (TYPE_SIZE (TREE_TYPE (arg0)),
> > - TYPE_SIZE (TREE_TYPE (arg1)),
> > + if (!operand_equal_p (TYPE_SIZE (type0),
> > + TYPE_SIZE (type1),
> > flags & ~OEP_ADDRESS_OF))
> > return false;
> >
> > @@ -3509,15 +3523,15 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> > if (!(flags & OEP_ADDRESS_OF))
> > {
> > /* Require equal access sizes */
> > - if (TYPE_SIZE (TREE_TYPE (arg0)) != TYPE_SIZE (TREE_TYPE
> > (arg1))
> > - && (!TYPE_SIZE (TREE_TYPE (arg0))
> > - || !TYPE_SIZE (TREE_TYPE (arg1))
> > - || !operand_equal_p (TYPE_SIZE (TREE_TYPE (arg0)),
> > - TYPE_SIZE (TREE_TYPE (arg1)),
> > + if (TYPE_SIZE (type0) != TYPE_SIZE (type1)
> > + && (!TYPE_SIZE (type0)
> > + || !TYPE_SIZE (type1)
> > + || !operand_equal_p (TYPE_SIZE (type0),
> > + TYPE_SIZE (type1),
> > flags)))
> > return false;
> > /* Verify that access happens in similar types. */
> > - if (!types_compatible_p (TREE_TYPE (arg0), TREE_TYPE (arg1)))
> > + if (!types_compatible_p (type0, type1))
> > return false;
> > /* Verify that accesses are TBAA compatible. */
> > if (!alias_ptr_types_compatible_p
> > @@ -3529,8 +3543,8 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> > != MR_DEPENDENCE_BASE (arg1)))
> > return false;
> > /* Verify that alignment is compatible. */
> > - if (TYPE_ALIGN (TREE_TYPE (arg0))
> > - != TYPE_ALIGN (TREE_TYPE (arg1)))
> > + if (TYPE_ALIGN (type0)
> > + != TYPE_ALIGN (type1))
> > return false;
> > }
> > flags &= ~OEP_ADDRESS_OF;
> > @@ -3802,16 +3816,16 @@ operand_compare::operand_equal_p (const_tree
> arg0, const_tree arg1,
> > indexed in increasing order and form an initial sequence.
> >
> > We make no effort to compare nonconstant ones in GENERIC. */
> > - if (!VECTOR_TYPE_P (TREE_TYPE (arg0))
> > - || !VECTOR_TYPE_P (TREE_TYPE (arg1)))
> > + if (!VECTOR_TYPE_P (type0)
> > + || !VECTOR_TYPE_P (type1))
> > return false;
> >
> > /* Be sure that vectors constructed have the same representation.
> > We only tested element precision and modes to match.
> > Vectors may be BLKmode and thus also check that the number of
> > parts match. */
> > - if (maybe_ne (TYPE_VECTOR_SUBPARTS (TREE_TYPE (arg0)),
> > - TYPE_VECTOR_SUBPARTS (TREE_TYPE (arg1))))
> > + if (maybe_ne (TYPE_VECTOR_SUBPARTS (type0),
> > + TYPE_VECTOR_SUBPARTS (type1)))
> > return false;
> >
> > vec<constructor_elt, va_gc> *v0 = CONSTRUCTOR_ELTS (arg0);
> >
> >
> >
> >
> > --
