https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87054
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
Last reconfirmed| |2018-08-22
CC| |rguenth at gcc dot gnu.org
Ever confirmed|0 |1
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
there's no member 'unpacked'. With chaning 'c' to unpacked and using "+x" on
x86_64 I see
_2 = &b[i].unpacked;
_3 = *_2;
__asm__("" : "=x" *_2 : "0" _3);
which is probably what you refer to.
Can you test the following? The code indeed seems to be overly simplistic...
diff --git a/gcc/gimplify.c b/gcc/gimplify.c
index e35137aec2c..117e5fcedc9 100644
--- a/gcc/gimplify.c
+++ b/gcc/gimplify.c
@@ -12540,9 +12540,15 @@ gimplify_expr (tree *expr_p, gimple_seq *pre_p,
gimple_seq *post_p,
/* An lvalue will do. Take the address of the expression, store it
in a temporary, and replace the expression with an INDIRECT_REF of
that temporary. */
+ tree ref_alias_type = reference_alias_ptr_type (*expr_p);
+ unsigned int ref_align = get_object_alignment (*expr_p);
+ tree ref_type = TREE_TYPE (*expr_p);
tmp = build_fold_addr_expr_loc (input_location, *expr_p);
gimplify_expr (&tmp, pre_p, post_p, is_gimple_reg, fb_rvalue);
- *expr_p = build_simple_mem_ref (tmp);
+ if (TYPE_ALIGN (ref_type) != ref_align)
+ ref_type = build_aligned_type (ref_type, ref_align);
+ *expr_p = build2 (MEM_REF, ref_type,
+ tmp, build_zero_cst (ref_alias_type));
}
else if ((fallback & fb_rvalue) && is_gimple_reg_rhs_or_call (*expr_p))
{
