https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105690
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Component|c |ipa
Status|UNCONFIRMED |NEW
Blocks| |56456
Keywords| |wrong-debug
Ever confirmed|0 |1
Last reconfirmed| |2022-05-23
CC| |hubicka at gcc dot gnu.org,
| |marxin at gcc dot gnu.org
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
OK, so the diagnostic is misleading:
else if (vr && vr->kind () == VR_ANTI_RANGE)
{
if (up_bound
&& TREE_CODE (up_sub) == INTEGER_CST
&& (ignore_off_by_one
? tree_int_cst_lt (up_bound, up_sub)
: tree_int_cst_le (up_bound, up_sub))
&& TREE_CODE (low_sub) == INTEGER_CST
&& tree_int_cst_le (low_sub, low_bound))
warned = warning_at (location, OPT_Warray_bounds,
"array subscript [%E, %E] is outside "
"array bounds of %qT",
low_sub, up_sub, artype);
it should say
warning: array subscript ~[0, 19] is outside array bounds of ...
or given anti-ranges are odd maybe say
warning: array subscript [..., 0[ U ]19, ...] is outside of array bounds ...
or
warning: array subscript not in [0, 19] is outside of array ...
?
Note it's odd we somehow end up with
<bb 2> [local count: 1073741824]:
idx.2_8 = (unsigned int) idx_1(D);
if (idx.2_8 <= 19)
goto <bb 4>; [50.00%]
else
goto <bb 3>; [50.00%]
<bb 3> [local count: 536870912]:
_11 = gtable[idx_6].mode;
goto <bb 9>; [100.00%]
somehow function splitting exposes this, but since all calls to
.part.0 are properly guarded this is senseless doing (maybe IPA with VRP
info does something odd here with the signed/unsigned promotion)
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56456
[Bug 56456] [meta-bug] bogus/missing -Warray-bounds
