On Sat, Feb 24, 2018 at 3:30 PM, Christopher Li <sparse@xxxxxxxxxxx> wrote:
> On Sat, Feb 24, 2018 at 2:00 PM, Luc Van Oostenryck
> <luc.vanoostenryck@xxxxxxxxx> wrote:
>> Hi Chris,
>>
>> I saw that a few patches have been pushed on the head,
>> including the two I disagreed with and which I explained *why*
>> they where wrong. Well.
>>
>> Meanwhile, I see some regressions with the current head.
>> For example some simple programs like:
>> void a(void)
>> {
>> int b[] = { 8 };
>> int c;
>> for (;;)
>> b[c] = b[0];
>> }
>> now crashes when used with test-linearize.
>
> Oops, reproduced. Thanks for the report.
>
> I see pseudo->ident polluted by pseudo->size.
> Separate them out of the union seems to fix the crash.
Right, that is the exactly the course. I found the place that
use the ident without limit to PSEUDO_VAL.
found_dominator:
br = delete_last_instruction(&parent->insns);
phi = alloc_phi(parent, one->target, one->size);
phi->ident = phi->ident ? : one->target->ident; <=================
add_instruction(&parent->insns, br);
use_pseudo(insn, phi, add_pseudo(dominators, phi));
} END_FOR_EACH_PTR(parent);
So separate the "ident" and "size" in the union should fix this crash.
A little bit too aggressive on saving space.
Chris
--
To unsubscribe from this list: send the line "unsubscribe linux-sparse" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
