| Message ID | 20150821134029.GE399@x4 |
|---|---|
| State | New |
| Headers | show |
On 08/21/2015 07:40 AM, Markus Trippelsdorf wrote: > bootstrap-ubsan shows: > loop-iv.c:2626:14: runtime error: signed integer overflow: 9223372036854775806 - -9223372036854775808 cannot be represented in type 'long int' > > Fixed by moving the variables in question from signed to unsigned. > > Bootstrapped and tested on ppc64le. > OK for trunk and gcc-5 branch? > > Thanks. > > 2015-08-21 Markus Trippelsdorf <markus@trippelsdorf.de> > > PR rtl-optimization/61657 > * loop-iv.c (iv_number_of_iterations): Declare up and down as > unsigned. Remove superflous uint64_t cast. OK for trunk. AFAIK this isn't a regression, so Richi, Jakub or Joseph ought to explicitly ack for the gcc-5 branch. jeff >
diff --git a/gcc/loop-iv.c b/gcc/loop-iv.c index 6e9cc8c..1c9a159 100644 --- a/gcc/loop-iv.c +++ b/gcc/loop-iv.c @@ -2330,8 +2330,8 @@ iv_number_of_iterations (struct loop *loop, rtx_insn *insn, rtx condition, enum rtx_code cond; machine_mode mode, comp_mode; rtx mmin, mmax, mode_mmin, mode_mmax; - uint64_t s, size, d, inv, max; - int64_t up, down, inc, step_val; + uint64_t s, size, d, inv, max, up, down; + int64_t inc, step_val; int was_sharp = false; rtx old_niter; bool step_is_pow2; @@ -2621,7 +2621,7 @@ iv_number_of_iterations (struct loop *loop, rtx_insn *insn, rtx condition, down = INTVAL (CONST_INT_P (iv0.base) ? iv0.base : mode_mmin); - max = (uint64_t) (up - down) / inc + 1; + max = (up - down) / inc + 1; if (!desc->infinite && !desc->assumptions) record_niter_bound (loop, max, false, true);