Re: [PATCH] Fix VRP update_value_range and caller (PR tree-optimization/65875)

2015-04-27 Thread Richard Biener
On Mon, 27 Apr 2015, Jakub Jelinek wrote: > On Mon, Apr 27, 2015 at 10:05:21AM +0200, Richard Biener wrote: > > > --- gcc/tree-vrp.c.jj 2015-04-20 14:35:39.0 +0200 > > > +++ gcc/tree-vrp.c2015-04-24 18:10:41.321367440 +0200 > > > @@ -892,7 +892,12 @@ update_value_range (const_t

Re: [PATCH] Fix VRP update_value_range and caller (PR tree-optimization/65875)

2015-04-27 Thread Jakub Jelinek
On Mon, Apr 27, 2015 at 10:05:21AM +0200, Richard Biener wrote: > > --- gcc/tree-vrp.c.jj 2015-04-20 14:35:39.0 +0200 > > +++ gcc/tree-vrp.c 2015-04-24 18:10:41.321367440 +0200 > > @@ -892,7 +892,12 @@ update_value_range (const_tree var, valu > > UNDEFINED or from VARYING. */ >

Re: [PATCH] Fix VRP update_value_range and caller (PR tree-optimization/65875)

2015-04-27 Thread Richard Biener
On Fri, 24 Apr 2015, Jakub Jelinek wrote: > Hi! > > In vrp_visit_assignment_or_call we try to return SSA_PROP_VARYING > if update_value_range returned true and the new range is VR_VARYING, > but vrp_visit_phi_node fails to do that. > Another thing is that if update_value_range decides to > set_va

[PATCH] Fix VRP update_value_range and caller (PR tree-optimization/65875)

2015-04-24 Thread Jakub Jelinek
Hi! In vrp_visit_assignment_or_call we try to return SSA_PROP_VARYING if update_value_range returned true and the new range is VR_VARYING, but vrp_visit_phi_node fails to do that. Another thing is that if update_value_range decides to set_value_range_to_varying (old_vr); it doesn't update new_vr,