------- Comment #3 from ebotcazou at gcc dot gnu dot org 2006-03-22 10:05 ------- > I'm further digging but does this ring a bell to you, Richard?
I guess we should have a few V_MAY_DEFs for SFT.115, right? # VUSE <SFT.115_55>; q_17 = system__tasking__async_delays__timer_queue.succ; # q_1 = PHI <q_17(4), q_27(6)>; <L2>:; # VUSE <SMT.120_56>; D.1231_18 = q_1->resume_time; if (D.1231_18 >= t_13) goto <L4>; else goto <L3>; <L3>:; # VUSE <SMT.120_56>; q_26 = q_1->succ; q_27 = q_26; goto <bb 5> (<L2>); <L4>:; # SMT.120_58 = V_MAY_DEF <SMT.120_56>; d_12->succ = q_1; # VUSE <SMT.120_58>; D.1233_19 = q_1->pred; # SMT.120_59 = V_MAY_DEF <SMT.120_58>; d_12->pred = D.1233_19; # VUSE <SMT.120_59>; D.1234_20 = d_12->pred; # SMT.120_60 = V_MAY_DEF <SMT.120_59>; D.1234_20->succ = d_12; # SMT.120_61 = V_MAY_DEF <SMT.120_60>; q_1->pred = d_12; # VUSE <SFT.115_55>; D.1235_21 = system__tasking__async_delays__timer_queue.succ; if (D.1235_21 == d_12) goto <L5>; else goto <L6>; -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25737
