sanchayanmaity/minix
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity

pm - fix sched interaction

Browse source 
For coredumping processes, PM forgets to inform SCHED that the
process has vanished, causing future fork()s to fail.
This commit is contained in:
Ben Gras 2010-05-19 13:22:29 +00:00
parent 77d5243f4a
commit bcdaf033b5
2 changed files with 11 additions and 11 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

11
servers/pm/forkexit.c
View file

@ -364,6 +364,17 @@ int dump_core; /* flag indicating whether to dump core */
*/
int r;
if((r = sched_stop(rmp)) != OK) {
/* If the scheduler refuses to give up scheduling, there is
* little we can do, except report it. This may cause problems
* later on, if this scheduler is asked to schedule another proc
* that has an endpoint->schedproc mapping identical to the proc
* we just tried to stop scheduling.
*/
printf("PM: The scheduler did not want to give up "
"scheduling %s, ret=%d.\n", rmp->mp_name, r);
}
/* For core dumps, now is the right time to try to contact the parent. */
if (dump_core)
zombify(rmp);

11
servers/pm/main.c
View file

@ -479,17 +479,6 @@ PRIVATE void handle_fs_reply()
break;
case PM_EXIT_REPLY:
if((r = sched_stop(rmp)) != OK) {
/* If the scheduler refuses to give up scheduling, there is
* little we can do, except report it. This may cause problems
* later on, if this scheduler is asked to schedule another proc
* that has an endpoint->schedproc mapping identical to the proc
* we just tried to stop scheduling.
*/
printf("PM: The scheduler did not want to give up "
"scheduling %s, ret=%d.\n", rmp->mp_name, r);
}
exit_restart(rmp, FALSE /*dump_core*/);
break;