sim: handle the case when there're not enough HW thread contexts
In SE mode, since there's no OS scheduler, the number of active SW threads is limited by the number of HW thread contexts. Previously, if there is no spare HW thread context, the simulator just fails and stops. Instead, this patch returns EAGAIN error code from a clone syscall if there's no available HW thread context. Then it's up to the simulated program to handle the error. Linux man page reference: http://man7.org/linux/man-pages/man2/clone.2.html http://man7.org/linux/man-pages/man2/fork.2.html Change-Id: Ib4e092433e49de4dde376c8cb81f7d3f7851cbc0 Reviewed-on: https://gem5-review.googlesource.com/c/9628 Reviewed-by: Brandon Potter <Brandon.Potter@amd.com> Maintainer: Brandon Potter <Brandon.Potter@amd.com>
This commit is contained in:
@@ -1431,8 +1431,11 @@ cloneFunc(SyscallDesc *desc, int callnum, Process *p, ThreadContext *tc)
|
||||
return -EINVAL;
|
||||
|
||||
ThreadContext *ctc;
|
||||
if (!(ctc = p->findFreeContext()))
|
||||
fatal("clone: no spare thread context in system");
|
||||
if (!(ctc = p->findFreeContext())) {
|
||||
DPRINTF_SYSCALL(Verbose, "clone: no spare thread context in system"
|
||||
"[cpu %d, thread %d]", tc->cpuId(), tc->threadId());
|
||||
return -EAGAIN;
|
||||
}
|
||||
|
||||
/**
|
||||
* Note that ProcessParams is generated by swig and there are no other
|
||||
|
||||
Reference in New Issue
Block a user