Now works again, though I still get some crashes in `make check'. As
you can see, the bugs were really stupid...
Paolo
2008-03-19 Paolo Bonzini <
[hidden email]>
* libgst/prims.def: Mark VMpr_Process_suspend as a possible
source of interrupts.
* libgst/files.c: Load BlockClosure before the first closures
are executed (by the JIT).
diff --git a/libgst/files.c b/libgst/files.c
index e5e474a..e59bdf0 100644
--- a/libgst/files.c
+++ b/libgst/files.c
@@ -228,14 +228,14 @@ static const char standard_files[] = {
"ProcSched.st\0"
"Delay.st\0"
"SharedQueue.st\0"
- "Behavior.st\0"
- "ClassDesc.st\0"
- "Class.st\0"
- "Metaclass.st\0"
"ContextPart.st\0"
"MthContext.st\0"
"BlkContext.st\0"
"BlkClosure.st\0"
+ "Behavior.st\0"
+ "ClassDesc.st\0"
+ "Class.st\0"
+ "Metaclass.st\0"
"Continuation.st\0"
"Memory.st\0"
"MethodInfo.st\0"
diff --git a/libgst/prims.def b/libgst/prims.def
index 58842c5..02de21c 100644
--- a/libgst/prims.def
+++ b/libgst/prims.def
@@ -2814,7 +2814,7 @@ primitive VMpr_Semaphore_waitAfterSignalling [succeed,check_interrupt]
}
/* Process suspend */
-primitive VMpr_Process_suspend [succeed]
+primitive VMpr_Process_suspend [succeed,check_interrupt]
{
OOP oop1;
_gst_primitives_executed++;
_______________________________________________
help-smalltalk mailing list
[hidden email]
http://lists.gnu.org/mailman/listinfo/help-smalltalk
Locked
