GP-885: Better indication of ineffective breakpoints. Toggles on lower table.

This commit is contained in:
Dan
2021-04-23 16:12:22 -04:00
parent 128f8bc3d3
commit be8e27150a
34 changed files with 941 additions and 180 deletions

View File

@@ -22,8 +22,10 @@ import org.apache.commons.collections4.set.AbstractSetDecorator;
/**
* The kind of breakpoint
*
* <p>
* This identifies the sort of access that would trap execution
*
* <p>
* TODO: This is identical to {@code TargetBreakpointKind} (not in the classpath here). Is there a
* common place we could factor both? Should we? CAUTION: Encoding in a trace database depends on
* this enum's {@code bits} field, so we must take care not to introduce a dependency that would
@@ -32,8 +34,8 @@ import org.apache.commons.collections4.set.AbstractSetDecorator;
public enum TraceBreakpointKind {
READ(1 << 0),
WRITE(1 << 1),
EXECUTE(1 << 2),
SOFTWARE(1 << 3);
HW_EXECUTE(1 << 2),
SW_EXECUTE(1 << 3);
public static class TraceBreakpointKindSet extends AbstractSetDecorator<TraceBreakpointKind> {
public static TraceBreakpointKindSet of(TraceBreakpointKind... kinds) {