snapper gives negative wallclock numbers for some reason

gistfile1.txt

perl6 -e 'use Telemetry; snapper; snap "start"; my @fib-one = 1, 1, *+* ... * > 100_000; snap "fib 1 done"; my @fib-two = 1, 1, * + * ... * > 10_000; snap "fib2 done";'
Telemetry Report of Process #32508 (2019-04-16T17:37:40Z)
Number of Snapshots: 5
Initial/Final Size: 93472 / 98524 Kbytes
Total Time:           0.02 seconds
Total CPU Usage:      0.04 seconds
No supervisor thread has been running

wallclock  util%  max-rss
#-- start ----------------------------------------------------------------------
      -21  58.33
#-- fib 1 done -----------------------------------------------------------------
    19829  37.88     4972
#-- fib2 done ------------------------------------------------------------------
     2492  54.58       80
      176  25.14
--------- ------ --------
    22476  39.61     5052

Legend:
wallclock  Number of microseconds elapsed
    util%  Percentage of CPU utilization (0..100%)
  max-rss  Maximum resident set size (in Kbytes)

heapsnapshot.patch

diff --git a/lib/Telemetry.pm6 b/lib/Telemetry.pm6
index 29ce75258..dde8fbb42 100644
--- a/lib/Telemetry.pm6
+++ b/lib/Telemetry.pm6
@@ -708,6 +708,16 @@ multi sub snap(Str:D $message --> Nil) {
     T.message = $message;
     $snaps.push(T);
 }
+my int $snapshot-idx = 1;
+multi sub snap(:$heap! --> Nil) {
+    my str $filename = "heapsnapshot{$*PID}-{$snapshot-idx++}.mvmheap";
+    dd $filename;
+    nqp::mvmstartprofile(nqp::hash("kind", "heap", "filename", $filename));
+    nqp::mvmendprofile();
+    my \T := Telemetry.new;
+    T.message = $filename;
+    $snaps.push(T);
+}
 multi sub snap(@s --> Nil) {
     @s.push(Telemetry.new);
 }

output_with_heapsnapshots.txt

perl6-m -Ilib -e 'use Telemetry; snapper; snap :heap; my @fib-one = 1, 1, *+* ... * > 100_000; snap :heap; my @fib-two = 1, 1, * + * ... * > 10_000; snap :heap;'
Telemetry Report of Process #11539 (2019-04-16T18:07:04Z)
Number of Snapshots: 11
Initial/Final Size: 106712 / 211948 Kbytes
Total Time:           9.10 seconds
Total CPU Usage:      8.52 seconds
No supervisor thread has been running

wallclock  util%  max-rss
#-- taking heap snapshot... ----------------------------------------------------
       -3 641.67
  3148651  24.95    84612
#-- heapsnapshot11539-1.mvmheap ------------------------------------------------
      995  71.23
#-- taking heap snapshot... ----------------------------------------------------
     4085  49.37
  3248800  20.51     4080
#-- heapsnapshot11539-2.mvmheap ------------------------------------------------
      965  63.03
#-- taking heap snapshot... ----------------------------------------------------
     2564  54.06
  2692593  24.94    16544
#-- heapsnapshot11539-3.mvmheap ------------------------------------------------
      797  66.12
      245  25.10
--------- ------ --------
  9099692  23.39   105236

Legend:
wallclock  Number of microseconds elapsed
    util%  Percentage of CPU utilization (0..100%)
  max-rss  Maximum resident set size (in Kbytes)