Count instructions where timing cannot tell 3% apart (110)

016's second stage passed its criterion: six interleaved pairs against
ee21aec gave TPC-C (RAM, W=1) -0.6% at the median and the bank's
queries 0.0% to +2.2%, no run interfered. The ranks query was slower
all six times, and chasing it showed that on this Pi timing cannot
separate 1-3%: two servers asked in turns disagree with themselves
(ee21aec to e02bfc9 +1.6%, e02bfc9 to 481642f -1.2%, ee21aec to 481642f
+2.8%), and a query whose instructions grew 0.6% ran 3.5% faster.

Instructions do not move so: pcount.c counts the server's user-mode
instructions with perf_event_open (no perf here), and psample.c samples
where they go. ranks had grown 0.4% in instructions (the rest was where
code and data landed), per_branch 3.6% and tables 20 instructions a row,
which 07047c3 halved. The six pairs of 07047c3 were interfered (another
session's build, load 4-12) and are kept apart, unused.

1 files changed, 6 insertions, 0 deletions

diff --git a/research/experiments/025-i64-cost/2026-09-30-ab-one-process-linux-aarch64.txt b/research/experiments/025-i64-cost/2026-09-30-ab-one-process-linux-aarch64.txt
new file mode 100644
index 0000000..910487d
--- /dev/null
+++ b/research/experiments/025-i64-cost/2026-09-30-ab-one-process-linux-aarch64.txt
@@ -0,0 +1,6 @@
+ 17:01:12 up 1 day, 20:13,  3 users,  load average: 1.96, 2.72, 1.99
+{"query": "per_branch", "before_ms": 5.204, "after_ms": 5.284, "change_pct": 1.55, "rounds_after_slower": 10, "rounds": 12}
+{"query": "in_credit", "before_ms": 0.812, "after_ms": 0.811, "change_pct": -0.13, "rounds_after_slower": 6, "rounds": 12}
+{"query": "large", "before_ms": 0.366, "after_ms": 0.371, "change_pct": 1.44, "rounds_after_slower": 9, "rounds": 12}
+{"query": "ranks", "before_ms": 6.589, "after_ms": 6.772, "change_pct": 2.76, "rounds_after_slower": 11, "rounds": 12}
+ 17:01:55 up 1 day, 20:14,  3 users,  load average: 1.74, 2.57, 1.98