T21839r is failing on CI, 14–17% peak alloc
x86_64-linux-alpine3_12-validate+fully_static is failing despite multiple restarts with the following message:
Test Metric value New value Change
----------------------------------------------------------------------
MethSharing(normal) run/peak 1 1 +0.0%
T12791(normal) run/peak 1 1 +0.0%
T13218(normal) run/peak 1 1 +0.0%
T21839r(normal) run/peak 77 88 +14.3% BAD
T2762(normal) run/peak 1 1 +0.0%
T3586(normal) run/peak 17 17 +0.0%
T3738(normal) run/peak 1 1 +0.0%
T4029(ghci) run/peak 38 38 +0.0%
T4334(normal) run/peak 2 2 +0.0%
T5835(normal) run/peak 1 1 +0.0%
T7257(normal) run/peak 214 214 +0.0%
T7436(normal) run/peak 1 1 +0.0%
T7436b(normal) run/peak 1 1 +0.0%
T7850(normal) run/peak 2 2 +0.0%
lazy-bs-alloc(normal) run/peak 2 2 +0.0%
space_leak_001(normal) run/peak 2 2 +0.0%
geo. mean +0.8%
minimum +0.0%
maximum +14.3%
* All baselines were measured in the same environment as this test run
* All baseline commits are 14b5982a3ae
Some stats have changed If this is expected, allow changes by appending the git commit message with this:
-------------------------
Metric Increase:
T21839r
-------------------------
Unexpected results from:
TEST="T21839r"Affected MRs:
- !9751 (closed) (batch MR, no restarts)
- !9754 (closed) (batch MR, 3 restarts so far)
- !9753 (closed) (individual MR, 2 restarts so far)
That is 6 failed jobs in total. Based on the fact that we're seeing this in batch MRs where each individual change is has no effect on allocations (+0.0%), I'm inclined to believe that this is a problem with CI, not with the changes. Thus assigning to @chreekat