build: runtime failures

[gopherbot]

#!watchflakes
default <- repo == "go" && pkg == "" && test == "runtime"

Issue created automatically to collect these failures.

Example (log):

FAIL	runtime	1080.016s

— watchflakes

[gopherbot]

Found new dashboard test flakes for:

#!watchflakes
default <- repo == "go" && pkg == "" && test == "runtime"

2024-05-21 17:17 gotip-linux-ppc64_power10 go@dfb7073b runtime (log)

FAIL	runtime	1080.016s

2024-05-21 17:17 gotip-linux-ppc64_power8 go@dfb7073b runtime (log)

FAIL	runtime	1080.126s

— watchflakes

[rhysh]

The failures say:

    metrics_test.go:969: NumCPU 4
    metrics_test.go:970: GOMAXPROCS 4
panic: test timed out after 18m0s
	running tests:
		TestRuntimeLockMetricsAndProfile (17m46s)
		TestRuntimeLockMetricsAndProfile/runtime.lock (17m46s)
		TestRuntimeLockMetricsAndProfile/runtime.lock/sample-1 (17m46s)

and

    metrics_test.go:969: NumCPU 4
    metrics_test.go:970: GOMAXPROCS 4
panic: test timed out after 18m0s
	running tests:
		TestRuntimeLockMetricsAndProfile (17m44s)
		TestRuntimeLockMetricsAndProfile/runtime.lock (17m44s)
		TestRuntimeLockMetricsAndProfile/runtime.lock/sample-1 (17m44s)

It's ppc64, which gave me some trouble (on that same test) in #64253. But the build dashboard shows these only failing on one commit (in the middle of a patch series I landed today to update runtime-internal mutex profiles).

Because it was in the middle of the series, and so the code has changed yet again since the failure, I'm going to close this and trust the bot to re-open if the issue shows up again.