Attention is currently required from: Anastasia Klimchuk, Peter Marheine, Thomas Heijligen.
Brian Norris has posted comments on this change. ( https://review.coreboot.org/c/flashrom/+/81545?usp=email )
Change subject: udelay: only use OS time for delays, except on DOS ......................................................................
Patch Set 9: Code-Review+1
(3 comments)
Patchset:
PS9: LGTM % test question
File meson.build:
https://review.coreboot.org/c/flashrom/+/81545/comment/615b625a_b50dc7b2 : PS4, Line 121: DHAVE_NANOSLEEP=1
On reflection, this isn't even required. […]
👍
I was kinda wondering if we really cared to support non-POSIX.1-2008 here (outside of Windows and DOS, which build non-nanosleep code already).
File tests/udelay.c:
https://review.coreboot.org/c/flashrom/+/81545/comment/9e7989ec_59a9e340 : PS9, Line 41: This test could fail spuriously on a heavily-loaded system I see you acknowledge this already, but I just want to mention that this is definitely relevant for ChromeOS. We run tests from quite a few different projects in our CI environment, and sometimes this may yield a lot of system load at the same time. We've seen sleep-related latencies that are 5x their expected.
Specifically, we expected 100ms, so we tested for 500ms -- and even then, we saw frequent enough flakes that we expanded the margin to 1000ms.
That's to say, I expect this test will become flaky on ChromeOS CI. I don't know how that fits into the flashrom project expectations.
Anyway, personally, I'd recommend only testing the lower bound and not the upper bound.
---
For the record, public references:
https://chromium-review.googlesource.com/c/chromiumos/platform2/+/2207141 https://bugs.chromium.org/p/chromium/issues/detail?id=1083515
Non-public reference: b/333412847