Hi, Am Mittwoch, den 17.06.2015, 15:00 +0100 schrieb Edmund Grimley Evans: > > remove it there. > > Remove the assembly code, or remove the package? > > It wasn't clear to me whether the bit that reads an Intel performance > counter is central to the whole point of the package, or just a > feature that some users might find useful. If you modified that > function to return 0 on non-Intel architectures, would the package > then be useless on those other architectures?
I haven’t checked yet. > Also, I'm curious to know how useful this feature is on a multi-core > Intel system. Does the package arrange for processes to remain on a > single core? > > https://en.wikipedia.org/wiki/Time_Stamp_Counter says: "With the > advent of multi-core/hyper-threaded CPUs, systems with multiple CPUs, > and hibernating operating systems, the TSC cannot be relied on to > provide accurate results — unless great care is taken to correct the > possible flaws: rate of tick and whether all cores (processors) have > identical values in their time-keeping registers. There is no promise > that the timestamp counters of multiple CPUs on a single motherboard > will be synchronized." > Good points. They should probably be raised at https://github.com/bos/criterion/issues/80 though. -- Joachim "nomeata" Breitner Debian Developer nome...@debian.org | ICQ# 74513189 | GPG-Keyid: F0FBF51F JID: nome...@joachim-breitner.de | http://people.debian.org/~nomeata
signature.asc
Description: This is a digitally signed message part