Skip to content

Conversation

@mdaiter
Copy link

@mdaiter mdaiter commented Nov 1, 2016

I wanted to keep this as standard as possible. Within the Erlang docs, it says to add this flag to each method declared a "dirty function".

@kostis
Copy link
Contributor

kostis commented Nov 1, 2016

As you will (soon) see, the Travis tests fail on configurations that do not have support for dirty schedulers, so this needs to be corrected.
When this pull request has been corrected, please also change the commit message to fit in one line (e.g. "Add ERL_NIF_DIRTY_JOB_CPU_BOUND flag to the field declarations")

@TheGeorge
Copy link
Member

Nifty currently implements dirty schedulers in such a way that they are deactivated if compiled with a runtime that does not support them. The OTP version on Travis are not compiled with dirty scheduler support, that's the reason for the failing tests.

@kostis
Copy link
Contributor

kostis commented Nov 1, 2016

Actually, the Travis tests include only one OTP release (the last one, the one that does not fail in the tests of this PR) that is configured with --enable-dirty-schedulers.
For that configuration, and possibly for future Erlang/OTP releases that may have dirty schedulers enabled by default, the PR makes sense. For all others, not.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants