Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Incorrect cpuid config tests #1537

Closed
K-ballo opened this issue May 21, 2015 · 1 comment

Comments

Projects
None yet
2 participants
@K-ballo
Copy link
Member

commented May 21, 2015

Since db4a722, we do not execute our config tests anymore. This is causing all cpuid related config tests to return TRUE, as compilation always succeeds for them. This in turn causes RDSTCP to be always detected as supported, so we end up issuing a call to it for certain performance counters, which raises SIGILL.

We don't seem to have a documented rationale for disabling execution of config tests (@sithhell do you remember the motivation for that change?), but I think that at least for the cpuid related tests we would want to either execute them or fall back to false (for instance if cross-compiling). Our config tests infrastructure needs some changes in order to accommodate for that. Other config tests have to be checked to see they have not became useless/dangerous by disabling execution. I've looked at the cxx which are the ones I am familiarized with and those look fine.

Note this is currently stopping us from using the idle-rate counter on the old Sun machines.

@hkaiser hkaiser added this to the 0.9.11 milestone May 21, 2015

@hkaiser

This comment has been minimized.

Copy link
Member

commented May 21, 2015

I'd venture a guess and say that this change was accidental. @sithhell ?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.