Andreas Klöckner

> * `DeviceDataRecord.to_host_device_array` to `??`. Not sure about this one, but it seems like it should be deprecated and removed? Yes, I think that's the right approach. The need for...

> Update boxtree/tree_build.py I know it's tempting, but please modify the default commit message for accepting suggestions to describe a bit what's happening. Otherwise it's hard to know from a...

To cut down on noise in my email, I'm unsubscribing from this PR. When it next needs my attention, please @-mention me or hit the "request review" button. Otherwise, I...

#49 just landed (via squashing). Unfortunately, it looks like this has caused a few conflicts here. I hope you'll just be able to reapply these changes on top of that.

Are you running it the same way the CI does? https://github.com/inducer/ci-support/blob/f6803534383def5cf816002c1b08b2d128009871/ci-support.sh#L437= (I can reproduce the failure without a problem.)

Uh-oh. I was trying to get an early sense of (and maybe leave a few comments on) what was here that's not in #35, but this (for now) looks like...

https://github.com/jrprice/Oclgrind may be helpful, but there are some compatibility issues to sort out first.

Affected drivers are 370.x and 375.x. 367.x is what we had before.

Restoring to 367.x seems to have fixed things.

Thanks for the report, and sorry about the delay in responding. Could you take a look at #23 , especially since it disagrees with your proposed correct solution (the second...