Stream: t-compiler/const-eval

Topic: miri/pr/487


RalfJ (Oct 22 2018 at 06:47, on Zulip):

woah GitHub seems really broken currently. I tried to close and reopen the PR but even that did not help.

RalfJ (Oct 22 2018 at 06:50, on Zulip):

and it still shows the PR as open though I closed it twice. wtf.

RalfJ (Oct 22 2018 at 06:51, on Zulip):

it still says the PR is not mergeable even though it is a fast-forward from master?!?

RalfJ (Oct 22 2018 at 06:52, on Zulip):

oh I guess this explains it: https://status.github.com/messages

bjorn3 (Oct 22 2018 at 08:44, on Zulip):

Could you please make mod stacked_borrows public, so that I can use it in priroda?

bjorn3 (Oct 22 2018 at 08:46, on Zulip):

Can't leave a review comment due to that github problem :)

RalfJ (Oct 22 2018 at 09:33, on Zulip):

@bjorn3 what do you want to do with it? this is heavily WIP

RalfJ (Oct 22 2018 at 09:33, on Zulip):

(weird, the user bjorn3 is gone now?)

RalfJ (Oct 22 2018 at 09:36, on Zulip):

anyway @Oli would be nice if you could have a look at https://github.com/solson/miri/pull/487 so we can merge later when github is unborked. latest commit in there should be 0b22a1c .

oli (Oct 22 2018 at 10:36, on Zulip):

Github is so broken I can't even comment

oli (Oct 22 2018 at 10:36, on Zulip):

I'll try again later

RalfJ (Oct 22 2018 at 10:43, on Zulip):

you can comment here ;)

oli (Oct 22 2018 at 10:46, on Zulip):

What happens to immutable borrows in https://github.com/solson/miri/pull/487/files#diff-157b112a931b10b368e0e33da32c4a52R271

oli (Oct 22 2018 at 10:47, on Zulip):

I neither know what should be happening nor why it results in a raw mut borrow without interior mut

RalfJ (Oct 22 2018 at 10:48, on Zulip):

the problem is cases like &(i32, Cell<i32>)

RalfJ (Oct 22 2018 at 10:48, on Zulip):

then we want the memory covered by the first field to be treated with Frz, but the 2nd field should have Raw

oli (Oct 22 2018 at 10:49, on Zulip):

So the entire thing becomes raw until we go further in?

RalfJ (Oct 22 2018 at 10:49, on Zulip):

though I just realized I didnt even think this through, the ref only has one tag

RalfJ (Oct 22 2018 at 10:49, on Zulip):

I assume the ref would have Frz but that would be ignored for the Cell part

RalfJ (Oct 22 2018 at 10:49, on Zulip):

or so

RalfJ (Oct 22 2018 at 10:49, on Zulip):

I should likely rename Frz to Shr then

RalfJ (Oct 22 2018 at 10:49, on Zulip):

well, figuring out what exactly to do is future work :)

oli (Oct 22 2018 at 10:49, on Zulip):

Warrants a comment at least

RalfJ (Oct 22 2018 at 10:50, on Zulip):

and the comment is meant to indicate that there is work to be done. you want me to extend it?

oli (Oct 22 2018 at 10:51, on Zulip):

Well move it onto the arm at least

oli (Oct 22 2018 at 10:51, on Zulip):

I didn't figure out it only belonged to the second arm

RalfJ (Oct 22 2018 at 10:52, on Zulip):

https://github.com/solson/miri/commit/1a7fb7ec3ce5be7205fc1b3fb40642792985d501

oli (Oct 22 2018 at 11:04, on Zulip):

Otherwise I think I understood all functions but I don't think I have a full grokking of the overall design yet

oli (Oct 22 2018 at 11:05, on Zulip):

So r=me and I'll figure out things as they come

RalfJ (Oct 22 2018 at 13:04, on Zulip):

GH seems still broken? it doesnt trigger CI

oli (Oct 22 2018 at 13:07, on Zulip):

yea, it's on and off all the time

RalfJ (Oct 22 2018 at 13:08, on Zulip):

urg

Last update: Nov 15 2019 at 20:10UTC