Stream: t-compiler

Topic: pre-meeting triage 2020-01-09 #54818


pnkfelix (Jan 09 2020 at 04:22, on Zulip):

starting the pre-triage process many hours before the meeting this week

pnkfelix (Jan 09 2020 at 04:23, on Zulip):

(for the short term, I cannot block out the time immediate before the triage meeting the way I have done in the past.)

pnkfelix (Jan 09 2020 at 04:23, on Zulip):

there are 11 open unprioritized nominated issues

pnkfelix (Jan 09 2020 at 04:24, on Zulip):

unpri nom 1/11: "ICE: no type-dependent def or overloaded call" #68025

pnkfelix (Jan 09 2020 at 04:32, on Zulip):

the ICE itself at least has a span. Not sure about the correctness of the suggested patch from the issue

pnkfelix (Jan 09 2020 at 04:34, on Zulip):

lets call it P-high for now. Can downgrade later if need be.

pnkfelix (Jan 09 2020 at 04:35, on Zulip):

unpri nom 2/11: "Extra umlaut causes error in rustc 1.42.0-nightly" #68000

pnkfelix (Jan 09 2020 at 04:37, on Zulip):

triage: P-high, removing nomination

pnkfelix (Jan 09 2020 at 04:38, on Zulip):

unpri nom 3/11: "ICE: Future with placeholder, returning async closure" #67995

centril (Jan 09 2020 at 04:41, on Zulip):

o.O; this is very early!

centril (Jan 09 2020 at 04:42, on Zulip):

unpri nom 3/11: "ICE: Future with placeholder, returning async closure" #67995

P-high?

pnkfelix (Jan 09 2020 at 04:48, on Zulip):

yeah. I got caught up trying to use cargo-bisect-rustc to bisect it myself

pnkfelix (Jan 09 2020 at 04:48, on Zulip):

the tool apparently doesn't (yet?) understand how a compilation failure can regress to an ICE

pnkfelix (Jan 09 2020 at 04:48, on Zulip):

triage: P-high, removing nomination

pnkfelix (Jan 09 2020 at 04:49, on Zulip):

unpri nom 4/11: "clippy-driver no longer builds after rust-lang/rust#67970" #67994

pnkfelix (Jan 09 2020 at 04:49, on Zulip):

triage: P-medium, removing nomination

centril (Jan 09 2020 at 04:50, on Zulip):

should we just stop triaging these toolstate things?

pnkfelix (Jan 09 2020 at 04:50, on Zulip):

maybe?

centril (Jan 09 2020 at 04:50, on Zulip):

and let the tool maintainers deal with it?

pnkfelix (Jan 09 2020 at 04:51, on Zulip):

unpri nom 5/11: "ICE: No move_errors should be allowed in MIR borrowck" #67947

centril (Jan 09 2020 at 04:51, on Zulip):

has PR

pnkfelix (Jan 09 2020 at 04:52, on Zulip):

triage: P-medium (under assumption that one does tend to get useful diagnostics before hitting ICE). removing nomination.

pnkfelix (Jan 09 2020 at 04:53, on Zulip):

/me now has to go kill a spider on behalf of their spouse

pnkfelix (Jan 09 2020 at 04:57, on Zulip):

unpri nom 6/11: "SIGSEGV in LLVM when calling main from #[no_mangle] fn" #67946

pnkfelix (Jan 09 2020 at 05:07, on Zulip):

this is tough. My heart wants to triage this as P-medium, because it is so niche

pnkfelix (Jan 09 2020 at 05:08, on Zulip):

but the failure mode is pretty terrible. I imagine its super frustrating for someone who runs into it to even figure out what happened.

pnkfelix (Jan 09 2020 at 05:08, on Zulip):

I'm going to make it P-high and leave it nominated

pnkfelix (Jan 09 2020 at 05:09, on Zulip):

unpri nom 7/11: "[LLVM] Segfault / "llvm::thinLTOInternalizeModule ... Assertion `GS != DefinedGlobals.end()' failed", again" #67855

centril (Jan 09 2020 at 05:10, on Zulip):

let's assign to tmandry?

pnkfelix (Jan 09 2020 at 05:11, on Zulip):

still trying to decide about priority

pnkfelix (Jan 09 2020 at 05:12, on Zulip):

easiest to call it P-high and assign it as suggested.

pnkfelix (Jan 09 2020 at 05:13, on Zulip):

unpri nom 8/11: ""internal compiler error" on Higher-Ranked Trait Bound for impl Trait in return position" #67830

centril (Jan 09 2020 at 05:14, on Zulip):

something for Matthew or Niko to look at?

pnkfelix (Jan 09 2020 at 05:14, on Zulip):

this seems familiar, is this a duplicate?

pnkfelix (Jan 09 2020 at 05:21, on Zulip):

could not find one. I guess I'll call this P-high and cc @Matthew Jasper and @nikomatsakis on it

pnkfelix (Jan 09 2020 at 05:22, on Zulip):

unpri nom 9/11: " ICE "Cannot copy into already initialized unsized place"" #67696

pnkfelix (Jan 09 2020 at 05:22, on Zulip):

has PR. P-high. Removing nomination.

pnkfelix (Jan 09 2020 at 05:23, on Zulip):

unpri nom 10/11: "Project fails to link when using dylibs and the -Zshare-generics flag" #67276

pnkfelix (Jan 09 2020 at 05:24, on Zulip):

ah gad dylibs

pnkfelix (Jan 09 2020 at 05:24, on Zulip):

I'll leave this for prioritization in the meeting tomorrow

pnkfelix (Jan 09 2020 at 05:25, on Zulip):

unpri nom 11/11: "Compile-time stack overflow when trait impl contains extern crate" #55779

pnkfelix (Jan 09 2020 at 05:28, on Zulip):

hmm apparently there's some story here since @eddyb had a (WIP?) PR up a little over a year ago that would have resolved this (PR #56655), but it bitrotted and died...

pnkfelix (Jan 09 2020 at 05:29, on Zulip):

triage: P-high. Leaving nominated for discussion of how to resolve; in particular, would it be worthwhile to try to resurrect @eddyb 's branch Iwhich may not even be possible since the branch itself was deleted...)

pnkfelix (Jan 09 2020 at 05:31, on Zulip):

there are no beta-regressions without a P-label

pnkfelix (Jan 09 2020 at 05:31, on Zulip):

there are no nightly regressions without a P-label

pnkfelix (Jan 09 2020 at 05:31, on Zulip):

there are 17 nominated issues. Yikes.

pnkfelix (Jan 09 2020 at 05:37, on Zulip):

hey @oli , I was skimming over the beta-nominations ... do you still want PR #67630 to be nominated for a beta backport?

pnkfelix (Jan 09 2020 at 05:38, on Zulip):

oh, I guess it resolves issue #67612 (a beta regression)

pnkfelix (Jan 09 2020 at 05:38, on Zulip):

Okay I'm clearly too tired to continue doing this tonight. See you all in some number of hours

lqd (Jan 09 2020 at 12:24, on Zulip):

@pnkfelix I also encountered issues and had to debug cargo-bisect-rustc but did manage to bisect #67995 to #67597

pnkfelix (Jan 09 2020 at 14:48, on Zulip):

rust-lang/cargo-bisect-rustc#34 seems relevant

lqd (Jan 09 2020 at 14:52, on Zulip):

I've just always used something like this to look for ICEs :)

for line in io::stdin().lock().lines() {
        let line = line.unwrap();
        if line.contains("internal compiler error") {
            std::process::exit(1);
        }
    }

    std::process::exit(0);
eddyb (Jan 13 2020 at 15:21, on Zulip):

triage: P-high. Leaving nominated for discussion of how to resolve; in particular, would it be worthwhile to try to resurrect eddyb 's branch Iwhich may not even be possible since the branch itself was deleted...)

there's a "Restore branch" button. I believe I deleted it because it became irrelevant after the refactors?

eddyb (Jan 13 2020 at 15:22, on Zulip):

@pnkfelix I pressed the button, in case anyone it's useful to anyone shrug

Last update: May 26 2020 at 11:10UTC