Stream: t-compiler

Topic: pre-meeting triage 2019-10-31 #54818


pnkfelix (Oct 31 2019 at 09:11, on Zulip):

I will be doing pre-triage in this channel

pnkfelix (Oct 31 2019 at 09:15, on Zulip):

first up: we have 9 unprioritized nominated issues

pnkfelix (Oct 31 2019 at 09:16, on Zulip):

1/9: "yet another associated type ICE (probably more missing normalize calls)" #65934

pnkfelix (Oct 31 2019 at 09:16, on Zulip):

this is spawned off of #63154, because I wanted it to have its own bug.

pnkfelix (Oct 31 2019 at 09:19, on Zulip):

(however, #63154 does remain open, because the original crate that initiated the bug report is still exposing an ICE that we believe to be unrelated to #65934 )

pnkfelix (Oct 31 2019 at 09:20, on Zulip):

anyway, I'll triage #65934 as P-high (and remove the nomination label) for now. I'm debating about whether to assign it to myself or not

pnkfelix (Oct 31 2019 at 09:23, on Zulip):

2/9: " Document query modifiers" #65929

pnkfelix (Oct 31 2019 at 09:23, on Zulip):

it seems like there is good discussion happening here

pnkfelix (Oct 31 2019 at 09:25, on Zulip):

I'm going to leave it nominated, because I want the team to discuss whether the discussion has reached the point where someone from @WG-learning could pick it up and make a PR to the rustc-guide

pnkfelix (Oct 31 2019 at 09:26, on Zulip):

nom unpri 3/9: "ICE with impl Fn alias." #65918

pnkfelix (Oct 31 2019 at 09:27, on Zulip):

looks to me like a MCVE has been identified (or at least something very close to minimal)

pnkfelix (Oct 31 2019 at 09:27, on Zulip):

has a fix

pnkfelix (Oct 31 2019 at 09:28, on Zulip):

marking P-high and unnominating

pnkfelix (Oct 31 2019 at 09:29, on Zulip):

nom unpri 4/9: "miri no longer builds after rust-lang/rust#65885" #65889

pnkfelix (Oct 31 2019 at 09:29, on Zulip):

has a PR. marking P-medium and unnominating

pnkfelix (Oct 31 2019 at 09:30, on Zulip):

nom unpri 5/9: "Some features can no longer be controlled by conditional compilation" #65860

pnkfelix (Oct 31 2019 at 09:33, on Zulip):

long comment thread

pnkfelix (Oct 31 2019 at 09:35, on Zulip):

leaving nominated for discussion at team meeting. But I'm going to float this issue to the main thread to prepare people

pnkfelix (Oct 31 2019 at 09:36, on Zulip):

triage: P-high for making a decision about whether to back out the relevant change or not

pnkfelix (Oct 31 2019 at 09:37, on Zulip):

nom unpri 6/9: '"LLVM ERROR: Access past stack top!" when compiling without sse2' #65844

pnkfelix (Oct 31 2019 at 09:38, on Zulip):

sounds bad

pnkfelix (Oct 31 2019 at 09:39, on Zulip):

I wonder if it could be a candidate for ICE-breaker-LLVM ?

pnkfelix (Oct 31 2019 at 09:40, on Zulip):

certainly seems like a candidate, e.g. if we could turn the rust source into LLVM IR and/or look for a relevant bug upstream with LLVM repo

pnkfelix (Oct 31 2019 at 09:41, on Zulip):

heh. I think we need documentation somewhere as to what the protocol is from the T-compiler side for these ICE breakers.

pnkfelix (Oct 31 2019 at 09:42, on Zulip):

I'm pretty sure I'm just supposed to issue some command to triagebot ...

pnkfelix (Oct 31 2019 at 09:44, on Zulip):

found something here: https://github.com/rust-lang/triagebot/wiki/Pinging

pnkfelix (Oct 31 2019 at 09:45, on Zulip):

okay, think I did that right

pnkfelix (Oct 31 2019 at 09:45, on Zulip):

having said that, I'm not sure how heavily to prioritize this

pnkfelix (Oct 31 2019 at 09:46, on Zulip):

lets make it P-high for now

pnkfelix (Oct 31 2019 at 09:47, on Zulip):

nom unpri 7/9: "rls no longer builds after rust-lang/rust#65764" #65803

pnkfelix (Oct 31 2019 at 09:47, on Zulip):

triaging as P-medium and removing nomination

pnkfelix (Oct 31 2019 at 09:49, on Zulip):

nom unpri 8/9: "ICE: errors resolving bounds after type-checking for fmt::Display" #65774

pnkfelix (Oct 31 2019 at 09:51, on Zulip):

going to mark this as P-high for now, since it is not clear to me how easily one can figure out what to do to work-around the ICE from the diagnostics we are generating for this

pnkfelix (Oct 31 2019 at 09:55, on Zulip):

nom unpri 9/9: "Reduce the places where stable annotations are needed" #65515

pnkfelix (Oct 31 2019 at 09:56, on Zulip):

hmm there wasn't actual discussion of this in my absence last week

pnkfelix (Oct 31 2019 at 09:57, on Zulip):

my hope had been to actually reach some kind of consensus about whether to take action here or not

pnkfelix (Oct 31 2019 at 09:59, on Zulip):

(@nikomatsakis had said during the meeting that they were going to remove the nomination, but they didn't, and that oversight is all for the better, in my opinion...)

pnkfelix (Oct 31 2019 at 10:00, on Zulip):

pre-pass: there is 1 unprioritized beta-regression

pnkfelix (Oct 31 2019 at 10:00, on Zulip):

ah I forgot to mark this P-high last week

pnkfelix (Oct 31 2019 at 10:01, on Zulip):

assigning to self, if we're going to fix this in beta, it needs to happen really fast

pnkfelix (Oct 31 2019 at 10:02, on Zulip):

pre-pass: there are 2 unprioritized nightly regressions

pnkfelix (Oct 31 2019 at 10:03, on Zulip):

unpri nightly regr 1/2: " Pre-expansion gating collects spans from unSuccessful macro matchers" #65846

pnkfelix (Oct 31 2019 at 10:04, on Zulip):

has fix in #65974

pnkfelix (Oct 31 2019 at 10:05, on Zulip):

(and is related to already nominated #65860, as noted by @simulacrum )

pnkfelix (Oct 31 2019 at 10:07, on Zulip):

anyway, triage P-high

pnkfelix (Oct 31 2019 at 10:14, on Zulip):

unpri nightly regr 2/2: "Panicking doesn't work on arm-unknown-linux-musleabi" #65765

pnkfelix (Oct 31 2019 at 10:17, on Zulip):

has a fix. triage P-high

pnkfelix (Oct 31 2019 at 10:17, on Zulip):

ah this was T-libs not T-compiler

pnkfelix (Oct 31 2019 at 10:17, on Zulip):

so I'll leave it nominted

pnkfelix (Oct 31 2019 at 10:18, on Zulip):

next prepass: all nominations (to see if any need their nomination label removed.)

pnkfelix (Oct 31 2019 at 10:18, on Zulip):

already discussed #65929 #65860 and #65515

pnkfelix (Oct 31 2019 at 10:19, on Zulip):

only other nomination is "Rust 1.38 regressions weren't fully triaged" #65577

pnkfelix (Oct 31 2019 at 10:19, on Zulip):

@simulacrum says it was a process failure

pnkfelix (Oct 31 2019 at 10:20, on Zulip):

removing nomination label

pnkfelix (Oct 31 2019 at 10:22, on Zulip):

we have 34 open P-high issues

pnkfelix (Oct 31 2019 at 10:23, on Zulip):

I don't have time to walk them right now

Last update: Nov 16 2019 at 01:05UTC