Stream: t-compiler

Topic: pre-meeting triage 2019-05-09 #54818


nikomatsakis (May 09 2019 at 12:29, on Zulip):

Good morning @T-compiler/meeting -- the triage meeting is in 90 minutes. I'll be doing pre-triage today. According to the working group calendar, we're scheduled to have a check-in from @WG-nll and @WG-diagnostics. Who wants to do those updates?

Perhaps, @Matthew Jasper, are you around to describe current status of NLL?

Perhaps @Esteban K├╝ber for diagnostics?

nikomatsakis (May 09 2019 at 12:32, on Zulip):

# Unprioritized I-nominated compiler issues

nikomatsakis (May 09 2019 at 12:32, on Zulip):

There are 18 :(

oli (May 09 2019 at 12:32, on Zulip):

I can do diagnostics

nikomatsakis (May 09 2019 at 12:33, on Zulip):

ICE with generic associated type #60654

nikomatsakis (May 09 2019 at 12:33, on Zulip):

Seems like the surprising thing here is that we permit type Item<'b> at all

nikomatsakis (May 09 2019 at 12:33, on Zulip):

Ah, ok, we don't

nikomatsakis (May 09 2019 at 12:35, on Zulip):

I've added to tracking issue for GATs and removed nomination

nikomatsakis (May 09 2019 at 12:35, on Zulip):

ICE in LlvmArchiveBuilder when cross-compiling openssl for Android ARM64 #60652

nikomatsakis (May 09 2019 at 12:36, on Zulip):

"There is an ICE in nightly that I can trigger by adding a dependency on vendored openssl when cross-compiling for aarch64 android with the NDK. It's possible the problem here isn't really rustc's fault, but it does crash."

nikomatsakis (May 09 2019 at 12:37, on Zulip):

To what extent do we support vendered openssl?

nikomatsakis (May 09 2019 at 12:37, on Zulip):

I'll leave this open to discuss in the main meeting

nikomatsakis (May 09 2019 at 12:37, on Zulip):

compiler panic "randomly" with incremental build #60629

nikomatsakis (May 09 2019 at 12:39, on Zulip):

Seems like a real incremental issue, not very well clarified, but perhaps close. It might be nice to get this to an exact set of steps. I'm going to try and cover it during meeting today, but cc @mw for now. (re: #60629)

nikomatsakis (May 09 2019 at 12:39, on Zulip):

Compiler crash when dealing with Any and Ref #60622

nikomatsakis (May 09 2019 at 12:40, on Zulip):

regression from stable-to-stable

nikomatsakis (May 09 2019 at 12:41, on Zulip):

it's been bisected down but to a very large range of commits

nikomatsakis (May 09 2019 at 12:41, on Zulip):

too large to be actionable

nikomatsakis (May 09 2019 at 12:41, on Zulip):

I guess I will mark as P-high

nikomatsakis (May 09 2019 at 12:41, on Zulip):

(regression, ICE) -- the only caveat is that this pattern is deprecated (future compat warning)_

nikomatsakis (May 09 2019 at 12:43, on Zulip):

Compiler panic #60621

nikomatsakis (May 09 2019 at 12:44, on Zulip):

not enough info and OP cannot reproduce, going to close for now

nikomatsakis (May 09 2019 at 12:45, on Zulip):

Infinite "recursive requirements" compiler error #60603

nikomatsakis (May 09 2019 at 12:47, on Zulip):

Hmm, I'm marking as medium, but honestly I don't know what we ought to do to fix this. Tempted to close instead.

nikomatsakis (May 09 2019 at 12:47, on Zulip):

ICE with associated existential type #60564

nikomatsakis (May 09 2019 at 12:49, on Zulip):

Marking as P-medium, unstable feature.

nikomatsakis (May 09 2019 at 12:49, on Zulip):

Release build taking forever on 1.34.1 #60548

nikomatsakis (May 09 2019 at 12:51, on Zulip):

regression, P-high, but it seems @Matthew Jasper has a fix pending, though the full nature of the problem is not deeply understood

nikomatsakis (May 09 2019 at 12:52, on Zulip):

miri no longer builds after rust-lang/rust#60462 #60533

nikomatsakis (May 09 2019 at 12:52, on Zulip):

@RalfJ seems to be o nit, blocked on something or other, P-medium

nikomatsakis (May 09 2019 at 12:52, on Zulip):

Compiler-internal lint for public libcore items not reexported in libstd #60479

nikomatsakis (May 09 2019 at 12:53, on Zulip):

Not sure why this is tagged as T-compiler

centril (May 09 2019 at 12:53, on Zulip):

@nikomatsakis I think because "do the impl work of enforcing the lint"

centril (May 09 2019 at 12:53, on Zulip):

or just "write the lint"

nikomatsakis (May 09 2019 at 12:54, on Zulip):

I think the lint exists and enforcing it is more T-libs

nikomatsakis (May 09 2019 at 12:54, on Zulip):

but it's not really clear what is even being requested here

centril (May 09 2019 at 12:54, on Zulip):

oh? TIL :P

nikomatsakis (May 09 2019 at 12:54, on Zulip):

i.e., what lint?

nikomatsakis (May 09 2019 at 12:54, on Zulip):

anyway, I removed T-compiler :)

nikomatsakis (May 09 2019 at 12:54, on Zulip):

ICE impl_trait_in_bindings with elided lifetime #60473

nikomatsakis (May 09 2019 at 12:54, on Zulip):

i.e., what lint?

to be clear, if the lint is a lint for "public but not visible from outside", I believe it exists as part of rust 2018 compability

centril (May 09 2019 at 12:55, on Zulip):

i.e., what lint?

An internal lint or a tidy check that makes sure that a stable libcore path exists in liballoc and libstd

centril (May 09 2019 at 12:56, on Zulip):

@nikomatsakis btw, if you have noticed, I've been doing a pre-pre-Triage of issues for this meeting where I suggest a P-* label for you; like it?

nikomatsakis (May 09 2019 at 12:57, on Zulip):

I noticed you had nominated a lot :) and in some cases included a suggested priority

nikomatsakis (May 09 2019 at 12:57, on Zulip):

is that what you mean?

centril (May 09 2019 at 12:57, on Zulip):

@nikomatsakis yeah

centril (May 09 2019 at 12:57, on Zulip):

(ICEs get nominated automatically ^^)

nikomatsakis (May 09 2019 at 12:57, on Zulip):

An internal lint or a tidy check that makes sure that a stable libcore path exists in liballoc and libstd

Ah, that might not exist. I suppose that could be T-compiler.

nikomatsakis (May 09 2019 at 12:57, on Zulip):

I do like having a suggestion, yes :)

centril (May 09 2019 at 12:58, on Zulip):

cool; then I'll continue with it :slight_smile:

nikomatsakis (May 09 2019 at 12:58, on Zulip):

I think I've agreed with them all so far

nikomatsakis (May 09 2019 at 12:58, on Zulip):

I think we need a better way to handle "unstable buggy features"

nikomatsakis (May 09 2019 at 12:58, on Zulip):

in general, our prioritization scheme neeeds love, but we know that

nikomatsakis (May 09 2019 at 12:59, on Zulip):

Regression: Typemap type mismatch in 1.34.0+ #60375

centril (May 09 2019 at 13:00, on Zulip):

Me and Felix didn't understand that one

centril (May 09 2019 at 13:00, on Zulip):

Don't believe that has changed; perhaps you have more insights

RalfJ (May 09 2019 at 13:01, on Zulip):

RalfJ seems to be o nit, blocked on something or other, P-medium

blocked on the lang team looking at https://github.com/rust-lang/rust/pull/60469 ;)

RalfJ (May 09 2019 at 13:01, on Zulip):

or rather deciding about it

nikomatsakis (May 09 2019 at 13:02, on Zulip):

Re: #60375, hmm, seems like it needs investigation. <-- this is another category we don't have a great way to handle, I think we want to categorize them as "help wanted" or something

nikomatsakis (May 09 2019 at 13:02, on Zulip):

I guess i'll call it P-high for now (regression, unknown cause)

centril (May 09 2019 at 13:03, on Zulip):

blocked on the lang team looking at https://github.com/rust-lang/rust/pull/60469 ;)

@RalfJ you mean libs team?

centril (May 09 2019 at 13:03, on Zulip):

@nikomatsakis I-needs-investigation ?

nikomatsakis (May 09 2019 at 13:03, on Zulip):

Stable rustc always panics on arm/musl #60297

nikomatsakis (May 09 2019 at 13:04, on Zulip):

wait, is triagebot active?

nikomatsakis (May 09 2019 at 13:04, on Zulip):

@centril perhaps, yeah

centril (May 09 2019 at 13:04, on Zulip):

wait, is triagebot active?

Sure; the @rustbot accepts commands-

nikomatsakis (May 09 2019 at 13:05, on Zulip):

it didn't seem to hear me declaring triage: P-high

nikomatsakis (May 09 2019 at 13:05, on Zulip):

Stable rustc always panics on arm/musl #60297

So arm-unknown-linux-musleabihf is tier 2.

nikomatsakis (May 09 2019 at 13:05, on Zulip):

that means, I think, we guarantee it will build, but we don't run tests

Zoxc (May 09 2019 at 13:06, on Zulip):

Does https://github.com/rust-lang/rust/pull/59064 escape triage searches?

nikomatsakis (May 09 2019 at 13:06, on Zulip):

@Zoxc I haven't gotten that far yet

centril (May 09 2019 at 13:06, on Zulip):

@nikomatsakis https://github.com/rust-lang/triagebot/wiki/Labeling

nikomatsakis (May 09 2019 at 13:07, on Zulip):

did we remove the triage: P-high notation?

nikomatsakis (May 09 2019 at 13:08, on Zulip):

You used to be able to just say triage: p-high and it would know to remove the nominated tag etc

centril (May 09 2019 at 13:08, on Zulip):

don't think it was a thing? cc @simulacrum

nikomatsakis (May 09 2019 at 13:08, on Zulip):

I think it maybe done by highfive, I don't know

nikomatsakis (May 09 2019 at 13:09, on Zulip):

Decouple nightly RLS from Clippy #59761

seems like p-medium

nikomatsakis (May 09 2019 at 13:09, on Zulip):

RLS ICE: "node_type: no type for node" when building bitflags #59134

nikomatsakis (May 09 2019 at 13:10, on Zulip):

"This is now causing widespread ICEs in the RLS. Nominating to get this fixed (maybe @Xanewok has an idea?). See #60484 (comment) for a simpler reproducer."

nikomatsakis (May 09 2019 at 13:10, on Zulip):

p-high, assign to xanewok

nikomatsakis (May 09 2019 at 13:11, on Zulip):

Update dependency versions in the various Cargo.tomls to the version that we actually use #57443

Pietro Albini (May 09 2019 at 13:11, on Zulip):

@nikomatsakis I think the triage: P-high was the old triagebot

Pietro Albini (May 09 2019 at 13:11, on Zulip):

which is not on rust infra and I have no idea on where it lives (maybe nrc's own server?)

Pietro Albini (May 09 2019 at 13:12, on Zulip):

we could reimplement that on the new triagebot though

nikomatsakis (May 09 2019 at 13:12, on Zulip):

ok. I'll try to learn the new syntax, but it might be nice to have a similar shorthand.

nikomatsakis (May 09 2019 at 13:12, on Zulip):

So #57443 seems like it was nominated for discusison. I guess i'll call it p-medium

Zoxc (May 09 2019 at 13:13, on Zulip):

@nikomatsakis It was nominated a few weeks ago, so I was just wondering if it escaped the search filters

Pietro Albini (May 09 2019 at 13:13, on Zulip):

it's @rustbot modify labels: P-high -I-nominated right now

Pietro Albini (May 09 2019 at 13:13, on Zulip):

what we could do is make I-nominated and P-* exclusive

nikomatsakis (May 09 2019 at 13:13, on Zulip):

@Zoxc I don't think so, we just have a really hard time kepeing up

nikomatsakis (May 09 2019 at 13:14, on Zulip):

@Pietro Albini I feel like it could be @rustbot triage P-high or something; i.e., I think I'd prefer if that syntax were as simple as possible

nikomatsakis (May 09 2019 at 13:14, on Zulip):

but I don't have a strong opinion

nikomatsakis (May 09 2019 at 13:14, on Zulip):

but there arE times you want something to both be nominated

Pietro Albini (May 09 2019 at 13:14, on Zulip):

@simulacrum ^

nikomatsakis (May 09 2019 at 13:14, on Zulip):

and have a priority

nikomatsakis (May 09 2019 at 13:14, on Zulip):

Panics in destructors can cause the return value to be leaked #47949

nikomatsakis (May 09 2019 at 13:15, on Zulip):

This .. definitely feels like a bug .. but it's hard to call it p-high ..

nikomatsakis (May 09 2019 at 13:15, on Zulip):

I guess p-medium but I think it might be nice if we clearly defined the expected drop order at least and then documented known bugs

nikomatsakis (May 09 2019 at 13:16, on Zulip):

Implement converting an AST to a token tree #43081

nikomatsakis (May 09 2019 at 13:16, on Zulip):

I think @eddyb nominatedthis for discussion and design

Pietro Albini (May 09 2019 at 13:16, on Zulip):

@nikomatsakis https://github.com/rust-lang/triagebot/issues/10

nikomatsakis (May 09 2019 at 13:16, on Zulip):

I'm going to mark it as p-medium

nikomatsakis (May 09 2019 at 13:17, on Zulip):

Ugh, ok, finally got through the nominated things without a priority

centril (May 09 2019 at 13:17, on Zulip):

oh no; not another drop order issue :P

nikomatsakis (May 09 2019 at 13:53, on Zulip):

Argh, I was not able to do the rest of pre-triage

nikomatsakis (May 09 2019 at 13:53, on Zulip):

Got called away. but let me try to do a quick sweep.

nikomatsakis (May 09 2019 at 13:53, on Zulip):

# Stable/beta backport nominations

nikomatsakis (May 09 2019 at 13:54, on Zulip):

https://github.com/rust-lang/rust/pull/60649 -- an RLS fix

nikomatsakis (May 09 2019 at 13:54, on Zulip):

# P-high issues

nikomatsakis (May 09 2019 at 13:54, on Zulip):

ZOMG there are 30. We've got to do something.

nikomatsakis (May 09 2019 at 13:54, on Zulip):

But partly I think this is a categorization challenge.

nikomatsakis (May 09 2019 at 13:55, on Zulip):

Encountered error Unimplemented selecting <...> during codegen #58375

I'm going to assign to myself to investigate.

nikomatsakis (May 09 2019 at 13:56, on Zulip):

rustc 1.32.0 produces faulty wasm code #58548

nikomatsakis (May 09 2019 at 13:59, on Zulip):

asked for a reproduction

nikomatsakis (May 09 2019 at 14:00, on Zulip):

Compiler panic when using a slice pattern #59016

assigned to @oli -- looks like we hvae a PR

oli (May 09 2019 at 14:02, on Zulip):

I'll rebase that, but I'm unsure how to proceed (lazy norm questions came up)

RalfJ (May 09 2019 at 14:09, on Zulip):

blocked on the lang team looking at https://github.com/rust-lang/rust/pull/60469 ;)

RalfJ you mean libs team?

sorry, yes, I do

Last update: Nov 22 2019 at 05:35UTC