Stream: t-compiler

Topic: pre-meeting triage 2019-06-20 #54818


pnkfelix (Jun 20 2019 at 13:28, on Zulip):

I will be doing pre-triage in this channel.

pnkfelix (Jun 20 2019 at 13:31, on Zulip):

first up: nominated unprioritized issues

pnkfelix (Jun 20 2019 at 13:31, on Zulip):

"Forgone caching in cycles caused much overflow in trait solving" #61960 -- i filed this last night. Centril since nominated it, presumably to get it prioritized. I think it probably gets the same priority that related issue #60846 has.

pnkfelix (Jun 20 2019 at 13:32, on Zulip):

So, triaging #61960 as P-high, removing nomination tag.

pnkfelix (Jun 20 2019 at 13:32, on Zulip):

next: "type_name ICE" #61894

centril (Jun 20 2019 at 13:33, on Zulip):

Should be P-high if we want to stabilize

pnkfelix (Jun 20 2019 at 13:33, on Zulip):

currently relies on feature(core_intrinsics) ...

pnkfelix (Jun 20 2019 at 13:34, on Zulip):

but as @centril says, the stabilization of type_name becomes an issue

pnkfelix (Jun 20 2019 at 13:35, on Zulip):

okay, and the FCP (w/ inclination to merge) for RFC 1428 recently ended like 3 days ago.

pnkfelix (Jun 20 2019 at 13:35, on Zulip):

so yes, seems like a blocker for stabilization

pnkfelix (Jun 20 2019 at 13:35, on Zulip):

but also I think I'd prefer for us to immediately prioiritize things related to the release

eddyb (Jun 20 2019 at 13:36, on Zulip):

oh I think I forgot to explain to @oli that you can have parameterization in scope that's not part of monomorphization. for name mangling we produce something that demangles to _ in that case

pnkfelix (Jun 20 2019 at 13:36, on Zulip):

because there remain a large set of beta-regressions that I have not yet managed to get all the way through (see yet another topic)

centril (Jun 20 2019 at 13:37, on Zulip):

Maybe @eddyb would like to take on the ICE?

pnkfelix (Jun 20 2019 at 13:37, on Zulip):

okay well lets call this (#61894) P-high for now

eddyb (Jun 20 2019 at 13:38, on Zulip):

I'd say @oli but if he doesn't have time, me

pnkfelix (Jun 20 2019 at 13:38, on Zulip):

next: "Using extra lifetime parameter when implementing a trait is allowed" #61888

centril (Jun 20 2019 at 13:38, on Zulip):

@eddyb he doesn't

pnkfelix (Jun 20 2019 at 13:39, on Zulip):

I ... don't know what to say about #61888

pnkfelix (Jun 20 2019 at 13:40, on Zulip):

I'm going to assume that its nominated for lang team discussion

pnkfelix (Jun 20 2019 at 13:40, on Zulip):

next: "rustc runs out of memory when compiling an if statement with no block, enclosed in parentheses" #61858

pnkfelix (Jun 20 2019 at 13:41, on Zulip):

that's pretty awesome

pnkfelix (Jun 20 2019 at 13:41, on Zulip):

(the expression in question is (if foobar))

pnkfelix (Jun 20 2019 at 13:41, on Zulip):

I have to assume this is some sort of error recovery going terribly terribly wrong.

centril (Jun 20 2019 at 13:42, on Zulip):

/me looks at @Esteban Küber :slight_smile:

pnkfelix (Jun 20 2019 at 13:42, on Zulip):

but it also ... doesn't feel P-high to me.

pnkfelix (Jun 20 2019 at 13:42, on Zulip):

marking #61858 as P-medium for now. Though I do want to assign it. I'll assign to self and @Esteban Küber . And remove the nomation tag.

pnkfelix (Jun 20 2019 at 13:44, on Zulip):

next: " type parameter T/#1 (T/1) out of range when substituting" #61808

eddyb (Jun 20 2019 at 13:45, on Zulip):

looks like #61041 needs to be finished?

pnkfelix (Jun 20 2019 at 13:45, on Zulip):

@oli do you know if #61808 would be resolved by PR #61041 ?

pnkfelix (Jun 20 2019 at 13:46, on Zulip):

it would be nice to reduce it to a smaller standalone test case, in any case

pnkfelix (Jun 20 2019 at 13:46, on Zulip):

but oh does it rely on clippy anyway?

pnkfelix (Jun 20 2019 at 13:46, on Zulip):

I guess so. So never mind.

pnkfelix (Jun 20 2019 at 13:47, on Zulip):

I don't know how best to prioritize it. It feels P-medium at the moment but maybe @oli would convince me otherwise.

pnkfelix (Jun 20 2019 at 13:47, on Zulip):

I'll mark it P-medium and assign to self and @oli

pnkfelix (Jun 20 2019 at 13:48, on Zulip):

next: "Codegen ICE/regression with 2019-06-12 nightly when using async fn<T: Fn()>(&self, T)" #61793

pnkfelix (Jun 20 2019 at 13:49, on Zulip):

how are we currently triaging async-await bugs?

pnkfelix (Jun 20 2019 at 13:50, on Zulip):

I'm going to assign this to niko to let them figure out the approprite prioirty and any other async-await specific tags.

pnkfelix (Jun 20 2019 at 13:52, on Zulip):

next: "Size mismatch when const transmuting" #61774

centril (Jun 20 2019 at 13:52, on Zulip):

next: "Codegen ICE/regression with 2019-06-12 nightly when using async fn<T: Fn()>(&self, T)" #61793

Seems Async-Await-Blocking to me

pnkfelix (Jun 20 2019 at 13:52, on Zulip):

so, even though #61774 does ICE with or without the feature gate in place

pnkfelix (Jun 20 2019 at 13:53, on Zulip):

we do issue the diagnostic saying that it is feature gated if you are missing the feature gate

pnkfelix (Jun 20 2019 at 13:53, on Zulip):

which to me means this still falls under the umbrella of "gated"

pnkfelix (Jun 20 2019 at 13:53, on Zulip):

and thus not something to worry about in stable rustc

centril (Jun 20 2019 at 13:53, on Zulip):

P-medium + assign varkor?

pnkfelix (Jun 20 2019 at 13:54, on Zulip):

yeah, triage #61774 P-medium, remove nomination, assign self and varkor.

pnkfelix (Jun 20 2019 at 13:55, on Zulip):

next: "Regression: implementing Clone for trait objects" #61759

pnkfelix (Jun 20 2019 at 13:56, on Zulip):

Did we ever have a lint for "obvious" infinite recursion? I cannot remember.

nagisa (Jun 20 2019 at 13:57, on Zulip):

yes.

nagisa (Jun 20 2019 at 13:58, on Zulip):

well, not at a type level, but at regular code level.

pnkfelix (Jun 20 2019 at 13:58, on Zulip):

I'm not even sure if the infinite recursion here would qualify as "obvious" anyway

nagisa (Jun 20 2019 at 13:58, on Zulip):

https://doc.rust-lang.org/nightly/nightly-rustc/rustc_lint/builtin/static.UNCONDITIONAL_RECURSION.html

pnkfelix (Jun 20 2019 at 13:58, on Zulip):

anyway, stable-to-stable regression

pnkfelix (Jun 20 2019 at 13:59, on Zulip):

only question is whether to immediately write #61759 off as "expected fallout from #59500", or to dig deeper into the matter.

pnkfelix (Jun 20 2019 at 14:00, on Zulip):

okay well time is up, lets go over to the meeting.

Last update: Nov 16 2019 at 01:05UTC