Add a `fatal_cycle` attribute for queries which indicates that they will cause a fatal error on query cycles #47906

Zoxc · 2018-01-31T15:13:25Z

This moves us towards the goal of having cycle errors be non-fatal by not relying on the default implementation of ty::maps::values::Value which aborts on errors.

r? @nikomatsakis

Zoxc · 2018-01-31T15:16:00Z

src/librustc/ty/maps/plumbing.rs

@@ -193,7 +205,7 @@ macro_rules! define_maps {

        define_map_struct! {
            tcx: $tcx,
-            input: ($(([$($modifiers)*] [$($attr)*] [$name]))*)
+            input: ($(([$($attr)*] [$name]))*)


I had to remove this use of $modifiers otherwise I got this error:

error: no rules expected the token `nocycle` --> librustc\ty\maps\mod.rs:270:6 | 270 | [nocycle] fn is_panic_runtime: IsPanicRuntime(CrateNum) -> bool, | ^^^^^^^ error: aborting due to previous error

I'm not sure why this happens. define_map_struct should accept any tokens there. 😕

see my comment below :)

pietroalbini · 2018-02-05T10:44:41Z

@nikomatsakis this PR needs your review :)

eddyb · 2018-02-05T11:20:23Z

I've mentioned on IRC that I prefer calling this [fatal_cycle] and using the current default implementation of Value::from_cycle_error in those cases, to avoid regressing UX with it.

nikomatsakis · 2018-02-05T18:48:15Z

Sorry for being slow. I meant to leave a comment on this branch -- I guess in the end I don't have strong opinions here, but being able to identify queries that we really don't expect to participate in a cycle seems like a fine thing to me. I'd be tempted to reverse the default, and instead mark those queries that do expect to be in cycles, but @eddyb thinks that will wind up being brittle -- that maybe so. Still, I know there are a lot of queries where I've made no effort to make a nice "debug msg" precisely because I don't expect them to show up in cycle error messages. I guess I could mark all of those as no-cycle.

nikomatsakis · 2018-02-05T18:49:16Z

src/librustc/ty/maps/plumbing.rs

@@ -583,7 +595,7 @@ macro_rules! define_maps {

 macro_rules! define_map_struct {
    (tcx: $tcx:tt,
-     input: ($(([$(modifiers:tt)*] [$($attr:tt)*] [$name:ident]))*)) => {


the problem was here. This should have been:

$($modifiers:tt)*

note the $ before modifiers

nikomatsakis · 2018-02-05T18:51:30Z

I have no opinion about nocycle vs fatal_cycle -- well, I guess I mildly prefer fatal_cycle, since fatal carries rather specific connotations (and I prefer an _ between words).

r=me with the name changed.

I'm indifferent about whether we remove the [$($modifiers:tt)*] from define_map_struct -- we can always add it back in if we need it again.

…ill cause a fatal error on query cycles

Zoxc · 2018-02-10T04:53:48Z

@bors r=nikomatsakis

bors · 2018-02-10T04:53:49Z

📌 Commit e236994 has been approved by nikomatsakis

eddyb · 2018-02-10T08:28:41Z

@Zoxc Can you also update the PR title/description?

bors · 2018-02-13T03:05:53Z

⌛ Testing commit e236994 with merge dbc75ddfb1e9515b08a941111938dbf570dc5911...

bors · 2018-02-13T04:58:12Z

💔 Test failed - status-travis

kennytm · 2018-02-13T06:21:46Z

@bors retry #48116

bors · 2018-02-13T06:48:01Z

⌛ Testing commit e236994 with merge d5e730fbb979a1923adc034005740d3b8cbcc0d1...

bors · 2018-02-13T09:49:04Z

💔 Test failed - status-appveyor

kennytm · 2018-02-13T10:04:08Z

@bors retry

3 hour timeout

bors · 2018-02-13T12:53:58Z

⌛ Testing commit e236994 with merge dc971d16ed76f9e9485b642cd90616a24de23640...

bors · 2018-02-13T14:51:58Z

💔 Test failed - status-appveyor

pietroalbini · 2018-02-13T14:55:00Z

[01:55:45] failures:
[01:55:45]     [compile-fail] compile-fail\rfc-2126-extern-in-paths\single-segment.rs
[01:55:45]     [compile-fail] compile-fail\use-keyword.rs
[01:55:45]     [compile-fail] compile-fail\use-mod-2.rs

cc #48116 @kennytm

kennytm · 2018-02-13T14:56:29Z

@bors retry #48116

bors · 2018-02-17T02:58:27Z

⌛ Testing commit e236994 with merge 554fe71...

@nikomatsakis

Add a `fatal_cycle` attribute for queries which indicates that they will cause a fatal error on query cycles This moves us towards the goal of having cycle errors be non-fatal by not relying on the default implementation of `ty::maps::values::Value` which aborts on errors. r? @nikomatsakis

bors · 2018-02-17T05:53:50Z

☀️ Test successful - status-appveyor, status-travis
Approved by: nikomatsakis
Pushing 554fe71 to master...

rust-highfive assigned nikomatsakis Jan 31, 2018

kennytm added T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Jan 31, 2018

Zoxc commented Jan 31, 2018

View reviewed changes

nikomatsakis reviewed Feb 5, 2018

View reviewed changes

nikomatsakis added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Feb 6, 2018

Add a fatal_cycle attribute for queries which indicates that they w…

e236994

…ill cause a fatal error on query cycles

Zoxc force-pushed the nocycle branch from da6a434 to e236994 Compare February 10, 2018 02:38

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Feb 10, 2018

Zoxc changed the title ~~Add a nocycle attribute for queries which indicates that they cannot result in cycle errors~~ Add a fatal_cycle attribute for queries which indicates that they will cause a fatal error on query cycles Feb 10, 2018

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Feb 13, 2018

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Feb 13, 2018

kennytm mentioned this pull request Feb 13, 2018

Spurious (?) error in "compile-fail\rfc-2126-extern-in-paths\single-segment.rs" etc #48116

Closed

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Feb 13, 2018

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Feb 13, 2018

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Feb 13, 2018

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Feb 13, 2018

bors merged commit e236994 into rust-lang:master Feb 17, 2018

aturon mentioned this pull request Aug 16, 2019

Audit sources of shared state across rustc #63643

Open

37 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add a `fatal_cycle` attribute for queries which indicates that they will cause a fatal error on query cycles #47906

Add a `fatal_cycle` attribute for queries which indicates that they will cause a fatal error on query cycles #47906

Zoxc commented Jan 31, 2018

Zoxc Jan 31, 2018

nikomatsakis Feb 5, 2018

pietroalbini commented Feb 5, 2018

eddyb commented Feb 5, 2018 •

edited

Loading

nikomatsakis commented Feb 5, 2018

nikomatsakis Feb 5, 2018

Zoxc Feb 5, 2018

nikomatsakis commented Feb 5, 2018

Zoxc commented Feb 10, 2018

bors commented Feb 10, 2018

eddyb commented Feb 10, 2018

bors commented Feb 13, 2018

bors commented Feb 13, 2018

kennytm commented Feb 13, 2018

bors commented Feb 13, 2018

bors commented Feb 13, 2018

kennytm commented Feb 13, 2018

bors commented Feb 13, 2018

bors commented Feb 13, 2018

pietroalbini commented Feb 13, 2018

kennytm commented Feb 13, 2018

bors commented Feb 17, 2018

bors commented Feb 17, 2018

Add a fatal_cycle attribute for queries which indicates that they will cause a fatal error on query cycles #47906

Add a fatal_cycle attribute for queries which indicates that they will cause a fatal error on query cycles #47906

Conversation

Zoxc commented Jan 31, 2018

Zoxc Jan 31, 2018

Choose a reason for hiding this comment

nikomatsakis Feb 5, 2018

Choose a reason for hiding this comment

pietroalbini commented Feb 5, 2018

eddyb commented Feb 5, 2018 • edited Loading

nikomatsakis commented Feb 5, 2018

nikomatsakis Feb 5, 2018

Choose a reason for hiding this comment

Zoxc Feb 5, 2018

Choose a reason for hiding this comment

nikomatsakis commented Feb 5, 2018

Zoxc commented Feb 10, 2018

bors commented Feb 10, 2018

eddyb commented Feb 10, 2018

bors commented Feb 13, 2018

bors commented Feb 13, 2018

kennytm commented Feb 13, 2018

bors commented Feb 13, 2018

bors commented Feb 13, 2018

kennytm commented Feb 13, 2018

bors commented Feb 13, 2018

bors commented Feb 13, 2018

pietroalbini commented Feb 13, 2018

kennytm commented Feb 13, 2018

bors commented Feb 17, 2018

bors commented Feb 17, 2018

Add a `fatal_cycle` attribute for queries which indicates that they will cause a fatal error on query cycles #47906

Add a `fatal_cycle` attribute for queries which indicates that they will cause a fatal error on query cycles #47906

eddyb commented Feb 5, 2018 •

edited

Loading