[WIP] remove `Location::All` #50938

nikomatsakis · 2018-05-21T13:26:58Z

When adding the facts for the new analysis, I added this idea of a Locations::All which enforces an outlives relation at all points in the CFG. This was kind of a hack to deal with a problem in the return slot _0. What would happen is this. We would unify the type of _0 with the return type from the signature at the entry point of the function. Then later, we would write to _0:

...
_0 = ...
RETURN

This means that _0 — which is otherwise treated like a regular local variable — would be dead at the entry point. In the existing analysis, in part because of its imprecision, this still works out ok, but in the new analysis it did not: in effect, the type of _0 was severed from the signature. So the hack I added basically re-equates the type of _0 (and all parameters) with the signature at all points in the CFG. But this winds up adding a ton of redundant outlives relations (hence rust-lang/polonius#24).

So now I am taking a different approach. I equate the return type with the type from the signature only at the return location. This has the same soundness effect but without a hacky notion like Location::All. In the case of yields, I equate at all points. The argument types are equated on function entry.

I think this is the right general direction; we may want to think about the use of liveness in general, as it does mean that something like this will type-check:

fn foo(mut x: &'static u32) {
    let p = 22;
    x = &p; // here the old value of `x` is dead hence the region takes on a new, shorter value
    println!("x = {}", x);
}

This might seem surprising, since the type of x is declared as &'static u32, but &p cannot live for 'static lifetime. But it works because x is dead at the point x = &p and hence it effectively gets a "fresh set of regions" from its initial value.

I think we can address this concern in other ways, though, and not via the Locations::All hack (e.g., perhaps we should not use liveness to decide when regions in local variable types go dead, but rather going out of scope?). But anyway I'd rather we focus on that question in a separate PR.

I'm marking this PR as WIP, because IIRC it still had a few problems in travis, but I wanted to get it posted for feedback purposes.

r? @pnkfelix
cc @rust-lang/wg-compiler-nll

rust-highfive · 2018-05-21T14:13:41Z

The job x86_64-gnu-llvm-3.9 of your PR failed on Travis (raw log). Through arcane magic we have determined that the following fragments from the build log may contain information about the problem.

Click to expand the log.

[00:44:38] ....................................................................................................
[00:44:43] ....................................................................................................
[00:44:48] ....................................................................................................
[00:44:53] .....................................................................i..............................
[00:44:57] ..........F...................................i.....................................................
[00:45:07] ....................................................................................................
[00:45:14] ...................................................................i................................
[00:45:16] .............................................
[00:45:16] failures:
[00:45:16] failures:
[00:45:16] 
[00:45:16] thread 'main' panicked at 'Some tests failed', tools/compiletest/src/main.rs:498:22
[00:45:16] ---- [ui] ui/nll/generator-distinct-lifetime.rs stdout ----
[00:45:16] 
[00:45:16] error: test compilation failed although it shouldn't!
[00:45:16] status: exit code: 101
[00:45:16] command: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin/rustc" "/checkout/src/test/ui/nll/generator-distinct-lifetime.rs" "--target=x86_64-unknown-linux-gnu" "--error-format" "json" "-Zui-testing" "-C" "prefer-dynamic" "-o" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/nll/generator-distinct-lifetime/a" "-Crpath" "-O" "-Zunstable-options" "-Lnative=/checkout/obj/build/x86_64-unknown-linux-gnu/native/rust-test-helpers" "-L" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/nll/generator-distinct-lifetime/auxiliary" "-A" "unused"
[00:45:16] ------------------------------------------
[00:45:16] 
[00:45:16] ------------------------------------------
[00:45:16] stderr:
[00:45:16] stderr:
[00:45:16] ------------------------------------------
[00:45:16] {"message":"librustc_mir/transform/generator.rs:494: Broken MIR: generator contains type std::ops::GeneratorState<(), ()> in MIR, but typeck only knows about for<'r, 's, 't0> {&'r mut u32, u32, &'s mut u32, (), &'t0 mut u32}","code":null,"level":"error: internal compiler error","spans":[{"file_name":"/checkout/src/test/ui/nll/generator-distinct-lifetime.rs","byte_start":835,"byte_end":965,"line_start":22,"line_end":30,"column_start":5,"column_end":6,"is_primary":true,"text":[{"text":"    move || {","highlight_start":5,"highlight_end":14},{"text":"        let s = &mut *x;","highlight_start":1,"highlight_end":25},{"text":"        yield;","highlight_start":1,"highlight_end":15},{"text":"        *s += 1;","highlight_start":1,"highlight_end":17},{"text":"","highlight_start":1,"highlight_end":1},{"text":"        let t = &mut *x;","highlight_start":1,"highlight_end":25},{"text":"        yield;","highlight_start":1,"highlight_end":15},{"text":"        *t += 1;","highlight_start":1,"highlight_end":17},{"text":"    };","highlight_start":1,"highlight_end":6}],"label":null,"suggested_replacement":null,"expansion":null}],"children":[],"rendered":"error: internal compiler error: librustc_mir/transform/generator.rs:494: Broken MIR: generator contains type std::ops::GeneratorState<(), ()> in MIR, but typeck only knows about for<'r, 's, 't0> {&'r mut u32, u32, &'s mut u32, (), &'t0 mut u32}\n  --> /checkout/src/test/ui/nll/generator-distinct-lifetime.rs:22:5\n   |\nLL | /     move || {\nLL | |         let s = &mut *x;\nLL | |         yield;\nLL | |         *s += 1;\n...  |\nLL | |         *t += 1;\nLL | |     };\n   | |_____^\n\n"}
[00:45:16] note: Run with `RUST_BACKTRACE=1` for a backtrace.
[00:45:16] note: Run with `RUST_BACKTRACE=1` for a backtrace.
[00:45:16] {"message":"aborting due to previous error","code":null,"level":"error","spans":[],"children":[],"rendered":"error: aborting due to previous error\n\n"}
[00:45:16] 
[00:45:16] note: the compiler unexpectedly panicked. this is a bug.
[00:45:16] 
[00:45:16] note: we would appreciate a bug report: https://github.com/rust-lang/rust/blob/master/CONTRIBUTING.md#bug-reports
[00:45:16] note: rustc 1.28.0-dev running on x86_64-unknown-linux-gnu
[00:45:16] 
[00:45:16] 
[00:45:16] note: compiler flags: -Z ui-testing -Z unstable-options -C prefer-dynamic -C rpath
[00:45:16] 
[00:45:16] ------------------------------------------
[00:45:16] 
[00:45:16] thread '[ui] ui/nll/generator-distinct-lifetime.rs' panicked at 'explicit panic', tools/compiletest/src/runtest.rs:3042:9
---
[00:45:16] test result: FAILED. 1437 passed; 1 failed; 7 ignored; 0 measured; 0 filtered out
[00:45:16] 
[00:45:16] 
[00:45:16] 
[00:45:16] command did not execute successfully: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0-tools-bin/compiletest" "--compile-lib-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/lib" "--run-lib-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/lib/rustlib/x86_64-unknown-linux-gnu/lib" "--rustc-path" "/checkout/obj/build/x86_64-unknown-linux-

The command "stamp sh -x -c "$RUN_SCRIPT"" exited with 2.
travis_time:start:0eb197a8
$ date && (curl -fs --head https://google.com | grep ^Date: | sed 's/Date: //g' || true)

I'm a bot! I can only do what humans tell me to, so if this was not helpful or you have suggestions for improvements, please ping or otherwise contact @TimNN. (Feature Requests)

rust-highfive · 2018-05-23T20:21:22Z

The job x86_64-gnu-llvm-3.9 of your PR failed on Travis (raw log). Through arcane magic we have determined that the following fragments from the build log may contain information about the problem.

Click to expand the log.

[00:47:11] ....................................................................................................
[00:47:16] ....................................................................................................
[00:47:20] ....................................................................................................
[00:47:26] ...........................................................................i........................
[00:47:30] ................F...................................i...............................................
[00:47:40] ....................................................................................................
[00:47:46] ............................................................................i.................iiiiii
[00:47:48] iii...................................................
[00:47:48] failures:
[00:47:48] failures:
[00:47:48] 
[00:47:48] ---- [ui] ui/nll/generator-distinct-lifetime.rs stdout ----
[00:47:48] 
[00:47:48] error: test compilation failed although it shouldn't!
[00:47:48] status: exit code: 101
[00:47:48] command: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin/rustc" "/checkout/src/test/ui/nll/generator-distinct-lifetime.rs" "--target=x86_64-unknown-linux-gnu" "--error-format" "json" "-Zui-testing" "-C" "prefer-dynamic" "-o" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/nll/generator-distinct-lifetime/a" "-Crpath" "-O" "-Zunstable-options" "-Lnative=/checkout/obj/build/x86_64-unknown-linux-gnu/native/rust-test-helpers" "-L" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/nll/generator-distinct-lifetime/auxiliary" "-A" "unused"
[00:47:48] ------------------------------------------
[00:47:48] 
[00:47:48] ------------------------------------------
[00:47:48] stderr:
[00:47:48] stderr:
[00:47:48] ------------------------------------------
[00:47:48] {"message":"librustc_mir/transform/generator.rs:494: Broken MIR: generator contains type std::ops::GeneratorState<(), ()> in MIR, but typeck only knows about for<'r, 's, 't0> {&'r mut u32, u32, &'s mut u32, (), &'t0 mut u32}","code":null,"level":"error: internal compiler error","spans":[{"file_name":"/checkout/src/test/ui/nll/generator-distinct-lifetime.rs","byte_start":835,"byte_end":965,"line_start":22,"line_end":30,"column_start":5,"column_end":6,"is_primary":true,"text":[{"text":"    move || {","highlight_start":5,"highlight_end":14},{"text":"        let s = &mut *x;","highlight_start":1,"highlight_end":25},{"text":"        yield;","highlight_start":1,"highlight_end":15},{"text":"        *s += 1;","highlight_start":1,"highlight_end":17},{"text":"","highlight_start":1,"highlight_end":1},{"text":"        let t = &mut *x;","highlight_start":1,"highlight_end":25},{"text":"        yield;","highlight_start":1,"highlight_end":15},{"text":"        *t += 1;","highlight_start":1,"highlight_end":17},{"text":"    };","highlight_start":1,"highlight_end":6}],"label":null,"suggested_replacement":null,"suggestion_applicability":null,"expansion":null}],"children":[],"rendered":"error: internal compiler error: librustc_mir/transform/generator.rs:494: Broken MIR: generator contains type std::ops::GeneratorState<(), ()> in MIR, but typeck only knows about for<'r, 's, 't0> {&'r mut u32, u32, &'s mut u32, (), &'t0 mut u32}\n  --> /checkout/src/test/ui/nll/generator-distinct-lifetime.rs:22:5\n   |\nLL | /     move || {\nLL | |         let s = &mut *x;\nLL | |         yield;\nLL | |         *s += 1;\n...  |\nLL | |         *t += 1;\nLL | |     };\n   | |_____^\n\n"}
[00:47:48] note: Run with `RUST_BACKTRACE=1` for a backtrace.
[00:47:48] note: Run with `RUST_BACKTRACE=1` for a backtrace.
[00:47:48] {"message":"aborting due to previous error","code":null,"level":"error","spans":[],"children":[],"rendered":"error: aborting due to previous error\n\n"}
[00:47:48] 
[00:47:48] note: the compiler unexpectedly panicked. this is a bug.
[00:47:48] 
[00:47:48] note: we would appreciate a bug report: https://github.com/rust-lang/rust/blob/master/CONTRIBUTING.md#bug-reports
[00:47:48] note: rustc 1.28.0-dev running on x86_64-unknown-linux-gnu
[00:47:48] 
[00:47:48] 
[00:47:48] note: compiler flags: -Z ui-testing -Z unstable-options -C prefer-dynamic -C rpath
[00:47:48] 
[00:47:48] ------------------------------------------
[00:47:48] 
[00:47:48] thread '[ui] ui/nll/generator-distinct-lifetime.rs' panicked at 'explicit panic', tools/compiletest/src/runtest.rs:3044:9
---
[00:47:48] 
[00:47:48] thread 'main' panicked at 'Some tests failed', tools/compiletest/src/main.rs:498:22
[00:47:48] 
[00:47:48] 
[00:47:48] command did not execute successfully: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0-tools-bin/compiletest" "--compile-lib-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/lib" "--run-lib-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/lib/rustlib/x86_64-unknown-linux-gnu/lib" "--rustc-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin/rustc" "--src-base" "/checkout/src/test/ui" "--build-base" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui" "--stage-id" "stage2-x86_64-unknown-linux-gnu" "--mode" "ui" "--target" "x86_64-unknown-linux-gnu" "--host" "x86_64-unknown-linux-gnu" "--llvm-filecheck" "/usr/lib/llvm-3.9/bin/FileCheck" "--host-rustcflags" "-Crpath -O -Zunstable-options " "--target-rustcflags" "-Crpath -O -Zunstable-options  -Lnative=/checkout/obj/build/x86_64-unknown-linux-gnu/native/rust-test-helpers" "--docck-python" "/usr/bin/python2.7" "--lldb-python" "/usr/bin/python2.7" "--gdb" "/usr/bin/gdb" "--quiet" "--llvm-version" "3.9.1\n" "--system-llvm" "--cc" "" "--cxx" "" "--cflags" "" "--llvm-components" "" "--llvm-cxxflags" "" "--adb-path" "adb" "--adb-test-dir" "/data/tmp/work" "--android-cross-path" "" "--color" "always"
[00:47:48] 
[00:47:48] 
[00:47:48] failed to run: /checkout/obj/build/bootstrap/debug/bootstrap test
[00:47:48] Build completed unsuccessfully in 0:02:34
[00:47:48] Build completed unsuccessfully in 0:02:34
[00:47:48] Makefile:58: recipe for target 'check' failed
[00:47:48] make: *** [check] Error 1

The command "stamp sh -x -c "$RUN_SCRIPT"" exited with 2.
travis_time:start:1ab378ec
$ date && (curl -fs --head https://google.com | grep ^Date: | sed 's/Date: //g' || true)

I'm a bot! I can only do what humans tell me to, so if this was not helpful or you have suggestions for improvements, please ping or otherwise contact @TimNN. (Feature Requests)

bors · 2018-05-26T14:37:30Z

☔ The latest upstream changes (presumably #51082) made this pull request unmergeable. Please resolve the merge conflicts.

pnkfelix · 2018-05-30T10:15:36Z

This might seem surprising, since the type of x is declared as &'static u32, but &p cannot live for 'static lifetime. But it works because x is dead at the point x = &p and hence it effectively gets a "fresh set of regions" from its initial value.

I think we can address this concern in other ways, though, and not via the Locations::All hack (e.g., perhaps we should not use liveness to decide when regions in local variable types go dead, but rather going out of scope?). But anyway I'd rather we focus on that question in a separate PR.

I found the above interesting.

Do you think the way we might address that concern is perhaps related to finding a complete resolution to #47184 ?

pnkfelix · 2018-05-30T11:06:01Z

src/librustc_mir/borrow_check/nll/subtype_constraint_generation.rs

@@ -112,23 +108,19 @@ impl<'cx, 'tcx> SubtypeConstraintGenerator<'cx, 'tcx> {
                // reverse direction, because `regioncx` talks about
                // "outlives" (`>=`) whereas the region constraints
                // talk about `<=`.
-                self.regioncx.add_outlives(span, b_vid, a_vid, at_location);
+                self.regioncx.add_outlives(span, b_vid, a_vid);


Am I right in my understanding that one side-effect of no longer passing at_location into add_outlives is that our diagnostics stop being able to point to this location as the place where a particular constraint was introduced, which is what is leading to problems like the regression in diagnostic quality for e.g. a904c55#diff-40124ba66f06953794b406e3bb131b95L6 ?)

(And if that hypothesis is correct, is your thinking that we're going to have to solve that problem more generally anyway, perhaps as part of #51188, but that finding a solution to that problem should not block landing PR's like this one?)

hey @nikomatsakis do you have an answer to my Q above ?

pnkfelix · 2018-05-30T11:08:15Z

src/librustc_mir/borrow_check/nll/type_check/input_output.rs

        );
        if let Some(mir_yield_ty) = mir.yield_ty {
            let ur_yield_ty = universal_regions.yield_ty.unwrap();
-            self.equate_normalized_input_or_output(ur_yield_ty, mir_yield_ty);
+            self.equate_normalized_input_or_output(Locations::Yield, ur_yield_ty, mir_yield_ty);


I think I need a comment here saying this is the place where effectively you are equating at all points. (or am I even right about that?)

@nikomatsakis I still would like to see a comment in the source code about the effect of this change. (Either my hypothesis is right, which means its a subtle effect. Or my hypothesis is wrong, which means I need to figure out where the equate-at-all-points is happening for Yield...)

pnkfelix · 2018-05-30T11:11:14Z

So this all seems acceptable to me.

Its a shame about the regressions in diagnostic quality, but we have big problems to solve there on other test cases anyway, and thus I'm willing to assume we'll work on getting the quality back in these cases when we address those.

There aren't that many universal regions. This was a premature optimization and it makes the code much messier.

rust-highfive · 2018-05-30T15:06:38Z

The job x86_64-gnu-llvm-3.9 of your PR failed on Travis (raw log). Through arcane magic we have determined that the following fragments from the build log may contain information about the problem.

Click to expand the log.

[00:47:39] ....................................................................................................
[00:47:44] ....................................................................................................
[00:47:48] ....................................................................................................
[00:47:54] .............................................................................................i......
[00:48:00] ..................................F...................................i.............................
[00:48:10] ....................................................................................................
[00:48:16] ...................................................................................................i
[00:48:16] ...................................................................................................i
[00:48:19] .................iiiiiiiii...................................................
[00:48:19] 
[00:48:19] ---- [ui] ui/nll/generator-distinct-lifetime.rs stdout ----
[00:48:19] 
[00:48:19] thread 'main' panicked at 'Some tests failed', tools/compiletest/src/main.rs:498:22
[00:48:19] thread 'main' panicked at 'Some tests failed', tools/compiletest/src/main.rs:498:22
[00:48:19] error: test compilation failed although it shouldn't!
[00:48:19] status: exit code: 101
[00:48:19] command: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin/rustc" "/checkout/src/test/ui/nll/generator-distinct-lifetime.rs" "--target=x86_64-unknown-linux-gnu" "--error-format" "json" "-Zui-testing" "-C" "prefer-dynamic" "-o" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/nll/generator-distinct-lifetime/a" "-Crpath" "-O" "-Zunstable-options" "-Lnative=/checkout/obj/build/x86_64-unknown-linux-gnu/native/rust-test-helpers" "-L" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/nll/generator-distinct-lifetime/auxiliary" "-A" "unused"
[00:48:19] ------------------------------------------
[00:48:19] 
[00:48:19] ------------------------------------------
[00:48:19] stderr:
[00:48:19] stderr:
[00:48:19] ------------------------------------------
[00:48:19] {"message":"librustc_mir/transform/generator.rs:494: Broken MIR: generator contains type std::ops::GeneratorState<(), ()> in MIR, but typeck only knows about for<'r, 's, 't0> {&'r mut u32, u32, &'s mut u32, (), &'t0 mut u32}","code":null,"level":"error: internal compiler error","spans":[{"file_name":"/checkout/src/test/ui/nll/generator-distinct-lifetime.rs","byte_start":835,"byte_end":965,"line_start":22,"line_end":30,"column_start":5,"column_end":6,"is_primary":true,"text":[{"text":"    move || {","highlight_start":5,"highlight_end":14},{"text":"        let s = &mut *x;","highlight_start":1,"highlight_end":25},{"text":"        yield;","highlight_start":1,"highlight_end":15},{"text":"        *s += 1;","highlight_start":1,"highlight_end":17},{"text":"","highlight_start":1,"highlight_end":1},{"text":"        let t = &mut *x;","highlight_start":1,"highlight_end":25},{"text":"        yield;","highlight_start":1,"highlight_end":15},{"text":"        *t += 1;","highlight_start":1,"highlight_end":17},{"text":"    };","highlight_start":1,"highlight_end":6}],"label":null,"suggested_replacement":null,"suggestion_applicability":null,"expansion":null}],"children":[],"rendered":"error: internal compiler error: librustc_mir/transform/generator.rs:494: Broken MIR: generator contains type std::ops::GeneratorState<(), ()> in MIR, but typeck only knows about for<'r, 's, 't0> {&'r mut u32, u32, &'s mut u32, (), &'t0 mut u32}\n  --> /checkout/src/test/ui/nll/generator-distinct-lifetime.rs:22:5\n   |\nLL | /     move || {\nLL | |         let s = &mut *x;\nLL | |         yield;\nLL | |         *s += 1;\n...  |\nLL | |         *t += 1;\nLL | |     };\n   | |_____^\n\n"}
[00:48:19] note: Run with `RUST_BACKTRACE=1` for a backtrace.
[00:48:19] note: Run with `RUST_BACKTRACE=1` for a backtrace.
[00:48:19] {"message":"aborting due to previous error","code":null,"level":"error","spans":[],"children":[],"rendered":"error: aborting due to previous error\n\n"}
[00:48:19] 
[00:48:19] note: the compiler unexpectedly panicked. this is a bug.
[00:48:19] 
[00:48:19] note: we would appreciate a bug report: https://github.com/rust-lang/rust/blob/master/CONTRIBUTING.md#bug-reports
[00:48:19] note: rustc 1.28.0-dev running on x86_64-unknown-linux-gnu
[00:48:19] 
[00:48:19] 
[00:48:19] note: compiler flags: -Z ui-testing -Z unstable-options -C prefer-dynamic -C rpath
[00:48:19] 
[00:48:19] ------------------------------------------
[00:48:19] 
[00:48:19] thread '[ui] ui/nll/generator-distinct-lifetime.rs' panicked at 'explicit panic', tools/compiletest/src/runtest.rs:3053:9
---
[00:48:19] test result: FAILED. 1462 passed; 1 failed; 14 ignored; 0 measured; 0 filtered out
[00:48:19] 
[00:48:19] 
[00:48:19] 
[00:48:19] command did not execute successfully: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0-tools-bin/compiletest" "--compile-lib-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/lib" "--run-lib-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/lib/rustlib/x86_64-unknown-linux-gnu/lib" "--rustc-path" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage2/bin/rustc" "--src-base" "/checkout/src/test/ui" "--build-base" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui" "--stage-id" "stage2-x86_64-unknown-linux-gnu" "--mode" "ui" "--target" "x86_64-unknown-linux-gnu" "--host" "x86_64-unknown-linux-gnu" "--llvm-filecheck" "/usr/lib/llvm-3.9/bin/FileCheck" "--host-rustcflags" "-Crpath -O -Zunstable-options " "--target-rustcflags" "-Crpath -O -Zunstable-options  -Lnative=/checkout/obj/build/x86_64-unknown-linux-gnu/native/rust-test-helpers" "--docck-python" "/usr/bin/python2.7" "--lldb-python" "/usr/bin/python2.7" "--gdb" "/usr/bin/gdb" "--quiet" "--llvm-version" "3.9.1\n" "--system-llvm" "--cc" "" "--cxx" "" "--cflags" "" "--llvm-components" "" "--llvm-cxxflags" "" "--adb-path" "adb" "--adb-test-dir" "/data/tmp/work" "--android-cross-path" "" "--color" "always"
[00:48:19] 
[00:48:19] 
[00:48:19] failed to run: /checkout/obj/build/bootstrap/debug/bootstrap test
[00:48:19] Build completed unsuccessfully in 0:02:40
[00:48:19] Build completed unsuccessfully in 0:02:40
[00:48:19] make: *** [check] Error 1
[00:48:19] Makefile:58: recipe for target 'check' failed

The command "stamp sh -x -c "$RUN_SCRIPT"" exited with 2.
travis_time:start:00b1c4dc
$ date && (curl -fs --head https://google.com | grep ^Date: | sed 's/Date: //g' || true)

I'm a bot! I can only do what humans tell me to, so if this was not helpful or you have suggestions for improvements, please ping or otherwise contact @TimNN. (Feature Requests)

pietroalbini · 2018-06-04T10:49:33Z

Ping from triage @pnkfelix! The author pushed new commits.

pnkfelix · 2018-06-04T12:55:49Z

src/test/ui/nll/closure-requirements/region-lbr-anon-does-not-outlive-static.stderr

-LL |     &*x
-   |     ^^^ lifetime `ReStatic` required
+LL |   fn foo(x: &u32) -> &'static u32 {
+   |   ^      - consider changing the type of `x` to `&ReStatic u32`


(this sort of thing is the reason I'm not really thrilled with ui/ tests that use -Z verbose ... though I guess seeing ReStatic in supposed user output is not as egregious as other potential leakage of internal details...)

pnkfelix · 2018-06-04T12:57:52Z

src/test/ui/nll/ty-outlives/impl-trait-captures.stderr

-LL |     x
-   |     ^ lifetime `ReEarlyBound(0, 'a)` required
+LL |   fn foo<'a, T>(x: &T) -> impl Foo<'a> {
+   |   ^             - consider changing the type of `x` to `&ReEarlyBound(0, 'a) T`


Like a similar comment elsewhere: this sort of thing is the reason I'm not really thrilled with ui/ tests that use -Z verbose

I recognize its a pre-existing issue with the test, but I think it would clean things up if you took the -Z verbose out the test file. (Unless you think the output is serving a useful purpose here? Provides hint to compiler devs about the internal region structure in play?)

pnkfelix · 2018-06-04T13:00:29Z

This all still seems fine. I had some questions about the code. and of course travis is faillng (I think due to a panic from ui/nll/generator-distinct-lifetime.rs?), so that needs to be addressed.

But mostly I think the PR is still in @nikomatsakis 's court. Its still marked as WIP.

pietroalbini · 2018-06-18T09:04:00Z

Ping from triage @nikomatsakis! It's been a while since we heard from you on this, will you have time to work on this again?

bors · 2018-06-18T19:00:37Z

☔ The latest upstream changes (presumably #51460) made this pull request unmergeable. Please resolve the merge conflicts.

nikomatsakis · 2018-06-19T18:47:58Z

I will get back to it, but right now it's relatively low priority as this is more concerned with polonius than anything else I guess.

pietroalbini · 2018-06-25T08:51:17Z

Ping from triage @nikomatsakis! Are you ok with closing this until you have time for it?

nikomatsakis · 2018-06-29T14:10:08Z

I'm going to close this PR until I have time to get back to it 😿

rust-highfive assigned pnkfelix May 21, 2018

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label May 21, 2018

nikomatsakis mentioned this pull request May 21, 2018

Revisit the "Location::All" hack rust-lang/polonius#24

Open

nikomatsakis force-pushed the nll-alias-analysis-no-loc-all branch from 377430c to a904c55 Compare May 23, 2018 19:32

nikomatsakis mentioned this pull request May 28, 2018

NLL ignores lifetimes bounds derived from Sized requirements #50716

Closed

pnkfelix reviewed May 30, 2018

View reviewed changes

nikomatsakis added 7 commits May 30, 2018 10:15

remove point from outlives

1b10566

remove point from TypeTest

cbdb330

add a span method to Locations

09df848

kill Locations::All and move towards specific locations

9d40370

replace universal_region with adding entries to region_live_at

9a34822

There aren't that many universal regions. This was a premature optimization and it makes the code much messier.

make Return terminators "use" the RETURN_PLACE

3ff3116

update tests with new error messages (mostly worse, I'm afraid)

fd78a2f

nikomatsakis force-pushed the nll-alias-analysis-no-loc-all branch from a904c55 to fd78a2f Compare May 30, 2018 14:16

pnkfelix reviewed Jun 4, 2018

View reviewed changes

pnkfelix 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 Jun 4, 2018

TimNN added A-allocators Area: Custom and system allocators and removed A-allocators Area: Custom and system allocators labels Jun 12, 2018

nikomatsakis closed this Jun 29, 2018

ecstatic-morse mentioned this pull request Nov 9, 2021

Can I create and use a *mut to an uninitialized local? #90718

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[WIP] remove `Location::All` #50938

[WIP] remove `Location::All` #50938

nikomatsakis commented May 21, 2018

rust-highfive commented May 21, 2018

rust-highfive commented May 23, 2018

bors commented May 26, 2018

pnkfelix commented May 30, 2018

pnkfelix May 30, 2018

pnkfelix Jun 4, 2018

pnkfelix May 30, 2018

pnkfelix Jun 4, 2018 •

edited

Loading

pnkfelix commented May 30, 2018

rust-highfive commented May 30, 2018

pietroalbini commented Jun 4, 2018

pnkfelix Jun 4, 2018

pnkfelix Jun 4, 2018

pnkfelix commented Jun 4, 2018

pietroalbini commented Jun 18, 2018

bors commented Jun 18, 2018

nikomatsakis commented Jun 19, 2018

pietroalbini commented Jun 25, 2018

nikomatsakis commented Jun 29, 2018

[WIP] remove Location::All #50938

[WIP] remove Location::All #50938

Conversation

nikomatsakis commented May 21, 2018

rust-highfive commented May 21, 2018

rust-highfive commented May 23, 2018

bors commented May 26, 2018

pnkfelix commented May 30, 2018

pnkfelix May 30, 2018

Choose a reason for hiding this comment

pnkfelix Jun 4, 2018

Choose a reason for hiding this comment

pnkfelix May 30, 2018

Choose a reason for hiding this comment

pnkfelix Jun 4, 2018 • edited Loading

Choose a reason for hiding this comment

pnkfelix commented May 30, 2018

rust-highfive commented May 30, 2018

pietroalbini commented Jun 4, 2018

pnkfelix Jun 4, 2018

Choose a reason for hiding this comment

pnkfelix Jun 4, 2018

Choose a reason for hiding this comment

pnkfelix commented Jun 4, 2018

pietroalbini commented Jun 18, 2018

bors commented Jun 18, 2018

nikomatsakis commented Jun 19, 2018

pietroalbini commented Jun 25, 2018

nikomatsakis commented Jun 29, 2018

[WIP] remove `Location::All` #50938

[WIP] remove `Location::All` #50938

pnkfelix Jun 4, 2018 •

edited

Loading