Deduplicate pretty printing of constants #67133

oli-obk · 2019-12-08T00:29:09Z

r? @eddyb for the pretty printing logic
cc @RalfJung

src/librustc/ty/context.rs

src/test/mir-opt/const_prop/discriminant.rs

src/librustc/ty/print/pretty.rs

RalfJung · 2019-12-09T12:21:20Z

src/librustc/ty/print/pretty.rs

@@ -1425,6 +1492,19 @@ impl<F: fmt::Write> PrettyPrinter<'tcx> for FmtPrinter<'_, 'tcx, F> {
            ty::ReClosureBound(_) => true,
        }
    }
+
+    fn pretty_print_const_pointer(


So is this used both for MIR dumps and something user-visible? Or why do we only sometimes print the AllocId?

Yes, the pretty printer is used for both MIR dumps and error messages, and the "show alloc id" flag is only set by MIR dumps

But there's also something it is not used for and that goes through the other pretty_print_const_pointer?

Yes, the SymbolPrinter and the AbsolutePathPrinter. The latter is used for type_name, which we probably don't want to make users see the alloc ids.

SymbolPrinter... May actually need the ids. cc @eddyb

SymbolPrinter... May actually need the ids. cc @eddyb

Which one? legacy symbols don't matter much, the hash contains all of the information, they should just be consistent across crates. You don't have to show the const's value if you can't do it consistently cross-crate.

v0 ones are trickier, and I feel like the push for #[structural_match] in const generics is going to force us to implement some sort of general ADT representation for constants in v0 symbols, sooner than later.
(it'd likely be similar to how all values of #[structural_match] types can technically be turned into some sort of pattern)

cc @michaelwoerister @varkor @yodaldevoid

For reference, the dedicated issue for the last comment is #61486.

I could set it to print just integer types for now and bug! out for other types so we don't silently introduce bugs

src/librustc/mir/mod.rs

src/librustc/ty/print/pretty.rs

RalfJung · 2019-12-09T21:30:16Z

src/librustc/ty/print/pretty.rs

+            (Scalar::Raw { data: 0, .. }, ty::RawPtr(_)) => p!(write("{{null pointer}}")),
+            // This is UB, but we still print it
+            (Scalar::Raw { data: 0, ..}, ty::Ref(_, ty, _))
+                => p!(write("{{null reference to "), print(ty), write("}}")),
+            (Scalar::Raw { data, .. }, ty::Ref(..)) |
+            (Scalar::Raw { data, .. }, ty::RawPtr(_)) => {
+                let pointer_width = self.tcx().data_layout.pointer_size.bytes();
+                p!(write("0x{:01$x}", data, pointer_width as usize * 2))
+            },
+            (Scalar::Ptr(p), ty::Ref(..)) |
+            (Scalar::Ptr(p), ty::RawPtr(_)) => self = self.pretty_print_const_pointer(p)?,


So here we (mostly) do not print types. I am fine with that, I think, but it seems somewhat inconsistent as we are otherwise so careful to have type information (e.g. with the integer types).

It is inconsistent, it just seemed very noisy considering that there are no size differences that can be relevant like with the integer types.

src/test/ui/const-generics/raw-ptr-const-param.rs

src/test/mir-opt/const_prop/discriminant.rs

src/librustc/mir/mod.rs

src/librustc/ty/print/pretty.rs

src/test/mir-opt/const_prop/read_immutable_static.rs

src/librustc/ty/print/pretty.rs

eddyb · 2020-03-11T15:58:28Z

I can't easily check old comments, but I think #67133 (comment) and #67133 (comment) are the only remaining ones, and after that you can r=me.

src/librustc/ty/print/pretty.rs

oli-obk · 2020-03-13T09:56:49Z

@bors r=eddyb

bors · 2020-03-13T09:56:51Z

📌 Commit 6ca65bd has been approved by eddyb

bors · 2020-03-13T09:56:51Z

🌲 The tree is currently closed for pull requests below priority 1000, this pull request will be tested once the tree is reopened

@eddyb

Add support for LLVM globals corresponding to miri allocations should be named alloc123 Adds support for this request from @eddyb in #69134: > That is, if -Zfewer-names is false (usually only because of --emit=llvm-ir), we should use the same name for LLVM globals we generate out of miri allocs as #67133 does in MIR output (allocN). > >This way, we can easily see the mapping between MIR and LLVM IR (and it shouldn't be any costlier for regular compilation, which would continue to use unnamed globals). r? @eddyb cc @oli-obk

bors · 2020-03-16T16:31:30Z

⌛ Testing commit 6ca65bd with merge dd67187...

bors · 2020-03-16T19:50:25Z

☀️ Test successful - checks-azure
Approved by: eddyb
Pushing dd67187 to master...

rust-highfive · 2020-03-16T19:51:44Z

📣 Toolstate changed by #67133!

Tested on commit dd67187.
Direct link to PR: #67133

💔 miri on windows: test-pass → test-fail (cc @oli-obk @eddyb @RalfJung).
💔 miri on linux: test-pass → test-fail (cc @oli-obk @eddyb @RalfJung).

@oli-obk

Tested on commit rust-lang/rust@dd67187. Direct link to PR: <rust-lang/rust#67133> 💔 miri on windows: test-pass → test-fail (cc @oli-obk @eddyb @RalfJung). 💔 miri on linux: test-pass → test-fail (cc @oli-obk @eddyb @RalfJung).

rust-highfive assigned eddyb Dec 8, 2019

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Dec 8, 2019

eddyb reviewed Dec 8, 2019

View reviewed changes

src/librustc/ty/context.rs Outdated Show resolved Hide resolved

This comment has been minimized.

Sign in to view

RalfJung reviewed Dec 9, 2019

View reviewed changes

src/test/mir-opt/const_prop/discriminant.rs Outdated Show resolved Hide resolved

RalfJung reviewed Dec 9, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

RalfJung reviewed Dec 9, 2019

View reviewed changes

oli-obk commented Dec 9, 2019

View reviewed changes

src/librustc/mir/mod.rs Outdated Show resolved Hide resolved

RalfJung reviewed Dec 9, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

RalfJung reviewed Dec 9, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

RalfJung reviewed Dec 9, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

RalfJung reviewed Dec 9, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

RalfJung reviewed Dec 9, 2019

View reviewed changes

eddyb reviewed Dec 10, 2019

View reviewed changes

src/test/ui/const-generics/raw-ptr-const-param.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/test/mir-opt/const_prop/discriminant.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/librustc/mir/mod.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/librustc/ty/print/pretty.rs Show resolved Hide resolved

eddyb reviewed Dec 10, 2019

View reviewed changes

src/test/mir-opt/const_prop/read_immutable_static.rs Outdated Show resolved Hide resolved

This comment has been minimized.

Sign in to view

oli-obk force-pushed the it_must_be_a_sign branch from aeeac8d to 0e08b7f Compare December 13, 2019 14:41

This comment has been minimized.

Sign in to view

oli-obk force-pushed the it_must_be_a_sign branch from 0e08b7f to 14f4f81 Compare December 21, 2019 09:54

This comment has been minimized.

Sign in to view

oli-obk force-pushed the it_must_be_a_sign branch from 14f4f81 to 69f6301 Compare December 23, 2019 14:18

oli-obk added 5 commits March 11, 2020 09:10

Reuse type ascripted printing for type cast printing

b78dbf4

Print function pointer type for function pointer const generics

fff2e0f

Don't print literal type suffixes if print_ty is false

cc9ca64

Generalize typed value printing and use for undef printing

2e91065

Address review comments

db9ddf1

oli-obk force-pushed the it_must_be_a_sign branch from 44eec64 to db9ddf1 Compare March 11, 2020 08:40

oli-obk added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Mar 11, 2020

eddyb reviewed Mar 11, 2020

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

oli-obk added 2 commits March 12, 2020 11:06

Comment nit

d0b1211

Remove fn special casing in const printing

6e73a14

eddyb reviewed Mar 12, 2020

View reviewed changes

src/librustc/ty/print/pretty.rs Outdated Show resolved Hide resolved

oli-obk added 2 commits March 13, 2020 10:48

Print ConstKind::Bound the same as TyKind::Bound

306df94

Print ConstKind::Placeholder just like TyKind::Placeholder

6ca65bd

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 Mar 13, 2020

bors added the merged-by-bors This PR was explicitly merged by bors. label Mar 16, 2020

bors mentioned this pull request Mar 16, 2020

Enable blessing of mir opt tests #69916

Merged

bors merged commit dd67187 into rust-lang:master Mar 16, 2020

rust-highfive mentioned this pull request Mar 16, 2020

miri no longer builds after rust-lang/rust#67133 #70055

Closed

oli-obk deleted the it_must_be_a_sign branch March 16, 2021 12:14

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Deduplicate pretty printing of constants #67133

Deduplicate pretty printing of constants #67133

oli-obk commented Dec 8, 2019

This comment has been minimized.

This comment has been minimized.

RalfJung Dec 9, 2019

oli-obk Dec 9, 2019

RalfJung Dec 9, 2019

oli-obk Dec 9, 2019

eddyb Dec 10, 2019

varkor Jan 11, 2020

oli-obk Jan 24, 2020

RalfJung Dec 9, 2019

oli-obk Dec 10, 2019

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

eddyb commented Mar 11, 2020

oli-obk commented Mar 13, 2020

bors commented Mar 13, 2020

bors commented Mar 13, 2020

bors commented Mar 16, 2020

bors commented Mar 16, 2020

rust-highfive commented Mar 16, 2020

Deduplicate pretty printing of constants #67133

Deduplicate pretty printing of constants #67133

Conversation

oli-obk commented Dec 8, 2019

This comment has been minimized.

This comment has been minimized.

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

eddyb commented Mar 11, 2020

oli-obk commented Mar 13, 2020

bors commented Mar 13, 2020

bors commented Mar 13, 2020

bors commented Mar 16, 2020

bors commented Mar 16, 2020

rust-highfive commented Mar 16, 2020