Refactor rustc_codegen_ssa #56198

bjorn3 · 2018-11-24T15:06:08Z

cc #56108 (not all things are done yet)

This removes an unsafe method from cg_ssa.

src/librustc_codegen_ssa/traits/statics.rs

eddyb · 2018-11-24T15:08:54Z

src/librustc_codegen_llvm/common.rs

@@ -98,7 +98,6 @@ impl BackendTypes for CodegenCx<'ll, 'tcx> {
 type Value = &'ll Value;
 type BasicBlock = &'ll BasicBlock;
 type Type = &'ll Type;
- type Context = &'ll llvm::Context;


Whoa, this is nice!

eddyb · 2018-11-24T15:10:36Z

src/librustc_codegen_llvm/intrinsic.rs

@@ -763,6 +763,95 @@ impl IntrinsicCallMethods<'tcx> for Builder<'a, 'll, 'tcx> {
 }
 }
 }
+
+ fn abort(&mut self) {
+ let fnname = self.cx().get_intrinsic(&("llvm.trap"));


@sunfishcode I think this builder method should be called trap, wdyt?

In cranelift the Interrupt trap can be resumed.

Yeah, Cranelift should add an explicit way to request an "abort" trap.

So what should we call this?

I like how bx.abort() lines up with mir::TerminatorKind::Abort. But either way works for me.

eddyb · 2018-11-24T15:12:17Z

src/librustc_codegen_ssa/traits/intrinsic.rs

+#[derive(Copy, Clone)]
+pub enum OverflowOp {
+ Add, Sub, Mul
+}


@denismerigoux put some enums in common.rs but I'm not sure, this might be better.

src/librustc_codegen_ssa/traits/intrinsic.rs

eddyb · 2018-11-25T10:23:08Z

@bors r+

bors · 2018-11-25T10:23:09Z

📌 Commit 36baa97 has been approved by eddyb

eddyb · 2018-11-27T18:17:06Z

src/librustc_mir/interpret/traits.rs

+ Usize(u64),
+ Nullptr,
+ FnPtr(Instance<'tcx>),
+}


Note, in case @oli-obk sees this in the meanwhile: I've already told @bjorn3 they can instead use miri to create the whole vtable allocation and just convert that in the backend uniformly.

Note that this should be done without an EvalContext. With my recent changes to Allocations this should be mostly possible. I was hoping on getting https://github.com/rust-lang/rust/pull/55392/files#diff-2eb1005fe45cbe73c0ddfcde017b5f73R83 into master before having to address vtable generation, because that function would make such an operation trivial. You can still look at that function as a reference for creating the function that generates a vtable Allocation

The code should probably move to rustc::mir::interpret::traits, too

I am backing out the vtable generation changes for now, because they are a bit too much work for me right now.

…ackend

…::checked_binop

rust-highfive · 2018-11-29T17:53:25Z

The job x86_64-gnu-llvm-5.0 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.

travis_time:end:0099f33c:start=1543513236546488962,finish=1543513238839969904,duration=2293480942
$ git checkout -qf FETCH_HEAD
travis_fold:end:git.checkout

Encrypted environment variables have been removed for security reasons.
See https://docs.travis-ci.com/user/pull-requests/#Pull-Requests-and-Security-Restrictions
$ export SCCACHE_BUCKET=rust-lang-ci-sccache2
$ export SCCACHE_REGION=us-west-1
Setting environment variables from .travis.yml
$ export IMAGE=x86_64-gnu-llvm-5.0
---

[00:03:00] travis_fold:start:tidy
travis_time:start:tidy
tidy check
[00:03:00] tidy error: /checkout/src/librustc_codegen_llvm/consts.rs:344: TODO is deprecated; use FIXME
[00:03:01] some tidy checks failed
[00:03:01] 
[00:03:01] 
[00:03:01] command did not execute successfully: "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0-tools-bin/tidy" "/checkout/src" "/checkout/obj/build/x86_64-unknown-linux-gnu/stage0/bin/cargo" "--no-vendor" "--quiet"
[00:03:01] 
[00:03:01] 
[00:03:01] failed to run: /checkout/obj/build/bootstrap/debug/bootstrap test src/tools/tidy
[00:03:01] Build completed unsuccessfully in 0:00:54
[00:03:01] Build completed unsuccessfully in 0:00:54
[00:03:01] make: *** [tidy] Error 1
[00:03:01] Makefile:79: recipe for target 'tidy' failed
The command "stamp sh -x -c "$RUN_SCRIPT"" exited with 2.
travis_time:start:06e8c59c
$ date && (curl -fs --head https://google.com | grep ^Date: | sed 's/Date: //g' || true)
Thu Nov 29 17:43:49 UTC 2018
---
travis_time:end:18261aaa:start=1543513429885035375,finish=1543513429889595057,duration=4559682
travis_fold:end:after_failure.3
travis_fold:start:after_failure.4
travis_time:start:064d16f0
$ ln -s . checkout && for CORE in obj/cores/core.*; do EXE=$(echo $CORE | sed 's|obj/cores/core\.[0-9]*\.!checkout!\(.*\)|\1|;y|!|/|'); if [ -f "$EXE" ]; then printf travis_fold":start:crashlog\n\033[31;1m%s\033[0m\n" "$CORE"; gdb --batch -q -c "$CORE" "$EXE" -iex 'set auto-load off' -iex 'dir src/' -iex 'set sysroot .' -ex bt -ex q; echo travis_fold":"end:crashlog; fi; done || true
travis_fold:end:after_failure.4
travis_fold:start:after_failure.5
travis_time:start:03cf10ef
travis_time:start:03cf10ef
$ cat ./obj/build/x86_64-unknown-linux-gnu/native/asan/build/lib/asan/clang_rt.asan-dynamic-i386.vers || true
cat: ./obj/build/x86_64-unknown-linux-gnu/native/asan/build/lib/asan/clang_rt.asan-dynamic-i386.vers: No such file or directory
travis_fold:end:after_failure.5
travis_fold:start:after_failure.6
travis_time:start:22737450
$ dmesg | grep -i kill

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)

eddyb · 2018-12-02T14:46:41Z

@bors r+

bors · 2018-12-02T14:46:42Z

📌 Commit d108a91 has been approved by eddyb

bors · 2018-12-02T18:02:27Z

⌛ Testing commit d108a91 with merge 21f2684...

@eddyb

Refactor rustc_codegen_ssa cc #56108 (not all things are done yet) This removes an unsafe method from cg_ssa. r? @eddyb cc @sunfishcode

bors · 2018-12-02T20:39:25Z

☀️ Test successful - status-appveyor, status-travis
Approved by: eddyb
Pushing 21f2684 to master...

rust-highfive assigned eddyb Nov 24, 2018

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Nov 24, 2018

eddyb requested changes Nov 24, 2018

View reviewed changes

eddyb approved these changes Nov 25, 2018

View reviewed changes

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 Nov 25, 2018

eddyb reviewed Nov 27, 2018

View reviewed changes

bjorn3 force-pushed the cg_ssa_refactor branch 2 times, most recently from 8a67d20 to cd66200 Compare November 28, 2018 14:41

memoryruins mentioned this pull request Nov 28, 2018

1.30.0 fails to build for target powerpc-unknown-netbsd #55465

Closed

kennytm 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-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Nov 29, 2018

bjorn3 force-pushed the cg_ssa_refactor branch from cd66200 to 1626e59 Compare November 29, 2018 17:19

bjorn3 added 15 commits November 29, 2018 18:19

Some refactorings

b3b6e4d

Make ConstMethods and StaticMethods require BackendTypes instead of B…

436eff5

…ackend

Remove static_replace_all_uses and statics_to_rauw from cg_ssa

2d46ee2

Remove __build_diagnostic_array! from cg_utils

8698f5c

Don't use llvm intrinsic names in cg_ssa

15a5009

Rustfmt on cg_ssa/traits

66c3195

Use BackendTypes instead of Backend or HasCodegen in a few places

187c4cf

Remove call_lifetime_intrinsic from cg_ssa

9a90455

Move IntrinsicCallMethods::call_overflow_intrinsics to BuilderMethods…

2d4c96d

…::checked_binop

Remove an unnecessary reference

b8d55d4

Remove static_bitcast from cg_ssa

59682d3

Rename StaticMethods::static_ptrcast to ConstMethods::const_ptrcast

aaca5a3

Remove static_addr_of_mut from cg_ssa

e8da3c6

Rename static_bitcast to const_bitcast

f47505e

Require Deref to CodegenCx for HasCodegen

e457330

bjorn3 added 2 commits November 29, 2018 18:19

Use implicit deref instead of BuilderMethods::cx()

ceb29e2

Move get_static from CodegenCx to Builder

d108a91

bjorn3 force-pushed the cg_ssa_refactor branch from 1626e59 to 0fc95f9 Compare November 29, 2018 17:39

bjorn3 force-pushed the cg_ssa_refactor branch from 0fc95f9 to d108a91 Compare November 29, 2018 18:01

eddyb approved these changes Dec 2, 2018

View reviewed changes

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 Dec 2, 2018

bors added a commit that referenced this pull request Dec 2, 2018

Auto merge of #56198 - bjorn3:cg_ssa_refactor, r=eddyb

21f2684

Refactor rustc_codegen_ssa cc #56108 (not all things are done yet) This removes an unsafe method from cg_ssa. r? @eddyb cc @sunfishcode

bors merged commit d108a91 into rust-lang:master Dec 2, 2018

This was referenced Dec 2, 2018

Use pinning for generators to make trait safe #55704

Merged

rustc: move debug info from LocalDecl and UpvarDecl into a dedicated VarDebugInfo. #56231

Merged

Future-proof MIR for dedicated debuginfo. #56278

Merged

bjorn3 deleted the cg_ssa_refactor branch December 3, 2018 07:25

bjorn3 mentioned this pull request Jul 27, 2019

rustc_codegen_ssa: streamline interactions between Builder's and CodegenCx, layout, etc. #56122

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Refactor rustc_codegen_ssa #56198

Refactor rustc_codegen_ssa #56198

bjorn3 commented Nov 24, 2018

eddyb Nov 24, 2018

eddyb Nov 24, 2018

bjorn3 Nov 24, 2018

sunfishcode Nov 24, 2018

eddyb Nov 24, 2018

sunfishcode Nov 25, 2018

eddyb Nov 24, 2018

eddyb commented Nov 25, 2018

bors commented Nov 25, 2018

eddyb Nov 27, 2018 •

edited

Loading

oli-obk Nov 27, 2018

oli-obk Nov 27, 2018

bjorn3 Nov 29, 2018

rust-highfive commented Nov 29, 2018

eddyb commented Dec 2, 2018

bors commented Dec 2, 2018

bors commented Dec 2, 2018

bors commented Dec 2, 2018

Refactor rustc_codegen_ssa #56198

Refactor rustc_codegen_ssa #56198

Conversation

bjorn3 commented Nov 24, 2018

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

eddyb commented Nov 25, 2018

bors commented Nov 25, 2018

eddyb Nov 27, 2018 • edited Loading

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

rust-highfive commented Nov 29, 2018

eddyb commented Dec 2, 2018

bors commented Dec 2, 2018

bors commented Dec 2, 2018

bors commented Dec 2, 2018

eddyb Nov 27, 2018 •

edited

Loading