incr.comp.: Switch to red/green change tracking, remove legacy system. #44901

michaelwoerister · Sep 28, 2017

This PR finally switches incremental compilation to red/green tracking and completely removes the legacy dependency graph implementation -- which includes a few quite costly passes that are simply not needed with the new system anymore.

There's still some documentation to be done and there's certainly still lots of optimizing and tuning ahead -- but the foundation for red/green is in place with this PR. This has been in the making for a long time :)

r? @nikomatsakis
cc @alexcrichton, @rust-lang/compiler

alexcrichton · Sep 28, 2017

alexcrichton reviewed Sep 28, 2017

View changes

alexcrichton · Sep 28, 2017

src/librustc_trans/back/symbol_export.rs

@@ -79,9 +79,7 @@ pub fn provide_local(providers: &mut Providers) {
    providers.is_exported_symbol = |tcx, id| {
        // FIXME(#42293) needs red/green to not break a bunch of incremental
        // tests
-        tcx.dep_graph.with_ignore(|| {


:D

(you can probably delete the FIXME above this as well)

alexcrichton · Sep 28, 2017

🎉 🎉 🎉 🎉 🎉 🎉 🌮 🎉 🎉 🎉 🎉 🎉 🎉 🎉

kennytm · Sep 28, 2017

kennytm reviewed Sep 28, 2017

View changes

kennytm · Sep 28, 2017

src/librustc/ty/maps/plumbing.rs

+
+                profq_msg!(tcx,
+                    ProfileQueriesMsg::QueryBegin(
+                        span.clone(),


Should be span.data(). Why needs the clone()?

[00:06:30] error[E0308]: mismatched types [00:06:30] --> /checkout/src/librustc/ty/maps/plumbing.rs:654:25 [00:06:30] | [00:06:30] 654 | span.clone(), [00:06:30] | ^^^^^^^^^^^^ expected struct `syntax_pos::SpanData`, found struct `syntax_pos::Span` [00:06:30] ... [00:06:30] 710 | DepKind::RegionScopeTree => { force!(region_scope_tree, def_id!()); } [00:06:30] | ------------------------------------- in this macro invocation [00:06:30] | [00:06:30] = note: expected type `syntax_pos::SpanData` [00:06:30] found type `syntax_pos::Span` [00:06:30] = help: here are some functions which might fulfill your needs: [00:06:30] - .data() ... [00:06:32] error: aborting due to 110 previous errors [00:06:32] [00:06:32] error: Could not compile `rustc`.

Looks like this needs a rebase...

sgrif · Sep 28, 2017

Speaking of red/green

that's a lot of both. ;)

Mark-Simulacrum · Sep 29, 2017

Let me know if you want to check perf (try build + ping me).

michaelwoerister · Sep 29, 2017

@Mark-Simulacrum Awesome, thank you!

michaelwoerister · Sep 29, 2017

@bors try

bors · Sep 29, 2017

⌛️ Trying commit 762b735 with merge b669ce1...

bors · Sep 29, 2017

☀️ Test successful - status-travis
State: approved= try=True

Mark-Simulacrum · Sep 29, 2017

Looks like wins across the board:

http://perf.rust-lang.org/compare.html?commit_a=0253d98382fa351d53880605a346d0ea2e941a07&commit_b=b669ce1863e5451453545f892d1014eb16b9a878&stat=instructions%3Au

michaelwoerister · Sep 29, 2017

Thanks, @Mark-Simulacrum! The link does not seem to work for me though at the moment (I see no results listed).

Mark-Simulacrum · Sep 29, 2017

Should work now. I think perf.rlo was just not updating for some reason...

michaelwoerister · Oct 2, 2017

Thanks, @Mark-Simulacrum! Looks promising.

nikomatsakis · Oct 3, 2017

nikomatsakis reviewed Oct 3, 2017

View changes

nikomatsakis · Oct 3, 2017

src/librustc/dep_graph/graph.rs

+                DepNodeColor::Red
+            };
+
+            assert!(data.colors.borrow_mut().insert(key, color).is_none());


Nit: Although I am often tempted to write an assert! like this, I tend to think that putting crucial side-effect operations inside of an assert! is bad-style. Too easy to overlook. I think I'd prefer to see something like:

let old_value = data.colors.borrow_mut().insert(key, color); assert!(old_value.is_none(), "key {:?} executed twice", key);

nikomatsakis · Oct 3, 2017

nikomatsakis approved these changes Oct 3, 2017

View changes

r=me with comments and nits addressed

nikomatsakis · Oct 3, 2017

src/librustc/ty/maps/plumbing.rs

+                                                                        f)
+                    }
+
+                    if let Some(dep_node_index) = tcx.dep_graph.try_mark_green(tcx, &dep_node) {


I'm curious, why not make try_mark_green check the cache, so we can collapse these two ifs into one? They seem sort of like fundamentally the same case (marked green already vs can mark green now)

Wouldn't that mean that we always load the result from the cache, even if it is never asked for?

nikomatsakis · Oct 3, 2017

src/librustc/ty/maps/plumbing.rs

+                debug_assert!(tcx.dep_graph.is_green(dep_node_index));
+
+                // We don't do any caching yet, so recompute
+                let (result, diagnostics) = tcx.cycle_check(span, Query::$name(key), || {


Nit: feels like this maybe wants a subroutine, kind of duplicates the code above.

this happens later, I saw, sorry :)

nikomatsakis · Oct 3, 2017

src/librustc/ty/maps/plumbing.rs

+    // We should never get into the situation of having to force this from the DepNode.
+    // Since we cannot reconstruct the query key, we would always end up having to evaluate
+    // the first caller of this query that *is* reconstructible. This might very well be
+    // compile_codegen_unit() in which case we'd lose all re-use.


I find this comment pretty unclear. What is "this", in the first sentence, referring to? What makes CodegenUnit worse or different from other kinds of depnodes?

Hm, yeah, I seem to have moved this comment out of context without updating it. CodegenUnit is different because there's no way to reconstruct the query-key from the DepNode, so the "parent query" would have to be re-done. That parent query is compile_codegen_unit(), so we would always have to recompile the codegen unit just to find out that it would not have been necessary.

I'll update the comment and try to make it clearer.

nikomatsakis · Oct 3, 2017

src/librustc/ty/maps/plumbing.rs

+    match dep_node.kind {
+        // These are inputs that are expected to be pre-allocated and that
+        // should therefore always be red or green already
+        DepKind::AllLocalTraitImpls |


This match feels mildly unfortunate -- i.e., it feels like we ought to be able to derive this information from other macros and not require it to be "re-entered" here. But oh well, I haven't thought about how to actually do that. Might be nice to file a FIXME issue about it though.

Yeah, it's definitely not pretty. But I didn't want spend too much time on making this macro generated before we have a proof of concept implementation of red/green. I'll add a FIXME.

nikomatsakis · Oct 3, 2017

src/librustc/dep_graph/graph.rs

@@ -644,15 +664,26 @@ pub(super) struct CurrentDepGraph {
    edges: IndexVec<DepNodeIndexNew, Vec<DepNodeIndexNew>>,
    node_to_node_index: FxHashMap<DepNode, DepNodeIndexNew>,

+    anon_id_seed: Fingerprint,


this feels like it merits a nice comment

Agreed :)

michaelwoerister · Oct 3, 2017

Thanks for the review, @nikomatsakis!

michaelwoerister · Oct 4, 2017

@bors r=nikomatsakis

bors · Oct 4, 2017

📌 Commit 0454a41 has been approved by nikomatsakis

bors · Oct 4, 2017

⌛️ Testing commit 0454a41 with merge d7e73e4...

bors · Oct 4, 2017

☀️ Test successful - status-appveyor, status-travis
Approved by: nikomatsakis
Pushing d7e73e4 to master...

alexcrichton · Oct 5, 2017

@michaelwoerister

I think we've recovered on all perf benchmarks except two. Everything else either improved since the original regression or was within 5% that I figured it would come out in the wash.

Interestingly I think those two benchmarks are the "from scratch build up incremental first time" benchmark, right?

michaelwoerister · Oct 6, 2017

perf.rlo is showing me only flat lines right now :/

I think I'll write up a tracking issue for possible performance improvements.

rust-highfive assigned nikomatsakis Sep 28, 2017

alexcrichton added the S-waiting-on-review label Sep 28, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from a1d7f6c to b88c694 Sep 29, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from b88c694 to 762b735 Sep 29, 2017

michaelwoerister referenced this pull request Sep 29, 2017
Closed
Tracking Issue for incr. comp. red/green testing #44716
3 of 3 tasks complete

michaelwoerister referenced this pull request Sep 29, 2017
Closed
incr.comp.: Update fingerprint-based auto tests for red/green tracking. #44924
27 of 27 tasks complete

michaelwoerister referenced this pull request Oct 2, 2017
Merged
rustc: Implement ThinLTO #44841

incr.comp.: Do some cleanup.

dbab705

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from 762b735 to 563d30e Oct 2, 2017

incr.comp.: Fix some merge fallout.

c96d0bf

michaelwoerister referenced this pull request Oct 4, 2017
Open
incr.comp.: Implement force_from_dep_node() in a macro. #45015

incr.comp.: Address review comments.

Loading status checks…

0454a41

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from 563d30e to 0454a41 Oct 4, 2017

michaelwoerister referenced this pull request Oct 4, 2017
Merged
incr.comp.: Fix infinite recursion in Debug implementation of DepNode #45018

bors merged commit 0454a41 into rust-lang:master Oct 4, 2017
2 checks passed

2 checks passed

continuous-integration/travis-ci/pr The Travis CI build passed
Details

homu Test successful
Details

michaelwoerister referenced this pull request Oct 23, 2017
Closed
incr.comp.: Move Metadata-hash loading to DepGraph. #44702

rust-lang/rust

Join GitHub today

incr.comp.: Switch to red/green change tracking, remove legacy system. #44901

Conversation

michaelwoerister commented Sep 28, 2017

rust-highfive assigned nikomatsakis Sep 28, 2017

alexcrichton reviewed Sep 28, 2017 View changes

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

alexcrichton commented Sep 28, 2017

kennytm reviewed Sep 28, 2017 View changes

This comment has been minimized.

This comment has been minimized.

alexcrichton added the S-waiting-on-review label Sep 28, 2017

This comment has been minimized.

sgrif commented Sep 28, 2017

This comment has been minimized.

Mark-Simulacrum commented Sep 29, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from a1d7f6c to b88c694 Sep 29, 2017

This comment has been minimized.

michaelwoerister commented Sep 29, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from b88c694 to 762b735 Sep 29, 2017

This comment has been minimized.

michaelwoerister commented Sep 29, 2017

This comment has been minimized.

bors commented Sep 29, 2017

bors added a commit that referenced this pull request Sep 29, 2017

michaelwoerister referenced this pull request Sep 29, 2017

This comment has been minimized.

bors commented Sep 29, 2017

michaelwoerister referenced this pull request Sep 29, 2017

This comment has been minimized.

Mark-Simulacrum commented Sep 29, 2017

This comment has been minimized.

michaelwoerister commented Sep 29, 2017

This comment has been minimized.

Mark-Simulacrum commented Sep 29, 2017

This comment has been minimized.

michaelwoerister commented Oct 2, 2017

michaelwoerister referenced this pull request Oct 2, 2017

michaelwoerister added some commits Sep 25, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from 762b735 to 563d30e Oct 2, 2017

nikomatsakis reviewed Oct 3, 2017 View changes

This comment has been minimized.

This comment has been minimized.

nikomatsakis approved these changes Oct 3, 2017 View changes

nikomatsakis left a comment

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

michaelwoerister commented Oct 3, 2017

michaelwoerister referenced this pull request Oct 4, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from 563d30e to 0454a41 Oct 4, 2017

This comment has been minimized.

michaelwoerister commented Oct 4, 2017

This comment has been minimized.

bors commented Oct 4, 2017

michaelwoerister referenced this pull request Oct 4, 2017

This comment has been minimized.

bors commented Oct 4, 2017

bors added a commit that referenced this pull request Oct 4, 2017

This comment has been minimized.

bors commented Oct 4, 2017

Hide details View details bors merged commit 0454a41 into rust-lang:master Oct 4, 2017 2 checks passed

2 checks passed

This was referenced Oct 4, 2017

This comment has been minimized.

alexcrichton commented Oct 5, 2017

This comment has been minimized.

michaelwoerister commented Oct 6, 2017

kennytm added a commit to kennytm/rust that referenced this pull request Oct 8, 2017

alexcrichton reviewed Sep 28, 2017

View changes

kennytm reviewed Sep 28, 2017

View changes

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from `a1d7f6c` to `b88c694` Sep 29, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from `b88c694` to `762b735` Sep 29, 2017

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from `762b735` to `563d30e` Oct 2, 2017

nikomatsakis reviewed Oct 3, 2017

View changes

nikomatsakis approved these changes Oct 3, 2017

View changes

michaelwoerister force-pushed the michaelwoerister:on-demand-eval branch from `563d30e` to `0454a41` Oct 4, 2017

bors merged commit `0454a41` into rust-lang:master Oct 4, 2017
2 checks passed