Remove NULL ObjectReference #1064

wks · 2024-01-08T12:08:41Z

We changed ObjectReference so that it is backed by NonZeroUsize, and can no longer represent a NULL reference. For cases where an ObjectReferences may or may not exist, Option<ObjectReference> should be used instead.

This is an API-breaking change.

ObjectReference::NULL and ObjectReference::is_null() are removed.
ObjectReference::from_raw_address() now returns Option<ObjectReference>. The unsafe method ObjectReference::from_raw_address_unchecked() assumes the address is not zero, and returns ObjectReference.
API functions that may or may not return a valid ObjectReference now return Option<ObjectReference>. Examples include Edge::load() and ReferenceGlue::get_referent().

If a VM binding needs to expose Option<ObjectReference> to native (C/C++) programs, a convenient type crate::util::api_util::NullableObjectReference is provided.

Fixes: #1043

wks · 2024-01-19T17:06:19Z

Here are some preliminary benchmark results. Two builds:

build1: master
build2: this PR + the mmtk-openjdk counterpart (Remove NULL ObjectReference mmtk-openjdk#265)

lusearch from DaCapo Chopin, 40 invocations, 5 iterations, 2.5x and 3x min heap size, ran on shrew.moma (Coffee Lake) and vole.moma (Ryzen 4).

I ran it with Immix and GenCopy. However GenCopy overflowed the phase counter. As a result, the stats of subsequent GCs are discarded from the result. (See #1074) So I am only showing the results for Immix.

https://squirrel.anu.edu.au/plotty/wks/noproject/#0|shrew-2024-01-19-Fri-080239^shrew-2024-01-19-Fri-113438^vole-2024-01-19-Fri-080218^vole-2024-01-19-Fri-113411&benchmark^build^hfac^invocation^iteration^logfile^mmtk_gc&GC^time^time.other^time.stw&|10&iteration^1^4&mmtk_gc^1^Immix|20&1^invocation|30&1&benchmark^hfac^logfile&build;build1|60&hfac&logfile|70&hfac-logfile^shrew-vole-253x|40&Histogram%20(with%20CI)^build^hfac-logfile&

From the data, the impact on total time is between -0.5% to +0.5%, and the error bar is large. Interestingly, it always shows slowdown on shrew, and always speed-up on vole. For STW time, "shrew 2.5x" has a slow-down of 1%, and "vole 3x" has a speed-up of 1.5%, but their error bars are too large.

I think the result of forcing ObjectReference to be non-zero is insignificant for lusearch with Immix.

wks · 2024-01-22T03:57:49Z

More benchmark results. I ran the same binary on shrew.moma and mole.moma (identical to vole.moma), 20 invocations, 3x min heap size, using Immix and StickyImmix GC.

Plotty link: https://squirrel.anu.edu.au/plotty/wks/noproject/#0|mole-2024-01-20-Sat-113636^shrew-2024-01-20-Sat-113610&benchmark^build^invocation^iteration^logfile^mmtk_gc&GC^time^time.other^time.stw&|10&iteration^1^4|20&1^invocation|30&1&benchmark^logfile^mmtk_gc&build;build1|60&build&logfile&mmtk_gc|70&build-logfile-mmtk_gc^blg-2024-01-22|41&Histogram%20(with%20CI)^build-logfile-mmtk_gc^benchmark&

Total time: normalized to build1 of the corresponding machine-plan pair:

For each machine-plan pair, the geomean of total time of build2 is between 99.6% to 100.3% of that of build1.

Other time:

Geomeans are between 99.8% to 100.3% of build1.

STW time:

STW times are between 97.9% to 101.0% of build1. But given the noise level, it is hard to conclude there is any improvement. We can see significant speed-up for build2-shrew-StickyImmix for biojava, jython and pmd, but build2-mole-StickyImmix does not have the same speed-up for those benchmarks. This is a bit hard to explain.

Parent PR: mmtk/mmtk-core#1064

It's used by multiple bindings.

Parent PR: mmtk/mmtk-core#1064

wks · 2024-04-24T14:31:50Z

binding-refs
OPENJDK_BINDING_REPO=wks/mmtk-openjdk
OPENJDK_BINDING_REF=fix/no-null-objref
JIKESRVM_BINDING_REPO=wks/mmtk-jikesrvm
JIKESRVM_BINDING_REF=fix/no-null-objref
V8_BINDING_REPO=wks/mmtk-v8
V8_BINDING_REF=fix/no-null-objref
RUBY_BINDING_REPO=wks/mmtk-ruby
RUBY_BINDING_REF=fix/no-null-objref
JULIA_BINDING_REPO=wks/mmtk-julia
JULIA_BINDING_REF=fix/no-null-objref

wks · 2024-04-24T16:52:26Z

This PR is now ready for review. Associated PRs for the bindings:

qinsoon

Looks good to me. There are some minor points.

Also a quick check: is there any change that may affect performance for OpenJDK and MMTk since the evaluation in #1064 (comment)?

qinsoon · 2024-04-26T03:21:21Z

src/util/address.rs

@@ -479,28 +480,40 @@ use crate::vm::VMBinding;
 /// the opaque `ObjectReference` type, and we haven't seen a use case for now.
 #[repr(transparent)]
 #[derive(Copy, Clone, Eq, Hash, PartialOrd, Ord, PartialEq, NoUninit)]
-pub struct ObjectReference(usize);
+pub struct ObjectReference(NonZeroUsize);


The comments for the type should note that ObjectReference cannot be null, and for places where we may have an invalid object reference, we use Option<ObjectReference>. NonZeroUsize makes sure that Option<ObjectReference> has the same length as ObjectReference.

qinsoon · 2024-04-26T03:31:29Z

src/util/mod.rs

@@ -11,6 +11,8 @@ pub mod address;
 /// Allocators
 // This module is made public so the binding could implement allocator slowpaths if they would like to.
 pub mod alloc;
+/// Helpers for making native APIs.
+pub mod apiutils;


Could use the name api_util to be consistent with other existing modules such as rust_util/test_util.

wks · 2024-04-26T06:03:21Z

Also a quick check: is there any change that may affect performance for OpenJDK and MMTk since the evaluation in #1064 (comment)?

I don't think so, but I'll re-run the benchmarks just to be safe.

wks · 2024-04-27T11:48:48Z

I re-ran the benchmarks with the same setting as in #1064 (comment)

Plotty: https://squirrel.anu.edu.au/plotty/wks/noproject/#0|shrew-2024-04-26-Fri-091850^mole-2024-04-26-Fri-091914&benchmark^build^invocation^iteration^logfile^mmtk_gc&GC^time^time.other^time.stw&|10&iteration^1^4|20&1^invocation|30&1&benchmark^logfile^mmtk_gc&build;build1|60&build&mmtk_gc&logfile|70&build-mmtk_gc-logfile^mole-shrew-20240426|40&Histogram%20(with%20CI)^build-mmtk_gc-logfile^benchmark&

Total time: (geomens of build2 are between 99.5% to 100.2% of build1, varying among plan * machine pairs)

Other time: (99.6% to 100.1%)

STW time: (99.2% to 100.8%)

The differences are within 1%. This result is consistent with the last evaluation.

There were a few comments that still mentioned null ObjectReference. They have been removed.

Parent PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

Upstream PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

Parent PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

This was referenced Jan 8, 2024

Remove NULL ObjectReference mmtk/mmtk-openjdk#265

Merged

Remove NULL ObjectReference #1043

Closed

wks force-pushed the fix/no-null-objref branch from 0d1c7d3 to a5035a2 Compare January 19, 2024 07:23

wks added 2 commits April 15, 2024 17:49

Make ObjectReferenct non-nullable

2da71a0

Fix clippy warnings

03fb6d4

wks force-pushed the fix/no-null-objref branch from a5035a2 to 03fb6d4 Compare April 15, 2024 10:05

wks added 6 commits April 17, 2024 10:04

Merge branch 'master' into fix/no-null-objref

2908eb1

Fix sanity

738354f

Make malloc_native_mimalloc compile

adbd6ae

Merge branch 'master' into fix/no-null-objref

bac848d

Fix mock tests

3075710

Fix doc

19b8641

wks added a commit to wks/mmtk-jikesrvm that referenced this pull request Apr 19, 2024

Remove NULL ObjectReference

59cd0aa

Parent PR: mmtk/mmtk-core#1064

wks mentioned this pull request Apr 19, 2024

Remove NULL ObjectReference mmtk/mmtk-jikesrvm#169

Merged

wks added 4 commits April 24, 2024 10:32

Merge branch 'master' into fix/no-null-objref

b7a8b22

Updated style and comment

7eed7db

Move NullableObjectReference to mmtk-core

6901642

It's used by multiple bindings.

Merge branch 'master' into fix/no-null-objref

6aa75cd

wks added a commit to wks/mmtk-ruby that referenced this pull request Apr 24, 2024

Remove NULL ObjectReference

1bf6794

Parent PR: mmtk/mmtk-core#1064

wks mentioned this pull request Apr 24, 2024

Remove NULL ObjectReference mmtk/mmtk-ruby#62

Merged

wks added a commit to wks/mmtk-ruby that referenced this pull request Apr 24, 2024

Remove NULL ObjectReference

2b9873e

Parent PR: mmtk/mmtk-core#1064

Fix link

9909808

wks added a commit to wks/mmtk-v8 that referenced this pull request Apr 24, 2024

Remove NULL ObjectReference

b74e44f

Parent PR: mmtk/mmtk-core#1064

wks mentioned this pull request Apr 24, 2024

Remove NULL ObjectReference mmtk/mmtk-v8#84

Merged

wks added a commit to wks/mmtk-julia that referenced this pull request Apr 24, 2024

Remove NULL ObjectReference

f89caea

Parent PR: mmtk/mmtk-core#1064

wks mentioned this pull request Apr 24, 2024

Remove NULL ObjectReference mmtk/mmtk-julia#146

Merged

wks added the PR-extended-testing Run extended tests for the pull request label Apr 24, 2024

wks marked this pull request as ready for review April 24, 2024 16:48

wks requested a review from qinsoon April 24, 2024 16:49

qinsoon approved these changes Apr 26, 2024

View reviewed changes

wks added 2 commits April 26, 2024 14:43

Update the doc of ObjectReference

fe80f3e

Rename apiutils to api_util

edd0d0a

Remove comments about ObjectReference::NULL.

3474c98

There were a few comments that still mentioned null ObjectReference. They have been removed.

wks enabled auto-merge April 27, 2024 13:09

wks added this pull request to the merge queue Apr 27, 2024

Merged via the queue into mmtk:master with commit a02803b Apr 27, 2024
22 of 24 checks passed

wks deleted the fix/no-null-objref branch April 27, 2024 14:30

mmtkgc-bot added a commit to mmtk/mmtk-v8 that referenced this pull request Apr 27, 2024

Remove NULL ObjectReference (#84)

6a90d22

Parent PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

mmtkgc-bot added a commit to mmtk/mmtk-ruby that referenced this pull request Apr 27, 2024

Remove NULL ObjectReference (#62)

eacde09

Parent PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

mmtkgc-bot added a commit to mmtk/mmtk-openjdk that referenced this pull request Apr 27, 2024

Remove NULL ObjectReference (#265)

3d98c8b

Upstream PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

mmtkgc-bot added a commit to mmtk/mmtk-jikesrvm that referenced this pull request Apr 27, 2024

Remove NULL ObjectReference (#169)

1b0959f

Upstream PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

mmtkgc-bot added a commit to mmtk/mmtk-julia that referenced this pull request Apr 27, 2024

Remove NULL ObjectReference (#146)

309a6ae

Parent PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

qinsoon pushed a commit to qinsoon/mmtk-julia that referenced this pull request May 22, 2024

Remove NULL ObjectReference (mmtk#146)

54b8e44

Parent PR: mmtk/mmtk-core#1064 --------- Co-authored-by: mmtkgc-bot <mmtkgc.bot@gmail.com>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Remove NULL ObjectReference #1064

Remove NULL ObjectReference #1064

wks commented Jan 8, 2024 •

edited

Loading

wks commented Jan 19, 2024

wks commented Jan 22, 2024

wks commented Apr 24, 2024

wks commented Apr 24, 2024

qinsoon left a comment

qinsoon Apr 26, 2024

qinsoon Apr 26, 2024

wks commented Apr 26, 2024

wks commented Apr 27, 2024

Remove NULL ObjectReference #1064

Remove NULL ObjectReference #1064

Conversation

wks commented Jan 8, 2024 • edited Loading

wks commented Jan 19, 2024

wks commented Jan 22, 2024

wks commented Apr 24, 2024

wks commented Apr 24, 2024

qinsoon left a comment

Choose a reason for hiding this comment

qinsoon Apr 26, 2024

Choose a reason for hiding this comment

qinsoon Apr 26, 2024

Choose a reason for hiding this comment

wks commented Apr 26, 2024

wks commented Apr 27, 2024

wks commented Jan 8, 2024 •

edited

Loading