[Git][ghc/ghc][wip/T19627] 50 commits: rts: Drop redundant prototype

Apoorv Ingle (@ani) gitlab at gitlab.haskell.org
Wed Mar 8 21:10:32 UTC 2023



Apoorv Ingle pushed to branch wip/T19627 at Glasgow Haskell Compiler / GHC


Commits:
321d46d9 by Ben Gamari at 2023-03-08T15:02:30-05:00
rts: Drop redundant prototype

- - - - -
abb6070f by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Fix style

- - - - -
be278901 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Deduplicate assertion

- - - - -
b9034639 by Ben Gamari at 2023-03-08T15:02:30-05:00
rts: Fix type issues in Sparks.h

Adds explicit casts to satisfy a C++ compiler.

- - - - -
da7b2b94 by Ben Gamari at 2023-03-08T15:02:30-05:00
rts: Use release ordering when storing thread labels

Since this makes the ByteArray# visible from other cores.

- - - - -
5b7f6576 by Ben Gamari at 2023-03-08T15:02:30-05:00
rts/BlockAlloc: Allow disabling of internal assertions

These can be quite expensive and it is sometimes useful to compile a
DEBUG RTS without them.

- - - - -
6283144f by Ben Gamari at 2023-03-08T15:02:30-05:00
rts/Sanity: Mark pinned_object_blocks

- - - - -
9b528404 by Ben Gamari at 2023-03-08T15:02:30-05:00
rts/Sanity: Look at nonmoving saved_filled lists

- - - - -
0edc5438 by Ben Gamari at 2023-03-08T15:02:30-05:00
Evac: Squash data race in eval_selector_chain

- - - - -
7eab831a by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Clarify implementation

This makes the intent of this implementation a bit clearer.

- - - - -
532262b9 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Clarify comment

- - - - -
bd9cd84b by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Add missing no-op in busy-wait loop

- - - - -
c4e6bfc8 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Don't push empty arrays to update remembered set

Previously the write barrier of resizeSmallArray# incorrectly handled
resizing of zero-sized arrays, pushing an invalid pointer to the update
remembered set.

Fixes #22931.

- - - - -
92227b60 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Fix handling of weak pointers

This fixes an interaction between aging and weak pointer handling which
prevented the finalization of some weak pointers. In particular, weak
pointers could have their keys incorrectly marked by the preparatory
collector, preventing their finalization by the subsequent concurrent
collection.

While in the area, we also significantly improve the assertions
regarding weak pointers.

Fixes #22327.

- - - - -
ba7e7972 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Sanity check nonmoving large objects and compacts

- - - - -
71b038a1 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Sanity check mutable list

Assert that entries in the nonmoving generation's generational
remembered set (a.k.a. mutable list) live in nonmoving generation.

- - - - -
99d144d5 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Don't show occupancy if we didn't collect live words

- - - - -
81d6cc55 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Fix tracking of FILLED_SWEEPING segments

Previously we only updated the state of the segment at the head of each
allocator's filled list.

- - - - -
58e53bc4 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Assert state of swept segments

- - - - -
2db92e01 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Handle new closures in nonmovingIsNowAlive

We must conservatively assume that new closures are reachable since we
are not guaranteed to mark such blocks.

- - - - -
e4c3249f by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Don't clobber update rem sets of old capabilities

Previously `storageAddCapabilities` (called by `setNumCapabilities`) would
clobber the update remembered sets of existing capabilities when
increasing the capability count. Fix this by only initializing the
update remembered sets of the newly-created capabilities.

Fixes #22927.

- - - - -
1b069671 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Add missing write barriers in selector optimisation

This fixes the selector optimisation, adding a few write barriers which
are necessary for soundness. See the inline comments for details.

Fixes #22930.

- - - - -
d4032690 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Post-sweep sanity checking

- - - - -
0baa8752 by Ben Gamari at 2023-03-08T15:02:30-05:00
nonmoving: Avoid n_caps race

- - - - -
5d3232ba by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Don't push if nonmoving collector isn't enabled

- - - - -
0a7eb0aa by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Be more paranoid in segment tracking

Previously we left various segment link pointers dangling. None of this
wrong per se, but it did make it harder than necessary to debug.

- - - - -
7c817c0a by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Sync-phase mark budgeting

Here we significantly improve the bound on sync phase pause times by
imposing a limit on the amount of work that we can perform during the
sync. If we find that we have exceeded our marking budget then we allow
the mutators to resume, return to concurrent marking, and try
synchronizing again later.

Fixes #22929.

- - - - -
ce22a3e2 by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Allow pinned gen0 objects to be WEAK keys

- - - - -
78746906 by Ben Gamari at 2023-03-08T15:02:31-05:00
rts: Reenable assertion

- - - - -
b500867a by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Move current segment array into Capability

The current segments are conceptually owned by the mutator, not the
collector. Consequently, it was quite tricky to prove that the mutator
would not race with the collect due to this shared state. It turns out
that such races are possible: when resizing the current segment array
we may concurrently try to take a heap census. This will attempt to walk
the current segment array, causing a data race.

Fix this by moving the current segment array into `Capability`, where it
belongs.

Fixes #22926.

- - - - -
56e669c1 by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Fix Note references

Some references to Note [Deadlock detection under the non-moving
collector] were missing an article.

- - - - -
4a7650d7 by Ben Gamari at 2023-03-08T15:02:31-05:00
rts/Sanity: Fix block count assertion with non-moving collector

The nonmoving collector does not use `oldest_gen->blocks` to track its
block list. However, it nevertheless updates `oldest_gen->n_blocks` to
ensure that its size is accounted for by the storage manager.
Consequently, we must not attempt to assert consistency between the two.

- - - - -
96a5aaed by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Don't call prepareUnloadCheck

When the nonmoving GC is in use we do not call `checkUnload` (since we
don't unload code) and therefore should not call `prepareUnloadCheck`,
lest we run into assertions.

- - - - -
6c6674ca by Ben Gamari at 2023-03-08T15:02:31-05:00
rts: Encapsulate block allocator spinlock

This makes it a bit easier to add instrumentation on this spinlock
while debugging.

- - - - -
e84f7167 by Ben Gamari at 2023-03-08T15:02:31-05:00
testsuite: Skip some tests when sanity checking is enabled

- - - - -
3ae0f368 by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Fix unregisterised build

- - - - -
4eb9d06b by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Ensure that sanity checker accounts for saved_filled segments

- - - - -
f0cf384d by Ben Gamari at 2023-03-08T15:02:31-05:00
hadrian: Add +boot_nonmoving_gc flavour transformer

For using GHC bootstrapping to validate the non-moving GC.

- - - - -
581e58ac by Ben Gamari at 2023-03-08T15:02:31-05:00
gitlab-ci: Add job bootstrapping with nonmoving GC

- - - - -
487a8b58 by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Move allocator into new source file

- - - - -
8f374139 by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Split out nonmovingAllocateGC

- - - - -
662b6166 by Ben Gamari at 2023-03-08T15:02:31-05:00
testsuite: Only run T22795* in the normal way

It doesn't make sense to run these in multiple ways as they merely test
whether `-threaded`/`-single-threaded` flags.

- - - - -
0af21dfa by Ben Gamari at 2023-03-08T15:02:31-05:00
rts: Rename clear_segment(_free_blocks)?

To reflect the fact that these are to do with the nonmoving collector,
now since they are exposed no longer static.

- - - - -
7bcb192b by Ben Gamari at 2023-03-08T15:02:31-05:00
rts: Fix incorrect STATIC_INLINE

This should be INLINE_HEADER lest we get unused declaration warnings.

- - - - -
f1fd3ffb by Ben Gamari at 2023-03-08T15:02:31-05:00
testsuite: Mark ffi023 as broken due to #23089

- - - - -
a57f12b3 by Ben Gamari at 2023-03-08T15:02:31-05:00
testsuite: Skip T7160 in the nonmoving way

Finalization order is different under the nonmoving collector.

- - - - -
f6f12a36 by Ben Gamari at 2023-03-08T15:02:31-05:00
rts: Capture GC configuration in a struct

The number of distinct arguments passed to GarbageCollect was getting a
bit out of hand.

- - - - -
ba73a807 by Ben Gamari at 2023-03-08T15:02:31-05:00
nonmoving: Non-concurrent collection

- - - - -
7c813d06 by Alexis King at 2023-03-08T15:03:10-05:00
hadrian: Fix flavour compiler stage options off-by-one error

!9193 pointed out that ghcDebugAssertions was supposed to be a predicate
on the stage of the built compiler, but in practice it was a predicate
on the stage of the compiler used to build. Unfortunately, while it
fixed that issue for ghcDebugAssertions, it documented every other
similar option as behaving the same way when in fact they all used the
old behavior.

The new behavior of ghcDebugAssertions seems more intuitive, so this
commit changes the interpretation of every other option to match. It
also improves the enableProfiledGhc and debugGhc flavour transformers by
making them more selective about which stages in which they build
additional library/RTS ways.

- - - - -
61e58ea3 by Apoorv Ingle at 2023-03-08T21:10:28+00:00
Fixes #19627.

Previously the solver failed with an unhelpful "solver reached too may iterations" error.
With the fix for #21909 in place we no longer have the possibility of generating such an error if we have `-fconstraint-solver-iteration` > `-fgivens-fuel > `-fwanteds-fuel`. This is true by default, and the said fix also gives programmers a knob to control how hard the solver should try before giving up.

This commit adds:
* Reference to ticket #19627 in the Note [Expanding Recursive Superclasses and ExpansionFuel]
* Test `typecheck/should_fail/T19627.hs` for regression purposes

- - - - -


30 changed files:

- .gitlab/gen_ci.hs
- .gitlab/jobs.yaml
- compiler/GHC/Tc/Solver.hs
- hadrian/doc/user-settings.md
- hadrian/src/Expression.hs
- hadrian/src/Flavour.hs
- hadrian/src/Oracles/Flavour.hs
- hadrian/src/Settings/Builders/Ghc.hs
- hadrian/src/Settings/Builders/RunTest.hs
- hadrian/src/Settings/Flavours/Development.hs
- hadrian/src/Settings/Flavours/Quick.hs
- hadrian/src/Settings/Packages.hs
- libraries/ghc-heap/tests/all.T
- rts/Capability.c
- rts/Capability.h
- rts/PrimOps.cmm
- rts/RtsStartup.c
- rts/RtsUtils.h
- rts/Schedule.c
- rts/Sparks.h
- rts/ThreadLabels.c
- rts/include/rts/storage/GC.h
- rts/include/rts/storage/MBlock.h
- rts/rts.cabal.in
- rts/sm/BlockAlloc.c
- rts/sm/Evac.c
- rts/sm/GC.c
- rts/sm/GC.h
- rts/sm/GCUtils.c
- rts/sm/GCUtils.h


The diff was not included because it is too large.


View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a470a0e41ac9498f08261e9979888e3026313f25...61e58ea3bdceff3c7aa13eecb23f25b619e2ad02

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/a470a0e41ac9498f08261e9979888e3026313f25...61e58ea3bdceff3c7aa13eecb23f25b619e2ad02
You're receiving this email because of your account on gitlab.haskell.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20230308/ff4910cc/attachment-0001.html>


More information about the ghc-commits mailing list