2022-04-05 20:57:21 +03:00
|
|
|
module Base
|
|
|
|
( module Test.Tasty,
|
|
|
|
module Test.Tasty.HUnit,
|
2022-07-08 14:59:45 +03:00
|
|
|
module Juvix.Prelude,
|
2022-04-05 20:57:21 +03:00
|
|
|
module Base,
|
2022-12-20 15:05:40 +03:00
|
|
|
module Juvix.Extra.Paths,
|
2023-01-05 19:48:26 +03:00
|
|
|
module Juvix.Prelude.Env,
|
2023-10-23 14:38:52 +03:00
|
|
|
module Juvix.Compiler.Pipeline.Run,
|
|
|
|
module Juvix.Compiler.Pipeline.EntryPoint.IO,
|
2022-04-05 20:57:21 +03:00
|
|
|
)
|
|
|
|
where
|
2022-02-15 16:12:53 +03:00
|
|
|
|
2024-07-15 11:02:48 +03:00
|
|
|
import Control.Exception qualified as E
|
2022-07-25 18:51:42 +03:00
|
|
|
import Control.Monad.Extra as Monad
|
2022-05-05 16:12:17 +03:00
|
|
|
import Data.Algorithm.Diff
|
|
|
|
import Data.Algorithm.DiffOutput
|
2024-07-12 19:31:09 +03:00
|
|
|
import GHC.Generics qualified as GHC
|
2023-12-06 20:24:59 +03:00
|
|
|
import Juvix.Compiler.Internal.Translation.FromInternal.Analysis.Termination
|
2023-10-23 14:38:52 +03:00
|
|
|
import Juvix.Compiler.Pipeline.EntryPoint.IO
|
2023-12-30 22:15:35 +03:00
|
|
|
import Juvix.Compiler.Pipeline.Loader.PathResolver
|
2023-10-23 14:38:52 +03:00
|
|
|
import Juvix.Compiler.Pipeline.Run
|
2023-12-06 20:24:59 +03:00
|
|
|
import Juvix.Data.Effect.TaggedLock
|
2023-10-30 16:05:52 +03:00
|
|
|
import Juvix.Extra.Paths hiding (rootBuildDir)
|
2024-10-29 19:32:59 +03:00
|
|
|
import Juvix.Prelude hiding (assert, readProcess)
|
2023-01-05 19:48:26 +03:00
|
|
|
import Juvix.Prelude.Env
|
2024-11-05 16:28:28 +03:00
|
|
|
import Juvix.Prelude.Pretty
|
2024-05-29 14:34:04 +03:00
|
|
|
import System.Process qualified as P
|
2022-02-15 16:12:53 +03:00
|
|
|
import Test.Tasty
|
2024-11-05 16:28:28 +03:00
|
|
|
import Test.Tasty.HUnit hiding (assertFailure, testCase)
|
2024-01-11 15:04:38 +03:00
|
|
|
import Test.Tasty.HUnit qualified as HUnit
|
2022-02-15 16:12:53 +03:00
|
|
|
|
2022-04-05 20:57:21 +03:00
|
|
|
data AssertionDescr
|
|
|
|
= Single Assertion
|
2022-02-18 15:01:42 +03:00
|
|
|
| Steps ((String -> IO ()) -> Assertion)
|
|
|
|
|
2022-04-05 20:57:21 +03:00
|
|
|
data TestDescr = TestDescr
|
2022-04-07 19:10:53 +03:00
|
|
|
{ _testName :: String,
|
2022-12-20 15:05:40 +03:00
|
|
|
_testRoot :: Path Abs Dir,
|
2022-04-05 20:57:21 +03:00
|
|
|
-- | relative to root
|
2022-04-07 19:10:53 +03:00
|
|
|
_testAssertion :: AssertionDescr
|
2022-02-15 16:12:53 +03:00
|
|
|
}
|
|
|
|
|
2022-08-02 19:58:45 +03:00
|
|
|
newtype WASMInfo = WASMInfo
|
2022-12-20 15:05:40 +03:00
|
|
|
{ _wasmInfoActual :: Path Abs File -> IO Text
|
2022-08-01 13:53:19 +03:00
|
|
|
}
|
|
|
|
|
2022-04-07 19:10:53 +03:00
|
|
|
makeLenses ''TestDescr
|
|
|
|
|
2023-01-05 19:48:26 +03:00
|
|
|
data StdlibMode
|
|
|
|
= StdlibInclude
|
|
|
|
| StdlibExclude
|
2022-06-30 12:31:08 +03:00
|
|
|
deriving stock (Show, Eq)
|
|
|
|
|
2023-01-05 19:48:26 +03:00
|
|
|
data CompileMode
|
|
|
|
= WASI StdlibMode
|
|
|
|
| WASM WASMInfo
|
2022-08-01 13:53:19 +03:00
|
|
|
|
2022-02-15 16:12:53 +03:00
|
|
|
mkTest :: TestDescr -> TestTree
|
2022-04-07 19:10:53 +03:00
|
|
|
mkTest TestDescr {..} = case _testAssertion of
|
2023-04-13 12:27:39 +03:00
|
|
|
Single assertion -> testCase _testName (withCurrentDir _testRoot assertion)
|
2022-12-20 15:05:40 +03:00
|
|
|
Steps steps -> testCaseSteps _testName (withCurrentDir _testRoot . steps)
|
2022-05-05 16:12:17 +03:00
|
|
|
|
2024-07-15 11:02:48 +03:00
|
|
|
withPrecondition :: Assertion -> IO TestTree -> IO TestTree
|
|
|
|
withPrecondition assertion ifSuccess = do
|
|
|
|
E.catch (assertion >> ifSuccess) $ \case
|
2024-11-05 16:28:28 +03:00
|
|
|
E.SomeException e -> return (testCase @String "Precondition failed" (assertFailure (show e)))
|
2024-07-15 11:02:48 +03:00
|
|
|
|
2023-02-14 18:27:11 +03:00
|
|
|
assertEqDiffText :: String -> Text -> Text -> Assertion
|
|
|
|
assertEqDiffText = assertEqDiff unpack
|
|
|
|
|
2023-08-25 19:37:23 +03:00
|
|
|
assertEqDiff :: (Eq a) => (a -> String) -> String -> a -> a -> Assertion
|
2023-02-14 18:27:11 +03:00
|
|
|
assertEqDiff show_ msg a b
|
2022-05-05 16:12:17 +03:00
|
|
|
| a == b = return ()
|
|
|
|
| otherwise = do
|
|
|
|
putStrLn (pack $ ppDiff (getGroupedDiff pa pb))
|
|
|
|
putStrLn "End diff"
|
2022-07-25 18:51:42 +03:00
|
|
|
Monad.fail msg
|
2022-05-05 16:12:17 +03:00
|
|
|
where
|
2023-02-14 18:27:11 +03:00
|
|
|
pa = lines $ show_ a
|
|
|
|
pb = lines $ show_ b
|
|
|
|
|
|
|
|
assertEqDiffShow :: (Eq a, Show a) => String -> a -> a -> Assertion
|
|
|
|
assertEqDiffShow = assertEqDiff show
|
2022-05-05 16:12:17 +03:00
|
|
|
|
2022-12-20 15:05:40 +03:00
|
|
|
assertCmdExists :: Path Rel File -> Assertion
|
2022-05-30 14:40:52 +03:00
|
|
|
assertCmdExists cmd =
|
2022-12-20 15:05:40 +03:00
|
|
|
assertBool ("Command: " <> toFilePath cmd <> " is not present on $PATH")
|
2022-05-30 14:40:52 +03:00
|
|
|
. isJust
|
|
|
|
=<< findExecutable cmd
|
2023-12-06 20:24:59 +03:00
|
|
|
|
2024-03-20 11:56:00 +03:00
|
|
|
testTaggedLockedToIO :: (MonadIO m) => Sem PipelineAppEffects a -> m a
|
2023-12-06 20:24:59 +03:00
|
|
|
testTaggedLockedToIO =
|
2024-03-21 15:09:34 +03:00
|
|
|
runM
|
2024-07-22 18:14:37 +03:00
|
|
|
. ignoreLogger
|
2024-06-07 08:57:27 +03:00
|
|
|
. runReader testPipelineOptions
|
2023-12-06 20:24:59 +03:00
|
|
|
. runTaggedLock LockModeExclusive
|
|
|
|
|
|
|
|
testRunIO ::
|
2024-03-20 11:56:00 +03:00
|
|
|
forall a m.
|
|
|
|
(MonadIO m) =>
|
2023-12-06 20:24:59 +03:00
|
|
|
EntryPoint ->
|
|
|
|
Sem (PipelineEff PipelineAppEffects) a ->
|
2024-03-20 11:56:00 +03:00
|
|
|
m (ResolverState, PipelineResult a)
|
Parallel pipeline (#2779)
This pr introduces parallelism in the pipeline to gain performance. I've
included benchmarks at the end.
- Closes #2750.
# Flags:
There are two new global flags:
1. `-N / --threads`. It is used to set the number of capabilities.
According to [GHC
documentation](https://hackage.haskell.org/package/base-4.20.0.0/docs/GHC-Conc.html#v:setNumCapabilities):
_Set the number of Haskell threads that can run truly simultaneously (on
separate physical processors) at any given time_. When compiling in
parallel, we create this many worker threads. The default value is `-N
auto`, which sets `-N` to half the number of logical cores, capped at 8.
2. `--dev-show-thread-ids`. When given, the thread id is printed in the
compilation progress log. E.g.
![image](https://github.com/anoma/juvix/assets/5511599/9359fae2-0be1-43e5-8d74-faa82cba4034)
# Parallel compilation
1. I've added `src/Parallel/ParallelTemplate.hs` which contains all the
concurrency related code. I think it is good to keep this code separated
from the actual compiler code.
2. I've added a progress log (only for the parallel driver) that outputs
a log of the compilation progress, similar to what stack/cabal do.
# Code changes:
1. I've removed the `setup` stage where we were registering
dependencies. Instead, the dependencies are registered when the
`pathResolver` is run for the first time. This way it is safer.
1. Now the `ImportTree` is needed to run the pipeline. Cycles are
detected during the construction of this tree, so I've removed `Reader
ImportParents` from the pipeline.
3. For the package pathresolver, we do not support parallelism yet (we
could add support for it in the future, but the gains will be small).
4. When `-N1`, the pipeline remains unchanged, so performance should be
the same as in the main branch (except there is a small performance
degradation due to adding the `-threaded` flag).
5. I've introduced `PipelineOptions`, which are options that are used to
pass options to the effects in the pipeline.
6. `PathResolver` constraint has been removed from the `upTo*` functions
in the pipeline due to being redundant.
7. I've added a lot of `NFData` instances. They are needed to force the
full evaluation of `Stored.ModuleInfo` in each of the threads.
2. The `Cache` effect uses
[`SharedState`](https://hackage.haskell.org/package/effectful-core-2.3.0.1/docs/Effectful-State-Static-Shared.html)
as opposed to
[`LocalState`](https://hackage.haskell.org/package/effectful-core-2.3.0.1/docs/Effectful-Writer-Static-Local.html).
Perhaps we should provide different versions.
3. I've added a `Cache` handler that accepts a setup function. The setup
is triggered when a miss is detected. It is used to lazily compile the
modules in parallel.
# Tests
1. I've adapted the smoke test suite to ignore the progress log in the
stderr.
5. I've had to adapt `tests/positive/Internal/Lambda.juvix`. Due to
laziness, a crash happening in this file was not being caught. The
problem is that in this file we have a lambda function with different
number of patterns in their clauses, which we currently do not support
(https://github.com/anoma/juvix/issues/1706).
6. I've had to comment out the definition
```
x : Box ((A : Type) → A → A) := box λ {A a := a};
```
From the test as it was causing a crash
(https://github.com/anoma/juvix/issues/2247).
# Future Work
1. It should be investigated how much performance we lose by fully
evaluating the `Stored.ModuleInfo`, since some information in it will be
discarded. It may be possible to be more fine-grained when forcing
evaluation.
8. The scanning of imports to build the import tree is sequential. Now,
we build the import tree from the entry point module and only the
modules that are imported from it are in the tree. However, we have
discussed that at some point we should make a distinction between
`juvix` _the compiler_ and `juvix` _the build tool_. When using `juvix`
as a build tool it makes sense to typecheck/compile (to stored core) all
modules in the project. When/if we do this, scanning imports in all
modules in parallel becomes trivial.
9. The implementation of the `ParallelTemplate` uses low level
primitives such as
[forkIO](https://hackage.haskell.org/package/base-4.20.0.0/docs/Control-Concurrent.html#v:forkIO).
At some point it should be refactored to use safer functions from the
[`Effectful.Concurrent.Async`](https://hackage.haskell.org/package/effectful-2.3.0.0/docs/Effectful-Concurrent-Async.html)
module.
10. The number of cores and worker threads that we spawn is determined
by the command line. Ideally, we could use to import tree to compute an
upper bound to the ideal number of cores to use.
11. We could add an animation that displays which modules are being
compiled in parallel and which have finished being compiled.
# Benchmarks
On some benchmarks, I include the GHC runtime option
[`-A`](https://downloads.haskell.org/ghc/latest/docs/users_guide/runtime_control.html#rts-flag--A%20%E2%9F%A8size%E2%9F%A9),
which sometimes makes a good impact on performance. Thanks to
@paulcadman for pointing this out. I've figured a good combination of
`-N` and `-A` through trial and error (but this oviously depends on the
cpu and juvix projects).
## Typecheck the standard library
### Clean run (88% faster than main):
```
hyperfine --warmup 1 --prepare 'juvix clean' 'juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432' 'juvix -N 4 typecheck Stdlib/Prelude.juvix' 'juvix-main typecheck Stdlib/Prelude.juvix'
Benchmark 1: juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
Time (mean ± σ): 444.1 ms ± 6.5 ms [User: 1018.0 ms, System: 77.7 ms]
Range (min … max): 432.6 ms … 455.9 ms 10 runs
Benchmark 2: juvix -N 4 typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 628.3 ms ± 23.9 ms [User: 1227.6 ms, System: 69.5 ms]
Range (min … max): 584.7 ms … 670.6 ms 10 runs
Benchmark 3: juvix-main typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 835.9 ms ± 12.3 ms [User: 788.5 ms, System: 31.9 ms]
Range (min … max): 816.0 ms … 853.6 ms 10 runs
Summary
juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432 ran
1.41 ± 0.06 times faster than juvix -N 4 typecheck Stdlib/Prelude.juvix
1.88 ± 0.04 times faster than juvix-main typecheck Stdlib/Prelude.juvix
```
### Cached run (43% faster than main):
```
hyperfine --warmup 1 'juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432' 'juvix -N 4 typecheck Stdlib/Prelude.juvix' 'juvix-main typecheck Stdlib/Prelude.juvix'
Benchmark 1: juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
Time (mean ± σ): 241.3 ms ± 7.3 ms [User: 538.6 ms, System: 101.3 ms]
Range (min … max): 231.5 ms … 251.3 ms 11 runs
Benchmark 2: juvix -N 4 typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 235.1 ms ± 12.0 ms [User: 405.3 ms, System: 87.7 ms]
Range (min … max): 216.1 ms … 253.1 ms 12 runs
Benchmark 3: juvix-main typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 336.7 ms ± 13.3 ms [User: 269.5 ms, System: 67.1 ms]
Range (min … max): 316.9 ms … 351.8 ms 10 runs
Summary
juvix -N 4 typecheck Stdlib/Prelude.juvix ran
1.03 ± 0.06 times faster than juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
1.43 ± 0.09 times faster than juvix-main typecheck Stdlib/Prelude.juvix
```
## Typecheck the test suite of the containers library
At the moment this is the biggest juvix project that we have.
### Clean run (105% faster than main)
```
hyperfine --warmup 1 --prepare 'juvix clean' 'juvix -N 6 typecheck Main.juvix +RTS -A67108864' 'juvix -N 4 typecheck Main.juvix' 'juvix-main typecheck Main.juvix'
Benchmark 1: juvix -N 6 typecheck Main.juvix +RTS -A67108864
Time (mean ± σ): 1.006 s ± 0.011 s [User: 2.171 s, System: 0.162 s]
Range (min … max): 0.991 s … 1.023 s 10 runs
Benchmark 2: juvix -N 4 typecheck Main.juvix
Time (mean ± σ): 1.584 s ± 0.046 s [User: 2.934 s, System: 0.149 s]
Range (min … max): 1.535 s … 1.660 s 10 runs
Benchmark 3: juvix-main typecheck Main.juvix
Time (mean ± σ): 2.066 s ± 0.010 s [User: 1.939 s, System: 0.089 s]
Range (min … max): 2.048 s … 2.077 s 10 runs
Summary
juvix -N 6 typecheck Main.juvix +RTS -A67108864 ran
1.57 ± 0.05 times faster than juvix -N 4 typecheck Main.juvix
2.05 ± 0.03 times faster than juvix-main typecheck Main.juvix
```
### Cached run (54% faster than main)
```
hyperfine --warmup 1 'juvix -N 6 typecheck Main.juvix +RTS -A33554432' 'juvix -N 4 typecheck Main.juvix' 'juvix-main typecheck Main.juvix'
Benchmark 1: juvix -N 6 typecheck Main.juvix +RTS -A33554432
Time (mean ± σ): 551.8 ms ± 13.2 ms [User: 1419.8 ms, System: 199.4 ms]
Range (min … max): 535.2 ms … 570.6 ms 10 runs
Benchmark 2: juvix -N 4 typecheck Main.juvix
Time (mean ± σ): 636.7 ms ± 17.3 ms [User: 1006.3 ms, System: 196.3 ms]
Range (min … max): 601.6 ms … 655.3 ms 10 runs
Benchmark 3: juvix-main typecheck Main.juvix
Time (mean ± σ): 847.2 ms ± 58.9 ms [User: 710.1 ms, System: 126.5 ms]
Range (min … max): 731.1 ms … 890.0 ms 10 runs
Summary
juvix -N 6 typecheck Main.juvix +RTS -A33554432 ran
1.15 ± 0.04 times faster than juvix -N 4 typecheck Main.juvix
1.54 ± 0.11 times faster than juvix-main typecheck Main.juvix
```
2024-05-31 14:41:30 +03:00
|
|
|
testRunIO e =
|
|
|
|
testTaggedLockedToIO
|
|
|
|
. runIO defaultGenericOptions e
|
2023-12-06 20:24:59 +03:00
|
|
|
|
2024-03-20 11:56:00 +03:00
|
|
|
testDefaultEntryPointIO :: (MonadIO m) => Path Abs Dir -> Path Abs File -> m EntryPoint
|
Parallel pipeline (#2779)
This pr introduces parallelism in the pipeline to gain performance. I've
included benchmarks at the end.
- Closes #2750.
# Flags:
There are two new global flags:
1. `-N / --threads`. It is used to set the number of capabilities.
According to [GHC
documentation](https://hackage.haskell.org/package/base-4.20.0.0/docs/GHC-Conc.html#v:setNumCapabilities):
_Set the number of Haskell threads that can run truly simultaneously (on
separate physical processors) at any given time_. When compiling in
parallel, we create this many worker threads. The default value is `-N
auto`, which sets `-N` to half the number of logical cores, capped at 8.
2. `--dev-show-thread-ids`. When given, the thread id is printed in the
compilation progress log. E.g.
![image](https://github.com/anoma/juvix/assets/5511599/9359fae2-0be1-43e5-8d74-faa82cba4034)
# Parallel compilation
1. I've added `src/Parallel/ParallelTemplate.hs` which contains all the
concurrency related code. I think it is good to keep this code separated
from the actual compiler code.
2. I've added a progress log (only for the parallel driver) that outputs
a log of the compilation progress, similar to what stack/cabal do.
# Code changes:
1. I've removed the `setup` stage where we were registering
dependencies. Instead, the dependencies are registered when the
`pathResolver` is run for the first time. This way it is safer.
1. Now the `ImportTree` is needed to run the pipeline. Cycles are
detected during the construction of this tree, so I've removed `Reader
ImportParents` from the pipeline.
3. For the package pathresolver, we do not support parallelism yet (we
could add support for it in the future, but the gains will be small).
4. When `-N1`, the pipeline remains unchanged, so performance should be
the same as in the main branch (except there is a small performance
degradation due to adding the `-threaded` flag).
5. I've introduced `PipelineOptions`, which are options that are used to
pass options to the effects in the pipeline.
6. `PathResolver` constraint has been removed from the `upTo*` functions
in the pipeline due to being redundant.
7. I've added a lot of `NFData` instances. They are needed to force the
full evaluation of `Stored.ModuleInfo` in each of the threads.
2. The `Cache` effect uses
[`SharedState`](https://hackage.haskell.org/package/effectful-core-2.3.0.1/docs/Effectful-State-Static-Shared.html)
as opposed to
[`LocalState`](https://hackage.haskell.org/package/effectful-core-2.3.0.1/docs/Effectful-Writer-Static-Local.html).
Perhaps we should provide different versions.
3. I've added a `Cache` handler that accepts a setup function. The setup
is triggered when a miss is detected. It is used to lazily compile the
modules in parallel.
# Tests
1. I've adapted the smoke test suite to ignore the progress log in the
stderr.
5. I've had to adapt `tests/positive/Internal/Lambda.juvix`. Due to
laziness, a crash happening in this file was not being caught. The
problem is that in this file we have a lambda function with different
number of patterns in their clauses, which we currently do not support
(https://github.com/anoma/juvix/issues/1706).
6. I've had to comment out the definition
```
x : Box ((A : Type) → A → A) := box λ {A a := a};
```
From the test as it was causing a crash
(https://github.com/anoma/juvix/issues/2247).
# Future Work
1. It should be investigated how much performance we lose by fully
evaluating the `Stored.ModuleInfo`, since some information in it will be
discarded. It may be possible to be more fine-grained when forcing
evaluation.
8. The scanning of imports to build the import tree is sequential. Now,
we build the import tree from the entry point module and only the
modules that are imported from it are in the tree. However, we have
discussed that at some point we should make a distinction between
`juvix` _the compiler_ and `juvix` _the build tool_. When using `juvix`
as a build tool it makes sense to typecheck/compile (to stored core) all
modules in the project. When/if we do this, scanning imports in all
modules in parallel becomes trivial.
9. The implementation of the `ParallelTemplate` uses low level
primitives such as
[forkIO](https://hackage.haskell.org/package/base-4.20.0.0/docs/Control-Concurrent.html#v:forkIO).
At some point it should be refactored to use safer functions from the
[`Effectful.Concurrent.Async`](https://hackage.haskell.org/package/effectful-2.3.0.0/docs/Effectful-Concurrent-Async.html)
module.
10. The number of cores and worker threads that we spawn is determined
by the command line. Ideally, we could use to import tree to compute an
upper bound to the ideal number of cores to use.
11. We could add an animation that displays which modules are being
compiled in parallel and which have finished being compiled.
# Benchmarks
On some benchmarks, I include the GHC runtime option
[`-A`](https://downloads.haskell.org/ghc/latest/docs/users_guide/runtime_control.html#rts-flag--A%20%E2%9F%A8size%E2%9F%A9),
which sometimes makes a good impact on performance. Thanks to
@paulcadman for pointing this out. I've figured a good combination of
`-N` and `-A` through trial and error (but this oviously depends on the
cpu and juvix projects).
## Typecheck the standard library
### Clean run (88% faster than main):
```
hyperfine --warmup 1 --prepare 'juvix clean' 'juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432' 'juvix -N 4 typecheck Stdlib/Prelude.juvix' 'juvix-main typecheck Stdlib/Prelude.juvix'
Benchmark 1: juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
Time (mean ± σ): 444.1 ms ± 6.5 ms [User: 1018.0 ms, System: 77.7 ms]
Range (min … max): 432.6 ms … 455.9 ms 10 runs
Benchmark 2: juvix -N 4 typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 628.3 ms ± 23.9 ms [User: 1227.6 ms, System: 69.5 ms]
Range (min … max): 584.7 ms … 670.6 ms 10 runs
Benchmark 3: juvix-main typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 835.9 ms ± 12.3 ms [User: 788.5 ms, System: 31.9 ms]
Range (min … max): 816.0 ms … 853.6 ms 10 runs
Summary
juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432 ran
1.41 ± 0.06 times faster than juvix -N 4 typecheck Stdlib/Prelude.juvix
1.88 ± 0.04 times faster than juvix-main typecheck Stdlib/Prelude.juvix
```
### Cached run (43% faster than main):
```
hyperfine --warmup 1 'juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432' 'juvix -N 4 typecheck Stdlib/Prelude.juvix' 'juvix-main typecheck Stdlib/Prelude.juvix'
Benchmark 1: juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
Time (mean ± σ): 241.3 ms ± 7.3 ms [User: 538.6 ms, System: 101.3 ms]
Range (min … max): 231.5 ms … 251.3 ms 11 runs
Benchmark 2: juvix -N 4 typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 235.1 ms ± 12.0 ms [User: 405.3 ms, System: 87.7 ms]
Range (min … max): 216.1 ms … 253.1 ms 12 runs
Benchmark 3: juvix-main typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 336.7 ms ± 13.3 ms [User: 269.5 ms, System: 67.1 ms]
Range (min … max): 316.9 ms … 351.8 ms 10 runs
Summary
juvix -N 4 typecheck Stdlib/Prelude.juvix ran
1.03 ± 0.06 times faster than juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
1.43 ± 0.09 times faster than juvix-main typecheck Stdlib/Prelude.juvix
```
## Typecheck the test suite of the containers library
At the moment this is the biggest juvix project that we have.
### Clean run (105% faster than main)
```
hyperfine --warmup 1 --prepare 'juvix clean' 'juvix -N 6 typecheck Main.juvix +RTS -A67108864' 'juvix -N 4 typecheck Main.juvix' 'juvix-main typecheck Main.juvix'
Benchmark 1: juvix -N 6 typecheck Main.juvix +RTS -A67108864
Time (mean ± σ): 1.006 s ± 0.011 s [User: 2.171 s, System: 0.162 s]
Range (min … max): 0.991 s … 1.023 s 10 runs
Benchmark 2: juvix -N 4 typecheck Main.juvix
Time (mean ± σ): 1.584 s ± 0.046 s [User: 2.934 s, System: 0.149 s]
Range (min … max): 1.535 s … 1.660 s 10 runs
Benchmark 3: juvix-main typecheck Main.juvix
Time (mean ± σ): 2.066 s ± 0.010 s [User: 1.939 s, System: 0.089 s]
Range (min … max): 2.048 s … 2.077 s 10 runs
Summary
juvix -N 6 typecheck Main.juvix +RTS -A67108864 ran
1.57 ± 0.05 times faster than juvix -N 4 typecheck Main.juvix
2.05 ± 0.03 times faster than juvix-main typecheck Main.juvix
```
### Cached run (54% faster than main)
```
hyperfine --warmup 1 'juvix -N 6 typecheck Main.juvix +RTS -A33554432' 'juvix -N 4 typecheck Main.juvix' 'juvix-main typecheck Main.juvix'
Benchmark 1: juvix -N 6 typecheck Main.juvix +RTS -A33554432
Time (mean ± σ): 551.8 ms ± 13.2 ms [User: 1419.8 ms, System: 199.4 ms]
Range (min … max): 535.2 ms … 570.6 ms 10 runs
Benchmark 2: juvix -N 4 typecheck Main.juvix
Time (mean ± σ): 636.7 ms ± 17.3 ms [User: 1006.3 ms, System: 196.3 ms]
Range (min … max): 601.6 ms … 655.3 ms 10 runs
Benchmark 3: juvix-main typecheck Main.juvix
Time (mean ± σ): 847.2 ms ± 58.9 ms [User: 710.1 ms, System: 126.5 ms]
Range (min … max): 731.1 ms … 890.0 ms 10 runs
Summary
juvix -N 6 typecheck Main.juvix +RTS -A33554432 ran
1.15 ± 0.04 times faster than juvix -N 4 typecheck Main.juvix
1.54 ± 0.11 times faster than juvix-main typecheck Main.juvix
```
2024-05-31 14:41:30 +03:00
|
|
|
testDefaultEntryPointIO cwd mainFile =
|
|
|
|
testTaggedLockedToIO $
|
2024-11-01 17:42:18 +03:00
|
|
|
defaultEntryPointIO cwd (Just mainFile)
|
2023-12-06 20:24:59 +03:00
|
|
|
|
|
|
|
testDefaultEntryPointNoFileIO :: Path Abs Dir -> IO EntryPoint
|
2024-11-01 17:42:18 +03:00
|
|
|
testDefaultEntryPointNoFileIO cwd = testTaggedLockedToIO (defaultEntryPointIO cwd Nothing)
|
2023-12-06 20:24:59 +03:00
|
|
|
|
|
|
|
testRunIOEither ::
|
|
|
|
EntryPoint ->
|
|
|
|
Sem (PipelineEff PipelineAppEffects) a ->
|
2023-12-30 22:15:35 +03:00
|
|
|
IO (Either JuvixError (ResolverState, PipelineResult a))
|
Parallel pipeline (#2779)
This pr introduces parallelism in the pipeline to gain performance. I've
included benchmarks at the end.
- Closes #2750.
# Flags:
There are two new global flags:
1. `-N / --threads`. It is used to set the number of capabilities.
According to [GHC
documentation](https://hackage.haskell.org/package/base-4.20.0.0/docs/GHC-Conc.html#v:setNumCapabilities):
_Set the number of Haskell threads that can run truly simultaneously (on
separate physical processors) at any given time_. When compiling in
parallel, we create this many worker threads. The default value is `-N
auto`, which sets `-N` to half the number of logical cores, capped at 8.
2. `--dev-show-thread-ids`. When given, the thread id is printed in the
compilation progress log. E.g.
![image](https://github.com/anoma/juvix/assets/5511599/9359fae2-0be1-43e5-8d74-faa82cba4034)
# Parallel compilation
1. I've added `src/Parallel/ParallelTemplate.hs` which contains all the
concurrency related code. I think it is good to keep this code separated
from the actual compiler code.
2. I've added a progress log (only for the parallel driver) that outputs
a log of the compilation progress, similar to what stack/cabal do.
# Code changes:
1. I've removed the `setup` stage where we were registering
dependencies. Instead, the dependencies are registered when the
`pathResolver` is run for the first time. This way it is safer.
1. Now the `ImportTree` is needed to run the pipeline. Cycles are
detected during the construction of this tree, so I've removed `Reader
ImportParents` from the pipeline.
3. For the package pathresolver, we do not support parallelism yet (we
could add support for it in the future, but the gains will be small).
4. When `-N1`, the pipeline remains unchanged, so performance should be
the same as in the main branch (except there is a small performance
degradation due to adding the `-threaded` flag).
5. I've introduced `PipelineOptions`, which are options that are used to
pass options to the effects in the pipeline.
6. `PathResolver` constraint has been removed from the `upTo*` functions
in the pipeline due to being redundant.
7. I've added a lot of `NFData` instances. They are needed to force the
full evaluation of `Stored.ModuleInfo` in each of the threads.
2. The `Cache` effect uses
[`SharedState`](https://hackage.haskell.org/package/effectful-core-2.3.0.1/docs/Effectful-State-Static-Shared.html)
as opposed to
[`LocalState`](https://hackage.haskell.org/package/effectful-core-2.3.0.1/docs/Effectful-Writer-Static-Local.html).
Perhaps we should provide different versions.
3. I've added a `Cache` handler that accepts a setup function. The setup
is triggered when a miss is detected. It is used to lazily compile the
modules in parallel.
# Tests
1. I've adapted the smoke test suite to ignore the progress log in the
stderr.
5. I've had to adapt `tests/positive/Internal/Lambda.juvix`. Due to
laziness, a crash happening in this file was not being caught. The
problem is that in this file we have a lambda function with different
number of patterns in their clauses, which we currently do not support
(https://github.com/anoma/juvix/issues/1706).
6. I've had to comment out the definition
```
x : Box ((A : Type) → A → A) := box λ {A a := a};
```
From the test as it was causing a crash
(https://github.com/anoma/juvix/issues/2247).
# Future Work
1. It should be investigated how much performance we lose by fully
evaluating the `Stored.ModuleInfo`, since some information in it will be
discarded. It may be possible to be more fine-grained when forcing
evaluation.
8. The scanning of imports to build the import tree is sequential. Now,
we build the import tree from the entry point module and only the
modules that are imported from it are in the tree. However, we have
discussed that at some point we should make a distinction between
`juvix` _the compiler_ and `juvix` _the build tool_. When using `juvix`
as a build tool it makes sense to typecheck/compile (to stored core) all
modules in the project. When/if we do this, scanning imports in all
modules in parallel becomes trivial.
9. The implementation of the `ParallelTemplate` uses low level
primitives such as
[forkIO](https://hackage.haskell.org/package/base-4.20.0.0/docs/Control-Concurrent.html#v:forkIO).
At some point it should be refactored to use safer functions from the
[`Effectful.Concurrent.Async`](https://hackage.haskell.org/package/effectful-2.3.0.0/docs/Effectful-Concurrent-Async.html)
module.
10. The number of cores and worker threads that we spawn is determined
by the command line. Ideally, we could use to import tree to compute an
upper bound to the ideal number of cores to use.
11. We could add an animation that displays which modules are being
compiled in parallel and which have finished being compiled.
# Benchmarks
On some benchmarks, I include the GHC runtime option
[`-A`](https://downloads.haskell.org/ghc/latest/docs/users_guide/runtime_control.html#rts-flag--A%20%E2%9F%A8size%E2%9F%A9),
which sometimes makes a good impact on performance. Thanks to
@paulcadman for pointing this out. I've figured a good combination of
`-N` and `-A` through trial and error (but this oviously depends on the
cpu and juvix projects).
## Typecheck the standard library
### Clean run (88% faster than main):
```
hyperfine --warmup 1 --prepare 'juvix clean' 'juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432' 'juvix -N 4 typecheck Stdlib/Prelude.juvix' 'juvix-main typecheck Stdlib/Prelude.juvix'
Benchmark 1: juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
Time (mean ± σ): 444.1 ms ± 6.5 ms [User: 1018.0 ms, System: 77.7 ms]
Range (min … max): 432.6 ms … 455.9 ms 10 runs
Benchmark 2: juvix -N 4 typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 628.3 ms ± 23.9 ms [User: 1227.6 ms, System: 69.5 ms]
Range (min … max): 584.7 ms … 670.6 ms 10 runs
Benchmark 3: juvix-main typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 835.9 ms ± 12.3 ms [User: 788.5 ms, System: 31.9 ms]
Range (min … max): 816.0 ms … 853.6 ms 10 runs
Summary
juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432 ran
1.41 ± 0.06 times faster than juvix -N 4 typecheck Stdlib/Prelude.juvix
1.88 ± 0.04 times faster than juvix-main typecheck Stdlib/Prelude.juvix
```
### Cached run (43% faster than main):
```
hyperfine --warmup 1 'juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432' 'juvix -N 4 typecheck Stdlib/Prelude.juvix' 'juvix-main typecheck Stdlib/Prelude.juvix'
Benchmark 1: juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
Time (mean ± σ): 241.3 ms ± 7.3 ms [User: 538.6 ms, System: 101.3 ms]
Range (min … max): 231.5 ms … 251.3 ms 11 runs
Benchmark 2: juvix -N 4 typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 235.1 ms ± 12.0 ms [User: 405.3 ms, System: 87.7 ms]
Range (min … max): 216.1 ms … 253.1 ms 12 runs
Benchmark 3: juvix-main typecheck Stdlib/Prelude.juvix
Time (mean ± σ): 336.7 ms ± 13.3 ms [User: 269.5 ms, System: 67.1 ms]
Range (min … max): 316.9 ms … 351.8 ms 10 runs
Summary
juvix -N 4 typecheck Stdlib/Prelude.juvix ran
1.03 ± 0.06 times faster than juvix -N 4 typecheck Stdlib/Prelude.juvix +RTS -A33554432
1.43 ± 0.09 times faster than juvix-main typecheck Stdlib/Prelude.juvix
```
## Typecheck the test suite of the containers library
At the moment this is the biggest juvix project that we have.
### Clean run (105% faster than main)
```
hyperfine --warmup 1 --prepare 'juvix clean' 'juvix -N 6 typecheck Main.juvix +RTS -A67108864' 'juvix -N 4 typecheck Main.juvix' 'juvix-main typecheck Main.juvix'
Benchmark 1: juvix -N 6 typecheck Main.juvix +RTS -A67108864
Time (mean ± σ): 1.006 s ± 0.011 s [User: 2.171 s, System: 0.162 s]
Range (min … max): 0.991 s … 1.023 s 10 runs
Benchmark 2: juvix -N 4 typecheck Main.juvix
Time (mean ± σ): 1.584 s ± 0.046 s [User: 2.934 s, System: 0.149 s]
Range (min … max): 1.535 s … 1.660 s 10 runs
Benchmark 3: juvix-main typecheck Main.juvix
Time (mean ± σ): 2.066 s ± 0.010 s [User: 1.939 s, System: 0.089 s]
Range (min … max): 2.048 s … 2.077 s 10 runs
Summary
juvix -N 6 typecheck Main.juvix +RTS -A67108864 ran
1.57 ± 0.05 times faster than juvix -N 4 typecheck Main.juvix
2.05 ± 0.03 times faster than juvix-main typecheck Main.juvix
```
### Cached run (54% faster than main)
```
hyperfine --warmup 1 'juvix -N 6 typecheck Main.juvix +RTS -A33554432' 'juvix -N 4 typecheck Main.juvix' 'juvix-main typecheck Main.juvix'
Benchmark 1: juvix -N 6 typecheck Main.juvix +RTS -A33554432
Time (mean ± σ): 551.8 ms ± 13.2 ms [User: 1419.8 ms, System: 199.4 ms]
Range (min … max): 535.2 ms … 570.6 ms 10 runs
Benchmark 2: juvix -N 4 typecheck Main.juvix
Time (mean ± σ): 636.7 ms ± 17.3 ms [User: 1006.3 ms, System: 196.3 ms]
Range (min … max): 601.6 ms … 655.3 ms 10 runs
Benchmark 3: juvix-main typecheck Main.juvix
Time (mean ± σ): 847.2 ms ± 58.9 ms [User: 710.1 ms, System: 126.5 ms]
Range (min … max): 731.1 ms … 890.0 ms 10 runs
Summary
juvix -N 6 typecheck Main.juvix +RTS -A33554432 ran
1.15 ± 0.04 times faster than juvix -N 4 typecheck Main.juvix
1.54 ± 0.11 times faster than juvix-main typecheck Main.juvix
```
2024-05-31 14:41:30 +03:00
|
|
|
testRunIOEither entry =
|
|
|
|
testTaggedLockedToIO
|
|
|
|
. runIOEither entry
|
2023-12-06 20:24:59 +03:00
|
|
|
|
|
|
|
testRunIOEitherTermination ::
|
|
|
|
EntryPoint ->
|
|
|
|
Sem (Termination ': PipelineEff PipelineAppEffects) a ->
|
2023-12-30 22:15:35 +03:00
|
|
|
IO (Either JuvixError (ResolverState, PipelineResult a))
|
2023-12-06 20:24:59 +03:00
|
|
|
testRunIOEitherTermination entry =
|
2023-12-30 22:15:35 +03:00
|
|
|
testRunIOEither entry
|
2023-12-06 20:24:59 +03:00
|
|
|
. evalTermination iniTerminationState
|
2024-01-11 15:04:38 +03:00
|
|
|
|
|
|
|
assertFailure :: (MonadIO m) => String -> m a
|
|
|
|
assertFailure = liftIO . HUnit.assertFailure
|
2024-05-29 14:34:04 +03:00
|
|
|
|
2024-11-05 16:28:28 +03:00
|
|
|
runSimpleErrorHUnit :: (Members '[EmbedIO] r) => Sem (Error SimpleError ': r) a -> Sem r a
|
|
|
|
runSimpleErrorHUnit m = do
|
|
|
|
res <- runError m
|
|
|
|
case res of
|
|
|
|
Left (SimpleError msg) -> assertFailure (toPlainString msg)
|
|
|
|
Right r -> return r
|
|
|
|
|
2024-07-12 19:31:09 +03:00
|
|
|
wantsError ::
|
|
|
|
forall err b.
|
|
|
|
(Generic err, GenericHasConstructor (GHC.Rep err)) =>
|
|
|
|
(b -> err) ->
|
|
|
|
Path Abs File ->
|
|
|
|
err ->
|
|
|
|
Maybe String
|
|
|
|
wantsError wanted file actualErr
|
|
|
|
| genericSameConstructor wantedErr actualErr = Nothing
|
|
|
|
| otherwise =
|
|
|
|
Just
|
|
|
|
( "In "
|
|
|
|
<> prettyString file
|
|
|
|
<> "\nExpected "
|
|
|
|
<> genericConstructorName wantedErr
|
|
|
|
<> "\nFound "
|
|
|
|
<> genericConstructorName actualErr
|
|
|
|
)
|
|
|
|
where
|
|
|
|
wantedErr :: err
|
|
|
|
wantedErr = wanted impossible
|
|
|
|
|
2024-05-29 14:34:04 +03:00
|
|
|
-- | The same as `P.readProcess` but instead of inheriting `stderr` redirects it
|
|
|
|
-- to the child's `stdout`.
|
|
|
|
readProcess :: FilePath -> [String] -> Text -> IO Text
|
2024-06-07 08:57:27 +03:00
|
|
|
readProcess = readProcessCwd' Nothing Nothing
|
|
|
|
|
|
|
|
readProcessWithEnv :: [(String, String)] -> FilePath -> [String] -> Text -> IO Text
|
|
|
|
readProcessWithEnv env = readProcessCwd' (Just env) Nothing
|
2024-05-29 14:34:04 +03:00
|
|
|
|
|
|
|
readProcessCwd :: FilePath -> FilePath -> [String] -> Text -> IO Text
|
2024-06-07 08:57:27 +03:00
|
|
|
readProcessCwd cwd = readProcessCwd' Nothing (Just cwd)
|
2024-05-29 14:34:04 +03:00
|
|
|
|
2024-06-07 08:57:27 +03:00
|
|
|
readProcessCwd' :: Maybe [(String, String)] -> Maybe FilePath -> FilePath -> [String] -> Text -> IO Text
|
|
|
|
readProcessCwd' menv mcwd cmd args stdinText =
|
2024-05-29 14:34:04 +03:00
|
|
|
withTempDir'
|
|
|
|
( \dirPath -> do
|
|
|
|
(_, hin) <- openTempFile dirPath "stdin"
|
|
|
|
(_, hout) <- openTempFile dirPath "stdout"
|
|
|
|
hPutStr hin stdinText
|
|
|
|
hSeek hin AbsoluteSeek 0
|
|
|
|
(_, _, _, ph) <-
|
|
|
|
P.createProcess_
|
|
|
|
"readProcess"
|
|
|
|
(P.proc cmd args)
|
|
|
|
{ P.std_in = P.UseHandle hin,
|
|
|
|
P.std_out = P.UseHandle hout,
|
|
|
|
P.std_err = P.UseHandle hout,
|
2024-06-07 08:57:27 +03:00
|
|
|
P.cwd = mcwd,
|
|
|
|
P.env = menv
|
2024-05-29 14:34:04 +03:00
|
|
|
}
|
|
|
|
P.waitForProcess ph
|
|
|
|
hSeek hout AbsoluteSeek 0
|
|
|
|
r <- hGetContents hout
|
|
|
|
hClose hin
|
|
|
|
hClose hout
|
|
|
|
return r
|
|
|
|
)
|
2024-11-05 16:28:28 +03:00
|
|
|
|
|
|
|
to3DigitString :: Int -> Text
|
|
|
|
to3DigitString n
|
|
|
|
| n < 10 = "00" <> show n
|
|
|
|
| n < 100 = "0" <> show n
|
|
|
|
| n < 1000 = show n
|
|
|
|
| otherwise = error ("The given number has more than 3 digits. Given number = " <> prettyText n)
|
|
|
|
|
|
|
|
-- | E.g. Test001: str
|
|
|
|
numberedTestName :: Int -> Text -> Text
|
|
|
|
numberedTestName i str = "Test" <> to3DigitString i <> ": " <> str
|
|
|
|
|
|
|
|
testCase :: (HasTextBackend str) => str -> Assertion -> TestTree
|
|
|
|
testCase name = HUnit.testCase (toPlainString name)
|