| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
| |
[skip ci]
Fixes #15429
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is a one line change. It is a fixup from MR!7325, was pointed out
in review of MR!7442, specifically: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/7442#note_406581
The change removes isNCG check from cgTopBinding. Instead it changes the
type of binBlobThresh in DynFlags from Word to Maybe Word, where a Just
0 or a Nothing indicates an infinite threshold and thus the disable
CmmFileEmbed case in the original check.
This improves the cohesion of the module because more NCG related
Backend stuff is moved into, and checked in, StgToCmm.Config. Note, that
the meaning of a Just 0 or a Nothing in binBlobThresh is indicated in a
comment next to its field in GHC.StgToCmm.Config.
DynFlags: binBlobThresh: Word -> Maybe Word
StgToCmm.Config: binBlobThesh add not ncg check
DynFlags.binBlob: move Just 0 check to dflags init
StgToCmm.binBlob: only check isNCG, Just 0 check to dflags
StgToCmm.Config: strictify binBlobThresh
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This adds a number of changes to ticky-ticky profiling.
When an executable is profiled with IPE profiling it's now possible to
associate id-related ticky counters to their source location.
This works by emitting the info table address as part of the counter
which can be looked up in the IPE table.
Add a `-ticky-ap-thunk` flag. This flag prevents the use of some standard thunks
which are precompiled into the RTS. This means reduced cache locality
and increased code size. But it allows better attribution of execution
cost to specific source locations instead of simple attributing it to
the standard thunk.
ticky-ticky now uses the `arg` field to emit additional information
about counters in json format. When ticky-ticky is used in combination
with the eventlog eventlog2html can be used to generate a html table
from the eventlog similar to the old text output for ticky-ticky.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We were not looking through floatable ticks when collecting arguments in
Core Lint, which caused `checkCanEtaExpand` to fail on something like:
```haskell
reallyUnsafePtrEquality
= \ @a ->
(src<loc> reallyUnsafePtrEquality#)
@Lifted @a @Lifted @a
```
We fix this by using `collectArgsTicks tickishFloatable` instead of
`collectArgs`, to be consistent with the behaviour of eta expansion
outlined in Note [Eta expansion and source notes] in GHC.Core.Opt.Arity.
Fixes #21152.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
As noted in #21071 we were missing adding this edge so there were
situations where the .hs file would get compiled before the .hs-boot
file which leads to issues with -j.
I fixed this properly by adding the edge in downsweep so the definition
of nodeDependencies can be simplified to avoid adding this dummy edge
in.
There are plenty of tests which seem to have these redundant boot files
anyway so no new test. #21094 tracks the more general issue of
identifying redundant hs-boot and SOURCE imports.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In the case when we tell moduleGraphNodes to drop hs-boot files the idea
is to collapse hs-boot files into their hs file nodes. In the old code
* nodeDependencies changed edges from IsBoot to NonBoot
* moduleGraphNodes just dropped boot file nodes
The net result is that any dependencies of the hs-boot files themselves
were dropped. The correct thing to do is
* nodeDependencies changes edges from IsBoot to NonBoot
* moduleGraphNodes merges dependencies of IsBoot and NonBoot nodes.
The result is a properly quotiented dependency graph which contains no
hs-boot files nor hs-boot file edges.
Why this didn't cause endless issues when compiling with boot files, we
will never know.
|
| |
|
|
|
| |
It seems like a commented out section of code was accidentally included
in the docstring for a field.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These flags were added to help with the high linking cost of the old
split-objs mode. Now we are using split-sections these flags appear to
make no difference to memory usage or time taken to link.
I tested various configurations linking together the ghc library with
-split-sections enabled.
| linker | time (s) |
| ------ | ------ |
| gold | 0.95 |
| ld | 1.6 |
| ld (hash-size = 31, reduce-memory-overheads) | 1.6 |
| ldd | 0.47 |
Fixes #20967
|
| |
|
|
|
|
|
|
| |
Remove these smart constructors for these reasons:
* mkLocalClosureTableLabel : Does the same as the non-local variant.
* mkLocalClosureLabel : Does the same as the non-local variant.
* mkLocalInfoTableLabel : Decide if we make a local label based on the name
and just use mkInfoTableLabel everywhere.
|
| |
|
|
|
|
|
| |
If a setting is missing from the configuration file it's likely the user
needs to reconfigure.
Fixes #20476
|
| |
|
|
|
| |
This does not fix all hlint issues as the criticised index and
length expressions seem to be fine in context.
|
| |
|
|
| |
fix #18963
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch adds a check to Core Lint, checkCanEtaExpand,
which ensures that primops and other wired-in functions with
no binding such as unsafeCoerce#, oneShot, rightSection...
can always be eta-expanded, by checking that the remaining
argument types have a fixed RuntimeRep.
Two subtleties came up:
- the notion of arity in Core looks through newtypes, so we may
need to unwrap newtypes in this check,
- we want to avoid calling hasNoBinding on something whose unfolding
we are in the process of linting, as this would cause a loop;
to avoid this we add some information to the Core Lint environment
that holds this information.
Fixes #20480
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The code in tcAnonWildCardOcc assumed that it could never encounter
anonymous wildcards in illegal positions, because the renamer would
have ruled them out. However, it's possible to sneak past the checks
in the renamer by using Template Haskell. It isn't possible to simply
pass on additional information when renaming Template Haskell
brackets, because we don't know in advance in what context the bracket
will be spliced in (see test case T15433b). So we accept that we might
encounter these bogus wildcards in the typechecker and throw the
appropriate error.
This patch also migrates the error messages for illegal wildcards in
types to use the diagnostic infrastructure.
Fixes #15433
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Essentially we apply the identity:
> lookupNameEnv n (plusNameEnv_C (++) rb1 rb2)
> = lookupNameEnv n rb1 ++ lookupNameEnv n rb2
The latter being more efficient as we don't construct an intermediate
map.
This is now quite important as each time we try and apply rules we need
to combine the current EPS RuleBase with the HPT and ModGuts rule bases.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As the prescient (now deleted) note warns in simplifyPgmIO we have to be a bit careful
about when we gather rules from the EPS so that we get the rules for
imported bindings.
```
-- Get any new rules, and extend the rule base
-- See Note [Overall plumbing for rules] in GHC.Core.Rules
-- We need to do this regularly, because simplification can
-- poke on IdInfo thunks, which in turn brings in new rules
-- behind the scenes. Otherwise there's a danger we'll simply
-- miss the rules for Ids hidden inside imported inlinings
```
Given the previous commit, the loading of unfoldings is now even more
delayed so we need to be more careful to read the EPS rule base closer to the point
where we decide to try rules.
Without this fix GHC performance regressed by a noticeably amount
because the `zip` rule was not brought into scope eagerly enough which
led to a further series of unfortunate events in the simplifer which
tipped `substTyWithCoVars` over the edge of the size threshold, stopped
it being inlined and increased allocations by 10% in some cases.
Furthermore, this change is noticeably in the testsuite as it changes
T19790 so that the `length` rules from GHC.List fires earlier.
-------------------------
Metric Increase:
T9961
-------------------------
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
The old logic was unecessarily strict in loading unfoldings because when
reading the unfolding we would case on the result of attempting to load
the template before commiting to which type of unfolding we were
producing. Hence trying to inspect any of the information about an
unfolding would force the template to be loaded.
This also removes a potentially hard to discover bug where if the
template failed to be typechecked for some reason then we would just not
return an unfolding. Instead we now panic so these bad situations which
should never arise can be identified.
|
| |
|
|
|
|
| |
Now it takes a better account of precise vs. imprecise exception semantics.
Fixes #19854.
|
| |
|
|
| |
This is used by the JS backend for serialization.
|
| |
|
|
|
| |
This is a minor change in preparation for the JS backend: CgInfos aren't
mandatory and the JS backend won't return them.
|
| |
|
|
| |
Taken from !3658
|
| |
|
|
|
|
|
| |
The parser didn't allow qualified constructor names to appear
in COMPLETE pragmas. This patch fixes that.
Fixes #20551
|
| | |
|
| |
|
|
|
|
|
|
|
| |
This patch adds some derived instances to
`GHC.Data.Bool.OverridingBool`. It also changes the order of the
constructors, so that the derived `Ord` instance matches the
behaviour for `Maybe Bool`.
Fixes #20326
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Note [Matching lets] in GHC.Core.Rules claims the following:
> We use GHC.Core.Subst.substBind to freshen the binding, using an
> in-scope set that is the original in-scope variables plus the
> rs_bndrs (currently floated let-bindings).
However, previously the implementation didn't actually do extend the
in-scope set with rs_bndrs. This appears to be a regression which was
introduced by 4ff4d434e9a90623afce00b43e2a5a1ccbdb4c05.
Moreover, the originally reasoning was subtly wrong: we must rather use
the in-scope set from rv_lcl, extended with rs_bndrs, not that of
`rv_fltR`
Fixes #21122.
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
Tag inference included a way to collect stats about avoided tag-checks.
This was dony by emitting "dummy" ticky entries with counts corresponding
to predicted/unpredicated tag checks.
This behaviour for ticky is now gated behind -fticky-tag-checks.
I also documented ticky-LNE in the process.
|
| |
|
|
|
|
| |
@nrnrnr points out that on his machine ld.lld rejects text relocations.
Generalize the Darwin text-relocation avoidance logic to account for
this.
|
| |
|
|
|
|
|
|
|
| |
This patch skips performing type normalisation when we haven't
fully instantiated the type. That is, in tcRnExpr
(used only for :type in GHCi), skip normalisation if
the result type responds True to isSigmaTy.
Fixes #20974
|
| |
|
|
|
|
| |
Fix two issues regarding printing numeric literals.
Fixing #20454.
|
| | |
|
| |
|
|
|
|
|
| |
operations (#20214)
Previously, when trying to load module with SIMD vector operations, ghci would panic
in 'GHC.StgToByteCode.findPushSeq'. Now, a more helpful message is displayed.
|
| |
|
|
|
| |
It turns out this job hasn't been running for quite a while (perhaps
ever) so there are quite a few failures when running the linter locally.
|
| |
|
|
|
|
| |
If you are running the `lint:{base/compiler}` command locally then this
improves the responsiveness because we don't re-run configure everytime
if the header file already exists.
|
| |
|
|
|
|
|
|
|
|
| |
This MR moves the GHC linters into the tree, so that they can be run directly using Hadrian.
* Query all files tracked by Git instead of using changed files, so that we can run the exact same linting step locally and in a merge request.
* Only check that the changelogs don't contain TBA when RELEASE=YES.
* Add hadrian/lint script, which runs all the linting steps.
* Ensure the hlint job exits with a failure if hlint is not installed (otherwise we were ignoring the failure). Given that hlint doesn't seem to be available in CI at the moment, I've temporarily allowed failure in the hlint job.
* Run all linting tests in CI using hadrian.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It is unclear what `TestEquality` is for. There are 3 possible choices.
Assuming
```haskell
data Tag a where
TagInt1 :: Tag Int
TagInt2 :: Tag Int
```
Weakest -- type param equality semi-decidable
---------------------------------------------
`Just Refl` merely means the type params are equal, the values being compared might not be.
`Nothing` means the type params may or may not be not equal.
```haskell
instance TestEquality Tag where
testEquality TagInt1 TagInt1 = Nothing -- oopsie is allowed
testEquality TagInt1 TagInt2 = Just Refl
testEquality TagInt2 TagInt1 = Just Refl
testEquality TagInt2 TagInt2 = Just Refl
```
This option is better demonstrated with a different type:
```haskell
data Tag' a where
TagInt1 :: Tag Int
TagInt2 :: Tag a
```
```haskell
instance TestEquality Tag' where
testEquality TagInt1 TagInt1 = Just Refl
testEquality TagInt1 TagInt2 = Nothing -- can't be sure
testEquality TagInt2 TagInt1 = Nothing -- can't be sure
testEquality TagInt2 TagInt2 = Nothing -- can't be sure
```
Weaker -- type param equality decidable
---------------------------------------
`Just Refl` merely means the type params are equal, the values being compared might not be.
`Nothing` means the type params are not equal.
```haskell
instance TestEquality Tag where
testEquality TagInt1 TagInt1 = Just Refl
testEquality TagInt1 TagInt2 = Just Refl
testEquality TagInt2 TagInt1 = Just Refl
testEquality TagInt2 TagInt2 = Just Refl
```
Strong -- Like `Eq`
-------------------
`Just Refl` means the type params are equal, and the values are equal according to `Eq`.
```haskell
instance TestEquality Tag where
testEquality TagInt1 TagInt1 = Just Refl
testEquality TagInt2 TagInt2 = Just Refl
testEquality _ _ = Nothing
```
Strongest -- unique value concrete type
---------------------------------------
`Just Refl` means the type params are equal, and the values are equal, and the class assume if the type params are equal the values must also be equal. In other words, the type is a singleton type when the type parameter is a closed term.
```haskell
-- instance TestEquality -- invalid instance because two variants for `Int`
```
------
The discussion in
https://github.com/haskell/core-libraries-committee/issues/21 has
decided on the "Weaker" option (confusingly formerly called the
"Weakest" option). So that is what is implemented.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The doc says that the last statement of an ado-block can be one of
`return E`, `return $ E`, `pure E` and `pure $ E`. But `return`
is not accepted in a few cases such as:
```haskell
-- The ado-block only has one statement
x :: F ()
x = do
return ()
-- The ado-block only has let-statements besides the `return`
y :: F ()
y = do
let a = True
return ()
```
These currently require `Monad` instances. This MR fixes it.
Normally `return` is accepted as the last statement because it is
stripped in constructing an `ApplicativeStmt`, but this cannot be
done in the above cases, so instead we replace `return` by `pure`.
A similar but different issue (when the ado-block contains `BindStmt`
or `BodyStmt`, the second last statement cannot be `LetStmt`, even if
the last statement uses `pure`) is fixed in !6786.
|
| |
|
|
|
|
|
| |
Pointers to closures must be untagged before use.
Produce closures of different types so we get different info tables.
Fixes #21112
|
| |
|
|
|
|
|
|
| |
The outputs of these tests vary on the order interface files are loaded
so we normalise the output to correct for these inconsequential
differences.
Fixes #21121
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The idea of the needsTemplateHaskellOrQQ query is to check if any of the
modules in a module graph need Template Haskell then enable -dynamic-too
if necessary. This is quite imprecise though as it will enable
-dynamic-too for all modules in the module graph even if only one module
uses template haskell, with multiple home units, this is obviously even
worse.
With -fno-code we already have similar logic to enable code generation
just for the modules which are dependeded on my TemplateHaskell modules
so we use the same code path to decide whether to enable -dynamic-too
rather than using this big hammer.
This is part of the larger overall goal of moving as much statically
known configuration into the downsweep as possible in order to have
fully decided the build plan and all the options before starting to
build anything.
I also included a fix to #21095, a long standing bug with with the logic
which is supposed to enable the external interpreter if we don't have
the internal interpreter.
Fixes #20696 #21095
|
| |
|
|
|
| |
It was unused in the compiler so I have removed it to streamline
ModuleGraph.
|
| | |
|
| |
|
|
|
|
|
|
| |
Several 64-bit operation were implemented with FFI calls on 32-bit
architectures but we can easily implement them with inline assembly
code.
Also remove unused hs_int64ToWord64 and hs_word64ToInt64 C functions.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
* add getLocalRegReg to avoid allocating a CmmLocal just to call
getRegisterReg
* 64-bit registers: in the general case we must always use the virtual
higher part of the register, so we might as well always return it with
the lower part. The only exception is to implement 64-bit to 32-bit
conversions. We now have to explicitly discard the higher part when
matching on Reg64/RegCode64 datatypes instead of explicitly fetching
the higher part from the lower part: much safer default.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Preliminary work done to make working on #5444 easier.
Mostly make make control-flow easier to follow:
* renamed genCCall into genForeignCall
* split genForeignCall into the part dispatching on PrimTarget (genPrim) and
the one really generating code for a C call (cf ForeignTarget and genCCall)
* made genPrim/genSimplePrim only dispatch on MachOp: each MachOp now
has its own code generation function.
* out-of-line primops are not handled in a partial `outOfLineCmmOp`
anymore but in the code generation functions directly. Helper
functions have been introduced (e.g. genLibCCall) for code sharing.
* the latter two bullets make code generated for primops that are only
sometimes out-of-line (e.g. Pdep or Memcpy) and the logic to select
between inline/out-of-line much more localized
* avoided passing is32bit as an argument as we can easily get it from NatM
state when we really need it
* changed genCCall type to avoid it being partial (it can't handle
PrimTarget)
* globally removed 12 calls to `panic` thanks to better control flow and
types ("parse, don't validate" ftw!).
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The old return type `(RecompRequired, Maybe _)`, was confusing
because it was inhabited by values like `(UpToDate, Nothing)` that made
no sense.
The new type ensures:
- you must provide a value if it is up to date.
- you must provide a reason if you don't provide a value.
it is used as the return value of:
- `checkOldIface`
- `checkByteCode`
- `checkObjects`
|
