| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
| |
As noted in #20569 the documentation for search path was wrong because
it seemed to indicate that `-i` dirs were important when looking for
interface files in `-c` mode, but they are not important if `-hidir` is
set.
Fixes #20569
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
How things should work:
* -i is the search path for source files
* -hidir explicitly sets the search path for interface files and the output location for interface files.
* -odir sets the search path and output location for object files.
Before in one shot mode we would look for the interface file in the
search locations given by `-i`, but then set the path to be in the
`hidir`, so in unusual situations the finder could find an interface
file in the `-i` dir but later fail because it tried to read the
interface file from the `-hidir`.
A bug identified by #20569
|
| |
|
|
| |
filter has fusion rules that filterOut lacks
|
| | |
|
| |
|
|
|
|
| |
Remove GHC.Driver.Session imports that weren't considered as redundant
because of the reexport of PlatformConstants. Also remove this reexport
as modules using this datatype should import GHC.Platform instead.
|
| |
|
|
|
|
|
|
| |
Previously `markCAFs` would call `markObjectCode` even in non-major GCs.
This is problematic since `prepareUnloadCheck` is not called in such
GCs, meaning that the section index has not been updated.
Fixes #21254
|
| |
|
|
|
|
|
|
| |
Previously we failed to untag the function closure when scavenging the
payload of a PAP, resulting in an invalid closure pointer being passed
to scavenge_large_bitmap and consequently #21254. Fix this.
Fixes #21254
|
| |
|
|
| |
Change the wording to make it clear that the upper bound is non-inclusive.
|
| |
|
|
|
|
|
|
|
|
| |
In !7604 we started placing adjustor templates in the data section on
Linux as some toolchains there reject relocations in the text section.
However, it turns out that OpenBSD also exhibits this restriction.
Fix this by *always* placing adjustor templates in the data section.
Fixes #21155.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Ticket #21110 points out that -Wunused-packages behaves a bit unusually
in GHCi. Now we define the semantics for -Wunused-packages in
interactive mode as follows:
* If you use -Wunused-packages on an initial load then the warning is reported.
* If you explicitly set -Wunused-packages on the command line then the
warning is displayed (until it is disabled)
* If you then subsequently modify the set of available targets by using
:load or :cd (:cd unloads everything) then the warning is (silently)
turned off.
This means that every :r the warning is printed if it's turned on (but you did ask for it).
Fixes #21110
|
| | |
|
| |
|
|
|
| |
LLVM 13.0.0 is released in Oct 2021, and latest head validates against
LLVM 13 just fine if LlvmMaxVersion is bumped.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Names appearing in Haddock docstrings are lexed and renamed like any other names
appearing in the AST. We currently rename names irrespective of the namespace,
so both type and constructor names corresponding to an identifier will appear in
the docstring. Haddock will select a given name as the link destination based on
its own heuristics.
This patch also restricts the limitation of `-haddock` being incompatible with
`Opt_KeepRawTokenStream`.
The export and documenation structure is now computed in GHC and serialised in
.hi files. This can be used by haddock to directly generate doc pages without
reparsing or renaming the source. At the moment the operation of haddock
is not modified, that's left to a future patch.
Updates the haddock submodule with the minimum changes needed.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This patch improves code generation for derived Eq instances.
The idea is to use 'dataToTag' to evaluate both arguments.
This allows to 'short-circuit' when tags do not match.
Unfortunately, inner evals are still present when we branch
on tags. This is due to the way 'dataToTag#' primop
evaluates its argument in the code generator. #21207 was
created to explore further optimizations.
Metric Decrease:
LargeRecord
|
| | |
|
| |
|
|
| |
Fixes #21251
|
| |
|
|
|
|
|
| |
GHC.Event.Internal was not listed in base.cabal on Windows. This caused
undefined reference errors. This patch adds it back, by moving it out
of the OS-specific logic in base.cabal.
Fixes #21245.
|
| |
|
|
|
|
|
|
| |
The pretty printer for regular data types already accounted for these,
and had some duplication with the newtype pretty printer.
Factoring the logic out into a common function and using it for both
newtypes and data declarations is enough to fix the bug.
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
Previously while constructing the jump table index we would
zero-extend the discriminant before subtracting the start of the
jump-table. This goes subtly wrong in the case of a sub-word, signed
discriminant, as described in the included Note. Fix this in both the
PPC and X86 NCGs.
Fixes #21186.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Rewrite the critical notes and fix outdated ones,
use `HsQuote GhcRn` (in `HsBracketTc`) for desugaring regardless of the
bracket being typed or untyped,
remove unused `EpAnn` from `Hs*Bracket GhcRn`,
zonkExpr factor out common brackets code,
ppr_expr factor out common brackets code,
and fix tests,
to finish MR https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4782.
-------------------------
Metric Decrease:
hard_hole_fits
-------------------------
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
An untyped bracket `HsQuote p` can never be constructed with
`p ~ GhcTc`. This is because we don't typecheck `HsQuote` at all.
That's OK, because we also never use `HsQuote GhcTc`.
To enforce this at the type level we make `HsQuote GhcTc` isomorphic
to `NoExtField` and impossible to construct otherwise, by using TTG field
extensions to make all constructors, except for `XQuote` (which takes `NoExtField`),
unconstructable, with `DataConCantHappen`
This is explained more in detail in Note [The life cycle of a TH quotation]
Related discussion: https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4782
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
When desugaring a bracket we want to desugar /renamed/ rather than
/typechecked/ code; So in (HsExpr GhcTc) tree, we must
have a (HsExpr GhcRn) for the quotation itself.
This commit reworks the TTG refactor on typed and untyped brackets by
storing the /renamed/ code in the bracket field extension rather than in
the constructor extension in `HsQuote` (previously called
`HsUntypedBracket`)
See Note [The life cycle of a TH quotation] and https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4782
|
| |
|
|
|
|
|
|
|
|
|
| |
Split HsBracket into HsTypedBracket and HsUntypedBracket.
Unfortunately, we still cannot get rid of
instance XXTypedBracket GhcTc = HsTypedBracket GhcRn
despite no longer requiring it for typechecking, but rather because the
TH desugarer works on GhcRn rather than GhcTc (See GHC.HsToCore.Quote)
|
| |
|
|
|
|
|
|
|
|
| |
When HsExpr GhcTc, the HsBracket constructor should hold a HsBracket
GhcRn, rather than an HsBracket GhcTc.
We make use of the HsBracket p extension constructor (XBracket
(XXBracket p)) to hold an HsBracket GhcRn when the pass is GhcTc
See !4782 https://gitlab.haskell.org/ghc/ghc/-/merge_requests/4782
|
| | |
|
| |
|
|
|
|
| |
The late cost centre flag was renamed from -fprof-late-ccs
to -fprof-late in 7fe07143, but this change hadn't been
propagated to Hadrian.
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
This problem was due to a bug in cloneWanted, which was incorrectly
creating a coercion hole to hold an evidence variable.
This bug was introduced by 8bb52d91 and fixed in 81740ce8.
Fixes #21130
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, we let `Unboxed` win in `lubBoxity`, which is unsoundly optimistic
in terms ob Boxity analysis. "Unsoundly" in the sense that we sometimes unbox
parameters that we better shouldn't unbox. Examples are #18907 and T19871.absent.
Until now, we thought that this hack pulled its weight becuase it worked around
some shortcomings of the phase separation between Boxity analysis and CPR
analysis. But it is a gross hack which caused regressions itself that needed all
kinds of fixes and workarounds. See for example #20767. It became impossible to
work with in !7599, so I want to remove it.
For example, at the moment, `lubDmd B dmd` will not unbox `dmd`,
but `lubDmd A dmd` will. Given that `B` is supposed to be the bottom element of
the lattice, it's hardly justifiable to get a better demand when `lub`bing with
`A`.
The consequence of letting `Boxed` win in `lubBoxity` is that we *would* regress
#2387, #16040 and parts of #5075 and T19871.sumIO, until Boxity and CPR
are able to communicate better. Fortunately, that is not the case since I could
tweak the other source of optimism in Boxity analysis that is described in
`Note [Unboxed demand on function bodies returning small products]` so that
we *recursively* assume unboxed demands on function bodies returning small
products. See the updated Note.
`Note [Boxity for bottoming functions]` describes why we need bottoming
functions to have signatures that say that they deeply unbox their arguments.
In so doing, I had to tweak `finaliseArgBoxities` so that it will never unbox
recursive data constructors. This is in line with our handling of them in CPR.
I updated `Note [Which types are unboxed?]` to reflect that.
In turn we fix #21119, #20767, #18907, T19871.absent and get a much simpler
implementation (at least to think about). We can also drop the very ad-hoc
definition of `deferAfterPreciseException` and its Note in favor of the
simple, intuitive definition we used to have.
Metric Decrease:
T16875
T18223
T18698a
T18698b
hard_hole_fits
Metric Increase:
LargeRecord
MultiComponentModulesRecomp
T15703
T8095
T9872d
Out of all the regresions, only the one in T9872d doesn't vanish in a perf
build, where the compiler is bootstrapped with -O2 and thus SpecConstr.
Reason for regressions:
* T9872d is due to `ty_co_subst` taking its `LiftingContext` boxed.
That is because the context is passed to a function argument, for
example in `liftCoSubstTyVarBndrUsing`.
* In T15703, LargeRecord and T8095, we get a bit more allocations in
`expand_syn` and `piResultTys`, because a `TCvSubst` isn't unboxed.
In both cases that guards against reboxing in some code paths.
* The same is true for MultiComponentModulesRecomp, where we get less unboxing
in `GHC.Unit.Finder.$wfindInstalledHomeModule`. In a perf build, allocations
actually *improve* by over 4%!
Results on NoFib:
--------------------------------------------------------------------------------
Program Allocs Instrs
--------------------------------------------------------------------------------
awards -0.4% +0.3%
cacheprof -0.3% +2.4%
fft -1.5% -5.1%
fibheaps +1.2% +0.8%
fluid -0.3% -0.1%
ida +0.4% +0.9%
k-nucleotide +0.4% -0.1%
last-piece +10.5% +13.9%
lift -4.4% +3.5%
mandel2 -99.7% -99.8%
mate -0.4% +3.6%
parser -1.0% +0.1%
puzzle -11.6% +6.5%
reverse-complem -3.0% +2.0%
scs -0.5% +0.1%
sphere -0.4% -0.2%
wave4main -8.2% -0.3%
--------------------------------------------------------------------------------
Summary excludes mandel2 because of excessive bias
Min -11.6% -5.1%
Max +10.5% +13.9%
Geometric Mean -0.2% +0.3%
--------------------------------------------------------------------------------
Not bad for a bug fix.
The regression in `last-piece` could become a win if SpecConstr would work on
non-recursive functions. The regression in `fibheaps` is due to
`Note [Reboxed crud for bottoming calls]`, e.g., #21128.
|
| |
|
|
|
|
| |
We still don't allow negative overloaded patterns. Earler all negative patterns
were treated as negative overloaded patterns. Now, we expliclty check the
extension field to see if the pattern is actually a negative overloaded pattern
|
| |
|
|
|
|
| |
- Use extension suggestion hints instead of suggesting extensions in the
error message body for several FFI errors.
- Adds a test case for `TcRnForeignImportPrimExtNotSet`
|
| |
|
|
| |
Converts all uses of 'TcRnUnknownMessage' to proper diagnostics.
|
| |
|
|
|
|
|
|
|
|
| |
* Users can define their own (~) type operator
* Haddock can display documentation for the built-in (~)
* New transitional warnings implemented:
-Wtype-equality-out-of-scope
-Wtype-equality-requires-operators
Updates the haddock submodule.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As `Note [Demand analysis for recursive data constructors]` describes, we now
refrain from unboxing recursive data type arguments, for two reasons:
1. Relating to run/alloc perf: Similar to
`Note [CPR for recursive data constructors]`, it seldomly improves run/alloc
performance if we just unbox a finite number of layers of a potentially huge
data structure.
2. Relating to ghc/alloc perf: Inductive definitions on single-product
recursive data types like the one in T11545 will (diverge, and) have very
deep demand signatures before any other abortion mechanism in Demand
analysis is triggered. That leads to great and unnecessary churn on Demand
analysis when ultimately we will never make use of any nested strictness
information anyway.
Conclusion: Discard nested demand and boxity information on such recursive types
with the help of `Note [Detecting recursive data constructors]`.
I also implemented `GHC.Types.Unique.MemoFun.memoiseUniqueFun` in order to avoid
the overhead of repeated calls to `GHC.Core.Opt.WorkWrap.Utils.isRecDataCon`.
It's nice and simple and guards against some smaller regressions in T9233 and
T16577.
ghc/alloc performance-wise, this patch is a very clear win:
Test Metric value New value Change
---------------------------------------------------------------------------------------
LargeRecord(normal) ghc/alloc 6,141,071,720 6,099,871,216 -0.7%
MultiLayerModulesTH_OneShot(normal) ghc/alloc 2,740,973,040 2,705,146,640 -1.3%
T11545(normal) ghc/alloc 945,475,492 85,768,928 -90.9% GOOD
T13056(optasm) ghc/alloc 370,245,880 326,980,632 -11.7% GOOD
T18304(normal) ghc/alloc 90,933,944 76,998,064 -15.3% GOOD
T9872a(normal) ghc/alloc 1,800,576,840 1,792,348,760 -0.5%
T9872b(normal) ghc/alloc 2,086,492,432 2,073,991,848 -0.6%
T9872c(normal) ghc/alloc 1,750,491,240 1,737,797,832 -0.7%
TcPlugin_RewritePerf(normal) ghc/alloc 2,286,813,400 2,270,957,896 -0.7%
geo. mean -2.9%
No noteworthy change in run/alloc either.
NoFib results show slight wins, too:
--------------------------------------------------------------------------------
Program Allocs Instrs
--------------------------------------------------------------------------------
constraints -1.9% -1.4%
fasta -3.6% -2.7%
reverse-complem -0.3% -0.9%
treejoin -0.0% -0.3%
--------------------------------------------------------------------------------
Min -3.6% -2.7%
Max +0.1% +0.1%
Geometric Mean -0.1% -0.1%
Metric Decrease:
T11545
T13056
T18304
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
As #20837 pointed out, `isLiftedType_maybe` returned `Just False` in
many situations where it should return `Nothing`, because it didn't
take into account type families or type variables.
In this patch, we fix this issue. We rename `isLiftedType_maybe` to
`typeLevity_maybe`, which now returns a `Levity` instead of a boolean.
We now return `Nothing` for types with kinds of the form
`TYPE (F a1 ... an)` for a type family `F`, as well as
`TYPE (BoxedRep l)` where `l` is a type variable.
This fix caused several other problems, as other parts of the compiler
were relying on `isLiftedType_maybe` returning a `Just` value, and were
now panicking after the above fix. There were two main situations in
which panics occurred:
1. Issues involving the let/app invariant. To uphold that invariant,
we need to know whether something is lifted or not. If we get an
answer of `Nothing` from `isLiftedType_maybe`, then we don't know
what to do. As this invariant isn't particularly invariant, we
can change the affected functions to not panic, e.g. by behaving
the same in the `Just False` case and in the `Nothing` case
(meaning: no observable change in behaviour compared to before).
2. Typechecking of data (/newtype) constructor patterns. Some programs
involving patterns with unknown representations were accepted, such
as T20363. Now that we are stricter, this caused further issues,
culminating in Core Lint errors. However, the behaviour was
incorrect the whole time; the incorrectness only being revealed by
this change, not triggered by it.
This patch fixes this by overhauling where the representation
polymorphism involving pattern matching are done. Instead of doing
it in `tcMatches`, we instead ensure that the `matchExpected`
functions such as `matchExpectedFunTys`, `matchActualFunTySigma`,
`matchActualFunTysRho` allow return argument pattern types which
have a fixed RuntimeRep (as defined in Note [Fixed RuntimeRep]).
This ensures that the pattern matching code only ever handles types
with a known runtime representation. One exception was that
patterns with an unknown representation type could sneak in via
`tcConPat`, which points to a missing representation-polymorphism
check, which this patch now adds.
This means that we now reject the program in #20363, at least until
we implement PHASE 2 of FixedRuntimeRep (allowing type families in
RuntimeRep positions). The aforementioned refactoring, in which
checks have been moved to `matchExpected` functions, is a first
step in implementing PHASE 2 for patterns.
Fixes #20837
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
AbsBinds and ABExport both depended on the typechecker, and were thus
removed from the main AST Expr.
CollectPass now has a new function `collectXXHsBindsLR` used for the new
HsBinds extension point
Bumped haddock submodule to work with AST changes.
The removed Notes from Language.Haskell.Syntax.Binds were duplicated
(and not referenced) and the copies in GHC.Hs.Binds are kept (and
referenced there). (See #19252)
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Move GHC-specific comments from Language.Haskell.Syntax.Binds to
GHC.Hs.Binds
It looks like the Note was deleted but there were actually two copies of
it. L.H.S.B no longer references it, and GHC.Hs.Binds keeps an updated
copy. (See #19252)
There are other duplicated notes -- they will be fixed in the next
commit
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Issue #21150 shows that worker/wrapper allocated a worker function for a
function with multiple calls that said "called at most once" when the first
argument was absent. That's bad!
This patch makes it so that WW preserves at least one non-one-shot value lambda
(see `Note [Preserving float barriers]`) by passing around `void#` in place of
absent arguments.
Fixes #21150.
Since the fix is pretty similar to `Note [Protecting the last value argument]`,
I put the logic in `mkWorkerArgs`. There I realised (#21204) that
`-ffun-to-thunk` is basically useless with `-ffull-laziness`, so I deprecated
the flag, simplified and split into `needsVoidWorkerArg`/`addVoidWorkerArg`.
SpecConstr is another client of that API.
Fixes #21204.
Metric Decrease:
T14683
|
| |
|
|
|
|
|
|
|
|
|
| |
Note [Weak loop breakers] explains why we need to track variables free
in RHS of rules. But we need to do this for /inactive/ rules as well
as active ones, unlike the rhs_fv_env stuff.
So we now have two fields in node Details, one for free vars of
active rules, and one for free vars of all rules.
This was shown up by #20820, which is now fixed.
|
| | |
|
| | |
|
| |
|
|
| |
Added for no reason in 7ce1b694f7be7fbf6e2d7b7eb0639e61fbe358c6
|
| | |
|
| |
|
|
|
|
|
|
| |
Cf ticky report before the change:
Entries Alloc Alloc'd Non-void Arguments STG Name
--------------------------------------------------------------------------------
696987 29044128 0 1 L main:Target.trackArgument_go5{v r24kY} (fun)
|
| | |
|
