-
00478944
by Simon Peyton Jones at 2025-08-27T16:48:30+01:00
Comments only
-
a7884589
by Simon Peyton Jones at 2025-08-28T11:08:23+01:00
Type-family occurs check in unification
The occurs check in `GHC.Core.Unify.uVarOrFam` was inadequate in dealing
with type families.
Better now. See Note [The occurs check in the Core unifier].
As I did this I realised that the whole apartness thing is trickier than I
thought: see the new Note [Shortcomings of the apartness test]
-
8adfc222
by sheaf at 2025-08-28T19:47:17-04:00
Fix orientation in HsWrapper composition (<.>)
This commit fixes the order in which WpCast HsWrappers are composed,
fixing a bug introduced in commit 56b32c5a2d5d7cad89a12f4d74dc940e086069d1.
Fixes #26350
-
eb2ab1e2
by Oleg Grenrus at 2025-08-29T11:00:53-04:00
Generalise thNameToGhcName by adding HasHscEnv
There were multiple single monad-specific `getHscEnv` across codebase.
HasHscEnv is modelled on HasDynFlags.
My first idea was to simply add thNameToGhcNameHsc and
thNameToGhcNameTc, but those would been exactly the same
as thNameToGhcName already.
Also add an usage example to thNameToGhcName and mention that it's
recommended way of looking up names in GHC plugins
-
2d575a7f
by fendor at 2025-08-29T11:01:36-04:00
configure: Bump minimal bootstrap GHC version to 9.10
-
716274a5
by Simon Peyton Jones at 2025-08-29T17:27:12-04:00
Fix deep subsumption again
This commit fixed #26255:
commit 56b32c5a2d5d7cad89a12f4d74dc940e086069d1
Author: sheaf <sam.derbyshire@gmail.com>
Date: Mon Aug 11 15:50:47 2025 +0200
Improve deep subsumption
This commit improves the DeepSubsumption sub-typing implementation
in GHC.Tc.Utils.Unify.tc_sub_type_deep by being less eager to fall back
to unification.
But alas it still wasn't quite right for view patterns: #26331
This MR does a generalisation to fix it. A bit of a sledgehammer to crack
a nut, but nice.
* Add a field `ir_inst :: InferInstFlag` to `InferResult`, where
```
data InferInstFlag = IIF_Sigma | IIF_ShallowRho | IIF_DeepRho
```
* The flag says exactly how much `fillInferResult` should instantiate
before filling the hole.
* We can also use this to replace the previous very ad-hoc `tcInferSigma`
that was used to implement GHCi's `:type` command.
-
27206c5e
by sheaf at 2025-08-29T17:28:14-04:00
Back-compat for TH SpecialiseP data-con of Pragma
This commit improves the backwards-compatibility story for the
SpecialiseP constructor of the Template Haskell 'Pragma' datatype.
Instead of keeping the constructor but deprecating it, this commit makes
it into a bundled pattern synonym of the Pragma datatype. We no longer
deprecate it; it's useful for handling old-form specialise pragmas.
-
26dbcf61
by fendor at 2025-08-30T05:10:08-04:00
Move stack decoding logic from ghc-heap to ghc-internal
The stack decoding logic in `ghc-heap` is more sophisticated than the one
currently employed in `CloneStack`. We want to use the stack decoding
implementation from `ghc-heap` in `base`.
We cannot simply depend on `ghc-heap` in `base` due do bootstrapping
issues.
Thus, we move the code that is necessary to implement stack decoding to
`ghc-internal`. This is the right location, as we don't want to add a
new API to `base`.
Moving the stack decoding logic and re-exposing it in ghc-heap is
insufficient, though, as we have a dependency cycle between.
* ghc-heap depends on stage1:ghc-internal
* stage0:ghc depends on stage0:ghc-heap
To fix this, we remove ghc-heap from the set of `stage0` dependencies.
This is not entirely straight-forward, as a couple of boot dependencies,
such as `ghci` depend on `ghc-heap`.
Luckily, the boot compiler of GHC is now >=9.10, so we can migrate `ghci`
to use `ghc-internal` instead of `ghc-heap`, which already exports the
relevant modules.
However, we cannot 100% remove ghc's dependency on `ghc-heap`, since
when we compile `stage0:ghc`, `stage1:ghc-internal` is not yet
available.
Thus, when we compile with the boot-compiler, we still depend on an
older version of `ghc-heap`, and only use the modules from `ghc-internal`,
if the `ghc-internal` version is recent enough.
-------------------------
Metric Increase:
T24602_perf_size
T25046_perf_size_gzip
T25046_perf_size_unicode
T25046_perf_size_unicode_gzip
size_hello_artifact
size_hello_artifact_gzip
size_hello_unicode
size_hello_unicode_gzip
-------------------------
These metric increases are unfortunate, they are most likely caused by
the larger (literally in terms of lines of code) stack decoder implementation
that are now linked into hello-word binaries.
On linux, it is almost a 10% increase, which is considerable.
-
bd80bb70
by fendor at 2025-08-30T05:10:08-04:00
Implement `decode` in terms of `decodeStackWithIpe`
Uses the more efficient stack decoder implementation.
-
24441165
by fendor at 2025-08-30T05:10:08-04:00
Remove stg_decodeStackzh
-
fb9cc882
by Simon Peyton Jones at 2025-08-30T05:10:51-04:00
Fix a long standing bug in the coercion optimiser
We were mis-optimising ForAllCo, leading to #26345
Part of the poblem was the tricky tower of abstractions leading to
the dreadful
GHC.Core.TyCo.Subst.substForAllCoTyVarBndrUsing
This function was serving two masters: regular substitution, but also
coercion optimsation. So tricky was it that it did so wrong.
In this MR I locate all the fancy footwork for coercion optimisation
in GHC.Core.Coercion.Opt, where it belongs. That leaves substitution
free to be much simpler.
-
6c78de2d
by Sylvain Henry at 2025-09-01T08:46:19-04:00
Driver: substitute virtual Prim module in --make mode too
When we build ghc-internal with --make (e.g. with cabal-install), we
need to be careful to substitute the virtual interface file for
GHC.Internal.Prim:
- after code generation (we generate code for an empty module, so we get
an empty interface)
- when we try to reload its .hi file
-
26e0db16
by fendor at 2025-09-01T08:47:01-04:00
Expose Stack Annotation frames in IPE backtraces by default
When decoding the Haskell-native call stack and displaying the IPE information
for the stack frames, we print the `StackAnnotation` of the `AnnFrame` by default.
This means, when an exception is thrown, any intermediate stack annotations will
be displayed in the `IPE Backtrace`.
Example backtrace:
```
Exception: ghc-internal:GHC.Internal.Exception.ErrorCall:
Oh no!
IPE backtrace:
annotateCallStackIO, called at app/Main.hs:48:10 in backtrace-0.1.0.0-inplace-server:Main
annotateCallStackIO, called at app/Main.hs:46:13 in backtrace-0.1.0.0-inplace-server:Main
Main.handler (app/Main.hs:(46,1)-(49,30))
Main.liftIO (src/Servant/Server/Internal/Handler.hs:30:36-42)
Servant.Server.Internal.Delayed.runHandler' (src/Servant/Server/Internal/Handler.hs:27:31-41)
Control.Monad.Trans.Resource.runResourceT (./Control/Monad/Trans/Resource.hs:(192,14)-(197,18))
Network.Wai.Handler.Warp.HTTP1.processRequest (./Network/Wai/Handler/Warp/HTTP1.hs:195:20-22)
Network.Wai.Handler.Warp.HTTP1.processRequest (./Network/Wai/Handler/Warp/HTTP1.hs:(195,5)-(203,31))
Network.Wai.Handler.Warp.HTTP1.http1server.loop (./Network/Wai/Handler/Warp/HTTP1.hs:(141,9)-(157,42))
HasCallStack backtrace:
error, called at app/Main.hs:48:32 in backtrace-0.1.0.0-inplace-server:Main
```
The first two entries have been added by `annotateCallStackIO`, defined in `annotateCallStackIO`.
-
a1567efd
by Sylvain Henry at 2025-09-01T23:01:35-04:00
RTS: rely less on Hadrian for flag setting (#25843)
Hadrian used to pass -Dfoo command-line flags directly to build the rts.
We can replace most of these flags with CPP based on cabal flags.
It makes building boot libraries with cabal-install simpler (cf #25843).
-
ca5b0283
by Sergey Vinokurov at 2025-09-01T23:02:23-04:00
Remove unnecessary irrefutable patterns from Bifunctor instances for tuples
Implementation of https://github.com/haskell/core-libraries-committee/issues/339
Metric Decrease:
mhu-perf
-
2da84b7a
by sheaf at 2025-09-01T23:03:23-04:00
Only use active rules when simplifying rule RHSs
When we are simplifying the RHS of a rule, we make sure to only apply
rewrites from rules that are active throughout the original rule's
range of active phases.
For example, if a rule is always active, we only fire rules that are
themselves always active when simplifying the RHS. Ditto for inline
activations.
This is achieved by setting the simplifier phase to a range of phases,
using the new SimplPhaseRange constructor. Then:
1. When simplifying the RHS of a rule, or of a stable unfolding,
we set the simplifier phase to a range of phases, computed from
the activation of the RULE/unfolding activation, using the
function 'phaseFromActivation'.
The details are explained in Note [What is active in the RHS of a RULE?]
in GHC.Core.Opt.Simplify.Utils.
2. The activation check for other rules and inlinings is then:
does the activation of the other rule/inlining cover the whole
phase range set in sm_phase? This continues to use the 'isActive'
function, which now accounts for phase ranges.
On the way, this commit also moves the exact-print SourceText annotation
from the Activation datatype to the ActivationAnn type. This keeps the
main Activation datatype free of any extra cruft.
Fixes #26323
-
79816cc4
by Rodrigo Mesquita at 2025-09-02T12:19:59-04:00
cleanup: Move dehydrateCgBreakInfo to Stg2Bc
This no longer has anything to do with Core.
-
53da94ff
by Rodrigo Mesquita at 2025-09-02T12:19:59-04:00
rts/Disassembler: Fix spacing of BRK_FUN
-
08c0cf85
by Rodrigo Mesquita at 2025-09-02T12:19:59-04:00
debugger: Fix bciPtr in Step-out
We need to use `BCO_NEXT` to move bciPtr to ix=1, because ix=0 points to
the instruction itself!
I do not understand how this didn't crash before.
-
e7e021fa
by Rodrigo Mesquita at 2025-09-02T12:19:59-04:00
debugger: Allow BRK_FUNs to head case continuation BCOs
When we start executing a BCO, we may want to yield to the scheduler:
this may be triggered by a heap/stack check, context switch, or a
breakpoint. To yield, we need to put the stack in a state such that
when execution is resumed we are back to where we yielded from.
Previously, a BKR_FUN could only head a function BCO because we only
knew how to construct a valid stack for yielding from one -- simply add
`apply_interp_info` + the BCO to resume executing. This is valid because
the stack at the start of run_BCO is headed by that BCO's arguments.
However, in case continuation BCOs (as per Note [Case continuation BCOs]),
we couldn't easily reconstruct a valid stack that could be resumed
because we dropped too soon the stack frames regarding the value
returned (stg_ret) and received (stg_ctoi) by that continuation.
This is especially tricky because of the variable type and size return
frames (e.g. pointer ret_p/ctoi_R1p vs a tuple ret_t/ctoi_t2).
The trick to being able to yield from a BRK_FUN at the start of a case
cont BCO is to stop removing the ret frame headers eagerly and instead
keep them until the BCO starts executing. The new layout at the start of
a case cont. BCO is described by the new Note [Stack layout when entering run_BCO].
Now, we keep the ret_* and ctoi_* frames when entering run_BCO.
A BRK_FUN is then executed if found, and the stack is yielded as-is with
the preserved ret and ctoi frames.
Then, a case cont BCO's instructions always SLIDE off the headers of the
ret and ctoi frames, in StgToByteCode.doCase, turning a stack like
| .... |
+---------------+
| fv2 |
+---------------+
| fv1 |
+---------------+
| BCO |
+---------------+
| stg_ctoi_ret_ |
+---------------+
| retval |
+---------------+
| stg_ret_..... |
+---------------+
into
| .... |
+---------------+
| fv2 |
+---------------+
| fv1 |
+---------------+
| retval |
+---------------+
for the remainder of the BCO.
Moreover, this more uniform approach of keeping the ret and ctoi frames
means we need less ad-hoc logic concerning the variable size of
ret_tuple vs ret_p/np frames in the code generator and interpreter:
Always keep the return to cont. stack intact at the start of run_BCO,
and the statically generated instructions will take care of adjusting
it.
Unlocks BRK_FUNs at the start of case cont. BCOs which will enable a
better user-facing step-out (#26042) which is free of the bugs the
current BRK_ALTS implementation suffers from (namely, using BRK_FUN
rather than BRK_ALTS in a case cont. means we'll never accidentally end
up in a breakpoint "deeper" than the continuation, because we stop at
the case cont itself rather than on the first breakpoint we evaluate
after it).
-
ade3c1e6
by Rodrigo Mesquita at 2025-09-02T12:19:59-04:00
BRK_FUN with InternalBreakLocs for code-generation time breakpoints
At the start of a case continuation BCO, place a BRK_FUN.
This BRK_FUN uses the new "internal breakpoint location" -- allowing us
to come up with a valid source location for this breakpoint that is not associated with a source-level tick.
For case continuation BCOs, we use the last tick seen before it as the
source location. The reasoning is described in Note [Debugger: Stepout internal break locs].
Note how T26042c, which was broken because it displayed the incorrect
behavior of the previous step out when we'd end up at a deeper level
than the one from which we initiated step-out, is now fixed.
As of this commit, BRK_ALTS is now dead code and is thus dropped.
Note [Debugger: Stepout internal break locs]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Step-out tells the interpreter to run until the current function
returns to where it was called from, and stop there.
This is achieved by enabling the BRK_FUN found on the first RET_BCO
frame on the stack (See [Note Debugger: Step-out]).
Case continuation BCOs (which select an alternative branch) must
therefore be headed by a BRK_FUN. An example:
f x = case g x of <--- end up here
1 -> ...
2 -> ...
g y = ... <--- step out from here
- `g` will return a value to the case continuation BCO in `f`
- The case continuation BCO will receive the value returned from g
- Match on it and push the alternative continuation for that branch
- And then enter that alternative.
If we step-out of `g`, the first RET_BCO on the stack is the case
continuation of `f` -- execution should stop at its start, before
selecting an alternative. (One might ask, "why not enable the breakpoint
in the alternative instead?", because the alternative continuation is
only pushed to the stack *after* it is selected by the case cont. BCO)
However, the case cont. BCO is not associated with any source-level
tick, it is merely the glue code which selects alternatives which do
have source level ticks. Therefore, we have to come up at code
generation time with a breakpoint location ('InternalBreakLoc') to
display to the user when it is stopped there.
Our solution is to use the last tick seen just before reaching the case
continuation. This is robust because a case continuation will thus
always have a relevant breakpoint location:
- The source location will be the last source-relevant expression
executed before the continuation is pushed
- So the source location will point to the thing you've just stepped
out of
- Doing :step-local from there will put you on the selected
alternative (which at the source level may also be the e.g. next
line in a do-block)
Examples, using angle brackets (<<...>>) to denote the breakpoint span:
f x = case <<g x>> {- step in here -} of
1 -> ...
2 -> ...>
g y = <<...>> <--- step out from here
...
f x = <<case g x of <--- end up here, whole case highlighted
1 -> ...
2 -> ...>>
doing :step-local ...
f x = case g x of
1 -> <<...>> <--- stop in the alternative
2 -> ...
A second example based on T26042d2, where the source is a do-block IO
action, optimised to a chain of `case expressions`.
main = do
putStrLn "hello1"
<<f>> <--- step-in here
putStrLn "hello3"
putStrLn "hello4"
f = do
<<putStrLn "hello2.1">> <--- step-out from here
putStrLn "hello2.2"
...
main = do
putStrLn "hello1"
<<f>> <--- end up here again, the previously executed expression
putStrLn "hello3"
putStrLn "hello4"
doing step/step-local ...
main = do
putStrLn "hello1"
f
<<putStrLn "hello3">> <--- straight to the next line
putStrLn "hello4"
Finishes #26042
-
c66910c0
by Rodrigo Mesquita at 2025-09-02T12:19:59-04:00
debugger: Re-use the last BreakpointId whole in step-out
Previously, to come up with a location to stop at for `:stepout`, we
would store the location of the last BreakpointId surrounding the
continuation, as described by Note [Debugger: Stepout internal break locs].
However, re-using just the location from the last source breakpoint
isn't sufficient to provide the necessary information in the break
location. Specifically, it wouldn't bind any variables at that location.
Really, there is no reason not to re-use the last breakpoint wholesale,
and re-use all the information we had there. Step-out should behave just
as if we had stopped at the call, but s.t. continuing will not
re-execute the call.
This commit updates the CgBreakInfo to always store a BreakpointId, be
it the original one or the one we're emulating (for step-out).
It makes variable bindings on :stepout work
-
e4abed7b
by sheaf at 2025-09-02T12:20:40-04:00
Revert accidental changes to hie.yaml
-
003b715b
by meooow25 at 2025-09-02T23:48:51+02:00
Adjust the strictness of Data.List.iterate'
* Don't force the next element in advance when generating a (:).
* Force the first element to WHNF like every other element.
Now every element in the output list is forced to WHNF when the (:)
containing it is forced.
CLC proposal:
https://github.com/haskell/core-libraries-committee/issues/335
-
b2f6aad0
by Simon Hengel at 2025-09-03T04:36:10-04:00
Refactoring: More consistently use logOutput, logInfo, fatalErrorMsg
-
60a16db7
by Rodrigo Mesquita at 2025-09-03T10:55:50+01:00
bytecode: Don't PUSH_L 0; SLIDE 1 1
While looking through bytecode I noticed a quite common unfortunate
pattern:
...
PUSH_L 0
SLIDE 1 1
We do this often by generically constructing a tail call from a function
atom that may be somewhere arbitrary on the stack.
However, for the special case that the function can be found directly on
top of the stack, as part of the arguments, it's plain redundant to push
then slide it.
In this commit we add a small optimisation to the generation of
tailcalls in bytecode. Simply: lookahead for the function in the stack.
If it is the first thing on the stack and it is part of the arguments
which would be dropped as we entered the tail call, then don't push then
slide it.
In a simple example (T26042b), this already produced a drastic
improvement in generated code (left is old, right is with this patch):
```diff
3c3
< 2025-07-29 10:14:02.081277 UTC
---
> 2025-07-29 10:50:36.560949 UTC
160,161c160
< PUSH_L 0
< SLIDE 1 2
---
> SLIDE 1 1
164,165d162
< PUSH_L 0
< SLIDE 1 1
175,176c172
< PUSH_L 0
< SLIDE 1 2
---
> SLIDE 1 1
179,180d174
< PUSH_L 0
< SLIDE 1 1
206,207d199
< PUSH_L 0
< SLIDE 1 1
210,211d201
< PUSH_L 0
< SLIDE 1 1
214,215d203
< PUSH_L 0
< SLIDE 1 1
218,219d205
< PUSH_L 0
< SLIDE 1 1
222,223d207
< PUSH_L 0
< SLIDE 1 1
...
600,601c566
< PUSH_L 0
< SLIDE 1 2
---
> SLIDE 1 1
604,605d568
< PUSH_L 0
< SLIDE 1 1
632,633d594
< PUSH_L 0
< SLIDE 1 1
636,637d596
< PUSH_L 0
< SLIDE 1 1
640,641d598
< PUSH_L 0
< SLIDE 1 1
644,645d600
< PUSH_L 0
< SLIDE 1 1
648,649d602
< PUSH_L 0
< SLIDE 1 1
652,653d604
< PUSH_L 0
< SLIDE 1 1
656,657d606
< PUSH_L 0
< SLIDE 1 1
660,661d608
< PUSH_L 0
< SLIDE 1 1
664,665d610
< PUSH_L 0
< SLIDE 1 1
```
I also compiled lib:Cabal to bytecode and counted the number of bytecode
lines with `find dist-newstyle -name "*.dump-BCOs" -exec wc {} +`:
with unoptimized core:
1190689 lines (before) - 1172891 lines (now)
= 17798 less redundant instructions (-1.5% lines)
with optimized core:
1924818 lines (before) - 1864836 lines (now)
= 59982 less redundant instructions (-3.1% lines)
-
8b2c72c0
by L0neGamer at 2025-09-04T06:32:03-04:00
Add Control.Monad.thenM and Control.Applicative.thenA
-
39e1b7cb
by Teo Camarasu at 2025-09-04T06:32:46-04:00
ghc-internal: invert dependency of GHC.Internal.TH.Syntax on Data.Data
This means that Data.Data no longer blocks building TH.Syntax, which
allows greater parallelism in our builds.
We move the Data.Data.Data instances to Data.Data. Quasi depends on
Data.Data for one of its methods, so,
we split the Quasi/Q, etc definition out of GHC.Internal.TH.Syntax
into its own module. This has the added benefit of splitting up this
quite large module.
Previously TH.Syntax was a bottleneck when compiling ghc-internal. Now
it is less of a bottle-neck and is also slightly quicker to
compile (since it no longer contains these instances) at the cost of
making Data.Data slightly more expensive to compile.
TH.Lift which depends on TH.Syntax can also compile quicker and no
longer blocks ghc-internal finishing to compile.
Resolves #26217
-------------------------
Metric Decrease:
MultiLayerModulesTH_OneShot
T13253
T21839c
T24471
Metric Increase:
T12227
-------------------------
-
bdf82fd2
by Teo Camarasu at 2025-09-04T06:32:46-04:00
compiler: delete unused names in Builtins.Names.TH
returnQ and bindQ are no longer used in the compiler.
There was also a very old comment that referred to them that I have modernized
-
41a448e5
by Ben Gamari at 2025-09-04T19:21:43-04:00
hadrian: Pass lib & include directories to ghc `Setup configure`
-
46bb9a79
by Ben Gamari at 2025-09-04T19:21:44-04:00
rts/IPE: Fix compilation when zstd is enabled
This was broken by the refactoring undertaken in
c80dd91c0bf6ac034f0c592f16c548b9408a8481.
Closes #26312.
-
138a6e34
by sheaf at 2025-09-04T19:22:46-04:00
Make mkCast assertion a bit clearer
This commit changes the assertion message that gets printed when one
calls mkCast with a coercion whose kind does not match the type of the
inner expression. I always found the assertion message a bit confusing,
as it didn't clearly state what exactly was the error.
-
9d626be1
by sheaf at 2025-09-04T19:22:46-04:00
Simplifier/rules: fix mistakes in Notes & comments
-
94b62aa7
by Simon Peyton Jones at 2025-09-08T03:37:14-04:00
Refactor ForAllCo
This is a pure refactor, addressing #26389.
It arranges that the kind coercion in a ForAllCo is a MCoercion, rather
than a plain Coercion, thus removing redundancy in the common case.
See (FC8) in Note [ForAllCo]
It's a nice cleanup.
-
624afa4a
by sheaf at 2025-09-08T03:38:05-04:00
Use tcMkScaledFunTys in matchExpectedFunTys
We should use tcMkScaledFunTys rather than mkScaledFunTys in
GHC.Tc.Utils.Unify.matchExpectedFunTys, as the latter crashes
when the kind of the result type is a bare metavariable.
We know the result is always Type-like, so we don't need scaledFunTys
to try to rediscover that from the kind.
Fixes #26277
-
0975d2b6
by sheaf at 2025-09-08T03:38:54-04:00
Revert "Remove hptAllFamInstances usage during upsweep"
This reverts commit 3bf6720eff5e86e673568e756161e6d6150eb440.
-
0cf34176
by soulomoon at 2025-09-08T03:38:54-04:00
Family consistency checks: add test for #26154
This commit adds the test T26154, to make sure that GHC doesn't crash
when performing type family consistency checks. This test case
was extracted from Agda.
Fixes #26154
-
ba210d98
by Simon Peyton Jones at 2025-09-08T16:26:36+01:00
Report solid equality errors before custom errors
This MR fixes #26255 by
* Reporting solid equality errors like
Int ~ Bool
before "custom type errors". See comments in `report1` in
`reportWanteds`
* Suppressing errors that arise from superclasses of
Wanteds. See (SCE1) in Note [Suppressing confusing errors]
More details in #26255.
-
b6249140
by Simon Peyton Jones at 2025-09-10T10:42:38-04:00
Fix a scoping error in Specialise
This small patch fixes #26329, which triggered a scoping error.
Test is in T21391, with -fpolymorphic-specialisation enabled
-
45305ab8
by sheaf at 2025-09-10T10:43:29-04:00
Make rationalTo{Float,Double} inline in phase 0
We hold off on inlining these until phase 0 to allow constant-folding
rules to fire. However, once we get to phase 0, we should inline them,
e.g. to expose unboxing opportunities.
See CLC proposal #356.
-
0959d4bc
by Andreas Klebinger at 2025-09-10T10:44:12-04:00
Add regression test for #26056
-
dc79593d
by sheaf at 2025-09-10T10:45:01-04:00
Deep subsumption: unify mults without tcEqMult
As seen in #26332, we may well end up with a non-reflexive multiplicity
coercion when doing deep subsumption. We should do the same thing that
we do without deep subsumption: unify the multiplicities normally,
without requiring that the coercion is reflexive (which is what
'tcEqMult' was doing).
Fixes #26332
-
4bfe2269
by sheaf at 2025-09-10T10:45:50-04:00
lint-codes: fixup MSYS drive letter on Windows
This change ensures that System.Directory.listDirectory doesn't trip up
on an MSYS-style path like '/c/Foo' when trying to list all testsuite
stdout/stderr files as required for testing coverage of GHC diagnostic
codes in the testsuite.
Fixes #25178
-
56540775
by Ben Gamari at 2025-09-10T10:46:32-04:00
gitlab-ci: Disable split sections on FreeBSD
Due to #26303.
-
1537784b
by Moritz Angermann at 2025-09-10T10:47:13-04:00
Improve mach-o relocation information
This change adds more information about the symbol and addresses
we try to relocate in the linker. This significantly helps when
deubbging relocation issues reported by users.
-
4e67855b
by Moritz Angermann at 2025-09-10T10:47:54-04:00
test.mk expect GhcLeadingUnderscore, not LeadingUnderscore (in line with the other Ghc prefixed variables.
-
c1cdd265
by Moritz Angermann at 2025-09-10T10:48:35-04:00
testsuite: Fix broken exec_signals_child.c
There is no signal 0. The signal mask is 1-32.
-
99ac335c
by Moritz Angermann at 2025-09-10T10:49:15-04:00
testsuite: clarify Windows/Darwin locale rationale for skipping T6037 T2507 T8959a
-
0e8fa77a
by Moritz Angermann at 2025-09-10T10:49:56-04:00
Skip broken tests on macOS (due to leading underscore not handled properly in the expected output.)
-
28570c59
by Zubin Duggal at 2025-09-10T10:50:37-04:00
docs(sphinx): fix links to reverse flags when using the :ghc-flag:`-fno-<flag>` syntax
This solution is rather hacky and I suspect there is a better way to do this but I don't know
enough about Sphinx to do better.
Fixes #26352
-
d17257ed
by Cheng Shao at 2025-09-10T17:01:27+02:00
rel-eng: update alpine images to 3.22
This patch is a part of #25876 and updates alpine images to 3.22,
while still retaining 3.12 for x86_64 fully_static bindists.
-------------------------
Metric Decrease:
MultiComponentModulesRecomp
-------------------------
-
db3276bb
by Sylvain Henry at 2025-09-11T11:27:28-04:00
T16180: indicate that the stack isn't executable
-
11eeeba7
by Sylvain Henry at 2025-09-11T11:27:28-04:00
Fix some tests (statically linked GHC vs libc)
When GHC is linked statically, the stdout C global variable that GHC uses
isn't shared with the stdout C global variable used by loaded code.
As a consequence, the latter must be explicitly flushed because GHC
won't flush it before exiting.
-
80a07571
by Sylvain Henry at 2025-09-11T11:28:18-04:00
Testsuite: fix debug_rts detection
Running the testsuite without Hadrian should set config.debug_rts
correctly too.
-
62ae97de
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Handle heap allocation failure in I/O primops
The current I/O managers do not use allocateMightFail, but future ones
will. To support this properly we need to be able to return to the
primop with a failure. We simply use a bool return value.
Currently however, we will just throw an exception rather than calling
the GC because that's what all the other primops do too.
For the general issue of primops invoking GC and retrying, see
https://gitlab.haskell.org/ghc/ghc/-/issues/24105
-
cb9093f5
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Move (and rename) scheduleStartSignalHandlers into RtsSignals.h
Previously it was a local helper (static) function in Schedule.c.
Rename it to startPendingSignalHandlers and deifine it as an inline
header function in RtsSignals.h. So it should still be fast.
Each (new style) I/O manager is going to need to do the same, so eliminating
the duplication now makes sense.
-
9736d44a
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Reduce detail in printThreadBlockage I/O blocking cases
The printThreadBlockage is used in debug tracing output.
For the cases BlockedOn{Read,Write,Delay} the output previously included
the fd that was being waited on, and the delay target wake time.
Superficially this sounds useful, but it's clearly not that useful
because it was already wrong for the Win32 non-threaded I/O manager. In
that situation it will print garbage (the async_result pointer, cast to
a fd or a time).
So given that it apparently never mattered that the information was
accurate, then it's hardly a big jump to say it doesn't matter if it is
present at all.
A good reason to remove it is that otherwise we have to make a new
API and a per-I/O manager implementation to fetch the information. And
for some I/O manager implementations, this information is not available.
It is not available in the win32 non-threaded I/O manager. And for some
future Linux ones, there is no need for the fd to be stored, so storing
it would be just extra space used for very little gain.
So the simplest thing is to just remove the detail.
-
bc0f2d5d
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Add TimeoutQueue.{c,h} and corresponding tests
A data structure used to efficiently manage a collection of timeouts.
It is a priority queue based on absolute expiry time. It uses 64bit
high-precision Time for the keys. The values are normal closures which
allows for example using MVars for unblocking.
It is common in many applications for timeouts to be created and then
deleted or altered before they expire. Thus the choice of data structure
for timeouts should support this efficiently. The implementation choice
here is a leftist heap with the extra feature that it supports deleting
arbitrary elements, provided the caller retain a pointer to the element.
While the deleteMin operation takes O(log n) time, as in all heap
structures, the delete operation for arbitrary elements /typically/
takes O(1), and only O(log n) in the worst case. In practice, when
managing thousands of timeouts it can be a factor of 10 faster to delete
a random timeout queue element than to remove the minimum element. This
supports the common use case.
The plan is to use it in some of the RTS-side I/O managers to support
their timer functionality. In this use case the heap value will be an
MVar used for each timeout to unblock waiting threads.
-
d1679c9d
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Add ClosureTable.{c,h} and corresponding tests
A table of pointers to closures on the GC heap with stable indexes.
It provides O(1) alloc, free and lookup. The table can be expanded
using a simple doubling strategy: in which case allocation is typically
O(1) and occasionally O(n) for overall amortised O(1). No shrinking is
used.
The table itself is heap allocated, and points to other heap objects.
As such it's necessary to use markClosureTable to ensure the table is
used as a GC root to keep the table entries alive, and maintain proper
pointers to them as the GC moves heap objects about.
It is designed to be allocated and accesses exclusively from a single
capability, enabling it to work without any locking. It is thus similar
to the StablePtr table, but per-capability which removes the need for
locking. It _should_ also provide lower GC pause times with the
non-moving GC by spending only O(1) time in markClosureTable, vs O(n)
for markStablePtrTable.
The plan is to use it in some of the I/O managers to keep track of
in-flight I/O operations (but not timers). This allows the tracking
info to be kept on the (unpinned) GC heap, and shared with Haskell
code, and by putting a pointer to the tracking information in a table,
the index remains stable and can be passed via foreign code (like the
kernel).
-
78cb8dd5
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Add the StgAsyncIOOp closure type
This is intended to be used by multiple I/O managers to help with
tracking in-flight I/O operations.
It is called asynchronous because from the point of view of the RTS we
have many such operations in progress at once. From the point of view of
a Haskell thread of course it can look synchronous.
-
a2839896
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Add StgAsyncIOOp and StgTimeoutQueue to tso->block_info
These will be used by new I/O managers, for threads blocked on I/O or
timeouts.
-
fdc2451c
by Duncan Coutts at 2025-09-12T13:23:33-04:00
Add a new I/O manager based on poll()
This is a proof of concept I/O manager, to show how to add new ones
neatly, using the ClosureTable and TimeoutQueue infrastructure.
It uses the old unix poll() API, so it is of course limited in
performance by that, but it should have the benefit of wide
compatibility. Also we neatly avoid a name clash with the existing
select() I/O manager.
Compared to the select() I/O manager:
1. beause it uses poll() it is not limited to 1024 file descriptors
(but it's still O(n) so don't expect great performance);
2. it should have much faster threadDelay (when using it in lots of
threads at once) because it's based on the new TimeoutQueue which is
O(log n) rather than O(n).
Some of the code related to timers/timouts is put into a shared module
rts/posix/Timeout.{h,c} since it is intended to be shared with other
similar I/O managers.
-
6c273b76
by Duncan Coutts at 2025-09-12T13:23:34-04:00
Document the I/O managers in the user guide
and note the new poll I/O manager in the release notes.
-
824fab74
by Duncan Coutts at 2025-09-12T13:23:34-04:00
Use the poll() I/O manager by default
That is, for the non-threaded RTS, prefer the poll I/O manager over the
legacy select() one, if both can be enabled.
This patch is primarily for CI testing, so we should probably remove
this patch before merging. We can change defaults later after wider
testing and feedback.
-
39392532
by Luite Stegeman at 2025-09-12T13:24:16-04:00
Support larger unboxed sums
Change known constructor encoding for sums in interfaces to use
11 bits for both the arity and the alternative (up from 8 and 6,
respectively)
-
2af12e21
by Luite Stegeman at 2025-09-12T13:24:16-04:00
Decompose padding smallest-first in Cmm toplevel data constructors
This makes each individual padding value aligned
-
418fa78f
by Luite Stegeman at 2025-09-12T13:24:16-04:00
Use slots smaller than word as tag for smaller unboxed sums
This packs unboxed sums more efficiently by allowing
Word8, Word16 and Word32 for the tag field if the number of
constructors is small enough
-
8d7e912f
by Rodrigo Mesquita at 2025-09-12T17:57:24-04:00
ghc-toolchain: Use ByteOrder rather than new Endianness
Don't introduce a duplicate datatype when the previous one is equivalent
and already used elsewhere. This avoids unnecessary translation between
the two.
-
7d378476
by Rodrigo Mesquita at 2025-09-12T17:57:24-04:00
Read Toolchain.Target files rather than 'settings'
This commit makes GHC read `lib/targets/default.target`, a file with a
serialized value of `ghc-toolchain`'s `GHC.Toolchain.Target`.
Moreover, it removes all the now-redundant entries from `lib/settings`
that are configured as part of a `Target` but were being written into
`settings`.
This makes it easier to support multiple targets from the same compiler
(aka runtime retargetability). `ghc-toolchain` can be re-run many times
standalone to produce a `Target` description for different targets, and,
in the future, GHC will be able to pick at runtime amongst different
`Target` files.
This commit only makes it read the default `Target` configured in-tree
or configured when installing the bindist.
The remaining bits of `settings` need to be moved to `Target` in follow
up commits, but ultimately they all should be moved since they are
per-target relevant.
Fixes #24212
On Windows, the constant overhead of parsing a slightly more complex
data structure causes some small-allocation tests to wiggle around 1 to
2 extra MB (1-2% in these cases).
-------------------------
Metric Increase:
MultiLayerModulesTH_OneShot
T10421
T10547
T12234
T12425
T13035
T18140
T18923
T9198
TcPlugin_RewritePerf
-------------------------
-
e0780a16
by Rodrigo Mesquita at 2025-09-12T17:57:24-04:00
ghc-toolchain: Move TgtHasLibm to per-Target file
TargetHasLibm is now part of the per-target configuration
Towards #26227
-
8235dd8c
by Rodrigo Mesquita at 2025-09-12T17:57:24-04:00
ghc-toolchain: Move UseLibdw to per-Target file
To support DWARF unwinding, the RTS must be built with the -f+libdw flag
and with the -DUSE_LIBDW macro definition. These flags are passed on
build by Hadrian when --enable-dwarf-unwinding is specified at configure
time.
Whether the RTS was built with support for DWARF is a per-target
property, and as such, it was moved to the per-target
GHC.Toolchain.Target.Target file.
Additionally, we keep in the target file the include and library paths
for finding libdw, since libdw should be checked at configure time (be
it by configure, or ghc-toolchain, that libdw is properly available).
Preserving the user-given include paths for libdw facilitates in the
future building the RTS on demand for a given target (if we didn't keep
that user input, we couldn't)
Towards #26227
-
d5ecf2e8
by Rodrigo Mesquita at 2025-09-12T17:57:25-04:00
ghc-toolchain: Make "Support SMP" a query on a Toolchain.Target
"Support SMP" is merely a function of target, so we can represent it as
such in `ghc-toolchain`.
Hadrian queries the Target using this predicate to determine how to
build GHC, and GHC queries the Target similarly to report under --info
whether it "Support SMP"
Towards #26227
-
e07b031a
by Rodrigo Mesquita at 2025-09-12T17:57:25-04:00
ghc-toolchain: Make "tgt rts linker only supports shared libs" function on Target
Just like with "Support SMP", "target RTS linker only supports shared
libraries" is a predicate on a `Target` so we can just compute it when
necessary from the given `Target`.
Towards #26227
-
14123ee6
by Simon Peyton Jones at 2025-09-12T17:58:07-04:00
Solve forall-constraints via an implication, again
In this earlier commit:
commit 953fd8f1dc080f1c56e3a60b4b7157456949be29
Author: Simon Peyton Jones <simon.peytonjones@gmail.com>
Date: Mon Jul 21 10:06:43 2025 +0100
Solve forall-constraints immediately, or not at all
I used a all-or-nothing strategy for quantified constraints
(aka forall-constraints). But alas that fell foul of #26315,
and #26376.
So this MR goes back to solving a quantified constraint by
turning it into an implication; UNLESS we are simplifying
constraints from a SPECIALISE pragma, in which case the
all-or-nothing strategy is great. See:
Note [Solving a Wanted forall-constraint]
Other stuff in this MR:
* TcSMode becomes a record of flags, rather than an enumeration
type; much nicer.
* Some fancy footwork to avoid error messages worsening again
(The above MR made them better; we want to retain that.)
See `GHC.Tc.Errors.Ppr.pprQCOriginExtra`.
-------------------------
Metric Decrease:
T24471
-------------------------
-
e6c192e2
by Simon Peyton Jones at 2025-09-12T17:58:07-04:00
Add a test case for #26396
...same bug ast #26315
-
8f3d80ff
by Luite Stegeman at 2025-09-13T08:43:09+02:00
Use mkVirtHeapOffsets for reconstructing terms in RTTI
This makes mkVirtHeapOffsets the single source of truth for
finding field offsets in closures.
-
eb389338
by Luite Stegeman at 2025-09-13T08:43:09+02:00
Sort non-pointer fields by size for more efficient packing
This sorts non-pointer fields in mkVirtHeapOffsets, always
storing the largest field first. The relative order of
equally sized fields remains unchanged.
This reduces wasted padding/alignment space in closures with
differently sized fields.
-
99b233f4
by Alison at 2025-09-13T16:51:04-04:00
ghc-heap: Fix race condition with profiling builds
Apply the same fix from Closures.hs (64fd0fac83) to Heap.hs by adding
empty imports to make way-dependent dependencies visible to `ghc -M`.
Fixes #15197, #26407
-
77deaa7a
by Cheng Shao at 2025-09-14T21:29:45-04:00
hadrian: build in-tree gmp with -fvisibility=hidden
When hadrian builds in-tree gmp, it should build the shared objects
with -fvisibility=hidden. The gmp symbols are only used by bignum
logic in ghc-internal and shouldn't be exported by the ghc-internal
shared library. We should always strive to keep shared library symbol
table lean, which benefits platforms with slow dynamic linker or even
hard limits about how many symbols can be exported (e.g. macos dyld,
win32 dll and wasm dyld).
-
42a18960
by Cheng Shao at 2025-09-14T21:30:26-04:00
Revert "wasm: add brotli compression for ghci browser mode"
This reverts commit 731217ce68a1093b5f9e26a07d5bd2cdade2b352.
Benchmarks show non-negligible overhead when browser runs on the same
host, which is the majority of actual use cases.
-
e6755b9f
by Cheng Shao at 2025-09-14T21:30:26-04:00
wasm: remove etag logic in ghci browser mode web server
This commit removes the etag logic in dyld script's ghci browser mode
web server. It was meant to support caching logic of wasm shared
libraries, but even if the port is manually specified to make caching
even relevant, for localhost the extra overhead around etag logic is
simply not worth it according to benchmarks.
-
ac5859b9
by sheaf at 2025-09-16T14:58:38-04:00
Add 'Outputable Natural' instance
This commit adds an Outputable instance for the Natural natural-number type,
as well as a "natural :: Natural -> SDoc" function that mirrors the existing
"integer" function.
-
d48ebc23
by Cheng Shao at 2025-09-16T14:59:18-04:00
autoconf: emit warning instead of error for FIND_PYTHON logic
This patch makes FIND_PYTHON logic emit warning instead of error, so
when the user doesn't expect to run the testsuite driver (especially
when installing a bindist), python would not be mandatory. Fixes #26347.
-
54b5950e
by Sylvain Henry at 2025-09-17T04:45:18-04:00
Print fully qualified unit names in name mismatch
It's more user-friendly to directly print the right thing instead of
requiring the user to retry with the additional `-dppr-debug` flag.
-
403cb665
by Ben Gamari at 2025-09-17T04:46:00-04:00
configure: Fix consistency between distrib and source CC check
Previously distrib/configure.ac did not
include `cc`.
Closes #26394.
-
2dcd4cb9
by Oleg Grenrus at 2025-09-17T04:46:41-04:00
Use isPrint in showUnique
The comment say
```
-- Avoid emitting non-printable characters in pretty uniques. See #25989.
```
so let the code do exactly that.
There are tags (at least : and 0 .. 9) which weren't in A .. z range.
-
e5dd754b
by Oleg Grenrus at 2025-09-17T04:46:42-04:00
Shorten in-module links in hyperlinked source
Instead of href="This.Module#ident" to just "#ident"
-
63189b2c
by Oleg Grenrus at 2025-09-17T04:46:42-04:00
Use showUnique in internalAnchorIdent
Showing the key of Unique as a number is generally not a great idea.
GHC Unique has a tag in high bits, so the raw number is unnecessarily
big.
So now we have
```html
<a href="#l-rvgK"><span class="hs-identifier hs-var hs-var">bar</span></a>
```
instead of
```html
<a href="#local-6989586621679015689"><span class="hs-identifier hs-var hs-var">bar</span></a>
```
Together with previous changes of shorter intra-module links the effect
on compressed files is not huge, that is expected as we simply remove
repetitive contents which pack well.
```
12_694_206 Agda-2.9.0-docs-orig.tar.gz
12_566_065 Agda-2.9.0-docs.tar.gz
```
However when unpacked, the difference can be significant,
e.g. Agda's largest module source got 5% reduction:
```
14_230_117 Agda.Syntax.Parser.Parser.html
13_422_109 Agda.Syntax.Parser.Parser.html
```
The whole hyperlinked source code directory got similar reduction
```
121M Agda-2.9.0-docs-orig/src
114M Agda-2.9.0-docs/src
```
For the reference, sources are about 2/3 of the generated haddocks
```
178M Agda-2.9.0-docs-old
172M Agda-2.9.0-docs
```
so we get around 3.5% size reduction overall. Not bad for a small local
changes.
-
6f63f57b
by Stefan Schulze Frielinghaus at 2025-09-17T04:47:22-04:00
rts: Fix alignment for gen_workspace #26334
After a0fa4941903272c48b050d24e93eec819eff51bd bootstrap is broken on
s390x and errors out with
rts/sm/GCThread.h:207:5: error:
error: alignment of array elements is greater than element size
207 | gen_workspace gens[];
| ^~~~~~~~~~~~~
The alignment constraint is applied via the attribute to the type
gen_workspace and leaves the underlying type struct gen_workspace_
untouched. On Aarch64, x86, and s390x the struct has a size of 128
bytes. On Aarch64 and x86 the alignments of 128 and 64 are divisors of
the size, respectively, which is why the type is a viable member type
for an array. However, on s390x, the alignment is 256 and therefore is
not a divisor of the size and hence cannot be used for arrays.
Basically I see two fixes here. Either decrease the alignment
requirement on s390x, or by applying the alignment constraint on the
struct itself. The former might affect performance as noted in
a0fa4941903272c48b050d24e93eec819eff51bd. The latter introduces padding
bits whenever necessary in order to ensure that
sizeof(gen_workspace[N])==N*sizeof(gen_workspace) holds which is done by
this patch.
-
06d25623
by Cheng Shao at 2025-09-17T19:32:27-04:00
ghci: add :shell command
This patch adds a new :shell command to ghci which works similarly to
:!, except it guarantees to run the command via sh -c. On POSIX hosts
the behavior is identical to :!, but on Windows it uses the msys2
shell instead of system cmd.exe shell. This is convenient when writing
simple ghci scripts that run simple POSIX commands, and the behavior
can be expected to be coherent on both Windows and POSIX.
Co-authored-by: Codex <codex@openai.com>
-
186054f7
by Cheng Shao at 2025-09-17T19:32:27-04:00
testsuite: remove legacy :shell trick
This commit makes use of the built-in :shell functionality in ghci in
the test cases, and remove the legacy :shell trick.
-
0a3a4aa3
by Cheng Shao at 2025-09-17T19:32:27-04:00
docs: document :shell in ghci
This commit documents the :shell command in ghci.
Co-authored-by: Codex <codex@openai.com>
-
a4ff12bb
by Cheng Shao at 2025-09-17T19:33:09-04:00
ghc-internal: fix codepages program
codepages was not properly updated during the base -> ghc-internal
migration, this commit fixes it.
-
7e094def
by Cheng Shao at 2025-09-17T19:33:09-04:00
ghc-internal: relax ucd2haskell cabal upper bounds
This commit relaxes ucd2haskell cabal upper bounds to make it runnable
via ghc 9.12/9.14.
-
7077c9f7
by Cheng Shao at 2025-09-17T19:33:09-04:00
ghc-internal: update to unicode 17.0.0
This commit updates the generated code in ghc-internal to match
unicode 17.0.0.
-
cef8938f
by sheaf at 2025-09-17T19:34:09-04:00
Bad record update msg: allow out-of-scope datacons
This commit ensures that, when we encounter an invalid record update
(because no constructor exists which contains all of the record fields
mentioned in the record update), we graciously handle the situation in
which the constructors themselves are not in scope. In that case,
instead of looking up the constructors in the GlobalRdrEnv, directly
look up their GREInfo using the lookupGREInfo function.
Fixes #26391
-
a2d9d7c2
by sheaf at 2025-09-17T19:34:09-04:00
Improve Notes about disambiguating record updates
This commit updates the notes [Disambiguating record updates] and
[Type-directed record disambiguation], in particular adding more
information about the deprecation status of type-directed disambiguation
of record updates.
-
de44e69e
by sheaf at 2025-09-19T05:16:51-04:00
Enable TcM plugins in initTc
This commit ensures that we run typechecker plugins and defaulting
plugins whenever we call initTc.
In particular, this ensures that the pattern-match checker, which calls
'initTcDsForSolver' which calls 'initTc', runs with typechecker plugins
enabled. This matters for situations like:
merge :: Vec n a -> Vec n a -> Vec (2 * n) a
merge Nil Nil = Nil
merge (a <: as) (b <: bs) = a :< (b <: merge as bs)
in which we need the typechecker plugin to run in order to tell us that
the Givens would be inconsistent in the additional equation
merge (_ <: _) Nil
and thus that the equation is not needed.
Fixes #26395
-
2c378ad2
by Cheng Shao at 2025-09-19T05:17:33-04:00
rel-eng: update fedora image to 42
This patch is a part of #25876 and updates fedora image to 42.
-
0a9d9ffc
by Sylvain Henry at 2025-09-19T13:12:14-04:00
Fix output of T14999 (#23685)
Fix output of T14999 to:
- take into account the +1 offset to DW_AT_low_pc (see Note [Info Offset])
- always use Intel's syntax to force consistency: it was reported that
sometimes GDB prints `jmpq` instead of `jmp` with the AT&T syntax
-
1480872a
by Vladislav Zavialov at 2025-09-19T13:12:54-04:00
Fix PREP_MAYBE_LIBRARY in prep_target_file.m4
This change fixes a configure error introduced in:
commit 8235dd8c4945db9cb03e3be3c388d729d576ed1e
ghc-toolchain: Move UseLibdw to per-Target file
Now the build no longer fails with:
acghc-toolchain: Failed to read a valid Target value from hadrian/cfg/default.target
-
d1d9e39e
by Ben Gamari at 2025-09-19T18:24:52-04:00
StgToByteCode: Don't assume that data con workers are nullary
Previously StgToByteCode assumed that all data-con workers were of a
nullary representation. This is not a valid assumption, as seen
in #23210, where an unsaturated application of a unary data
constructor's worker resulted in invalid bytecode. Sadly, I have not yet
been able to reduce a minimal testcase for this.
Fixes #23210.
-
3eeecd50
by Ben Gamari at 2025-09-19T18:24:53-04:00
testsuite: Mark T23146* as unbroken
-
2e73f342
by sheaf at 2025-09-19T18:24:53-04:00
Add test for #26216
-
c2efb912
by Sven Tennie at 2025-09-19T18:25:36-04:00
Generate correct test header
This increases convenience when copying & pasting...
-
d2fb811e
by Sven Tennie at 2025-09-19T18:25:36-04:00
foundation test: Fix shift amount (#26248)
Shift primops' results are only defined for shift amounts of 0 to word
size - 1. The approach is similar to testing div-like operations (which
have a constraint regarding zero operands.)
This was partly vibe coded (https://github.com/supersven/ghc/pull/1) but
then heavily refactored.
-
a62ce115
by Andreas Klebinger at 2025-09-19T18:26:18-04:00
Tweak jspace test
I've given it a longer timeout, and tweaked the test file generation
to speed it up a bit. Hopefully that is enough to make it constentily pass.
Last but not least it now also always uses three threads.
-
0f034942
by Cheng Shao at 2025-09-19T18:26:59-04:00
rts: remove obsolete CC_SUPPORTS_TLS logic
This patch removes obsolete CC_SUPPORTS_TLS logic throughout the rts,
given __thread is now uniformly supported by C toolchains of all
platforms we currently support.
-
ef705655
by Cheng Shao at 2025-09-19T18:27:41-04:00
rts: remove obsolete HAS_VISIBILITY_HIDDEN logic
This patch removes obsolete HAS_VISIBILITY_HIDDEN logic throughout the
rts, given __attribute__((visibility("hidden"))) is uniformly
supported by C toolchains of all platforms we currently support.
-
9fdc1f7d
by Cheng Shao at 2025-09-19T18:28:21-04:00
rts: remove -O3 pragma hack in Hash.c
This patch removes an obsolete gcc pragma to specify -O3 in Hash.c.
Hadrian already passes the right flag.
-
b8cfa8f7
by Cheng Shao at 2025-09-19T18:29:01-04:00
rts: remove obsolete COMPILING_WINDOWS_DLL logic
This patch removes obsolete COMPILING_WINDOWS_DLL logic throughout the
rts. They were once used for compiling to win32 DLLs, but we haven't
been able to compile Haskell units to win32 DLLs for many years now,
due to PE format's restriction of no more than 65536 exported symbols
in a single DLL.
-
bb760611
by Cheng Shao at 2025-09-19T18:29:42-04:00
wasm: bump browser_wasi_shim to 0.4.2
This patch bumps the browser_wasi_shim dependency of wasm dyld script
to 0.4.2.
-
8b0940db
by Cheng Shao at 2025-09-20T06:48:05-04:00
compiler: move Binary instance of Map to GHC.Utils.Binary
This patch moves `Binary` instance of `Map` from `haddock-api` to
`GHC.Utils.Binary`. This also allows us to remove a redundant instance
defined for `NameEntityInfo`, which is a type synonym for `Map`.
-
4a8fed75
by Vladislav Zavialov at 2025-09-20T06:48:47-04:00
Fix keyword in ExplicitNamespaces error message (#26418)
Consider this module header and the resulting error:
{-# LANGUAGE NoExplicitNamespaces #-}
module T26418 (data HeadC) where
-- error: [GHC-47007]
-- Illegal keyword 'type'
Previously, the error message would mention 'type' (as shown above),
even though the user wrote 'data'. This has now been fixed.
The error location has also been corrected: it is now reported at the
keyword position rather than at the position of the associated
import/export item.
-
867c2675
by Cheng Shao at 2025-09-20T06:49:28-04:00
wasm: fix dyld handling for forward declared GOT.func items
This patch fixes wasm shared linker's handling of forward declared
GOT.func items, see linked issue for details. Also adds T26430 test to
witness the fix. Fixes #26430.
Co-authored-by: Codex <codex@openai.com>
-
e7df6cc0
by Simon Peyton Jones at 2025-09-23T14:34:39-04:00
Improve pretty printer for HsExpr
Given a very deeply-nested application, it just kept printing
deeper and deeper. This small change makes it cut off.
Test is in #26330, but we also get a dramatic decrease in compile
time for perf/compiler/InstanceMatching:
InstanceMatching 4,086,884,584 1,181,767,232 -71.1% GOOD
Why? Because before we got a GIGANTIC error message that took
ages to pretty-print; now we get this much more civilised message
(I have removed some whitespace.)
Match.hs:1007:1: error:
• No instance for ‘Show (F001 a)’ arising from a use of ‘showsPrec’
• In the second argument of ‘showString’, namely
‘(showsPrec
11 b1
(GHC.Internal.Show.showSpace
(showsPrec
11 b2
(GHC.Internal.Show.showSpace
(showsPrec
11 b3
(GHC.Internal.Show.showSpace
(showsPrec
11 b4
(GHC.Internal.Show.showSpace
(showsPrec
11 b5
(GHC.Internal.Show.showSpace
(showsPrec
11 b6
(GHC.Internal.Show.showSpace (showsPrec ...)))))))))))))’
-----------------------
The main payload is
* At the start of `pprExpr`
* In the defn of `pprApp`
A little bit of refactoring:
* It turned out that we were setting the default cut-off depth to a
fixed value in two places, so changing one didn't change the other.
See defaultSDocDepth and defaultSDocCols
* I refactored `pprDeeperList` a bit so I could understand it better.
Because the depth calculation has changed, there are lots of small
error message wibbles.
Metric Decrease:
InstanceMatching
-
209f0158
by Simon Peyton Jones at 2025-09-23T14:34:39-04:00
Use Outputable.ellipsis rather than text "..."
-
64bb0e37
by Sylvain Henry at 2025-09-23T14:35:56-04:00
deriveConstants: automatically pass -fcommon CC flag (#26393)
By mistake we tried to use deriveConstants without passing
`--gcc-flag -fcommon` (which Hadrian does) and it failed.
This patch:
1. adds parsing support for constants stored in the .bss section (i.e.
when -fcommon isn't passed)
2. enables passing `-fcommon` automatically to the C compiler because
Windows requires this for subtle reasons
3. Documents the subtle reasons
(1) isn't strictly necessary because we always do (2) but it does no
harm and it is still useful if the CC flags ever contain -fno-common
-
afcdf92f
by Oleg Grenrus at 2025-09-23T14:36:41-04:00
Don't wrap spaces in <span>s
Doing similar comparison as in 63189b2ceca07edf4e179f4180ca60d470c62cb3
With this change the gzipped documentation is now 2% smaller (previously 1%)
12_694_206 Agda-2.9.0-docs-orig.tar.gz
12_436_829 Agda-2.9.0-docs.tar.gz
Unzipped docs are 5% smaller (previously 3%)
178M Agda-2.9.0-docs-orig
169M Agda-2.9.0-docs
Individual hyperlinked sources are around 7-10% smaller (previously 5%)
(`Parser` module is generated by happy and has relatively little whitespace)
14_230_117 Agda.Syntax.Parser.Parser.html
13_220_758 Agda.Syntax.Parser.Parser.html
Agda's hyperlinked sources are 9% smaller now:
121M Agda-2.9.0-docs-orig/src
110M Agda-2.9.0-docs/src
-
67de53a6
by Cheng Shao at 2025-09-23T14:37:31-04:00
rts: remove obsolete __GNUC__ related logic
This patch removes obsolete `__GNUC__` related logic, given on any
currently supported platform and toolchain, `__GNUC__ >= 4` is
universally true. Also pulls some other weeds and most notably, use
`__builtin___clear_cache` for clang as well, since clang has supported
this gcc intrinsic since 2014, see
https://github.com/llvm/llvm-project/commit/c491a8d4577052bc6b3b4c72a7db6a7cfcbc2ed0.
-
c4d32493
by Sven Tennie at 2025-09-23T20:40:57-04:00
RV64: Fix: Add missing truncation to MO_S_Shr (#26248)
Sub-double word (<W64) registers need to be truncated after the
operation.
-
41dce477
by Sven Tennie at 2025-09-23T20:40:57-04:00
RV64: Cleanup shift emitting cases/code
Remove overlapping cases to make the shift logic easier to understand.
-
0a601c30
by Alex Washburn at 2025-09-23T20:41:41-04:00
Correcting LLVM linking of Intel BMI intrinsics pdep{8,16} and pext{8,16}.
This patch fixes #26065.
The LLVM interface does not expose bindings to:
- llvm.x86.bmi.pdep.8
- llvm.x86.bmi.pdep.16
- llvm.x86.bmi.pext.8
- llvm.x86.bmi.pext.16
So calls are instead made to llvm.x86.bmi.{pdep,pext}.32 in these cases,
with pre/post-operation truncation to constrain the logical value range.
-
89e8ff3d
by Peng Fan at 2025-09-23T20:42:37-04:00
NCG/LA64: Implement MO_BSwap and MO_BRev with bit-manipulation Instructions
-
50f6be09
by Sylvain Henry at 2025-09-23T20:43:29-04:00
Allow Core plugins to access unoptimized Core (#23337)
Make the first simple optimization pass after desugaring a real CoreToDo
pass. This allows CorePlugins to decide whether they want to be executed
before or after this pass.
-
30ef0aac
by Simon Hengel at 2025-09-23T20:44:12-04:00
docs: Fix typo in scoped_type_variables.rst
-
f8919262
by Cheng Shao at 2025-09-23T20:44:54-04:00
ghci: fix bootstrapping with 9.12.3-rc1 and above
This patch fixes bootstrapping GHC with 9.12.3-rc1 and above. ghci
defines `Binary` instance for `HalfWord` in `ghc-heap`, which is a
proper `newtype` in 9.14 and starting from 9.12.3. Given we don't
build `ghc-heap` in stage0, we need to fix this predicate so that it
corresponds to the boot ghc versions that contain the right version of
`ghc-heap`.
-
a7f15858
by sheaf at 2025-09-24T09:49:53-04:00
User's guide: clarify optimisation of INLINABLE unfoldings
This updates the user's guide section on INLINABLE pragmas to explain how
the unfoldings of inlineable functions are optimised. The user's guide incorrectly
stated that the RHS was not optimised at all, but this is not true. Instead, GHC
is careful about phase control to optmise the RHS while retaining the guarantee
that GHC behaves as if the original RHS had been written.
-
495886d9
by Rodrigo Mesquita at 2025-09-24T09:50:35-04:00
cleanup: Delete historical artifact of COMPILING_WINDOWS_DLL
Namely, drop the obsolete
- DLL_IMPORT_RTS
- DLL_IMPORT_DATA_VAR
- DLL_IMPORT_DATA_VARNAME
- DLL_IMPORT_DATA_REF
These macros were not doing anything and placed inconsistently
Looking at the git logs reveal these macros were used to support
dynamic libraries on Win32, a feature that was dropped
in b8cfa8f741729ef123569fb321c4b2ab4a1a941c
This allows us to get rid of the rts/DLL.h file too.
-
5ae89054
by Sylvain Henry at 2025-09-24T17:07:00-04:00
Allow disabling builtin rules (#20298)
Add a way to disable built-in rules programmatically and with a debug flag.
I also took the opportunity to add a debug flag to disable bignum rules,
which was only possible programmatically (e.g. in a plugin).
-
135242ca
by Rodrigo Mesquita at 2025-09-24T17:07:44-04:00
Don't use build CFLAGS and friends as target settings
In the GHC in tree configure, `CFLAGS`, `CXXFLAGS`, and similar tool
configuration flags apply to the BUILD phase of the compiler, i.e. to
the tools run to compile GHC itself.
Notably, they should /not/ be carried over to the Target settings, i.e.
these flags should /not/ apply to the tool which GHC invokes at runtime.
Fixes #25637
-
b418408b
by Irene Knapp at 2025-09-25T09:47:54-04:00
Document etymology of "bind" as the name for `>>=`
It took me twenty years of contemplation to realize why it's called that.
I therefore feel that it may not be obvious to beginners.
-
e9c5e46f
by Brandon Chinn at 2025-09-25T09:48:36-04:00
Fix tabs in string gaps (#26415)
Tabs in string gaps were broken in bb030d0d because previously, string gaps were manually parsed, but now it's lexed by the usual Alex grammar and post-processed after successful lexing.
It broke because of a discrepancy between GHC's lexer grammar and the Haskell Report. The Haskell Report includes tabs in whitechar:
whitechar → newline | vertab | space | tab | uniWhite
$whitechar used to include tabs until 18 years ago, when it was removed in order to exclude tabs from $white_no_nl in order to warn on tabs: 6e202120. In this MR, I'm adding \t back into $whitechar, and explicitly excluding \t from the $white_no_nl+ rule ignoring all whitespace in source code, which more accurately colocates the "ignore all whitespace except tabs, which is handled in the next line" logic.
As a side effect of this MR, tabs are now allowed in pragmas; currently, a pragma written as {-# \t LANGUAGE ... #-} is interpreted as the tab character being the pragma name, and GHC warns "Unrecognized pragma". With this change, tabs are ignored as whitespace, which more closely matches the Report anyway.
-
8bf5b309
by Cheng Shao at 2025-09-25T09:49:18-04:00
wasm: remove the --no-turbo-fast-api-calls hack from dynamic linker shebang
This patch removes the `--no-turbo-fast-api-calls` hack from the dyld
script shebang; it was used to workaround v8 fast call coredumps in
nodejs and no longer needed, and comes with a performance penalty,
hence the removal.
-
c1cab0c3
by Sylvain Henry at 2025-09-26T10:36:30-04:00
Revert "Add necessary flag for js linking"
This reverts commit 84f68e2231b2eddb2e1dc4e90af394ef0f2e803f.
This commit didn't have the expected effect. See discussion in #26290.
Instead we export HEAP8 and HEAPU8 from rts/js/mem.js
-
0a434a80
by Sylvain Henry at 2025-09-26T10:36:30-04:00
JS: export HEAPU8 (#26290)
This is now required by newer Emscripten versions.
-
b10296a9
by Andreas Klebinger at 2025-09-26T10:37:11-04:00
sizeExpr: Improve Tick handling.
When determining if we scrutinize a function argument we
now properly look through ticks. Fixes #26444.
-
b6ef1bfd
by Wolfgang Jeltsch at 2025-09-26T21:09:30+03:00
Add operations for obtaining operating-system handles
This resolves #26265.