-
9a9ae4df
by Duncan Coutts at 2026-05-05T14:44:37-04:00
Use __attribute__((dllimport)) for external RTS symbol declarations
This is needed to be hygenic about DLL symbol imports and exports.
The attribute is ignored on platforms other than Windows.
Use of the attribute however means that external data symbols do not
have a compile-time constant address (they are loaded using an
indirection). This means we have to adjust the rtsSyms initial linker
table so that it is a local constant in a function, rather than a global
constant. We now define it within a function that pre-populates the
symbol table with the RTS symbols.
-
2ad3e01e
by Duncan Coutts at 2026-05-05T14:44:37-04:00
Fix the rts linker declarations for a few data symbols
and ensure that the (windows only) rts_IOManagerIsWin32Native data
symbol is marked as externally visible.
-
8ff4fdb5
by David Eichmann at 2026-05-05T14:44:37-04:00
Hadrian: Disable runtime pseudo relocations for RTS on windows hosts
-
96974723
by Teo Camarasu at 2026-05-05T14:45:20-04:00
ghci/TH: refactor to use IORef QState
This is a pure refactor and shouldn't modify semantics at all
-
eff6bfaf
by Teo Camarasu at 2026-05-05T14:45:20-04:00
iserv: recover/getQ/putQ should behave same as internal interpreter
The internal and external interpreter should behave the same when
handling `recover`, the exeception recovery method of Q.
In practice, they diverge. In case of failure, the internal interpreter
only restores error message state to before the computation, wheras the
external interperter restores error message state *and* the state of putQ/getQ.
As far as I can tell this is a simple mistake in the implementation.
Note [TH recover with -fexternal-interpreter] describes the correct
behaviour but the implementation doesn't mirror this.
This change restores the correct behaviour by keeping the effects of
putQ in the erroring computation.
This is a breaking change since it modifies the behaviour of programs
that rely on recover ignoring putQ from failling computations when used
with the external interpreter. Although I highly doubt anyone relies on
this behaviour.
This divergence was first introduced in d00c308633fe7d216d31a1087e00e63532d87d6d.
As far as I can tell this was unintentional and tha commit was trying to solve a different bug.
Resolves #27022
-
1cb1d672
by Wen Kokke at 2026-05-06T09:53:40-04:00
rts: Add dynamic trace flags API
This commit adds an API to the RTS (exposed via Rts.h) that allows users to dynamically change the trace flags.
Prior to this commit, users were able to stop and start the profiling and heap profiling timers (via startProfTimer/stopProfTimer and startHeapProfTimer/stopHeapProfTimer).
This extends that functionality to also cover the core event types.
The getTraceFlag/setTraceFlag functions read and write the values of the trace flag cache, which is allocated by Trace.c, rather than modifying the members of RtsFlags.TraceFlags.
This is done under the assumption that the members of RtsFlags should not be modified after RTS initialisation.
Consequently, if the user modifies the trace flags using setTraceFlag, the object returned by getTraceFlags (from base) will not reflect these changes.
The trace flags are not protected by locks of any sort.
Hence, these functions are not thread-safe.
However, the trace flags are not modified by the RTS after initialisation, only read, so the race conditions introduced by one user modifying them are most likely benign.
This PR also puts the trace flag cache in a single global struct, as opposed to a collection of global variables, and changes the types of the individual flags from uint8_t to bool, as these have the same size on both Clang and GCC and are a better semantic match.
Prior to the change to uint8_t, they had type int, see 42c47cd6.
Even with its deprecation in C23, I don't think there should be any issue depending on stdbool.h.
The TRACE_X macros are redefined to access the global struct, with values cast to const bool to ensure they are read-only.
-
9d54dc94
by Wen Kokke at 2026-05-06T09:53:40-04:00
rts: Ensure TRACE_X values are used in place of RtsFlags.TraceFlags.X
-
418d737b
by Wen Kokke at 2026-05-06T09:53:40-04:00
rts: Fix nonmoving-GC tracing
The current nonmoving-GC tracing functions were written in a different
style from the other tracing functions. They were directly implemented
as, e.g., a traceConcMarkEnd function that called postConcMarkEnd.
The other tracing functions are implemented as, e.g., traceThreadLabel_,
a function that posts the thread label event, and traceThreadLabel, a
macro that checks whether TRACE_scheduler is set. This commit fixes that
implementation, and ensures that the nonmoving-GC tracing functions only
emit events if nonmoving-GC tracing is enabled.
-
99f4afa4
by Wen Kokke at 2026-05-06T09:53:40-04:00
rts: Add SymI_HasProto for get/setTraceFlag
-
7e9eb8b9
by Wen Kokke at 2026-05-06T09:53:40-04:00
rts: Add SymI_HasProto for start/endEventLogging
-
3a3045fb
by Wen Kokke at 2026-05-06T09:53:41-04:00
rts: Add changelog entry
-
a3b339a4
by Teo Camarasu at 2026-05-06T09:54:25-04:00
interface-stability/base: don't distinguish ws-32
The interface of base is identical when the Word size is 32bits.
Therefore, there is no need to have another file for this case.
So, we delete it.
Step towards: #26752
-
eb922183
by Duncan Coutts at 2026-05-07T14:28:50+01:00
Add a rts posix FdWakup utility module
This will be used to implement wakeupIOManager for in-RTS I/O managers.
It provides a notification/wakeup mechanism using FDs, suitable for
situations when a thread is blocked on a set of fds anyway. It uses the
classic self-pipe trick, or equivalently eventfd on supported platforms.
This will initially be used to implement prompt interrupt or shutdown of
the posix ticker thread.
-
01b0e233
by Duncan Coutts at 2026-05-07T14:28:50+01:00
Add prompt shutdown to the pthread ticker implementation.
The Linux timerfd ticker monitors a pipe which is used by exitTicker to
ensure a prompt wakeup and shutdown. The pthread ticker lacked this and
so would only exit at the next ticker wakeup (10ms by default).
This patch adds the same mechanism to the pthread ticker.
This changes the pthread ticker from waiting by using nanosleep() to
waiting using either ppoll() or select(), so that it can wait on both
a time and a file descriptor. On Linux at least, a test program to
compare the timing jitter of these APIs shows that using nanpsleep,
ppoll or select makes no statistical difference to the maximum or
average jitter.
This is a step towards unifying the posix ticker implementations, so
that we can have just one portable one (albeit with some limited cpp).
It is also a step towards using the ticker as part of a more general
implementation of wakeUpRts, since this will require a method to wake
the rts from a signal handler context (ctl-c handler).
-
bc41d646
by Duncan Coutts at 2026-05-07T14:28:50+01:00
Update ticker header commentary
It was antique and didn't apply even to the previous implementation, and
certainly not to the updated one.
-
4ed9a386
by Duncan Coutts at 2026-05-07T14:28:50+01:00
Remove the timerfd-based ticker implementation
There does not appear to be any remaining advantage on Linux to using
the timerfd ticker implementation over the portable one (using ppoll on
Linux for precise timing).
The eventfd implementation was originally added at a time when Linux was
still using a signal based implementation. So it made sense at the time.
See (closed) issue #10840.
-
97504fa6
by Duncan Coutts at 2026-05-07T14:28:50+01:00
Consolidate to a single posix ticker implementation
Previously we had four implementations, two using signals and two using
threads. Having just one should make behaviour more consistent between
platforms, and should make maintenance easier.
-
1e60023b
by Facundo Domínguez at 2026-05-07T18:01:16-04:00
Generalize so_inline to specify which bindings should be preserved
This commit generalizes the so_inline option of the simple optimizer
so we can indicate with a predicate the specific bindings that should
be kept.
This feature is important for the LiquidHaskell plugin, which relies on the
simple optimizer to make core programs easier to read, but needs to preserve
bindings that are relevant for verification.
See https://gitlab.haskell.org/ghc/ghc/-/issues/24386 for the full discussion.
-
44cf9cd7
by Wolfgang Jeltsch at 2026-05-12T09:48:18-04:00
Move the `Text.Read` implementation into `base`
-
4ac3f7d6
by Vladislav Zavialov at 2026-05-12T09:49:03-04:00
EPA: Use AnnParen for tuples and sums
Summary of changes
* Do not use AnnParen in XListTy, replace it with EpToken "[" and "]"
* Specialise AnnParen to tuple/sums by dropping the AnnParensSquare
and keeping only AnnParens and AnnParensHash
* Use AnnParen in XExplicitTuple
* Use AnnParen in XExplicitTupleTy
* Use AnnParen in XTuplePat
* Use AnnParen in XExplicitSum (via AnnExplicitSum)
* Use AnnParen in XSumPat (via EpAnnSumPat)
This is a refactoring with no user-facing changes.
-
1bdcddec
by Duncan Coutts at 2026-05-12T09:49:48-04:00
Add minimal dlltool support to ghc-toolchain
The dlltool is a tool that can create dll import libraries from .def
files. These .def files list the exported symbols of dlls. Its somewhat
like gnu linker scripts, but more limited.
We will need dlltool to build the rts and ghc-internal libraries as DLLs
on Windows. The rts and ghc-internal libraries have a recursive
dependency on each other. Import libraries can be used to resolve
recursive dependencies between dlls. We will use an import library for
the rts when linking the ghc-internal library.
-
f7fc3770
by Duncan Coutts at 2026-05-12T09:49:48-04:00
Add minimal dlltool support into ./configure
Find dlltool, and hopefully support finding it within the bundled llvm
toolchain on windows.
-
e4e22bfb
by Duncan Coutts at 2026-05-12T09:49:48-04:00
Update the default host and target files for dlltool support
-
5666c8f9
by Duncan Coutts at 2026-05-12T09:49:48-04:00
Add dlltool as a hadrian builder
Optional except on windows.
-
5e14fe3f
by Duncan Coutts at 2026-05-12T09:49:48-04:00
Update and generate libHSghc-internal.def from .def.in file
The only symbol that the rts imports from the ghc-internal package now
is init_ghc_hs_iface. So the rts only needs an import lib that defines
that one symbol.
Also, remove the libHSghc-prim.def because it is redundant. The rts no
longer imports anything from ghc-prim.
Keep libHSffi.def for now. We may yet need it once it is clear how
libffi is going to be built/used for ghc.
-
3d91e4a6
by Duncan Coutts at 2026-05-12T09:49:48-04:00
Add rule to build libHSghc-internal.dll.a and link into the rts
On windows only, with dynamic linking.
This is needed because on windows, all symbols in dlls must be resolved.
No dangling symbols allowed. References to external symbols must be
explicit. We resolve this with an import library. We create an import
library for ghc-internal, a .dll.a file. This is a static archive
containing .o files that define the symbols we need, and crucially have
".idata" sections that specifies the symbols the dll imports and from
where.
Note that we do not install this libHSghc-internal.dll.a, and it does
not need to list all the symbols exported by that package. We create a
special purpose import lib and only use it when linking the rts dll, so
it only has to list the symbols that the rts uses from ghc-internal
(which is exactly one symbol: init_ghc_hs_iface).
-
c8dae539
by Alice Rixte at 2026-05-12T09:50:52-04:00
Script for downloading and copying `base-exports` file
-
dea6fdbf
by Simon Hengel at 2026-05-13T19:15:39+07:00
Don't use global variables to address concurrency bugs! (fixes #27234)
This was originally introduce with
88f38b03025386f0f1e8f5861eed67d80495168a to address #17922.
In this specific case a better fix would have been to synchronize on
stderr:
withHandle_ "stderrSupportsAnsiColors" stderr $ \ _ -> do
...
But apparently the dependency on `terminfo` was removed in
32ab07bf3d6ce45e8ea5b55e8095174a6b42a7f0, preventing #17922 in the first
place.