August 2025 - ghc-commits

[Git][ghc/ghc][master] Allow defining HasField instances for naughty fields
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 1107af89 by Oleg Grenrus at 2025-08-13T21:08:06-04:00 Allow defining HasField instances for naughty fields Resolves #26295 ... as HasField solver doesn't solve for fields with "naughty" selectors, we could as well allow defining HasField instances for these fields. - - - - - 4 changed files: - compiler/GHC/Tc/Validity.hs - + testsuite/tests/overloadedrecflds/should_run/T26295.hs - + testsuite/tests/overloadedrecflds/should_run/T26295.stdout - testsuite/tests/overloadedrecflds/should_run/all.T Changes: ===================================== compiler/GHC/Tc/Validity.hs ===================================== @@ -31,6 +31,7 @@ import GHC.Tc.Instance.Family import GHC.Tc.Types.Origin import GHC.Tc.Types.Rank import GHC.Tc.Errors.Types +import GHC.Tc.Utils.Env (tcLookupId) import GHC.Tc.Utils.TcType import GHC.Tc.Utils.Monad import GHC.Tc.Zonk.TcType @@ -60,6 +61,8 @@ import qualified GHC.LanguageExtensions as LangExt import GHC.Types.Error import GHC.Types.Basic ( TypeOrKind(..), UnboxedTupleOrSum(..) , unboxedTupleOrSumExtension ) +import GHC.Types.Id (isNaughtyRecordSelector) +import GHC.Types.FieldLabel (flSelector) import GHC.Types.Name import GHC.Types.Var.Env import GHC.Types.Var.Set @@ -1718,8 +1721,17 @@ checkHasFieldInst cls tys@[_k_ty, _r_rep, _a_rep, lbl_ty, r_ty, _a_ty] = | otherwise -> case isStrLitTy lbl_ty of Just lbl | let lbl_str = FieldLabelString lbl - , isJust (lookupTyConFieldLabel lbl_str tc) - -> add_err $ IllegalHasFieldInstanceTyConHasField tc lbl_str + , Just fl <- lookupTyConFieldLabel lbl_str tc + -> do + -- GHC does not provide HasField instances for naughty record selectors + -- (see Note [Naughty record selectors] in GHC.Tc.TyCl.Utils), + -- so don't prevent the user from writing such instances. + -- See GHC.Tc.Instance.Class.matchHasField. + -- Test case: T26295. + sel_id <- tcLookupId $ flSelector fl + if isNaughtyRecordSelector sel_id + then return () + else add_err $ IllegalHasFieldInstanceTyConHasField tc lbl_str | otherwise -> return () Nothing ===================================== testsuite/tests/overloadedrecflds/should_run/T26295.hs ===================================== @@ -0,0 +1,26 @@ +{-# LANGUAGE GHC2021 #-} +{-# LANGUAGE GADTs #-} +{-# LANGUAGE DataKinds #-} +{-# LANGUAGE OverloadedRecordDot #-} +{-# LANGUAGE UndecidableInstances #-} +import GHC.Records + +-- large-records mangles record definitions to look like below to +-- prevent selector function generation (even implicit ones) +data R = forall a b. (a ~ Int, b ~ Char) => MkR + { field_a :: a + , field_b :: b + } + +-- fields in R are naughty, so we can define custom HasField instancs for them +instance a ~ Int => HasField "field_a" R a where + getField (MkR a _) = a + +ex :: Int +ex = r.field_a + where + r :: R + r = MkR 42 'x' + +main :: IO () +main = print ex ===================================== testsuite/tests/overloadedrecflds/should_run/T26295.stdout ===================================== @@ -0,0 +1 @@ +42 ===================================== testsuite/tests/overloadedrecflds/should_run/all.T ===================================== @@ -20,3 +20,4 @@ test('T12243', normal, compile_and_run, ['']) test('T11228', normal, compile_and_run, ['']) test('T11671_run', normal, compile_and_run, ['']) test('T17551b', [req_th], compile_and_run, ['']) +test('T26295', [], compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1107af899a1222f0d6101dacddca7e6… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/1107af899a1222f0d6101dacddca7e6… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] 2 commits: rts: ensure MessageBlackHole.link is always a valid closure
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: a8b2fbae by Teo Camarasu at 2025-08-13T21:07:24-04:00 rts: ensure MessageBlackHole.link is always a valid closure We turn a MessageBlackHole into an StgInd in wakeBlockingQueue(). Therefore it's important that the link field, which becomes the indirection field, always points to a valid closure. It's unclear whether it's currently possible for the previous behaviour to lead to a crash, but it's good to be consistent about this invariant nonetheless. Co-authored-by: Andreas Klebinger <klebinger.andreas(a)gmx.at> - - - - - 4021181e by Teo Camarasu at 2025-08-13T21:07:24-04:00 rts: spin if we see a WHITEHOLE in messageBlackHole When a BLACKHOLE gets cancelled in raiseAsync, we indirect to a THUNK. GC can then shortcut this, replacing our BLACKHOLE with a fresh THUNK. This THUNK is not guaranteed to have a valid indirectee field. If at the same time, a message intended for the previous BLACKHOLE is processed and concurrently we BLACKHOLE the THUNK, thus temporarily turning it into a WHITEHOLE, we can get a segfault, since we look at the undefined indirectee field of the THUNK The fix is simple: spin if we see a WHITEHOLE, and it will soon be replaced with a valid BLACKHOLE. Resolves #26205 - - - - - 3 changed files: - rts/Messages.c - rts/StgMiscClosures.cmm - rts/Updates.h Changes: ===================================== rts/Messages.c ===================================== @@ -180,13 +180,22 @@ uint32_t messageBlackHole(Capability *cap, MessageBlackHole *msg) bh_info != &stg_CAF_BLACKHOLE_info && bh_info != &__stg_EAGER_BLACKHOLE_info && bh_info != &stg_WHITEHOLE_info) { - // if it is a WHITEHOLE, then a thread is in the process of - // trying to BLACKHOLE it. But we know that it was once a - // BLACKHOLE, so there is at least a valid pointer in the - // payload, so we can carry on. return 0; } + // If we see a WHITEHOLE then we should wait for it to turn into a BLACKHOLE. + // Otherwise we might look at the indirectee and segfault. + // See "Exception handling" in Note [Thunks, blackholes, and indirections] + // We might be looking at a *fresh* THUNK being WHITEHOLE-d so we can't + // guarantee that the indirectee is a valid pointer. +#if defined(THREADED_RTS) + if (bh_info == &stg_WHITEHOLE_info) { + while(ACQUIRE_LOAD(&bh->header.info) == &stg_WHITEHOLE_info) { + busy_wait_nop(); + } + } +#endif + // The blackhole must indirect to a TSO, a BLOCKING_QUEUE, an IND, // or a value. StgClosure *p; ===================================== rts/StgMiscClosures.cmm ===================================== @@ -31,6 +31,7 @@ import CLOSURE ENT_VIA_NODE_ctr; import CLOSURE RtsFlags; import CLOSURE stg_BLOCKING_QUEUE_CLEAN_info; import CLOSURE stg_BLOCKING_QUEUE_DIRTY_info; +import CLOSURE stg_END_TSO_QUEUE_closure; import CLOSURE stg_IND_info; import CLOSURE stg_MSG_BLACKHOLE_info; import CLOSURE stg_TSO_info; @@ -588,6 +589,9 @@ retry: MessageBlackHole_tso(msg) = CurrentTSO; MessageBlackHole_bh(msg) = node; + // Ensure that the link field is a valid closure, + // since we might turn this into an indirection in wakeBlockingQueue() + MessageBlackHole_link(msg) = stg_END_TSO_QUEUE_closure; SET_HDR(msg, stg_MSG_BLACKHOLE_info, CCS_SYSTEM); // messageBlackHole has appropriate memory barriers when this object is exposed. // See Note [Heap memory barriers]. ===================================== rts/Updates.h ===================================== @@ -333,6 +333,10 @@ * `AP_STACK` closure recording the aborted execution state. * See `RaiseAsync.c:raiseAsync` for details. * + * This can combine with indirection shortcutting during GC to replace a BLACKHOLE + * with a fresh THUNK. We should be very careful here since the THUNK will have an + * undefined value in the indirectee field. Looking at the indirectee field can then + * lead to a segfault such as #26205. * * CAFs * ---- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dfe6f464839e5dfae143b6d4ba1478… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/dfe6f464839e5dfae143b6d4ba1478… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] Refactoring: Don't misuse `MCDiagnostic` for lint messages
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: dfe6f464 by Simon Hengel at 2025-08-13T21:06:43-04:00 Refactoring: Don't misuse `MCDiagnostic` for lint messages `MCDiagnostic` is meant to be used for compiler diagnostics. Any code that creates `MCDiagnostic` directly, without going through `GHC.Driver.Errors.printMessage`, side steps `-fdiagnostics-as-json` (see e.g. !14475, !14492 !14548). To avoid this in the future I want to control more narrowly who creates `MCDiagnostic` (see #24113). Some parts of the compiler use `MCDiagnostic` purely for formatting purposes, without creating any real compiler diagnostics. This change introduces a helper function, `formatDiagnostic`, that can be used in such cases instead of constructing `MCDiagnostic`. - - - - - 4 changed files: - compiler/GHC/Core/Lint.hs - compiler/GHC/Stg/Lint.hs - compiler/GHC/Types/Error.hs - compiler/GHC/Utils/Error.hs Changes: ===================================== compiler/GHC/Core/Lint.hs ===================================== @@ -3418,7 +3418,7 @@ addMsg show_context env msgs msg [] -> noSrcSpan (s:_) -> s !diag_opts = le_diagOpts env - mk_msg msg = mkLocMessage (mkMCDiagnostic diag_opts WarningWithoutFlag Nothing) msg_span + mk_msg msg = mkLintWarning diag_opts msg_span (msg $$ context) addLoc :: LintLocInfo -> LintM a -> LintM a ===================================== compiler/GHC/Stg/Lint.hs ===================================== @@ -107,7 +107,6 @@ import GHC.Core.Type import GHC.Types.Basic ( TopLevelFlag(..), isTopLevel, isMarkedCbv ) import GHC.Types.CostCentre ( isCurrentCCS ) -import GHC.Types.Error ( DiagnosticReason(WarningWithoutFlag) ) import GHC.Types.Id import GHC.Types.Var.Set import GHC.Types.Name ( getSrcLoc, nameIsLocalOrFrom ) @@ -116,7 +115,7 @@ import GHC.Types.SrcLoc import GHC.Utils.Logger import GHC.Utils.Outputable -import GHC.Utils.Error ( mkLocMessage, DiagOpts ) +import GHC.Utils.Error ( DiagOpts ) import qualified GHC.Utils.Error as Err import GHC.Unit.Module ( Module ) @@ -540,7 +539,7 @@ addErr diag_opts errs_so_far msg locs = errs_so_far `snocBag` mk_msg locs where mk_msg (loc:_) = let (l,hdr) = dumpLoc loc - in mkLocMessage (Err.mkMCDiagnostic diag_opts WarningWithoutFlag Nothing) + in Err.mkLintWarning diag_opts l (hdr $$ msg) mk_msg [] = msg ===================================== compiler/GHC/Types/Error.hs ===================================== @@ -72,6 +72,7 @@ module GHC.Types.Error , pprMessageBag , mkLocMessage , mkLocMessageWarningGroups + , formatDiagnostic , getCaretDiagnostic , jsonDiagnostic @@ -495,11 +496,11 @@ data MessageClass -- ^ Diagnostics from the compiler. This constructor is very powerful as -- it allows the construction of a 'MessageClass' with a completely -- arbitrary permutation of 'Severity' and 'DiagnosticReason'. As such, - -- users are encouraged to use the 'mkMCDiagnostic' smart constructor + -- users are encouraged to use higher level primitives -- instead. Use this constructor directly only if you need to construct -- and manipulate diagnostic messages directly, for example inside -- 'GHC.Utils.Error'. In all the other circumstances, /especially/ when - -- emitting compiler diagnostics, use the smart constructor. + -- emitting compiler diagnostics, use higher level primitives. -- -- The @Maybe 'DiagnosticCode'@ field carries a code (if available) for -- this diagnostic. If you are creating a message not tied to any @@ -656,32 +657,51 @@ mkLocMessageWarningGroups -> SrcSpan -- ^ location -> SDoc -- ^ message -> SDoc - -- Always print the location, even if it is unhelpful. Error messages - -- are supposed to be in a standard format, and one without a location - -- would look strange. Better to say explicitly "<no location info>". mkLocMessageWarningGroups show_warn_groups msg_class locn msg - = sdocOption sdocColScheme $ \col_scheme -> - let locn' = sdocOption sdocErrorSpans $ \case - True -> ppr locn - False -> ppr (srcSpanStart locn) - + = case msg_class of + MCDiagnostic severity reason code -> formatDiagnostic show_warn_groups locn severity reason code msg + _ -> sdocOption sdocColScheme $ \col_scheme -> + let msg_colour = getMessageClassColour msg_class col_scheme - col = coloured msg_colour . text msg_title = coloured msg_colour $ case msg_class of - MCDiagnostic SevError _ _ -> text "error" - MCDiagnostic SevWarning _ _ -> text "warning" MCFatal -> text "fatal" _ -> empty + in formatLocMessageWarningGroups locn msg_title empty empty msg + +formatDiagnostic + :: Bool -- ^ Print warning groups? + -> SrcSpan -- ^ location + -> Severity + -> ResolvedDiagnosticReason + -> Maybe DiagnosticCode + -> SDoc -- ^ message + -> SDoc +formatDiagnostic show_warn_groups locn severity reason code msg + = sdocOption sdocColScheme $ \col_scheme -> + let + msg_colour :: Col.PprColour + msg_colour = getSeverityColour severity col_scheme + + col :: String -> SDoc + col = coloured msg_colour . text + + msg_title :: SDoc + msg_title = coloured msg_colour $ + case severity of + SevError -> text "error" + SevWarning -> text "warning" + SevIgnore -> empty + + warning_flag_doc :: SDoc warning_flag_doc = - case msg_class of - MCDiagnostic sev reason _code - | Just msg <- flag_msg sev (resolvedDiagnosticReason reason) - -> brackets msg - _ -> empty + case flag_msg severity (resolvedDiagnosticReason reason) of + Nothing -> empty + Just msg -> brackets msg + ppr_with_hyperlink :: DiagnosticCode -> SDoc ppr_with_hyperlink code = -- this is a bit hacky, but we assume that if the terminal supports colors -- then it should also support links @@ -691,10 +711,11 @@ mkLocMessageWarningGroups show_warn_groups msg_class locn msg then ppr $ LinkedDiagCode code else ppr code + code_doc :: SDoc code_doc = - case msg_class of - MCDiagnostic _ _ (Just code) -> brackets (ppr_with_hyperlink code) - _ -> empty + case code of + Just code -> brackets (ppr_with_hyperlink code) + Nothing -> empty flag_msg :: Severity -> DiagnosticReason -> Maybe SDoc flag_msg SevIgnore _ = Nothing @@ -725,13 +746,35 @@ mkLocMessageWarningGroups show_warn_groups msg_class locn msg vcat [ text "locn:" <+> ppr locn , text "msg:" <+> ppr msg ] + warn_flag_grp :: [WarningGroup] -> SDoc warn_flag_grp groups | show_warn_groups, not (null groups) = text $ "(in " ++ intercalate ", " (map (("-W"++) . warningGroupName) groups) ++ ")" | otherwise = empty + in formatLocMessageWarningGroups locn msg_title code_doc warning_flag_doc msg + +formatLocMessageWarningGroups + :: SrcSpan -- ^ location + -> SDoc -- ^ title + -> SDoc -- ^ diagnostic code + -> SDoc -- ^ warning groups + -> SDoc -- ^ message + -> SDoc +formatLocMessageWarningGroups locn msg_title code_doc warning_flag_doc msg + = sdocOption sdocColScheme $ \col_scheme -> + let + -- Always print the location, even if it is unhelpful. Error messages + -- are supposed to be in a standard format, and one without a location + -- would look strange. Better to say explicitly "<no location info>". + locn' :: SDoc + locn' = sdocOption sdocErrorSpans $ \case + True -> ppr locn + False -> ppr (srcSpanStart locn) + -- Add prefixes, like Foo.hs:34: warning: -- <the warning message> + header :: SDoc header = locn' <> colon <+> msg_title <> colon <+> code_doc <+> warning_flag_doc @@ -741,11 +784,16 @@ mkLocMessageWarningGroups show_warn_groups msg_class locn msg msg) getMessageClassColour :: MessageClass -> Col.Scheme -> Col.PprColour -getMessageClassColour (MCDiagnostic SevError _reason _code) = Col.sError -getMessageClassColour (MCDiagnostic SevWarning _reason _code) = Col.sWarning +getMessageClassColour (MCDiagnostic severity _reason _code) = getSeverityColour severity getMessageClassColour MCFatal = Col.sFatal getMessageClassColour _ = const mempty +getSeverityColour :: Severity -> Col.Scheme -> Col.PprColour +getSeverityColour severity = case severity of + SevError -> Col.sError + SevWarning -> Col.sWarning + SevIgnore -> const mempty + getCaretDiagnostic :: MessageClass -> SrcSpan -> IO SDoc getCaretDiagnostic _ (UnhelpfulSpan _) = pure empty getCaretDiagnostic msg_class (RealSrcSpan span _) = ===================================== compiler/GHC/Utils/Error.hs ===================================== @@ -32,7 +32,7 @@ module GHC.Utils.Error ( emptyMessages, mkDecorated, mkLocMessage, mkMsgEnvelope, mkPlainMsgEnvelope, mkPlainErrorMsgEnvelope, mkErrorMsgEnvelope, - mkMCDiagnostic, diagReasonSeverity, + mkLintWarning, diagReasonSeverity, mkPlainError, mkPlainDiagnostic, @@ -160,12 +160,10 @@ diag_reason_severity opts reason = fmap ResolvedDiagnosticReason $ case reason o ErrorWithoutFlag -> (SevError, reason) --- | Make a 'MessageClass' for a given 'DiagnosticReason', consulting the --- 'DiagOpts'. -mkMCDiagnostic :: DiagOpts -> DiagnosticReason -> Maybe DiagnosticCode -> MessageClass -mkMCDiagnostic opts reason code = MCDiagnostic sev reason' code +mkLintWarning :: DiagOpts -> SrcSpan -> SDoc -> SDoc +mkLintWarning opts span = formatDiagnostic True span severity reason Nothing where - (sev, reason') = diag_reason_severity opts reason + (severity, reason) = diag_reason_severity opts WarningWithoutFlag -- -- Creating MsgEnvelope(s) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/dfe6f464839e5dfae143b6d4ba14785… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/dfe6f464839e5dfae143b6d4ba14785… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] testsuite: Be more permissive when filtering out GNU_PROPERTY_TYPE linker warnings
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 51c701fe by Zubin Duggal at 2025-08-13T21:06:00-04:00 testsuite: Be more permissive when filtering out GNU_PROPERTY_TYPE linker warnings The warning text is slightly different with ld.bfd. Fixes #26249 - - - - - 1 changed file: - testsuite/driver/testlib.py Changes: ===================================== testsuite/driver/testlib.py ===================================== @@ -2988,7 +2988,7 @@ def normalise_errmsg(s: str) -> str: # filter out unsupported GNU_PROPERTY_TYPE (5), which is emitted by LLVM10 # and not understood by older binutils (ar, ranlib, ...) - s = modify_lines(s, lambda l: re.sub(r'^(.+)warning: (.+): unsupported GNU_PROPERTY_TYPE $5$ type: 0xc000000(.*)$', '', l)) + s = modify_lines(s, lambda l: re.sub(r'^(.+)warning: (.+): unsupported GNU_PROPERTY_TYPE (?:$5$ )?type: 0xc000000(.*)$', '', l)) s = re.sub(r'ld: warning: passed .* min versions $.*$ for platform macOS. Using [\.0-9]+.','',s) s = re.sub('ld: warning: -sdk_version and -platform_version are not compatible, ignoring -sdk_version','',s) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/51c701fef034e2062809eed5de3a51b… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/51c701fef034e2062809eed5de3a51b… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] Report -pgms as a deprecated flag
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 3506fa7d by Simon Hengel at 2025-08-13T21:05:18-04:00 Report -pgms as a deprecated flag (instead of reporting an unspecific warning) Before: on the commandline: warning: Object splitting was removed in GHC 8.8 After: on the commandline: warning: [GHC-53692] [-Wdeprecated-flags] -pgms is deprecated: Object splitting was removed in GHC 8.8 - - - - - 1 changed file: - compiler/GHC/Driver/Session.hs Changes: ===================================== compiler/GHC/Driver/Session.hs ===================================== @@ -1145,8 +1145,8 @@ dynamic_flags_deps = [ $ noArgM $ \d -> do deprecate $ "use -pgml-supports-no-pie instead" pure $ alterToolSettings (\s -> s { toolSettings_ccSupportsNoPie = True }) d) - , make_ord_flag defFlag "pgms" - (HasArg (\_ -> addWarn "Object splitting was removed in GHC 8.8")) + , make_dep_flag defFlag "pgms" + (HasArg (\_ -> return ())) "Object splitting was removed in GHC 8.8" , make_ord_flag defFlag "pgma" $ hasArg $ \f -> alterToolSettings $ \s -> s { toolSettings_pgm_a = (f,[]) } , make_ord_flag defFlag "pgml" View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3506fa7d4417ecd7225e413d3d77039… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/3506fa7d4417ecd7225e413d3d77039… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] ghc-internal: Fix naturalAndNot for NB/NS case
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: a766286f by Reed Mullanix at 2025-08-13T21:04:36-04:00 ghc-internal: Fix naturalAndNot for NB/NS case When the first argument to `naturalAndNot` is larger than a `Word` and the second is `Word`-sized, `naturalAndNot` will truncate the result: ``` >>> naturalAndNot ((2 ^ 65) .|. (2 ^ 3)) (2 ^ 3) 0 ``` In contrast, `naturalAndNot` does not truncate when both arguments are larger than a `Word`, so this appears to be a bug. Luckily, the fix is pretty easy: we just need to call `bigNatAndNotWord#` instead of truncating. Fixes #26230 - - - - - 6 changed files: - libraries/base/changelog.md - libraries/ghc-bignum/changelog.md - libraries/ghc-internal/src/GHC/Internal/Bignum/Natural.hs - + testsuite/tests/numeric/should_run/T26230.hs - + testsuite/tests/numeric/should_run/T26230.stdout - testsuite/tests/numeric/should_run/all.T Changes: ===================================== libraries/base/changelog.md ===================================== @@ -5,6 +5,7 @@ * Fix issues with toRational for types capable to represent infinite and not-a-number values ([CLC proposal #338](https://github.com/haskell/core-libraries-committee/issues/338)) * Modify the implementation of `Data.List.sortOn` to use `(>)` instead of `compare`. ([CLC proposal #332](https://github.com/haskell/core-libraries-committee/issues/332)) * `GHC.Exts.IOPort#` and its related operations have been removed ([CLC #213](https://github.com/haskell/core-libraries-committee/issues/213)) + * Fix bug where `naturalAndNot` was incorrectly truncating results ([CLC proposal #350](github.com/haskell/core-libraries-committee/issues/350)) ## 4.22.0.0 *TBA* * Shipped with GHC 9.14.1 ===================================== libraries/ghc-bignum/changelog.md ===================================== @@ -5,6 +5,8 @@ - `ghc-bignum`'s implementation has been merged into `ghc-internal`. Downstream users should import `GHC.Num.{Integer,Natural,BigNat}` stable modules from `base` instead. +- Fix bug where `naturalAndNot` was incorrectly truncating results (#26230) + ## 1.3 ===================================== libraries/ghc-internal/src/GHC/Internal/Bignum/Natural.hs ===================================== @@ -488,7 +488,7 @@ naturalAndNot :: Natural -> Natural -> Natural {-# NOINLINE naturalAndNot #-} naturalAndNot (NS n) (NS m) = NS (n `and#` not# m) naturalAndNot (NS n) (NB m) = NS (n `and#` not# (bigNatToWord# m)) -naturalAndNot (NB n) (NS m) = NS (bigNatToWord# n `and#` not# m) +naturalAndNot (NB n) (NS m) = NB (bigNatAndNotWord# n m) naturalAndNot (NB n) (NB m) = naturalFromBigNat# (bigNatAndNot n m) naturalOr :: Natural -> Natural -> Natural ===================================== testsuite/tests/numeric/should_run/T26230.hs ===================================== @@ -0,0 +1,8 @@ +import Data.Bits +import GHC.Num.Natural + +main = do + print $ naturalAndNot ((2 ^ 4) .|. (2 ^ 3)) (2 ^ 3) + print $ naturalAndNot ((2 ^ 129) .|. (2 ^ 65)) (2 ^ 65) + print $ naturalAndNot ((2 ^ 4) .|. (2 ^ 3)) ((2 ^ 65) .|. (2 ^ 3)) + print $ naturalAndNot ((2 ^ 65) .|. (2 ^ 3)) (2 ^ 3) ===================================== testsuite/tests/numeric/should_run/T26230.stdout ===================================== @@ -0,0 +1,4 @@ +16 +680564733841876926926749214863536422912 +16 +36893488147419103232 ===================================== testsuite/tests/numeric/should_run/all.T ===================================== @@ -88,3 +88,4 @@ test('div01', normal, compile_and_run, ['']) test('T24245', normal, compile_and_run, ['']) test('T25653', normal, compile_and_run, ['']) test('T18619', exit_code(1), compile_and_run, ['']) +test('T26230', normal, compile_and_run, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a766286fe759251eceb304c54ba5284… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a766286fe759251eceb304c54ba5284… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] 3 commits: Store `StackTrace` and `StackSnapshot` in `Backtraces`
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: c91e2650 by fendor at 2025-08-13T21:03:43-04:00 Store `StackTrace` and `StackSnapshot` in `Backtraces` Instead of decoding the stack traces when collecting the `Backtraces`, defer this decoding until actually showing the `Backtraces`. This allows users to customise how `Backtraces` are displayed by using a custom implementation of `displayExceptionWithInfo`, overwriting the default implementation for `Backtraces` (`displayBacktraces`). - - - - - dee28cdd by fendor at 2025-08-13T21:03:43-04:00 Allow users to customise the collection of exception annotations Add a global `CollectExceptionAnnotationMechanism` which determines how `ExceptionAnnotation`s are collected upon throwing an `Exception`. This API is exposed via `ghc-experimental`. By overriding how we collect `Backtraces`, we can control how the `Backtraces` are displayed to the user by newtyping `Backtraces` and giving a different instance for `ExceptionAnnotation`. A concrete use-case for this feature is allowing us to experiment with alternative stack decoders, without having to modify `base`, which take additional information from the stack frames. This commit does not modify how `Backtraces` are currently collected or displayed. - - - - - 66024722 by fendor at 2025-08-13T21:03:43-04:00 Expose Backtraces internals from ghc-experimental Additionally, expose the same API `base:Control.Exception.Backtrace` to make it easier to use as a drop-in replacement. - - - - - 32 changed files: - + libraries/ghc-experimental/src/GHC/Exception/Backtrace/Experimental.hs - libraries/ghc-internal/src/GHC/Internal/Exception.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Backtrace.hs - libraries/ghc-internal/src/GHC/Internal/Exception/Backtrace.hs-boot - + libraries/ghc-internal/tests/Makefile - + libraries/ghc-internal/tests/all.T - + libraries/ghc-internal/tests/backtraces/Makefile - + libraries/ghc-internal/tests/backtraces/T14532a.hs - + libraries/ghc-internal/tests/backtraces/T14532a.stdout - + libraries/ghc-internal/tests/backtraces/T14532b.hs - + libraries/ghc-internal/tests/backtraces/T14532b.stdout - + libraries/ghc-internal/tests/backtraces/all.T - testsuite/tests/arrows/should_compile/T21301.stderr - testsuite/tests/deSugar/should_fail/DsStrictFail.stderr - testsuite/tests/deSugar/should_run/T20024.stderr - testsuite/tests/deSugar/should_run/dsrun005.stderr - testsuite/tests/deSugar/should_run/dsrun007.stderr - testsuite/tests/deSugar/should_run/dsrun008.stderr - testsuite/tests/deriving/should_run/T9576.stderr - testsuite/tests/ghci/scripts/Defer02.stderr - testsuite/tests/ghci/scripts/T15325.stderr - testsuite/tests/patsyn/should_run/ghci.stderr - testsuite/tests/quotes/LiftErrMsgDefer.stderr - testsuite/tests/safeHaskell/safeLanguage/SafeLang15.stderr - testsuite/tests/type-data/should_run/T22332a.stderr - testsuite/tests/typecheck/should_run/T10284.stderr - testsuite/tests/typecheck/should_run/T13838.stderr - testsuite/tests/typecheck/should_run/T9497a-run.stderr - testsuite/tests/typecheck/should_run/T9497b-run.stderr - testsuite/tests/typecheck/should_run/T9497c-run.stderr - testsuite/tests/unsatisfiable/T23816.stderr - testsuite/tests/unsatisfiable/UnsatDefer.stderr The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/50842f83f467ff54dd22470559a7af… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/50842f83f467ff54dd22470559a7af… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] Make unexpected LLVM versions a warning rather than an error.
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 50842f83 by Andreas Klebinger at 2025-08-13T21:03:01-04:00 Make unexpected LLVM versions a warning rather than an error. Typically a newer LLVM version *will* work so erroring out if a user uses a newer LLVM version is too aggressive. Fixes #25915 - - - - - 1 changed file: - compiler/GHC/Driver/Errors/Ppr.hs Changes: ===================================== compiler/GHC/Driver/Errors/Ppr.hs ===================================== @@ -281,7 +281,7 @@ instance Diagnostic DriverMessage where ++ llvmVersionStr supportedLlvmVersionLowerBound ++ " and " ++ llvmVersionStr supportedLlvmVersionUpperBound - ++ ") and reinstall GHC to make -fllvm work") + ++ ") and reinstall GHC to ensure -fllvm works") diagnosticReason = \case DriverUnknownMessage m @@ -352,7 +352,7 @@ instance Diagnostic DriverMessage where DriverInstantiationNodeInDependencyGeneration {} -> ErrorWithoutFlag DriverNoConfiguredLLVMToolchain - -> ErrorWithoutFlag + -> WarningWithoutFlag diagnosticHints = \case DriverUnknownMessage m View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/50842f83f467ff54dd22470559a7af7… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/50842f83f467ff54dd22470559a7af7… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] rts: Clarify rationale for undefined atomic wrappers
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: a5e4b7d9 by Ben Gamari at 2025-08-13T21:02:18-04:00 rts: Clarify rationale for undefined atomic wrappers Since c06e3f46d24ef69f3a3d794f5f604cb8c2a40cbc the RTS has declared various atomic operation wrappers defined by ghc-internal as undefined. While the rationale for this isn't clear from the commit message, I believe that this is necessary due to the unregisterised backend. Specifically, the code generator will reference these symbols when compiling RTS Cmm sources. - - - - - 2 changed files: - rts/external-symbols.list.in - rts/rts.cabal Changes: ===================================== rts/external-symbols.list.in ===================================== @@ -54,6 +54,12 @@ ghczminternal_GHCziInternalziWord_W16zh_con_info ghczminternal_GHCziInternalziWord_W32zh_con_info ghczminternal_GHCziInternalziWord_W64zh_con_info ghczminternal_GHCziInternalziStable_StablePtr_con_info +ghczminternal_GHCziInternalziStackziCloneStack_StackSnapshot_closure + +#if 0 +N.B. These symbols are defined in ghc-internal and may be referenced by the +unregisterised code-generator when compiling RTS Cmm sources. +#endif hs_atomic_add8 hs_atomic_add16 hs_atomic_add32 @@ -94,4 +100,3 @@ hs_atomicwrite8 hs_atomicwrite16 hs_atomicwrite32 hs_atomicwrite64 -ghczminternal_GHCziInternalziStackziCloneStack_StackSnapshot_closure ===================================== rts/rts.cabal ===================================== @@ -596,8 +596,7 @@ library -- GHC already does `-undefined dynamic_lookup`, so we just do that for now, but -- we might try to get more precise with `-U` in the future. -- --- Note that the RTS also `-u`s some atomics symbols that *are* defined --- and --- defined within the RTS! It is not immediately clear why this is needed. This --- dates back to c06e3f46d24ef69f3a3d794f5f604cb8c2a40cbc which mentions a build --- failure that it was suggested that this fix, but the precise reasoning is not --- explained. +-- Note that the RTS also `-u`s some symbols for atomic operations that *are* defined +-- by `ghc-internal`. This is needed when the RTS is compiled when the RTS is +-- compiled with, e.g., the unregisterised backend since the code-generator will +-- reference these symbols when compiling atomic primops in Cmm sources. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a5e4b7d9fa56e8357f0b26c5c1a29c1… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/a5e4b7d9fa56e8357f0b26c5c1a29c1… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] 2 commits: Extend record-selector usage ticking to all binds using a record field
by Marge Bot (＠marge-bot) 14 Aug '25

14 Aug '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 62899117 by Florian Ragwitz at 2025-08-13T21:01:34-04:00 Extend record-selector usage ticking to all binds using a record field This extends the previous handling of ticking for RecordWildCards and NamedFieldPuns to all var bindings that involve record selectors. Note that certain patterns such as `Foo{foo = 42}` will currently not tick the `foo` selector, as ticking is triggered by `HsVar`s. Closes #26191. - - - - - b37b3af7 by Florian Ragwitz at 2025-08-13T21:01:34-04:00 Add release notes for 9.16.1 and move description of latest HPC changes there. - - - - - 6 changed files: - compiler/GHC/HsToCore/Ticks.hs - − docs/users_guide/9.14.1-notes.rst - + docs/users_guide/9.16.1-notes.rst - docs/users_guide/release-notes.rst - testsuite/tests/hpc/recsel/recsel.hs - testsuite/tests/hpc/recsel/recsel.stdout Changes: ===================================== compiler/GHC/HsToCore/Ticks.hs ===================================== @@ -251,7 +251,7 @@ addTickLHsBind (L pos (XHsBindsLR bind@(AbsBinds { abs_binds = binds add_rec_sels env = env{ recSelBinds = recSelBinds env `extendVarEnvList` - [ (abe_mono, abe_poly) + [ (abe_mono, unitDVarSet abe_poly) | ABE{ abe_poly, abe_mono } <- abs_exports , RecSelId{} <- [idDetails abe_poly] ] } @@ -270,8 +270,8 @@ addTickLHsBind (L pos (funBind@(FunBind { fun_id = L _ id, fun_matches = matches case tickish of { ProfNotes | inline -> return (L pos funBind); _ -> do -- See Note [Record-selector ticks] - selTick <- recSelTick id - case selTick of { Just tick -> tick_rec_sel tick; _ -> do + selTicks <- recSelTick id + case selTicks of { Just ticks -> tick_rec_sel ticks; _ -> do (fvs, mg) <- getFreeVars $ @@ -303,8 +303,8 @@ addTickLHsBind (L pos (funBind@(FunBind { fun_id = L _ id, fun_matches = matches } } where -- See Note [Record-selector ticks] - tick_rec_sel tick = - pure $ L pos $ funBind { fun_ext = second (tick :) (fun_ext funBind) } + tick_rec_sel ticks = + pure $ L pos $ funBind { fun_ext = second (ticks ++) (fun_ext funBind) } -- Note [Record-selector ticks] @@ -319,9 +319,8 @@ addTickLHsBind (L pos (funBind@(FunBind { fun_id = L _ id, fun_matches = matches -- coverage purposes to improve the developer experience. -- -- This is done by keeping track of which 'Id's are effectively bound to --- record fields (using NamedFieldPuns or RecordWildCards) in 'TickTransEnv's --- 'recSelBinds', and making 'HsVar's corresponding to those fields tick the --- appropriate box when executed. +-- record fields in 'TickTransEnv's 'recSelBinds', and making 'HsVar's +-- corresponding to those fields tick the appropriate box when executed. -- -- To enable that, we also treat 'FunBind's for record selector functions -- specially. We only create a TopLevelBox for the record selector function, @@ -329,11 +328,11 @@ addTickLHsBind (L pos (funBind@(FunBind { fun_id = L _ id, fun_matches = matches -- of ticks for the same record selector, and is done by not recursing into -- the fun_matches match group for record selector functions. -- +-- Note that due to the use of 'HsVar's for ticking, certain patterns such +-- as `Foo{foo = 42}` will not cause the `foo` selector to be ticked. +-- -- This scheme could be extended further in the future, making coverage for --- constructor fields (named or even positional) mean that the field was --- accessed at run-time. For the time being, we only cover NamedFieldPuns and --- RecordWildCards binds to cover most practical use-cases while keeping it --- simple. +-- positional constructor fields mean that the field was accessed at run-time. -- TODO: Revisit this addTickLHsBind (L pos (pat@(PatBind { pat_lhs = lhs @@ -519,7 +518,7 @@ addTickHsExpr :: HsExpr GhcTc -> TM (HsExpr GhcTc) -- See Note [Record-selector ticks] addTickHsExpr e@(HsVar _ (L _ id)) = freeVar id >> recSelTick id >>= pure . maybe e wrap - where wrap tick = XExpr . HsTick tick . noLocA $ e + where wrap = foldr (\tick -> XExpr . HsTick tick . noLocA) e addTickHsExpr e@(HsIPVar {}) = return e addTickHsExpr e@(HsOverLit {}) = return e addTickHsExpr e@(HsOverLabel{}) = return e @@ -1086,7 +1085,7 @@ data TickTransEnv = TTE { fileName :: FastString , blackList :: Set RealSrcSpan , this_mod :: Module , tickishType :: TickishType - , recSelBinds :: IdEnv Id + , recSelBinds :: IdEnv DVarSet } -- deriving Show @@ -1241,11 +1240,12 @@ allocTickBox boxLabel countEntries topOnly pos m tickish <- mkTickish boxLabel countEntries topOnly pos fvs (declPath env) return (this_loc (XExpr $ HsTick tickish $ this_loc e)) -recSelTick :: Id -> TM (Maybe CoreTickish) +recSelTick :: Id -> TM (Maybe [CoreTickish]) recSelTick id = ifDensity TickForCoverage maybe_tick (pure Nothing) where maybe_tick = getEnv >>= - maybe (pure Nothing) tick . (`lookupVarEnv` id) . recSelBinds + maybe (pure Nothing) tick_all . (`lookupVarEnv` id) . recSelBinds + tick_all = fmap (Just . catMaybes) . mapM tick . dVarSetElems tick sel = getState >>= maybe (alloc sel) (pure . Just) . (`lookupVarEnv` sel) . recSelTicks alloc sel = allocATickBox (box sel) False False (getSrcSpan sel) noFVs @@ -1367,7 +1367,7 @@ class CollectBinders a where -- -- See Note [Record-selector ticks]. class CollectFldBinders a where - collectFldBinds :: a -> IdEnv Id + collectFldBinds :: a -> IdEnv DVarSet instance CollectBinders (LocatedA (Pat GhcTc)) where collectBinds = collectPatBinders CollNoDictBinders @@ -1385,41 +1385,37 @@ instance (CollectFldBinders a) => CollectFldBinders [a] where instance (CollectFldBinders e) => CollectFldBinders (GenLocated l e) where collectFldBinds = collectFldBinds . unLoc instance CollectFldBinders (Pat GhcTc) where - collectFldBinds ConPat{ pat_args = RecCon HsRecFields{ rec_flds, rec_dotdot } } = - collectFldBinds rec_flds `plusVarEnv` plusVarEnvList (zipWith fld_bnds [0..] rec_flds) - where n_explicit | Just (L _ (RecFieldsDotDot n)) <- rec_dotdot = n - | otherwise = length rec_flds - fld_bnds n (L _ HsFieldBind{ hfbLHS = L _ FieldOcc{ foLabel = L _ sel } - , hfbRHS = L _ (VarPat _ (L _ var)) - , hfbPun }) - | hfbPun || n >= n_explicit = unitVarEnv var sel - fld_bnds _ _ = emptyVarEnv - collectFldBinds ConPat{ pat_args = PrefixCon pats } = collectFldBinds pats - collectFldBinds ConPat{ pat_args = InfixCon p1 p2 } = collectFldBinds [p1, p2] - collectFldBinds (LazyPat _ pat) = collectFldBinds pat - collectFldBinds (BangPat _ pat) = collectFldBinds pat - collectFldBinds (AsPat _ _ pat) = collectFldBinds pat - collectFldBinds (ViewPat _ _ pat) = collectFldBinds pat - collectFldBinds (ParPat _ pat) = collectFldBinds pat - collectFldBinds (ListPat _ pats) = collectFldBinds pats - collectFldBinds (TuplePat _ pats _) = collectFldBinds pats - collectFldBinds (SumPat _ pats _ _) = collectFldBinds pats - collectFldBinds (SigPat _ pat _) = collectFldBinds pat - collectFldBinds (XPat exp) = collectFldBinds exp - collectFldBinds VarPat{} = emptyVarEnv - collectFldBinds WildPat{} = emptyVarEnv - collectFldBinds OrPat{} = emptyVarEnv - collectFldBinds LitPat{} = emptyVarEnv - collectFldBinds NPat{} = emptyVarEnv - collectFldBinds NPlusKPat{} = emptyVarEnv - collectFldBinds SplicePat{} = emptyVarEnv - collectFldBinds EmbTyPat{} = emptyVarEnv - collectFldBinds InvisPat{} = emptyVarEnv -instance (CollectFldBinders r) => CollectFldBinders (HsFieldBind l r) where - collectFldBinds = collectFldBinds . hfbRHS -instance CollectFldBinders XXPatGhcTc where - collectFldBinds (CoPat _ pat _) = collectFldBinds pat - collectFldBinds (ExpansionPat _ pat) = collectFldBinds pat + collectFldBinds = go emptyDVarSet where + go sels ConPat{ pat_args = RecCon HsRecFields{ rec_flds } } = + plusVarEnvList (map fld_binds rec_flds) + where fld_binds (L _ HsFieldBind{ hfbLHS = L _ FieldOcc{ foLabel = L _ sel } + , hfbRHS = L _ rhs }) + = go (extendDVarSet sels sel) rhs + go sels ConPat{ pat_args = PrefixCon ps } = + plusVarEnvList (map (go sels . unLoc) ps) + go sels ConPat{ pat_args = InfixCon (L _ p1) (L _ p2) } = + go sels p1 `plusVarEnv` go sels p2 + go sels (VarPat _ (L _ var)) | isEmptyDVarSet sels = emptyVarEnv + | otherwise = unitVarEnv var sels + go sels (LazyPat _ (L _ p)) = go sels p + go sels (BangPat _ (L _ p)) = go sels p + go sels (AsPat _ _ (L _ p)) = go sels p + go sels (ViewPat _ _ (L _ p)) = go sels p + go sels (ParPat _ (L _ p)) = go sels p + go sels (SigPat _ (L _ p) _) = go sels p + go sels (SumPat _ (L _ p) _ _) = go sels p + go sels (XPat (CoPat _ p _)) = go sels p + go sels (XPat (ExpansionPat _ p)) = go sels p + go sels (ListPat _ ps) = plusVarEnvList (map (go sels . unLoc) ps) + go sels (TuplePat _ ps _) = plusVarEnvList (map (go sels . unLoc) ps) + go _ WildPat{} = emptyVarEnv + go _ OrPat{} = emptyVarEnv + go _ LitPat{} = emptyVarEnv + go _ NPat{} = emptyVarEnv + go _ NPlusKPat{} = emptyVarEnv + go _ SplicePat{} = emptyVarEnv + go _ EmbTyPat{} = emptyVarEnv + go _ InvisPat{} = emptyVarEnv instance CollectFldBinders (HsLocalBinds GhcTc) where collectFldBinds (HsValBinds _ bnds) = collectFldBinds bnds collectFldBinds HsIPBinds{} = emptyVarEnv @@ -1430,9 +1426,9 @@ instance CollectFldBinders (HsValBinds GhcTc) where instance CollectFldBinders (HsBind GhcTc) where collectFldBinds PatBind{ pat_lhs } = collectFldBinds pat_lhs collectFldBinds (XHsBindsLR AbsBinds{ abs_exports, abs_binds }) = - mkVarEnv [ (abe_poly, sel) + mkVarEnv [ (abe_poly, sels) | ABE{ abe_poly, abe_mono } <- abs_exports - , Just sel <- [lookupVarEnv monos abe_mono] ] + , Just sels <- [lookupVarEnv monos abe_mono] ] where monos = collectFldBinds abs_binds collectFldBinds VarBind{} = emptyVarEnv collectFldBinds FunBind{} = emptyVarEnv ===================================== docs/users_guide/9.14.1-notes.rst deleted ===================================== @@ -1,290 +0,0 @@ -.. _release-9-14-1: - -Version 9.14.1 -============== - -The significant changes to the various parts of the compiler are listed in the -following sections. See the `migration guide -<https://gitlab.haskell.org/ghc/ghc/-/wikis/migration/9.14>`_ on the GHC Wiki -for specific guidance on migrating programs to this release. - -Language -~~~~~~~~ - -* `GHC proposal 493: allow expressions in SPECIALISE pragmas <https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0493-s…>`_ - has been implemented. SPECIALISE pragmas now allow arbitrary expressions such as: :: - - {-# SPECIALISE f @Int False :: Int -> Char #-} - - The ability to specify multiple specialisations in a single SPECIALISE pragma, - with syntax of the form (note the comma between the type signatures): :: - - {-# SPECIALISE g : Int -> Int, Float -> Float #-} - - has been deprecated, and is scheduled to be removed in GHC 9.18. - This deprecation is controlled by the newly introduced ``-Wdeprecated-pragmas`` - flag in ``-Wdefault``. - -* ``-Wincomplete-record-selectors`` is now part of `-Wall`, as specified - by `GHC Proposal 516: add warning for incomplete record selectors <https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0516-i…>`_. - Hence, if a library is compiled with ``-Werror``, compilation may now fail. Solution: fix the library. - Workaround: add ``-Werror=no-incomplete-record-selectors``. - - Note that this warning is at least - as serious as a warning about missing patterns from a function definition, perhaps even - more so, since it is invisible in the source program. - -* The combination of :extension:`ScopedTypeVariables` and :extension:`TypeApplications` - no longer enables type applications in patterns, which now always requires - :extension:`TypeAbstractions`. The warning flag``deprecated-type-abstractions`` - has also been removed from the compiler. - -* :extension:`OverloadedRecordUpdate` now passes the arguments to a ``setField`` function - in the flipped order, as specified by `GHC Proposal 583: HasField redesign <https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0583-h…>`_. - - Previously GHC expected ``setField`` to have this type: :: - - setField :: forall (fld :: Symbol) a r. r -> a -> r - - And that's what GHC expects now: :: - - setField :: forall (fld :: Symbol) a r. a -> r -> r - - That will break the combination of :extension:`OverloadedRecordUpdate` with :extension:`RebindableSyntax`. - -* Multiline strings are now accepted in foreign imports. (#25157) - -* GHC now does a better job at inferring types in calls to ``coerce``: instead of - complaining about ambiguous type variables, GHC will consider that such type - variables are determined by the ``Coercible`` constraints they appear in. - -* With :extension:`LinearTypes` record fields can now be non-linear. This means that - the following record declaration is now valid: - - :: - - data Record = Rec { x %'Many :: Int, y :: Char } - - This causes the constructor to have type ``Rec :: Int %'Many -> Char %1 -> Record``. - -* The :extension:`ExplicitNamespaces` extension now allows the ``data`` - namespace specifier in import and export lists. - -* The ``-Wdata-kinds-tc`` warning has been deprecated, and the use of promoted - data types in kinds is now an error (rather than a warning) unless the - :extension:`DataKinds` extension is enabled. For example, the following code - will be rejected unless :extension:`DataKinds` is on: - - import Data.Kind (Type) - import GHC.TypeNats (Nat) - - -- Nat shouldn't be allowed here without DataKinds - data Vec :: Nat -> Type -> Type - - (The ``-Wdata-kinds-tc`` warning was introduced in GHC 9.10 as part of a fix - for an accidental oversight in which programs like the one above were - mistakenly accepted without the use of :extension:`DataKinds`.) - -* The :extension:`MonadComprehensions` extension now implies :extension:`ParallelListComp` as was originally intended (see `Monad Comprehensions <https://ghc.gitlab.haskell.org/ghc/doc/users_guide/exts/monad_comprehension…>`_). - -* In accordance with `GHC Proposal #281 <https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0281-v…>`_, - section 4.7 "Data constructors", the :extension:`RequiredTypeArguments` - extension now allows visible forall in types of data constructors - (:ghc-ticket:`25127`). The following declaration is now accepted by GHC: - - :: - - data T a where - Typed :: forall a -> a -> T a - - See :ref:`visible-forall-in-gadts` for details. - -Compiler -~~~~~~~~ - -- An improved error message is introduced to refer users to the heap-controlling flags of the RTS when there is a heap overflow during compilation. (#25198) - -- The kind checker now does a better job of finding type family instances for - use in the kinds of other declarations in the same module. This fixes a number - of tickets: - :ghc-ticket:`12088`, :ghc-ticket:`12239`, :ghc-ticket:`14668`, :ghc-ticket:`15561`, - :ghc-ticket:`16410`, :ghc-ticket:`16448`, :ghc-ticket:`16693`, :ghc-ticket:`19611`, - :ghc-ticket:`20875`, :ghc-ticket:`21172`, :ghc-ticket:`22257`, :ghc-ticket:`25238`, - :ghc-ticket:`25834`. - -- The compiler no longer accepts invalid ``type`` namespace specifiers in - subordinate import lists (:ghc-ticket:`22581`). - -- A new flag, :ghc-flag:`-Wuseless-specialisations`, controls warnings emitted when GHC - determines that a SPECIALISE pragma would have no effect. - -- A new flag, :ghc-flag:`-Wrule-lhs-equalities`, controls warnings emitted for RULES - whose left-hand side attempts to quantify over equality constraints that - previous GHC versions accepted quantifying over. GHC will now drop such RULES, - emitting a warning message controlled by this flag. - - This warning is intended to give visibility to the fact that the RULES that - previous GHC versions generated in such circumstances could never fire. - -- A new flag, :ghc-flag:`-Wunusable-unpack-pragmas`, controls warnings emitted - when GHC is unable to unpack a data constructor field annotated by the - ``{-# UNPACK #-}`` pragma. - - Previous GHC versions issued this warning unconditionally. Now it is possible - to disable it with ``-Wno-unusable-unpack-pragmas`` or turn it into an error - with ``-Werror=unusable-unpack-pragmas``. - -- Introduce a new warning :ghc-flag:`-Wpattern-namespace-specifier` to detect - uses of the now deprecated ``pattern`` namespace specifier in import/export - lists. See `GHC Proposal #581, section 2.3 <https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0581-n…>`_. - -- Code coverage (:ghc-flag:`-fhpc`) now treats uses of record fields via - :extension:`RecordWildCards` or :extension:`NamedFieldPuns` as if the fields - were accessed using the generated record selector functions, marking the fields - as covered in coverage reports (:ghc-ticket:`17834`). - -- SIMD support in the X86 native code generator has been extended with 128-bit - integer operations. Also, ``shuffleFloatX4#`` and ``shuffleDoubleX2#`` no longer - require ``-mavx``. - -- JSON diagnostics produced with (:ghc-flag:`-fdiagnostics-as-json`) now - include the `rendered` diagnostics message, in the exact same format as what - GHC would have produced without -fdiagnostics-as-json (including ANSI escape - sequences). - -GHCi -~~~~ - -- :ghci-cmd:`:info` now outputs type declarations with @-binders that are - considered semantically significant. See the documentation for :ghci-cmd:`:info` - itself for a more detailed explanation. - -- GHCi errors and warnings now have their own numeric error codes that are - displayed alongside the error. - -Runtime system -~~~~~~~~~~~~~~ - -- Add new runtime flag :rts-flag:`--optimistic-linking` which instructs the - runtime linker to continue in the presence of unknown symbols. By default this - flag is not passed, preserving previous behavior. - -Cmm -~~~ - -``base`` library -~~~~~~~~~~~~~~~~ - -``ghc-prim`` library -~~~~~~~~~~~~~~~~~~~~ - -``ghc`` library -~~~~~~~~~~~~~~~ - -* The `UnknownDiagnostic` constructor now takes an additional type argument - for the type of hints corresponding to the diagnostic, and an additional - value-level argument used for existential wrapping of the hints of the inner - diagnostic. - -* Changes to the HPT and HUG interface: - - - `addToHpt` and `addListToHPT` were moved from `GHC.Unit.Home.ModInfo` to `GHC.Unit.Home.PackageTable` and deprecated in favour of `addHomeModInfoToHpt` and `addHomeModInfosToHpt`. - - `UnitEnvGraph` and operations `unitEnv_lookup_maybe`, `unitEnv_foldWithKey, `unitEnv_singleton`, `unitEnv_adjust`, `unitEnv_insert`, `unitEnv_new` were moved from `GHC.Unit.Env` to `GHC.Unit.Home.Graph`. - - The HomePackageTable (HPT) is now exported from `GHC.Unit.Home.PackageTable`, - and is now backed by an IORef to avoid by construction very bad memory leaks. - This means the API to the HPT now is for the most part in IO. For instance, - `emptyHomePackageTable` and `addHomeModInfoToHpt` are now in IO. - - `mkHomeUnitEnv` was moved to `GHC.Unit.Home.PackageTable`, and now takes two - extra explicit arguments. To restore previous behaviour, pass `emptyUnitState` - and `Nothing` as the first two arguments additionally. - - `hugElts` was removed. Users should prefer `allUnits` to get the keys of the - HUG (the typical use case), or `traverse` or `unitEnv_foldWithKey` in other - cases. - -* Changes to `Language.Haskell.Syntax.Expr` - - - The `ParStmtBlock` list argument of the `ParStmt` constructor of `StmtLR` is now `NonEmpty`. - -* As part of the implementation of `GHC proposal 493 <https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0493-s…>`_, - the `SpecSig` constructor of `Sig` has been deprecated. It is replaced by - the constructor `SpecSigE` which supports expressions at the head, rather than - a lone variable. - -``ghc-heap`` library -~~~~~~~~~~~~~~~~~~~~ - -* The functions `getClosureInfoTbl_maybe`, `getClosureInfoTbl`, - `getClosurePtrArgs` and `getClosurePtrArgs_maybe` have been added to allow - reading of the relevant Closure attributes without reliance on incomplete - selectors. - -``ghc-experimental`` library -~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - -- ``ghc-experimental`` now exposes ``GHC.RTS.Flags`` and ``GHC.Stats`` as - ``GHC.RTS.Flags.Experimental`` and ``GHC.Stats.Experimental``. These are - *also* exposed in ``base``, however the ``base`` versions will be deprecated as - part of the split base project. See `CLC proposal 289 - <https://github.com/haskell/core-libraries-committee/issues/289>`__. - Downstream consumers of these flags are encouraged to migrate to the - ``ghc-experimental`` versions. - - - -``template-haskell`` library -~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - -- As part of the implementation of `GHC proposal 493 <https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0493-s…>`_, - the ``SpecialiseP`` constructor of the Template Haskell ``Pragma`` type, as - well as the helpers ``pragSpecD`` and ``pragSpecInlD``, have been deprecated. - - They are replaced, respectively, by ``SpecialiseEP``, ``pragSpecED`` and - ``pragSpecInlED``. - -Included libraries -~~~~~~~~~~~~~~~~~~ - -The package database provided with this distribution also contains a number of -packages other than GHC itself. See the changelogs provided with these packages -for further change information. - -.. ghc-package-list:: - - libraries/array/array.cabal: Dependency of ``ghc`` library - libraries/base/base.cabal: Core library - libraries/binary/binary.cabal: Dependency of ``ghc`` library - libraries/bytestring/bytestring.cabal: Dependency of ``ghc`` library - libraries/Cabal/Cabal/Cabal.cabal: Dependency of ``ghc-pkg`` utility - libraries/Cabal/Cabal-syntax/Cabal-syntax.cabal: Dependency of ``ghc-pkg`` utility - libraries/containers/containers/containers.cabal: Dependency of ``ghc`` library - libraries/deepseq/deepseq.cabal: Dependency of ``ghc`` library - libraries/directory/directory.cabal: Dependency of ``ghc`` library - libraries/exceptions/exceptions.cabal: Dependency of ``ghc`` and ``haskeline`` library - libraries/filepath/filepath.cabal: Dependency of ``ghc`` library - compiler/ghc.cabal: The compiler itself - libraries/ghci/ghci.cabal: The REPL interface - libraries/ghc-boot/ghc-boot.cabal: Internal compiler library - libraries/ghc-boot-th/ghc-boot-th.cabal: Internal compiler library - libraries/ghc-compact/ghc-compact.cabal: Core library - libraries/ghc-heap/ghc-heap.cabal: GHC heap-walking library - libraries/ghc-prim/ghc-prim.cabal: Core library - utils/haddock/haddock-api/haddock-api.cabal: Dependency of ``haddock`` executable - utils/haddock/haddock-library/haddock-library.cabal: Dependency of ``haddock`` executable - libraries/haskeline/haskeline.cabal: Dependency of ``ghci`` executable - libraries/hpc/hpc.cabal: Dependency of ``hpc`` executable - libraries/integer-gmp/integer-gmp.cabal: Core library - libraries/mtl/mtl.cabal: Dependency of ``Cabal`` library - libraries/parsec/parsec.cabal: Dependency of ``Cabal`` library - libraries/pretty/pretty.cabal: Dependency of ``ghc`` library - libraries/process/process.cabal: Dependency of ``ghc`` library - libraries/stm/stm.cabal: Dependency of ``haskeline`` library - libraries/template-haskell/template-haskell.cabal: Core library - libraries/terminfo/terminfo.cabal: Dependency of ``haskeline`` library - libraries/text/text.cabal: Dependency of ``Cabal`` library - libraries/time/time.cabal: Dependency of ``ghc`` library - libraries/transformers/transformers.cabal: Dependency of ``ghc`` library - libraries/unix/unix.cabal: Dependency of ``ghc`` library - libraries/Win32/Win32.cabal: Dependency of ``ghc`` library - libraries/xhtml/xhtml.cabal: Dependency of ``haddock`` executable - libraries/os-string/os-string.cabal: Dependency of ``filepath`` library - libraries/file-io/file-io.cabal: Dependency of ``directory`` library ===================================== docs/users_guide/9.16.1-notes.rst ===================================== @@ -0,0 +1,99 @@ +.. _release-9-16-1: + +Version 9.16.1 +============== + +The significant changes to the various parts of the compiler are listed in the +following sections. See the `migration guide +<https://gitlab.haskell.org/ghc/ghc/-/wikis/migration/9.16>`_ on the GHC Wiki +for specific guidance on migrating programs to this release. + +Language +~~~~~~~~ + +Compiler +~~~~~~~~ + +- Code coverage's (:ghc-flag:`-fhpc`) treatment of record fields now extends + beyond record fields accessed via :extension:`RecordWildCards` and + :extension:`NamedFieldPuns`, and also handles access to nested record fields. + That is, in a pattern such as ``Foo{bar = Bar{baz = b}}`` both ``bar`` and + ``baz`` will now be marked as covered if ``b`` is evaluated. Note that this + currently only works when record fields (or values contained within them) are + bound to variables. The very similar pattern ``Foo{bar = Bar{baz = 42}}`` + will will not yet mark ``bar`` or ``baz`` as covered. + +GHCi +~~~~ + +Runtime system +~~~~~~~~~~~~~~ + +Cmm +~~~ + +``base`` library +~~~~~~~~~~~~~~~~ + +``ghc-prim`` library +~~~~~~~~~~~~~~~~~~~~ + +``ghc`` library +~~~~~~~~~~~~~~~ + +``ghc-heap`` library +~~~~~~~~~~~~~~~~~~~~ + +``ghc-experimental`` library +~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +``template-haskell`` library +~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Included libraries +~~~~~~~~~~~~~~~~~~ + +The package database provided with this distribution also contains a number of +packages other than GHC itself. See the changelogs provided with these packages +for further change information. + +.. ghc-package-list:: + + libraries/array/array.cabal: Dependency of ``ghc`` library + libraries/base/base.cabal: Core library + libraries/binary/binary.cabal: Dependency of ``ghc`` library + libraries/bytestring/bytestring.cabal: Dependency of ``ghc`` library + libraries/Cabal/Cabal/Cabal.cabal: Dependency of ``ghc-pkg`` utility + libraries/Cabal/Cabal-syntax/Cabal-syntax.cabal: Dependency of ``ghc-pkg`` utility + libraries/containers/containers/containers.cabal: Dependency of ``ghc`` library + libraries/deepseq/deepseq.cabal: Dependency of ``ghc`` library + libraries/directory/directory.cabal: Dependency of ``ghc`` library + libraries/exceptions/exceptions.cabal: Dependency of ``ghc`` and ``haskeline`` library + libraries/filepath/filepath.cabal: Dependency of ``ghc`` library + compiler/ghc.cabal: The compiler itself + libraries/ghci/ghci.cabal: The REPL interface + libraries/ghc-boot/ghc-boot.cabal: Internal compiler library + libraries/ghc-boot-th/ghc-boot-th.cabal: Internal compiler library + libraries/ghc-compact/ghc-compact.cabal: Core library + libraries/ghc-heap/ghc-heap.cabal: GHC heap-walking library + libraries/ghc-prim/ghc-prim.cabal: Core library + utils/haddock/haddock-api/haddock-api.cabal: Dependency of ``haddock`` executable + utils/haddock/haddock-library/haddock-library.cabal: Dependency of ``haddock`` executable + libraries/haskeline/haskeline.cabal: Dependency of ``ghci`` executable + libraries/hpc/hpc.cabal: Dependency of ``hpc`` executable + libraries/integer-gmp/integer-gmp.cabal: Core library + libraries/mtl/mtl.cabal: Dependency of ``Cabal`` library + libraries/parsec/parsec.cabal: Dependency of ``Cabal`` library + libraries/pretty/pretty.cabal: Dependency of ``ghc`` library + libraries/process/process.cabal: Dependency of ``ghc`` library + libraries/stm/stm.cabal: Dependency of ``haskeline`` library + libraries/template-haskell/template-haskell.cabal: Core library + libraries/terminfo/terminfo.cabal: Dependency of ``haskeline`` library + libraries/text/text.cabal: Dependency of ``Cabal`` library + libraries/time/time.cabal: Dependency of ``ghc`` library + libraries/transformers/transformers.cabal: Dependency of ``ghc`` library + libraries/unix/unix.cabal: Dependency of ``ghc`` library + libraries/Win32/Win32.cabal: Dependency of ``ghc`` library + libraries/xhtml/xhtml.cabal: Dependency of ``haddock`` executable + libraries/os-string/os-string.cabal: Dependency of ``filepath`` library + libraries/file-io/file-io.cabal: Dependency of ``directory`` library ===================================== docs/users_guide/release-notes.rst ===================================== @@ -4,4 +4,4 @@ Release notes .. toctree:: :maxdepth: 1 - 9.14.1-notes + 9.16.1-notes ===================================== testsuite/tests/hpc/recsel/recsel.hs ===================================== @@ -10,7 +10,8 @@ import Trace.Hpc.Tix import Trace.Hpc.Reflect data Foo = Foo { fooA, fooB, fooC, fooD, fooE, fooF, fooG, fooH, fooI - , fooJ, fooK, fooL, fooM, fooN, fooO :: Int } + , fooJ, fooK, fooL, fooM, fooN, fooO :: Int + , fooP, fooQ :: Maybe Int } data Bar = Bar { barFoo :: Foo } fAB Foo{..} = fooA + fooB @@ -35,14 +36,17 @@ fL = runIdentity . runKleisli (proc f -> do fM f | Foo{..} <- f = fooM fN f = fooN f fO = runIdentity . runKleisli (proc Foo{..} -> returnA -< fooO) +fP Foo{fooP = Just x} = x +fP _ = 0 +fQ Foo{fooQ = Just 42} = 1 recSel (n, TopLevelBox [s]) | any (`isPrefixOf` s) ["foo", "bar"] = Just (n, s) recSel _ = Nothing main = do - let foo = Foo 42 23 0 1 2 3 4 5 6 7 0xaffe 9 10 11 12 + let foo = Foo 42 23 0 1 2 3 4 5 6 7 0xaffe 9 10 11 12 (Just 13) (Just 42) mapM_ (print . ($ foo)) - [fAB, fC, fD False, fE . Bar, fF, fG, fH, fI, fJ, fK, fL, fM, fN, fO] + [fAB, fC, fD False, fE . Bar, fF, fG, fH, fI, fJ, fK, fL, fM, fN, fO, fP, fQ] (Mix _ _ _ _ mixs) <- readMix [".hpc"] (Left "Main") let sels = mapMaybe recSel . zip [0..] $ map snd mixs (Tix [TixModule "Main" _ _ tix]) <- examineTix ===================================== testsuite/tests/hpc/recsel/recsel.stdout ===================================== @@ -12,13 +12,15 @@ 10 11 12 -(0,"barFoo") +13 +1 +(1,"barFoo") (1,"fooA") (1,"fooB") (1,"fooC") (0,"fooD") (1,"fooE") -(0,"fooF") +(1,"fooF") (1,"fooG") (1,"fooH") (1,"fooI") @@ -28,3 +30,5 @@ (1,"fooM") (1,"fooN") (1,"fooO") +(1,"fooP") +(0,"fooQ") View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0434af813cd5aa629ef7566cd267d7… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0434af813cd5aa629ef7566cd267d7… You're receiving this email because of your account on gitlab.haskell.org.

1 0