July 2025 - ghc-commits

[Git][ghc/ghc][wip/teo/allow-newer-ghc-paths] cabal.project-reinstall: use correct package to specify in-ghc-tree flag for haddock
by Teo Camarasu (＠teo) 08 Jul '25

08 Jul '25

Teo Camarasu pushed to branch wip/teo/allow-newer-ghc-paths at Glasgow Haskell Compiler / GHC Commits: 2d4f262e by Teo Camarasu at 2025-07-08T18:49:59+01:00 cabal.project-reinstall: use correct package to specify in-ghc-tree flag for haddock - - - - - 1 changed file: - cabal.project-reinstall Changes: ===================================== cabal.project-reinstall ===================================== @@ -58,7 +58,7 @@ packages: ./compiler constraints: ghc +internal-interpreter +dynamic-system-linke, ghc-bin +internal-interpreter +threaded, ghci +internal-interpreter, - haddock +in-ghc-tree, + haddock-api +in-ghc-tree, any.array installed, any.base installed, any.deepseq installed, View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2d4f262e40e27f8f6e06622667570c6… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/2d4f262e40e27f8f6e06622667570c6… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] Fix typo in using.rst
by Marge Bot (＠marge-bot) 08 Jul '25

08 Jul '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 03e047f9 by Simon Hengel at 2025-07-08T13:25:25-04:00 Fix typo in using.rst - - - - - 1 changed file: - docs/users_guide/using.rst Changes: ===================================== docs/users_guide/using.rst ===================================== @@ -548,7 +548,7 @@ The available mode flags are: Print ``YES`` if GHC was compiled to use symbols with leading underscores in object files, ``NO`` otherwise. - This is usually atarget platform dependent. + This is usually target platform dependent. .. ghc-flag:: --print-libdir :shortdesc: display GHC library directory View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/03e047f9ed3227ba697770454c483ed… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/03e047f9ed3227ba697770454c483ed… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] Implement user-defined allocation limit handlers
by Marge Bot (＠marge-bot) 08 Jul '25

08 Jul '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: ea2c6673 by Teo Camarasu at 2025-07-08T13:24:43-04:00 Implement user-defined allocation limit handlers Allocation Limits allow killing a thread if they allocate more than a user-specified limit. We extend this feature to allow more versatile behaviour. - We allow not killing the thread if the limit is exceeded. - We allow setting a custom handler to be called when the limit is exceeded. User-specified allocation limit handlers run in a fresh thread and are passed the ThreadId of the thread that exceeded its limit. We introduce utility functions for getting and setting the allocation limits of other threads, so that users can reset the limit of a thread from a handler. Both of these are somewhat coarse-grained as we are unaware of the allocations in the current nursery chunk. We provide several examples of usages in testsuite/tests/rts/T22859.hs Resolves #22859 - - - - - 28 changed files: - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - libraries/ghc-experimental/ghc-experimental.cabal.in - + libraries/ghc-experimental/src/System/Mem/Experimental.hs - libraries/ghc-internal/ghc-internal.cabal.in - + libraries/ghc-internal/src/GHC/Internal/AllocationLimitHandler.hs - rts/Prelude.h - rts/PrimOps.cmm - rts/RtsStartup.c - rts/RtsSymbols.c - rts/Schedule.c - rts/external-symbols.list.in - rts/include/rts/storage/GC.h - rts/include/rts/storage/TSO.h - rts/include/stg/MiscClosures.h - testsuite/tests/backpack/cabal/bkpcabal08/bkpcabal08.stdout - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 - testsuite/tests/interface-stability/ghc-experimental-exports.stdout - testsuite/tests/interface-stability/ghc-experimental-exports.stdout-mingw32 - testsuite/tests/interface-stability/ghc-prim-exports.stdout - testsuite/tests/interface-stability/ghc-prim-exports.stdout-mingw32 - + testsuite/tests/rts/T22859.hs - + testsuite/tests/rts/T22859.stderr - testsuite/tests/rts/all.T Changes: ===================================== compiler/GHC/Builtin/primops.txt.pp ===================================== @@ -4065,6 +4065,15 @@ primop SetThreadAllocationCounter "setThreadAllocationCounter#" GenPrimOp effect = ReadWriteEffect out_of_line = True +primop SetOtherThreadAllocationCounter "setOtherThreadAllocationCounter#" GenPrimOp + Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld + { Sets the allocation counter for the another thread to the given value. + This doesn't take allocations into the current nursery chunk into account. + Therefore it is only accurate if the other thread is not currently running. } + with + effect = ReadWriteEffect + out_of_line = True + primtype StackSnapshot# { Haskell representation of a @StgStack*@ that was created (cloned) with a function in "GHC.Stack.CloneStack". Please check the ===================================== compiler/GHC/StgToCmm/Prim.hs ===================================== @@ -1775,6 +1775,7 @@ emitPrimOp cfg primop = TraceEventBinaryOp -> alwaysExternal TraceMarkerOp -> alwaysExternal SetThreadAllocationCounter -> alwaysExternal + SetOtherThreadAllocationCounter -> alwaysExternal KeepAliveOp -> alwaysExternal where ===================================== compiler/GHC/StgToJS/Prim.hs ===================================== @@ -1173,6 +1173,7 @@ genPrim prof bound ty op = case op of WhereFromOp -> unhandledPrimop op -- should be easily implementable with o.f.n SetThreadAllocationCounter -> unhandledPrimop op + SetOtherThreadAllocationCounter -> unhandledPrimop op ------------------------------- Vector ----------------------------------------- -- For now, vectors are unsupported on the JS backend. Simply put, they do not ===================================== libraries/ghc-experimental/ghc-experimental.cabal.in ===================================== @@ -38,6 +38,7 @@ library GHC.RTS.Flags.Experimental GHC.Stats.Experimental Prelude.Experimental + System.Mem.Experimental if arch(wasm32) exposed-modules: GHC.Wasm.Prim other-extensions: ===================================== libraries/ghc-experimental/src/System/Mem/Experimental.hs ===================================== @@ -0,0 +1,10 @@ +module System.Mem.Experimental + ( setGlobalAllocationLimitHandler + , AllocationLimitKillBehaviour(..) + , getAllocationCounterFor + , setAllocationCounterFor + , enableAllocationLimitFor + , disableAllocationLimitFor + ) + where +import GHC.Internal.AllocationLimitHandler ===================================== libraries/ghc-internal/ghc-internal.cabal.in ===================================== @@ -122,6 +122,7 @@ Library rts == 1.0.* exposed-modules: + GHC.Internal.AllocationLimitHandler GHC.Internal.ClosureTypes GHC.Internal.Control.Arrow GHC.Internal.Control.Category ===================================== libraries/ghc-internal/src/GHC/Internal/AllocationLimitHandler.hs ===================================== @@ -0,0 +1,117 @@ +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE UnboxedTuples #-} +{-# LANGUAGE UnliftedFFITypes #-} +{-# LANGUAGE GHCForeignImportPrim #-} +{-# OPTIONS_HADDOCK not-home #-} +module GHC.Internal.AllocationLimitHandler + ( runAllocationLimitHandler + , setGlobalAllocationLimitHandler + , AllocationLimitKillBehaviour(..) + , getAllocationCounterFor + , setAllocationCounterFor + , enableAllocationLimitFor + , disableAllocationLimitFor + ) + where +import GHC.Internal.Base +import GHC.Internal.Conc.Sync (ThreadId(..)) +import GHC.Internal.Data.IORef (IORef, readIORef, writeIORef, newIORef) +import GHC.Internal.Foreign.C.Types +import GHC.Internal.IO (unsafePerformIO) +import GHC.Internal.Int (Int64(..)) + + +{-# NOINLINE allocationLimitHandler #-} +allocationLimitHandler :: IORef (ThreadId -> IO ()) +allocationLimitHandler = unsafePerformIO (newIORef defaultHandler) + +defaultHandler :: ThreadId -> IO () +defaultHandler _ = pure () + +foreign import ccall "setAllocLimitKill" setAllocLimitKill :: CBool -> CBool -> IO () + +runAllocationLimitHandler :: ThreadId# -> IO () +runAllocationLimitHandler tid = do + hook <- getAllocationLimitHandler + hook $ ThreadId tid + +getAllocationLimitHandler :: IO (ThreadId -> IO ()) +getAllocationLimitHandler = readIORef allocationLimitHandler + +data AllocationLimitKillBehaviour = + KillOnAllocationLimit + -- ^ Throw a @AllocationLimitExceeded@ async exception to the thread when the + -- allocation limit is exceeded. + | DontKillOnAllocationLimit + -- ^ Do not throw an exception when the allocation limit is exceeded. + +-- | Define the behaviour for handling allocation limits. +-- The default behaviour is to throw an @AllocationLimitExceeded@ async exception to the thread. +-- This can be overriden using @AllocationLimitKillBehaviour@. +-- +-- We can set a user-specified handler, which can be run in addition to +-- or in place of the exception. +-- This allows for instance logging on the allocation limit being exceeded, +-- or dynamically determining whether to terminate the thread. +-- The handler is not guaranteed to run before the thread is terminated or restarted. +-- +-- Note: that if you don't terminate the thread, then the allocation limit gets +-- removed. +-- If you wish to keep the allocation limit you will have to reset it using +-- @setAllocationCounter@ and @enableAllocationLimit@. +setGlobalAllocationLimitHandler :: AllocationLimitKillBehaviour -> Maybe (ThreadId -> IO ()) -> IO () +setGlobalAllocationLimitHandler killBehaviour mHandler = do + shouldRunHandler <- case mHandler of + Just hook -> do + writeIORef allocationLimitHandler hook + pure 1 + Nothing -> do + writeIORef allocationLimitHandler defaultHandler + pure 0 + let shouldKill = + case killBehaviour of + KillOnAllocationLimit -> 1 + DontKillOnAllocationLimit -> 0 + setAllocLimitKill shouldKill shouldRunHandler + +-- | Retrieves the allocation counter for the another thread. +foreign import prim "stg_getOtherThreadAllocationCounterzh" getOtherThreadAllocationCounter# + :: ThreadId# + -> State# RealWorld + -> (# State# RealWorld, Int64# #) + +-- | Get the allocation counter for a different thread. +-- +-- Note: this doesn't take the current nursery chunk into account. +-- If the thread is running then it may underestimate allocations by the size of a nursery thread. +getAllocationCounterFor :: ThreadId -> IO Int64 +getAllocationCounterFor (ThreadId t#) = IO $ \s -> + case getOtherThreadAllocationCounter# t# s of (# s', i# #) -> (# s', I64# i# #) + +-- | Set the allocation counter for a different thread. +-- This can be combined with 'enableAllocationLimitFor' to enable allocation limits for another thread. +-- You may wish to do this during a user-specified allocation limit handler. +-- +-- Note: this doesn't take the current nursery chunk into account. +-- If the thread is running then it may overestimate allocations by the size of a nursery thread, +-- and trigger the limit sooner than expected. +setAllocationCounterFor :: Int64 -> ThreadId -> IO () +setAllocationCounterFor (I64# i#) (ThreadId t#) = IO $ \s -> + case setOtherThreadAllocationCounter# i# t# s of s' -> (# s', () #) + + +-- | Enable allocation limit processing the thread @t@. +enableAllocationLimitFor :: ThreadId -> IO () +enableAllocationLimitFor (ThreadId t) = do + rts_enableThreadAllocationLimit t + +-- | Disable allocation limit processing the thread @t@. +disableAllocationLimitFor :: ThreadId -> IO () +disableAllocationLimitFor (ThreadId t) = do + rts_disableThreadAllocationLimit t + +foreign import ccall unsafe "rts_enableThreadAllocationLimit" + rts_enableThreadAllocationLimit :: ThreadId# -> IO () + +foreign import ccall unsafe "rts_disableThreadAllocationLimit" + rts_disableThreadAllocationLimit :: ThreadId# -> IO () ===================================== rts/Prelude.h ===================================== @@ -67,6 +67,7 @@ PRELUDE_CLOSURE(ghczminternal_GHCziInternalziEventziWindows_processRemoteComplet PRELUDE_CLOSURE(ghczminternal_GHCziInternalziTopHandler_flushStdHandles_closure); PRELUDE_CLOSURE(ghczminternal_GHCziInternalziTopHandler_runMainIO_closure); +PRELUDE_CLOSURE(ghczminternal_GHCziInternalziAllocationLimitHandler_runAllocationLimitHandler_closure); PRELUDE_INFO(ghczminternal_GHCziInternalziCString_unpackCStringzh_info); PRELUDE_INFO(ghczminternal_GHCziInternalziTypes_Czh_con_info); @@ -102,6 +103,7 @@ PRELUDE_INFO(ghczminternal_GHCziInternalziStable_StablePtr_con_info); #if defined(mingw32_HOST_OS) #define processRemoteCompletion_closure DLL_IMPORT_DATA_REF(ghczminternal_GHCziInternalziEventziWindows_processRemoteCompletion_closure) #endif +#define runAllocationLimitHandler_closure DLL_IMPORT_DATA_REF(ghczminternal_GHCziInternalziAllocationLimitHandler_runAllocationLimitHandler_closure) #define flushStdHandles_closure DLL_IMPORT_DATA_REF(ghczminternal_GHCziInternalziTopHandler_flushStdHandles_closure) #define runMainIO_closure DLL_IMPORT_DATA_REF(ghczminternal_GHCziInternalziTopHandler_runMainIO_closure) ===================================== rts/PrimOps.cmm ===================================== @@ -2838,6 +2838,11 @@ stg_getThreadAllocationCounterzh () return (StgTSO_alloc_limit(CurrentTSO) - TO_I64(offset)); } +stg_getOtherThreadAllocationCounterzh ( gcptr t ) +{ + return (StgTSO_alloc_limit(t)); +} + stg_setThreadAllocationCounterzh ( I64 counter ) { // Allocation in the current block will be subtracted by @@ -2850,6 +2855,12 @@ stg_setThreadAllocationCounterzh ( I64 counter ) return (); } +stg_setOtherThreadAllocationCounterzh ( I64 counter, gcptr t ) +{ + StgTSO_alloc_limit(t) = counter; + return (); +} + #define KEEP_ALIVE_FRAME_FIELDS(w_,p_,info_ptr,p1,p2,c) \ w_ info_ptr, \ ===================================== rts/RtsStartup.c ===================================== @@ -224,6 +224,7 @@ static void initBuiltinGcRoots(void) * GHC.Core.Make.mkExceptionId. */ getStablePtr((StgPtr)absentSumFieldError_closure); + getStablePtr((StgPtr)runAllocationLimitHandler_closure); } void ===================================== rts/RtsSymbols.c ===================================== @@ -748,6 +748,7 @@ extern char **environ; SymI_HasProto(rts_enableThreadAllocationLimit) \ SymI_HasProto(rts_disableThreadAllocationLimit) \ SymI_HasProto(rts_setMainThread) \ + SymI_HasProto(setAllocLimitKill) \ SymI_HasProto(setProgArgv) \ SymI_HasProto(startupHaskell) \ SymI_HasProto(shutdownHaskell) \ @@ -922,7 +923,9 @@ extern char **environ; SymI_HasDataProto(stg_traceMarkerzh) \ SymI_HasDataProto(stg_traceBinaryEventzh) \ SymI_HasDataProto(stg_getThreadAllocationCounterzh) \ + SymI_HasDataProto(stg_getOtherThreadAllocationCounterzh) \ SymI_HasDataProto(stg_setThreadAllocationCounterzh) \ + SymI_HasDataProto(stg_setOtherThreadAllocationCounterzh) \ SymI_HasProto(getMonotonicNSec) \ SymI_HasProto(lockFile) \ SymI_HasProto(unlockFile) \ ===================================== rts/Schedule.c ===================================== @@ -41,6 +41,7 @@ #include "Threads.h" #include "Timer.h" #include "ThreadPaused.h" +#include "ThreadLabels.h" #include "Messages.h" #include "StablePtr.h" #include "StableName.h" @@ -94,6 +95,10 @@ StgWord recent_activity = ACTIVITY_YES; */ StgWord sched_state = SCHED_RUNNING; + +bool allocLimitKill = true; +bool allocLimitRunHook = false; + /* * This mutex protects most of the global scheduler data in * the THREADED_RTS runtime. @@ -1125,19 +1130,36 @@ schedulePostRunThread (Capability *cap, StgTSO *t) } } - // - // If the current thread's allocation limit has run out, send it - // the AllocationLimitExceeded exception. + // Handle the current thread's allocation limit running out, if (PK_Int64((W_*)&(t->alloc_limit)) < 0 && (t->flags & TSO_ALLOC_LIMIT)) { - // Use a throwToSelf rather than a throwToSingleThreaded, because - // it correctly handles the case where the thread is currently - // inside mask. Also the thread might be blocked (e.g. on an - // MVar), and throwToSingleThreaded doesn't unblock it - // correctly in that case. - throwToSelf(cap, t, allocationLimitExceeded_closure); - ASSIGN_Int64((W_*)&(t->alloc_limit), - (StgInt64)RtsFlags.GcFlags.allocLimitGrace * BLOCK_SIZE); + if(allocLimitKill) { + // Throw the AllocationLimitExceeded exception. + // Use a throwToSelf rather than a throwToSingleThreaded, because + // it correctly handles the case where the thread is currently + // inside mask. Also the thread might be blocked (e.g. on an + // MVar), and throwToSingleThreaded doesn't unblock it + // correctly in that case. + throwToSelf(cap, t, allocationLimitExceeded_closure); + ASSIGN_Int64((W_*)&(t->alloc_limit), + (StgInt64)RtsFlags.GcFlags.allocLimitGrace * BLOCK_SIZE); + } else { + // If we aren't killing the thread, we must disable the limit + // otherwise we will immediatelly retrigger it. + // User defined handlers should re-enable it if wanted. + t->flags = t->flags & ~TSO_ALLOC_LIMIT; + } + + if(allocLimitRunHook) + { + // Create a thread to run the allocation limit handler. + StgClosure* c = rts_apply(cap, runAllocationLimitHandler_closure, (StgClosure*)t); + StgTSO* hookThread = createIOThread(cap, RtsFlags.GcFlags.initialStkSize, c); + setThreadLabel(cap, hookThread, "allocation limit handler thread"); + // Schedule the handler to be run immediatelly. + pushOnRunQueue(cap, hookThread); + } + } /* some statistics gathering in the parallel case */ @@ -3342,3 +3364,9 @@ resurrectThreads (StgTSO *threads) } } } + +void setAllocLimitKill(bool shouldKill, bool shouldHook) +{ + allocLimitKill = shouldKill; + allocLimitRunHook = shouldHook; +} ===================================== rts/external-symbols.list.in ===================================== @@ -43,6 +43,7 @@ ghczminternal_GHCziInternalziTypes_Izh_con_info ghczminternal_GHCziInternalziTypes_Fzh_con_info ghczminternal_GHCziInternalziTypes_Dzh_con_info ghczminternal_GHCziInternalziTypes_Wzh_con_info +ghczminternal_GHCziInternalziAllocationLimitHandler_runAllocationLimitHandler_closure ghczminternal_GHCziInternalziPtr_Ptr_con_info ghczminternal_GHCziInternalziPtr_FunPtr_con_info ghczminternal_GHCziInternalziInt_I8zh_con_info ===================================== rts/include/rts/storage/GC.h ===================================== @@ -279,6 +279,10 @@ void flushExec(W_ len, AdjustorExecutable exec_addr); // Used by GC checks in external .cmm code: extern W_ large_alloc_lim; +// Should triggering an allocation limit kill the thread +// and should we run a user-defined hook when it is triggered. +void setAllocLimitKill(bool, bool); + /* ----------------------------------------------------------------------------- Performing Garbage Collection -------------------------------------------------------------------------- */ ===================================== rts/include/rts/storage/TSO.h ===================================== @@ -157,9 +157,10 @@ typedef struct StgTSO_ { /* * The allocation limit for this thread, which is updated as the * thread allocates. If the value drops below zero, and - * TSO_ALLOC_LIMIT is set in flags, we raise an exception in the - * thread, and give the thread a little more space to handle the - * exception before we raise the exception again. + * TSO_ALLOC_LIMIT is set in flags, then a handler is triggerd. + * Either we raise an exception in the thread, and give the thread + * a little more space to handle the exception before we raise the + * exception again; or we run a user defined handler. * * This is an integer, because we might update it in a place where * it isn't convenient to raise the exception, so we want it to ===================================== rts/include/stg/MiscClosures.h ===================================== @@ -604,7 +604,9 @@ RTS_FUN_DECL(stg_traceEventzh); RTS_FUN_DECL(stg_traceBinaryEventzh); RTS_FUN_DECL(stg_traceMarkerzh); RTS_FUN_DECL(stg_getThreadAllocationCounterzh); +RTS_FUN_DECL(stg_getOtherThreadAllocationCounterzh); RTS_FUN_DECL(stg_setThreadAllocationCounterzh); +RTS_FUN_DECL(stg_setOtherThreadAllocationCounterzh); RTS_FUN_DECL(stg_castWord64ToDoublezh); RTS_FUN_DECL(stg_castDoubleToWord64zh); ===================================== testsuite/tests/backpack/cabal/bkpcabal08/bkpcabal08.stdout ===================================== @@ -13,13 +13,13 @@ Building library 'q' instantiated with for bkpcabal08-0.1.0.0... [2 of 4] Compiling B[sig] ( q/B.hsig, nothing ) [3 of 4] Compiling M ( q/M.hs, nothing ) [A changed] -[4 of 4] Instantiating bkpcabal08-0.1.0.0-CoQJNXLfoYQ4TyvApzFHv-p +[4 of 4] Instantiating bkpcabal08-0.1.0.0-LjROTJ30vjpHLvYNUnY1dD-p Preprocessing library 'q' for bkpcabal08-0.1.0.0... Building library 'q' instantiated with - A = bkpcabal08-0.1.0.0-5HJrxUERN7CD204UZeT4Ws-impl:A - B = bkpcabal08-0.1.0.0-5HJrxUERN7CD204UZeT4Ws-impl:B + A = bkpcabal08-0.1.0.0-KxkxGUgMJM6Dqo87AQu6V5-impl:A + B = bkpcabal08-0.1.0.0-KxkxGUgMJM6Dqo87AQu6V5-impl:B for bkpcabal08-0.1.0.0... -[1 of 3] Compiling A[sig] ( q/A.hsig, dist/build/bkpcabal08-0.1.0.0-1DQJ9DKc4h59P07qcb0kBc-q+J5mAfRWG9IgLmFQVftCb8t/A.o ) [Prelude package changed] -[2 of 3] Compiling B[sig] ( q/B.hsig, dist/build/bkpcabal08-0.1.0.0-1DQJ9DKc4h59P07qcb0kBc-q+J5mAfRWG9IgLmFQVftCb8t/B.o ) [Prelude package changed] +[1 of 3] Compiling A[sig] ( q/A.hsig, dist/build/bkpcabal08-0.1.0.0-9Cd344YIpXKFRnHN0O8bjR-q+BXOBBkqkS8ELakzWUKHCOx/A.o ) [Prelude package changed] +[2 of 3] Compiling B[sig] ( q/B.hsig, dist/build/bkpcabal08-0.1.0.0-9Cd344YIpXKFRnHN0O8bjR-q+BXOBBkqkS8ELakzWUKHCOx/B.o ) [Prelude package changed] Preprocessing library 'r' for bkpcabal08-0.1.0.0... Building library 'r' for bkpcabal08-0.1.0.0... ===================================== testsuite/tests/interface-stability/base-exports.stdout ===================================== @@ -4607,6 +4607,7 @@ module GHC.Base where sequence :: forall (m :: * -> *) a. Monad m => [m a] -> m [a] setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# @@ -6693,6 +6694,7 @@ module GHC.Exts where seq# :: forall a s. a -> State# s -> (# State# s, a #) setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# ===================================== testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs ===================================== @@ -4607,6 +4607,7 @@ module GHC.Base where sequence :: forall (m :: * -> *) a. Monad m => [m a] -> m [a] setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# @@ -6665,6 +6666,7 @@ module GHC.Exts where seq# :: forall a s. a -> State# s -> (# State# s, a #) setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# ===================================== testsuite/tests/interface-stability/base-exports.stdout-mingw32 ===================================== @@ -4610,6 +4610,7 @@ module GHC.Base where sequence :: forall (m :: * -> *) a. Monad m => [m a] -> m [a] setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# @@ -6836,6 +6837,7 @@ module GHC.Exts where seq# :: forall a s. a -> State# s -> (# State# s, a #) setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# ===================================== testsuite/tests/interface-stability/base-exports.stdout-ws-32 ===================================== @@ -4607,6 +4607,7 @@ module GHC.Base where sequence :: forall (m :: * -> *) a. Monad m => [m a] -> m [a] setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# @@ -6693,6 +6694,7 @@ module GHC.Exts where seq# :: forall a s. a -> State# s -> (# State# s, a #) setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# ===================================== testsuite/tests/interface-stability/ghc-experimental-exports.stdout ===================================== @@ -5873,6 +5873,7 @@ module GHC.PrimOps where seq# :: forall a s. a -> State# s -> (# State# s, a #) setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# @@ -10916,6 +10917,16 @@ module Prelude.Experimental where data Unit# = ... getSolo :: forall a. Solo a -> a +module System.Mem.Experimental where + -- Safety: None + type AllocationLimitKillBehaviour :: * + data AllocationLimitKillBehaviour = KillOnAllocationLimit | DontKillOnAllocationLimit + disableAllocationLimitFor :: GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO () + enableAllocationLimitFor :: GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO () + getAllocationCounterFor :: GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO GHC.Internal.Int.Int64 + setAllocationCounterFor :: GHC.Internal.Int.Int64 -> GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO () + setGlobalAllocationLimitHandler :: AllocationLimitKillBehaviour -> GHC.Internal.Maybe.Maybe (GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO ()) -> GHC.Internal.Types.IO () + -- Instances: instance GHC.Internal.Base.Alternative GHC.Internal.Types.IO -- Defined in ‘GHC.Internal.Base’ ===================================== testsuite/tests/interface-stability/ghc-experimental-exports.stdout-mingw32 ===================================== @@ -5876,6 +5876,7 @@ module GHC.PrimOps where seq# :: forall a s. a -> State# s -> (# State# s, a #) setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shiftL# :: Word# -> Int# -> Word# shiftRL# :: Word# -> Int# -> Word# @@ -10919,6 +10920,16 @@ module Prelude.Experimental where data Unit# = ... getSolo :: forall a. Solo a -> a +module System.Mem.Experimental where + -- Safety: None + type AllocationLimitKillBehaviour :: * + data AllocationLimitKillBehaviour = KillOnAllocationLimit | DontKillOnAllocationLimit + disableAllocationLimitFor :: GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO () + enableAllocationLimitFor :: GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO () + getAllocationCounterFor :: GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO GHC.Internal.Int.Int64 + setAllocationCounterFor :: GHC.Internal.Int.Int64 -> GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO () + setGlobalAllocationLimitHandler :: AllocationLimitKillBehaviour -> GHC.Internal.Maybe.Maybe (GHC.Internal.Conc.Sync.ThreadId -> GHC.Internal.Types.IO ()) -> GHC.Internal.Types.IO () + -- Instances: instance GHC.Internal.Base.Alternative GHC.Internal.Types.IO -- Defined in ‘GHC.Internal.Base’ ===================================== testsuite/tests/interface-stability/ghc-prim-exports.stdout ===================================== @@ -2505,6 +2505,7 @@ module GHC.Prim where seq :: forall {r :: GHC.Internal.Types.RuntimeRep} a (b :: TYPE r). a -> b -> b setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shrinkMutableByteArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> State# d shrinkSmallMutableArray# :: forall {l :: GHC.Internal.Types.Levity} d (a :: TYPE (GHC.Internal.Types.BoxedRep l)). SmallMutableArray# d a -> Int# -> State# d -> State# d @@ -3489,6 +3490,7 @@ module GHC.PrimopWrappers where retry# :: forall a_levpoly. GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> (# GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld, a_levpoly #) setAddrRange# :: GHC.Internal.Prim.Addr# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld setByteArray# :: forall s. GHC.Internal.Prim.MutableByteArray# s -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# s -> GHC.Internal.Prim.State# s + setOtherThreadAllocationCounter# :: GHC.Internal.Prim.Int64# -> GHC.Internal.Prim.ThreadId# -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld setThreadAllocationCounter# :: GHC.Internal.Prim.Int64# -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld shrinkMutableByteArray# :: forall s. GHC.Internal.Prim.MutableByteArray# s -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# s -> GHC.Internal.Prim.State# s shrinkSmallMutableArray# :: forall s a_levpoly. GHC.Internal.Prim.SmallMutableArray# s a_levpoly -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# s -> GHC.Internal.Prim.State# s ===================================== testsuite/tests/interface-stability/ghc-prim-exports.stdout-mingw32 ===================================== @@ -2505,6 +2505,7 @@ module GHC.Prim where seq :: forall {r :: GHC.Internal.Types.RuntimeRep} a (b :: TYPE r). a -> b -> b setAddrRange# :: Addr# -> Int# -> Int# -> State# RealWorld -> State# RealWorld setByteArray# :: forall d. MutableByteArray# d -> Int# -> Int# -> Int# -> State# d -> State# d + setOtherThreadAllocationCounter# :: Int64# -> ThreadId# -> State# RealWorld -> State# RealWorld setThreadAllocationCounter# :: Int64# -> State# RealWorld -> State# RealWorld shrinkMutableByteArray# :: forall d. MutableByteArray# d -> Int# -> State# d -> State# d shrinkSmallMutableArray# :: forall {l :: GHC.Internal.Types.Levity} d (a :: TYPE (GHC.Internal.Types.BoxedRep l)). SmallMutableArray# d a -> Int# -> State# d -> State# d @@ -3492,6 +3493,7 @@ module GHC.PrimopWrappers where retry# :: forall a_levpoly. GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> (# GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld, a_levpoly #) setAddrRange# :: GHC.Internal.Prim.Addr# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld setByteArray# :: forall s. GHC.Internal.Prim.MutableByteArray# s -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# s -> GHC.Internal.Prim.State# s + setOtherThreadAllocationCounter# :: GHC.Internal.Prim.Int64# -> GHC.Internal.Prim.ThreadId# -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld setThreadAllocationCounter# :: GHC.Internal.Prim.Int64# -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld -> GHC.Internal.Prim.State# GHC.Internal.Prim.RealWorld shrinkMutableByteArray# :: forall s. GHC.Internal.Prim.MutableByteArray# s -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# s -> GHC.Internal.Prim.State# s shrinkSmallMutableArray# :: forall s a_levpoly. GHC.Internal.Prim.SmallMutableArray# s a_levpoly -> GHC.Internal.Prim.Int# -> GHC.Internal.Prim.State# s -> GHC.Internal.Prim.State# s ===================================== testsuite/tests/rts/T22859.hs ===================================== @@ -0,0 +1,72 @@ +{-# LANGUAGE NumericUnderscores #-} +{-# LANGUAGE MagicHash #-} +{-# LANGUAGE UnboxedTuples #-} + +import Control.Exception +import Control.Exception.Backtrace +import Control.Concurrent +import Control.Concurrent.MVar +import System.Mem +import System.Mem.Experimental +import GHC.IO (IO (..)) +import GHC.Exts +import System.IO + +-- | Just do some work and hPutStrLn to stderr to indicate that we are making progress +worker :: IO () +worker = loop [] 2 + where + loop !m !n + | n > 30 = hPutStrLn stderr . show $ length m + | otherwise = do + let x = show n + hPutStrLn stderr x + -- just to bulk out the allocations + IO (\s -> case newByteArray# 900000# s of (# s', arr# #) -> (# s', () #)) + yield + loop (x:m) (n + 1) + +main :: IO () +main = do + hSetBuffering stderr LineBuffering -- necessary for Windows, otherwise our output gets garbled + done <- newMVar () -- we use this lock to wait for the worker to finish + started <- newEmptyMVar + let runWorker = do + forkIO . withMVar done $ \_ -> flip onException (hPutStrLn stderr "worker died") $ do + hPutStrLn stderr "worker starting" + putMVar started () + setAllocationCounter 1_000_000 + enableAllocationLimit + worker + hPutStrLn stderr "worker done" + takeMVar started + readMVar done + hFlush stderr + threadDelay 1000 + -- default behaviour: + -- kill it after the limit is exceeded + hPutStrLn stderr "default behaviour" + runWorker + hPutStrLn stderr "just log once on the hook being triggered" + setGlobalAllocationLimitHandler DontKillOnAllocationLimit (Just $ \_ -> hPutStrLn stderr "allocation limit triggered 1") + runWorker + hPutStrLn stderr "just log on the hook being triggered" + setGlobalAllocationLimitHandler DontKillOnAllocationLimit . Just $ \tid -> do + hPutStrLn stderr "allocation limit triggered 2" + -- re-enable the hook + setAllocationCounterFor 1_000_000 tid + enableAllocationLimitFor tid + runWorker + hPutStrLn stderr "kill from the hook" + setGlobalAllocationLimitHandler DontKillOnAllocationLimit . Just $ \tId -> throwTo tId AllocationLimitExceeded + runWorker + -- not super helpful, but let's test it anyway + hPutStrLn stderr "do nothing" + setGlobalAllocationLimitHandler DontKillOnAllocationLimit Nothing + runWorker + -- this is possible to handle using an exception handler instead. + hPutStrLn stderr "kill and log" + setGlobalAllocationLimitHandler KillOnAllocationLimit (Just $ \_ -> hPutStrLn stderr "allocation limit triggered 3") + runWorker + threadDelay 1000 + hPutStrLn stderr "done" ===================================== testsuite/tests/rts/T22859.stderr ===================================== @@ -0,0 +1,140 @@ +default behaviour +worker starting +2 +3 +worker died +T22859: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.SomeAsyncException: + +allocation limit exceeded +just log once on the hook being triggered +worker starting +2 +3 +allocation limit triggered 1 +4 +5 +6 +7 +8 +9 +10 +11 +12 +13 +14 +15 +16 +17 +18 +19 +20 +21 +22 +23 +24 +25 +26 +27 +28 +29 +30 +29 +worker done +just log on the hook being triggered +worker starting +2 +3 +allocation limit triggered 2 +4 +5 +allocation limit triggered 2 +6 +7 +allocation limit triggered 2 +8 +9 +allocation limit triggered 2 +10 +11 +allocation limit triggered 2 +12 +13 +allocation limit triggered 2 +14 +15 +allocation limit triggered 2 +16 +17 +allocation limit triggered 2 +18 +19 +allocation limit triggered 2 +20 +21 +allocation limit triggered 2 +22 +23 +allocation limit triggered 2 +24 +25 +allocation limit triggered 2 +26 +27 +allocation limit triggered 2 +28 +29 +allocation limit triggered 2 +30 +29 +worker done +kill from the hook +worker starting +2 +3 +worker died +T22859: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.SomeAsyncException: + +allocation limit exceeded +do nothing +worker starting +2 +3 +4 +5 +6 +7 +8 +9 +10 +11 +12 +13 +14 +15 +16 +17 +18 +19 +20 +21 +22 +23 +24 +25 +26 +27 +28 +29 +30 +29 +worker done +kill and log +worker starting +2 +3 +allocation limit triggered 3 +worker died +T22859: Uncaught exception ghc-internal:GHC.Internal.IO.Exception.SomeAsyncException: + +allocation limit exceeded +done ===================================== testsuite/tests/rts/all.T ===================================== @@ -643,3 +643,8 @@ test('T25280', [unless(opsys('linux'),skip),req_process,js_skip], compile_and_ru test('T25560', [req_c_rts, ignore_stderr], compile_and_run, ['']) test('TestProddableBlockSet', [req_c_rts], multimod_compile_and_run, ['TestProddableBlockSet.c', '-no-hs-main']) +test('T22859', + [js_skip, + # Allocation behaviour differs with the wasm backend so we get different output + when(arch('wasm32'), skip)], + compile_and_run, ['-with-rtsopts -A8K']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ea2c6673f71c321e5dc1f78e0b32cbd… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/ea2c6673f71c321e5dc1f78e0b32cbd… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc] Pushed new branch wip/romes/24212
by Rodrigo Mesquita (＠alt-romes) 08 Jul '25

08 Jul '25

Rodrigo Mesquita pushed new branch wip/romes/24212 at Glasgow Haskell Compiler / GHC -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/tree/wip/romes/24212 You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][wip/sol/add-rendered-to-json-diagnostics] 19 commits: debugger/rts: Allow toggling step-in per thread
by Simon Hengel (＠sol) 08 Jul '25

08 Jul '25

Simon Hengel pushed to branch wip/sol/add-rendered-to-json-diagnostics at Glasgow Haskell Compiler / GHC Commits: 8a69196e by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger/rts: Allow toggling step-in per thread The RTS global flag `rts_stop_next_breakpoint` globally sets the interpreter to stop at the immediate next breakpoint. With this commit, single step mode can additionally be set per thread in the TSO flag (TSO_STOP_NEXT_BREAKPOINT). Being able to toggle "stop at next breakpoint" per thread is an important requirement for implementing "stepping out" of a function in a multi-threaded context. And, more generally, having a per-thread flag for single-stepping paves the way for multi-threaded debugging. That said, when we want to enable "single step" mode for the whole interpreted program we still want to stop at the immediate next breakpoint, whichever thread it belongs to. That's why we also keep the global `rts_stop_next_breakpoint` flag, with `rts_enableStopNextBreakpointAll` and `rts_disableStopNextBreakpointAll` helpers. Preparation for #26042 - - - - - 73d3f864 by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 docs: Case continuation BCOs This commit documents a subtle interaction between frames for case BCOs and their parents frames. Namely, case continuation BCOs may refer to (non-local) variables that are part of the parent's frame. The note expanding a bit on these details is called [Case continuation BCOs] - - - - - d7aeddcf by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger: Implement step-out feature Implements support for stepping-out of a function (aka breaking right after returning from a function) in the interactive debugger. It also introduces a GHCi command :stepout to step-out of a function being debugged in the interpreter. The feature is described as: Stop at the first breakpoint immediately after returning from the current function scope. Known limitations: because a function tail-call does not push a stack frame, if step-out is used inside of a function that was tail-called, execution will not be returned to its caller, but rather its caller's first non-tail caller. On the other hand, it means the debugger follows the more realistic execution of the program. In the following example: .. code-block:: none f = do a b <--- (1) set breakpoint then step in here c b = do ... d <--- (2) step-into this tail call d = do ... something <--- (3) step-out here ... Stepping-out will stop execution at the `c` invokation in `f`, rather than stopping at `b`. The key idea is simple: When step-out is enabled, traverse the runtime stack until a continuation BCO is found -- and enable the breakpoint heading that BCO explicitly using its tick-index. The details are specified in `Note [Debugger: Step-out]` in `rts/Interpreter.c`. Since PUSH_ALTS BCOs (representing case continuations) were never headed by a breakpoint (unlike the case alternatives they push), we introduced the BRK_ALTS instruction to allow the debugger to set a case continuation to stop at the breakpoint heading the alternative that is taken. This is further described in `Note [Debugger: BRK_ALTS]`. Fixes #26042 - - - - - 5d9adf51 by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger: Filter step-out stops by SrcSpan To implement step-out, the RTS looks for the first continuation frame on the stack and explicitly enables its entry breakpoint. However, some continuations will be contained in the function from which step-out was initiated (trivial example is a case expression). Similarly to steplocal, we will filter the breakpoints at which the RTS yields to the debugger based on the SrcSpan. When doing step-out, only stop if the breakpoint is /not/ contained in the function from which we initiated it. This is especially relevant in monadic statements such as IO which is compiled to a long chain of case expressions. See Note [Debugger: Filtering step-out stops] - - - - - 7677adcc by Cheng Shao at 2025-07-08T07:40:29-04:00 compiler: make ModBreaks serializable - - - - - 14f67c6d by Rodrigo Mesquita at 2025-07-08T07:40:29-04:00 refactor: "Inspecting the session" moved from GHC Moved utilities for inspecting the session from the GHC module to GHC.Driver.Session.Inspect Purely a clean up - - - - - 9d3f484a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Pass the HUG to readModBreaks, not HscEnv A minor cleanup. The associated history and setupBreakpoint functions are changed accordingly. - - - - - b595f713 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move readModBreaks to GHC.Runtime.Interpreter With some small docs changes - - - - - d223227a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move interpreterProfiled to Interp.Types Moves interpreterProfiled and interpreterDynamic to GHC.Runtime.Interpreter.Types from GHC.Runtime.Interpreter. - - - - - 7fdd0a3d by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Don't import GHC in Debugger.Breakpoints Remove the top-level import GHC from GHC.Runtime.Debugger.Breakpoints This makes the module dependencies more granular and cleans up the qualified imports from the code. - - - - - 5e4da31b by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refactor: Use BreakpointId in Core and Ifaces - - - - - 741ac3a8 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 stg2bc: Derive BcM via ReaderT StateT A small refactor that simplifies GHC.StgToByteCode by deriving-via the Monad instances for BcM. This is done along the lines of previous similar refactors like 72b54c0760bbf85be1f73c1a364d4701e5720465. - - - - - 0414fcc9 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refact: Split InternalModBreaks out of ModBreaks There are currently two competing ways of referring to a Breakpoint: 1. Using the Tick module + Tick index 2. Using the Info module + Info index 1. The Tick index is allocated during desugaring in `mkModBreaks`. It is used to refer to a breakpoint associated to a Core Tick. For a given Tick module, there are N Ticks indexed by Tick index. 2. The Info index is allocated during code generation (in StgToByteCode) and uniquely identifies the breakpoints at runtime (and is indeed used to determine which breakpoint was hit at runtime). Why we need both is described by Note [Breakpoint identifiers]. For every info index we used to keep a `CgBreakInfo`, a datatype containing information relevant to ByteCode Generation, in `ModBreaks`. This commit splits out the `IntMap CgBreakInfo` out of `ModBreaks` into a new datatype `InternalModBreaks`. - The purpose is to separate the `ModBreaks` datatype, which stores data associated from tick-level information which is fixed after desugaring, from the unrelated `IntMap CgBreakInfo` information accumulated during bytecode generation. - We move `ModBreaks` to GHC.HsToCore.Breakpoints The new `InternalModBreaks` simply combines the `IntMap CgBreakInfo` with `ModBreaks`. After code generation we construct an `InternalModBreaks` with the `CgBreakInfo`s we accumulated and the existing `ModBreaks` and store that in the compiled BCO in `bc_breaks`. - Note that we previously only updated the `modBreaks_breakInfo` field of `ModBreaks` at this exact location, and then stored the updated `ModBreaks` in the same `bc_breaks`. - We put this new datatype in GHC.ByteCode.Breakpoints The rest of the pipeline for which CgBreakInfo is relevant is accordingly updated to also use `InternalModBreaks` - - - - - 2a097955 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Use BreakpointIds in bytecode gen Small clean up to use BreakpointId and InternalBreakpointId more uniformly in bytecode generation rather than using Module + Ix pairs - - - - - 0515cc2f by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 ghci: Allocate BreakArrays at link time only Previously, a BreakArray would be allocated with a slot for every tick in a module at `mkModBreaks`, in HsToCore. However, this approach has a few downsides: - It interleaves interpreter behaviour (allocating arrays for breakpoints) within the desugarer - It is inflexible in the sense it is impossible for the bytecode generator to add "internal" breakpoints that can be triggered at runtime, because those wouldn't have a source tick. (This is relevant for our intended implementation plan of step-out in #26042) - It ties the BreakArray indices to the *tick* indexes, while at runtime we would rather just have the *info* indexes (currently we have both because BreakArrays are indexed by the *tick* one). Paving the way for #26042 and #26064, this commit moves the allocation of BreakArrays to bytecode-loading time -- akin to what is done for CCS arrays. Since a BreakArray is allocated only when bytecode is linked, if a breakpoint is set (e.g. `:break 10`) before the bytecode is linked, there will exist no BreakArray to trigger the breakpoint in. Therefore, the function to allocate break arrays (`allocateBreakArrays`) is exposed and also used in GHC.Runtime.Eval to allocate a break array when a breakpoint is set, if it doesn't exist yet (in the linker env). - - - - - 8016561f by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add a test for T26176 - - - - - 454cd682 by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add test for #14010 This test started to work in GHC 9.6 and has worked since. This MR just adds a regression test - - - - - 431d8e39 by Simon Hengel at 2025-07-08T20:38:34+07:00 Refactoring - - - - - 81f2116d by Simon Hengel at 2025-07-08T20:38:34+07:00 Include the rendered message in -fdiagnostics-as-json output This implements #26173. - - - - - 95 changed files: - compiler/GHC.hs - compiler/GHC/ByteCode/Asm.hs - + compiler/GHC/ByteCode/Breakpoints.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Linker.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Map/Expr.hs - compiler/GHC/Core/Opt/OccurAnal.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Core/Ppr.hs - compiler/GHC/Core/Subst.hs - compiler/GHC/Core/Tidy.hs - compiler/GHC/Core/Utils.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/CoreToStg/Prep.hs - compiler/GHC/Driver/Config.hs - compiler/GHC/Driver/Errors.hs - + compiler/GHC/Driver/Session/Inspect.hs - compiler/GHC/HsToCore.hs - compiler/GHC/HsToCore/Breakpoints.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Syntax.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Debugger/Breakpoints.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Eval/Types.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Runtime/Interpreter/Types.hs - compiler/GHC/Stg/BcPrep.hs - compiler/GHC/Stg/FVs.hs - compiler/GHC/StgToByteCode.hs - − compiler/GHC/Types/Breakpoint.hs - compiler/GHC/Types/Error.hs - compiler/GHC/Types/Tickish.hs - compiler/GHC/Unit/Module/ModGuts.hs - compiler/GHC/Utils/Logger.hs - compiler/GHC/Utils/Ppr.hs - compiler/ghc.cabal.in - docs/users_guide/9.14.1-notes.rst - + docs/users_guide/diagnostics-as-json-schema-1_2.json - docs/users_guide/ghci.rst - docs/users_guide/using.rst - ghc/GHCi/UI.hs - libraries/ghc-heap/GHC/Exts/Heap/Closures.hs - libraries/ghc-heap/GHC/Exts/Heap/FFIClosures_ProfilingDisabled.hsc - libraries/ghc-heap/GHC/Exts/Heap/FFIClosures_ProfilingEnabled.hsc - libraries/ghc-heap/tests/parse_tso_flags.hs - + libraries/ghci/GHCi/Debugger.hs - libraries/ghci/GHCi/Message.hs - libraries/ghci/GHCi/Run.hs - libraries/ghci/ghci.cabal.in - rts/Disassembler.c - rts/Interpreter.c - rts/Interpreter.h - rts/RtsSymbols.c - rts/StgMiscClosures.cmm - rts/include/rts/Bytecodes.h - rts/include/rts/Constants.h - rts/include/rts/storage/Closures.h - testsuite/tests/count-deps/CountDepsAst.stdout - testsuite/tests/count-deps/CountDepsParser.stdout - testsuite/tests/driver/json.stderr - testsuite/tests/driver/json_warn.stderr - + testsuite/tests/ghci.debugger/scripts/T26042b.hs - + testsuite/tests/ghci.debugger/scripts/T26042b.script - + testsuite/tests/ghci.debugger/scripts/T26042b.stdout - + testsuite/tests/ghci.debugger/scripts/T26042c.hs - + testsuite/tests/ghci.debugger/scripts/T26042c.script - + testsuite/tests/ghci.debugger/scripts/T26042c.stdout - + testsuite/tests/ghci.debugger/scripts/T26042d.hs - + testsuite/tests/ghci.debugger/scripts/T26042d.script - + testsuite/tests/ghci.debugger/scripts/T26042d.stdout - + testsuite/tests/ghci.debugger/scripts/T26042e.hs - + testsuite/tests/ghci.debugger/scripts/T26042e.script - + testsuite/tests/ghci.debugger/scripts/T26042e.stdout - + testsuite/tests/ghci.debugger/scripts/T26042f.hs - + testsuite/tests/ghci.debugger/scripts/T26042f.script - + testsuite/tests/ghci.debugger/scripts/T26042f1.stderr - + testsuite/tests/ghci.debugger/scripts/T26042f1.stdout - + testsuite/tests/ghci.debugger/scripts/T26042f2.stdout - + testsuite/tests/ghci.debugger/scripts/T26042g.hs - + testsuite/tests/ghci.debugger/scripts/T26042g.script - + testsuite/tests/ghci.debugger/scripts/T26042g.stdout - testsuite/tests/ghci.debugger/scripts/all.T - + testsuite/tests/indexed-types/should_fail/T26176.hs - + testsuite/tests/indexed-types/should_fail/T26176.stderr - testsuite/tests/indexed-types/should_fail/all.T - + testsuite/tests/typecheck/should_compile/T14010.hs - testsuite/tests/typecheck/should_compile/all.T The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8e9f2be102e49c8452ceb14e416352… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/8e9f2be102e49c8452ceb14e416352… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][wip/sol/add-rendered-to-json-diagnostics] Include the rendered message in -fdiagnostics-as-json output
by Simon Hengel (＠sol) 08 Jul '25

08 Jul '25

Simon Hengel pushed to branch wip/sol/add-rendered-to-json-diagnostics at Glasgow Haskell Compiler / GHC Commits: 8e9f2be1 by Simon Hengel at 2025-07-08T20:37:37+07:00 Include the rendered message in -fdiagnostics-as-json output This implements #26173. - - - - - 8 changed files: - compiler/GHC/Driver/Errors.hs - compiler/GHC/Types/Error.hs - compiler/GHC/Utils/Logger.hs - docs/users_guide/9.14.1-notes.rst - + docs/users_guide/diagnostics-as-json-schema-1_2.json - docs/users_guide/using.rst - testsuite/tests/driver/json.stderr - testsuite/tests/driver/json_warn.stderr Changes: ===================================== compiler/GHC/Driver/Errors.hs ===================================== @@ -10,6 +10,7 @@ import GHC.Prelude import GHC.Types.SrcLoc (SrcSpan) import GHC.Types.SourceError import GHC.Types.Error +import GHC.Utils.Json import GHC.Utils.Error import GHC.Utils.Outputable import GHC.Utils.Logger @@ -22,9 +23,22 @@ printMessages logger msg_opts opts = mapM_ printMessage . sortMessages printMessage :: MsgEnvelope a -> IO () printMessage message - | log_diags_as_json = logJsonMsg logger messageClass message + | log_diags_as_json = do + decorated <- decorateDiagnostic logflags messageClass location doc + let + rendered :: String + rendered = renderWithContext (log_default_user_context logflags) decorated + + jsonMessage :: JsonDoc + jsonMessage = jsonDiagnostic rendered message + + logJsonMsg logger messageClass jsonMessage + | otherwise = logMsg logger messageClass location doc where + logflags :: LogFlags + logflags = logFlags logger + doc :: SDoc doc = updSDocContext (\_ -> ctx) (messageWithHints diagnostic) ===================================== compiler/GHC/Types/Error.hs ===================================== @@ -73,6 +73,9 @@ module GHC.Types.Error , mkLocMessage , mkLocMessageWarningGroups , getCaretDiagnostic + + , jsonDiagnostic + -- * Queries , isIntrinsicErrorMessage , isExtrinsicErrorMessage @@ -109,7 +112,7 @@ import GHC.Utils.Panic import GHC.Version (cProjectVersion) import Data.Bifunctor -import Data.Foldable ( fold, toList ) +import Data.Foldable import Data.List.NonEmpty ( NonEmpty (..) ) import qualified Data.List.NonEmpty as NE import Data.List ( intercalate ) @@ -171,9 +174,6 @@ instance Diagnostic e => Outputable (Messages e) where pprDiagnostic (errMsgDiagnostic envelope) ] -instance (Diagnostic e) => ToJson (Messages e) where - json msgs = JSArray . toList $ json <$> getMessages msgs - {- Note [Discarding Messages] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -573,7 +573,7 @@ instance ToJson DiagnosticCode where {- Note [Diagnostic Message JSON Schema] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The below instance of ToJson must conform to the JSON schema -specified in docs/users_guide/diagnostics-as-json-schema-1_1.json. +specified in docs/users_guide/diagnostics-as-json-schema-1_2.json. When the schema is altered, please bump the version. If the content is altered in a backwards compatible way, update the minor version (e.g. 1.3 ~> 1.4). @@ -586,15 +586,17 @@ https://json-schema.org -} schemaVersion :: String -schemaVersion = "1.1" +schemaVersion = "1.2" + -- See Note [Diagnostic Message JSON Schema] before editing! -instance Diagnostic e => ToJson (MsgEnvelope e) where - json m = JSObject $ [ +jsonDiagnostic :: forall e. Diagnostic e => String -> MsgEnvelope e -> JsonDoc +jsonDiagnostic rendered m = JSObject $ [ ("version", JSString schemaVersion), ("ghcVersion", JSString $ "ghc-" ++ cProjectVersion), ("span", json $ errMsgSpan m), ("severity", json $ errMsgSeverity m), ("code", maybe JSNull json (diagnosticCode diag)), + ("rendered", JSString rendered), ("message", JSArray $ map renderToJSString diagMsg), ("hints", JSArray $ map (renderToJSString . ppr) (diagnosticHints diag) ) ] ++ [ ("reason", reasonJson) ===================================== compiler/GHC/Utils/Logger.hs ===================================== @@ -62,6 +62,8 @@ module GHC.Utils.Logger , logJsonMsg , logDumpMsg + , decorateDiagnostic + -- * Dumping , defaultDumpAction , putDumpFile @@ -419,7 +421,7 @@ defaultLogActionWithHandles out err logflags msg_class srcSpan msg MCInfo -> printErrs msg MCFatal -> printErrs msg MCDiagnostic SevIgnore _ _ -> pure () -- suppress the message - MCDiagnostic _sev _rea _code -> printDiagnostics + MCDiagnostic _sev _rea _code -> decorateDiagnostic logflags msg_class srcSpan msg >>= printErrs where printOut :: SDoc -> IO () printOut = defaultLogActionHPrintDoc logflags False out @@ -430,23 +432,54 @@ defaultLogActionWithHandles out err logflags msg_class srcSpan msg putStrSDoc :: SDoc -> IO () putStrSDoc = defaultLogActionHPutStrDoc logflags False out +-- This function is used by `defaultLogActionWithHandles` for non-JSON output, +-- and also by `GHC.Driver.Errors.printMessages` to produce the `rendered` +-- message on `-fdiagnostics-as-json`. +-- +-- We would want to eventually consolidate this. However, this is currently +-- not feasible for the following reasons: +-- +-- 1. Some parts of the compiler sidestep `printMessages`, for that reason we +-- can not decorate the message in `printMessages`. +-- +-- 2. GHC uses two different code paths for JSON and non-JSON diagnostics. For +-- that reason we can not decorate the message in `defaultLogActionWithHandles`. +-- +-- See also Note [JSON Error Messages]: +-- +-- `jsonLogAction` should be removed along with -ddump-json +-- +-- Also note that (1) is the reason why some parts of the compiler produce +-- diagnostics that don't respect `-fdiagnostics-as-json`. +-- +-- The plan as I see it is as follows: +-- +-- 1. Refactor all places in the compiler that report diagnostics to go +-- through `GHC.Driver.Errors.printMessages`. +-- +-- (It's easy to find all those places by looking for who creates +-- MCDiagnostic, either directly or via `mkMCDiagnostic` or +-- `errorDiagnostic`.) +-- +-- 2. Get rid of `-ddump-json`, `jsonLogAction` and consolidate message +-- decoration at one place (either `printMessages` or +-- `defaultLogActionWithHandles`) +-- +decorateDiagnostic :: LogFlags -> MessageClass -> SrcSpan -> SDoc -> IO SDoc +decorateDiagnostic logflags msg_class srcSpan msg = addCaret + where -- Pretty print the warning flag, if any (#10752) message :: SDoc message = mkLocMessageWarningGroups (log_show_warn_groups logflags) msg_class srcSpan msg - printDiagnostics :: IO () - printDiagnostics = do + addCaret = do caretDiagnostic <- if log_show_caret logflags then getCaretDiagnostic msg_class srcSpan else pure empty - printErrs $ getPprStyle $ \style -> + return $ getPprStyle $ \style -> withPprStyle (setStyleColoured True style) (message $+$ caretDiagnostic $+$ blankLine) - -- careful (#2302): printErrs prints in UTF-8, - -- whereas converting to string first and using - -- hPutStr would just emit the low 8 bits of - -- each unicode char. -- | Like 'defaultLogActionHPutStrDoc' but appends an extra newline. defaultLogActionHPrintDoc :: LogFlags -> Bool -> Handle -> SDoc -> IO () @@ -611,8 +644,8 @@ defaultTraceAction logflags title doc x = logMsg :: Logger -> MessageClass -> SrcSpan -> SDoc -> IO () logMsg logger mc loc msg = putLogMsg logger (logFlags logger) mc loc msg -logJsonMsg :: ToJson a => Logger -> MessageClass -> a -> IO () -logJsonMsg logger mc d = putJsonLogMsg logger (logFlags logger) mc (json d) +logJsonMsg :: Logger -> MessageClass -> JsonDoc -> IO () +logJsonMsg logger mc = putJsonLogMsg logger (logFlags logger) mc -- | Dump something logDumpFile :: Logger -> PprStyle -> DumpFlag -> String -> DumpFormat -> SDoc -> IO () ===================================== docs/users_guide/9.14.1-notes.rst ===================================== @@ -143,6 +143,11 @@ Compiler were accessed using the generated record selector functions, marking the fields as covered in coverage reports (:ghc-ticket:`17834`). +- 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 ~~~~ ===================================== docs/users_guide/diagnostics-as-json-schema-1_2.json ===================================== @@ -0,0 +1,144 @@ +{ + "$schema": "https://json-schema.org/draft/2020-12/schema", + "title": "JSON Diagnostic Schema", + "description": "A Schema for specifying GHC diagnostics output as JSON", + "type": "object", + "properties": { + "version": { + "description": "The current JSON schema version this object conforms to", + "type": "string" + }, + "ghcVersion": { + "description": "The GHC version", + "type": "string" + }, + "span": { + "oneOf": [ + { "$ref": "#/$defs/span" }, + { "type": "null" } + ] + }, + "severity": { + "description": "The diagnostic severity", + "type": "string", + "enum": [ + "Warning", + "Error" + ] + }, + "code": { + "description": "The diagnostic code (if it exists)", + "type": [ + "integer", + "null" + ] + }, + "rendered": { + "description": "The rendered diagnostics message, in the exact same format as what GHC would have produced without -fdiagnostics-as-json (including ANSI escape sequences)", + "type": "string" + }, + "message": { + "description": "The string output of the diagnostic message by GHC", + "type": "array", + "items": { + "type": "string" + } + }, + "hints": { + "description": "The suggested fixes", + "type": "array", + "items": { + "type": "string" + } + }, + "reason" : { + "description": "The GHC flag that was responsible for the emission of the diagnostic message", + "oneOf": [ + { + "type": "object", + "description": "The diagnostic message was controlled by one or more GHC flags", + "properties": { + "flags": { + "type": "array", + "items": { + "description": "The name of a GHC flag controlling the diagnostic message", + "type": "string" + }, + "minItems": 1 + } + }, + "required": ["flags"] + }, + { + "type": "object", + "description": "The diagnostic message was controlled by a GHC diagnostic message category", + "properties": { + "category": { + "description": "The name of the GHC diagnostic message category controlling the diagnostic message", + "type": "string" + } + }, + "required": ["category"] + } + ] + } + }, + + "$comment": "NOTE: \"rendered\" is not a required field so that the schema is backward compatible with version 1.1. If you bump the schema version to 2.0 the please also add \"rendered\" to the \"required\" fields.", + "required": [ + "version", + "ghcVersion", + "span", + "severity", + "code", + "message", + "hints" + ], + + "additionalProperties": false, + "$defs": { + "span": { + "description": "The span of the diagnostic", + "type": "object", + "properties": { + "file": { + "description": "The file in which the diagnostic occurs", + "type": "string" + }, + "start": { + "description": "The start location of the diagnostic", + "$ref": "#/$defs/location" + }, + "end": { + "description": "The end location of the diagnostic", + "$ref": "#/$defs/location" + } + }, + "required": [ + "file", + "start", + "end" + ], + "additionalProperties": false + }, + "location": { + "description": "A location in a text file", + "type": "object", + "properties": { + "line": { + "description": "The line number", + "type": "integer" + }, + "column": { + "description": "The column number", + "type": "integer" + } + }, + "required": [ + "line", + "column" + ], + "additionalProperties": false + } + } +} ===================================== docs/users_guide/using.rst ===================================== @@ -1428,7 +1428,7 @@ messages and in GHCi: a new line. The structure of the output is described by a `JSON Schema <https://json-schema.org/>`_. - The schema can be downloaded :download:`here <diagnostics-as-json-schema-1_1.json>`. + The schema can be downloaded :download:`here <diagnostics-as-json-schema-1_2.json>`. .. ghc-flag:: -fdiagnostics-color=⟨always|auto|never⟩ :shortdesc: Use colors in error messages ===================================== testsuite/tests/driver/json.stderr ===================================== @@ -1 +1 @@ -{"version":"1.1","ghcVersion":"ghc-9.13.20250529","span":{"file":"json.hs","start":{"line":9,"column":11},"end":{"line":9,"column":21}},"severity":"Error","code":48010,"message":["Empty list of alternatives in case expression"],"hints":["Perhaps you intended to use the \u2018EmptyCase\u2019 extension"]} +{"version":"1.2","ghcVersion":"ghc-9.13.20250627","span":{"file":"json.hs","start":{"line":9,"column":11},"end":{"line":9,"column":21}},"severity":"Error","code":48010,"rendered":"json.hs:9:11: error: [GHC-48010]\n Empty list of alternatives in case expression\n Suggested fix:\n Perhaps you intended to use the \u2018EmptyCase\u2019 extension\n","message":["Empty list of alternatives in case expression"],"hints":["Perhaps you intended to use the \u2018EmptyCase\u2019 extension"]} ===================================== testsuite/tests/driver/json_warn.stderr ===================================== @@ -1,2 +1,2 @@ -{"version":"1.1","ghcVersion":"ghc-9.13.20250529","span":{"file":"json_warn.hs","start":{"line":4,"column":3},"end":{"line":4,"column":4}},"severity":"Warning","code":40910,"message":["Defined but not used: \u2018x\u2019"],"hints":[],"reason":{"flags":["unused-matches"]}} -{"version":"1.1","ghcVersion":"ghc-9.13.20250529","span":{"file":"json_warn.hs","start":{"line":7,"column":5},"end":{"line":7,"column":9}},"severity":"Warning","code":63394,"message":["In the use of \u2018head\u2019\n(imported from Prelude, but defined in GHC.Internal.List):\n\"This is a partial function, it throws an error on empty lists. Use pattern matching, 'Data.List.uncons' or 'Data.Maybe.listToMaybe' instead. Consider refactoring to use \"Data.List.NonEmpty\".\""],"hints":[],"reason":{"category":"x-partial"}} +{"version":"1.2","ghcVersion":"ghc-9.13.20250627","span":{"file":"json_warn.hs","start":{"line":4,"column":3},"end":{"line":4,"column":4}},"severity":"Warning","code":40910,"rendered":"json_warn.hs:4:3: warning: [GHC-40910] [-Wunused-matches (in -Wextra)]\n Defined but not used: \u2018x\u2019\n","message":["Defined but not used: \u2018x\u2019"],"hints":[],"reason":{"flags":["unused-matches"]}} +{"version":"1.2","ghcVersion":"ghc-9.13.20250627","span":{"file":"json_warn.hs","start":{"line":7,"column":5},"end":{"line":7,"column":9}},"severity":"Warning","code":63394,"rendered":"json_warn.hs:7:5: warning: [GHC-63394] [-Wx-partial (in -Wextended-warnings)]\n In the use of \u2018head\u2019\n (imported from Prelude, but defined in GHC.Internal.List):\n \"This is a partial function, it throws an error on empty lists. Use pattern matching, 'Data.List.uncons' or 'Data.Maybe.listToMaybe' instead. Consider refactoring to use \"Data.List.NonEmpty\".\"\n","message":["In the use of \u2018head\u2019\n(imported from Prelude, but defined in GHC.Internal.List):\n\"This is a partial function, it throws an error on empty lists. Use pattern matching, 'Data.List.uncons' or 'Data.Maybe.listToMaybe' instead. Consider refactoring to use \"Data.List.NonEmpty\".\""],"hints":[],"reason":{"category":"x-partial"}} View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8e9f2be102e49c8452ceb14e4163520… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/8e9f2be102e49c8452ceb14e4163520… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][wip/romes/step-out-10] 33 commits: Consider `PromotedDataCon` in `tyConStupidTheta`
by Rodrigo Mesquita (＠alt-romes) 08 Jul '25

08 Jul '25

Rodrigo Mesquita pushed to branch wip/romes/step-out-10 at Glasgow Haskell Compiler / GHC Commits: 8d33d048 by Berk Özkütük at 2025-07-07T20:42:20-04:00 Consider `PromotedDataCon` in `tyConStupidTheta` Haddock checks data declarations for the stupid theta so as not to pretty-print them as empty contexts. Type data declarations end up as `PromotedDataCon`s by the time Haddock performs this check, causing a panic. This commit extends `tyConStupidTheta` so that it returns an empty list for `PromotedDataCon`s. This decision was guided by the fact that type data declarations never have data type contexts (see (R1) in Note [Type data declarations]). Fixes #25739. - - - - - a26243fd by Ryan Hendrickson at 2025-07-07T20:43:07-04:00 haddock: Document instances from other packages When attaching instances to `Interface`s, it isn't enough just to look for instances in the list of `Interface`s being processed. We also need to look in the modules on which they depend, including those outside of this package. Fixes #25147. Fixes #26079. - - - - - 0fb24420 by Rodrigo Mesquita at 2025-07-07T20:43:49-04:00 hadrian: Fallback logic for internal interpreter When determining whether to build the internal interpreter, the `make` build system had a fallback case for platforms not in the list of explicitly-supported operating systems and architectures. This fallback says we should try to build the internal interpreter if building dynamic GHC programs (if the architecture is unknown). Fixes #24098 - - - - - fe925bd4 by Ben Gamari at 2025-07-07T20:44:30-04:00 users-guide: Reference Wasm FFI section - - - - - 5856284b by Ben Gamari at 2025-07-07T20:44:30-04:00 users-guide: Fix too-short heading warning - - - - - a48dcdf3 by Duncan Coutts at 2025-07-07T20:45:18-04:00 Reorganise documentation for allocate* functions Consolodate interface information into the .h file, keeping just implementation details in the .c file. Use Notes stlye in the .h file and refer to notes from the .c file. - - - - - de5b528c by Duncan Coutts at 2025-07-07T20:45:18-04:00 Introduce common utilities for allocating arrays The intention is to share code among the several places that do this already. - - - - - b321319d by Duncan Coutts at 2025-07-07T20:45:18-04:00 Use new array alloc utils in Heap.c The CMM primop can now report heap overflow. - - - - - 1d557ffb by Duncan Coutts at 2025-07-07T20:45:18-04:00 Use new array alloc utils in ThreadLabels.c Replacing a local utility. - - - - - e59a1430 by Duncan Coutts at 2025-07-07T20:45:18-04:00 Use new array alloc utils in Threads.c Replacing local open coded version. - - - - - 482df1c9 by Duncan Coutts at 2025-07-07T20:45:18-04:00 Add exitHeapOverflow helper utility This will be useful with the array alloc functions, since unlike allocate/allocateMaybeFail, they do not come in two versions. So if it's not convenient to propagate failure, then one can use this. - - - - - 4d3ec8f9 by Duncan Coutts at 2025-07-07T20:45:18-04:00 Use new array alloc utils in Weak.c Also add a cpp macro CCS_SYSTEM_OR_NULL which does what it says. The benefit of this is that it allows us to referece CCS_SYSTEM even when we're not in PROFILING mode. That makes abstracting over profiling vs normal mode a lot easier. - - - - - 0c4f2fde by Duncan Coutts at 2025-07-07T20:45:18-04:00 Convert the array alloc primops to use the new array alloc utils - - - - - a3354ad9 by Duncan Coutts at 2025-07-07T20:45:18-04:00 While we're at it, add one missing 'likely' hint To a cmm primops that raises an exception, like the others now do. - - - - - 33b546bd by meooow25 at 2025-07-07T20:46:09-04:00 Keep scanl' strict in the head on rewrite `scanl'` forces elements to WHNF when the corresponding `(:)`s are forced. The rewrite rule for `scanl'` missed forcing the first element, which is fixed here with a `seq`. - - - - - 8a69196e by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger/rts: Allow toggling step-in per thread The RTS global flag `rts_stop_next_breakpoint` globally sets the interpreter to stop at the immediate next breakpoint. With this commit, single step mode can additionally be set per thread in the TSO flag (TSO_STOP_NEXT_BREAKPOINT). Being able to toggle "stop at next breakpoint" per thread is an important requirement for implementing "stepping out" of a function in a multi-threaded context. And, more generally, having a per-thread flag for single-stepping paves the way for multi-threaded debugging. That said, when we want to enable "single step" mode for the whole interpreted program we still want to stop at the immediate next breakpoint, whichever thread it belongs to. That's why we also keep the global `rts_stop_next_breakpoint` flag, with `rts_enableStopNextBreakpointAll` and `rts_disableStopNextBreakpointAll` helpers. Preparation for #26042 - - - - - 73d3f864 by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 docs: Case continuation BCOs This commit documents a subtle interaction between frames for case BCOs and their parents frames. Namely, case continuation BCOs may refer to (non-local) variables that are part of the parent's frame. The note expanding a bit on these details is called [Case continuation BCOs] - - - - - d7aeddcf by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger: Implement step-out feature Implements support for stepping-out of a function (aka breaking right after returning from a function) in the interactive debugger. It also introduces a GHCi command :stepout to step-out of a function being debugged in the interpreter. The feature is described as: Stop at the first breakpoint immediately after returning from the current function scope. Known limitations: because a function tail-call does not push a stack frame, if step-out is used inside of a function that was tail-called, execution will not be returned to its caller, but rather its caller's first non-tail caller. On the other hand, it means the debugger follows the more realistic execution of the program. In the following example: .. code-block:: none f = do a b <--- (1) set breakpoint then step in here c b = do ... d <--- (2) step-into this tail call d = do ... something <--- (3) step-out here ... Stepping-out will stop execution at the `c` invokation in `f`, rather than stopping at `b`. The key idea is simple: When step-out is enabled, traverse the runtime stack until a continuation BCO is found -- and enable the breakpoint heading that BCO explicitly using its tick-index. The details are specified in `Note [Debugger: Step-out]` in `rts/Interpreter.c`. Since PUSH_ALTS BCOs (representing case continuations) were never headed by a breakpoint (unlike the case alternatives they push), we introduced the BRK_ALTS instruction to allow the debugger to set a case continuation to stop at the breakpoint heading the alternative that is taken. This is further described in `Note [Debugger: BRK_ALTS]`. Fixes #26042 - - - - - 5d9adf51 by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger: Filter step-out stops by SrcSpan To implement step-out, the RTS looks for the first continuation frame on the stack and explicitly enables its entry breakpoint. However, some continuations will be contained in the function from which step-out was initiated (trivial example is a case expression). Similarly to steplocal, we will filter the breakpoints at which the RTS yields to the debugger based on the SrcSpan. When doing step-out, only stop if the breakpoint is /not/ contained in the function from which we initiated it. This is especially relevant in monadic statements such as IO which is compiled to a long chain of case expressions. See Note [Debugger: Filtering step-out stops] - - - - - 7677adcc by Cheng Shao at 2025-07-08T07:40:29-04:00 compiler: make ModBreaks serializable - - - - - 14f67c6d by Rodrigo Mesquita at 2025-07-08T07:40:29-04:00 refactor: "Inspecting the session" moved from GHC Moved utilities for inspecting the session from the GHC module to GHC.Driver.Session.Inspect Purely a clean up - - - - - 9d3f484a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Pass the HUG to readModBreaks, not HscEnv A minor cleanup. The associated history and setupBreakpoint functions are changed accordingly. - - - - - b595f713 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move readModBreaks to GHC.Runtime.Interpreter With some small docs changes - - - - - d223227a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move interpreterProfiled to Interp.Types Moves interpreterProfiled and interpreterDynamic to GHC.Runtime.Interpreter.Types from GHC.Runtime.Interpreter. - - - - - 7fdd0a3d by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Don't import GHC in Debugger.Breakpoints Remove the top-level import GHC from GHC.Runtime.Debugger.Breakpoints This makes the module dependencies more granular and cleans up the qualified imports from the code. - - - - - 5e4da31b by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refactor: Use BreakpointId in Core and Ifaces - - - - - 741ac3a8 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 stg2bc: Derive BcM via ReaderT StateT A small refactor that simplifies GHC.StgToByteCode by deriving-via the Monad instances for BcM. This is done along the lines of previous similar refactors like 72b54c0760bbf85be1f73c1a364d4701e5720465. - - - - - 0414fcc9 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refact: Split InternalModBreaks out of ModBreaks There are currently two competing ways of referring to a Breakpoint: 1. Using the Tick module + Tick index 2. Using the Info module + Info index 1. The Tick index is allocated during desugaring in `mkModBreaks`. It is used to refer to a breakpoint associated to a Core Tick. For a given Tick module, there are N Ticks indexed by Tick index. 2. The Info index is allocated during code generation (in StgToByteCode) and uniquely identifies the breakpoints at runtime (and is indeed used to determine which breakpoint was hit at runtime). Why we need both is described by Note [Breakpoint identifiers]. For every info index we used to keep a `CgBreakInfo`, a datatype containing information relevant to ByteCode Generation, in `ModBreaks`. This commit splits out the `IntMap CgBreakInfo` out of `ModBreaks` into a new datatype `InternalModBreaks`. - The purpose is to separate the `ModBreaks` datatype, which stores data associated from tick-level information which is fixed after desugaring, from the unrelated `IntMap CgBreakInfo` information accumulated during bytecode generation. - We move `ModBreaks` to GHC.HsToCore.Breakpoints The new `InternalModBreaks` simply combines the `IntMap CgBreakInfo` with `ModBreaks`. After code generation we construct an `InternalModBreaks` with the `CgBreakInfo`s we accumulated and the existing `ModBreaks` and store that in the compiled BCO in `bc_breaks`. - Note that we previously only updated the `modBreaks_breakInfo` field of `ModBreaks` at this exact location, and then stored the updated `ModBreaks` in the same `bc_breaks`. - We put this new datatype in GHC.ByteCode.Breakpoints The rest of the pipeline for which CgBreakInfo is relevant is accordingly updated to also use `InternalModBreaks` - - - - - 2a097955 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Use BreakpointIds in bytecode gen Small clean up to use BreakpointId and InternalBreakpointId more uniformly in bytecode generation rather than using Module + Ix pairs - - - - - 0515cc2f by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 ghci: Allocate BreakArrays at link time only Previously, a BreakArray would be allocated with a slot for every tick in a module at `mkModBreaks`, in HsToCore. However, this approach has a few downsides: - It interleaves interpreter behaviour (allocating arrays for breakpoints) within the desugarer - It is inflexible in the sense it is impossible for the bytecode generator to add "internal" breakpoints that can be triggered at runtime, because those wouldn't have a source tick. (This is relevant for our intended implementation plan of step-out in #26042) - It ties the BreakArray indices to the *tick* indexes, while at runtime we would rather just have the *info* indexes (currently we have both because BreakArrays are indexed by the *tick* one). Paving the way for #26042 and #26064, this commit moves the allocation of BreakArrays to bytecode-loading time -- akin to what is done for CCS arrays. Since a BreakArray is allocated only when bytecode is linked, if a breakpoint is set (e.g. `:break 10`) before the bytecode is linked, there will exist no BreakArray to trigger the breakpoint in. Therefore, the function to allocate break arrays (`allocateBreakArrays`) is exposed and also used in GHC.Runtime.Eval to allocate a break array when a breakpoint is set, if it doesn't exist yet (in the linker env). - - - - - 8016561f by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add a test for T26176 - - - - - 454cd682 by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add test for #14010 This test started to work in GHC 9.6 and has worked since. This MR just adds a regression test - - - - - 6696180f by Rodrigo Mesquita at 2025-07-08T13:42:28+01:00 debugger: Uniquely identify breakpoints by internal id Since b85b11994e0130ff2401dd4bbdf52330e0bcf776 (support inlining breakpoints), a breakpoint has been identified at runtime by *two* pairs of <module,index>. - The first, aka a 'BreakpointId', uniquely identifies a breakpoint in the source of a module by using the Tick index. A Tick index can index into ModBreaks.modBreaks_xxx to fetch source-level information about where that tick originated. - When a user specifies e.g. a line breakpoint using :break, we'll reverse engineer what a Tick index for that line - We update the `BreakArray` of that module (got from the LoaderState) at that tick index to `breakOn`. - A BCO we can stop at is headed by a BRK_FUN instruction. This instruction stores in an operand the `tick index` it is associated to. We look it up in the associated `BreakArray` (also an operand) and check wheter it was set to `breakOn`. - The second, aka the `ibi_info_mod` + `ibi_info_ix` of the `InternalBreakpointId`, uniquely index into the `imodBreaks_breakInfo` -- the information we gathered during code generation about the existing breakpoint *ocurrences*. - Note that with optimisation there may be many occurrences of the same source-tick-breakpoint across different modules. The `ibi_info_ix` is unique per occurrence, but the `bi_tick_ix` may be shared. See Note [Breakpoint identifiers] about this. - Note that besides the tick ids, info ids are also stored in `BRK_FUN` so the break handler can refer to the associated `CgBreakInfo`. In light of that, the driving changes come from the desire to have the info_id uniquely identify the breakpoint at runtime, and the source tick id being derived from it: - An InternalBreakpointId should uniquely identify a breakpoint just from the code-generation identifiers of `ibi_info_ix` and `ibi_info_mod`. So we drop `ibi_tick_mod` and `ibi_tick_ix`. - A BRK_FUN instruction need only record the "internal breakpoint id", not the tick-level id. So we drop the tick mod and tick index operands. - A BreakArray should be indexed by InternalBreakpointId rather than BreakpointId That means we need to do some more work when setting a breakpoint. Specifically, we need to figure out the internal ids (occurrences of a breakpoint) from the source-level BreakpointId we want to set the breakpoint at (recall :break refers to breaks at the source level). Besides this change being an improvement to the handling of breakpoints (it's clearer to have a single unique identifier than two competing ones), it unlocks the possibility of generating "internal" breakpoints during Cg (needed for #26042). It should also be easier to introduce multi-threaded-aware `BreakArrays` following this change (needed for #26064). Se also the new Note [ModBreaks vs InternalModBreaks] - - - - - 120 changed files: - compiler/GHC.hs - compiler/GHC/ByteCode/Asm.hs - + compiler/GHC/ByteCode/Breakpoints.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Linker.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Map/Expr.hs - compiler/GHC/Core/Opt/OccurAnal.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Core/Ppr.hs - compiler/GHC/Core/Subst.hs - compiler/GHC/Core/Tidy.hs - compiler/GHC/Core/TyCon.hs - compiler/GHC/Core/Utils.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/CoreToStg/Prep.hs - compiler/GHC/Driver/Config.hs - + compiler/GHC/Driver/Session/Inspect.hs - compiler/GHC/HsToCore.hs - compiler/GHC/HsToCore/Breakpoints.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Syntax.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Debugger/Breakpoints.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Eval/Types.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Runtime/Interpreter/Types.hs - compiler/GHC/Stg/BcPrep.hs - compiler/GHC/Stg/FVs.hs - compiler/GHC/StgToByteCode.hs - − compiler/GHC/Types/Breakpoint.hs - compiler/GHC/Types/Tickish.hs - compiler/GHC/Unit/Module/ModGuts.hs - compiler/ghc.cabal.in - docs/users_guide/exts/doandifthenelse.rst - docs/users_guide/exts/ffi.rst - docs/users_guide/ghci.rst - ghc/GHCi/UI.hs - ghc/GHCi/UI/Monad.hs - hadrian/src/Oracles/Flag.hs - hadrian/src/Rules/Generate.hs - hadrian/src/Settings/Builders/Cabal.hs - hadrian/src/Settings/Packages.hs - hadrian/src/Settings/Program.hs - libraries/base/changelog.md - libraries/ghc-heap/GHC/Exts/Heap/Closures.hs - libraries/ghc-heap/GHC/Exts/Heap/FFIClosures_ProfilingDisabled.hsc - libraries/ghc-heap/GHC/Exts/Heap/FFIClosures_ProfilingEnabled.hsc - libraries/ghc-heap/tests/parse_tso_flags.hs - libraries/ghc-internal/src/GHC/Internal/List.hs - + libraries/ghci/GHCi/Debugger.hs - libraries/ghci/GHCi/Message.hs - libraries/ghci/GHCi/Run.hs - libraries/ghci/ghci.cabal.in - + rts/AllocArray.c - + rts/AllocArray.h - rts/Disassembler.c - rts/Exception.cmm - rts/Heap.c - rts/Interpreter.c - rts/Interpreter.h - rts/PrimOps.cmm - rts/RtsSymbols.c - rts/RtsUtils.c - rts/StgMiscClosures.cmm - rts/ThreadLabels.c - rts/Threads.c - rts/Weak.c - rts/include/Rts.h - rts/include/rts/Bytecodes.h - rts/include/rts/Constants.h - rts/include/rts/prof/CCS.h - rts/include/rts/storage/Closures.h - rts/include/rts/storage/GC.h - rts/include/rts/storage/Heap.h - rts/rts.cabal - rts/sm/Storage.c - testsuite/tests/count-deps/CountDepsAst.stdout - testsuite/tests/count-deps/CountDepsParser.stdout - + testsuite/tests/ghci.debugger/scripts/T26042b.hs - + testsuite/tests/ghci.debugger/scripts/T26042b.script - + testsuite/tests/ghci.debugger/scripts/T26042b.stdout - + testsuite/tests/ghci.debugger/scripts/T26042c.hs - + testsuite/tests/ghci.debugger/scripts/T26042c.script - + testsuite/tests/ghci.debugger/scripts/T26042c.stdout - + testsuite/tests/ghci.debugger/scripts/T26042d.hs - + testsuite/tests/ghci.debugger/scripts/T26042d.script - + testsuite/tests/ghci.debugger/scripts/T26042d.stdout - + testsuite/tests/ghci.debugger/scripts/T26042e.hs - + testsuite/tests/ghci.debugger/scripts/T26042e.script - + testsuite/tests/ghci.debugger/scripts/T26042e.stdout - + testsuite/tests/ghci.debugger/scripts/T26042f.hs - + testsuite/tests/ghci.debugger/scripts/T26042f.script - + testsuite/tests/ghci.debugger/scripts/T26042f1.stderr - + testsuite/tests/ghci.debugger/scripts/T26042f1.stdout - + testsuite/tests/ghci.debugger/scripts/T26042f2.stdout - + testsuite/tests/ghci.debugger/scripts/T26042g.hs - + testsuite/tests/ghci.debugger/scripts/T26042g.script - + testsuite/tests/ghci.debugger/scripts/T26042g.stdout - testsuite/tests/ghci.debugger/scripts/all.T - + testsuite/tests/indexed-types/should_fail/T26176.hs - + testsuite/tests/indexed-types/should_fail/T26176.stderr - testsuite/tests/indexed-types/should_fail/all.T - + testsuite/tests/typecheck/should_compile/T14010.hs - testsuite/tests/typecheck/should_compile/all.T - utils/haddock/CHANGES.md - utils/haddock/haddock-api/src/Haddock/Interface/AttachInstances.hs - utils/haddock/haddock-api/src/Haddock/Interface/Create.hs - utils/haddock/haddock-api/src/Haddock/Types.hs - utils/haddock/haddock-test/src/Test/Haddock/Config.hs - utils/haddock/html-test/ref/Bug1004.html - + utils/haddock/html-test/ref/Bug25739.html - + utils/haddock/html-test/src/Bug25739.hs The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/89c5a0885daccef17319e2c59fc1c3… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/89c5a0885daccef17319e2c59fc1c3… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][wip/marge_bot_batch_merge_job] 19 commits: debugger/rts: Allow toggling step-in per thread
by Marge Bot (＠marge-bot) 08 Jul '25

08 Jul '25

Marge Bot pushed to branch wip/marge_bot_batch_merge_job at Glasgow Haskell Compiler / GHC Commits: 8a69196e by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger/rts: Allow toggling step-in per thread The RTS global flag `rts_stop_next_breakpoint` globally sets the interpreter to stop at the immediate next breakpoint. With this commit, single step mode can additionally be set per thread in the TSO flag (TSO_STOP_NEXT_BREAKPOINT). Being able to toggle "stop at next breakpoint" per thread is an important requirement for implementing "stepping out" of a function in a multi-threaded context. And, more generally, having a per-thread flag for single-stepping paves the way for multi-threaded debugging. That said, when we want to enable "single step" mode for the whole interpreted program we still want to stop at the immediate next breakpoint, whichever thread it belongs to. That's why we also keep the global `rts_stop_next_breakpoint` flag, with `rts_enableStopNextBreakpointAll` and `rts_disableStopNextBreakpointAll` helpers. Preparation for #26042 - - - - - 73d3f864 by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 docs: Case continuation BCOs This commit documents a subtle interaction between frames for case BCOs and their parents frames. Namely, case continuation BCOs may refer to (non-local) variables that are part of the parent's frame. The note expanding a bit on these details is called [Case continuation BCOs] - - - - - d7aeddcf by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger: Implement step-out feature Implements support for stepping-out of a function (aka breaking right after returning from a function) in the interactive debugger. It also introduces a GHCi command :stepout to step-out of a function being debugged in the interpreter. The feature is described as: Stop at the first breakpoint immediately after returning from the current function scope. Known limitations: because a function tail-call does not push a stack frame, if step-out is used inside of a function that was tail-called, execution will not be returned to its caller, but rather its caller's first non-tail caller. On the other hand, it means the debugger follows the more realistic execution of the program. In the following example: .. code-block:: none f = do a b <--- (1) set breakpoint then step in here c b = do ... d <--- (2) step-into this tail call d = do ... something <--- (3) step-out here ... Stepping-out will stop execution at the `c` invokation in `f`, rather than stopping at `b`. The key idea is simple: When step-out is enabled, traverse the runtime stack until a continuation BCO is found -- and enable the breakpoint heading that BCO explicitly using its tick-index. The details are specified in `Note [Debugger: Step-out]` in `rts/Interpreter.c`. Since PUSH_ALTS BCOs (representing case continuations) were never headed by a breakpoint (unlike the case alternatives they push), we introduced the BRK_ALTS instruction to allow the debugger to set a case continuation to stop at the breakpoint heading the alternative that is taken. This is further described in `Note [Debugger: BRK_ALTS]`. Fixes #26042 - - - - - 5d9adf51 by Rodrigo Mesquita at 2025-07-08T07:39:47-04:00 debugger: Filter step-out stops by SrcSpan To implement step-out, the RTS looks for the first continuation frame on the stack and explicitly enables its entry breakpoint. However, some continuations will be contained in the function from which step-out was initiated (trivial example is a case expression). Similarly to steplocal, we will filter the breakpoints at which the RTS yields to the debugger based on the SrcSpan. When doing step-out, only stop if the breakpoint is /not/ contained in the function from which we initiated it. This is especially relevant in monadic statements such as IO which is compiled to a long chain of case expressions. See Note [Debugger: Filtering step-out stops] - - - - - 7677adcc by Cheng Shao at 2025-07-08T07:40:29-04:00 compiler: make ModBreaks serializable - - - - - 14f67c6d by Rodrigo Mesquita at 2025-07-08T07:40:29-04:00 refactor: "Inspecting the session" moved from GHC Moved utilities for inspecting the session from the GHC module to GHC.Driver.Session.Inspect Purely a clean up - - - - - 9d3f484a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Pass the HUG to readModBreaks, not HscEnv A minor cleanup. The associated history and setupBreakpoint functions are changed accordingly. - - - - - b595f713 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move readModBreaks to GHC.Runtime.Interpreter With some small docs changes - - - - - d223227a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move interpreterProfiled to Interp.Types Moves interpreterProfiled and interpreterDynamic to GHC.Runtime.Interpreter.Types from GHC.Runtime.Interpreter. - - - - - 7fdd0a3d by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Don't import GHC in Debugger.Breakpoints Remove the top-level import GHC from GHC.Runtime.Debugger.Breakpoints This makes the module dependencies more granular and cleans up the qualified imports from the code. - - - - - 5e4da31b by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refactor: Use BreakpointId in Core and Ifaces - - - - - 741ac3a8 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 stg2bc: Derive BcM via ReaderT StateT A small refactor that simplifies GHC.StgToByteCode by deriving-via the Monad instances for BcM. This is done along the lines of previous similar refactors like 72b54c0760bbf85be1f73c1a364d4701e5720465. - - - - - 0414fcc9 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refact: Split InternalModBreaks out of ModBreaks There are currently two competing ways of referring to a Breakpoint: 1. Using the Tick module + Tick index 2. Using the Info module + Info index 1. The Tick index is allocated during desugaring in `mkModBreaks`. It is used to refer to a breakpoint associated to a Core Tick. For a given Tick module, there are N Ticks indexed by Tick index. 2. The Info index is allocated during code generation (in StgToByteCode) and uniquely identifies the breakpoints at runtime (and is indeed used to determine which breakpoint was hit at runtime). Why we need both is described by Note [Breakpoint identifiers]. For every info index we used to keep a `CgBreakInfo`, a datatype containing information relevant to ByteCode Generation, in `ModBreaks`. This commit splits out the `IntMap CgBreakInfo` out of `ModBreaks` into a new datatype `InternalModBreaks`. - The purpose is to separate the `ModBreaks` datatype, which stores data associated from tick-level information which is fixed after desugaring, from the unrelated `IntMap CgBreakInfo` information accumulated during bytecode generation. - We move `ModBreaks` to GHC.HsToCore.Breakpoints The new `InternalModBreaks` simply combines the `IntMap CgBreakInfo` with `ModBreaks`. After code generation we construct an `InternalModBreaks` with the `CgBreakInfo`s we accumulated and the existing `ModBreaks` and store that in the compiled BCO in `bc_breaks`. - Note that we previously only updated the `modBreaks_breakInfo` field of `ModBreaks` at this exact location, and then stored the updated `ModBreaks` in the same `bc_breaks`. - We put this new datatype in GHC.ByteCode.Breakpoints The rest of the pipeline for which CgBreakInfo is relevant is accordingly updated to also use `InternalModBreaks` - - - - - 2a097955 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Use BreakpointIds in bytecode gen Small clean up to use BreakpointId and InternalBreakpointId more uniformly in bytecode generation rather than using Module + Ix pairs - - - - - 0515cc2f by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 ghci: Allocate BreakArrays at link time only Previously, a BreakArray would be allocated with a slot for every tick in a module at `mkModBreaks`, in HsToCore. However, this approach has a few downsides: - It interleaves interpreter behaviour (allocating arrays for breakpoints) within the desugarer - It is inflexible in the sense it is impossible for the bytecode generator to add "internal" breakpoints that can be triggered at runtime, because those wouldn't have a source tick. (This is relevant for our intended implementation plan of step-out in #26042) - It ties the BreakArray indices to the *tick* indexes, while at runtime we would rather just have the *info* indexes (currently we have both because BreakArrays are indexed by the *tick* one). Paving the way for #26042 and #26064, this commit moves the allocation of BreakArrays to bytecode-loading time -- akin to what is done for CCS arrays. Since a BreakArray is allocated only when bytecode is linked, if a breakpoint is set (e.g. `:break 10`) before the bytecode is linked, there will exist no BreakArray to trigger the breakpoint in. Therefore, the function to allocate break arrays (`allocateBreakArrays`) is exposed and also used in GHC.Runtime.Eval to allocate a break array when a breakpoint is set, if it doesn't exist yet (in the linker env). - - - - - 8016561f by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add a test for T26176 - - - - - 454cd682 by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add test for #14010 This test started to work in GHC 9.6 and has worked since. This MR just adds a regression test - - - - - 34f6b435 by Teo Camarasu at 2025-07-08T08:14:18-04:00 Implement user-defined allocation limit handlers Allocation Limits allow killing a thread if they allocate more than a user-specified limit. We extend this feature to allow more versatile behaviour. - We allow not killing the thread if the limit is exceeded. - We allow setting a custom handler to be called when the limit is exceeded. User-specified allocation limit handlers run in a fresh thread and are passed the ThreadId of the thread that exceeded its limit. We introduce utility functions for getting and setting the allocation limits of other threads, so that users can reset the limit of a thread from a handler. Both of these are somewhat coarse-grained as we are unaware of the allocations in the current nursery chunk. We provide several examples of usages in testsuite/tests/rts/T22859.hs Resolves #22859 - - - - - 099bc364 by Simon Hengel at 2025-07-08T08:14:20-04:00 Fix typo in using.rst - - - - - 114 changed files: - compiler/GHC.hs - compiler/GHC/Builtin/primops.txt.pp - compiler/GHC/ByteCode/Asm.hs - + compiler/GHC/ByteCode/Breakpoints.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Linker.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Map/Expr.hs - compiler/GHC/Core/Opt/OccurAnal.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Core/Ppr.hs - compiler/GHC/Core/Subst.hs - compiler/GHC/Core/Tidy.hs - compiler/GHC/Core/Utils.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/CoreToStg/Prep.hs - compiler/GHC/Driver/Config.hs - + compiler/GHC/Driver/Session/Inspect.hs - compiler/GHC/HsToCore.hs - compiler/GHC/HsToCore/Breakpoints.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Syntax.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Debugger/Breakpoints.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Eval/Types.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Runtime/Interpreter/Types.hs - compiler/GHC/Stg/BcPrep.hs - compiler/GHC/Stg/FVs.hs - compiler/GHC/StgToByteCode.hs - compiler/GHC/StgToCmm/Prim.hs - compiler/GHC/StgToJS/Prim.hs - − compiler/GHC/Types/Breakpoint.hs - compiler/GHC/Types/Tickish.hs - compiler/GHC/Unit/Module/ModGuts.hs - compiler/ghc.cabal.in - docs/users_guide/ghci.rst - docs/users_guide/using.rst - ghc/GHCi/UI.hs - libraries/ghc-experimental/ghc-experimental.cabal.in - + libraries/ghc-experimental/src/System/Mem/Experimental.hs - libraries/ghc-heap/GHC/Exts/Heap/Closures.hs - libraries/ghc-heap/GHC/Exts/Heap/FFIClosures_ProfilingDisabled.hsc - libraries/ghc-heap/GHC/Exts/Heap/FFIClosures_ProfilingEnabled.hsc - libraries/ghc-heap/tests/parse_tso_flags.hs - libraries/ghc-internal/ghc-internal.cabal.in - + libraries/ghc-internal/src/GHC/Internal/AllocationLimitHandler.hs - + libraries/ghci/GHCi/Debugger.hs - libraries/ghci/GHCi/Message.hs - libraries/ghci/GHCi/Run.hs - libraries/ghci/ghci.cabal.in - rts/Disassembler.c - rts/Interpreter.c - rts/Interpreter.h - rts/Prelude.h - rts/PrimOps.cmm - rts/RtsStartup.c - rts/RtsSymbols.c - rts/Schedule.c - rts/StgMiscClosures.cmm - rts/external-symbols.list.in - rts/include/rts/Bytecodes.h - rts/include/rts/Constants.h - rts/include/rts/storage/Closures.h - rts/include/rts/storage/GC.h - rts/include/rts/storage/TSO.h - rts/include/stg/MiscClosures.h - testsuite/tests/backpack/cabal/bkpcabal08/bkpcabal08.stdout - testsuite/tests/count-deps/CountDepsAst.stdout - testsuite/tests/count-deps/CountDepsParser.stdout - + testsuite/tests/ghci.debugger/scripts/T26042b.hs - + testsuite/tests/ghci.debugger/scripts/T26042b.script - + testsuite/tests/ghci.debugger/scripts/T26042b.stdout - + testsuite/tests/ghci.debugger/scripts/T26042c.hs - + testsuite/tests/ghci.debugger/scripts/T26042c.script - + testsuite/tests/ghci.debugger/scripts/T26042c.stdout - + testsuite/tests/ghci.debugger/scripts/T26042d.hs - + testsuite/tests/ghci.debugger/scripts/T26042d.script - + testsuite/tests/ghci.debugger/scripts/T26042d.stdout - + testsuite/tests/ghci.debugger/scripts/T26042e.hs - + testsuite/tests/ghci.debugger/scripts/T26042e.script - + testsuite/tests/ghci.debugger/scripts/T26042e.stdout - + testsuite/tests/ghci.debugger/scripts/T26042f.hs - + testsuite/tests/ghci.debugger/scripts/T26042f.script - + testsuite/tests/ghci.debugger/scripts/T26042f1.stderr - + testsuite/tests/ghci.debugger/scripts/T26042f1.stdout - + testsuite/tests/ghci.debugger/scripts/T26042f2.stdout - + testsuite/tests/ghci.debugger/scripts/T26042g.hs - + testsuite/tests/ghci.debugger/scripts/T26042g.script - + testsuite/tests/ghci.debugger/scripts/T26042g.stdout - testsuite/tests/ghci.debugger/scripts/all.T - + testsuite/tests/indexed-types/should_fail/T26176.hs - + testsuite/tests/indexed-types/should_fail/T26176.stderr - testsuite/tests/indexed-types/should_fail/all.T - testsuite/tests/interface-stability/base-exports.stdout - testsuite/tests/interface-stability/base-exports.stdout-javascript-unknown-ghcjs - testsuite/tests/interface-stability/base-exports.stdout-mingw32 - testsuite/tests/interface-stability/base-exports.stdout-ws-32 - testsuite/tests/interface-stability/ghc-experimental-exports.stdout - testsuite/tests/interface-stability/ghc-experimental-exports.stdout-mingw32 - testsuite/tests/interface-stability/ghc-prim-exports.stdout - testsuite/tests/interface-stability/ghc-prim-exports.stdout-mingw32 - + testsuite/tests/rts/T22859.hs - + testsuite/tests/rts/T22859.stderr - testsuite/tests/rts/all.T - + testsuite/tests/typecheck/should_compile/T14010.hs - testsuite/tests/typecheck/should_compile/all.T The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1e24e49a00b5253a2db89eea5214b6… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/1e24e49a00b5253a2db89eea5214b6… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] 2 commits: Add a test for T26176
by Marge Bot (＠marge-bot) 08 Jul '25

08 Jul '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 8016561f by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add a test for T26176 - - - - - 454cd682 by Simon Peyton Jones at 2025-07-08T07:41:13-04:00 Add test for #14010 This test started to work in GHC 9.6 and has worked since. This MR just adds a regression test - - - - - 5 changed files: - + testsuite/tests/indexed-types/should_fail/T26176.hs - + testsuite/tests/indexed-types/should_fail/T26176.stderr - testsuite/tests/indexed-types/should_fail/all.T - + testsuite/tests/typecheck/should_compile/T14010.hs - testsuite/tests/typecheck/should_compile/all.T Changes: ===================================== testsuite/tests/indexed-types/should_fail/T26176.hs ===================================== @@ -0,0 +1,32 @@ +{-# LANGUAGE DataKinds #-} +{-# LANGUAGE TypeFamilies #-} +module T26176 where + +-- In a previous version of GHC this test led to +-- a very misleading error message + +import GHC.TypeNats + +data T = X +type family FA t where FA X = 1 +type family FB t where FB X = 2 + +foo :: forall p (t :: T). p t +foo = undefined + where + a :: SNat 5 + a = b + + b :: SNat (FA t) + b = undefined + + c :: SNat 3 + c = d + + d :: SNat (FB t) + d = undefined + + x = bar @t + +bar :: FA t <= FB t => p t +bar = undefined ===================================== testsuite/tests/indexed-types/should_fail/T26176.stderr ===================================== @@ -0,0 +1,41 @@ +T26176.hs:18:7: error: [GHC-83865] + • Couldn't match type ‘FA t’ with ‘5’ + Expected: SNat 5 + Actual: SNat (FA t) + • In the expression: b + In an equation for ‘a’: a = b + In an equation for ‘foo’: + foo + = undefined + where + a :: SNat 5 + a = b + b :: SNat (FA t) + b = undefined + .... + • Relevant bindings include + b :: SNat (FA t) (bound at T26176.hs:21:3) + d :: SNat (FB t) (bound at T26176.hs:27:3) + x :: p0 t (bound at T26176.hs:29:3) + foo :: p t (bound at T26176.hs:15:1) + +T26176.hs:24:7: error: [GHC-83865] + • Couldn't match type ‘FB t’ with ‘3’ + Expected: SNat 3 + Actual: SNat (FB t) + • In the expression: d + In an equation for ‘c’: c = d + In an equation for ‘foo’: + foo + = undefined + where + a :: SNat 5 + a = b + b :: SNat (FA t) + b = undefined + .... + • Relevant bindings include + d :: SNat (FB t) (bound at T26176.hs:27:3) + x :: p0 t (bound at T26176.hs:29:3) + foo :: p t (bound at T26176.hs:15:1) + ===================================== testsuite/tests/indexed-types/should_fail/all.T ===================================== @@ -172,3 +172,4 @@ test('T21896', normal, compile_fail, ['']) test('HsBootFam', [extra_files(['HsBootFam_aux.hs','HsBootFam_aux.hs-boot'])], multimod_compile_fail, ['HsBootFam', '']) test('BadFamInstDecl', [extra_files(['BadFamInstDecl_aux.hs'])], multimod_compile_fail, ['BadFamInstDecl', '']) test('T19773', [], multimod_compile_fail, ['T19773', '-Wall']) +test('T26176', normal, compile_fail, ['']) ===================================== testsuite/tests/typecheck/should_compile/T14010.hs ===================================== @@ -0,0 +1,44 @@ +{-# LANGUAGE NoImplicitPrelude, PolyKinds, DataKinds, TypeFamilies, + UndecidableSuperClasses, RankNTypes, TypeOperators, + FlexibleContexts, TypeSynonymInstances, FlexibleInstances, + UndecidableInstances #-} +module Monolith where + +import Data.Kind (Type) +import GHC.Exts (Constraint) + +type family (~>) :: c -> c -> Type + +type instance (~>) = (->) +type instance (~>) = ArrPair + +type family Fst (p :: (a, b)) :: a where + Fst '(x, _) = x + +type family Snd (p :: (a, b)) :: b where + Snd '(_, y) = y + +data ArrPair a b = ArrPair (Fst a ~> Fst b) (Snd a ~> Snd b) + +type family Super c :: Constraint where + Super Type = () + Super (c, d) = (Category c, Category d) + +class Super cat => Category cat where + id :: forall (a :: cat). a ~> a + +instance Category Type where + id = \x -> x + +instance (Category c, Category d) => Category (c, d) where + id = ArrPair id id + +-- The commented out version worked +-- class Category (c, d) => Functor (f :: c -> d) where +class (Category c, Category d) => Functor (f :: c -> d) where + map :: (a ~> b) -> (f a ~> f b) + +data OnSnd f a b = OnSnd (f '(a, b)) + +instance Functor (f :: (c, d) -> Type) => Functor (OnSnd f a) where + map f (OnSnd x) = OnSnd (map (ArrPair id f) x) ===================================== testsuite/tests/typecheck/should_compile/all.T ===================================== @@ -939,3 +939,4 @@ test('T25960', normal, compile, ['']) test('T26020', normal, compile, ['']) test('T26020a', [extra_files(['T26020a_help.hs'])], multimod_compile, ['T26020a', '-v0']) test('T25992', normal, compile, ['']) +test('T14010', normal, compile, ['']) View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0515cc2fb2cd54d09786ab23cf9994… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/0515cc2fb2cd54d09786ab23cf9994… You're receiving this email because of your account on gitlab.haskell.org.

1 0

[Git][ghc/ghc][master] 11 commits: compiler: make ModBreaks serializable
by Marge Bot (＠marge-bot) 08 Jul '25

08 Jul '25

Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC Commits: 7677adcc by Cheng Shao at 2025-07-08T07:40:29-04:00 compiler: make ModBreaks serializable - - - - - 14f67c6d by Rodrigo Mesquita at 2025-07-08T07:40:29-04:00 refactor: "Inspecting the session" moved from GHC Moved utilities for inspecting the session from the GHC module to GHC.Driver.Session.Inspect Purely a clean up - - - - - 9d3f484a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Pass the HUG to readModBreaks, not HscEnv A minor cleanup. The associated history and setupBreakpoint functions are changed accordingly. - - - - - b595f713 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move readModBreaks to GHC.Runtime.Interpreter With some small docs changes - - - - - d223227a by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Move interpreterProfiled to Interp.Types Moves interpreterProfiled and interpreterDynamic to GHC.Runtime.Interpreter.Types from GHC.Runtime.Interpreter. - - - - - 7fdd0a3d by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Don't import GHC in Debugger.Breakpoints Remove the top-level import GHC from GHC.Runtime.Debugger.Breakpoints This makes the module dependencies more granular and cleans up the qualified imports from the code. - - - - - 5e4da31b by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refactor: Use BreakpointId in Core and Ifaces - - - - - 741ac3a8 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 stg2bc: Derive BcM via ReaderT StateT A small refactor that simplifies GHC.StgToByteCode by deriving-via the Monad instances for BcM. This is done along the lines of previous similar refactors like 72b54c0760bbf85be1f73c1a364d4701e5720465. - - - - - 0414fcc9 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 refact: Split InternalModBreaks out of ModBreaks There are currently two competing ways of referring to a Breakpoint: 1. Using the Tick module + Tick index 2. Using the Info module + Info index 1. The Tick index is allocated during desugaring in `mkModBreaks`. It is used to refer to a breakpoint associated to a Core Tick. For a given Tick module, there are N Ticks indexed by Tick index. 2. The Info index is allocated during code generation (in StgToByteCode) and uniquely identifies the breakpoints at runtime (and is indeed used to determine which breakpoint was hit at runtime). Why we need both is described by Note [Breakpoint identifiers]. For every info index we used to keep a `CgBreakInfo`, a datatype containing information relevant to ByteCode Generation, in `ModBreaks`. This commit splits out the `IntMap CgBreakInfo` out of `ModBreaks` into a new datatype `InternalModBreaks`. - The purpose is to separate the `ModBreaks` datatype, which stores data associated from tick-level information which is fixed after desugaring, from the unrelated `IntMap CgBreakInfo` information accumulated during bytecode generation. - We move `ModBreaks` to GHC.HsToCore.Breakpoints The new `InternalModBreaks` simply combines the `IntMap CgBreakInfo` with `ModBreaks`. After code generation we construct an `InternalModBreaks` with the `CgBreakInfo`s we accumulated and the existing `ModBreaks` and store that in the compiled BCO in `bc_breaks`. - Note that we previously only updated the `modBreaks_breakInfo` field of `ModBreaks` at this exact location, and then stored the updated `ModBreaks` in the same `bc_breaks`. - We put this new datatype in GHC.ByteCode.Breakpoints The rest of the pipeline for which CgBreakInfo is relevant is accordingly updated to also use `InternalModBreaks` - - - - - 2a097955 by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 cleanup: Use BreakpointIds in bytecode gen Small clean up to use BreakpointId and InternalBreakpointId more uniformly in bytecode generation rather than using Module + Ix pairs - - - - - 0515cc2f by Rodrigo Mesquita at 2025-07-08T07:40:30-04:00 ghci: Allocate BreakArrays at link time only Previously, a BreakArray would be allocated with a slot for every tick in a module at `mkModBreaks`, in HsToCore. However, this approach has a few downsides: - It interleaves interpreter behaviour (allocating arrays for breakpoints) within the desugarer - It is inflexible in the sense it is impossible for the bytecode generator to add "internal" breakpoints that can be triggered at runtime, because those wouldn't have a source tick. (This is relevant for our intended implementation plan of step-out in #26042) - It ties the BreakArray indices to the *tick* indexes, while at runtime we would rather just have the *info* indexes (currently we have both because BreakArrays are indexed by the *tick* one). Paving the way for #26042 and #26064, this commit moves the allocation of BreakArrays to bytecode-loading time -- akin to what is done for CCS arrays. Since a BreakArray is allocated only when bytecode is linked, if a breakpoint is set (e.g. `:break 10`) before the bytecode is linked, there will exist no BreakArray to trigger the breakpoint in. Therefore, the function to allocate break arrays (`allocateBreakArrays`) is exposed and also used in GHC.Runtime.Eval to allocate a break array when a breakpoint is set, if it doesn't exist yet (in the linker env). - - - - - 42 changed files: - compiler/GHC.hs - compiler/GHC/ByteCode/Asm.hs - + compiler/GHC/ByteCode/Breakpoints.hs - compiler/GHC/ByteCode/Instr.hs - compiler/GHC/ByteCode/Linker.hs - compiler/GHC/ByteCode/Types.hs - compiler/GHC/Core/FVs.hs - compiler/GHC/Core/Lint.hs - compiler/GHC/Core/Map/Expr.hs - compiler/GHC/Core/Opt/OccurAnal.hs - compiler/GHC/Core/Opt/Simplify/Iteration.hs - compiler/GHC/Core/Ppr.hs - compiler/GHC/Core/Subst.hs - compiler/GHC/Core/Tidy.hs - compiler/GHC/Core/Utils.hs - compiler/GHC/CoreToIface.hs - compiler/GHC/CoreToStg.hs - compiler/GHC/CoreToStg/Prep.hs - + compiler/GHC/Driver/Session/Inspect.hs - compiler/GHC/HsToCore.hs - compiler/GHC/HsToCore/Breakpoints.hs - compiler/GHC/HsToCore/Ticks.hs - compiler/GHC/Iface/Syntax.hs - compiler/GHC/Iface/Tidy.hs - compiler/GHC/IfaceToCore.hs - compiler/GHC/Linker/Loader.hs - compiler/GHC/Linker/Types.hs - compiler/GHC/Runtime/Debugger/Breakpoints.hs - compiler/GHC/Runtime/Eval.hs - compiler/GHC/Runtime/Eval/Types.hs - compiler/GHC/Runtime/Interpreter.hs - compiler/GHC/Runtime/Interpreter/Types.hs - compiler/GHC/Stg/BcPrep.hs - compiler/GHC/Stg/FVs.hs - compiler/GHC/StgToByteCode.hs - − compiler/GHC/Types/Breakpoint.hs - compiler/GHC/Types/Tickish.hs - compiler/GHC/Unit/Module/ModGuts.hs - compiler/ghc.cabal.in - ghc/GHCi/UI.hs - testsuite/tests/count-deps/CountDepsAst.stdout - testsuite/tests/count-deps/CountDepsParser.stdout The diff was not included because it is too large. View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5d9adf5103e7b99cf0c8c7d816d5a9… -- View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5d9adf5103e7b99cf0c8c7d816d5a9… You're receiving this email because of your account on gitlab.haskell.org.

1 0