reduce size of stage1 compiler with flags:

 minimal:          build without bytecode, js, wasm backends
 no-uncommon-ncgs: builf only the x86 and aarch64 ncgs

Author: luite

cabal.project.stage1

@@ -48,7 +48,7 @@ package *
   executable-static: False
 
 package ghc
-  flags: +bootstrap
+  flags: +bootstrap +minimal +no-uncommon-ncgs
 
 package ghci
   flags: +bootstrap

compiler/GHC/ByteCode/Linker.hs

@@ -1,3 +1,4 @@
+{-# LANGUAGE CPP #-}
 {-# LANGUAGE FlexibleInstances     #-}
 {-# LANGUAGE MagicHash             #-}
 {-# LANGUAGE MultiParamTypeClasses #-}
@@ -20,6 +21,7 @@ where
 
 import GHC.Prelude
 
+#if !defined(MINIMAL)
 import GHC.Runtime.Interpreter
 import GHC.ByteCode.Types
 import GHCi.RemoteTypes
@@ -238,3 +240,35 @@ primopToCLabel primop suffix = concat
     , zString (zEncodeFS (occNameFS (primOpOcc primop)))
     , '_':suffix
     ]
+#else
+import GHC.Runtime.Interpreter
+import GHC.ByteCode.Types hiding (ItblEnv, AddrEnv)
+import GHC.Linker.Types
+import GHC.Types.Name.Env
+import GHC.Types.Name
+import GHC.Data.FastString
+import GHCi.ResolvedBCO
+import Foreign.Ptr
+import GHC.Utils.Panic
+import GHC.Utils.Outputable
+
+linkBCO
+  :: Interp
+  -> PkgsLoaded
+  -> LinkerEnv
+  -> LinkedBreaks
+  -> NameEnv Int
+  -> UnlinkedBCO
+  -> IO ResolvedBCO
+linkBCO _ _ _ _ _ _ = panic "linkBCO: not available in MINIMAL build"
+
+lookupStaticPtr :: Interp -> FastString -> IO (Ptr ())
+lookupStaticPtr _ _ = panic "lookupStaticPtr: not available in MINIMAL build"
+
+lookupIE :: Interp -> PkgsLoaded -> ItblEnv -> Name -> IO (Ptr ())
+lookupIE _ _ _ _ = panic "lookupIE: not available in MINIMAL build"
+
+linkFail :: String -> SDoc -> IO a
+linkFail _ _ = panic "linkFail: not available in MINIMAL build"
+#endif
+

compiler/GHC/CmmToAsm.hs

@@ -1,3 +1,5 @@
+{-# LANGUAGE CPP #-}
+
 -- -----------------------------------------------------------------------------
 --
 -- (c) The University of Glasgow 1993-2004
@@ -27,7 +29,7 @@
 -- possible.
 --
 -- The machine-dependent bits are generally contained under
---  GHC/CmmToAsm/<Arch>/* and generally breaks down as follows:
+--  GHC/CmmToAsm/<Arch> and generally breaks down as follows:
 --
 --   * "Regs": Everything about the target platform's machine
 --     registers (and immediate operands, and addresses, which tend to
@@ -64,11 +66,18 @@ where
 import GHC.Prelude hiding (head)
 
 import qualified GHC.CmmToAsm.X86   as X86
-import qualified GHC.CmmToAsm.PPC   as PPC
 import qualified GHC.CmmToAsm.AArch64 as AArch64
-import qualified GHC.CmmToAsm.Wasm as Wasm32
+
+#if !defined(NO_UNCOMMON_NCGS)
+import qualified GHC.CmmToAsm.PPC   as PPC
 import qualified GHC.CmmToAsm.RV64  as RV64
 import qualified GHC.CmmToAsm.LA64 as LA64
+#endif
+
+#if !defined(MINIMAL)
+import qualified GHC.CmmToAsm.Wasm as Wasm32
+#endif
+
 
 import GHC.CmmToAsm.Reg.Liveness
 import qualified GHC.CmmToAsm.Reg.Linear                as Linear
@@ -143,19 +152,27 @@ nativeCodeGen logger ts config modLoc h cmms
    in case platformArch platform of
       ArchX86       -> nCG' (X86.ncgX86     config)
       ArchX86_64    -> nCG' (X86.ncgX86_64  config)
+#if !defined(NO_UNCOMMON_NCGS)
       ArchPPC       -> nCG' (PPC.ncgPPC     config)
       ArchPPC_64 _  -> nCG' (PPC.ncgPPC     config)
+#endif
       ArchS390X     -> panic "nativeCodeGen: No NCG for S390X"
       ArchARM {}    -> panic "nativeCodeGen: No NCG for ARM"
       ArchAArch64   -> nCG' (AArch64.ncgAArch64 config)
       ArchAlpha     -> panic "nativeCodeGen: No NCG for Alpha"
       ArchMipseb    -> panic "nativeCodeGen: No NCG for mipseb"
       ArchMipsel    -> panic "nativeCodeGen: No NCG for mipsel"
+#if !defined(NO_UNCOMMON_NCGS)
       ArchRISCV64   -> nCG' (RV64.ncgRV64 config)
       ArchLoongArch64 -> nCG' (LA64.ncgLA64 config)
+#endif
       ArchUnknown   -> panic "nativeCodeGen: No NCG for unknown arch"
       ArchJavaScript-> panic "nativeCodeGen: No NCG for JavaScript"
+#if !defined(MINIMAL)
       ArchWasm32    -> Wasm32.ncgWasm config logger platform ts modLoc h cmms
+#endif
+      _             -> panic "nativeCodeGen: No NCG for this architecture"
+
 
 -- | Data accumulated during code generation. Mostly about statistics,
 -- but also collects debug data for DWARF generation.

compiler/GHC/Driver/Backend.hs

@@ -1,4 +1,5 @@
 {-# LANGUAGE MultiWayIf, LambdaCase #-}
+{-# LANGUAGE CPP #-}
 
 {-|
 Module      : GHC.Driver.Backend
@@ -209,14 +210,21 @@ platformNcgSupported platform = if
       ncgValidArch = case platformArch platform of
          ArchX86       -> True
          ArchX86_64    -> True
+#if !defined(NO_UNCOMMON_NCGS)
          ArchPPC       -> True
          ArchPPC_64 {} -> True
+#endif
          ArchAArch64   -> True
+#if !defined(MINIMAL)
          ArchWasm32    -> True
+#endif
+#if !defined(NO_UNCOMMON_NCGS)
          ArchRISCV64   -> True
          ArchLoongArch64 -> True
+#endif
          _             -> False
 
+
 -- | Is the platform supported by the JS backend?
 platformJSSupported :: Platform -> Bool
 platformJSSupported platform

compiler/GHC/Driver/Main.hs

@@ -1,3 +1,4 @@
+{-# LANGUAGE CPP #-}
 {-# LANGUAGE LambdaCase #-}
 {-# LANGUAGE NondecreasingIndentation #-}
 {-# LANGUAGE GADTs #-}
@@ -134,18 +135,24 @@ import GHC.Driver.Config.Stg.Pipeline (initStgPipelineOpts)
 import GHC.Driver.Config.StgToCmm  (initStgToCmmConfig)
 import GHC.Driver.Config.Cmm       (initCmmConfig)
 import GHC.Driver.LlvmConfigCache  (initLlvmConfigCache)
+#if !defined(MINIMAL)
 import GHC.Driver.Config.StgToJS  (initStgToJSConfig)
+#endif
 import GHC.Driver.Config.Diagnostic
 import GHC.Driver.Config.Tidy
 import GHC.Driver.Hooks
 import GHC.Driver.GenerateCgIPEStub (generateCgIPEStub, lookupEstimatedTicks)
 
 import GHC.Runtime.Context
 import GHC.Runtime.Interpreter
+#if !defined(MINIMAL)
 import GHC.Runtime.Interpreter.JS
+#endif
 import GHC.Runtime.Loader      ( initializePlugins )
 import GHCi.RemoteTypes
+#if !defined(MINIMAL)
 import GHC.ByteCode.Types
+#endif
 
 import GHC.Linker.Loader
 import GHC.Linker.Types
@@ -157,11 +164,13 @@ import GHC.Hs.Stats         ( ppSourceStats )
 
 import GHC.HsToCore
 
+#if !defined(MINIMAL)
 import GHC.StgToByteCode    ( byteCodeGen )
 import GHC.StgToJS          ( stgToJS )
 import GHC.StgToJS.Ids
 import GHC.StgToJS.Types
 import GHC.JS.Syntax
+#endif
 
 import GHC.IfaceToCore  ( typecheckIface, typecheckWholeCoreBindings )
 
@@ -693,9 +702,13 @@ hsc_typecheck keep_rn mod_summary mb_rdr_module = do
                     Nothing -> hscParse' mod_summary
             tc_result0 <- tcRnModule' mod_summary keep_rn' hpm
             if hsc_src == HsigFile
+#if defined(MINIMAL)
+                then panic "hscSimpleIface not available in MINIMAL build"
+#else
                 then do (iface, _) <- liftIO $ hscSimpleIface hsc_env Nothing tc_result0 mod_summary
                         ioMsgMaybe $ hoistTcRnMessage $
                             tcRnMergeSignatures hsc_env hpm tc_result0 iface
+#endif
                 else return tc_result0
     -- TODO are we extracting anything when we merely instantiate a signature?
     -- If not, try to move this into the "else" case above.
@@ -732,11 +745,19 @@ tcRnModule' sum save_rn_syntax mod = do
     if not (safeHaskellOn dflags)
          || (safeInferOn dflags && not allSafeOK)
       -- if safe Haskell off or safe infer failed, mark unsafe
+#if defined(MINIMAL)
+      then return tcg_res
+#else
       then markUnsafeInfer tcg_res whyUnsafe
+#endif
 
       -- module (could be) safe, throw warning if needed
       else do
+#if defined(MINIMAL)
+          let tcg_res' = tcg_res
+#else
           tcg_res' <- hscCheckSafeImports tcg_res
+#endif
           safe <- liftIO $ readIORef (tcg_safe_infer tcg_res')
           when safe $
             case wopt Opt_WarnSafe dflags of
@@ -992,7 +1013,11 @@ initModDetails hsc_env iface =
     -- in one-shot mode, since we're not going to do
     -- any further typechecking.  It's much more useful
     -- in make mode, since this HMI will go into the HPT.
+#if defined(MINIMAL)
+    return (panic "genModDetails not available in MINIMAL build")
+#else
     genModDetails hsc_env iface
+#endif
 
 -- | Modify flags such that objects are compiled for the interpreter's way.
 -- This is necessary when building foreign objects for Template Haskell, since
@@ -1046,6 +1071,9 @@ loadIfaceByteCode ::
   ModLocation ->
   TypeEnv ->
   Maybe (IO Linkable)
+#if defined(MINIMAL)
+loadIfaceByteCode _ _ _ _ = Nothing
+#else
 loadIfaceByteCode hsc_env iface location type_env =
   compile <$> iface_core_bindings iface location
   where
@@ -1057,13 +1085,17 @@ loadIfaceByteCode hsc_env iface location type_env =
       if_time <- modificationTimeIfExists (ml_hi_file location)
       time <- maybe getCurrentTime pure if_time
       return $! Linkable time (mi_module iface) parts
+#endif
 
 loadIfaceByteCodeLazy ::
   HscEnv ->
   ModIface ->
   ModLocation ->
   TypeEnv ->
   IO (Maybe Linkable)
+#if defined(MINIMAL)
+loadIfaceByteCodeLazy _ _ _ _ = return Nothing
+#else
 loadIfaceByteCodeLazy hsc_env iface location type_env =
   case iface_core_bindings iface location of
     Nothing -> return Nothing
@@ -1078,6 +1110,7 @@ loadIfaceByteCodeLazy hsc_env iface location type_env =
       if_time <- modificationTimeIfExists (ml_hi_file location)
       time <- maybe getCurrentTime pure if_time
       return $!Linkable time (mi_module iface) parts
+#endif
 
 -- | If the 'Linkable' contains Core bindings loaded from an interface, replace
 -- them with a lazy IO thunk that compiles them to bytecode and foreign objects,
@@ -1107,6 +1140,9 @@ initWholeCoreBindings ::
   ModDetails ->
   Linkable ->
   IO Linkable
+#if defined(MINIMAL)
+initWholeCoreBindings _ _ _ l = return l
+#else
 initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = do
   Linkable utc_time this_mod <$> mapM (go hsc_env) uls
   where
@@ -1119,6 +1155,7 @@ initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = d
       l -> pure l
 
     type_env = md_types details
+#endif
 
 -- | Hydrate interface Core bindings and compile them to bytecode.
 --
@@ -1134,6 +1171,7 @@ initWholeCoreBindings hsc_env iface details (Linkable utc_time this_mod uls) = d
 --
 -- 3. Generating bytecode and foreign objects from the results of the previous
 --    steps using the usual pipeline actions.
+#if !defined(MINIMAL)
 compileWholeCoreBindings ::
   HscEnv ->
   TypeEnv ->
@@ -1168,6 +1206,7 @@ compileWholeCoreBindings hsc_env type_env wcb = do
     WholeCoreBindings {wcb_module, wcb_mod_location, wcb_foreign} = wcb
 
     logger = hsc_logger hsc_env
+#endif
 
 {-
 Note [ModDetails and --make mode]
@@ -1263,6 +1302,9 @@ hscDesugarAndSimplify summary (FrontendTypecheck tc_result) tc_warnings mb_old_h
   case mb_desugar of
       -- Just cause we desugared doesn't mean we are generating code, see above.
       Just desugared_guts | backendGeneratesCode bcknd -> do
+#if defined(MINIMAL)
+          panic "hscSimplify' not available in MINIMAL build"
+#else
           plugins <- liftIO $ readIORef (tcg_th_coreplugins tc_result)
           simplified_guts <- hscSimplify' plugins desugared_guts
 
@@ -1280,12 +1322,16 @@ hscDesugarAndSimplify summary (FrontendTypecheck tc_result) tc_warnings mb_old_h
                              hscs_partial_iface = partial_iface,
                              hscs_old_iface_hash = mb_old_hash
                            }
+#endif
 
       Just desugared_guts | gopt Opt_WriteIfSimplifiedCore dflags -> do
           -- If -fno-code is enabled (hence we fall through to this case)
           -- Running the simplifier once is necessary before doing byte code generation
           -- in order to inline data con wrappers but we honour whatever level of simplificication the
           -- user requested. See #22008 for some discussion.
+#if defined(MINIMAL)
+          panic "hscSimplify' not available in MINIMAL build"
+#else
           plugins <- liftIO $ readIORef (tcg_th_coreplugins tc_result)
           simplified_guts <- hscSimplify' plugins desugared_guts
           (cg_guts, _) <-
@@ -1301,11 +1347,15 @@ hscDesugarAndSimplify summary (FrontendTypecheck tc_result) tc_warnings mb_old_h
           if ms_mod summary == gHC_PRIM
             then return $ HscUpdate (getGhcPrimIface (hsc_hooks hsc_env))
             else return $ HscUpdate iface
+#endif
 
 
       -- We are not generating code or writing an interface with simplified core so we can skip simplification
       -- and generate a simple interface.
       _ -> do
+#if defined(MINIMAL)
+        panic "hscSimpleIface not available in MINIMAL build"
+#else
         (iface, _details) <- liftIO $
           hscSimpleIface hsc_env Nothing tc_result summary
 
@@ -1316,6 +1366,7 @@ hscDesugarAndSimplify summary (FrontendTypecheck tc_result) tc_warnings mb_old_h
         if ms_mod summary == gHC_PRIM
           then return $ HscUpdate (getGhcPrimIface (hsc_hooks hsc_env))
           else return $ HscUpdate iface
+#endif
 
 {-
 Note [Writing interface files]

compiler/GHC/Linker/Deps.hs

@@ -199,9 +199,12 @@ get_link_deps opts pls maybe_normal_osuf span mods = do
               DotO file ForeignObject -> pure (DotO file ForeignObject)
               DotA fp    -> panic ("adjust_ul DotA " ++ show fp)
               DotDLL fp  -> panic ("adjust_ul DotDLL " ++ show fp)
+#if !defined(MINIMAL)
               BCOs {}    -> pure part
               LazyBCOs{} -> pure part
+#endif
               CoreBindings WholeCoreBindings {wcb_module} ->
+
                 pprPanic "Unhydrated core bindings" (ppr wcb_module)