Vm-dev January 2021

vm-dev@lists.squeakfoundation.org

20 participants
161 discussions

VM Maker: VMMaker.oscog-eem.2937.mcz
by commits＠source.squeak.org 20 Jan '21

20 Jan '21

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.2937.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.2937 Author: eem Time: 19 January 2021, 8:45:06.642772 pm UUID: 9f7b8c1a-4f51-4b69-92f4-37bbdc45b73a Ancestors: VMMaker.oscog-eem.2936 doWithIndex: => withIndexDo: =============== Diff against VMMaker.oscog-eem.2936 =============== Item was changed: ----- Method: CCodeGenerator>>addClass: (in category 'public') ----- addClass: aClass "Add the variables and methods of the given class to the code base." aClass prepareToBeAddedToCodeGenerator: self. self checkClassForNameConflicts: aClass. self addClassVarsFor: aClass. "ikp..." self addPoolVarsFor: aClass. (aClass inheritsFrom: VMStructType) ifFalse: [variables addAll: (self instVarNamesForClass: aClass)]. self retainMethods: (aClass requiredMethodNames: self options). 'Adding Class ' , aClass name , '...' displayProgressAt: Sensor cursorPoint from: 0 to: aClass selectors size during: [:bar | + aClass selectors withIndexDo: - aClass selectors doWithIndex: [:sel :i | | source | bar value: i. self addMethodFor: aClass selector: sel]]. aClass declareCVarsIn: self! Item was changed: ----- Method: CCodeGenerator>>addStructClass: (in category 'public') ----- addStructClass: aClass "Add the non-accessor methods of the given struct class to the code base." aClass prepareToBeAddedToCodeGenerator: self. self addClassVarsFor: aClass. self addPoolVarsFor: aClass. self retainMethods: (aClass requiredMethodNames: self options). 'Adding Class ' , aClass name , '...' displayProgressAt: Sensor cursorPoint from: 0 to: aClass selectors size during: [:bar | + aClass selectors withIndexDo: - aClass selectors doWithIndex: [:sel :i | | source | bar value: i. self addStructMethodFor: aClass selector: sel]]. aClass declareCVarsIn: self! Item was changed: ----- Method: CCodeGenerator>>doBasicInlining: (in category 'inlining') ----- doBasicInlining: inlineFlagOrSymbol "Inline the bodies of all methods that are suitable for inlining. This method does only the basic inlining suitable for both the core VM and plugins - no bytecode inlining etc" | pass progress | self collectInlineList: inlineFlagOrSymbol. pass := 0. progress := true. [progress] whileTrue: [ "repeatedly attempt to inline methods until no further progress is made" progress := false. ('Inlining pass ', (pass := pass + 1) printString, '...') displayProgressAt: Sensor cursorPoint from: 0 to: methods size during: [:bar | + (self sortMethods: methods) withIndexDo: [:m :i | - (self sortMethods: methods) doWithIndex: [:m :i | bar value: i. currentMethod := m. (m tryToInlineMethodsIn: self) ifTrue: [progress := true]]]]. ! Item was changed: ----- Method: CCodeGenerator>>emitCMethods:on: (in category 'C code generator') ----- emitCMethods: methodList on: aStream 'Writing Translated Code...' displayProgressAt: Sensor cursorPoint from: 0 to: methods size during: [:bar | + methodList withIndexDo: - methodList doWithIndex: [ :m :i | bar value: i. (m isRealMethod and: [self shouldGenerateMethod: m]) ifTrue: [m emitCCodeOn: aStream generator: self]]]! Item was changed: ----- Method: CCodeGenerator>>removeAssertions (in category 'inlining') ----- removeAssertions "Remove all assertions in method bodies. This is for the benefit of inlining, which fails to recognise and disregard empty method bodies when checking the inlinability of sends." | newMethods | newMethods := Dictionary new. 'Removing assertions...' displayProgressAt: Sensor cursorPoint from: 0 to: methods size during: [ :bar | + methods withIndexDo: [ :m :i | - methods doWithIndex: [ :m :i | bar value: i. m isAssertion ifFalse: [ newMethods at: m selector put: m. m removeAssertions]]]. methods := newMethods.! Item was changed: ----- Method: CogARMv8Compiler class>>initialize (in category 'class initialization') ----- initialize "Initialize various ARM64 instruction-related constants." "self initialize" "main registers; a minor complication in reading the doc. ARM refer to the 64bit registers as X0...30 and use R0...30 to refer to the 32bit lower halves. They also use a whole suite of names for the floating point/SIMD registers. See ARMARM DDI0487 B1.2.1 etc for the gory details. Note that R30 (yes, yes, X30) is used as the link register and as such is not really a general purpose register. Also note that 31 in a general register field means R31, and that in most of these instructions R31 is the zero register named XZR in ARM doc. but in the rest of these instructions R31 the effective SP register. XZR is a pseudo-register that always reads as 0 and writes to /dev/null. And note that unlike the ARM32, there is no general purpose register for the PC; a big difference. See ARMARM DDI0487 C1.2.5. wrt to both the lack of a PC register and the XZR/SP distinction." "and initialize most sets of variables that run from 0 to N - 1..." #( "General registers, 0 to 31. We stick with R0...30 to refer to the 64 bit general regs and D0...31 (note the extra reg here!!) for the FP/SIMD regs" (R0 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31) "Floating-point registers, 0 to 31" (D0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D16 D17 D18 D19 D20 D21 D22 D23 D24 D25 D26 D27 D28 D29 D30 D31) "C argument registers, 0 to 6" (CArg0Reg CArg1Reg CArg2Reg CArg3Reg CArg4Reg CArg5Reg CArg6Reg) "Condition Codes 0 to 16. Note that cc=16rF is mapped back to AL in AARCH64. Generally it shouldn't be used." (EQ NE CS CC MI PL VS VC HI LS GE LT GT LE AL) "Logical Op Codes" (LogicalAnd LogicalOr LogicalXor LogicalAndS) "Arithmetic Opcodes" "ADD (shifted register) on page C6-763 10001011 ADDS (shifted register) on page C6-771 10101011 CMN (shifted register) on page C6-854 10101011 Rd=XZR SUB (shifted register) on page C6-1313 11001011 SUBS (shifted register) on page C6-1323 11101011 CMP (shifted register) on page C6-860 11101011 Rd=XZR NEG (shifted register) on page C6-1114 11001011 Rn=XZR NEGS on page C6-1116 11101011 Rn=XZR" (ArithmeticAdd ArithmeticAddS ArithmeticSub ArithmeticSubS) "Extension Methods " (UXTB UXTH UXTW UXTX "a.k.a. LSL" SXTB SXTH SXTW SXTX)) do: [:classVarNames| + classVarNames withIndexDo: - classVarNames doWithIndex: [:k :v| CogARMv8Compiler classPool at: k put: v - 1]]. SP := XZR := R31. LR := R30. FP := R29. "DC variant selectors; see concretizeDataCacheControl" DC_CISW := 13. DC_CIVAC := 14. DC_CSW := 15. DC_CVAC := 16. DC_CVAU := 19. DC_ISW := 26. DC_IVAC := 27. DC_ZVA := 28. "IC variant selectors; see concretizeInstructionCacheControl" IC_IALLU := 0. IC_IALLUIS := 1. IC_IVAU := 2. "DSB domains and types C6.2.81 DSB C6-891" DSB_OSH := 0. "Domain_OuterSharable" DSB_NSH := 1. "Domain_NonSharable" DSB_ISH := 2. "Domain_InnerSharable" DSB_SY := 3. "Domain_FullSystem" DSB_ALLSY := 0. "Types_All; domain = Domain_FullSystem" DSB_READS := 1. "Types_Reads" DSB_WRITES := 2. "Types_Writes" DSB_ALL := 3. "Types_All; domain ~= Domain_FullSystem" "Specific instructions" self initializeSpecificOpcodes: #(MulRRR MulOverflowRRR SMULHRRR DivRRR MSubRRR "N.B. ARMv8 has MSUBRRRR but we only support three operands" MoveAwRR MoveRRAw NativePushRR NativePopRR "these map to ldp/stp" "Cache control and memory barrier" "B2.3.7 Memory barriers B2-124" DC IC DMB DSB ISB MRS_CTR_EL0 MRS_ID_AA64ISAR0_EL1 CASAL CBNZ CBZ CCMPNE CSET CLREX LDAXR STLXR STLR) in: thisContext method! Item was changed: ----- Method: CogX64Compiler class>>initialize (in category 'class initialization') ----- initialize "Initialize various x64 instruction-related constants. [1] IA-32 Intel® Architecture Software Developer's Manual Volume 2A: Instruction Set Reference, A-M" "CogX64Compiler initialize" self ~~ CogX64Compiler ifTrue: [^self]. (InitializationOptions ifNil: [Dictionary new]) at: #ABI ifPresent: [:abi| SysV := abi asUppercase ~= #WIN64 and: [abi asUppercase ~= #'_WIN64']] ifAbsent: [SysV := true]. "Default ABI; set to true for SysV, false for WIN64/_WIN64" "Initialize most sets of variables that run from 0 to N - 1..." #( "16 integer registers, rcx, rdx, rbx bizarrely inconsistent" (RAX RCX RDX RBX RSP RBP RSI RDI R8 R9 R10 R11 R12 R13 R14 R15) "16 lower half floating point registers" (XMM0L XMM1L XMM2L XMM3L XMM4L XMM5L XMM6L XMM7L XMM8L XMM9L XMM10L XMM11L XMM12L XMM13L XMM14L XMM15L) "Mod R/M Mod fields. See [1] Sec 2.4, 2.5 & 2.6 & Table 2-2" (ModRegInd ModRegRegDisp8 ModRegRegDisp32 ModReg ModRegIndSIB ModRegIndDisp32) "SIB Scaled Index modes. See [1] Sec 2.4, 2.5 & 2.6 & Table 2-3" (SIB1 SIB2 SIB4 SIB8)) do: [:classVarNames| + classVarNames withIndexDo: - classVarNames doWithIndex: [:k :v| self classPool at: k put: v - 1]]. "Specific instructions" self initializeSpecificOpcodes: #(CPUID "Arithmetic:" CDQ IDIVR IMULRR XCHGRR "Bit count/data movement:" CLD REP MOVSB MOVSQ BSR "Multi-processing:" LFENCE MFENCE SFENCE LOCK CMPXCHGRMr SETE MoveRAwNoVBR) in: thisContext method extraClassVarNames: #(CArg0Reg CArg1Reg CArg2Reg CArg3Reg)! Item was changed: ----- Method: InterpreterSimulator>>ioLoadFunction:From: (in category 'plugin support') ----- ioLoadFunction: functionString From: pluginString "Load and return the requested function from a module" | plugin fnSymbol | fnSymbol := functionString asSymbol. transcript cr; show:'Looking for ', functionString, ' in '. pluginString isEmpty ifTrue:[transcript show: 'vm'] ifFalse:[transcript show: pluginString]. plugin := pluginList detect:[:any| any key = pluginString asString] ifNone:[self loadNewPlugin: pluginString]. plugin ifNil:[ "Transcript cr; show:'Failed ... no plugin found'." ^ 0]. plugin := plugin value. + mappedPluginEntries withIndexDo:[:pluginAndName :index| - mappedPluginEntries doWithIndex:[:pluginAndName :index| ((pluginAndName at: 1) == plugin and:[(pluginAndName at: 2) == fnSymbol]) ifTrue:[ "Transcript show:' ... okay'." ^ index]]. (plugin respondsTo: fnSymbol) ifFalse:[ "Transcript cr; show:'Failed ... primitive not in plugin'." ^ 0]. mappedPluginEntries := mappedPluginEntries copyWith: (Array with: plugin with: fnSymbol). "Transcript show:' ... okay'." ^ mappedPluginEntries size! Item was changed: ----- Method: KlattSynthesizerPlugin class>>initialize (in category 'class initialization') ----- initialize " KlattSynthesizerPlugin initialize " | parameterClass | parameterClass := (Smalltalk hasClassNamed: #KlattFrame) ifTrue: [Smalltalk at: #KlattFrame "original package structure"] ifFalse: [Smalltalk at: #KlattResonatorIndices "repackaged to separate plugin from Speech dependency"]. + parameterClass parameterNames withIndexDo: + [ :each :i | self classPool at: each capitalized asSymbol put: i - 1]. + PI := Float pi. - parameterClass parameterNames - doWithIndex: [ :each :i | self classPool at: each capitalized asSymbol put: i-1]. - PI := Float pi . Epsilon := 1.0e-04! Item was changed: ----- Method: StackInterpreter class>>table:from: (in category 'initialization') ----- table: anArray from: specArray "SpecArray is an array of one of (index selector) or (index1 index2 selector) or (index nil) or (index1 index2 nil). If selector then the entry is the selector, but if nil the entry is the index." | contiguous | contiguous := 0. specArray do: [:spec | (spec at: 1) = contiguous ifFalse: [self error: 'Non-contiguous table entry']. spec size = 2 ifTrue: [anArray at: (spec at: 1) + 1 put: ((spec at: 2) ifNil: [spec at: 1] ifNotNil: [:sym| sym]). contiguous := contiguous + 1] ifFalse: [(spec at: 1) to: (spec at: 2) do: [:i | anArray at: i + 1 put: ((spec at: 3) ifNil: [i] ifNotNil: [:sym| sym])]. contiguous := contiguous + ((spec at: 2) - (spec at: 1)) + 1]]. + anArray withIndexDo: - anArray doWithIndex: [:entry :index| entry isSymbol ifTrue: [(self shouldIncludeMethodForSelector: entry) ifFalse: [anArray at: index put: 0]]]! Item was changed: ----- Method: VMInterfaceConsistencyTests>>testInterpreterProxyInterfaceIsConsistentFor: (in category 'private-support') ----- testInterpreterProxyInterfaceIsConsistentFor: aClass | ccg | ccg := VMPluginCodeGenerator new. InterpreterProxy selectors asArray sort do: [:sel| | pm | ((#(initialize #'initialize-release' private) includes: (InterpreterProxy whichCategoryIncludesSelector: sel)) or: [ccg noteUsedPluginFunction: sel]) ifFalse: [pm := ccg compileToTMethodSelector: sel in: InterpreterProxy. (aClass whichClassIncludesSelector: sel) ifNotNil: [:impClass| | im pt it | im := ccg compileToTMethodSelector: sel in: impClass. "A number of functions are actually void in the interpreter but declared as sqInt in sqVirtualMachine.c, e.g. push" self assert: ((pt := pm returnType) = (it := im returnType) or: [pt = #sqInt and: [it = #void]]) description: 'inconsistent returnType for ', sel, ' in InterpreterProxy vs ', impClass, ' ', pt, ' vs ', it. + pm args withIndexDo: - pm args doWithIndex: [:pma :i| | ima | ima := im args at: i. self assert: (pt := pm typeFor: pma in: ccg) = (it := im typeFor: ima in: ccg) description: 'inconsistent ', i printString, (#('st' 'nd') at: i ifAbsent: 'th'), ' arg type for ', sel, ' in InterpreterProxy vs ', impClass, ' ', pt, ' vs ', it]]]]!

1 0

[OpenSmalltalk/opensmalltalk-vm] 253cbf: Make retrieve/storeLastKeyValue inline.
by Eliot Miranda 19 Jan '21

19 Jan '21

Branch: refs/heads/Cog Home: https://github.com/OpenSmalltalk/opensmalltalk-vm Commit: 253cbfc149cd7a3f6e04149c9002eec96530ef95 https://github.com/OpenSmalltalk/opensmalltalk-vm/commit/253cbfc149cd7a3f6e… Author: Eliot Miranda <eliot.miranda(a)gmail.com> Date: 2021-01-19 (Tue, 19 Jan 2021) Changed paths: M platforms/unix/vm-display-X11/sqUnixX11.c Log Message: ----------- Make retrieve/storeLastKeyValue inline.

1 0

Build failed: opensmalltalk-vm 1.0.2377
by AppVeyor 19 Jan '21

19 Jan '21

1 0

[OpenSmalltalk/opensmalltalk-vm]
by Eliot Miranda 19 Jan '21

19 Jan '21

Branch: refs/heads/mantis-7579 Home: https://github.com/OpenSmalltalk/opensmalltalk-vm

1 0

[OpenSmalltalk/opensmalltalk-vm] c3a43b: Fix missing KeyRelease events when multiple keys a...
by David T Lewis 19 Jan '21

19 Jan '21

Branch: refs/heads/Cog Home: https://github.com/OpenSmalltalk/opensmalltalk-vm Commit: c3a43bc998bad28e314cd3bd972015fe0a8e02aa https://github.com/OpenSmalltalk/opensmalltalk-vm/commit/c3a43bc998bad28e31… Author: David T Lewis <dtlewis290(a)gmail.com> Date: 2021-01-19 (Tue, 19 Jan 2021) Changed paths: M platforms/unix/vm-display-X11/sqUnixX11.c Log Message: ----------- Fix missing KeyRelease events when multiple keys are depressed. (#547) * Fix missing KeyRelease events when multiple keys are depressed. Reference Mantis 0007597 http://bugs.squeak.org/view.php?id=7597. Rather than keep a single lastKey to remember the last previously pressed key value, maintain an array size 256 of last key pressed values indexed by X11 KeyCode. Works for any number of simultaneous keys. * Initialize lastKeyValue array at compile time. For safety, mask index into lastKeyValue to ensure array bounds are honored. Co-authored-by: David T. Lewis <lewis(a)mail.msen.com>

1 0

[OpenSmalltalk/opensmalltalk-vm] Fix missing KeyRelease events when multiple keys are depressed. (#547)
by David T Lewis 19 Jan '21

19 Jan '21

Reference Mantis 0007597 http://bugs.squeak.org/view.php?id=7597. Rather than keep a single lastKey to remember the last previously pressed key value, maintain an array size 256 of last key pressed values indexed by X11 KeyCode. Works for any number of simultaneous keys. You can view, comment on, or merge this pull request online at: https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/547 -- Commit Summary -- * Fix missing KeyRelease events when multiple keys are depressed. -- File Changes -- M platforms/unix/vm-display-X11/sqUnixX11.c (73) -- Patch Links -- https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/547.patch https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/547.diff -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/547

2 4

[OpenSmalltalk/opensmalltalk-vm] 018c35: Initialize lastKeyValue array at compile time. For...
by David T Lewis 19 Jan '21

19 Jan '21

Branch: refs/heads/mantis-7579 Home: https://github.com/OpenSmalltalk/opensmalltalk-vm Commit: 018c35e194afb7ce21ee2a7340bcd39317f7b5b4 https://github.com/OpenSmalltalk/opensmalltalk-vm/commit/018c35e194afb7ce21… Author: David T. Lewis <lewis(a)mail.msen.com> Date: 2021-01-19 (Tue, 19 Jan 2021) Changed paths: M platforms/unix/vm-display-X11/sqUnixX11.c Log Message: ----------- Initialize lastKeyValue array at compile time. For safety, mask index into lastKeyValue to ensure array bounds are honored.

1 0

[OpenSmalltalk/opensmalltalk-vm] ea326f: Add a build directory for the v3 JIT on linux 32-b...
by Eliot Miranda 19 Jan '21

19 Jan '21

Branch: refs/heads/Cog Home: https://github.com/OpenSmalltalk/opensmalltalk-vm Commit: ea326f93baaed828f7590be14effe8dc0aa5cc8c https://github.com/OpenSmalltalk/opensmalltalk-vm/commit/ea326f93baaed828f7… Author: Eliot Miranda <eliot.miranda(a)gmail.com> Date: 2021-01-18 (Mon, 18 Jan 2021) Changed paths: A build.linux32ARMv6/squeak.cog.v3/build.assert/mvm A build.linux32ARMv6/squeak.cog.v3/build.debug/mvm A build.linux32ARMv6/squeak.cog.v3/build/mvm A build.linux32ARMv6/squeak.cog.v3/makeallclean A build.linux32ARMv6/squeak.cog.v3/makealldirty A build.linux32ARMv6/squeak.cog.v3/plugins.ext A build.linux32ARMv6/squeak.cog.v3/plugins.int Log Message: ----------- Add a build directory for the v3 JIT on linux 32-bit ARM. Inlcyde fbdev in all squeak linux 32-bit ARM builds (untested). Add -m32 to all linux 32-bit ARM builds, hoping for cross compilation on pi4.

1 0

[OpenSmalltalk/opensmalltalk-vm] 52d0b5: CogVM source as per VMMaker.oscog-eem.2936
by Eliot Miranda 19 Jan '21

19 Jan '21

Branch: refs/heads/Cog Home: https://github.com/OpenSmalltalk/opensmalltalk-vm Commit: 52d0b5d3b19974c3451d396aba88d33af3335f05 https://github.com/OpenSmalltalk/opensmalltalk-vm/commit/52d0b5d3b19974c345… Author: Eliot Miranda <eliot.miranda(a)gmail.com> Date: 2021-01-18 (Mon, 18 Jan 2021) Changed paths: M nsspur64src/vm/cogit.h M nsspur64src/vm/cogitARMv8.c M nsspur64src/vm/cogitX64SysV.c M nsspur64src/vm/cogitX64WIN64.c M nsspur64src/vm/cointerp.c M nsspur64src/vm/cointerp.h M nsspur64src/vm/gcc3x-cointerp.c M nsspursrc/vm/cogit.h M nsspursrc/vm/cogitARMv5.c M nsspursrc/vm/cogitIA32.c M nsspursrc/vm/cogitMIPSEL.c M nsspursrc/vm/cointerp.c M nsspursrc/vm/cointerp.h M nsspursrc/vm/gcc3x-cointerp.c M spur64src/vm/cogit.h M spur64src/vm/cogitARMv8.c M spur64src/vm/cogitX64SysV.c M spur64src/vm/cogitX64WIN64.c M spur64src/vm/cointerp.c M spur64src/vm/cointerp.h M spur64src/vm/gcc3x-cointerp.c M spursista64src/vm/cogit.h M spursista64src/vm/cogitARMv8.c M spursista64src/vm/cogitX64SysV.c M spursista64src/vm/cogitX64WIN64.c M spursista64src/vm/cointerp.c M spursista64src/vm/cointerp.h M spursista64src/vm/gcc3x-cointerp.c M spursistasrc/vm/cogit.h M spursistasrc/vm/cogitARMv5.c M spursistasrc/vm/cogitIA32.c M spursistasrc/vm/cogitMIPSEL.c M spursistasrc/vm/cointerp.c M spursistasrc/vm/cointerp.h M spursistasrc/vm/gcc3x-cointerp.c M spursrc/vm/cogit.h M spursrc/vm/cogitARMv5.c M spursrc/vm/cogitIA32.c M spursrc/vm/cogitMIPSEL.c M spursrc/vm/cointerp.c M spursrc/vm/cointerp.h M spursrc/vm/gcc3x-cointerp.c M src/vm/cogit.h M src/vm/cogitARMv5.c M src/vm/cogitIA32.c M src/vm/cogitMIPSEL.c M src/vm/cointerp.c M src/vm/cointerp.h M src/vm/gcc3x-cointerp.c Log Message: ----------- CogVM source as per VMMaker.oscog-eem.2936 Cogit ARMv5 code generator: increase method alignment to allow the entry alignment mask to be large enough, a la ARMv8. This is work in trying to provide a V3 32-bit ARM JIT.

1 0

VM Maker: VMMaker.oscog-eem.2936.mcz
by commits＠source.squeak.org 19 Jan '21

19 Jan '21

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.2936.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.2936 Author: eem Time: 18 January 2021, 5:00:46.89368 pm UUID: 69233536-ece4-4ea1-b523-c28c4aa7d4c2 Ancestors: VMMaker.oscog-eem.2935 Fix simulation of the ARMv5 code generator (increase method alignment to allow the entry alignment mask to be large enough, a la ARMv8). Fix simulation of the V3 simulator (needs to implement getStackPointer for SmartSyntaxPlugin simulation). Fix some speeling rorres in conemnts. =============== Diff against VMMaker.oscog-eem.2935 =============== Item was changed: ----- Method: CoInterpreter>>isCodeCompactingPrimitiveIndex: (in category 'primitive support') ----- isCodeCompactingPrimitiveIndex: primIndex "If instVarAt:, slotAt: or shallowCopy operate on a Context then they compute a + bytecode pc and hence may provoke a code compaction. Hence primitive invocation - bytecode pc and hence may provoke a code compaction. Hence primtiive invocation from these primitives must use a static return address (cePrimReturnEnterCogCode:)." <inline: true> self cCode: [] inSmalltalk: [#primitiveClone. #primitiveInstVarAt. #primitiveSlotAt]. "For senders..." ^primIndex = PrimNumberInstVarAt or: [primIndex = PrimNumberShallowCopy or: [primIndex = PrimNumberSlotAt]]! Item was added: + ----- Method: CogARMCompiler>>roundUpToMethodAlignment: (in category 'method zone and entry point alignment') ----- + roundUpToMethodAlignment: numBytes + "Determine the default alignment for the start of a CogMethod, which in turn + determines the size of the mask used to distinguish the checked and unchecked + entry-points, used to distinguish normal and super sends on method unlinking. + This is implemented here to allow processors with coarse instructions (ARM) to + increase the alignment if required." + <cmacro: '(ignored,numBytes) (((numBytes) + 15) & -16)'> "extra parens to placate gdb :-(" + ^numBytes + 15 bitAnd: -16! Item was changed: ----- Method: CogARMv8Compiler>>roundUpToMethodAlignment: (in category 'method zone and entry point alignment') ----- roundUpToMethodAlignment: numBytes + "Determine the default alignment for the start of a CogMethod, which in turn - "Determine the default alignment for the start of a CogMehtod, which in turn determines the size of the mask used to distinguish the checked and unchecked entry-points, used to distinguish normal and super sends on method unlinking. This is implemented here to allow processors with coarse instructions (ARM) to increase the alignment if required." <cmacro: '(ignored,numBytes) (((numBytes) + 15) & -16)'> "extra parens to placate gdb :-(" ^numBytes + 15 bitAnd: -16! Item was changed: ----- Method: CogAbstractInstruction>>roundUpToMethodAlignment: (in category 'method zone and entry point alignment') ----- roundUpToMethodAlignment: numBytes + "Determine the default alignment for the start of a CogMethod, which in turn - "Determine the default alignment for the start of a CogMehtod, which in turn determines the size of the mask used to distinguish the checked and unchecked entry-points, used to distinguish normal and super sends on method unlinking. This is implemented here to allow processors with coarse instructions (ARM) to increase the alignment if required." <cmacro: '(ignored,numBytes) ((numBytes) + 7 & -8)'> ^numBytes + 7 bitAnd: -8! Item was added: + ----- Method: NewObjectMemory>>getStackPointer (in category 'interpreter access') ----- + getStackPointer + "hack around the CoInterpreter/ObjectMemory split refactoring" + <doNotGenerate> + ^coInterpreter getStackPointer! Item was removed: - ----- Method: NewObjectMemorySimulator>>getStackPointer (in category 'interpreter access') ----- - getStackPointer - "hack around the CoInterpreter/ObjectMemory split refactoring" - ^coInterpreter getStackPointer! Item was changed: ----- Method: SimpleStackBasedCogit>>compileInterpreterPrimitive:flags: (in category 'primitive generators') ----- compileInterpreterPrimitive: primitiveRoutine flags: flags "Compile a call to an interpreter primitive. Call the C routine with the usual stack-switching dance, test the primFailCode and then either return on success or continue to the method body." <var: #primitiveRoutine declareC: 'void (*primitiveRoutine)(void)'> | jmp jmpSamplePrim continuePostSamplePrim jmpSampleNonPrim continuePostSampleNonPrim | - <var: #jmp type: #'AbstractInstruction *'> - <var: #jmpSamplePrim type: #'AbstractInstruction *'> - <var: #jmpSampleNonPrim type: #'AbstractInstruction *'> - <var: #continuePostSamplePrim type: #'AbstractInstruction *'> - <var: #continuePostSampleNonPrim type: #'AbstractInstruction *'> "Save processor fp, sp and return pc in the interpreter's frame stack and instruction pointers" self genExternalizePointersForPrimitiveCall. "Switch to the C stack." self genLoadCStackPointersForPrimCall. (flags anyMask: PrimCallCollectsProfileSamples) ifTrue: ["Test nextProfileTick for being non-zero and call checkProfileTick if so" objectMemory wordSize = 4 ifTrue: [self MoveAw: coInterpreter nextProfileTickAddress R: TempReg. self MoveAw: coInterpreter nextProfileTickAddress + objectMemory wordSize R: ClassReg. self OrR: TempReg R: ClassReg] ifFalse: [self MoveAw: coInterpreter nextProfileTickAddress R: TempReg. self CmpCq: 0 R: TempReg]. "If set, jump to record sample call." jmpSampleNonPrim := self JumpNonZero: 0. continuePostSampleNonPrim := self Label]. "Old full prim trace is in VMMaker-eem.550 and prior" self recordPrimTrace ifTrue: [self genFastPrimTraceUsing: ClassReg and: SendNumArgsReg]. "Clear the primFailCode and set argumentCount" self MoveCq: 0 R: TempReg. self MoveR: TempReg Aw: coInterpreter primFailCodeAddress. methodOrBlockNumArgs ~= 0 ifTrue: [self MoveCq: methodOrBlockNumArgs R: TempReg]. self MoveR: TempReg Aw: coInterpreter argumentCountAddress. "If required, set primitiveFunctionPointer and newMethod" (flags anyMask: PrimCallNeedsPrimitiveFunction) ifTrue: [self MoveCw: primitiveRoutine asInteger R: TempReg. primSetFunctionLabel := self MoveR: TempReg Aw: coInterpreter primitiveFunctionPointerAddress]. (flags anyMask: PrimCallNeedsNewMethod+PrimCallMayEndureCodeCompaction) ifTrue: ["The ceActivateFailingPrimitiveMethod: machinery can't handle framelessness." (flags anyMask: PrimCallMayEndureCodeCompaction) ifTrue: [needsFrame := true]. methodLabel addDependent: (self annotateAbsolutePCRef: (self MoveCw: methodLabel asInteger R: ClassReg)). self MoveMw: (self offset: CogMethod of: #methodObject) r: ClassReg R: TempReg. self MoveR: TempReg Aw: coInterpreter newMethodAddress]. "Invoke the primitive" self PrefetchAw: coInterpreter primFailCodeAddress. (flags anyMask: PrimCallMayEndureCodeCompaction) ifTrue: "Sideways call the C primitive routine so that we return through cePrimReturnEnterCogCode." ["On Spur ceActivateFailingPrimitiveMethod: would like to retry if forwarders are found. So insist on PrimCallNeedsPrimitiveFunction being set too." + objectMemory hasSpurMemoryManagerAPI ifTrue: + [self assert: (flags anyMask: PrimCallNeedsPrimitiveFunction)]. - self assert: (flags anyMask: PrimCallNeedsPrimitiveFunction). backEnd genMarshallNArgs: 0 arg: nil arg: nil arg: nil arg: nil; genSubstituteReturnAddress: ((flags anyMask: PrimCallCollectsProfileSamples) ifTrue: [cePrimReturnEnterCogCodeProfiling] ifFalse: [cePrimReturnEnterCogCode]). primInvokeInstruction := self JumpFullRT: primitiveRoutine asInteger. jmp := jmpSamplePrim := continuePostSamplePrim := nil] ifFalse: ["Call the C primitive routine." backEnd genMarshallNArgs: 0 arg: 0 arg: 0 arg: 0 arg: 0. primInvokeInstruction := self CallFullRT: primitiveRoutine asInteger. backEnd genRemoveNArgsFromStack: 0. (flags anyMask: PrimCallCollectsProfileSamples) ifTrue: [self assert: (flags anyMask: PrimCallNeedsNewMethod). "Test nextProfileTick for being non-zero and call checkProfileTick if so" objectMemory wordSize = 4 ifTrue: [self MoveAw: coInterpreter nextProfileTickAddress R: TempReg. self MoveAw: coInterpreter nextProfileTickAddress + objectMemory wordSize R: ClassReg. self OrR: TempReg R: ClassReg] ifFalse: [self MoveAw: coInterpreter nextProfileTickAddress R: TempReg. self CmpCq: 0 R: TempReg]. "If set, jump to record sample call." jmpSamplePrim := self JumpNonZero: 0. continuePostSamplePrim := self Label]. objectRepresentation maybeCompileRetryOnPrimitiveFail: primitiveIndex. "Switch back to the Smalltalk stack. Stack better be in either of these two states: success: stackPointer -> result (was receiver) arg1 ... argN return pc failure: receiver arg1 ... stackPointer -> argN return pc In either case we can push the instructionPointer or load it into the LinkRegister to reestablish the return pc" self MoveAw: coInterpreter instructionPointerAddress R: (backEnd hasLinkRegister ifTrue: [LinkReg] ifFalse: [ClassReg]). backEnd genLoadStackPointers. "Test primitive failure" self MoveAw: coInterpreter primFailCodeAddress R: TempReg. backEnd hasLinkRegister ifFalse: [self PushR: ClassReg]. "Restore return pc on CISCs" self flag: 'ask concrete code gen if move sets condition codes?'. self CmpCq: 0 R: TempReg. jmp := self JumpNonZero: 0. "Fetch result from stack" self MoveMw: (backEnd hasLinkRegister ifTrue: [0] ifFalse: [objectMemory wordSize]) r: SPReg R: ReceiverResultReg. self RetN: objectMemory wordSize]. "return to caller, popping receiver" (flags anyMask: PrimCallCollectsProfileSamples) ifTrue: ["The sample is collected by cePrimReturnEnterCogCode for external calls" jmpSamplePrim ifNotNil: ["Call ceCheckProfileTick: to record sample and then continue." jmpSamplePrim jmpTarget: self Label. self assert: (flags anyMask: PrimCallNeedsNewMethod). self CallFullRT: (self cCode: [#ceCheckProfileTick asUnsignedIntegerPtr] inSmalltalk: [self simulatedTrampolineFor: #ceCheckProfileTick]). "reenter the post-primitive call flow" self Jump: continuePostSamplePrim]. "Null newMethod and call ceCheckProfileTick: to record sample and then continue. ceCheckProfileTick will map null/0 to coInterpreter nilObject" jmpSampleNonPrim jmpTarget: self Label. self MoveCq: 0 R: TempReg. self MoveR: TempReg Aw: coInterpreter newMethodAddress. self CallFullRT: (self cCode: [#ceCheckProfileTick asUnsignedIntegerPtr] inSmalltalk: [self simulatedTrampolineFor: #ceCheckProfileTick]). "reenter the post-primitive call flow" self Jump: continuePostSampleNonPrim]. jmp ifNotNil: ["Jump to restore of receiver reg and proceed to frame build for failure." jmp jmpTarget: self Label. "Restore receiver reg from stack. If on RISCs ret pc is in LinkReg, if on CISCs ret pc is on stack." self MoveMw: objectMemory wordSize * (methodOrBlockNumArgs + (backEnd hasLinkRegister ifTrue: [0] ifFalse: [1])) r: SPReg R: ReceiverResultReg]. ^0!

1 0

← Newer
1
2
3
4
5
6
7
...
17
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Vm-dev January 2021