Vm-dev March 2015

vm-dev@lists.squeakfoundation.org

37 participants
175 discussions

About primitiveBitAnd / Or / Xor
by Nicolas Cellier 01 Apr '15

01 Apr '15

Hi again, I've noticed the COG code currently branches on objectMemory wordSize = 8 condition, to either load self positive64BitValueOf: integerArgument. or to load self positive32BitValueOf: integerArgument. Of course, this could be resolved to a single integerArgument := self positiveMachineIntegerValueOf: integerArgument which will care of the branch by itself... That's what I did. But maybe the intention was rather to deal with signed64BitValueOf in the 64 bit branch like what was done to primitiveBitShift? Of course, all this is assuming that the machine uses a 2-complement implementation, otherwise it won't match the Smalltalk semantic. That's not so rare ;) - but theoretically not fully portable - contrarily to actual strategy which is fully portable. Since other parts of the VM also rely on such specific, maybe we can trade a little portability tribute for the sake of speed. Thoughts? Nicolas

2 2

Re: [squeak-dev] Bug whith contextOn:do:
by Eliot Miranda 01 Apr '15

01 Apr '15

Hi Tobias, first let me sympathize. This is such horrible code :-(. Ovr the years, getting this code to work with Cog's context-to-stack-mapping machinery has given me headaches :-). On Tue, Mar 31, 2015 at 9:01 AM, Tobias Pape <Das.Linux(a)gmx.de> wrote: > Hey fellow Squeakers > > [Attention: low level lengthy stuff] > > I today encountered a strange behavior. > When running the Startup code, somewhen ContextPart>>#complete is called, > which, in the process issues #contextOn:do:, which subsequently somewhere > interanlly > does a jump: > > contextOn: exceptionClass do: block > "Create an #on:do: context that is ready to return from executing > its receiver" > > | ctxt chain | > ctxt := thisContext. > [chain := thisContext sender cut: ctxt. ctxt jump] on: > exceptionClass do: block. > "jump above will resume here without unwinding chain" > ^ chain > > The idea is that we end up right before '^ chain' as the comment indicates. > so what happens is that ctxt is the context of #contextOn:do: itself and > it gets > send #jump in the closure. > > Jump now does the following: > > jump > "Abandon thisContext and resume self instead (using the same > current process). You may want to save thisContext's sender before calling > this so you can jump back to it. > Self MUST BE a top context (ie. a suspended context or a abandoned > context that was jumped out of). A top context already has its return > value on its stack (see Interpreter>>primitiveSuspend and other suspending > primitives). > thisContext's sender is converted to a top context (by pushing a > nil return value on its stack) so it can be jump back to." > > | top | > "Make abandoned context a top context (has return value (nil)) so > it can be jumped back to" > thisContext sender push: nil. > > "Pop self return value then return it to self (since we jump to > self by returning to it)" > stackp = 0 ifTrue: [self stepToSendOrReturn]. > stackp = 0 ifTrue: [self push: nil]. "must be quick return > self/constant" > top := self pop. > thisContext privSender: self. > ^ top > > So. bytecode for #contextOn:do: is: > > 29 <8A 01> push: (Array new: 1) > 31 <6B> popIntoTemp: 3 > 32 <89> pushThisContext: > 33 <6A> popIntoTemp: 2 > 34 <12> pushTemp: 2 > 35 <13> pushTemp: 3 > 36 <8F 20 00 0A> closureNumCopied: 2 numArgs: 0 bytes 40 to 49 > 40 <89> pushThisContext: > 41 <D2> send: sender > 42 <10> pushTemp: 0 > 43 <E1> send: cut: > 44 <8E 00 01> popIntoTemp: 0 inVectorAt: 1 > 47 <10> pushTemp: 0 > 48 <D3> send: jump > 49 <7D> blockReturn > 50 <10> pushTemp: 0 > 51 <11> pushTemp: 1 > 52 <F0> send: on:do: > 53 <87> pop > 54 <8C 00 03> pushTemp: 0 inVectorAt: 3 > 57 <7C> returnTop > > > The jump lands right at 53 and does a pop. > HOWEVER, at this point the stack of this context is empty and the pop > actually pops the 3rd temp > from the temps that 'just happens' to be right under the stack. This > should be fatal. > HOWEVER again, Squeak actually does not pop but only decrement the SP so > the temp access still > works(this _could_ be fine but some implementations (Eg, RSqueak) tried > to separate temps and > stack; which is not possible currently). > > What could be the problem here? > - are the 'stackp = 0'-checks in #jump wrong and they actually should > check for the actual stack depth _after_ temps? > It does look like it. I would have expected this to be more correct: jump "Abandon thisContext and resume self instead (using the same current process). You may want to save thisContext's sender before calling this so you can jump back to it. Self MUST BE a top context (ie. a suspended context or a abandoned context that was jumped out of). A top context already has its return value on its stack (see Interpreter>>primitiveSuspend and other suspending primitives). thisContext's sender is converted to a top context (by pushing a nil return value on its stack) so it can be jump back to." | top | "Make abandoned context a top context (has return value (nil)) so it can be jumped back to" thisContext sender push: nil. "Pop self return value then return it to self (since we jump to self by returning to it)" stackp <= self numTemps ifTrue: [self stepToSendOrReturn]. (stackp <= self numTemps and: [self willJustPop]) ifTrue: [self push: nil]. "must be quick return self/constant" top := self pop. thisContext privSender: self. ^top > - should we put in a "sacrificial anode" in #contextOn:do: so that the pop > does not pop the empty stack? (like this: > > contextOn: exceptionClass do: block > "Create an #on:do: context that is ready to return from executing > its receiver" > > | ctxt chain | > ctxt := thisContext. > [chain := thisContext sender cut: ctxt. > ctxt push: nil. "sacrifical anode" > ctxt jump > ] on: exceptionClass do: block. > "jump above will resume here without unwinding chain" > ^ chain > That looks right. Once the on:do: is sent the thisContext of contextOn:do:'s stack contains only exceptionClass, block, context and the indirection vector containing chain. So it is in the stack = self numTemps case. > > - Or is there an even better way? > I'm not sure the other ways are any better. The way to transfer to a context without disturbing its stack is to do a process switch. So you do something equivalent to jump "Abandon thisContext and resume self instead (using the same current process)." | process semaphore | process := Processor activeProcess. semaphore := Semaphore new. [process suspendedContext unwindTo: self. process suspendedContext: self. semaphore signal] fork. semaphore wait This way no bizarre stack manipulations are going on, and no return value is pushed, because there isn't a return. One may get away with: jump "Abandon thisContext and resume self instead (using the same current process)." [| process | process := Processor activeProcess. process suspendedContext unwindTo: self. process suspendedContext: self] fork. Processor yield I'd be interested in your experience using either of these. One of the advantages the process switch versions have is in not updating the receiving context sp there's a chance the context-to-stack mapping machinery won't flush the context to the heap. In the end it might actually be faster. -- best, Eliot

2 2

NoDbgRegParms bug bites again--this time with CMake
by gettimothy 01 Apr '15

01 Apr '15

Just a heads up/progress report. The "fix" of putting a #define NoDbgRegParms "<--empty definition" in config.h.in for autotools does NOT translate to CMake because CMake builds its own config.h based on some tests/probing in a config.cmake file. This then creates its own config.h.in (cmake version) and then config.h When it does so, it does NOT respect the "empty definition" and just leave things alone, but "helps" by placing a /* #undef NoDbgRegParms */ in config.h There are 3 strategies I have in mind. 1. Have the user hand edit config.h (unacceptable, rejected) 2. Find a CMake workaround to generate the #define NoDbgRegParms 3. Take this as a clue about how autotools expects things and make the logic in Cog/src/vm/cogit.h Cog/src/vm/cointerp.h etc logic so that it works with #if !PRODUCTION && defined(__GNUC__) && !defined(NoDbgRegParms) # define NoDbgRegParms __attribute__ ((regparm (0))) #endif I will be focusing on approach 3 for a bit. "Maybe" its a. something about a comparison between the !defines and the !PRODUCTION b. a conflict created when autotools does its macro-expansion since the test is in multiple .h files. cheers. tty.

3 6

VM Maker: BytecodeSets.spur-eem.19.mcz
by commits＠source.squeak.org 01 Apr '15

01 Apr '15

Eliot Miranda uploaded a new version of BytecodeSets to project VM Maker: http://source.squeak.org/VMMaker/BytecodeSets.spur-eem.19.mcz ==================== Summary ==================== Name: BytecodeSets.spur-eem.19 Author: eem Time: 31 March 2015, 4:59:45.812 pm UUID: 03b6e370-a91c-4eb6-a80c-8b2fab6a4f4e Ancestors: BytecodeSets.spur-eem.18 Fix SistaV1 decode for the extended push and dup bytecodes. Add the new absent self and absent outer bytecodes to Newspeak V4 decode. =============== Diff against BytecodeSets.spur-eem.18 =============== Item was changed: ----- Method: InstructionStream>>interpretNext2ByteNSV4Instruction:for:extA:extB:startPC: (in category '*BytecodeSets-NewsqueakV4-decoding') ----- interpretNext2ByteNSV4Instruction: bytecode for: client extA: extA extB: extB startPC: startPC "Send to the argument, client, a message that specifies the next instruction. This method handles the two-byte codes. For a table of the bytecode set, see EncoderForNewsqueakV4's class comment." | byte method | method := self method. byte := self method at: pc. pc := pc + 1. "We do an inline quasi-binary search on bytecode" bytecode < 235 ifTrue: [bytecode < 231 ifTrue: [bytecode < 229 ifTrue: [| literal | bytecode = 226 ifTrue: [^client pushReceiverVariable: (extA bitShift: 8) + byte]. literal := method literalAt: (extA bitShift: 8) + byte + 1. bytecode = 227 ifTrue: [^client pushLiteralVariable: literal]. ^client pushConstant: literal]. bytecode = 229 ifTrue: [^client pushConstant: (extB bitShift: 8) + byte]. ^client pushTemporaryVariable: byte]. bytecode = 231 ifTrue: [^byte < 128 ifTrue: [client pushNewArrayOfSize: byte] ifFalse: [client pushConsArrayWithElements: byte - 128]]. bytecode = 232 ifTrue: [^client storeIntoReceiverVariable: (extA bitShift: 8) + byte]. bytecode = 233 ifTrue: [^client storeIntoLiteralVariable: (method literalAt: (extA bitShift: 8) + byte + 1)]. ^client storeIntoTemporaryVariable: byte]. bytecode < 238 ifTrue: [bytecode = 235 ifTrue: [^client popIntoReceiverVariable: (extA bitShift: 8) + byte]. bytecode = 236 ifTrue: [^client popIntoLiteralVariable: (method literalAt: (extA bitShift: 8) + byte + 1)]. ^client popIntoTemporaryVariable: byte]. bytecode < 242 ifTrue: [| selector numArgs | selector := method literalAt: (extA bitShift: 5) + (byte // 8) + 1. numArgs := (extB bitShift: 3) + (byte \\ 8). bytecode = 238 ifTrue: [^client send: selector super: false numArgs: numArgs]. bytecode = 239 ifTrue: [^client send: selector super: true numArgs: numArgs]. bytecode = 240 ifTrue: [^client sendToAbsentImplicitReceiver: selector numArgs: numArgs]. ^client sendToAbsentDynamicSuperclass: selector numArgs: numArgs]. bytecode < 245 ifTrue: [bytecode = 242 ifTrue: [^client jump: (extB bitShift: 8) + byte]. ^client jump: (extB bitShift: 8) + byte if: bytecode = 243]. + bytecode = 245 ifTrue: + [| selector numArgs | + selector := method literalAt: (extA bitShift: 5) + (byte // 8) + 1. + numArgs := (extB bitShift: 3) + (byte \\ 8). + ^client sendToAbsentSelf: selector numArgs: numArgs]. - "245 11110101 xxxxxxxx UNASSIGNED" "246-247 1111011 i xxxxxxxx UNASSIGNED 248-249 1111100 i xxxxxxxx UNASSIGNED" ^self unusedBytecode: client at: startPC! Item was changed: ----- Method: InstructionStream>>interpretNext3ByteNSV4Instruction:for:extA:extB:startPC: (in category '*BytecodeSets-NewsqueakV4-decoding') ----- interpretNext3ByteNSV4Instruction: bytecode for: client extA: extA extB: extB startPC: startPC "Send to the argument, client, a message that specifies the next instruction. This method handles the three-byte codes. For a table of the bytecode set, see EncoderForNewsqueakV4's class comment." | method byte2 byte3 | method := self method. byte2 := method at: pc. byte3 := method at: pc + 1. pc := pc + 2. "we search the bytecodes by static frequency" bytecode = 253 ifTrue: ["253 11111101 eeiiikkk jjjjjjjj Push Closure Num Copied iii (+ Ext A // 16 * 8) Num Args kkk (+ Ext A \\ 16 * 8) BlockSize jjjjjjjj (+ Ext B * 256). ee = num extensions" ^client pushClosureCopyNumCopiedValues: ((byte2 bitShift: -3) bitAnd: 7) + (extA // 16 bitShift: 3) numArgs: (byte2 bitAnd: 7) + (extA \\ 16 bitShift: 3) blockSize: byte3 + (extB bitShift: 8)]. bytecode = 250 ifTrue: [^client pushRemoteTemp: byte2 inVectorAt: byte3]. bytecode = 252 ifTrue: [^client popIntoRemoteTemp: byte2 inVectorAt: byte3]. bytecode = 251 ifTrue: [^client storeIntoRemoteTemp: byte2 inVectorAt: byte3]. bytecode = 249 ifTrue: + [^client callPrimitive: byte2 + (byte3 bitShift: 8)]. + bytecode = 254 ifTrue: + [| selector numArgs depth | + selector := method literalAt: (extA bitShift: 5) + (byte2 // 8) + 1. + numArgs := (extB bitShift: 3) + (byte2 \\ 8). + depth := byte3. + ^client sendToAbsentOuter: selector numArgs: numArgs depth: depth]. + "255-255 11111111 xxxxxxxx yyyyyyyy UNASSIGNED" - [^client callPrimitive: byte2 + (byte3 bitShift: 8)]. - "254-255 1111111 i xxxxxxxx yyyyyyyy UNASSIGNED" ^self unusedBytecode: client at: startPC! Item was changed: ----- Method: InstructionStream>>interpretNextSistaV1InstructionFor: (in category '*BytecodeSets-SistaV1-decoding') ----- interpretNextSistaV1InstructionFor: client "Send to the argument, client, a message that specifies the next instruction." | byte div16 offset method extA extB savedPC | method := self method. "For a table of the bytecode set, see EncoderForSistaV1's class comment." "consume and compute any extensions first." extA := extB := 0. savedPC := pc. [byte := self method at: pc. pc := pc + 1. byte >= 16rE0 and: [byte <= 16rE1]] whileTrue: [| extByte | extByte := self method at: pc. pc := pc + 1. byte = 16rE0 ifTrue: [extA := (extA bitShift: 8) + extByte] ifFalse: [extB := (extB = 0 and: [extByte > 127]) ifTrue: [extByte - 256] ifFalse: [(extB bitShift: 8) + extByte]]]. div16 := byte // 16. offset := byte \\ 16. "We do an inline quasi-binary search on each of the possible 16 values of div16" div16 < 11 ifTrue: [div16 < 6 ifTrue: [div16 < 4 ifTrue: [div16 < 2 ifTrue: [div16 = 0 ifTrue: [^client pushReceiverVariable: offset]. ^client pushLiteralVariable: (method literalAt: offset + 1)]. "div16 = 1" ^client pushConstant: (method literalAt: byte \\ 32 + 1)]. div16 = 4 ifTrue: [offset < 12 ifTrue: [^client pushTemporaryVariable: offset]. offset = 12 ifTrue: [^client pushReceiver]. offset = 13 ifTrue: [^client pushConstant: true]. offset = 14 ifTrue: [^client pushConstant: false]. offset = 15 ifTrue: [^client pushConstant: nil]]. "div16 = 5" offset < 2 ifTrue: [^client pushConstant: offset]. + offset = 2 ifTrue: - offset = 3 ifTrue: [^self interpretSistaV1ExtendedPush: extB for: client]. + offset = 3 ifTrue: - offset = 4 ifTrue: [^client doDup]. + offset = 8 ifTrue: [^client methodReturnReceiver]. offset = 9 ifTrue: [^client methodReturnConstant: true]. offset = 10 ifTrue: [^client methodReturnConstant: false]. offset = 11 ifTrue: [^client methodReturnConstant: nil]. offset = 12 ifTrue: [^client methodReturnTop]. offset = 13 ifTrue: [^client blockReturnConstant: nil]. offset = 14 ifTrue: [^client blockReturnTop]. offset = 15 ifTrue: [^client doNop]. ^self unusedBytecode: client at: savedPC]. "short sends" div16 = 6 ifTrue: [^client send: (Smalltalk specialSelectorAt: offset + 1) super: false numArgs: (Smalltalk specialNargsAt: offset + 1)]. div16 = 7 ifTrue: [^client send: (Smalltalk specialSelectorAt: offset + 17) super: false numArgs: (Smalltalk specialNargsAt: offset + 17)]. ^client send: (method literalAt: offset + 1) super: false numArgs: div16 - 8]. "div16 >= 11; bytecode >= 176" div16 < 14 ifTrue: [div16 = 11 ifTrue: [offset < 8 ifTrue: [^client jump: offset + 1]. ^client jump: offset - 7 if: true]. div16 = 12 ifTrue: [offset < 8 ifTrue: [^client jump: offset + 1 if: false]. ^client popIntoReceiverVariable: offset - 8]. "div16 = 13" offset < 8 ifTrue: [^client popIntoTemporaryVariable: offset]. offset = 9 ifTrue: [^client doDup]. ^self unusedBytecode: client at: savedPC]. "2 byte and 3 byte codes" byte < 248 ifTrue: [^self interpretNext2ByteSistaV1Instruction: byte for: client extA: extA extB: extB startPC: savedPC]. ^self interpretNext3ByteSistaV1Instruction: byte for: client extA: extA extB: extB startPC: savedPC!

1 0

VM Maker: VMMaker.oscog-eem.1141.mcz
by commits＠source.squeak.org 01 Apr '15

01 Apr '15

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.1141.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.1141 Author: eem Time: 31 March 2015, 4:54:17.46 pm UUID: a20d15d4-58a4-4dcb-83d5-d20904fc76ef Ancestors: VMMaker.oscog-eem.1140 Make the fake address generation stuff for in-image compilation a little less confusing. =============== Diff against VMMaker.oscog-eem.1140 =============== Item was changed: ----- Method: CurrentImageCoInterpreterFacade>>addressForLabel: (in category 'labels') ----- addressForLabel: l ^variables at: l ifAbsentPut: + [(self isLabelRelativeToCogitVarBaseReg: l) + ifTrue: [cogit fakeAddressFor: l index: variables size] + ifFalse: [variables size * 4 + self variablesBase]]! - [(coInterpreter class instVarIndexFor: l ifAbsent: [objectMemory class instVarIndexFor: l ifAbsent: [#(CStackPointer CFramePointer) includes: l]]) - ifNil: [variables size * 4 + self variablesBase] - ifNotNil: [cogit fakeAddressFor: l index: (variables keys select: [:k| (coInterpreter class instVarIndexFor: l ifAbsent: []) notNil]) size]]! Item was added: + ----- Method: CurrentImageCoInterpreterFacade>>isLabelRelativeToCogitVarBaseReg: (in category 'labels') ----- + isLabelRelativeToCogitVarBaseReg: l + coInterpreter class + instVarIndexFor: l + ifAbsent: + [objectMemory class + instVarIndexFor: l + ifAbsent: [^false]]. + ^true!

1 0

VM Maker: VMMaker.oscog-eem.1140.mcz
by commits＠source.squeak.org 01 Apr '15

01 Apr '15

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.1140.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.1140 Author: eem Time: 31 March 2015, 4:45:47.233 pm UUID: b5079ad1-0ca2-4184-bb6f-4c61f4390d1c Ancestors: VMMaker.oscog-eem.1139 Fix code generation for ARM concretizeMoveMbrR/RMbr Fix in-image compilation for ARM and varBaseReg. Beware! There may be discrepancies between in-image compiled code and simulated compiled code. We need to reduce these, ideally to none. =============== Diff against VMMaker.oscog-eem.1139 =============== Item was changed: ----- Method: CogARMCompiler>>concretizeMoveMbrR (in category 'generate machine code - concretize') ----- concretizeMoveMbrR "Will get inlined into concretizeAt: switch." "ldrb destReg, [srcReg, #immediate] or ldrb destReg, [srcReg, ConcreteIPReg]" <inline: true> | srcReg offset destReg instrOffset| offset := operands at: 0. srcReg := self concreteRegister: (operands at: 1). destReg := self concreteRegister: (operands at: 2). self is12BitValue: offset ifTrue: [ :u :immediate | self machineCodeAt: 0 "ldrb destReg, [srcReg, #immediate]" put: (self ldrb: destReg rn: srcReg plus: u imm: immediate). ^machineCodeSize := 4] ifFalse: [(self isAddressRelativeToVarBase: offset) ifTrue: + [self machineCodeAt: 0 put: (self adds: ConcreteIPReg rn: ConcreteVarBaseReg imm: offset - cogit varBaseAddress ror: 0). - [self machineCodeAt: 0 put: (self ldr: ConcreteIPReg rn: ConcreteVarBaseReg plusImm: offset - cogit varBaseAddress). instrOffset := 4] ifFalse: [instrOffset := self at: 0 moveCw: offset intoR: ConcreteIPReg]. "ldrb destReg, [srcReg, ConcreteIPReg]" self machineCodeAt: instrOffset put: (self ldrb: destReg rn: srcReg rm: ConcreteIPReg). ^machineCodeSize := instrOffset + 4]! Item was changed: ----- Method: CogARMCompiler>>concretizeMoveRMbr (in category 'generate machine code - concretize') ----- concretizeMoveRMbr "Will get inlined into concretizeAt: switch." <inline: true> | srcReg offset baseReg instrOffset| srcReg := self concreteRegister: (operands at: 0). offset := operands at: 1. baseReg := self concreteRegister: (operands at: 2). self is12BitValue: offset ifTrue: [ :u :immediate | self machineCodeAt: 0 "strb srcReg, [baseReg, #immediate]" put: (self strb: srcReg rn: baseReg plus: u imm: immediate). ^machineCodeSize := 4] ifFalse: [(self isAddressRelativeToVarBase: offset) ifTrue: + [self machineCodeAt: 0 put: (self adds: ConcreteIPReg rn: ConcreteVarBaseReg imm: offset - cogit varBaseAddress ror: 0). - [self machineCodeAt: 0 put: (self ldr: ConcreteIPReg rn: ConcreteVarBaseReg plusImm: offset - cogit varBaseAddress). instrOffset := 4] ifFalse: [instrOffset := self at: 0 moveCw: offset intoR: ConcreteIPReg]. "strb srcReg, [baseReg, ConcreteIPReg]" self machineCodeAt: instrOffset put: (self strb: srcReg rn: baseReg rm: ConcreteIPReg). ^machineCodeSize := instrOffset + 4 ]! Item was changed: ----- Method: Cogit>>fakeAddressFor:index: (in category 'initialization') ----- fakeAddressFor: anObject index: index "Answer a fake address for some variable based on some index. The index will usually be the size of simulatedAddresses, but in determining the varBaseAddress we take a guess at the final size of simulatedAddresses." <doNotGenerate> ^(index + 101 * objectMemory wordSize) negated + bitAnd: ((backEnd notNil + and: [backEnd wantsNearAddressFor: anObject]) - bitAnd: ((backEnd wantsNearAddressFor: anObject) ifTrue: [self addressSpaceMask] ifFalse: [self allButTopBitOfAddressSpaceMask])! Item was changed: ----- Method: CurrentImageCoInterpreterFacade>>addressForLabel: (in category 'labels') ----- addressForLabel: l + ^variables + at: l + ifAbsentPut: + [(coInterpreter class instVarIndexFor: l ifAbsent: [objectMemory class instVarIndexFor: l ifAbsent: [#(CStackPointer CFramePointer) includes: l]]) + ifNil: [variables size * 4 + self variablesBase] + ifNotNil: [cogit fakeAddressFor: l index: (variables keys select: [:k| (coInterpreter class instVarIndexFor: l ifAbsent: []) notNil]) size]]! - ^variables at: l ifAbsentPut: [variables size * 4 + self variablesBase]! Item was changed: ----- Method: CurrentImageCoInterpreterFacade>>cogit: (in category 'initialize-release') ----- cogit: aCogit cogit := aCogit. + cogit objectMemory ifNil: + [cogit instVarNamed: 'objectMemory' put: objectMemory]. coInterpreter cogit: aCogit. (objectMemory respondsTo: #cogit:) ifTrue: [objectMemory cogit: aCogit]. (objectMemory respondsTo: #coInterpreter:) ifTrue: + [objectMemory coInterpreter: coInterpreter]. + #('stackLimit') do: + [:l| self addressForLabel: l]. + self initializeObjectMap! - [objectMemory coInterpreter: coInterpreter]! Item was changed: ----- Method: CurrentImageCoInterpreterFacade>>initialize (in category 'initialize-release') ----- initialize memory := (VMBIGENDIAN ifTrue: [Bitmap] ifFalse: [LittleEndianBitmap]) new: 1024*1024/4. objectMemory := self class objectMemoryClass simulatorClass new. objectMemory memory: memory. objectMemory initializeFreeSpaceForFacadeFrom: self startOfMemory to: self variablesBase. coInterpreter := CoInterpreter new. coInterpreter instVarNamed: 'objectMemory' put: objectMemory; instVarNamed: 'primitiveTable' put: (CArrayAccessor on: CoInterpreter primitiveTable copy). + variables := Dictionary new! - variables := Dictionary new. - #('stackLimit') do: - [:l| self addressForLabel: l]. - self initializeObjectMap!

1 0

VM-safety missing guard clauses
by Nicolas Cellier 01 Apr '15

01 Apr '15

I can crash VM with Object adoptInstance: 24. Presumably, it's a missing guard in primitiveAdoptInstance Find correction attached.

1 0

VM safety: missing failing guards in SmartSyntaxPlugin
by Nicolas Cellier 01 Apr '15

01 Apr '15

Hi again, with clang analyzer I noticed a bunch of failing guards missing. For example, take this one: [image: Images intégrées 1]This code comes from: primitiveSocketConnectionStatus: socket | s status | <var: #s type: 'SocketPtr '> self primitive: 'primitiveSocketConnectionStatus' parameters: #(Oop). s := self socketValueOf: socket. interpreterProxy failed ifFalse: [ status := self sqSocketConnectionStatus: s]. ^ status asSmallIntegerObj Maybe it's innocuous to generate an integerObjectOf: somethingNotInitialized, but who knows what more dangerous expression might be returned... This is the SmartSyntaxPluginCodeGenerator at work. looking at this SmartSyntaxPluginTMethod method, I feel like it is a missing guard near the bottom, find correction attached: fixUpReturnOneStmt: stmt on: sStream stmt isReturn ifFalse: [^sStream nextPut: stmt]. (stmt expression isSend and: [#('primitiveFail' 'primitiveFailFor:') includes: stmt expression selector]) ifTrue: ["failure return" sStream nextPut: stmt expression. sStream nextPut: self nullReturnExpr. ^nil]. (stmt expression isVariable and: ['nil' = stmt expression name]) ifTrue: ["^ nil -- this is never right unless automatically generated" sStream nextPut: stmt. ^nil]. (stmt expression isVariable and: ['self' = stmt expression name]) ifTrue: ["^ self" self generateFailureGuardOn: sStream. fullArgs isEmpty ifFalse:[ sStream nextPut: (self popExpr: fullArgs size)]. sStream nextPut: self nullReturnExpr. ^nil]. (stmt expression isVariable | stmt expression isConstant | suppressingFailureGuards) ifTrue: ["^ variable or ^ constant or ^ expr without guardchecking" self generateFailureGuardOn: sStream. sStream nextPut: (self pop: fullArgs size + 1 thenReturnExpr: stmt expression). sStream nextPut: self nullReturnExpr. ^nil]. "^ expr with guardchecking" "vvvvvv LOOK HERE vvvvvvvvvvvvvv" self generateFailureGuardOn: sStream. "^^^^^^ THIS GUARD WAS MISSING I HAD TO ADD IT" sStream nextPut: (self assign: (self oopVariable: '_return_value') expression: stmt expression). self generateFailureGuardOn: sStream. sStream nextPut: (self pop: fullArgs size + 1 thenReturnExpr: (self oopVariable: '_return_value')). sStream nextPut: self nullReturnExpr

1 0

VM Maker: VMMaker.oscog-eem.1139.mcz
by commits＠source.squeak.org 01 Apr '15

01 Apr '15

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.1139.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.1139 Author: eem Time: 31 March 2015, 3:22:28.078 pm UUID: e8eb9b7b-a9f8-40f2-9b93-d0f13e1204f4 Ancestors: VMMaker.oscog-eem.1138 Fix register map printing given the new register assignments on ARM, and new abstract registers. =============== Diff against VMMaker.oscog-eem.1138 =============== Item was changed: ----- Method: CogARMCompiler>>generalPurposeRegisterMap (in category 'disassembly') ----- generalPurposeRegisterMap <doNotGenerate> "Answer a Dictionary from register getter to register index." ^Dictionary newFromPairs: + { #r0. R0. + #r4. R4. - { #r4. R4. #r5. R5. #r6. R6. #r7. R7. #r8. R8. + #r10. R10. + #r11. R11. + #r12. R12 }! - #r9. R9 }! Item was changed: ----- Method: CogRTLOpcodes class>>nameForRegister: (in category 'debug printing') ----- nameForRegister: reg "<Integer>" ^#(Arg0Reg Arg1Reg ClassReg FPReg ReceiverResultReg SPReg SendNumArgsReg TempReg + DPFPReg0 DPFPReg1 DPFPReg2 DPFPReg3 DPFPReg4 DPFPReg5 DPFPReg6 DPFPReg7 + LinkReg RISCTempReg VarBaseReg PCReg) - DPFPReg0 DPFPReg1 DPFPReg2 DPFPReg3 DPFPReg4 DPFPReg5 DPFPReg6 DPFPReg7 LinkReg PCReg) detect: [:sym| (classPool at: sym) = reg]!

1 0

VM Maker: VMMaker.oscog-eem.1138.mcz
by commits＠source.squeak.org 31 Mar '15

31 Mar '15

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.1138.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.1138 Author: eem Time: 31 March 2015, 11:51:36.459 am UUID: 78f53b2b-2b10-4896-a1f4-8ab0e5a9a820 Ancestors: VMMaker.oscog-eem.1137 Fix a C compiler warning. =============== Diff against VMMaker.oscog-eem.1137 =============== Item was changed: ----- Method: StackInterpreter>>callPrimitiveBytecode (in category 'miscellaneous bytecodes') ----- callPrimitiveBytecode "V4: 249 11111001 i i i i i i i i jjjjjjjj Call Primitive #iiiiiiii + (jjjjjjjj * 256) SistaV1: 248 11111000 iiiiiiii mjjjjjjj Call Primitive #iiiiiiii + (jjjjjjj * 256) m=1 means inlined primitive, no hard return after execution. V3/Spur: 139 10001011 i i i i i i i i jjjjjjjj Call Primitive #iiiiiiii + (jjjjjjjj * 256)" "Note that we simply skip a callPrimitiveBytecode at the start of a method that contains a primitive. This because methods like Context(Part)>>reset have to be updated to skip the callPrimtiive bytecode otherwise." self cppIf: SistaVM ifTrue: [| byte1 byte2 prim header | byte1 := self fetchByte. byte2 := self fetchByte. self fetchNextBytecode. byte2 < 128 ifTrue: [header := objectMemory methodHeaderOf: method. ((self methodHeaderHasPrimitive: header) + and: [localIP asUnsignedInteger + = (self initialPCForHeader: header method: method) + (self sizeOfCallPrimitiveBytecode: header)]) ifTrue: - and: [localIP = (self initialPCForHeader: header method: method) + (self sizeOfCallPrimitiveBytecode: header)]) ifTrue: [^self]. localIP := localIP - 3. ^self respondToUnknownBytecode]. prim := byte2 - 128 << 8 + byte1. prim < 1000 ifTrue: [^self nullaryInlinePrimitive: prim]. prim < 2000 ifTrue: [^self unaryInlinePrimitive: prim - 1000]. prim < 3000 ifTrue: [^self binaryInlinePrimitive: prim - 2000]. prim < 4000 ifTrue: [^self trinaryInlinePrimitive: prim - 3000]. localIP := localIP - 3. ^self respondToUnknownBytecode] ifFalse: [| header | header := objectMemory methodHeaderOf: method. ((self methodHeaderHasPrimitive: header) and: [localIP asInteger = (self initialPCForHeader: header method: method)]) ifTrue: [localIP := localIP + (self sizeOfCallPrimitiveBytecode: header) - 1. ^self fetchNextBytecode] ifFalse: [^self respondToUnknownBytecode]]!

1 0

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 March 2015