Vm-dev May 2017

vm-dev@lists.squeakfoundation.org

39 participants
158 discussions

[OpenSmalltalk/opensmalltalk-vm] 0364ee: CogVM source as per VMMaker.oscog-eem.2227
by GitHub 29 May '17

29 May '17

Branch: refs/heads/Cog Home: https://github.com/OpenSmalltalk/opensmalltalk-vm Commit: 0364ee70e95e68d2acea04067147f102b1510ce0 https://github.com/OpenSmalltalk/opensmalltalk-vm/commit/0364ee70e95e68d2ac… Author: Eliot Miranda <eliot.miranda(a)gmail.com> Date: 2017-05-29 (Mon, 29 May 2017) Changed paths: M nsspur64src/vm/cogit.h M nsspur64src/vm/cogitX64SysV.c M nsspur64src/vm/cogitX64WIN64.c M nsspursrc/vm/cogit.h M nsspursrc/vm/cogitARMv5.c M nsspursrc/vm/cogitIA32.c M nsspursrc/vm/cogitMIPSEL.c M spur64src/vm/cogit.h M spur64src/vm/cogitX64SysV.c M spur64src/vm/cogitX64WIN64.c M spurlowcode64src/vm/cogit.h M spurlowcode64src/vm/cogitX64SysV.c M spurlowcode64src/vm/cogitX64WIN64.c M spurlowcodesrc/vm/cogit.h M spurlowcodesrc/vm/cogitARMv5.c M spurlowcodesrc/vm/cogitIA32.c M spurlowcodesrc/vm/cogitMIPSEL.c M spursista64src/vm/cogit.h M spursista64src/vm/cogitX64SysV.c M spursista64src/vm/cogitX64WIN64.c M spursistasrc/vm/cogit.h M spursistasrc/vm/cogitARMv5.c M spursistasrc/vm/cogitIA32.c M spursistasrc/vm/cogitMIPSEL.c M spursrc/vm/cogit.h M spursrc/vm/cogitARMv5.c M spursrc/vm/cogitIA32.c M spursrc/vm/cogitMIPSEL.c M src/plugins/FilePlugin/FilePlugin.c M src/vm/cogit.h M src/vm/cogitARMv5.c M src/vm/cogitIA32.c M src/vm/cogitMIPSEL.c Log Message: ----------- CogVM source as per VMMaker.oscog-eem.2227 Cogit: Use genMarshallNArgs:arg:arg:arg:arg: when compiling an interpreter primitive call to allow adding the shadow parameter space on Win64. Remove a couple of unused and dubious methods in the CogAbstractInstruction hierarchy. Inline arg removal to match inlining of arg marshalling, since removal is either no or one instruction. Fix generation of JumpR on x64. I forgot a rex byte. This doesn't affect SysV because we only ever jump through ClassReg (in an open PIC dispatch) which is RCX (1) on SysV but R8 (8) on Win64. With the fix to JumpR a Win64 image simulates correctly (with smashing of the shadow parameter space). FilePlugin: Incorporate Monty's file plugin changes: add a new primitiveFileOpenNew to open an unexisting file atomically without race condition.

3 3

Debugging Win64 Cog Spur
by Eliot Miranda 29 May '17

29 May '17

Hi Nicolas, the VM gets quite far before some unknown problem in path name manipulation. I'm drunning the debug VM under gdb via (gdb) run -trace=259 trunk50-64.image (See Cogit>>sendTrace: for a definition of the flags) and this is the output ... UnixFileDirectory class>pathNameDelimiter Array(Object)>at: BlockClosure>value: AcornFileDirectory class>isActiveDirectoryClass SmalltalkImage>getSystemAttribute: ByteString(String)>isString ByteString(ArrayedCollection)>size ByteString(ArrayedCollection)>size SmallInteger>= Array(Object)>at: BlockClosure>value: MacFileDirectory class>isActiveDirectoryClass MacFileDirectory class>pathNameDelimiter Character>= Array(Object)>at: BlockClosure>value: DosFileDirectory class(FileDirectory class)>isActiveDirectoryClass DosFileDirectory class>pathNameDelimiter DosFileDirectory class(FileDirectory class)>primPathNameDelimiter Character>= FilePath class>pathName: FilePath class>pathName:isEncoded: Alas there's no debug information to be had: (gdb) where #0 0x00000000000008d4 in ?? () Backtrace stopped: previous frame identical to this frame (corrupt stack?) So my next step is to put a breakpoint for the selector #pathName:isEncoded: and step from there. (gdb) b warning Breakpoint 1 at 0x4015d9: file ../../spur64src/vm/gcc3x-cointerp.c, line 44. (gdb) run -breaksel pathName:isEncoded: trunk50-64.image The program being debugged has been started already. Start it from the beginning? (y or n) y Starting program: /cygdrive/z/oscogvm/build.win64x64/squeak.cog.spur/builddbg/vm/Squeak.exe -breaksel pathName:isEncoded: trunk50-64.image [New Thread 4080.0x5ec] [New Thread 4080.0xb30] etc... _,,,^..^,,,_ best, Eliot

2 7

VM Maker: VMMaker.oscog-eem.2227.mcz
by commits＠source.squeak.org 29 May '17

29 May '17

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.2227.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.2227 Author: eem Time: 29 May 2017, 10:27:16.31959 am UUID: 7ea146b4-39ce-4de7-afa3-a76ed1d1da35 Ancestors: VMMaker.oscog-eem.2225 Inline arg removal to match inlining of arg marshalling, since removal is either no or one instruction. =============== Diff against VMMaker.oscog-eem.2225 =============== Item was changed: ----- Method: CogARMCompiler>>genRemoveNArgsFromStack: (in category 'abi') ----- genRemoveNArgsFromStack: n "This is a no-op on ARM since the ABI passes up to 4 args in registers and trampolines currently observe that limit." + <inline: true> self assert: n <= 4. ^0! Item was changed: ----- Method: CogIA32Compiler>>genRemoveNArgsFromStack: (in category 'abi') ----- + genRemoveNArgsFromStack: n + <inline: true> - genRemoveNArgsFromStack: n n = 0 ifFalse: [cogit AddCq: n * 4 R: ESP]. ^0! Item was changed: ----- Method: CogIA32Compiler>>genRemoveNFloatArgsFromStack: (in category 'abi') ----- + genRemoveNFloatArgsFromStack: n + <inline: true> - genRemoveNFloatArgsFromStack: n n > 0 ifTrue: [cogit AddCq: n * 8 R: ESP]. ^0! Item was changed: ----- Method: CogMIPSELCompiler>>genRemoveNArgsFromStack: (in category 'abi') ----- genRemoveNArgsFromStack: n "This is a no-op on MIPS since the ABI passes up to 4 args in registers and trampolines currently observe that limit." + <inline: true> self assert: n <= 4. ^0! Item was changed: ----- Method: CogX64Compiler>>genRemoveNArgsFromStack: (in category 'abi') ----- genRemoveNArgsFromStack: n "This is a no-op on x64 SysV since the ABI passes up to 6 args in registers and trampolines currently observe a limit of 4. But the WIN64 ABI allways reserve shadow space for saving up to 4 parameter registers (even if less than 4 args)." + <inline: true> self assert: n <= 4. SysV ifFalse: [cogit AddCq: 32 R: RSP]. ^0! Item was changed: ----- Method: CogX64Compiler>>genRemoveNFloatArgsFromStack: (in category 'abi') ----- genRemoveNFloatArgsFromStack: n "This is a no-op on x64 SysV since the ABI passes up to 6 args in registers and trampolines currently observe a limit of 4. But the WIN64 ABI allways reserve shadow space for saving up to 4 parameter registers (even if less than 4 args)." + <inline: true> self assert: n <= 4. SysV ifFalse: [cogit AddCq: 32 R: RSP]. ^0!

1 0

VM Maker: VMMaker.oscog-eem.2226.mcz
by commits＠source.squeak.org 29 May '17

29 May '17

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.2226.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.2226 Author: eem Time: 29 May 2017, 10:13:01.322144 am UUID: 5e9f2691-638f-4ed0-bb8b-72491e1abf1b Ancestors: VMMaker.oscog-eem.2225 Rewrite genMarshallNArgs:arg:arg:arg:arg: to get better dead code removal in compileInterpreterPrimitive:flags:. =============== Diff against VMMaker.oscog-eem.2225 =============== Item was changed: ----- Method: CogARMCompiler>>genMarshallNArgs:arg:arg:arg:arg: (in category 'abi') ----- genMarshallNArgs: numArgs arg: regOrConst0 arg: regOrConst1 arg: regOrConst2 arg: regOrConst3 "Generate the code to pass up to four arguments in a C run-time call. Hack: each argument is either a negative number, which encodes a constant, or a non-negative number, that of a register. Run-time calls have no more than four arguments, so chosen so that on ARM, where in its C ABI the first four integer arguments are passed in registers, all arguments can be passed in registers. We defer to the back end to generate this code not so much that the back end knows whether it uses the stack or registers to pass arguments (it does, but...). In fact we defer for an extremely evil reason. Doing so allows the x64 (where up to 6 args are passed) to assign the register arguments in an order that allows some of the argument registers to be used for specific abstract registers, specifically ReceiverResultReg and ClassReg. This is evil, evil, evil, but also it's really nice to keep using the old register assignments the original author has grown accustomed to." <inline: true> + numArgs > 0 ifTrue: "Has to be written like this for good dead code removal" + [(cogit isTrampolineArgConstant: regOrConst0) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: CArg0Reg] + ifFalse: [cogit MoveR: regOrConst0 R: CArg0Reg]. + numArgs = 1 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst1) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: CArg1Reg] + ifFalse: [cogit MoveR: regOrConst1 R: CArg1Reg]. + numArgs = 2 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst2) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: CArg2Reg] + ifFalse: [cogit MoveR: regOrConst2 R: CArg2Reg]. + numArgs = 3 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst3) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: CArg3Reg] + ifFalse: [cogit MoveR: regOrConst3 R: CArg3Reg]]! - numArgs = 0 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst0) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: CArg0Reg] - ifFalse: [cogit MoveR: regOrConst0 R: CArg0Reg]. - numArgs = 1 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst1) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: CArg1Reg] - ifFalse: [cogit MoveR: regOrConst1 R: CArg1Reg]. - numArgs = 2 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst2) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: CArg2Reg] - ifFalse: [cogit MoveR: regOrConst2 R: CArg2Reg]. - numArgs = 3 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst3) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: CArg3Reg] - ifFalse: [cogit MoveR: regOrConst3 R: CArg3Reg]! Item was changed: ----- Method: CogIA32Compiler>>genMarshallNArgs:arg:arg:arg:arg: (in category 'abi') ----- genMarshallNArgs: numArgs arg: regOrConst0 arg: regOrConst1 arg: regOrConst2 arg: regOrConst3 "Generate the code to pass up to four arguments in a C run-time call. Hack: each argument is either a negative number, which encodes a constant, or a non-negative number, that of a register. Run-time calls have no more than four arguments, so chosen so that on ARM, where in its C ABI the first four integer arguments are passed in registers, all arguments can be passed in registers. We defer to the back end to generate this code not so much that the back end knows whether it uses the stack or registers to pass arguments (it does, but...). In fact we defer for an extremely evil reason. Doing so allows the x64 (where up to 6 args are passed) to assign the register arguments in an order that allows some of the argument registers to be used for specific abstract registers, specifically ReceiverResultReg and ClassReg. This is evil, evil, evil, but also it's really nice to keep using the old register assignments the original author has grown accustomed to." <inline: true> + numArgs > 0 ifTrue: "Has to be written like this for good dead code removal" + [numArgs > 1 ifTrue: + [numArgs > 2 ifTrue: + [numArgs > 3 ifTrue: + [(cogit isTrampolineArgConstant: regOrConst3) + ifFalse: [cogit PushR: regOrConst3] + ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst3)]]. + (cogit isTrampolineArgConstant: regOrConst2) + ifFalse: [cogit PushR: regOrConst2] + ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst2)]]. + (cogit isTrampolineArgConstant: regOrConst1) + ifFalse: [cogit PushR: regOrConst1] + ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst1)]]. + (cogit isTrampolineArgConstant: regOrConst0) + ifFalse: [cogit PushR: regOrConst0] + ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst0)]]! - numArgs = 0 ifTrue: - [^self]. - numArgs > 1 ifTrue: - [numArgs > 2 ifTrue: - [numArgs > 3 ifTrue: - [(cogit isTrampolineArgConstant: regOrConst3) - ifFalse: [cogit PushR: regOrConst3] - ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst3)]]. - (cogit isTrampolineArgConstant: regOrConst2) - ifFalse: [cogit PushR: regOrConst2] - ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst2)]]. - (cogit isTrampolineArgConstant: regOrConst1) - ifFalse: [cogit PushR: regOrConst1] - ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst1)]]. - (cogit isTrampolineArgConstant: regOrConst0) - ifFalse: [cogit PushR: regOrConst0] - ifTrue: [cogit PushCq: (cogit trampolineArgValue: regOrConst0)]! Item was changed: ----- Method: CogMIPSELCompiler>>genMarshallNArgs:arg:arg:arg:arg: (in category 'abi') ----- genMarshallNArgs: numArgs arg: regOrConst0 arg: regOrConst1 arg: regOrConst2 arg: regOrConst3 "Generate the code to pass up to four arguments in a C run-time call. Hack: each argument is either a negative number, which encodes a constant, or a non-negative number, that of a register. Run-time calls have no more than four arguments, so chosen so that on ARM, where in its C ABI the first four integer arguments are passed in registers, all arguments can be passed in registers. We defer to the back end to generate this code not so much that the back end knows whether it uses the stack or registers to pass arguments (it does, but...). In fact we defer for an extremely evil reason. Doing so allows the x64 (where up to 6 args are passed) to assign the register arguments in an order that allows some of the argument registers to be used for specific abstract registers, specifically ReceiverResultReg and ClassReg. This is evil, evil, evil, but also it's really nice to keep using the old register assignments the original author has grown accustomed to." <inline: true> self flag: #OABI. + numArgs > 0 ifTrue: "Has to be written like this for good dead code removal" + [(cogit isTrampolineArgConstant: regOrConst0) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: A0] + ifFalse: [cogit MoveR: regOrConst0 R: A0]. + numArgs = 1 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst1) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: A1] + ifFalse: [cogit MoveR: regOrConst1 R: A1]. + numArgs = 2 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst2) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: A2] + ifFalse: [cogit MoveR: regOrConst2 R: A2]. + numArgs = 3 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst3) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: A3] + ifFalse: [cogit MoveR: regOrConst3 R: A3]]! - numArgs = 0 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst0) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: A0] - ifFalse: [cogit MoveR: regOrConst0 R: A0]. - numArgs = 1 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst1) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: A1] - ifFalse: [cogit MoveR: regOrConst1 R: A1]. - numArgs = 2 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst2) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: A2] - ifFalse: [cogit MoveR: regOrConst2 R: A2]. - numArgs = 3 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst3) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: A3] - ifFalse: [cogit MoveR: regOrConst3 R: A3]! Item was changed: ----- Method: CogX64Compiler>>genMarshallNArgs:arg:arg:arg:arg: (in category 'abi') ----- genMarshallNArgs: numArgs arg: regOrConst0 arg: regOrConst1 arg: regOrConst2 arg: regOrConst3 "Generate the code to pass up to four arguments in a C run-time call. Hack: each argument is either a negative number, which encodes a constant, or a non-negative number, that of a register. Run-time calls have no more than four arguments, so chosen so that on ARM, where in its C ABI the first four integer arguments are passed in registers, all arguments can be passed in registers. We defer to the back end to generate this code not so much that the back end knows whether it uses the stack or registers to pass arguments (it does, but...). In fact we defer for an extremely evil reason. Doing so allows the x64 (where up to 6 args are passed) to assign the register arguments in an order that allows some of the argument registers to be used for specific abstract registers, specifically ReceiverResultReg and ClassReg. This is evil, evil, evil, but also it's really nice to keep using the old register assignments the original author has grown accustomed to. How can this possibly work? Look at Cogit class>>runtime for a list of the run-time calls and their arguments, including which arguments are passed in which registers. Look at CogX64Compiler's subclass implementations of initializeAbstractRegisters. There are no calls in which ReceiverResultReg (RDX) and/or ClassReg (RCX) are passed along with Arg0Reg and Arg1Reg, and none in which the use of either ReceiverResultReg or ClassReg conflict for args 3 & 4. So if args are assigned in order, the registers do not get overwritten. Yes, this is evil, but it's so nice to continue to use RCX & RDX. Argument registers for args 0 to 3 in SysV are RDI RSI RDX RCX, and in Win64 are RCX RDX R8 R9" <inline: true> SysV ifFalse: "WIN64 ABI allways reserve shadow space on the stack for callee to save up to 4 register parameters" [cogit SubCq: 32 R: RSP]. + numArgs > 0 ifTrue: "Has to be written like this for good dead code removal" + [SysV + ifTrue: + [(cogit isTrampolineArgConstant: regOrConst0) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: RDI] "a.k.a. Arg0Reg" - numArgs = 0 ifTrue: [^self]. - SysV - ifTrue: - [(cogit isTrampolineArgConstant: regOrConst0) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: RDI] "a.k.a. Arg0Reg" - ifFalse: - [regOrConst0 ~= RDI ifTrue: - [cogit MoveR: regOrConst0 R: RDI]]. - numArgs = 1 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst1) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: RSI] "a.k.a. Arg1Reg" - ifFalse: - [regOrConst1 ~= RSI ifTrue: - [cogit MoveR: regOrConst1 R: RSI]]. - numArgs = 2 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst2) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: RDX] "a.k.a. ReceiverResultReg" - ifFalse: - [regOrConst2 ~= RDX ifTrue: - [cogit MoveR: regOrConst2 R: RDX]]. - numArgs = 3 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst3) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: RCX] "a.k.a. ClassReg" ifFalse: + [regOrConst0 ~= RDI ifTrue: + [cogit MoveR: regOrConst0 R: RDI]]. + numArgs = 1 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst1) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: RSI] "a.k.a. Arg1Reg" - [regOrConst3 ~= RCX ifTrue: - [cogit MoveR: regOrConst3 R: RCX]]] - ifFalse: - [(cogit isTrampolineArgConstant: regOrConst0) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: RCX] "a.k.a. Arg0Reg" - ifFalse: - [regOrConst0 ~= RCX ifTrue: - [cogit MoveR: regOrConst0 R: RCX]]. - numArgs = 1 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst1) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: RDX] "a.k.a. Arg1Reg" - ifFalse: - [regOrConst1 ~= RDX ifTrue: - [cogit MoveR: regOrConst1 R: RDX]]. - numArgs = 2 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst2) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: R8] "a.k.a. RISCTempReg in CogInLineLiteralsX64Compiler and Extra6Reg in CogOutOfLineLiteralsX64Compiler" - ifFalse: - [regOrConst2 ~= R8 ifTrue: - [cogit MoveR: regOrConst2 R: R8]]. - numArgs = 3 ifTrue: [^self]. - (cogit isTrampolineArgConstant: regOrConst3) - ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: R9] "a.k.a. SendNumArgsReg" ifFalse: + [regOrConst1 ~= RSI ifTrue: + [cogit MoveR: regOrConst1 R: RSI]]. + numArgs = 2 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst2) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: RDX] "a.k.a. ReceiverResultReg" + ifFalse: + [regOrConst2 ~= RDX ifTrue: + [cogit MoveR: regOrConst2 R: RDX]]. + numArgs = 3 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst3) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: RCX] "a.k.a. ClassReg" + ifFalse: + [regOrConst3 ~= RCX ifTrue: + [cogit MoveR: regOrConst3 R: RCX]]] + ifFalse: + [(cogit isTrampolineArgConstant: regOrConst0) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst0) R: RCX] "a.k.a. Arg0Reg" + ifFalse: + [regOrConst0 ~= RCX ifTrue: + [cogit MoveR: regOrConst0 R: RCX]]. + numArgs = 1 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst1) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst1) R: RDX] "a.k.a. Arg1Reg" + ifFalse: + [regOrConst1 ~= RDX ifTrue: + [cogit MoveR: regOrConst1 R: RDX]]. + numArgs = 2 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst2) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst2) R: R8] "a.k.a. RISCTempReg in CogInLineLiteralsX64Compiler and Extra6Reg in CogOutOfLineLiteralsX64Compiler" + ifFalse: + [regOrConst2 ~= R8 ifTrue: + [cogit MoveR: regOrConst2 R: R8]]. + numArgs = 3 ifTrue: [^self]. + (cogit isTrampolineArgConstant: regOrConst3) + ifTrue: [cogit MoveCq: (cogit trampolineArgValue: regOrConst3) R: R9] "a.k.a. SendNumArgsReg" + ifFalse: + [regOrConst3 ~= R9 ifTrue: + [cogit MoveR: regOrConst3 R: R9]]]. + self assert: numArgs <= 4]! - [regOrConst3 ~= R9 ifTrue: - [cogit MoveR: regOrConst3 R: R9]]]. - self assert: numArgs <= 4!

1 0

VM Maker: VMMaker.oscog-eem.2225.mcz
by commits＠source.squeak.org 29 May '17

29 May '17

Eliot Miranda uploaded a new version of VMMaker to project VM Maker: http://source.squeak.org/VMMaker/VMMaker.oscog-eem.2225.mcz ==================== Summary ==================== Name: VMMaker.oscog-eem.2225 Author: eem Time: 29 May 2017, 9:34:28.198165 am UUID: a3791777-370b-419b-9708-9e73bf94e39b Ancestors: VMMaker.oscog-nice.2224 Cogit: Use genMarshallNArgs:arg:arg:arg:arg: when compiling an interpreter primitive call to allow adding the shadow parameter space on Win64. Fix generation of JumpR on x64. I forgot a rex byte. This doesn't affect SysV because we only ever jump through ClassReg (in an open PIC dispatch) which is RCX (1) on SysV but R8 (8) on Win64. Add tests for CallR and JumpR generation on x64. Have printRumpCStackTo: include the sp and fp if in the range. With the fix to JumpR a Win64 image simulates correctly (with smashing of the shadow parameter space). =============== Diff against VMMaker.oscog-nice.2224 =============== Item was changed: ----- Method: CogVMSimulator>>printRumpCStackTo: (in category 'rump c stack') ----- printRumpCStackTo: address self assert: (self isOnRumpCStack: address). heapBase - objectMemory wordSize to: address by: objectMemory wordSize negated do: [:addr| + self printHex: addr. + addr = cogit processor sp + ifTrue: [self print: ' sp->'] + ifFalse: + [addr = cogit processor fp + ifTrue: [self print: ' fp->'] + ifFalse: [self tab]]. + self tab; printHex: (objectMemory longAt: addr); cr]! - self printHex: addr; tab; printHex: (objectMemory longAt: addr); cr]! Item was changed: ----- Method: CogX64Compiler>>computeMaximumSize (in category 'generate machine code') ----- (excessive size, no diff calculated) Item was changed: ----- Method: CogX64Compiler>>concretizeJumpR (in category 'generate machine code') ----- concretizeJumpR <inline: true> | reg | reg := operands at: 0. machineCode + at: 0 put: (self rexR: 0 x: 0 b: reg); + at: 1 put: 16rFF; + at: 2 put: (self mod: ModReg RM: reg RO: 4). + ^machineCodeSize := 3! - at: 0 put: 16rFF; - at: 1 put: (self mod: ModReg RM: reg RO: 4). - ^machineCodeSize := 2! Item was added: + ----- Method: CogX64CompilerTests>>testCallR (in category 'tests') ----- + testCallR + "self new testCallR" + self concreteCompilerClass registersWithNamesDo: + [:reg :regname| | inst len| + inst := self gen: CallR operand: reg. + len := inst concretizeAt: 0. + self processor + disassembleInstructionAt: 0 + In: inst machineCode object + into: [:str :sz| | plainJane herIntended | + "Convert e.g. '00000000: movl %eax, 0x2(%eax) : 89 40 02' to 'movl %eax, 0x2(%eax)'" + plainJane := self strip: str. + herIntended := 'call ', regname. + self assert: herIntended equals: plainJane. + self assert: len = sz]]! Item was added: + ----- Method: CogX64CompilerTests>>testJumpR (in category 'tests') ----- + testJumpR + "self new testJumpR" + self concreteCompilerClass registersWithNamesDo: + [:reg :regname| | inst len| + inst := self gen: JumpR operand: reg. + len := inst concretizeAt: 0. + self processor + disassembleInstructionAt: 0 + In: inst machineCode object + into: [:str :sz| | plainJane herIntended | + "Convert e.g. '00000000: movl %eax, 0x2(%eax) : 89 40 02' to 'movl %eax, 0x2(%eax)'" + plainJane := self strip: str. + herIntended := 'jmp ', regname. + self assert: herIntended equals: plainJane. + self assert: len = sz]]! 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+PrimCallMayCallBack) ifTrue: ["The ceActivateFailingPrimitiveMethod: machinery can't handle framelessness." (flags anyMask: PrimCallMayCallBack) 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: PrimCallMayCallBack) 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." self assert: (flags anyMask: PrimCallNeedsPrimitiveFunction). + backEnd + genMarshallNArgs: 0 arg: 0 arg: 0 arg: 0 arg: 0; + genSubstituteReturnAddress: + ((flags anyMask: PrimCallCollectsProfileSamples) + ifTrue: [cePrimReturnEnterCogCodeProfiling] + ifFalse: [cePrimReturnEnterCogCode]). - backEnd 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. self maybeCompileAllocFillerCheck. "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

VM Maker: Cog-eem.343.mcz
by commits＠source.squeak.org 29 May '17

29 May '17

Eliot Miranda uploaded a new version of Cog to project VM Maker: http://source.squeak.org/VMMaker/Cog-eem.343.mcz ==================== Summary ==================== Name: Cog-eem.343 Author: eem Time: 29 May 2017, 8:52:07.276334 am UUID: 1930114c-32c6-4acc-8c08-0b9e2aca3bfb Ancestors: Cog-eem.341 Fix saving the argument registers to the shadow stack space in smashCallerSavedRegistersWithValuesFrom:by:in:; aMemory is one-relative, not 0-relative. =============== Diff against Cog-eem.341 =============== Item was changed: ----- Method: BochsX64Alien>>smashCallerSavedRegistersWithValuesFrom:by:in: (in category 'accessing-abstract') ----- smashCallerSavedRegistersWithValuesFrom: base by: step in: aMemory | volatileRegisters | CogX64Compiler isSysV ifTrue: [volatileRegisters := #(rax: rcx: rdx: rsi: rdi: r8: r9: r10: r11:)] ifFalse: "We should also save the registers to the shadow parameter space." [volatileRegisters := #(rax: rcx: rdx: r8: r9: r10: r11:). #(rcx rdx r8 r9) withIndexDo: [:getter :index| aMemory + unsignedLong64At: self rbp + 9 + (index * 8) "skip saved fp and retpc; aMemory is 1-relative" - unsignedLong64At: self rbp + 8 + (index * 8) "skip saved fp and retpc" put: (self perform: getter)]]. volatileRegisters withIndexDo: [:setter :index| self perform: setter with: index - 1 * step + base]!

1 0

VM Maker: CogTools-sk.79.mcz
by commits＠source.squeak.org 29 May '17

29 May '17

Sophie Kaleba uploaded a new version of CogTools to project VM Maker: http://source.squeak.org/VMMaker/CogTools-sk.79.mcz ==================== Summary ==================== Name: CogTools-sk.79 Author: sk Time: 29 May 2017, 12:03:07.991919 pm UUID: 03ac5973-4cc7-4c4d-bf6f-c82b01d333bc Ancestors: CogTools-sk.78 add some class comments cleaning the code =============== Diff against CogTools-sk.78 =============== Item was changed: ----- Method: PharoVMFileSystem>>deleteContentsOf: (in category 'as yet unclassified') ----- deleteContentsOf: aDirectory + "not yet implemented. The temp directory created during the profiling will not be deleted when quitting the image" + ! - "have to find something similar"! Item was changed: VMProfiler subclass: #PharoVMProfiler instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'CogTools-VMProfiler'! + + !PharoVMProfiler commentStamp: 'sk 5/29/2017 11:56' prior: 0! + I am the VMProfiler called when using a Pharo image. + + I can only be used headless for the moment. + Try inspecting on : String streamContents: [ :s | VMProfiler headlessSpyOn: [ 1 to: 10000000 do: [ :i | Object new ] ] reportOn: s ] + + ! Item was changed: VMProfiler subclass: #SqueakVMProfiler instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'CogTools-VMProfiler'! + + !SqueakVMProfiler commentStamp: 'sk 5/29/2017 11:56' prior: 0! + I am the VMProfiler called when using a Squeak image. + + I can be used + - headful (WorldMenu>open>VMProfiler) + - or headless : for instance, inspect on : String streamContents: [ :s | VMProfiler headlessSpyOn: [ 1 to: 10000000 do: [ :i | Object new ] ] reportOn: s ] + ! Item was changed: Object subclass: #VMFileSystem instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'CogTools-VMProfiler'! + !VMFileSystem commentStamp: 'sk 5/29/2017 11:27' prior: 0! + I dispatch between the file systems of Pharo (FileSystem) and Squeak (FileDirectory)! - !VMFileSystem commentStamp: 'SophieKaleba 5/15/2017 01:49' prior: 0! - I am a bridge between the file systems of Pharo and Squeak. - I am used in the VMProfiler to keep it generic.! Item was changed: ----- Method: VMFileSystem>>nameFordirPath:plus: (in category 'as yet unclassified') ----- nameFordirPath: aDirPath plus: aProcessId + "answers a FileSystem/Directory object representing a directory. The name of this directory is its path+the name of a process id" self subclassResponsibility ! Item was changed: ----- Method: VMProfilerMacSymbolsManager>>parseSymbolsFor: (in category 'parsing') ----- parseSymbolsFor: module | proc symtab symStream | (VMFileSystem exists: tempDir fullName, '/', module shortName) ifFalse: - "(tempDir fileExists: module shortName) ifFalse:" ["N.B. Don't use the -f option (which meant flat symbols) as in El Capitan it is misinterpreted to mean -format." proc := OSProcess thisOSProcess command: 'cd ', tempDir fullName, ';nm -n -arch ', self archName, " -f" ' "', module name, '" | grep -v " [aAU] " >"', module shortName, '"']. symStream := (Array new: 1000) writeStream. symStream nextPut: module. + proc ifNotNil: [[proc isComplete] whileFalse: [(Delay forMilliseconds: 25) wait]]. - proc ifNotNil:[(Delay forMilliseconds: 25) wait]. - "[[proc isComplete] whileFalse: [(Delay forMilliseconds: 25) wait]]." "infinite loop" symtab := [StandardFileStream readOnlyFileNamed: (tempDir fullName, '/', module shortName)] on: Error do: [:ex| "Handle flaky OSProcess stuff by reporting error and failing to parse" Transcript print: ex; flush. ^nil]. "Have caller eliminate modules with no text." symtab size = 0 ifTrue: [^nil]. module shortName = 'HIToolbox' ifTrue: [self halt]. [| prev | prev := self parseSymbolsFrom: symtab to: symStream. "CoreAUC has a huge chunk of data at the end of its text segment that causes the profiler to spend ages counting zeros. Hack fix by setting the end of the last symbol in the text segment to a little less than 1Mb." "00000000000f1922 retq" "Mavericks 13.4" "00000000000f3b21 retq" "Yosemite 14.5" module shortName = 'CoreAUC' ifTrue: [prev limit: 16rf8000]. symbolsByModule at: module put: (self relocateSymbols: symStream contents allButFirst inModule: module). (prev notNil and: [prev limit isNil]) ifTrue: [prev limit: module limit]] ensure: [symtab close]!

1 0

multiple crashes on macOS Sierra
by Tudor Girba 29 May '17

29 May '17

Hi, Today I experienced multiple crashes on macOS Sierra. These happened during regular programming, mostly while typing code. I attached here the crash dump(s). I believe I heard Phil having the same issue with the stable Pharo VM. Does anyone else has the same issue? The VM is: CoInterpreter VMMaker.oscog-eem.2203 uuid: 12d4afae-8498-4e76-8efe-60eba6ef4db2 May 2 2017 StackToRegisterMappingCogit VMMaker.oscog-eem.2203 uuid: 12d4afae-8498-4e76-8efe-60eba6ef4db2 May 2 2017 VM: 201705021552 https://github.com/OpenSmalltalk/opensmalltalk-vm.git $ Date: Tue May 2 08:52:41 2017 -0700 $ Plugins: 201705021552 https://github.com/OpenSmalltalk/opensmalltalk-vm.git $ The Pharo update number is: 60482 Cheers, Doru -- www.tudorgirba.com www.feenk.com "From an abstract enough point of view, any two things are similar."

8 14

[OpenSmalltalk/opensmalltalk-vm] 0f07ff: [iOS] Dont autorelease what we not own
by GitHub 29 May '17

29 May '17

Branch: refs/heads/Cog Home: https://github.com/OpenSmalltalk/opensmalltalk-vm Commit: 0f07ffbedcc184391ed973636f8968c93b63f2cc https://github.com/OpenSmalltalk/opensmalltalk-vm/commit/0f07ffbedcc184391e… Author: Tobias Pape <tobias(a)netshed.de> Date: 2017-05-29 (Mon, 29 May 2017) Changed paths: M platforms/iOS/vm/OSX/sqSqueakOSXCGView.m M platforms/iOS/vm/OSX/sqSqueakOSXOpenGLView.m Log Message: ----------- [iOS] Dont autorelease what we not own

1 0

VM Maker: Cog-eem.342.mcz
by commits＠source.squeak.org 29 May '17

29 May '17

Eliot Miranda uploaded a new version of Cog to project VM Maker: http://source.squeak.org/VMMaker/Cog-eem.342.mcz ==================== Summary ==================== Name: Cog-eem.342 Author: eem Time: 28 May 2017, 10:45:46.436283 pm UUID: c50cecd8-9ed1-4c1c-962d-96c3f8af7b0d Ancestors: Cog-eem.341 Smashing the shadow parameter space must take PostBuildStackDelta into account. =============== Diff against Cog-eem.341 =============== Item was changed: ----- Method: BochsX64Alien>>smashCallerSavedRegistersWithValuesFrom:by:in: (in category 'accessing-abstract') ----- smashCallerSavedRegistersWithValuesFrom: base by: step in: aMemory | volatileRegisters | CogX64Compiler isSysV ifTrue: [volatileRegisters := #(rax: rcx: rdx: rsi: rdi: r8: r9: r10: r11:)] ifFalse: "We should also save the registers to the shadow parameter space." [volatileRegisters := #(rax: rcx: rdx: r8: r9: r10: r11:). #(rcx rdx r8 r9) withIndexDo: [:getter :index| aMemory + unsignedLong64At: self rbp + PostBuildStackDelta + 8 + (index * 8) "skip saved fp and retpc" - unsignedLong64At: self rbp + 8 + (index * 8) "skip saved fp and retpc" put: (self perform: getter)]]. volatileRegisters withIndexDo: [:setter :index| self perform: setter with: index - 1 * step + base]!

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 May 2017