Squeak-dev March 2024

squeak-dev@lists.squeakfoundation.org

27 participants
58 discussions

The Trunk: KernelTests-ct.457.mcz
by commits＠source.squeak.org 03 Mar '24

03 Mar '24

Christoph Thiede uploaded a new version of KernelTests to project The Trunk: http://source.squeak.org/trunk/KernelTests-ct.457.mcz ==================== Summary ==================== Name: KernelTests-ct.457 Author: ct Time: 3 March 2024, 7:54:10.139958 pm UUID: 47913553-ecfa-2943-a05e-bc27cfebac31 Ancestors: KernelTests-ct.456 Complements Kernel-ct.1559 (merges SimulationSideEffectWarning.5.cs, step 2/2). =============== Diff against KernelTests-ct.456 =============== Item was added: + ----- Method: ContextTest>>invokeSimulationGuard (in category 'private') ----- + invokeSimulationGuard + <simulationGuard> + "Nothing to see here, please move along!!" + ^ 42! Item was added: + ----- Method: ContextTest>>testSimulationSideEffectWarningControl (in category 'tests - simulation') ----- + testSimulationSideEffectWarningControl + + self + should: [Context runSimulated: [[] fork]] + raise: SimulationSideEffectWarning + withExceptionDo: [:warning | + self assert: warning notNil. + self assert: warning isControlPrimitive. + self assert: warning suppressed].! Item was added: + ----- Method: ContextTest>>testSimulationSideEffectWarningGuard (in category 'tests - simulation') ----- + testSimulationSideEffectWarningGuard + + self + should: + [Context runSimulated: + [self invokeSimulationGuard]] + raise: SimulationSideEffectWarning + withExceptionDo: [:warning | + self assert: warning notNil. + self assert: warning isSimulationGuard. + self deny: warning suppressed].! Item was added: + ----- Method: ContextTest>>testSimulationSideEffectWarningGuardResume (in category 'tests - simulation') ----- + testSimulationSideEffectWarningGuardResume + + self + assert: self invokeSimulationGuard + equals: + ([Context runSimulated: + [self invokeSimulationGuard]] + on: SimulationSideEffectWarning + do: [:ex | ex resume]).! Item was added: + ----- Method: ContextTest>>testSimulationSideEffectWarningGuardSkip (in category 'tests - simulation') ----- + testSimulationSideEffectWarningGuardSkip + + self + assert: self + equals: + ([Context runSimulated: + [self invokeSimulationGuard]] + on: SimulationSideEffectWarning + do: [:ex | ex skipPrimitive]).! Item was added: + ----- Method: ContextTest>>testSimulationSideEffectWarningSkip (in category 'tests - simulation') ----- + testSimulationSideEffectWarningSkip + + self + should: + [[(SimulationSideEffectWarning forPrimitive: 42) + unsuppress; + signalIfSkipped: [self fail]. + true] + on: SimulationSideEffectWarning do: [:ex | ex resume]]. + + [:exit | + [(SimulationSideEffectWarning forPrimitive: 42) + unsuppress; + signalIfSkipped: exit. + self fail] + on: SimulationSideEffectWarning do: [:ex | ex skipPrimitive]] + valueWithExit.! Item was added: + ----- Method: ContextTest>>testSimulationSideEffectWarningSuppress (in category 'tests - simulation') ----- + testSimulationSideEffectWarningSuppress + + self + shouldnt: [(SimulationSideEffectWarning forPrimitive: 42) + suppress; + defaultAction] raise: UnhandledWarning; + should: [(SimulationSideEffectWarning forPrimitive: 42) + unsuppress; + defaultAction] raise: UnhandledWarning.!

1 0

The Trunk: Kernel-ct.1559.mcz
by commits＠source.squeak.org 03 Mar '24

03 Mar '24

Christoph Thiede uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-ct.1559.mcz ==================== Summary ==================== Name: Kernel-ct.1559 Author: ct Time: 3 March 2024, 7:49:53.952958 pm UUID: 12cceefa-01f6-fa46-917b-cc2ec6f91312 Ancestors: Kernel-ct.1558 Merges SimulationSideEffectWarning.5.cs (step 2/2): In Context>>doPrimitive:method:receiver:args:, replaces generic "simulation guard" warnings for undebuggable methods by new specific, suppressable SimulationSideEffectWarnings. Also warns when primitive 87 (primitiveResume) is hit. Refactors <primitive: 19> simulation guards into a named alias pragma <simulationGuard>. Revision: Revise documentation and document emergence and usage of SimulationSideEffectWarning in Context>>runSimulated:.... Fix resumption value of SimulationSideEffectWarning when proceeding from a debugger. Avoid infinite recursion due to process-faithful debugging when debugging a SimulationSideEffectWarning from a debugger. Improves tests. Step 2/2: After the new <simulationGuard> pragma has been added to the compiler (Compiler-ct.497), we can use it in the tests and the ST80 package. Thanks to Marcel (mt) for the advice! See: https://lists.squeakfoundation.org/archives/list/squeak-dev%40lists.squeakf… =============== Diff against Kernel-ct.1558 =============== Item was changed: ----- Method: Context class>>runSimulated: (in category 'simulation') ----- runSimulated: aBlock + "Simulate the execution of aBlock, until it ends or is curtailed. Answer the result it returns. + + If aBlock attempts any action that escapes the control of the simulator (e.g., resuming another process), signal a SimulationSideEffectWarning. The client of the simulator (i.e., the sender of this method) may handle those warnings to remain control, or otherwise, these actions are allowed. Examples: + Context runSimulated: [[] fork]. --> fork is executed + [Context runSimulated: [[] fork]] + on: SimulationSideEffectWarning + do: [:ex | + ex isControlPrimitive ifTrue: [ex return: #forbidden]. + ex pass]. --> fork is prevented" - "Simulate the execution of aBlock, until it ends or is curtailed. Answer the result it returns." ^thisContext runSimulated: aBlock contextAtEachStep: [] "Context runSimulated: [Pen new ifNotNil: [:pen| pen defaultNib: 5. 4 timesRepeat: [pen go: 100; turn: 90]]]" "Here's a fun example, reaching into the computation to squash the Display>>fillWhite that mandala: begins with..." "thisContext runSimulated: [Pen new mandala: 45] contextAtEachStep: [:ctxt| ctxt selector == #fillWhite ifTrue: [ctxt scanFor: [:ign| ctxt willReturn]]]"! Item was changed: ----- Method: Context>>doPrimitive:method:receiver:args: (in category 'private') ----- doPrimitive: primitiveIndex method: meth receiver: receiver args: arguments "Simulate a primitive method whose index is primitiveIndex. The simulated receiver and arguments are given as arguments to this message. If successful, push result and return resuming context, else ^ {errCode, PrimitiveFailToken}. Any primitive which provokes execution needs to be intercepted and simulated to avoid execution running away." | value | + "Test for unsimulatable side effects (that is, code that will be triggered in the image outside of the simulator range). This includes simulation guards, which are traditionally flagged using primitive 19 (a null primitive that doesn't do anything), as well as certain control primitives that might trigger code on other processes. If a side effect is detected, raise a warning to give the user/client a chance to cancel or virtualize the operation." + "#(19 87) do: [:primitive | self systemNavigation browseAllSelect: [:m | m primitive = primitive]]" + (primitiveIndex = 19 "simulationGuard" or: [primitiveIndex = 87 "primitiveResume"]) ifTrue: + [[(SimulationSideEffectWarning forPrimitive: primitiveIndex) + context: self method: meth receiver: receiver arguments: arguments; + signalIfSkipped: [^ self]] + ifCurtailed: + [self push: receiver "Cheap fix of the context's internal state. Note that unwinding the receiver -- so that the next step would invoke the primitive again -- would be challenging due to to the variety of senders to this method."]]. - "Judicious use of primitive 19 (a null primitive that doesn't do anything) prevents - the debugger from entering various run-away activities such as spawning a new - process, etc. Injudicious use results in the debugger not being able to debug - interesting code, such as the debugger itself. Hence use primitive 19 with care :-)" - "SystemNavigation new browseAllSelect: [:m| m primitive = 19]" - primitiveIndex = 19 ifTrue: [ - [self notify: ('The code being simulated is trying to control a process ({1}). Process controlling cannot be simulated. If you proceed, things may happen outside the observable area of the simulator.' translated format: {meth reference})] - ifCurtailed: [self push: nil "Cheap fix of the context's internal state"]]. ((primitiveIndex between: 201 and: 222) and: [(self objectClass: receiver) includesBehavior: BlockClosure]) ifTrue: [(primitiveIndex = 206 or: [primitiveIndex = 208]) ifTrue: "[Full]BlockClosure>>valueWithArguments:" [^receiver simulateValueWithArguments: arguments first caller: self]. ((primitiveIndex between: 201 and: 209) "[Full]BlockClosure>>value[:value:...]" or: [primitiveIndex between: 221 and: 222]) ifTrue: "[Full]BlockClosure>>valueNoContextSwitch[:]" [^receiver simulateValueWithArguments: arguments caller: self]]. primitiveIndex = 83 ifTrue: "afr 9/11/1998 19:50" "Object>>perform:[with:...]" [| selector | selector := arguments at: 1 ifAbsent: [^ self class primitiveFailTokenFor: #'bad argument']. ^self send: selector to: receiver with: arguments allButFirst]. primitiveIndex = 84 ifTrue: "afr 9/11/1998 19:50 & eem 8/18/2009 17:04" "Object>>perform:withArguments:" [| selector args | arguments size = 2 ifFalse: [^ self class primitiveFailTokenFor: #'bad argument']. selector := arguments first. args := arguments second. args isArray ifFalse: [^ self class primitiveFailTokenFor: #'bad argument']. ^self send: selector to: receiver with: args]. primitiveIndex = 100 ifTrue: "eem 8/18/2009 16:57" "Object>>perform:withArguments:inSuperclass:" [| rcvr selector args superclass | arguments size caseOf: { [3] -> [ rcvr := receiver. selector := arguments first. args := arguments second. superclass := arguments third]. [4] -> ["mirror primitive" rcvr := arguments first. selector := arguments second. args := arguments third. superclass := arguments fourth] } otherwise: [^ self class primitiveFailTokenFor: #'bad number of arguments']. args isArray ifFalse: [^ self class primitiveFailTokenFor: #'bad argument']. ((self objectClass: rcvr) includesBehavior: superclass) ifFalse: [^ self class primitiveFailTokenFor: #'bad argument']. ^self send: selector to: rcvr with: args lookupIn: superclass]. "Mutex>>primitiveEnterCriticalSection Mutex>>primitiveTestAndSetOwnershipOfCriticalSection" (primitiveIndex = 186 or: [primitiveIndex = 187]) ifTrue: [| effective | effective := Processor activeProcess effectiveProcess. "active == effective" value := primitiveIndex = 186 ifTrue: [receiver primitiveEnterCriticalSectionOnBehalfOf: effective] ifFalse: [receiver primitiveTestAndSetOwnershipOfCriticalSectionOnBehalfOf: effective]. ^(self isPrimFailToken: value) ifTrue: [value] ifFalse: [self push: value]]. (primitiveIndex = 188 or: [primitiveIndex = 189]) ifTrue: [| n args methodArg thisReceiver | primitiveIndex caseOf: {[188 "primitiveExecuteMethodArgsArray"] -> ["Object>>withArgs:executeMethod: CompiledMethod class>>receiver:withArguments:executeMethod: VMMirror>>ifFail:object:with:executeMethod: et al" ((n := arguments size) between: 2 and: 4) ifFalse: [^self class primitiveFailTokenFor: #'unsupported operation']. ((self objectClass: (args := arguments at: n - 1)) == Array and: [(self objectClass: (methodArg := arguments at: n)) includesBehavior: CompiledMethod]) ifFalse: [^self class primitiveFailTokenFor: #'bad argument']. thisReceiver := arguments at: n - 2 ifAbsent: [receiver]]. [189 "primitiveExecuteMethod"] -> ["Object>>executeMethod: Object>>with:...executeMethod:" (arguments size > 0) ifFalse: [^self class primitiveFailTokenFor: #'bad argument']. ((self objectClass: (methodArg := arguments atLast: 1)) includesBehavior: CompiledMethod) ifFalse: [^self class primitiveFailTokenFor: #'bad argument']. args := arguments allButLast. thisReceiver := receiver]}. methodArg numArgs = args size ifFalse: [^self class primitiveFailTokenFor: #'bad number of arguments']. methodArg primitive > 0 ifTrue: [methodArg isQuick ifTrue: [^self push: (methodArg valueWithReceiver: thisReceiver arguments: args)]. ^self doPrimitive: methodArg primitive method: methodArg receiver: thisReceiver args: args]. ^self activateMethod: methodArg withArgs: args receiver: thisReceiver]. primitiveIndex = 118 ifTrue: "[receiver:]tryPrimitive:withArgs:; avoid recursing in the VM" [(arguments size = 3 and: [(self objectClass: arguments second) == SmallInteger and: [(self objectClass: arguments last) == Array]]) ifTrue: [^self doPrimitive: arguments second method: meth receiver: arguments first args: arguments last]. (arguments size = 2 and: [(self objectClass: arguments first) == SmallInteger and: [(self objectClass: arguments last) == Array]]) ifFalse: [^self class primitiveFailTokenFor: -3]. ^self doPrimitive: arguments first method: meth receiver: receiver args: arguments last]. value := primitiveIndex = 120 "FFI method" ifTrue: [(meth literalAt: 1) tryInvokeWithArguments: arguments] ifFalse: [primitiveIndex = 117 "named primitives" ifTrue: [self tryNamedPrimitiveIn: meth for: receiver withArgs: arguments] ifFalse: "should use self receiver: receiver tryPrimitive: primitiveIndex withArgs: arguments but this is only in later VMs (and appears to be broken)" [receiver tryPrimitive: primitiveIndex withArgs: arguments]]. ^(self isPrimFailToken: value) ifTrue: [value] ifFalse: [self push: value]! Item was changed: ----- Method: Context>>runSimulated:contextAtEachStep: (in category 'system simulation') ----- runSimulated: aBlock contextAtEachStep: anotherBlock + "Simulate the execution of the argument, aBlock, until it ends or is curtailed. If any exception is signaled during the execution, simulate it being handled on the present caller stack. Evaluate anotherBlock with the current context prior to each instruction executed. Answer the simulated value of aBlock. + + If aBlock attempts any action that escapes the control of the simulator (e.g., resuming another process), signal a SimulationSideEffectWarning. The client of the simulator (i.e., the sender of this method) may handle those warnings to remain control, or otherwise, these actions are allowed. Examples: + Context runSimulated: [[] fork]. --> fork is executed + [Context runSimulated: [[] fork]] + on: SimulationSideEffectWarning + do: [:ex | + ex isControlPrimitive ifTrue: [ex return: #forbidden]. + ex pass]. --> fork is prevented" - "Simulate the execution of the argument, aBlock, until it ends or is curtailed. If any exception is signaled during the execution, simulate it being handled on the present caller stack. Evaluate anotherBlock with the current context prior to each instruction executed. Answer the simulated value of aBlock." | current resume ensure | resume := false. "Affect the context stack of the receiver during the simulation of aBlock." current := aBlock asContextWithSender: self. "Insert outer context denoting the end of the simulation." ensure := (ensure := current) insertSender: (Context contextEnsure: [resume := true. ensure privSender: thisContext home sender]). (anotherBlock numArgs = 0 ifTrue: ["optimized" [resume]] ifFalse: ["stop execution on time, don't expose simulation details to caller" [current == ensure or: ["Context >> #resume:" current size >= 2 and: [(current at: 2) == ensure]]] ]) whileFalse: [anotherBlock cull: current. current := current step]. "Continue with the execution in the previous context." ^ current jump! Item was added: + Warning subclass: #SimulationSideEffectWarning + instanceVariableNames: 'primitiveIndex context method receiver arguments suppressed' + classVariableNames: '' + poolDictionaries: '' + category: 'Kernel-Exceptions'! + + !SimulationSideEffectWarning commentStamp: 'ct 3/3/2024 19:17' prior: 0! + I am signaled to notify the client of the simulator (i.e., a sender of Context>>step) about potential side effects of the next instruction to be executed that would escape the control of the simulator. For example, I am signaled before the simulated code starts another process. See Context>>doPrimitive:method:receiver:args:, my messageText, and Parser>>simulationGuard for more information.! Item was added: + ----- Method: SimulationSideEffectWarning class>>forPrimitive: (in category 'instance creation') ----- + forPrimitive: primitiveIndex + + ^ self new primitive: primitiveIndex! Item was added: + ----- Method: SimulationSideEffectWarning>>arguments (in category 'accessing') ----- + arguments + + ^ arguments! Item was added: + ----- Method: SimulationSideEffectWarning>>context (in category 'accessing') ----- + context + + ^ context! Item was added: + ----- Method: SimulationSideEffectWarning>>context:method:receiver:arguments: (in category 'initialize-release') ----- + context: aContext method: aCompiledMethod receiver: rcvr arguments: args + + context := aContext. + method := aCompiledMethod. + receiver := rcvr. + arguments := args.! Item was added: + ----- Method: SimulationSideEffectWarning>>defaultAction (in category 'priv handling') ----- + defaultAction + + self suppressed ifFalse: [super defaultAction]. + self flag: #forLater. "When we support explicit exception handler invocation (e.g., #resume, #retry) from the debugger, this exception should publish a #resume handler rather than relying on the weakly defined proceed semantics of the debugger." + ^ self defaultResumeValue! Item was added: + ----- Method: SimulationSideEffectWarning>>defaultResumeValue (in category 'defaults') ----- + defaultResumeValue + + ^ true! Item was added: + ----- Method: SimulationSideEffectWarning>>isControlPrimitive (in category 'testing') ----- + isControlPrimitive + "See StackInterpreter class>>#initializePrimitiveTable." + + ^ self primitive between: 80 and: 89! Item was added: + ----- Method: SimulationSideEffectWarning>>isSimulationGuard (in category 'testing') ----- + isSimulationGuard + "See Parser >> #simulationGuard." + + ^ self primitive = 19! Item was added: + ----- Method: SimulationSideEffectWarning>>messageText (in category 'printing') ----- + messageText + + ^ messageText ifNil: [ + 'The code being simulated is trying to control a process ({1}). {2}' translated format: { + self context method reference. + self isSimulationGuard + ifTrue: ['If you proceed, your image may be locked. Continue at own risk, and better save your image before.' translated] + ifFalse: ['Process controlling cannot be simulated. If you proceed, side effects may occur outside the observable area of the simulator.' translated]}]! Item was added: + ----- Method: SimulationSideEffectWarning>>method (in category 'accessing') ----- + method + + ^ method! Item was added: + ----- Method: SimulationSideEffectWarning>>primitive (in category 'accessing') ----- + primitive + + ^ primitiveIndex! Item was added: + ----- Method: SimulationSideEffectWarning>>primitive: (in category 'initialize-release') ----- + primitive: anInteger + + primitiveIndex := anInteger.! Item was added: + ----- Method: SimulationSideEffectWarning>>signalIfSkipped: (in category 'signaling') ----- + signalIfSkipped: skipBlock + + ^ self signal ifFalse: skipBlock! Item was added: + ----- Method: SimulationSideEffectWarning>>skipPrimitive (in category 'handling') ----- + skipPrimitive + + ^ self resume: false! Item was added: + ----- Method: SimulationSideEffectWarning>>suppress (in category 'accessing') ----- + suppress + + suppressed := true.! Item was added: + ----- Method: SimulationSideEffectWarning>>suppressed (in category 'accessing') ----- + suppressed + + ^ suppressed ifNil: [self isSimulationGuard not]! Item was added: + ----- Method: SimulationSideEffectWarning>>theReceiver (in category 'accessing') ----- + theReceiver + + ^ receiver! Item was added: + ----- Method: SimulationSideEffectWarning>>unsuppress (in category 'accessing') ----- + unsuppress + + suppressed := false.!

1 0

Error when compiling the cog-vm with CLANG 16 on FreeBSD 14
by Edwin Ancaer 03 Mar '24

03 Mar '24

Hello, I recently upgraded to FreeBSD 14, with clang version 'FreeBSD clang version 16.0.6 (https://github.com/llvm/llvm-project.git llvmorg-16.0.6-0-g7cbf1a259152)' If I understand correctly, Clang promoted some warnings to errors and in special the warning Wincompatible-function-pointer-types. It looks as if this prevents me to compile the linux-cog-spur virtual machine: clang -Wall -g -O2 -DNDEBUG -DDEBUGVM=0 -msse2 -DCOGMTVM=0 -pthread -DLSB_FIRST=1 -m64 -Wno-missing-braces -Wno-unknown-pragmas -Wno-unused-value -Wno-unused-label -Wno-unused-function -Wno-unused-variable -Wno-unused-but-set-variable -DHAVE_CONFIG_H -DSQUEAK_BUILTIN_PLUGIN -I/home/edwin/Squeak/opensmalltalk-vm/building/linux64x64/squeak.cog.spur/build -I/home/edwin/Squeak/opensmalltalk-vm/building/linux64x64/squeak.cog.spur/build -I/home/edwin/Squeak/opensmalltalk-vm/platforms/unix/vm -I/home/edwin/Squeak/opensmalltalk-vm/platforms/Cross/vm -I/home/edwin/Squeak/opensmalltalk-vm/src/spur64.cog -I/home/edwin/Squeak/opensmalltalk-vm/platforms/Cross/vm -I/home/edwin/Squeak/opensmalltalk-vm/platforms/unix/vm -I/home/edwin/Squeak/opensmalltalk-vm/src/spur64.cog -I/home/edwin/Squeak/opensmalltalk-vm/platforms/Cross/plugins/FilePlugin -I/home/edwin/Squeak/opensmalltalk-vm/platforms/unix/plugins/B3DAcceleratorPlugin -m64 -Wno-missing-braces -Wno-unknown-pragmas -Wno-unused-value -Wno-unused-label -Wno-unused-function -Wno-unused-variable -Wno-unused-but-set-variable -c -o sqUnixMain.o /home/edwin/Squeak/opensmalltalk-vm/platforms/unix/vm/sqUnixMain.c /home/edwin/Squeak/opensmalltalk-vm/platforms/unix/vm/sqUnixMain.c:1832:10: warning: format string is not a string literal (potentially insecure) [-Wformat-security] printf(optionStrings[count]); ^~~~~~~~~~~~~~~~~~~~ /home/edwin/Squeak/opensmalltalk-vm/platforms/unix/vm/sqUnixMain.c:1832:10: note: treat the string as an argument to avoid this printf(optionStrings[count]); ^ "%s", /home/edwin/Squeak/opensmalltalk-vm/platforms/unix/vm/sqUnixMain.c:2324:38: error: incompatible function pointer types assigning to 'void (*)(int, struct __siginfo *, void *)' from 'void (int, siginfo_t *, ucontext_t *)' (aka 'void (int, struct __siginfo *, struct __ucontext *)') [-Wincompatible-function-pointer-types] sigsegv_handler_action.sa_sigaction = sigsegv; ^ ~~~~~~~ /home/edwin/Squeak/opensmalltalk-vm/platforms/unix/vm/sqUnixMain.c:2331:38: error: incompatible function pointer types assigning to 'void (*)(int, struct __siginfo *, void *)' from 'void (int, siginfo_t *, ucontext_t *)' (aka 'void (int, struct __siginfo *, struct __ucontext *)') [-Wincompatible-function-pointer-types] sigusr1_handler_action.sa_sigaction = sigusr1; ^ ~~~~~~~ 1 warning and 2 errors generated. *** Error code 1 Stop. make[1]: stopped in /home/edwin/Squeak/opensmalltalk-vm/building/linux64x64/squeak.cog.spur/build/vm *** Error code 1 Stop. make: stopped in /home/edwin/Squeak/opensmalltalk-vm/building/linux64x64/squeak.cog.spur/build [edwin@ottopedi ~/Squeak/opensmalltalk-vm/building/linux64x64/squeak.cog.spur/build]$ If my assumption is correct, the 'solution seems to downgrade the error to a warning with a compiler option: -Wno-error=, but I have no idea how to do this in the whole configuration and compilation process. Can someone tell how this should be done, or is it better to signal this as an error and wait until it gets corrected. Kind regards Edwin Ancaer

4 13

The Trunk: Tools-ct.1248.mcz
by commits＠source.squeak.org 02 Mar '24

02 Mar '24

Christoph Thiede uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-ct.1248.mcz ==================== Summary ==================== Name: Tools-ct.1248 Author: ct Time: 2 March 2024, 10:10:36.218052 pm UUID: e6b53eab-8cfd-3148-812e-4997678d65f5 Ancestors: Tools-ct.1247 In inspectors, fixes MNU for pressing Cmd+v in the value list. =============== Diff against Tools-ct.1247 =============== Item was changed: ----- Method: Inspector>>inspectorKey:from: (in category 'menu') ----- inspectorKey: aChar from: view "Respond to a Command key issued while the cursor is over my field list" ^ aChar caseOf: { [$x] -> [self removeSelection]. [$i] -> [self inspectSelection]. [$I] -> [self exploreSelection]. [$b] -> [self browseClass]. [$h] -> [self browseClassHierarchy]. [$p] -> [self browseFullProtocol]. [$r] -> [self browseVariableReferences]. [$a] -> [self browseVariableAssignments]. [$N] -> [self browseClassRefs]. [$c] -> [self copyName]. [$t] -> [self tearOffTile]. + [$v] -> [self selectionOrObject beViewed] } - [$v] -> [self viewerForValue] } otherwise: [self arrowKey: aChar from: view]!

1 0

source.squeak.org very slow today
by christoph.thiede＠student.hpi.uni-potsdam.de 02 Mar '24

02 Mar '24

Hi all, installing the update stream or even downloading a single version takes much longer than usually today (30 seconds or longer per version). Is there anyone who could press a magic button to make the server fast again? :-) Best, Christoph --- Sent from Squeak Inbox Talk

1 1

The Trunk: Kernel-ct.1558.mcz
by commits＠source.squeak.org 02 Mar '24

02 Mar '24

Christoph Thiede uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-ct.1558.mcz ==================== Summary ==================== Name: Kernel-ct.1558 Author: ct Time: 2 March 2024, 8:30:30.372052 pm UUID: b8e15dd5-2d06-b148-b83b-083c3e6aaaa9 Ancestors: Kernel-ct.1557 Repairs the comment in Object>>instVarNamed:put:. =============== Diff against Kernel-ct.1557 =============== Item was changed: ----- Method: Object>>instVarNamed:put: (in category 'system primitives') ----- instVarNamed: aString put: aValue + "Store the value into the instance variable in me with that name. Slow and unclean, but very useful. " - "Store into the value of the instance variable in me of that name. Slow and unclean, but very useful. " ^self instVarAt: (self class instVarIndexFor: aString asString ifAbsent: [self error: 'no such inst var' translated]) put: aValue !

1 0

Context >> #tempNamed:(put:)
by Thiede, Christoph 02 Mar '24

02 Mar '24

Hi all, I know we are in feature freeze, but that won't stop me from having {fun with, new ideas for} Squeak :-) So please have a look at the following mini-changeset. It implements #tempNamed: and #tempNamed:put: on Context, which I personally find quite useful when debugging and working with Context objects. Print it: thisContext sender tempNamed: #value put: 42 :-) Feedback welcome as always! Best, Christoph <http://www.hpi.de/>

4 3

The Trunk: Tools-ct.1247.mcz
by commits＠source.squeak.org 02 Mar '24

02 Mar '24

Christoph Thiede uploaded a new version of Tools to project The Trunk: http://source.squeak.org/trunk/Tools-ct.1247.mcz ==================== Summary ==================== Name: Tools-ct.1247 Author: ct Time: 2 March 2024, 8:28:46.009052 pm UUID: 7ce4a4da-cc0f-1646-8421-a10fe85212de Ancestors: Tools-ct.1246 Merges temputils.1.cs: Adds Context>>tempNamed: and Context>>tempNamed:put:. Thanks to Eliot (emm) for the review! =============== Diff against Tools-ct.1246 =============== Item was added: + ----- Method: Context>>tempNamed: (in category '*Tools-debugger access') ----- + tempNamed: aString + "Answer the value of the temporary variable in me with that name. Slower than #tempAt:, but very useful." + + ^ self debuggerMap tempNamed: aString in: self! Item was added: + ----- Method: Context>>tempNamed:put: (in category '*Tools-debugger access') ----- + tempNamed: aString put: aValue + "Store the value into the temporary variable in me of that name. Slower than #tempAt:put:, but very useful." + + ^ self debuggerMap + tempNamed: aString + put: aValue + in: self! Item was added: + ----- Method: DebuggerMethodMap>>tempIndexFor:in:ifAbsent: (in category 'accessing') ----- + tempIndexFor: tempName in: aContext ifAbsent: aBlock + + ^ (self tempNamesForContext: aContext) + indexOf: tempName + ifAbsent: aBlock! Item was added: + ----- Method: DebuggerMethodMap>>tempNamed:in: (in category 'accessing') ----- + tempNamed: aString in: aContext + "Answer the value of the temporary variable in aContext with that name. Slower than Context>>tempAt:, but very useful." + + ^ self + namedTempAt: (self + tempIndexFor: aString asString + in: aContext + ifAbsent: [self error: 'no such temp']) + in: aContext! Item was added: + ----- Method: DebuggerMethodMap>>tempNamed:put:in: (in category 'accessing') ----- + tempNamed: aString put: aValue in: aContext + "Store into the value of the temporary variable in aContext of that name. Slower than Context>>tempAt:put:, but very useful." + + ^ self + namedTempAt: (self + tempIndexFor: aString asString + in: aContext + ifAbsent: [self error: 'no such temp']) + put: aValue + in: aContext!

1 0

The Trunk: Kernel-ct.1410.mcz
by commits＠source.squeak.org 02 Mar '24

02 Mar '24

Christoph Thiede uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-ct.1410.mcz ==================== Summary ==================== Name: Kernel-ct.1410 Author: ct Time: 30 August 2021, 8:43:40.47076 pm UUID: 46b520fa-b225-e349-8f11-a423359f0466 Ancestors: Kernel-eem.1408 Proposal: Makes IllegalResumeAttempt an error. Please find my argument in [1]. [1] http://lists.squeakfoundation.org/pipermail/squeak-dev/2021-May/215520.html =============== Diff against Kernel-eem.1408 =============== Item was changed: + Error subclass: #IllegalResumeAttempt - Exception subclass: #IllegalResumeAttempt instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'Kernel-Exceptions'! !IllegalResumeAttempt commentStamp: '<historical>' prior: 0! This class is private to the EHS implementation. An instance of it is signaled whenever an attempt is made to resume from an exception which answers false to #isResumable.! Item was removed: - ----- Method: IllegalResumeAttempt>>defaultAction (in category 'handling') ----- - defaultAction - "No one has handled this error, but now give them a chance to decide how to debug it. If none handle this either then open debugger (see UnhandedError-defaultAction)" - - UnhandledError signalForException: self! Item was removed: - ----- Method: IllegalResumeAttempt>>isResumable (in category 'handling') ----- - isResumable - - ^ false! Item was removed: - ----- Method: IllegalResumeAttempt>>readMe (in category 'comment') ----- - readMe - - "Never handle this exception!!"!

1 0

Why is IllegalResumeAttempt not an Error?
by Thiede, Christoph 02 Mar '24

02 Mar '24

Hi all! I was just wondering why IllegalResumeAttempt is not an error. An IllegalResumeAttempt is signaled when you try to #resume from an exception that is not resumable. In my opinion, this should pretty clearly be an Error. Just like "nil + 1", "#() at: 0", or "'hello' at: 1 put: $H", the programmer is trying to perform an illegal operation so the EHS notices this error. Why should "Error new resume" (pseudo, of course) not be an Error? I could imagine that in ancient days, when nested exception handling did not yet work as flawlessly as today (thanks Jaromir and Nicolas for all the work on this field!), the original designers tried to prevent the case in which a second exception was handled on the stack. Nevertheless, this limitation does not exist any longer today. It would not even be a problem (although not a recommended pattern) to say something like: [self error + 1] on: Error do: [:ex1 | [ex1 resume: 42] on: IllegalResumeAttempt do: [:ex2 | "Well, then return instead" ex1 return: 42] ]. So why don't we just go and make IllegalResumeAttempt an Error? Its defaultAction is already completely compatible. Best, Christoph

4 4