Vm-dev December 2010

vm-dev@lists.squeakfoundation.org

42 participants
109 discussions

[COG] Random crash with markAndTrace
by Mariano Martinez Peck 14 Dec '10

14 Dec '10

Hi folks. I have a crash in Cog, and I am not sure if it is "my fault" or not. Basically, it crashes while during the markAndTrace of an incrementalGC. In my case, it always fails in the method #startObj , here: startObj "Start tracing the object 'child' and answer the next action. The object may be anywhere in the middle of being swept itself. See comment in markAndTrace for explanation of tracer state variables, and aComment for gory details." | header lastFieldOffset | <inline: true> statMarkCountLocal := statMarkCountLocal + 1. ((self oop: child isLessThan: youngStartLocal) "old object; skip it" or: [((header := self longAt: child) bitAnd: MarkBit) ~= 0]) "already marked; skip it" when it does the if.. In C is: if (((((usqInt) child)) < (((usqInt) youngStartLocal))) || ((((header1 = longAt(child))) & MarkBit) != 0)) { child is 1887948912 youngStartLocal is 466502688 I attach a screenshot with the XCode GDB. Thanks for any help in advance Mariano

3 4

Is SmallInteger a really too dangerous class?
by Mariano Martinez Peck 14 Dec '10

14 Dec '10

Hi folks. Sorry for the cross-post, I am not sure if it is image or vm problem. If I implement this: SmallInteger >> initialize self addInstVarName: 'Zaraza'. I have the error: 'SmallInteger cannot be changed' This is because SmallInteger is in the ClassBuilder >> tooDangerousClasses and the error is in ClassBuilder >> name: className inEnvironment: env subclassOf: newSuper type: type instanceVariableNames: instVarString classVariableNames: classVarString poolDictionaries: poolString category: category unsafe: unsafe Now....since this is Smalltalk, I can just comment something, continue, or any other kind of hack so that I can add class variables to SmallInetger. For my CompiledMethod proxies stuff, I am using SmallIntegers and I have some crashes (one is due to a corrupt heap and Eliot said something about sizes)...I wonder if this is related to this...So the question is: is really SMallInteger a too dangerous class so that it cannot be changed? not even adding a class variable??? does the VM assume that the class SmallInetger has a fixed size? if true, where? Thanks in advance Mariano

4 7

cog and crash.dmp
by Alexander Lazarević 14 Dec '10

14 Dec '10

Hi! Not my turf, so I'm a little puzzled! Does the latest cog vm for windows (7) also write a crash.dmp file to the vmpath when it crashes? The dialog says so, but I can't find the file in that directory. I have permission to create/write files in that directory.Is there some sort of switch/ini-Setting I have to set to make the vm dump that file on a crash? Alex

2 1

Cog on Mac. Problems with mprotect()?
by Igor Stasenko 14 Dec '10

14 Dec '10

I have no idea why mprotect() returns -1 (failure), and then perror prints following: mprotect(x,y,PROT_READ | PROT_WRITE | PROT_EXEC): Cannot allocate memory mprotect(x,y,PROT_READ | PROT_WRITE): Invalid argument While it looks like Cog VM runs stable at the same time. But it crashing at crucial point of NB tests, needless to say, that same tests are running on other platforms and everything ok. So, there is something wrong with it. The only question is what :) While in general, i can run native code, generated by NB, which means that memory execution is enabled, something leads to memory access error during GC, in test which checking if relocated native code will continue to run from expected point as if it was not relocated. What is strange with this thing, that i expected to see errors in my code. But why it crashes in the middle of pow() function, which is called from initializeMemoryFirstFree()?? Thread 0 Crashed: Dispatch queue: com.apple.main-thread 0 libSystem.B.dylib 0x98273176 __kill + 10 1 libSystem.B.dylib 0x98273168 kill$UNIX2003 + 32 2 libSystem.B.dylib 0x9830589d raise + 26 3 libSystem.B.dylib 0x9831b9bc abort + 93 4 com.teleplace.Teleplace 0x00090f86 error + 86 5 libSystem.B.dylib 0x9827846b _sigtramp + 43 6 ??? 0xffffffff 0 + 4294967295 7 com.teleplace.Teleplace 0x0010d468 initializeMemoryFirstFree + 184 8 com.teleplace.Teleplace 0x00113381 incCompBody + 1057 9 com.teleplace.Teleplace 0x001229ec fullGC + 1356 10 ??? 0x17203026 0 + 387985446 11 com.teleplace.Teleplace 0x0014fe98 primitiveNativeCall + 216 12 com.teleplace.Teleplace 0x00132422 executeNewMethod + 130 13 com.teleplace.Teleplace 0x00133111 primitiveExecuteMethodArgsArray + 497 14 com.teleplace.Teleplace 0x001365f6 interpret + 1078 15 com.teleplace.Teleplace 0x0013d94e enterSmalltalkExecutiveImplementation + 110 16 com.teleplace.Teleplace 0x0013db81 initStackPagesAndInterpret + 545 17 com.teleplace.Teleplace 0x00079494 EventLoopEventHandler + 132 The stack frame #11 is a primitive which calls my native code (frame #10) then the native code deliberately calls fullGC() (frame #9) and it seems like works, till something bad happen in pow() function, called by initializeMemoryFirstFree() Here the code near call to pow() function: 0x0010d43d <+0141> shr %ebx 0x0010d43f <+0143> add $0x1,%eax 0x0010d442 <+0146> mov 0x182be8,%edi 0x0010d448 <+0152> add %ebx,%eax 0x0010d44a <+0154> cvtsi2sd %eax,%xmm0 0x0010d44e <+0158> movsd %xmm0,0x8(%esp) 0x0010d454 <+0164> movl $0x0,(%esp) 0x0010d45b <+0171> movl $0x40000000,0x4(%esp) 0x0010d463 <+0179> call 0x163a84 <dyld_stub_pow> 0x0010d468 <+0184> fstpl -0x30(%ebp) 0x0010d46b <+0187> cvttsd2si -0x30(%ebp),%eax 0x0010d470 <+0192> add -0x1c(%ebp),%eax and here the code where it crashing: 0x98262f56 <+1078> jmp 0x98262b66 <pow$fenv_access_off+70> 0x98262f5b <+1083> movapd %xmm2,%xmm1 0x98262f5f <+1087> andpd 0x13e049(%ebx),%xmm1 0x98262f67 <+1095> movapd %xmm2,%xmm0 0x98262f6b <+1099> andpd 0x13e079(%ebx),%xmm0 *** 0x98262f73 <+1107> movapd %xmm0,-0x78(%ebp) 0x98262f78 <+1112> movapd %xmm1,%xmm0 0x98262f7c <+1116> lea 0x13e069(%ebx),%esi 0x98262f82 <+1122> addpd (%esi),%xmm0 0x98262f86 <+1126> andpd -0x78(%ebp),%xmm0 0x98262f8b <+1131> movapd %xmm0,-0x78(%ebp) 0x98262f90 <+1136> paddq 0x13cfb9(%ebx),%xmm1 i can only guess, that it either a problem with mprotect(), or that due to heavy use of MMX/SSE instructions, i miss some precautions (not saving some registers/processor state) between the calls in my code. Any ideas? -- Best regards, Igor Stasenko AKA sig.

3 5

VM Maker: VMMaker-dtl.210.mcz
by squeak-dev-noreply＠lists.squeakfoundation.org 14 Dec '10

14 Dec '10

Dave Lewis uploaded a new version of VMMaker to project VM Maker: http://www.squeaksource.com/VMMaker/VMMaker-dtl.210.mcz ==================== Summary ==================== Name: VMMaker-dtl.210 Author: dtl Time: 14 December 2010, 8:19:29 am UUID: 6bdedd54-4a61-4355-b8bc-4f65042f1776 Ancestors: VMMaker-dtl.209 VMMaker 4.4.1 Flush the at cache but not the entire method cache in #primitiveAdoptInstance. Fix a BaseHeaderSize reference that slipped in to the last update (should be self baseHeaderSize now).

1 0

[COG] primitiveIndexOf: aMethod crash in Cog when using SmallInteger
by Mariano Martinez Peck 14 Dec '10

14 Dec '10

Hi Eliot. I am not sure if you are interested in completly support SmallInteger as methods. If not, please let me know and I don't bather again. I was trying them in Cog, and I have a crash in primitiveFlushCacheByMethod() when it calls primitiveIndexOf: because primitiveIndexOf: methodPointer "Note: We now have 10 bits of primitive index, but they are in two places for temporary backward compatibility. The time to unpack is negligible, since the derived primitive function pointer is stored in the method cache." <api> | primBits | primBits := ((self headerOf: methodPointer) >> 1) bitAnd: 16r100001FF. ^(primBits bitAnd: 16r1FF) + (primBits >> 19) But....."methodPointer" can be a Smallnteger...so headerOf: fails, because rawHeaderOf: fails because it does a "objectMemory fetchPointer: HeaderIndex ofObject: methodPointer" which send ObjectMemory >> fetchPointer: fieldIndex ofObject: oop which does the final "self longAt: oop + BaseHeaderSize + (fieldIndex << ShiftForWord)" ok, you got it ;) Now....do you consider this a bug? how could I fix it ? Thanks Mariano

1 1

Works in SqueakVM but not Cog
by Igor Stasenko 14 Dec '10

14 Dec '10

| newNil | newNil := UndefinedObject basicNew. nil becomeForward: newNil. (Array new: 1) first == newNil In Cog it just hangs VM (under Mac) Any ideas how to fix that? -- Best regards, Igor Stasenko AKA sig.

3 6

Re: [Pharo-project] Fwd: [Vm-dev] Image freeze because handleTimerEvent and Seaside process gone?!
by Eliot Miranda 13 Dec '10

13 Dec '10

On Tue, Dec 7, 2010 at 3:41 AM, Lukas Renggli <renggli(a)gmail.com> wrote: > > One change that I don't understand, although it probably is unrelated, is > in [2]: > > > > LargePositiveInteger removeSelector: #=! > > LargePositiveInteger removeSelector: #bitAnd:! > > LargePositiveInteger removeSelector: #bitOr:! > > LargePositiveInteger removeSelector: #bitShift:! > > LargePositiveInteger removeSelector: #bitXor:! > > LargePositiveInteger removeSelector: #'~='! > > > > Why would one want to remove these primitive calls from large integers? > > AFAIK, Cog needs those. > That's right. Those methods used SmallInteger primitives (7, 8, 14, 15, 16 & 17) and in Cog SmallInteger primitives only work on SmallIntegers, not on up to 64-bit LargeIntegers; i.e. they don't bother to test the receiver for being a SmallInteger and hence crash if installed on LargeInteger. I probably made a mistake in deciding to save a tag test, but I did. Instead these methods should use the relevant LargeInteger primitives. The issue is I think those are missing in the standard VM. See primitives 27 28 34 35 36 & 37 primitiveEqualLargeIntegers, primitiveNotEqualLargeIntegers, primitiveBitAndLargeIntegers, primitiveBitOrLargeIntegers, primitiveBitXorLargeIntegers, primitiveBitShiftLargeIntegers in the Cog VM and perhaps the standard VM. If they're missing in teh standard VM it should be fine to redefine those methods to include the 27 28 34 35 36 & 37 primitive numbers. If they're defined differently someone needs to check that using those numbers works in the standard VM. HTH Eliot > > Lukas > > -- > Lukas Renggli > www.lukas-renggli.ch > >

2 1

[OT] why CoInterpreter and not CogInterpreter ?
by Mariano Martinez Peck 12 Dec '10

12 Dec '10

Co is for co-exist ? and not for cog ? thanks mariano

2 2

Re: [squeak-dev] Re: OSProcess on Mac?
by David T. Lewis 12 Dec '10

12 Dec '10

Forwarding to vm-dev: Does anyone have a recent OSProcessPlugin for Mac that they can provide to Bernhard? Thanks, Dave On Sun, Dec 12, 2010 at 05:53:12PM +0100, Bernhard Pieber wrote: > Hi David, > > Thanks for your detailed explanations and your help. > > Am 12.12.2010 um 17:46 schrieb David T. Lewis: > > Some of the recent Mac VMs were built with a very out of date version > > of OSProcess that causes problems on Mac, including crashes when > > performing #forkSqueak. I'm not sure if the 4.2.5 VM has this issue, > > but you check by evaluating "OSProcess accessor osppModuleVersionString". > > If the result is '4.0.1' then you have the outdated plugin. You need > > at least version 4.0.2 to run safely on a Mac (the current version of > > OSPP is 4.4.0). > This seems to be my problem. I have the outdated 4.0.1. Can I get the most recent one built for Mac from somewhere? > > Cheers, > Bernhard

1 0

← Newer
1
...
5
6
7
8
9
10
11
Older →

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Vm-dev December 2010