January 2018 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] [ANN] Fari â a shell tool to fetch & prepare your fresh images

by Stephane Ducasse

On Fri, Jan 19, 2018 at 7:02 PM, Damien Pollet <damien.pollet(a)gmail.com> wrote: > Here it is, tell me what you think: https://github.com/cdlm/fari.sh > > We discussed it with Christophe and of course there is some overlap with > PharoLauncher, and in the future, it might get replaced by a command-line > version of PharoLauncher based on Clap. > Yes we want that too. Guille was talking about this too :) Stef > > But for the moment it works and fits my twisted mind better ð > > -- > Damien Pollet > type less, do more [ | ] http://people.untyped.org/damien.pollet >

Jan. 20, 2018

Re: [Pharo-dev] Random corrupted data when copying from very large byte array

by ClÃ©ment Bera

Ok so the bug happens in VMMaker 2265 while the change in prim 105 was introduced in 2273, so that's not it (I was surprised, there is lots of tests for this). It seems each time I have the problem, it happens when totalReadCount reach 18186240. I guess that matches some specific array growth strategy ? There are multiple different paths in the code dealing with this case. --- We had issues in the past with signed vs unsigned operations and things like that which would make some operations fails for large integers but it does not seem to be related in this case (I don't see any integer overflowing in the 32 bits range instead of 31 bits). I don't think that's related to the C compiler either since the bug happen in Mac and Linux which are compiled with llvm and gcc respectively. --- I tried in Pharo 5, with recent VM it fails, older VMs it works. I tried once for each though. There is no simple code to debug, so I don't really know where to look. I tried to wrap prim 105 with assertion and they all pass so the primitive may not be faulty (I could check the diff with old versions though...). On Fri, Jan 19, 2018 at 9:53 PM, Alistair Grant <akgrant0710(a)gmail.com> wrote: > Hi ClÃ©ment, > > On 19 January 2018 at 17:21, Alistair Grant <akgrant0710(a)gmail.com> wrote: > > Hi ClÃ©ment, > > > > On 19 January 2018 at 17:04, ClÃ©ment Bera <bera.clement(a)gmail.com> > wrote: > >> Does not seem to be related to prim 105. > >> > >> I am confused. Has the size of the array an impact at all ? > > > > Yes, I tried reducing the size of the array by a factor of 10 and > > wasn't able to reproduce the problem at all. > > > > With the full size array it failed over half the time (32 bit). > > > > I ran the test about 180 times on 64 bit and didn't get a single failure. > > > >> It seems the > >> problem shows since the first copy of 16k elements. > >> > >> I can't really reproduce the bug - it happened once but I cannot do it > >> again. Does the bug happen with the StackVM/PharoS VM you can find here > the > >> 32 bits versions : http://files.pharo.org/vm/pharoS-spur32/ ? The > >> StackVM/PharoS VM is the VM without the JIT, it may be since the bug is > >> unreliable that it happens only in jitted code, so trying that out may > be > >> worth it. > > > > I'll try and have a look at this over the weekend. > > This didn't fail once in 55 runs. > > OS: Ubuntu 16.04 > Image: Pharo 6.0 Latest update: #60528 > VM: > 5.0 #1 Wed Oct 12 15:48:53 CEST 2016 gcc 4.6.3 [Production Spur ITHB VM] > StackInterpreter VMMaker.oscog-EstebanLorenzano.1881 uuid: > ed616067-a57c-409b-bfb6-dab51f058235 Oct 12 2016 > https://github.com/pharo-project/pharo-vm.git Commit: > 01a03276a2e2b243cd4a7d3427ba541f835c07d3 Date: 2016-10-12 14:31:09 > +0200 By: Esteban Lorenzano <estebanlm(a)gmail.com> Jenkins build #606 > Linux pharo-linux 3.2.0-31-generic-pae #50-Ubuntu SMP Fri Sep 7 > 16:39:45 UTC 2012 i686 i686 i386 GNU/Linux > plugin path: /home/alistair/pharo7/Issue20982/bin/ [default: > /home/alistair/pharo7/Issue20982/bin/] > > > I then went back and attempted to reproduce the failures in my regular > 32 bit image, but only got 1 corruption in 10 runs. I've been working > in this image without restarting for most of the day. > > Quitting out and restarting the image and then running the corruption > check resulted in 11 corruptions from 11 runs. > > > Image: Pharo 7.0 Build information: > Pharo-7.0+alpha.build.425.sha.eb0a6fb140ac4a42b1f158ed37717e0650f778b4 > (32 Bit) > VM: > 5.0-201801110739 Thursday 11 January 09:30:12 CET 2018 gcc 4.8.5 > [Production Spur VM] > CoInterpreter VMMaker.oscog-eem.2302 uuid: > 55ec8f63-cdbe-4e79-8f22-48fdea585b88 Jan 11 2018 > StackToRegisterMappingCogit VMMaker.oscog-eem.2302 uuid: > 55ec8f63-cdbe-4e79-8f22-48fdea585b88 Jan 11 2018 > VM: 201801110739 > alistair@alistair-xps13:snap/pharo-snap/pharo-vm/opensmalltalk-vm $ > Date: Wed Jan 10 23:39:30 2018 -0800 $ > Plugins: 201801110739 > alistair@alistair-xps13:snap/pharo-snap/pharo-vm/opensmalltalk-vm $ > Linux b07d7880072c 4.13.0-26-generic #29~16.04.2-Ubuntu SMP Tue Jan 9 > 22:00:44 UTC 2018 i686 i686 i686 GNU/Linux > plugin path: /snap/core/3748/lib/i386-linux-gnu/ [default: > /snap/core/3748/lib/i386-linux-gnu/] > > > So, as well as restarting the image before running the test, just > wondering if the gcc compiler version could have an impact? > > > HTH, > Alistair > > > > > Cheers, > > Alistair > > > > > > > >> On Thu, Jan 18, 2018 at 7:12 PM, ClÃ©ment Bera <bera.clement(a)gmail.com> > >> wrote: > >>> > >>> I would suspect a bug in primitive 105 on byte objects (it was changed > >>> recently in the VM), called by copyFrom: 1 to: readCount. The bug would > >>> likely by due to specific alignment in readCount or something like > that. > >>> (Assuming you're in 32 bits since the 4 bytes are corrupted). > >>> > >>> When I get better I can have a look (I am currently quite sick). > >>> > >>> On Thu, Jan 18, 2018 at 4:51 PM, Thierry Goubier > >>> <thierry.goubier(a)gmail.com> wrote: > >>>> > >>>> Hi Cyril, > >>>> > >>>> try with the last vms available at: > >>>> > >>>> https://bintray.com/opensmalltalk/vm/cog/ > >>>> > >>>> For example, the last Ubuntu 64bits vm is at: > >>>> > >>>> https://bintray.com/opensmalltalk/vm/cog/201801170946#files > >>>> > >>>> Regards, > >>>> > >>>> Thierry > >>>> > >>>> 2018-01-18 16:42 GMT+01:00 Cyrille Delaunay <cy.delaunay(a)gmail.com>: > >>>>> > >>>>> Hi everyone, > >>>>> > >>>>> I just added a new bug entry for an issue we are experimenting since > >>>>> some times: > >>>>> > >>>>> > >>>>> https://pharo.fogbugz.com/f/cases/20982/Random-corrupted- > data-when-copying-from-very-large-byte-array > >>>>> > >>>>> Here is the description: > >>>>> > >>>>> > >>>>> History: > >>>>> > >>>>> This issue has been spotted after experimenting strange behavior with > >>>>> seaside upload. > >>>>> After uploading a big file from a web browser, the modeled file > >>>>> generated within pharo image begins with 4 unexpected bytes. > >>>>> This issue occurs randomly: sometimes the first 4 bytes are right. > >>>>> Sometimes the first 4 bytes are wrong. > >>>>> This issue only occurs with Pharo 6. > >>>>> This issue occurs for all platforms (Mac, Ubuntu, Windows) > >>>>> > >>>>> Steps to reproduce: > >>>>> > >>>>> I have been able to set up a small scenario that highlight the issue. > >>>>> > >>>>> Download Pharo 6.1 on my Mac (Sierra 10.12.6): > >>>>> https://pharo.org/web/download > >>>>> Then, iterate over this process till spotting the issue: > >>>>> > >>>>> => start the pharo image > >>>>> => execute this piece of code in a playground > >>>>> > >>>>> 1: > >>>>> 2: > >>>>> 3: > >>>>> 4: > >>>>> 5: > >>>>> 6: > >>>>> > >>>>> ZnServer startDefaultOn: 1701. > >>>>> ZnServer default maximumEntitySize: 80* 1024 * 1024. > >>>>> '/Users/cdelaunay/myzip.zip' asFileReference writeStreamDo: [ :out | > >>>>> out binary; nextPutAll: #[80 75 3 4 10 0 0 0 0 0 125 83 67 73 0 0 0 > 0 0 > >>>>> 0]. > >>>>> 18202065 timesRepeat: [ out nextPut: 0 ] > >>>>> ]. > >>>>> > >>>>> => Open a web browser page on: http://localhost:1701/form-test-3 > >>>>> => Upload the file zip file previously generated ('myzip.zip'). > >>>>> => If the web page displays: "contents=000000000a00..." (or anything > >>>>> unexpected), THIS IS THE ISSUE ! > >>>>> => If the web page displays: "contents=504b03040a00..", the upload > >>>>> worked fine. You can close the image (without saving) > >>>>> > >>>>> > >>>>> > >>>>> Debugging: > >>>>> > >>>>> > >>>>> > >>>>> Bob Arning has been able to reproduce the issue with my scenario. > >>>>> He dived into the code involved during this process, till reaching > some > >>>>> "basic" methods where he saw the issue occuring. > >>>>> > >>>>> Here are the conclusion till there: > >>>>> => A corruption occurs while reading an input stream with method > ZnUtils > >>>>> class>>readUpToEnd:limit: > >>>>> The first 4 bytes may be altered randomely. > >>>>> => The first 4 bytes are initially correctly written to an > outputStream. > >>>>> But, the first 4 bytes of this outputStream gets altered (corrupted), > >>>>> sometimes when the inner byte array grows OR when performing the > final > >>>>> "outputStream contents" > >>>>> => Here is a piece of code that reproduce the issue (still randomely. > >>>>> stopping an restarting the image may change the behavior) > >>>>> > >>>>> 1: > >>>>> 2: > >>>>> 3: > >>>>> 4: > >>>>> 5: > >>>>> 6: > >>>>> 7: > >>>>> 8: > >>>>> 9: > >>>>> 10: > >>>>> 11: > >>>>> 12: > >>>>> 13: > >>>>> 14: > >>>>> 15: > >>>>> 16: > >>>>> 17: > >>>>> 18: > >>>>> 19: > >>>>> 20: > >>>>> > >>>>> test4"self test4" | species bufferSize buffer totalRead > outputStream > >>>>> answer inputStream ba byte1 | ba := ByteArray new: > 18202085. > >>>>> ba atAllPut: 99. 1 to: 20 do: [ :i | ba at: i put: (#[80 75 > 3 4 10 7 > >>>>> 7 7 7 7 125 83 67 73 7 7 7 7 7 7] at: i) ]. inputStream := ba > readStream. > >>>>> bufferSize := 16384. species := ByteArray. > >>>>> buffer := species new: bufferSize. > >>>>> totalRead := 0. > >>>>> outputStream := nil. > >>>>> [ inputStream atEnd ] whileFalse: [ | readCount | > >>>>> readCount := inputStream readInto: buffer startingAt: 1 > count: > >>>>> bufferSize. > >>>>> totalRead = 0 ifTrue: [ > >>>>> byte1 := buffer first. > >>>>> ]. > >>>>> totalRead := totalRead + readCount. > >>>>> > >>>>> outputStream ifNil: [ > >>>>> inputStream atEnd > >>>>> ifTrue: [ ^ buffer copyFrom: 1 to: readCount ] > >>>>> ifFalse: [ outputStream := (species new: bufferSize) > >>>>> writeStream ] ]. > >>>>> outputStream next: readCount putAll: buffer startingAt: 1. > >>>>> byte1 = outputStream contents first ifFalse: [ self halt ]. > >>>>> ]. > >>>>> answer := outputStream ifNil: [ species new ] ifNotNil: [ > >>>>> outputStream contents ]. > >>>>> byte1 = answer first ifFalse: [ self halt ]. ^answer > >>>>> > >>>>> > >>>>> > >>>>> suspicions > >>>>> > >>>>> This issue appeared with Pharo 6. > >>>>> > >>>>> Some people suggested that it could be a vm issue, and to try my > little > >>>>> scenario with the last available vm. > >>>>> > >>>>> I am not sure where to find the last available vm. > >>>>> > >>>>> I did the test using these elements: > >>>>> > >>>>> https://files.pharo.org/image/60/latest.zip > >>>>> > >>>>> https://files.pharo.org/get-files/70/pharo-mac-latest.zip/ > >>>>> > >>>>> > >>>>> > >>>>> The issue is still present > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> -- > >>>>> Cyrille Delaunay > >>>> > >>>> > >>> > >>> > >>> > >>> -- > >>> ClÃ©ment BÃ©ra > >>> Pharo consortium engineer > >>> https://clementbera.wordpress.com/ > >>> BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq > >> > >> > >> > >> > >> -- > >> ClÃ©ment BÃ©ra > >> Pharo consortium engineer > >> https://clementbera.wordpress.com/ > >> BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq > > -- ClÃ©ment BÃ©ra Pharo consortium engineer https://clementbera.wordpress.com/ BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq

Jan. 19, 2018

Re: [Pharo-dev] Random corrupted data when copying from very large byte array

by Eliot Miranda

Hi Alistair, Hi ClÃ©ment, On Fri, Jan 19, 2018 at 12:53 PM, Alistair Grant <akgrant0710(a)gmail.com> wrote: > Hi ClÃ©ment, > > On 19 January 2018 at 17:21, Alistair Grant <akgrant0710(a)gmail.com> wrote: > > Hi ClÃ©ment, > > > > On 19 January 2018 at 17:04, ClÃ©ment Bera <bera.clement(a)gmail.com> > wrote: > >> Does not seem to be related to prim 105. > >> > >> I am confused. Has the size of the array an impact at all ? > > > > Yes, I tried reducing the size of the array by a factor of 10 and > > wasn't able to reproduce the problem at all. > > > > With the full size array it failed over half the time (32 bit). > > > > I ran the test about 180 times on 64 bit and didn't get a single failure. > > > >> It seems the > >> problem shows since the first copy of 16k elements. > >> > >> I can't really reproduce the bug - it happened once but I cannot do it > >> again. Does the bug happen with the StackVM/PharoS VM you can find here > the > >> 32 bits versions : http://files.pharo.org/vm/pharoS-spur32/ ? The > >> StackVM/PharoS VM is the VM without the JIT, it may be since the bug is > >> unreliable that it happens only in jitted code, so trying that out may > be > >> worth it. > > > > I'll try and have a look at this over the weekend. > > This didn't fail once in 55 runs. > > OS: Ubuntu 16.04 > Image: Pharo 6.0 Latest update: #60528 > VM: > 5.0 #1 Wed Oct 12 15:48:53 CEST 2016 gcc 4.6.3 [Production Spur ITHB VM] > StackInterpreter VMMaker.oscog-EstebanLorenzano.1881 uuid: > ed616067-a57c-409b-bfb6-dab51f058235 Oct 12 2016 > https://github.com/pharo-project/pharo-vm.git Commit: > 01a03276a2e2b243cd4a7d3427ba541f835c07d3 Date: 2016-10-12 14:31:09 > +0200 By: Esteban Lorenzano <estebanlm(a)gmail.com> Jenkins build #606 > Linux pharo-linux 3.2.0-31-generic-pae #50-Ubuntu SMP Fri Sep 7 > 16:39:45 UTC 2012 i686 i686 i386 GNU/Linux > plugin path: /home/alistair/pharo7/Issue20982/bin/ [default: > /home/alistair/pharo7/Issue20982/bin/] > > > I then went back and attempted to reproduce the failures in my regular > 32 bit image, but only got 1 corruption in 10 runs. I've been working > in this image without restarting for most of the day. > > Quitting out and restarting the image and then running the corruption > check resulted in 11 corruptions from 11 runs. > > > Image: Pharo 7.0 Build information: > Pharo-7.0+alpha.build.425.sha.eb0a6fb140ac4a42b1f158ed37717e0650f778b4 > (32 Bit) > VM: > 5.0-201801110739 Thursday 11 January 09:30:12 CET 2018 gcc 4.8.5 > [Production Spur VM] > CoInterpreter VMMaker.oscog-eem.2302 uuid: > 55ec8f63-cdbe-4e79-8f22-48fdea585b88 Jan 11 2018 > StackToRegisterMappingCogit VMMaker.oscog-eem.2302 uuid: > 55ec8f63-cdbe-4e79-8f22-48fdea585b88 Jan 11 2018 > VM: 201801110739 > alistair@alistair-xps13:snap/pharo-snap/pharo-vm/opensmalltalk-vm $ > Date: Wed Jan 10 23:39:30 2018 -0800 $ > Plugins: 201801110739 > alistair@alistair-xps13:snap/pharo-snap/pharo-vm/opensmalltalk-vm $ > Linux b07d7880072c 4.13.0-26-generic #29~16.04.2-Ubuntu SMP Tue Jan 9 > 22:00:44 UTC 2018 i686 i686 i686 GNU/Linux > plugin path: /snap/core/3748/lib/i386-linux-gnu/ [default: > /snap/core/3748/lib/i386-linux-gnu/] > > > So, as well as restarting the image before running the test, just > wondering if the gcc compiler version could have an impact? > I suspect that the problem is the same compactor bug I've been trying to reproduce all week, and have just fixed. Could you try and reproduce with a VM built from the latest commit? Some details: The SpurPlanningCompactor works by using the fact that all Spur objects have room for a forwarding pointer. The compactor make three passes: - the first pass through memory works out where objects will go, replacig their first fields with where they will go, and saving their first fields in a buffer (savedFirstFieldsSpace). - the second pass scans all pointer objects, replacing their fields with where the objects referenced will go (following the forwarding pointers), and also relocates any pointer fields in savedFirstFieldsSpace - the final pass slides objects down, restoring their relocated first fields The buffer used for savedFirstFieldsSpace determines how many passes are used. The system will either use eden (which is empty when compaction occurs) or a large free chunk or allocate a new segment, depending on whatever yields the largest space. So in the right circumstances eden will be used and more than one pass required. The bug was that when multiple passes are used the compactor forgot to unmark the corpse left behind when the object was moved. Instead of the corpse being changed into free space it was retained, but its first field would be that of the forwarding pointer to its new location, not the actual first field. So on 32-bits a ByteArray that should have been collected would have its first 4 bytes appear to be invalid, and on 64-bits its first 8 bytes. Because the heap on 64-bits can grow larger it could be that the bug shows itself much less frequently than on 32-bits. When compaction can be completed in a single pass all corpses are correctly collected, so most of the time the bug is hidden. This is the commit: commit 0fe1e1ea108e53501a0e728736048062c83a66ce Author: Eliot Miranda <eliot.miranda(a)gmail.com> Date: Fri Jan 19 13:17:57 2018 -0800 CogVM source as per VMMaker.oscog-eem.2320 Spur: Fix a bad bug in SpurPlnningCompactor. unmarkObjectsFromFirstFreeObject, used when the compactor requires more than one pass due to insufficient savedFirstFieldsSpace, expects the corpse of a moved object to be unmarked, but copyAndUnmarkObject:to:bytes:firstField: only unmarked the target. Unmarking the corpse before the copy unmarks both. This fixes a crash with ReleaseBuilder class>>saveAsNewRelease when non-use of cacheDuring: creates lots of files, enough to push the system into the multi-pass regime. > > HTH, > Alistair > > > > > Cheers, > > Alistair > > > > > > > >> On Thu, Jan 18, 2018 at 7:12 PM, ClÃ©ment Bera <bera.clement(a)gmail.com> > >> wrote: > >>> > >>> I would suspect a bug in primitive 105 on byte objects (it was changed > >>> recently in the VM), called by copyFrom: 1 to: readCount. The bug would > >>> likely by due to specific alignment in readCount or something like > that. > >>> (Assuming you're in 32 bits since the 4 bytes are corrupted). > >>> > >>> When I get better I can have a look (I am currently quite sick). > >>> > >>> On Thu, Jan 18, 2018 at 4:51 PM, Thierry Goubier > >>> <thierry.goubier(a)gmail.com> wrote: > >>>> > >>>> Hi Cyril, > >>>> > >>>> try with the last vms available at: > >>>> > >>>> https://bintray.com/opensmalltalk/vm/cog/ > >>>> > >>>> For example, the last Ubuntu 64bits vm is at: > >>>> > >>>> https://bintray.com/opensmalltalk/vm/cog/201801170946#files > >>>> > >>>> Regards, > >>>> > >>>> Thierry > >>>> > >>>> 2018-01-18 16:42 GMT+01:00 Cyrille Delaunay <cy.delaunay(a)gmail.com>: > >>>>> > >>>>> Hi everyone, > >>>>> > >>>>> I just added a new bug entry for an issue we are experimenting since > >>>>> some times: > >>>>> > >>>>> > >>>>> https://pharo.fogbugz.com/f/cases/20982/Random-corrupted- > data-when-copying-from-very-large-byte-array > >>>>> > >>>>> Here is the description: > >>>>> > >>>>> > >>>>> History: > >>>>> > >>>>> This issue has been spotted after experimenting strange behavior with > >>>>> seaside upload. > >>>>> After uploading a big file from a web browser, the modeled file > >>>>> generated within pharo image begins with 4 unexpected bytes. > >>>>> This issue occurs randomly: sometimes the first 4 bytes are right. > >>>>> Sometimes the first 4 bytes are wrong. > >>>>> This issue only occurs with Pharo 6. > >>>>> This issue occurs for all platforms (Mac, Ubuntu, Windows) > >>>>> > >>>>> Steps to reproduce: > >>>>> > >>>>> I have been able to set up a small scenario that highlight the issue. > >>>>> > >>>>> Download Pharo 6.1 on my Mac (Sierra 10.12.6): > >>>>> https://pharo.org/web/download > >>>>> Then, iterate over this process till spotting the issue: > >>>>> > >>>>> => start the pharo image > >>>>> => execute this piece of code in a playground > >>>>> > >>>>> 1: > >>>>> 2: > >>>>> 3: > >>>>> 4: > >>>>> 5: > >>>>> 6: > >>>>> > >>>>> ZnServer startDefaultOn: 1701. > >>>>> ZnServer default maximumEntitySize: 80* 1024 * 1024. > >>>>> '/Users/cdelaunay/myzip.zip' asFileReference writeStreamDo: [ :out | > >>>>> out binary; nextPutAll: #[80 75 3 4 10 0 0 0 0 0 125 83 67 73 0 0 0 > 0 0 > >>>>> 0]. > >>>>> 18202065 timesRepeat: [ out nextPut: 0 ] > >>>>> ]. > >>>>> > >>>>> => Open a web browser page on: http://localhost:1701/form-test-3 > >>>>> => Upload the file zip file previously generated ('myzip.zip'). > >>>>> => If the web page displays: "contents=000000000a00..." (or anything > >>>>> unexpected), THIS IS THE ISSUE ! > >>>>> => If the web page displays: "contents=504b03040a00..", the upload > >>>>> worked fine. You can close the image (without saving) > >>>>> > >>>>> > >>>>> > >>>>> Debugging: > >>>>> > >>>>> > >>>>> > >>>>> Bob Arning has been able to reproduce the issue with my scenario. > >>>>> He dived into the code involved during this process, till reaching > some > >>>>> "basic" methods where he saw the issue occuring. > >>>>> > >>>>> Here are the conclusion till there: > >>>>> => A corruption occurs while reading an input stream with method > ZnUtils > >>>>> class>>readUpToEnd:limit: > >>>>> The first 4 bytes may be altered randomely. > >>>>> => The first 4 bytes are initially correctly written to an > outputStream. > >>>>> But, the first 4 bytes of this outputStream gets altered (corrupted), > >>>>> sometimes when the inner byte array grows OR when performing the > final > >>>>> "outputStream contents" > >>>>> => Here is a piece of code that reproduce the issue (still randomely. > >>>>> stopping an restarting the image may change the behavior) > >>>>> > >>>>> 1: > >>>>> 2: > >>>>> 3: > >>>>> 4: > >>>>> 5: > >>>>> 6: > >>>>> 7: > >>>>> 8: > >>>>> 9: > >>>>> 10: > >>>>> 11: > >>>>> 12: > >>>>> 13: > >>>>> 14: > >>>>> 15: > >>>>> 16: > >>>>> 17: > >>>>> 18: > >>>>> 19: > >>>>> 20: > >>>>> > >>>>> test4"self test4" | species bufferSize buffer totalRead > outputStream > >>>>> answer inputStream ba byte1 | ba := ByteArray new: > 18202085. > >>>>> ba atAllPut: 99. 1 to: 20 do: [ :i | ba at: i put: (#[80 75 > 3 4 10 7 > >>>>> 7 7 7 7 125 83 67 73 7 7 7 7 7 7] at: i) ]. inputStream := ba > readStream. > >>>>> bufferSize := 16384. species := ByteArray. > >>>>> buffer := species new: bufferSize. > >>>>> totalRead := 0. > >>>>> outputStream := nil. > >>>>> [ inputStream atEnd ] whileFalse: [ | readCount | > >>>>> readCount := inputStream readInto: buffer startingAt: 1 > count: > >>>>> bufferSize. > >>>>> totalRead = 0 ifTrue: [ > >>>>> byte1 := buffer first. > >>>>> ]. > >>>>> totalRead := totalRead + readCount. > >>>>> > >>>>> outputStream ifNil: [ > >>>>> inputStream atEnd > >>>>> ifTrue: [ ^ buffer copyFrom: 1 to: readCount ] > >>>>> ifFalse: [ outputStream := (species new: bufferSize) > >>>>> writeStream ] ]. > >>>>> outputStream next: readCount putAll: buffer startingAt: 1. > >>>>> byte1 = outputStream contents first ifFalse: [ self halt ]. > >>>>> ]. > >>>>> answer := outputStream ifNil: [ species new ] ifNotNil: [ > >>>>> outputStream contents ]. > >>>>> byte1 = answer first ifFalse: [ self halt ]. ^answer > >>>>> > >>>>> > >>>>> > >>>>> suspicions > >>>>> > >>>>> This issue appeared with Pharo 6. > >>>>> > >>>>> Some people suggested that it could be a vm issue, and to try my > little > >>>>> scenario with the last available vm. > >>>>> > >>>>> I am not sure where to find the last available vm. > >>>>> > >>>>> I did the test using these elements: > >>>>> > >>>>> https://files.pharo.org/image/60/latest.zip > >>>>> > >>>>> https://files.pharo.org/get-files/70/pharo-mac-latest.zip/ > >>>>> > >>>>> > >>>>> > >>>>> The issue is still present > >>>>> > >>>>> > >>>>> > >>>>> > >>>>> -- > >>>>> Cyrille Delaunay > >>>> > >>>> > >>> > >>> > >>> > >>> -- > >>> ClÃ©ment BÃ©ra > >>> Pharo consortium engineer > >>> https://clementbera.wordpress.com/ > >>> BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq > >> > >> > >> > >> > >> -- > >> ClÃ©ment BÃ©ra > >> Pharo consortium engineer > >> https://clementbera.wordpress.com/ > >> BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq > > -- _,,,^..^,,,_ best, Eliot

Jan. 19, 2018

Re: [Pharo-dev] Random corrupted data when copying from very large byte array

by Alistair Grant

Hi ClÃ©ment, On 19 January 2018 at 17:21, Alistair Grant <akgrant0710(a)gmail.com> wrote: > Hi ClÃ©ment, > > On 19 January 2018 at 17:04, ClÃ©ment Bera <bera.clement(a)gmail.com> wrote: >> Does not seem to be related to prim 105. >> >> I am confused. Has the size of the array an impact at all ? > > Yes, I tried reducing the size of the array by a factor of 10 and > wasn't able to reproduce the problem at all. > > With the full size array it failed over half the time (32 bit). > > I ran the test about 180 times on 64 bit and didn't get a single failure. > >> It seems the >> problem shows since the first copy of 16k elements. >> >> I can't really reproduce the bug - it happened once but I cannot do it >> again. Does the bug happen with the StackVM/PharoS VM you can find here the >> 32 bits versions : http://files.pharo.org/vm/pharoS-spur32/ ? The >> StackVM/PharoS VM is the VM without the JIT, it may be since the bug is >> unreliable that it happens only in jitted code, so trying that out may be >> worth it. > > I'll try and have a look at this over the weekend. This didn't fail once in 55 runs. OS: Ubuntu 16.04 Image: Pharo 6.0 Latest update: #60528 VM: 5.0 #1 Wed Oct 12 15:48:53 CEST 2016 gcc 4.6.3 [Production Spur ITHB VM] StackInterpreter VMMaker.oscog-EstebanLorenzano.1881 uuid: ed616067-a57c-409b-bfb6-dab51f058235 Oct 12 2016 https://github.com/pharo-project/pharo-vm.git Commit: 01a03276a2e2b243cd4a7d3427ba541f835c07d3 Date: 2016-10-12 14:31:09 +0200 By: Esteban Lorenzano <estebanlm(a)gmail.com> Jenkins build #606 Linux pharo-linux 3.2.0-31-generic-pae #50-Ubuntu SMP Fri Sep 7 16:39:45 UTC 2012 i686 i686 i386 GNU/Linux plugin path: /home/alistair/pharo7/Issue20982/bin/ [default: /home/alistair/pharo7/Issue20982/bin/] I then went back and attempted to reproduce the failures in my regular 32 bit image, but only got 1 corruption in 10 runs. I've been working in this image without restarting for most of the day. Quitting out and restarting the image and then running the corruption check resulted in 11 corruptions from 11 runs. Image: Pharo 7.0 Build information: Pharo-7.0+alpha.build.425.sha.eb0a6fb140ac4a42b1f158ed37717e0650f778b4 (32 Bit) VM: 5.0-201801110739 Thursday 11 January 09:30:12 CET 2018 gcc 4.8.5 [Production Spur VM] CoInterpreter VMMaker.oscog-eem.2302 uuid: 55ec8f63-cdbe-4e79-8f22-48fdea585b88 Jan 11 2018 StackToRegisterMappingCogit VMMaker.oscog-eem.2302 uuid: 55ec8f63-cdbe-4e79-8f22-48fdea585b88 Jan 11 2018 VM: 201801110739 alistair@alistair-xps13:snap/pharo-snap/pharo-vm/opensmalltalk-vm $ Date: Wed Jan 10 23:39:30 2018 -0800 $ Plugins: 201801110739 alistair@alistair-xps13:snap/pharo-snap/pharo-vm/opensmalltalk-vm $ Linux b07d7880072c 4.13.0-26-generic #29~16.04.2-Ubuntu SMP Tue Jan 9 22:00:44 UTC 2018 i686 i686 i686 GNU/Linux plugin path: /snap/core/3748/lib/i386-linux-gnu/ [default: /snap/core/3748/lib/i386-linux-gnu/] So, as well as restarting the image before running the test, just wondering if the gcc compiler version could have an impact? HTH, Alistair > Cheers, > Alistair > > > >> On Thu, Jan 18, 2018 at 7:12 PM, ClÃ©ment Bera <bera.clement(a)gmail.com> >> wrote: >>> >>> I would suspect a bug in primitive 105 on byte objects (it was changed >>> recently in the VM), called by copyFrom: 1 to: readCount. The bug would >>> likely by due to specific alignment in readCount or something like that. >>> (Assuming you're in 32 bits since the 4 bytes are corrupted). >>> >>> When I get better I can have a look (I am currently quite sick). >>> >>> On Thu, Jan 18, 2018 at 4:51 PM, Thierry Goubier >>> <thierry.goubier(a)gmail.com> wrote: >>>> >>>> Hi Cyril, >>>> >>>> try with the last vms available at: >>>> >>>> https://bintray.com/opensmalltalk/vm/cog/ >>>> >>>> For example, the last Ubuntu 64bits vm is at: >>>> >>>> https://bintray.com/opensmalltalk/vm/cog/201801170946#files >>>> >>>> Regards, >>>> >>>> Thierry >>>> >>>> 2018-01-18 16:42 GMT+01:00 Cyrille Delaunay <cy.delaunay(a)gmail.com>: >>>>> >>>>> Hi everyone, >>>>> >>>>> I just added a new bug entry for an issue we are experimenting since >>>>> some times: >>>>> >>>>> >>>>> https://pharo.fogbugz.com/f/cases/20982/Random-corrupted-data-when-copying-… >>>>> >>>>> Here is the description: >>>>> >>>>> >>>>> History: >>>>> >>>>> This issue has been spotted after experimenting strange behavior with >>>>> seaside upload. >>>>> After uploading a big file from a web browser, the modeled file >>>>> generated within pharo image begins with 4 unexpected bytes. >>>>> This issue occurs randomly: sometimes the first 4 bytes are right. >>>>> Sometimes the first 4 bytes are wrong. >>>>> This issue only occurs with Pharo 6. >>>>> This issue occurs for all platforms (Mac, Ubuntu, Windows) >>>>> >>>>> Steps to reproduce: >>>>> >>>>> I have been able to set up a small scenario that highlight the issue. >>>>> >>>>> Download Pharo 6.1 on my Mac (Sierra 10.12.6): >>>>> https://pharo.org/web/download >>>>> Then, iterate over this process till spotting the issue: >>>>> >>>>> => start the pharo image >>>>> => execute this piece of code in a playground >>>>> >>>>> 1: >>>>> 2: >>>>> 3: >>>>> 4: >>>>> 5: >>>>> 6: >>>>> >>>>> ZnServer startDefaultOn: 1701. >>>>> ZnServer default maximumEntitySize: 80* 1024 * 1024. >>>>> '/Users/cdelaunay/myzip.zip' asFileReference writeStreamDo: [ :out | >>>>> out binary; nextPutAll: #[80 75 3 4 10 0 0 0 0 0 125 83 67 73 0 0 0 0 0 >>>>> 0]. >>>>> 18202065 timesRepeat: [ out nextPut: 0 ] >>>>> ]. >>>>> >>>>> => Open a web browser page on: http://localhost:1701/form-test-3 >>>>> => Upload the file zip file previously generated ('myzip.zip'). >>>>> => If the web page displays: "contents=000000000a00..." (or anything >>>>> unexpected), THIS IS THE ISSUE ! >>>>> => If the web page displays: "contents=504b03040a00..", the upload >>>>> worked fine. You can close the image (without saving) >>>>> >>>>> >>>>> >>>>> Debugging: >>>>> >>>>> >>>>> >>>>> Bob Arning has been able to reproduce the issue with my scenario. >>>>> He dived into the code involved during this process, till reaching some >>>>> "basic" methods where he saw the issue occuring. >>>>> >>>>> Here are the conclusion till there: >>>>> => A corruption occurs while reading an input stream with method ZnUtils >>>>> class>>readUpToEnd:limit: >>>>> The first 4 bytes may be altered randomely. >>>>> => The first 4 bytes are initially correctly written to an outputStream. >>>>> But, the first 4 bytes of this outputStream gets altered (corrupted), >>>>> sometimes when the inner byte array grows OR when performing the final >>>>> "outputStream contents" >>>>> => Here is a piece of code that reproduce the issue (still randomely. >>>>> stopping an restarting the image may change the behavior) >>>>> >>>>> 1: >>>>> 2: >>>>> 3: >>>>> 4: >>>>> 5: >>>>> 6: >>>>> 7: >>>>> 8: >>>>> 9: >>>>> 10: >>>>> 11: >>>>> 12: >>>>> 13: >>>>> 14: >>>>> 15: >>>>> 16: >>>>> 17: >>>>> 18: >>>>> 19: >>>>> 20: >>>>> >>>>> test4"self test4" | species bufferSize buffer totalRead outputStream >>>>> answer inputStream ba byte1 | ba := ByteArray new: 18202085. >>>>> ba atAllPut: 99. 1 to: 20 do: [ :i | ba at: i put: (#[80 75 3 4 10 7 >>>>> 7 7 7 7 125 83 67 73 7 7 7 7 7 7] at: i) ]. inputStream := ba readStream. >>>>> bufferSize := 16384. species := ByteArray. >>>>> buffer := species new: bufferSize. >>>>> totalRead := 0. >>>>> outputStream := nil. >>>>> [ inputStream atEnd ] whileFalse: [ | readCount | >>>>> readCount := inputStream readInto: buffer startingAt: 1 count: >>>>> bufferSize. >>>>> totalRead = 0 ifTrue: [ >>>>> byte1 := buffer first. >>>>> ]. >>>>> totalRead := totalRead + readCount. >>>>> >>>>> outputStream ifNil: [ >>>>> inputStream atEnd >>>>> ifTrue: [ ^ buffer copyFrom: 1 to: readCount ] >>>>> ifFalse: [ outputStream := (species new: bufferSize) >>>>> writeStream ] ]. >>>>> outputStream next: readCount putAll: buffer startingAt: 1. >>>>> byte1 = outputStream contents first ifFalse: [ self halt ]. >>>>> ]. >>>>> answer := outputStream ifNil: [ species new ] ifNotNil: [ >>>>> outputStream contents ]. >>>>> byte1 = answer first ifFalse: [ self halt ]. ^answer >>>>> >>>>> >>>>> >>>>> suspicions >>>>> >>>>> This issue appeared with Pharo 6. >>>>> >>>>> Some people suggested that it could be a vm issue, and to try my little >>>>> scenario with the last available vm. >>>>> >>>>> I am not sure where to find the last available vm. >>>>> >>>>> I did the test using these elements: >>>>> >>>>> https://files.pharo.org/image/60/latest.zip >>>>> >>>>> https://files.pharo.org/get-files/70/pharo-mac-latest.zip/ >>>>> >>>>> >>>>> >>>>> The issue is still present >>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> Cyrille Delaunay >>>> >>>> >>> >>> >>> >>> -- >>> ClÃ©ment BÃ©ra >>> Pharo consortium engineer >>> https://clementbera.wordpress.com/ >>> BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq >> >> >> >> >> -- >> ClÃ©ment BÃ©ra >> Pharo consortium engineer >> https://clementbera.wordpress.com/ >> BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq

Jan. 19, 2018

Re: [Pharo-dev] [ANN] Fari â a shell tool to fetch & prepare your fresh images

by Damien Pollet

Le ven. 19 janv. 2018 Ã 20:59, HernÃ¡n Morales Durand < hernan.morales(a)gmail.com> a Ã©crit : > $ ./fari.sh > ls: cannot access './*.load.st': No such file or directory > > You may need to add a shell option in the script to include hidden file > names in expansions > > shopt -s dotglob > Ah no, this is about nullglob. I used it somewhere else but missed that spot. Will have a look later this evening. $./fari.sh: line 176: ./pharo-ui: No such file or directory > Missing feature ð youâll have to get the VM with zeroconf for now. I don't know but only sources file was downloaded. > Expected: the script stops as soon as a problem os encountered. 2018-01-19 15:02 GMT-03:00 Damien Pollet <damien.pollet(a)gmail.com>: > >> Here it is, tell me what you think: https://github.com/cdlm/fari.sh >> >> We discussed it with Christophe and of course there is some overlap with >> PharoLauncher, and in the future, it might get replaced by a command-line >> version of PharoLauncher based on Clap. >> >> But for the moment it works and fits my twisted mind better ð >> >> -- >> Damien Pollet >> type less, do more [ | ] http://people.untyped.org/damien.pollet >> > > -- Damien Pollet type less, do more [ | ] http://people.untyped.org/damien.pollet

Jan. 19, 2018

Re: [Pharo-dev] [ANN] Fari â a shell tool to fetch & prepare your fresh images

by HernÃ¡n Morales Durand

Hi Damien, Thanks for sharing this. This is what I did using mingw: $ wget https://github.com/cdlm/fari.sh/raw/master/fari.sh $ cat > .load.st Metacello new baseline: 'CCBC'; repository: 'gitlocal://./src'; load. $ ./fari.sh ls: cannot access './*.load.st': No such file or directory You may need to add a shell option in the script to include hidden file names in expansions shopt -s dotglob then $./fari.sh: line 176: ./pharo-ui: No such file or directory $ ll total 31861 drwxr-xr-x 1 mvs 197609 0 Jan 19 16:30 BaselineOfCCBC.package drwxr-xr-x 1 mvs 197609 0 Jan 19 16:44 CCBC.package drwxr-xr-x 1 mvs 197609 0 Jan 19 16:30 CCBC-Core.package -rw-r--r-- 1 mvs 197609 76951 Jan 19 16:30 'Collections - Strings.png' -rwxr-xr-x 1 mvs 197609 7399 Jan 19 16:53 fari.sh -rw-r--r-- 1 mvs 197609 44415 Jan 19 16:30 'Iceberg Class Critiques (Pharo 6.1).png' -rw-r--r-- 1 mvs 197609 1100 Jan 19 16:30 LICENSE -rw-r--r-- 1 mvs 197609 32371497 Jan 19 16:55 Pharo7.0-32bit-e359bc0.sources -rw-r--r-- 1 mvs 197609 549 Jan 19 16:30 README.md -rw-r--r-- 1 mvs 197609 101557 Jan 19 16:30 'Spec Method Critiques (Pharo 6.1).png' I don't know but only sources file was downloaded. Cheers, HernÃ¡n 2018-01-19 15:02 GMT-03:00 Damien Pollet <damien.pollet(a)gmail.com>: > Here it is, tell me what you think: https://github.com/cdlm/fari.sh > > We discussed it with Christophe and of course there is some overlap with > PharoLauncher, and in the future, it might get replaced by a command-line > version of PharoLauncher based on Clap. > > But for the moment it works and fits my twisted mind better ð > > -- > Damien Pollet > type less, do more [ | ] http://people.untyped.org/damien.pollet >

Jan. 19, 2018

[ANN] Fari â a shell tool to fetch & prepare your fresh images

by Damien Pollet

Here it is, tell me what you think: https://github.com/cdlm/fari.sh We discussed it with Christophe and of course there is some overlap with PharoLauncher, and in the future, it might get replaced by a command-line version of PharoLauncher based on Clap. But for the moment it works and fits my twisted mind better ð -- Damien Pollet type less, do more [ | ] http://people.untyped.org/damien.pollet

Jan. 19, 2018

Re: [Pharo-dev] Random corrupted data when copying from very large byte array

by Alistair Grant

Hi ClÃ©ment, On 19 January 2018 at 17:04, ClÃ©ment Bera <bera.clement(a)gmail.com> wrote: > Does not seem to be related to prim 105. > > I am confused. Has the size of the array an impact at all ? Yes, I tried reducing the size of the array by a factor of 10 and wasn't able to reproduce the problem at all. With the full size array it failed over half the time (32 bit). I ran the test about 180 times on 64 bit and didn't get a single failure. > It seems the > problem shows since the first copy of 16k elements. > > I can't really reproduce the bug - it happened once but I cannot do it > again. Does the bug happen with the StackVM/PharoS VM you can find here the > 32 bits versions : http://files.pharo.org/vm/pharoS-spur32/ ? The > StackVM/PharoS VM is the VM without the JIT, it may be since the bug is > unreliable that it happens only in jitted code, so trying that out may be > worth it. I'll try and have a look at this over the weekend. Cheers, Alistair > On Thu, Jan 18, 2018 at 7:12 PM, ClÃ©ment Bera <bera.clement(a)gmail.com> > wrote: >> >> I would suspect a bug in primitive 105 on byte objects (it was changed >> recently in the VM), called by copyFrom: 1 to: readCount. The bug would >> likely by due to specific alignment in readCount or something like that. >> (Assuming you're in 32 bits since the 4 bytes are corrupted). >> >> When I get better I can have a look (I am currently quite sick). >> >> On Thu, Jan 18, 2018 at 4:51 PM, Thierry Goubier >> <thierry.goubier(a)gmail.com> wrote: >>> >>> Hi Cyril, >>> >>> try with the last vms available at: >>> >>> https://bintray.com/opensmalltalk/vm/cog/ >>> >>> For example, the last Ubuntu 64bits vm is at: >>> >>> https://bintray.com/opensmalltalk/vm/cog/201801170946#files >>> >>> Regards, >>> >>> Thierry >>> >>> 2018-01-18 16:42 GMT+01:00 Cyrille Delaunay <cy.delaunay(a)gmail.com>: >>>> >>>> Hi everyone, >>>> >>>> I just added a new bug entry for an issue we are experimenting since >>>> some times: >>>> >>>> >>>> https://pharo.fogbugz.com/f/cases/20982/Random-corrupted-data-when-copying-… >>>> >>>> Here is the description: >>>> >>>> >>>> History: >>>> >>>> This issue has been spotted after experimenting strange behavior with >>>> seaside upload. >>>> After uploading a big file from a web browser, the modeled file >>>> generated within pharo image begins with 4 unexpected bytes. >>>> This issue occurs randomly: sometimes the first 4 bytes are right. >>>> Sometimes the first 4 bytes are wrong. >>>> This issue only occurs with Pharo 6. >>>> This issue occurs for all platforms (Mac, Ubuntu, Windows) >>>> >>>> Steps to reproduce: >>>> >>>> I have been able to set up a small scenario that highlight the issue. >>>> >>>> Download Pharo 6.1 on my Mac (Sierra 10.12.6): >>>> https://pharo.org/web/download >>>> Then, iterate over this process till spotting the issue: >>>> >>>> => start the pharo image >>>> => execute this piece of code in a playground >>>> >>>> 1: >>>> 2: >>>> 3: >>>> 4: >>>> 5: >>>> 6: >>>> >>>> ZnServer startDefaultOn: 1701. >>>> ZnServer default maximumEntitySize: 80* 1024 * 1024. >>>> '/Users/cdelaunay/myzip.zip' asFileReference writeStreamDo: [ :out | >>>> out binary; nextPutAll: #[80 75 3 4 10 0 0 0 0 0 125 83 67 73 0 0 0 0 0 >>>> 0]. >>>> 18202065 timesRepeat: [ out nextPut: 0 ] >>>> ]. >>>> >>>> => Open a web browser page on: http://localhost:1701/form-test-3 >>>> => Upload the file zip file previously generated ('myzip.zip'). >>>> => If the web page displays: "contents=000000000a00..." (or anything >>>> unexpected), THIS IS THE ISSUE ! >>>> => If the web page displays: "contents=504b03040a00..", the upload >>>> worked fine. You can close the image (without saving) >>>> >>>> >>>> >>>> Debugging: >>>> >>>> >>>> >>>> Bob Arning has been able to reproduce the issue with my scenario. >>>> He dived into the code involved during this process, till reaching some >>>> "basic" methods where he saw the issue occuring. >>>> >>>> Here are the conclusion till there: >>>> => A corruption occurs while reading an input stream with method ZnUtils >>>> class>>readUpToEnd:limit: >>>> The first 4 bytes may be altered randomely. >>>> => The first 4 bytes are initially correctly written to an outputStream. >>>> But, the first 4 bytes of this outputStream gets altered (corrupted), >>>> sometimes when the inner byte array grows OR when performing the final >>>> "outputStream contents" >>>> => Here is a piece of code that reproduce the issue (still randomely. >>>> stopping an restarting the image may change the behavior) >>>> >>>> 1: >>>> 2: >>>> 3: >>>> 4: >>>> 5: >>>> 6: >>>> 7: >>>> 8: >>>> 9: >>>> 10: >>>> 11: >>>> 12: >>>> 13: >>>> 14: >>>> 15: >>>> 16: >>>> 17: >>>> 18: >>>> 19: >>>> 20: >>>> >>>> test4"self test4" | species bufferSize buffer totalRead outputStream >>>> answer inputStream ba byte1 | ba := ByteArray new: 18202085. >>>> ba atAllPut: 99. 1 to: 20 do: [ :i | ba at: i put: (#[80 75 3 4 10 7 >>>> 7 7 7 7 125 83 67 73 7 7 7 7 7 7] at: i) ]. inputStream := ba readStream. >>>> bufferSize := 16384. species := ByteArray. >>>> buffer := species new: bufferSize. >>>> totalRead := 0. >>>> outputStream := nil. >>>> [ inputStream atEnd ] whileFalse: [ | readCount | >>>> readCount := inputStream readInto: buffer startingAt: 1 count: >>>> bufferSize. >>>> totalRead = 0 ifTrue: [ >>>> byte1 := buffer first. >>>> ]. >>>> totalRead := totalRead + readCount. >>>> >>>> outputStream ifNil: [ >>>> inputStream atEnd >>>> ifTrue: [ ^ buffer copyFrom: 1 to: readCount ] >>>> ifFalse: [ outputStream := (species new: bufferSize) >>>> writeStream ] ]. >>>> outputStream next: readCount putAll: buffer startingAt: 1. >>>> byte1 = outputStream contents first ifFalse: [ self halt ]. >>>> ]. >>>> answer := outputStream ifNil: [ species new ] ifNotNil: [ >>>> outputStream contents ]. >>>> byte1 = answer first ifFalse: [ self halt ]. ^answer >>>> >>>> >>>> >>>> suspicions >>>> >>>> This issue appeared with Pharo 6. >>>> >>>> Some people suggested that it could be a vm issue, and to try my little >>>> scenario with the last available vm. >>>> >>>> I am not sure where to find the last available vm. >>>> >>>> I did the test using these elements: >>>> >>>> https://files.pharo.org/image/60/latest.zip >>>> >>>> https://files.pharo.org/get-files/70/pharo-mac-latest.zip/ >>>> >>>> >>>> >>>> The issue is still present >>>> >>>> >>>> >>>> >>>> -- >>>> Cyrille Delaunay >>> >>> >> >> >> >> -- >> ClÃ©ment BÃ©ra >> Pharo consortium engineer >> https://clementbera.wordpress.com/ >> BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq > > > > > -- > ClÃ©ment BÃ©ra > Pharo consortium engineer > https://clementbera.wordpress.com/ > BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq

Jan. 19, 2018

Re: [Pharo-dev] Random corrupted data when copying from very large byte array

by ClÃ©ment Bera

Does not seem to be related to prim 105. I am confused. Has the size of the array an impact at all ? It seems the problem shows since the first copy of 16k elements. I can't really reproduce the bug - it happened once but I cannot do it again. Does the bug happen with the StackVM/PharoS VM you can find here the 32 bits versions : http://files.pharo.org/vm/pharoS-spur32/ ? The StackVM/PharoS VM is the VM without the JIT, it may be since the bug is unreliable that it happens only in jitted code, so trying that out may be worth it. On Thu, Jan 18, 2018 at 7:12 PM, ClÃ©ment Bera <bera.clement(a)gmail.com> wrote: > I would suspect a bug in primitive 105 on byte objects (it was changed > recently in the VM), called by copyFrom: 1 to: readCount. The bug would > likely by due to specific alignment in readCount or something like that. > (Assuming you're in 32 bits since the 4 bytes are corrupted). > > When I get better I can have a look (I am currently quite sick). > > On Thu, Jan 18, 2018 at 4:51 PM, Thierry Goubier < > thierry.goubier(a)gmail.com> wrote: > >> Hi Cyril, >> >> try with the last vms available at: >> >> https://bintray.com/opensmalltalk/vm/cog/ >> >> For example, the last Ubuntu 64bits vm is at: >> >> https://bintray.com/opensmalltalk/vm/cog/201801170946#files >> >> Regards, >> >> Thierry >> >> 2018-01-18 16:42 GMT+01:00 Cyrille Delaunay <cy.delaunay(a)gmail.com>: >> >>> Hi everyone, >>> >>> I just added a new bug entry for an issue we are experimenting since >>> some times: >>> >>> https://pharo.fogbugz.com/f/cases/20982/Random-corrupted-dat >>> a-when-copying-from-very-large-byte-array >>> >>> Here is the description: >>> >>> >>> *History:* >>> >>> This issue has been spotted after experimenting strange behavior with >>> seaside upload. >>> After uploading a big file from a web browser, the modeled file >>> generated within pharo image begins with 4 unexpected bytes. >>> This issue occurs randomly: sometimes the first 4 bytes are right. >>> Sometimes the first 4 bytes are wrong. >>> This issue only occurs with Pharo 6. >>> This issue occurs for all platforms (Mac, Ubuntu, Windows) >>> >>> *Steps to reproduce:* >>> >>> I have been able to set up a small scenario that highlight the issue. >>> >>> Download Pharo 6.1 on my Mac (Sierra 10.12.6): https://pharo.org/we >>> b/download >>> Then, iterate over this process till spotting the issue: >>> >>> => start the pharo image >>> => execute this piece of code in a playground >>> >>> 1: >>> 2: >>> 3: >>> 4: >>> 5: >>> 6: >>> >>> ZnServer startDefaultOn: 1701.ZnServer default maximumEntitySize: 80* 1024 * 1024.'/Users/cdelaunay/myzip.zip' asFileReference writeStreamDo: [ :out |out binary; nextPutAll: #[80 75 3 4 10 0 0 0 0 0 125 83 67 73 0 0 0 0 0 0].18202065 timesRepeat: [ out nextPut: 0 ]]. >>> >>> => Open a web browser page on: http://localhost:1701/form-test-3 >>> => Upload the file zip file previously generated ('myzip.zip'). >>> => If the web page displays: "contents=000000000a00 >>> <https://pharo.kilnhg.com/Search?search=000000000a00>..." (or anything >>> unexpected), THIS IS THE ISSUE ! >>> => If the web page displays: "contents=504b03040a00 >>> <https://pharo.kilnhg.com/Search?search=504b03040a00>..", the upload >>> worked fine. You can close the image (without saving) >>> >>> >>> >>> *Debugging:* >>> >>> >>> >>> Bob Arning has been able to reproduce the issue with my scenario. >>> He dived into the code involved during this process, till reaching some >>> "basic" methods where he saw the issue occuring. >>> >>> Here are the conclusion till there: >>> => A corruption occurs while reading an input stream with method ZnUtils >>> class>>readUpToEnd:limit: >>> The first 4 bytes may be altered randomely. >>> => The first 4 bytes are initially correctly written to an >>> outputStream. >>> But, the first 4 bytes of this outputStream gets altered (corrupted), >>> sometimes when the inner byte array grows OR when performing the >>> final "outputStream contents" >>> => Here is a piece of code that reproduce the issue (still randomely. >>> stopping an restarting the image may change the behavior) >>> >>> 1: >>> 2: >>> 3: >>> 4: >>> 5: >>> 6: >>> 7: >>> 8: >>> 9: >>> 10: >>> 11: >>> 12: >>> 13: >>> 14: >>> 15: >>> 16: >>> 17: >>> 18: >>> 19: >>> 20: >>> >>> test4"self test4" | species bufferSize buffer totalRead outputStream answer inputStream ba byte1 | ba := ByteArray new: 18202085. ba atAllPut: 99. 1 to: 20 do: [ :i | ba at: i put: (#[80 75 3 4 10 7 7 7 7 7 125 83 67 73 7 7 7 7 7 7] at: i) ]. inputStream := ba readStream. bufferSize := 16384. species := ByteArray. >>> buffer := species new: bufferSize. >>> totalRead := 0. >>> outputStream := nil. >>> [ inputStream atEnd ] whileFalse: [ | readCount | >>> readCount := inputStream readInto: buffer startingAt: 1 count: bufferSize. >>> totalRead = 0 ifTrue: [ >>> byte1 := buffer first. >>> ]. >>> totalRead := totalRead + readCount. >>> >>> outputStream ifNil: [ >>> inputStream atEnd >>> ifTrue: [ ^ buffer copyFrom: 1 to: readCount ] >>> ifFalse: [ outputStream := (species new: bufferSize) writeStream ] ]. >>> outputStream next: readCount putAll: buffer startingAt: 1. >>> byte1 = outputStream contents first ifFalse: [ self halt ]. >>> ]. >>> answer := outputStream ifNil: [ species new ] ifNotNil: [ outputStream contents ]. >>> byte1 = answer first ifFalse: [ self halt ]. ^answer >>> >>> >>> >>> *suspicions* >>> >>> This issue appeared with Pharo 6. >>> >>> Some people suggested that it could be a vm issue, and to try my little >>> scenario with the last available vm. >>> >>> I am not sure where to find the last available vm. >>> >>> I did the test using these elements: >>> https://files.pharo.org/image/60/latest.zip >>> >>> https://files.pharo.org/get-files/70/pharo-mac-latest.zip/ >>> <https://files.pharo.org/get-files/70/> >>> >>> >>> >>> The issue is still present >>> >>> >>> >>> -- >>> Cyrille Delaunay >>> >> >> > > > -- > ClÃ©ment BÃ©ra > Pharo consortium engineer > https://clementbera.wordpress.com/ > BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq > -- ClÃ©ment BÃ©ra Pharo consortium engineer https://clementbera.wordpress.com/ BÃ¢timent B 40, avenue Halley 59650 Villeneuve d'Ascq

Jan. 19, 2018

Re: [Pharo-dev] [IMPORTANT] PHARO CONTRIBUTION PROCESS UPDATED! PLEASE READ!

by Stephan Eggermont

Luke Gorrie <luke(a)snabb.co> wrote: > For me it's another story. I'm on an unsupported platform (NixOS Linux), > I'm building the VM from random git commits because the source releases are > all antiquated, Iceberg segfaults the moment I start it, and > epica+monticello+metacello+iceberg+fogbugz+jenkins feels like a series of > obstacles between me and maintaining my application. Isn't nix supported by travis? I suppose that means that you should be able to automatically build a variant of opensmalltalk I have absolutely no idea how much work that would be though. Stephan

Jan. 19, 2018