strange errors in jenkins jobs
Hi, Since yesterday, all Moose-related jobs are failing. These are built on top of Pharo 5.0 stable, downloaded like this: wget --quiet -O - http://get.pharo.org/50+vm | bash We did not change anything in the projects that fail, and locally on Mac, I can load any of the code without problems. The jobs are running Linux. So, I am wondering if anything else has changed in Pharo 5.0. In particular, the error is quite strange. It looks like this: [31mError: MessageNotUnderstood: receiver of "finish" is nil [0mUndefinedObject(Object)>>error: [ :err :rcvr | | errCtx errMorph | errCtx := thisContext. [ errCtx := errCtx sender. "Search the sender chain to find the morph causing the problem" [ errCtx notNil and: [ errCtx receiver isMorph not ] ] whileTrue: [ errCtx := errCtx sender ]. "If we're at the root of the context chain then we have a fatal drawing problem" errCtx ifNil: [ ^ self handleFatalDrawingError: err ]. errMorph := errCtx receiver. ... See more at: https://ci.inria.fr/moose/job/fame/1266/console Any idea of what might be wrong? Cheers, Doru -- www.tudorgirba.com www.feenk.com "We cannot reach the flow of things unless we let go."
2016-08-07 22:36 GMT+02:00 Tudor Girba <tudor@tudorgirba.com>:
Hi,
Since yesterday, all Moose-related jobs are failing. These are built on top of Pharo 5.0 stable, downloaded like this: wget --quiet -O - http://get.pharo.org/50+vm | bash
We did not change anything in the projects that fail, and locally on Mac, I can load any of the code without problems. The jobs are running Linux.
So, I am wondering if anything else has changed in Pharo 5.0.
In particular, the error is quite strange. It looks like this: [31mError: MessageNotUnderstood: receiver of "finish" is nil [0mUndefinedObject(Object)>>error: [ :err :rcvr | | errCtx errMorph | errCtx := thisContext. [ errCtx := errCtx sender. "Search the sender chain to find the morph causing the problem" [ errCtx notNil and: [ errCtx receiver isMorph not ] ] whileTrue: [ errCtx := errCtx sender ]. "If we're at the root of the context chain then we have a fatal drawing problem" errCtx ifNil: [ ^ self handleFatalDrawingError: err ]. errMorph := errCtx receiver. ...
I think this is the same reported here: http://forum.world.st/newly-introduced-problem-in-Pharo5-0-running-headless-... 18706 <https://pharo.fogbugz.com/f/cases/18706/Build-of-Pharo5-is-broken> Build of Pharo5 is broken
See more at: https://ci.inria.fr/moose/job/fame/1266/console
Any idea of what might be wrong?
Cheers, Doru
-- www.tudorgirba.com www.feenk.com
"We cannot reach the flow of things unless we let go."
On Aug 7, 2016, at 11:10 PM, Nicolai Hess <nicolaihess@gmail.com> wrote:
2016-08-07 22:36 GMT+02:00 Tudor Girba <tudor@tudorgirba.com>: Hi,
Since yesterday, all Moose-related jobs are failing. These are built on top of Pharo 5.0 stable, downloaded like this: wget --quiet -O - http://get.pharo.org/50+vm | bash
We did not change anything in the projects that fail, and locally on Mac, I can load any of the code without problems. The jobs are running Linux.
So, I am wondering if anything else has changed in Pharo 5.0.
In particular, the error is quite strange. It looks like this: [31mError: MessageNotUnderstood: receiver of "finish" is nil [0mUndefinedObject(Object)>>error: [ :err :rcvr | | errCtx errMorph | errCtx := thisContext. [ errCtx := errCtx sender. "Search the sender chain to find the morph causing the problem" [ errCtx notNil and: [ errCtx receiver isMorph not ] ] whileTrue: [ errCtx := errCtx sender ]. "If we're at the root of the context chain then we have a fatal drawing problem" errCtx ifNil: [ ^ self handleFatalDrawingError: err ]. errMorph := errCtx receiver. ...
I think this is the same reported here: http://forum.world.st/newly-introduced-problem-in-Pharo5-0-running-headless-... 18706 Build of Pharo5 is broken
Hmm, indeed it looks quite the same, but in our case, the problem appeared only today while the report of Dale is older. Yesterday, our jobs worked well. What could be at play to explain that? Doru
See more at: https://ci.inria.fr/moose/job/fame/1266/console
Any idea of what might be wrong?
Cheers, Doru
-- www.tudorgirba.com www.feenk.com
"We cannot reach the flow of things unless we let go."
_______________________________________________ Moose-dev mailing list Moose-dev@list.inf.unibe.ch https://www.list.inf.unibe.ch/listinfo/moose-dev
-- www.tudorgirba.com www.feenk.com "Value is always contextual."
Le 07/08/2016 à 23:12, Tudor Girba a écrit :
Hmm, indeed it looks quite the same, but in our case, the problem appeared only today while the report of Dale is older. Yesterday, our jobs worked well. What could be at play to explain that?
Hi, The 29/07 the FAMIX build failed for this reason. And it did not happened only one time. So this is not the first time.
Doru
-- www.tudorgirba.com www.feenk.com
"Value is always contextual."
-- Cyril Ferlicot http://www.synectique.eu 165 Avenue Bretagne Lille 59000 France
Yes this does look like the problem I have been seeing randomly for the last several weeks[1] ...The #finish MNU has been intermixed with segFaults ... The #finish looks like a random ui process that wakes up long after the worldstate has been replaced with something new ... depending upon where this zombie process was when when the image is saved, you can get a segfault if the #finish doesn't get you ... or you can be lucky and the processes dies a natural death. Dale [1] https://pharo.fogbugz.com/f/cases/18706#BugEvent.171768 On 8/7/16 2:12 PM, Tudor Girba wrote:
On Aug 7, 2016, at 11:10 PM, Nicolai Hess <nicolaihess@gmail.com> wrote:
2016-08-07 22:36 GMT+02:00 Tudor Girba <tudor@tudorgirba.com>: Hi,
Since yesterday, all Moose-related jobs are failing. These are built on top of Pharo 5.0 stable, downloaded like this: wget --quiet -O - http://get.pharo.org/50+vm | bash
We did not change anything in the projects that fail, and locally on Mac, I can load any of the code without problems. The jobs are running Linux.
So, I am wondering if anything else has changed in Pharo 5.0.
In particular, the error is quite strange. It looks like this: [31mError: MessageNotUnderstood: receiver of "finish" is nil [0mUndefinedObject(Object)>>error: [ :err :rcvr | | errCtx errMorph | errCtx := thisContext. [ errCtx := errCtx sender. "Search the sender chain to find the morph causing the problem" [ errCtx notNil and: [ errCtx receiver isMorph not ] ] whileTrue: [ errCtx := errCtx sender ]. "If we're at the root of the context chain then we have a fatal drawing problem" errCtx ifNil: [ ^ self handleFatalDrawingError: err ]. errMorph := errCtx receiver. ...
I think this is the same reported here: http://forum.world.st/newly-introduced-problem-in-Pharo5-0-running-headless-... 18706 Build of Pharo5 is broken
Hmm, indeed it looks quite the same, but in our case, the problem appeared only today while the report of Dale is older. Yesterday, our jobs worked well. What could be at play to explain that?
Doru
See more at: https://ci.inria.fr/moose/job/fame/1266/console
Any idea of what might be wrong?
Cheers, Doru
-- www.tudorgirba.com www.feenk.com
"We cannot reach the flow of things unless we let go."
_______________________________________________ Moose-dev mailing list Moose-dev@list.inf.unibe.ch https://www.list.inf.unibe.ch/listinfo/moose-dev -- www.tudorgirba.com www.feenk.com
"Value is always contextual."
participants (4)
-
Cyril Ferlicot D. -
Dale Henrichs -
Nicolai Hess -
Tudor Girba