Dear all,
I have just pushed a first beta release of 4.5.0. The package is available here:
http://www.dynare.org/beta-release
Things that remain to be done:
+ Update the NEWS file (I am almost done with the first draft, I should push it on Monday).
+ Update the reference manual.
+ Open a new page for the bugs in 4.5.x.
+ Update the microsoft code signature.
+ Check that the binaries are not detected as virus by the most common anti-virus (we had this issue with release of 4.4.0).
+ Fix the Octave bugs (those which can be fixed quickly).
I want to relase 4.5.0 for the summer school (12th of June). So it is likely that important bugs, related to block/bytecode, will remain in this release (Ferhat is already working on this). I think we will have to release 4.5.1 before the Summer holidays (end of July for me).
For the 4.5.x bug page (and also the New features page, 4.6-unstable) I propose to do it on Github (Wiki) instead of DynareWiki, the Markdown syntax (Github) is more elegant than moinmoin, and also this would greatly reduce the cost of updating the NEWS file.
Best, Stéphane.
Thanks Stéphane.
This is great.
All the best,
Michel
Stéphane Adjemian writes:
Dear all,
I have just pushed a first beta release of 4.5.0. The package is available here:
http://www.dynare.org/beta-release
Things that remain to be done:
- Update the NEWS file (I am almost done with the first draft, I
should push it on Monday).
Update the reference manual.
Open a new page for the bugs in 4.5.x.
Update the microsoft code signature.
Check that the binaries are not detected as virus by the most common
anti-virus (we had this issue with release of 4.4.0).
- Fix the Octave bugs (those which can be fixed quickly).
I want to relase 4.5.0 for the summer school (12th of June). So it is likely that important bugs, related to block/bytecode, will remain in this release (Ferhat is already working on this). I think we will have to release 4.5.1 before the Summer holidays (end of July for me).
For the 4.5.x bug page (and also the New features page, 4.6-unstable) I propose to do it on Github (Wiki) instead of DynareWiki, the Markdown syntax (Github) is more elegant than moinmoin, and also this would greatly reduce the cost of updating the NEWS file.
Best, Stéphane.
Hi Stéphane, what needs to be done for the manual and for Octave? Best, Johannes
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Stéphane Adjemian Gesendet: Samstag, 27. Mai 2017 16:12 An: dev@dynare.org Betreff: [DynareDev] Beta release of 4.5.0
Dear all,
I have just pushed a first beta release of 4.5.0. The package is available here:
http://www.dynare.org/beta-release
Things that remain to be done:
+ Update the NEWS file (I am almost done with the first draft, I should push it on Monday).
+ Update the reference manual.
+ Open a new page for the bugs in 4.5.x.
+ Update the microsoft code signature.
+ Check that the binaries are not detected as virus by the most common anti-virus (we had this issue with release of 4.4.0).
+ Fix the Octave bugs (those which can be fixed quickly).
I want to relase 4.5.0 for the summer school (12th of June). So it is likely that important bugs, related to block/bytecode, will remain in this release (Ferhat is already working on this). I think we will have to release 4.5.1 before the Summer holidays (end of July for me).
For the 4.5.x bug page (and also the New features page, 4.6-unstable) I propose to do it on Github (Wiki) instead of DynareWiki, the Markdown syntax (Github) is more elegant than moinmoin, and also this would greatly reduce the cost of updating the NEWS file.
Best, Stéphane.
-- Stéphane Adjemian Dynare Team _______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Hi Johannes,
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
@all We have now three pages on Dynare's Github/wiki:
- KnownBugs, for the non fixed bugs in the current stable. - FixedBugs, for the fixed bugs, to be published in the next release of the stable or the next major release, - NewFeatures, for documenting the new features to be published in the the next major release.
The corresponding pages on Dynare's MoinMoin wiki are frozen. In the FixedBug page, which I copy/paste from the updated NEWS file, I lost the external links to the Github issues (which are still available in the MoinMoin wiki). This is annoying and has to be fixed, but will take some time... Which I don't have right now.
You can clone the wiki pages locally:
git clone git@github.com:DynareTeam/dynare.wiki.git
modify the markdown files and push the patches on Github. There is a package called Gollum:
https://github.com/gollum/gollum
that will ease the development of the wiki pages, by allowing you to look at the results of your patches locally before committing and pushing on Gthub.
Best, Stéphane.
2017-05-27 20:11 GMT+02:00 Johannes Pfeifer jpfeifer@gmx.de:
Hi Stéphane, what needs to be done for the manual and for Octave? Best, Johannes
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Stéphane Adjemian Gesendet: Samstag, 27. Mai 2017 16:12 An: dev@dynare.org Betreff: [DynareDev] Beta release of 4.5.0
Dear all,
I have just pushed a first beta release of 4.5.0. The package is available here:
http://www.dynare.org/beta-release
Things that remain to be done:
- Update the NEWS file (I am almost done with the first draft, I should
push it on Monday).
Update the reference manual.
Open a new page for the bugs in 4.5.x.
Update the microsoft code signature.
Check that the binaries are not detected as virus by the most common
anti-virus (we had this issue with release of 4.4.0).
- Fix the Octave bugs (those which can be fixed quickly).
I want to relase 4.5.0 for the summer school (12th of June). So it is likely that important bugs, related to block/bytecode, will remain in this release (Ferhat is already working on this). I think we will have to release 4.5.1 before the Summer holidays (end of July for me).
For the 4.5.x bug page (and also the New features page, 4.6-unstable) I propose to do it on Github (Wiki) instead of DynareWiki, the Markdown syntax (Github) is more elegant than moinmoin, and also this would greatly reduce the cost of updating the NEWS file.
Best, Stéphane.
-- Stéphane Adjemian Dynare Team _______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Hi,
I added the links that got lost to the git-wiki. Are these links supposed to be added to the news-file as well?
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Stéphane Adjemian Gesendet: Montag, 29. Mai 2017 23:46 An: List for Dynare developers Betreff: Re: [DynareDev] Beta release of 4.5.0
Hi Johannes,
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
@all We have now three pages on Dynare's Github/wiki:
- KnownBugs, for the non fixed bugs in the current stable.
- FixedBugs, for the fixed bugs, to be published in the next release of the stable or the next major release,
- NewFeatures, for documenting the new features to be published in the the next major release.
The corresponding pages on Dynare's MoinMoin wiki are frozen. In the FixedBug page, which I copy/paste from the updated NEWS file, I lost the external links to the Github issues (which are still available in the MoinMoin wiki). This is annoying and has to be fixed, but will take some time... Which I don't have right now.
You can clone the wiki pages locally:
git clone git@github.com:DynareTeam/dynare.wiki.git
modify the markdown files and push the patches on Github. There is a package called Gollum:
https://github.com/gollum/gollum
that will ease the development of the wiki pages, by allowing you to look at the results of your patches locally before committing and pushing on Gthub.
Best,
Stéphane.
2017-05-27 20:11 GMT+02:00 Johannes Pfeifer jpfeifer@gmx.de:
Hi Stéphane, what needs to be done for the manual and for Octave? Best, Johannes
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Stéphane Adjemian Gesendet: Samstag, 27. Mai 2017 16:12 An: dev@dynare.org Betreff: [DynareDev] Beta release of 4.5.0
Dear all,
I have just pushed a first beta release of 4.5.0. The package is available here:
http://www.dynare.org/beta-release
Things that remain to be done:
+ Update the NEWS file (I am almost done with the first draft, I should push it on Monday).
+ Update the reference manual.
+ Open a new page for the bugs in 4.5.x.
+ Update the microsoft code signature.
+ Check that the binaries are not detected as virus by the most common anti-virus (we had this issue with release of 4.4.0).
+ Fix the Octave bugs (those which can be fixed quickly).
I want to relase 4.5.0 for the summer school (12th of June). So it is likely that important bugs, related to block/bytecode, will remain in this release (Ferhat is already working on this). I think we will have to release 4.5.1 before the Summer holidays (end of July for me).
For the 4.5.x bug page (and also the New features page, 4.6-unstable) I propose to do it on Github (Wiki) instead of DynareWiki, the Markdown syntax (Github) is more elegant than moinmoin, and also this would greatly reduce the cost of updating the NEWS file.
Best, Stéphane.
-- Stéphane Adjemian Dynare Team
_______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
_______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Thanks Johannes. No the links are only for the wiki page. I don't think we need them in the NEWS file. Best, Stéphane.
2017-06-06 15:59 GMT+02:00 Johannes Pfeifer jpfeifer@gmx.de:
Hi,
I added the links that got lost to the git-wiki. Are these links supposed to be added to the news-file as well?
*Von:* Dev [mailto:dev-bounces@dynare.org] *Im Auftrag von *Stéphane Adjemian *Gesendet:* Montag, 29. Mai 2017 23:46 *An:* List for Dynare developers *Betreff:* Re: [DynareDev] Beta release of 4.5.0
Hi Johannes,
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
@all We have now three pages on Dynare's Github/wiki:
KnownBugs, for the non fixed bugs in the current stable.
FixedBugs, for the fixed bugs, to be published in the next release of
the stable or the next major release,
- NewFeatures, for documenting the new features to be published in the
the next major release.
The corresponding pages on Dynare's MoinMoin wiki are frozen. In the FixedBug page, which I copy/paste from the updated NEWS file, I lost the external links to the Github issues (which are still available in the MoinMoin wiki). This is annoying and has to be fixed, but will take some time... Which I don't have right now.
You can clone the wiki pages locally:
git clone git@github.com:DynareTeam/dynare.wiki.git
modify the markdown files and push the patches on Github. There is a package called Gollum:
https://github.com/gollum/gollum
that will ease the development of the wiki pages, by allowing you to look at the results of your patches locally before committing and pushing on Gthub.
Best,
Stéphane.
2017-05-27 20:11 GMT+02:00 Johannes Pfeifer jpfeifer@gmx.de:
Hi Stéphane, what needs to be done for the manual and for Octave? Best, Johannes
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Stéphane Adjemian Gesendet: Samstag, 27. Mai 2017 16:12 An: dev@dynare.org Betreff: [DynareDev] Beta release of 4.5.0
Dear all,
I have just pushed a first beta release of 4.5.0. The package is available here:
http://www.dynare.org/beta-release
Things that remain to be done:
- Update the NEWS file (I am almost done with the first draft, I should
push it on Monday).
Update the reference manual.
Open a new page for the bugs in 4.5.x.
Update the microsoft code signature.
Check that the binaries are not detected as virus by the most common
anti-virus (we had this issue with release of 4.4.0).
- Fix the Octave bugs (those which can be fixed quickly).
I want to relase 4.5.0 for the summer school (12th of June). So it is likely that important bugs, related to block/bytecode, will remain in this release (Ferhat is already working on this). I think we will have to release 4.5.1 before the Summer holidays (end of July for me).
For the 4.5.x bug page (and also the New features page, 4.6-unstable) I propose to do it on Github (Wiki) instead of DynareWiki, the Markdown syntax (Github) is more elegant than moinmoin, and also this would greatly reduce the cost of updating the NEWS file.
Best, Stéphane.
-- Stéphane Adjemian Dynare Team
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
Stéphane Adjemian
Université du Maine, Dynare Team
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures : * estimation/fs2000_calibrated_covariance.mod * estimation/fs2000_model_comparison.mod * kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
--
Sébastien Villemot
Économiste
Sciences Po, OFCE
Dear Sébastien,
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get
*** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 ***
panic: Aborted -- stopping myself...
attempting to save variables to 'octave-workspace'...
save to 'octave-workspace' complete
Aborted
The first one seems to be a bug in Octave. I filed a bug report at https://savannah.gnu.org/bugs/index.php?51198
Best,
Johannes
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
* estimation/fs2000_calibrated_covariance.mod * estimation/fs2000_model_comparison.mod * kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker. So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod. Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0 Dear all, Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit : I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode. I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch. Basically there are only 3 test failures : estimation/fs2000_calibrated_covariance.modestimation/fs2000_model_co mparison.modkalman_filter_smoother/compare_results_simulation/fs2000_ ML.mod The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one). -- Sébastien Villemot Économiste Sciences Po, OFCE _______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
Sébastien Villemot
Économiste
Sciences Po, OFCE
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get
*** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 ***
panic: Aborted -- stopping myself...
attempting to save variables to 'octave-workspace'...
save to 'octave-workspace' complete
Aborted
The first one seems to be a bug in Octave. I filed a bug report at https://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.or mailto:dev-bounces@dynare.org g] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
* estimation/fs2000_calibrated_covariance.mod * estimation/fs2000_model_comparison.mod * kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI. Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode. I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
I consistently get it on Sedna. The message is
warning: matrix singular to machine precision, rcond = 1.34593e-17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double-linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI. Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode. I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Octave 4.2 on sedna was compiled by Stephane, though not necessarily correctly, because there is no Octave 4.2 backport to Jessie. We know this because, since 857fc3c4f4e3c1bc820b46ffbd0c3303abc1d143 (I think, though would need to check again), dynare fails to compile mex files against it while they compile correctly against Octave 4.2 on macOS.
Best, Houtan
On Jun 8, 2017, at 14:34, Johannes Pfeifer jpfeifer@gmx.de wrote:
I consistently get it on Sedna. The message is
warning: matrix singular to machine precision, rcond = 1.34593e-17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double-linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get
*** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted
The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
Sébastien Villemot
Économiste
Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
After Ferhat's fix, I was able to compile Dynare on Sedna. Before, the bytecode mex-file did not compile.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Houtan Gesendet: Donnerstag, 8. Juni 2017 14:47 An: List for Dynare developers Betreff: Re: [DynareDev] Beta release of 4.5.0
Octave 4.2 on sedna was compiled by Stephane, though not necessarily correctly, because there is no Octave 4.2 backport to Jessie. We know this because, since 857fc3c4f4e3c1bc820b46ffbd0c3303abc1d143 (I think, though would need to check again), dynare fails to compile mex files against it while they compile correctly against Octave 4.2 on macOS.
Best, Houtan
On Jun 8, 2017, at 14:34, Johannes Pfeifer jpfeifer@gmx.de wrote:
I consistently get it on Sedna. The message is
warning: matrix singular to machine precision, rcond = 1.34593e-17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double-linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get
*** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted
The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
Sébastien Villemot
Économiste
Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
_______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
I do not know how you are compiling it. The dynbot logs in /tmp that are run every night all show something like that attached.
That said, the test suite has been launched by dynbot twice yesterday and one of those compiled (though the test suite didn’t finish). Perhaps Stephane ran it as dynbot: Jun 7 17:03 /tmp/dynare-o-master-check-5YlKHyILFo.log ?
Best, Houtan
On Jun 8, 2017, at 14:48, Johannes Pfeifer jpfeifer@gmx.de wrote:
After Ferhat's fix, I was able to compile Dynare on Sedna. Before, the bytecode mex-file did not compile.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Houtan Gesendet: Donnerstag, 8. Juni 2017 14:47 An: List for Dynare developers Betreff: Re: [DynareDev] Beta release of 4.5.0
Octave 4.2 on sedna was compiled by Stephane, though not necessarily correctly, because there is no Octave 4.2 backport to Jessie. We know this because, since 857fc3c4f4e3c1bc820b46ffbd0c3303abc1d143 (I think, though would need to check again), dynare fails to compile mex files against it while they compile correctly against Octave 4.2 on macOS.
Best, Houtan
On Jun 8, 2017, at 14:34, Johannes Pfeifer jpfeifer@gmx.de wrote:
I consistently get it on Sedna. The message is
warning: matrix singular to machine precision, rcond = 1.34593e-17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double-linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get
*** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted
The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
Sébastien Villemot
Économiste
Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
I was exactly getting that message until Ferhat's fix. After his fix I did a pull and ran make all manually and it worked like a charm.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Houtan Gesendet: Donnerstag, 8. Juni 2017 15:00 An: List for Dynare developers Betreff: Re: [DynareDev] Beta release of 4.5.0
I do not know how you are compiling it. The dynbot logs in /tmp that are run every night all show something like that attached.
That said, the test suite has been launched by dynbot twice yesterday and one of those compiled (though the test suite didn’t finish). Perhaps Stephane ran it as dynbot: Jun 7 17:03 /tmp/dynare-o-master-check-5YlKHyILFo.log ?
Best, Houtan
I was wrong. We currently cannot rely on Octave on Sedna. Thus, I have no way to reproduce the problem. It does not appear on Windows.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Johannes Pfeifer Gesendet: Donnerstag, 8. Juni 2017 15:03 An: 'List for Dynare developers' Betreff: Re: [DynareDev] Beta release of 4.5.0
I was exactly getting that message until Ferhat's fix. After his fix I did a pull and ran make all manually and it worked like a charm.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Houtan Gesendet: Donnerstag, 8. Juni 2017 15:00 An: List for Dynare developers Betreff: Re: [DynareDev] Beta release of 4.5.0
I do not know how you are compiling it. The dynbot logs in /tmp that are run every night all show something like that attached.
That said, the test suite has been launched by dynbot twice yesterday and one of those compiled (though the test suite didn’t finish). Perhaps Stephane ran it as dynbot: Jun 7 17:03 /tmp/dynare-o-master-check-5YlKHyILFo.log ?
Best, Houtan
_______________________________________________ Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
I confirm that Dynare 4.5 does not compile on Debian Jessie, and that there are error messages in bytecode. But my understanding is that these messages are not related to the version of Octave, but rather to the version of g++, more precisely to the various flavours of C++ standards and their GNU interpretation (since the flag -std=gnu++11 is used at compile time). Le jeudi 08 juin 2017 à 15:23 +0200, Johannes Pfeifer a écrit :
I was wrong. We currently cannot rely on Octave on Sedna. Thus, I have no way to reproduce the problem. It does not appear on Windows.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Johannes Pfeifer Gesendet: Donnerstag, 8. Juni 2017 15:03 An: 'List for Dynare developers' Betreff: Re: [DynareDev] Beta release of 4.5.0
I was exactly getting that message until Ferhat's fix. After his fix I did a pull and ran make all manually and it worked like a charm.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Houtan Gesendet: Donnerstag, 8. Juni 2017 15:00 An: List for Dynare developers Betreff: Re: [DynareDev] Beta release of 4.5.0
I do not know how you are compiling it. The dynbot logs in /tmp that are run every night all show something like that attached.
That said, the test suite has been launched by dynbot twice yesterday and one of those compiled (though the test suite didn’t finish). Perhaps Stephane ran it as dynbot: Jun 7 17:03 /tmp/dynare-o-master- check-5YlKHyILFo.log ?
Best, Houtan
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
Sébastien Villemot
Économiste
Sciences Po, OFCE
I have recompiled Octave 4.2.1 and Dynare 4.5 on Debian Jessie (working around the bytecode issue), in order to mimic the environment of jessie, and I still don't get the crash. I don't understand how I got it in the first place. It does not look like a problem with consistently uninitialized memory space, because a check with valgrind could not unveil anything. So I'm giving up on this issue for now. Le jeudi 08 juin 2017 à 14:34 +0200, Johannes Pfeifer a écrit :
I consistently get it on Sedna. The message is warning: matrix singular to machine precision, rcond = 1.34593e-17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double- linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI. Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode. I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
Sébastien Villemot
Économiste
Sciences Po, OFCE
Did you run the entire testsuite or just the single .mod file?
Was your workaround for byte code something like the following in /mex/sources/bytecode/ErrorHandling.hh
+#ifdef __GNUC__ +# ifndef isnan +# define isnan(x) __builtin_isnan(x) +# endif +# ifndef isinf +# define isinf(x) __builtin_isinf(x) +# endif +#endif
Best, Houtan
On Jun 8, 2017, at 15:38, Sébastien Villemot <sebastien.villemot@sciencespo.fr mailto:sebastien.villemot@sciencespo.fr> wrote:
I have recompiled Octave 4.2.1 and Dynare 4.5 on Debian Jessie (working around the bytecode issue), in order to mimic the environment of jessie, and I still don't get the crash. I don't understand how I got it in the first place.
It does not look like a problem with consistently uninitialized memory space, because a check with valgrind could not unveil anything.
So I'm giving up on this issue for now.
Le jeudi 08 juin 2017 à 14:34 +0200, Johannes Pfeifer a écrit :
I consistently get it on Sedna. The message is
warning: matrix singular to machine precision, rcond = 1.34593e-17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double-linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org mailto:dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI.
Von: Dev [mailto:dev-bounces@dynare.org mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org mailto:dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get
*** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted
The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198 tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org mailto:dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org mailto:Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org mailto:Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev https://www.dynare.org/cgi-bin/mailman/listinfo/dev
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org mailto:Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Le jeudi 08 juin 2017 à 15:43 +0200, Houtan a écrit :
Did you run the entire testsuite or just the single .mod file?
On Debian Jessie I only ran the specific mod file.
Was your workaround for byte code something like the following in /mex/sources/bytecode/ErrorHandling.hh
Well, my workaround was more brutal :) I removed bytecode from the Makefiles , since it is not needed by the .mod file I'm interested in.
But indeed I guess the solution is to add or remove similar defines, or to add the std:: prefix.
+#ifdef __GNUC__ +# ifndef isnan +# define isnan(x) __builtin_isnan(x) +# endif +# ifndef isinf +# define isinf(x) __builtin_isinf(x) +# endif +#endif
Best, Houtan
On Jun 8, 2017, at 15:38, Sébastien Villemot <sebastien.villemot@sc iencespo.fr> wrote:
I have recompiled Octave 4.2.1 and Dynare 4.5 on Debian Jessie (working around the bytecode issue), in order to mimic the environment of jessie, and I still don't get the crash. I don't understand how I got it in the first place.
It does not look like a problem with consistently uninitialized memory space, because a check with valgrind could not unveil anything.
So I'm giving up on this issue for now.
Le jeudi 08 juin 2017 à 14:34 +0200, Johannes Pfeifer a écrit :
I consistently get it on Sedna. The message is warning: matrix singular to machine precision, rcond = 1.34593e- 17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double- linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI. Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod .
Von: Dev [mailto:dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode. I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev -- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
If this problem does not appear in Windows, we will fix it later. On sedna I install a chroot for stretch where I added the experimental repo and installed octave 4.2.1. In this environment dynare/octave compiles without issues, and we can run the testsuite. Sébastien is right, the issue is related to the C++ standard flavour, but I didn't have time to figure out which options should be passed to the configure script to have it right. I will give permissions to houtan and johannes to the stretch chroot, but I cannot do it now.
Best, Stéphane.
2017-06-08 15:38 GMT+02:00 Sébastien Villemot < sebastien.villemot@sciencespo.fr>:
I have recompiled Octave 4.2.1 and Dynare 4.5 on Debian Jessie (working around the bytecode issue), in order to mimic the environment of jessie, and I still don't get the crash. I don't understand how I got it in the first place.
It does not look like a problem with consistently uninitialized memory space, because a check with valgrind could not unveil anything.
So I'm giving up on this issue for now.
Le jeudi 08 juin 2017 à 14:34 +0200, Johannes Pfeifer a écrit :
I consistently get it on Sedna. The message is
warning: matrix singular to machine precision, rcond = 1.34593e-17 warning: called from dyn_first_order_solver at line 313 column 9 stochastic_solvers at line 264 column 18 resol at line 144 column 14 dynare_resolve at line 69 column 44 dsge_likelihood at line 266 column 57 hessian at line 92 column 34 dynare_estimation_1 at line 244 column 24 dynare_estimation at line 105 column 5 fs2000_ML at line 218 column 14 dynare at line 223 column 1 *** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x0000000002221150 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... *** Error in `/usr/local/bin/octave-cli-4.2.1': corrupted double-linked list: 0x00000000021fd560 *** panic: attempted clean up failed -- aborting...
Thus, the crash happens during or immediately after computing the Hessian. My hunch is that the Hessian again returns something complex with a NaN in it and we get the same crash as in the other mod-file when the inverse Hessian is supposed to be computed.
-----Ursprüngliche Nachricht----- Von: Dev [mailto:dev-bounces@dynare.org dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Donnerstag, 8. Juni 2017 14:18 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Well, I am no longer able to reproduce the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Do you have a way to consistently trigger it?
Le mercredi 07 juin 2017 à 18:16 +0200, Johannes Pfeifer a écrit :
Have you tried on Linux whether the problem also occurs on the Linux GUI? On the Windows CLI, everything runs smoothly, but there are still figures showing up on screen, which is different from the Linux CLI.
Von: Dev [mailto:dev-bounces@dynare.org dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 17:39 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Le mercredi 07 juin 2017 à 16:54 +0200, Johannes Pfeifer a écrit :
Thanks a lot. I cannot replicate the third crash on Windows. On Linux, I get
*** Error in `/usr/local/bin/octave-cli-4.2.1': double free or corruption (!prev): 0x000000000 26faad0 *** panic: Aborted -- stopping myself... attempting to save variables to 'octave-workspace'... save to 'octave-workspace' complete Aborted
The first one seems to be a bug in Octave. I filed a bug report at ht tps://savannah.gnu.org/bugs/index.php?51198
Thanks. I have added some technical information and a tentative patch to the bug tracker.
So I think we're done with the crash in estimation/fs2000_calibrated_covariance.mod.
Let's now investigate the crash in kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod.
Von: Dev [mailto:dev-bounces@dynare.org dev-bounces@dynare.org] Im Auftrag von Sébastien Villemot Gesendet: Mittwoch, 7. Juni 2017 12:10 An: dev@dynare.org Betreff: Re: [DynareDev] Beta release of 4.5.0
Dear all,
Le lundi 29 mai 2017 à 23:46 +0200, Stéphane Adjemian a écrit :
I do not receive integration tests reports for Octave since ages... I don't know why. Before the disappearance of the reports we had some failing tests, it would be nice to fix them. The difficulty with the Octave testsuite is that it often never ends. Some mod files, not always the same, run forever. I am convinced that this is not a bug in these files but rather a consequence of the memory leaks Ferhat has to fix in bytecode.
I have run the testsuite for Octave. The environment is Debian unstable, with Octave 4.2.1 from Debian experimental, and a few Octave forge packages recompiled by hand. The commit tested is d608588ef2 on the 4.5 branch.
Basically there are only 3 test failures :
estimation/fs2000_calibrated_covariance.mod estimation/fs2000_model_comparison.mod kalman_filter_smoother/compare_results_simulation/fs2000_ML.mod
The first and the third ones are crashes in Octave. I am going to investigate them. The second failure is simply the consequence of the first failure (the second test needs the output of the first one).
-- Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing listDev@dynare.orghttps://www.dynare.org/cgi-bin/mailman/listinfo/dev
Sébastien Villemot Économiste Sciences Po, OFCE
Dev mailing listDev@dynare.orghttps://www.dynare.org/cgi-bin/mailman/listinfo/dev
--
*Sébastien Villemot* Économiste Sciences Po, OFCE
Dev mailing list Dev@dynare.org https://www.dynare.org/cgi-bin/mailman/listinfo/dev
-
Houtan -
Johannes Pfeifer -
Michel Juillard -
Stéphane Adjemian -
Stéphane Adjemian -
Sébastien Villemot