Stopping GM Tests on Jenkins build for a bit.
Daniel Lee
what it is exactly, but hopefully I can figure it out tomorrow.
Daniel
---------- Forwarded message ----------
From: <mc.sta...@gmail.com>
Date: Thu, Feb 16, 2012 at 12:25 AM
Subject: Build failed in Jenkins: Stan GM Tests #35
To: stan-b...@googlegroups.com
See <http://d1m1s1b1.stat.columbia.edu:8080/job/Stan%20GM%20Tests/35/changes>
Changes:
[bea...@alum.mit.edu] precompiling model_header.hpp. Only works for
g++ for now.
[bea...@alum.mit.edu] removed warnings about missing .d.#### files
[bea...@alum.mit.edu] created src/stan/model/model_header.hpp for use
in generating models.
[bea...@alum.mit.edu] created src/stan/model/model_header.hpp for use
in generating models.
[bea...@alum.mit.edu] adding models/ to svn:ignore list
[bea...@alum.mit.edu] removing makefile.Jenkins.
[bea...@alum.mit.edu] moved src/stan/maths to src/stan/math
[guo...@gmail.com] bugs example vol2/cervix: add the version
integrating out binary latent variables
[bea...@alum.mit.edu] moved src/stan/maths to src/stan/math
[bea...@alum.mit.edu] moved src/stan/mcmc/prob_grad.hpp and
prob_grad_ad.hpp to src/stan/model/
[bea...@alum.mit.edu] added additional echo line to makefile.
[ca...@alias-i.com] added to-do and prob grad
[bea...@alum.mit.edu] moving makefile directory
------------------------------------------
Started by user Daniel Lee
Building in workspace
<http://d1m1s1b1.stat.columbia.edu:8080/job/Stan%20GM%20Tests/ws/>
Updating https://stan.googlecode.com/svn/trunk
D makefile.Jenkins
U doc\notes\TO-DO.txt
D src\makefile
D src\test\maths
U src\test\prob\transform_test.cpp
D src\test\file_includes\stan_mcmc_prob_grad_test.cpp
D src\test\file_includes\stan_mcmc_prob_grad_ad_test.cpp
U src\test\file_includes\stan_maths_matrix_error_handling_test.cpp
A src\test\file_includes\stan_model_prob_grad_test.cpp
U src\test\file_includes\stan_maths_error_handling_test.cpp
U src\test\file_includes\stan_maths_matrix_test.cpp
U src\test\file_includes\stan_maths_constants_test.cpp
U src\test\file_includes\stan_maths_util_test.cpp
A src\test\file_includes\stan_model_prob_grad_ad_test.cpp
A src\test\file_includes\stan_model_test.cpp
U src\test\file_includes\stan_maths_special_functions_test.cpp
U src\test\io\reader_test.cpp
U src\test\io\dump_test.cpp
U src\test\io\writer_test.cpp
A src\test\math
A src\test\math\matrix_test.cpp
AU src\test\math\special_functions_test.cpp
A src\test\math\constants_test.cpp
A src\test\math\error_handling_test.cpp
U src\test\agrad\matrix_test.cpp
U src\test\agrad\error_handling_test.cpp
D src\models\bugs_examples\vol1\leukfr\leukfr.stan.0
A src\models\bugs_examples\vol1\leukfr\leukfr.stan
D src\models\bugs_examples\vol1\leuk\leuk.stan.0
A src\models\bugs_examples\vol1\leuk\leuk.stan
U src\models\bugs_examples\vol2\cervix\post.R
A src\models\bugs_examples\vol2\cervix\cervix2.stan
A src\models\bugs_examples\vol2\cervix\cervix2.Rdata
U src\models\bugs_examples\vol2\cervix\makefile
U src\models\bugs_examples\vol3\fire\fire.stan
U src\models\bugs_examples\vol3\fire\makefile
U src\demo\bivar_norm.cpp
U src\demo\tmp.cpp
U src\demo\lda.cpp
U src\demo\logistic_regression_cauchy.cpp
U src\demo\norm_sample.cpp
U src\demo\isnmf.cpp
U src\demo\logistic_regression.cpp
U src\demo\model1.cpp
U src\demo\experiment.hpp
U src\demo\normal.cpp
U src\demo\bugs_rats.cpp
U src\demo\eight_schools.cpp
D src\stan\maths.hpp
D src\stan\maths
D src\stan\mcmc\prob_grad.hpp
D src\stan\mcmc\prob_grad_ad.hpp
U src\stan\mcmc\nuts.hpp
U src\stan\mcmc\adaptive_cdhmc.hpp
U src\stan\mcmc\util.hpp
U src\stan\mcmc\adaptive_hmc.hpp
U src\stan\mcmc\hmc.hpp
A src\stan\math.hpp
A src\stan\model.hpp
A src\stan\math
A src\stan\math\EigenDenseBaseAddons.hpp
A src\stan\math\constants.hpp
A src\stan\math\math.dox
A src\stan\math\error_handling.hpp
A src\stan\math\util.hpp
A src\stan\math\matrix_error_handling.hpp
A src\stan\math\boost_error_handling.hpp
A src\stan\math\matrix.hpp
A src\stan\math\special_functions.hpp
A src\stan\model
A src\stan\model\prob_grad.hpp
A src\stan\model\prob_grad_ad.hpp
A src\stan\model\model_header.hpp
U src\stan\prob\transform.hpp
U src\stan\prob\distributions\multivariate\discrete\multinomial.hpp
U src\stan\prob\distributions\multivariate\discrete\categorical.hpp
U src\stan\prob\distributions\multivariate\continuous\inv_wishart.hpp
U src\stan\prob\distributions\multivariate\continuous\wishart.hpp
U src\stan\prob\distributions\multivariate\continuous\multi_normal.hpp
U src\stan\prob\distributions\multivariate\continuous\multi_student_t.hpp
U src\stan\prob\distributions\multivariate\continuous\lkj_corr.hpp
U src\stan\prob\distributions\multivariate\continuous\dirichlet.hpp
U src\stan\prob\distributions\multivariate\continuous\lkj_cov.hpp
U src\stan\prob\distributions\univariate\discrete\poisson.hpp
U src\stan\prob\distributions\univariate\discrete\binomial.hpp
U src\stan\prob\distributions\univariate\discrete\bernoulli.hpp
U src\stan\prob\distributions\univariate\discrete\neg_binomial.hpp
U src\stan\prob\distributions\univariate\discrete\beta_binomial.hpp
U src\stan\prob\distributions\univariate\discrete\hypergeometric.hpp
U src\stan\prob\distributions\univariate\continuous\pareto.hpp
U src\stan\prob\distributions\univariate\continuous\lognormal.hpp
U src\stan\prob\distributions\univariate\continuous\inv_gamma.hpp
U src\stan\prob\distributions\univariate\continuous\inv_chi_square.hpp
U src\stan\prob\distributions\univariate\continuous\weibull.hpp
U src\stan\prob\distributions\univariate\continuous\student_t.hpp
U src\stan\prob\distributions\univariate\continuous\scaled_inv_chi_square.hpp
U src\stan\prob\distributions\univariate\continuous\beta.hpp
U src\stan\prob\distributions\univariate\continuous\cauchy.hpp
U src\stan\prob\distributions\univariate\continuous\logistic.hpp
U src\stan\prob\distributions\univariate\continuous\uniform.hpp
U src\stan\prob\distributions\univariate\continuous\double_exponential.hpp
U src\stan\prob\distributions\univariate\continuous\gamma.hpp
U src\stan\prob\distributions\univariate\continuous\chi_square.hpp
U src\stan\prob\distributions\univariate\continuous\exponential.hpp
U src\stan\prob\distributions\univariate\continuous\normal.hpp
U src\stan\gm\command.hpp
U src\stan\gm\generator.hpp
U src\stan\mcmc.hpp
U src\stan\io\writer.hpp
U src\stan\io\reader.hpp
U src\stan\io\dump.hpp
U src\stan\io\csv_writer.hpp
U src\stan\agrad\error_handling.hpp
U src\stan\agrad\matrix.hpp
U src\stan\agrad\special_functions.hpp
U src\stan.hpp
U makefile
A make
A make\linux
A make\mac
A make\tests
A make\models
A make\demo
A make\detect_os
A make\detect_cc
A make\doxygen
A make\win
U .
At revision 780
[workspace] $ cmd /c call C:\Windows\TEMP\hudson3729872063478268530.bat
<http://d1m1s1b1.stat.columbia.edu:8080/job/Stan%20GM%20Tests/ws/>call>
"C:\Program Files (x86)\Microsoft Visual Studio 10.0\VC\vcvarsall.bat"
Setting environment for using Microsoft Visual Studio 2010 x86 tools.
rm -f -r *.dSYM
rm -f -r models src/stan/model/model_header.hpp.gch
src/stan/model/model_header.hpp.pch
rm -f -r doc/api
rm -f -r demo
rm -f -r test
--- Precompiling model_header.hpp for g++ ---
g++ -I src -I lib -O3 -Wall -std=gnu++0x -Wall -Wno-unused-function
-c -o src/stan/model/model_header.hpp.gch
src/stan/model/model_header.hpp
In file included from src/stan/mcmc/nuts.hpp:15:0,
from src/stan/gm/command.hpp:17,
from src/stan/model/model_header.hpp:8:
src/stan/mcmc/dualaverage.hpp: In member function 'void
stan::mcmc::GrowingBatches::update(const std::vector<double>&,
std::vector<double>&)':
src/stan/mcmc/dualaverage.hpp:145:32: warning: unknown conversion type
character 'z' in format [-Wformat]
src/stan/mcmc/dualaverage.hpp:145:32: warning: format '%f' expects
argument of type 'double', but argument 3 has type 'size_t {aka long
long unsigned int}' [-Wformat]
src/stan/mcmc/dualaverage.hpp:145:32: warning: unknown conversion type
character 'z' in format [-Wformat]
src/stan/mcmc/dualaverage.hpp:145:32: warning: too many arguments for
format [-Wformat-extra-args]
In file included from src/stan/model/model_header.hpp:8:0:
src/stan/gm/command.hpp: In function 'int stan::gm::nuts_command(int,
const char**)':
src/stan/gm/command.hpp:262:57: warning: left shift count >= width of
type [enabled by default]
cl /EHsc /I src /I lib src/demo/gm.cpp /link /OUT:demo/gm.exe
Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 16.00.30319.01 for 80x86
Copyright (C) Microsoft Corporation. All rights reserved.
gm.cpp
Microsoft (R) Incremental Linker Version 10.00.30319.01
Copyright (C) Microsoft Corporation. All rights reserved.
/out:gm.exe
/OUT:demo/gm.exe
gm.obj
--- Translating Stan graphical model to C++ code ---
cat src/models/bugs_examples/vol1/blocker/blocker.stan | demo/gm.exe >
models/bugs_examples/vol1/blocker/blocker.cpp
--- Compiling C++ model ---
g++ -I src -I lib -O3 -Wall -std=gnu++0x -Wall -Wno-unused-function
-c -o models/bugs_examples/vol1/blocker/blocker.o
models/bugs_examples/vol1/blocker/blocker.cpp
models/bugs_examples/vol1/blocker/blocker.cpp: In member function
'void demo_model_namespace::demo_model::transform_inits(const
stan::io::var_context&, std::vector<int>&, std::vector<double>&)':
models/bugs_examples/vol1/blocker/blocker.cpp:159:39: warning:
comparison between signed and unsigned integer expressions
[-Wsign-compare]
models/bugs_examples/vol1/blocker/blocker.cpp:171:39: warning:
comparison between signed and unsigned integer expressions
[-Wsign-compare]
models/bugs_examples/vol1/blocker/blocker.cpp: In member function
'void demo_model_namespace::demo_model::write_csv_header(std::ostream&)':
models/bugs_examples/vol1/blocker/blocker.cpp:221:41: warning:
comparison between signed and unsigned integer expressions
[-Wsign-compare]
models/bugs_examples/vol1/blocker/blocker.cpp:225:41: warning:
comparison between signed and unsigned integer expressions
[-Wsign-compare]
make/models:27: recipe for target
`models/bugs_examples/vol1/blocker/blocker.o' failed
make: *** [models/bugs_examples/vol1/blocker/blocker.o] Error 1
Build step 'Execute Windows batch command' marked build as failure
Ben Goodrich
lot of RAM nowdays (like 4GB). Maybe the failures are due to that
getting maxed out? Or maybe those tests could be split into smaller
pieces?
Daniel Lee
models stopped building after refactoring.
Moving namespaces, changes to physical layout, and adding one header
file that includes others. Somewhere in there, it makes g++ on windows
fail without any intelligible error message. My mac can handle it
fine.
On a similar note, I can't get clang to use the precompiled header
without crashing. It complains about too many open files.
Bob Carpenter
science?
- Bob
On 2/16/12 12:03 PM, Daniel Lee wrote:
> This one is a little different. It's a compiler related issue. The
> models stopped building after refactoring.
>
> Moving namespaces, changes to physical layout, and adding one header
> file that includes others. Somewhere in there, it makes g++ on windows
> fail without any intelligible error message. My mac can handle it
> fine.
>
> On a similar note, I can't get clang to use the precompiled header
> without crashing. It complains about too many open files.
Might some file somewhere not have a header
guard and be getting continually compiled?
The only namespace shift was for the models
prob_grad and prob_grad_ad, which moved to
their own namespace, stan::model?
This would turn off argument-dependent lookup
for variables that are in the stan::mcmc namespace
getting passed to functions in the models.
We should also do a thorough "using" review.
I'm still worried there may be order dependencies
in the compiles.
- Bob
Daniel Lee
I'm not going to think too hard about why, but if I do find the
reason, I'll email the group.
Bob Carpenter
.hpp files that aren't including everything
they should, but rather relying on other files
to include it.
If we design our headers well, order of includes
shouldn't matter.
So we should definitely clean this up if we can
track it down. We should sit down and review
all the "using"s in the entire package.
- Bob
Daniel Lee
The precompiled header kills the compiler.