Bug#838789: Fwd: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined reference to `H5LTget_dataset_ndims'

[Gianfranco Costamagna]

Hi,

>and a porterbox won't work, either, since we cannot install non-free B-D
>there.

yep, just discovered that on plummer :/

G.

Bug#838789: Fwd: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined reference to `H5LTget_dataset_ndims'

[Andreas Beckmann]

On 2016-09-28 18:42, Gianfranco Costamagna wrote:
> But can a contrib package depend on stuff in non-free?

Yes, both Depends and Builds-Depends on non-free bist are fine.
The contrib package can also be under a free license, but as long as it
is not useful without some non-free bits, it has to go into contrib.
(e.g. game-engine-free/contrib + game-artwork/non-free, until someone
produces game-artwork-free/main).

B-D: non-free just means no autobuilding.

> I don't have any ppc64el machine :(

and a porterbox won't work, either, since we cannot install non-free B-D
there.


Andreas

Bug#838789: Fwd: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined reference to `H5LTget_dataset_ndims'

[Gianfranco Costamagna]

Hi Lumin, Mattia,



>Several weeks ago I posted a question on mentors, asking why buildd

>doesn't build caffe-contrib and saying "cuda" is missing. Now I know
>buildd won't pull non-free stuff and source-only upload may cause
>"partial missing" package.
>
>I noticed this FTBTS some time ago in package caffe (another
>version, free of cuda). However I didn't look into the reason of build failure
>because at that time there was an hdf5 transition.

oh indeed, so I'm forced to do binary-only uploads here...

But can a contrib package depend on stuff in non-free?


In any case, the package now is not building anymore, and I can do binary-only
uploads if:
1) the package is fixed
2) the arch is amd64.
I don't have any ppc64el machine :(

please fix the failure and let me know!

thanks you all

G.

Bug#838789: Fwd: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined reference to `H5LTget_dataset_ndims'

[lumin]

Control: affects -1 caffe

On Sat, 2016-09-24 at 22:51 +0200, Andreas Beckmann wrote:
> This is Bug#838789, but for some reason the BTS considers this as an
> unknown package, so you didn't get this forwarded ...

Thank you for forwarding. :-)

> Oh, you did a source-only upload. That does not work in this case, since
> you have Build-Depends in non-free. You'll always have to do
> source+binary uploads (for *all* architectures you want). And having a
> source package without binaries for some time seems to have made the
> package "partially disappear" from the archive.

Several weeks ago I posted a question on mentors, asking why buildd
doesn't build caffe-contrib and saying "cuda" is missing. Now I know
buildd won't pull non-free stuff and source-only upload may cause
"partial missing" package.

> caffe-contrib FTBFS in experimental:
> ../../../lib/libcaffe.so.1.0.0-rc3: undefined reference to
> `H5LTread_dataset_string'

I noticed this FTBTS some time ago in package caffe (another
version, free of cuda). However I didn't look into the reason of build failure
because at that time there was an hdf5 transition.

╰─>$ find /var/lib/dpkg/info/ -name '*.symbols' | grep hdf5 | xargs ack 
H5LTread_dataset_string

/var/lib/dpkg/info/libhdf5-openmpi-10:amd64.symbols
3734: H5LTread_dataset_string@HDF5_MPI_1.8.7 1.8.13
/var/lib/dpkg/info/libhdf5-10:amd64.symbols
3707: H5LTread_dataset_string@HDF5_SERIAL_1.8.7 1.8.13

╰─>$ apt list libhdf5-10 libhdf5-openmpi-10
Listing... Done
libhdf5-10/unstable,now 1.8.16+docs-8+b1 amd64 [installed,automatic]
libhdf5-openmpi-10/unstable,now 1.8.16+docs-8+b1 amd64 [installed]

Well, confusing.

> I'm *not* attaching the 160 MB build log. But I could look for details
> if needed.
> 
> In case it does matter:
> This was built against hdf5 1.8.16 from sid, not against 1.10 from
> experimental.

I used to test it against hdf5/sid. caffe* were uploaded to experimental
because runtime-dep python3-protobuf is still missing.

> IIRC some time ago apt changed how Build-Depends are resolved for
> packages in experimental, preferring packages from unstable unless
> a versioned constraint actually requires something from experimental.

But this reminds to to test whether it compiles against hdf5/experimental.

> I remember vaguely to have seen symbol changes during previous rebuilds,
> but I don't have the old longs any longer.

That's ok, I'll handle the changes if any, as long as I can pass the build.

> 
> Andreas

Bug#838789: Fwd: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined reference to `H5LTget_dataset_ndims'

[Mattia Rizzolo]

On Sat, Sep 24, 2016 at 10:51:37PM +0200, Andreas Beckmann wrote:
> This is Bug#838789, but for some reason the BTS considers this as an
> unknown package, so you didn't get this forwarded ...

the reason is still a regression of experimental not having
{Sources,Packages}.gz anymore, but having .xz only.
I poked dondelelcaro several times over IRC, but only get a partial fix
in place.
[this is also the reason this bug came to me: I'm in unknown-package@]

> Oh, you did a source-only upload. That does not work in this case, since
> you have Build-Depends in non-free. You'll always have to do
> source+binary uploads (for *all* architectures you want).

yes.  All good reason to not have such packages, if you ask me...
But for that you need to poke his sponsor, Gianfranco...

> And having a
> source package without binaries for some time seems to have made the
> package "partially disappear" from the archive.

umh?
AFAIK such kind of behaviour only happens with arch:all binaries (where
binaries actually get autoremoved)

>  Forwarded Message 
> Subject: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined
> reference to `H5LTget_dataset_ndims'
> Date: Sat, 24 Sep 2016 22:36:50 +0200
> From: Andreas Beckmann 
> To: Debian Bug Tracking System 
[...]
> In case it does matter:
> This was built against hdf5 1.8.16 from sid, not against 1.10 from
> experimental.
> IIRC some time ago apt changed how Build-Depends are resolved for
> packages in experimental, preferring packages from unstable unless
> a versioned constraint actually requires something from experimental.

that really depends on the resolver you use in your build daemon.  But
that always was the behaviour you had in experimental (unless you used
pbuilder and its broken idea of experimenal from version < 0.224)

-- 
regards,
Mattia Rizzolo

GPG Key: 66AE 2B4A FCCF 3F52 DA18  4D18 4B04 3FCD B944 4540  .''`.
more about me:  https://mapreri.org : :'  :
Launchpad user: https://launchpad.net/~mapreri  `. `'`
Debian QA page: https://qa.debian.org/developer.php?login=mattia  `-


signature.asc
Description: PGP signature

Bug#838789: Fwd: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined reference to `H5LTget_dataset_ndims'

[Andreas Beckmann]

This is Bug#838789, but for some reason the BTS considers this as an
unknown package, so you didn't get this forwarded ...

Oh, you did a source-only upload. That does not work in this case, since
you have Build-Depends in non-free. You'll always have to do
source+binary uploads (for *all* architectures you want). And having a
source package without binaries for some time seems to have made the
package "partially disappear" from the archive.

(source-only uploads works for nvidia-cuda-toolkit since it no longer
Build-Depends on something from the non-free driver)


Andreas

 Forwarded Message 
Subject: caffe-contrib: FTBFS: libcaffe.so.1.0.0-rc3: undefined
reference to `H5LTget_dataset_ndims'
Date: Sat, 24 Sep 2016 22:36:50 +0200
From: Andreas Beckmann 
To: Debian Bug Tracking System 

Source: caffe-contrib
Version: 1.0.0~rc3-2
Severity: serious
Justification: fails to build from source

Hi,

caffe-contrib FTBFS in experimental:

[...]
[100%] Linking CXX executable ../../../test/test.testbin
cd "/build/caffe-contrib-1.0.0~rc3/caffe_cuda_build/src/caffe/test" &&
/usr/bin/cmake -E cmake_link_script CMakeFiles/test.testbin.dir/link.txt
--verbose=1
/usr/bin/g++-5   -g -O2
-fdebug-prefix-map=/build/caffe-contrib-1.0.0~rc3=. -fPIE
-fstack-protector-strong -Wformat -Werror=format-security -Wall
-pedantic -Wdate-time -D_FORTIFY_SOURCE=2 -fPIC -Wall -Wno-sign-comp
are -Wno-uninitialized -O3 -DNDEBUG   -fPIE -pie -Wl,-z,relro -Wl,-z,now
-Wl,--as-needed CMakeFiles/test.testbin.dir/test_accuracy_layer.cpp.o
CMakeFiles/test.testbin.dir/test_argmax_layer.cpp.o CMakeFiles/test.tes
tbin.dir/test_batch_norm_layer.cpp.o
CMakeFiles/test.testbin.dir/test_batch_reindex_layer.cpp.o
CMakeFiles/test.testbin.dir/test_benchmark.cpp.o
CMakeFiles/test.testbin.dir/test_bias_layer.cpp.o CMakeFiles/test.tes
tbin.dir/test_blob.cpp.o
CMakeFiles/test.testbin.dir/test_caffe_main.cpp.o
CMakeFiles/test.testbin.dir/test_common.cpp.o
CMakeFiles/test.testbin.dir/test_concat_layer.cpp.o
CMakeFiles/test.testbin.dir/test_contrast
ive_loss_layer.cpp.o
CMakeFiles/test.testbin.dir/test_convolution_layer.cpp.o
CMakeFiles/test.testbin.dir/test_data_layer.cpp.o
CMakeFiles/test.testbin.dir/test_data_transformer.cpp.o
CMakeFiles/test.testbin.dir/te
st_db.cpp.o CMakeFiles/test.testbin.dir/test_deconvolution_layer.cpp.o
CMakeFiles/test.testbin.dir/test_dummy_data_layer.cpp.o
CMakeFiles/test.testbin.dir/test_eltwise_layer.cpp.o
CMakeFiles/test.testbin.dir/test_e
mbed_layer.cpp.o
CMakeFiles/test.testbin.dir/test_euclidean_loss_layer.cpp.o
CMakeFiles/test.testbin.dir/test_filler.cpp.o
CMakeFiles/test.testbin.dir/test_filter_layer.cpp.o
CMakeFiles/test.testbin.dir/test_flatte
n_layer.cpp.o
CMakeFiles/test.testbin.dir/test_gradient_based_solver.cpp.o
CMakeFiles/test.testbin.dir/test_hdf5_output_layer.cpp.o
CMakeFiles/test.testbin.dir/test_hdf5data_layer.cpp.o
CMakeFiles/test.testbin.dir/
test_hinge_loss_layer.cpp.o
CMakeFiles/test.testbin.dir/test_im2col_layer.cpp.o
CMakeFiles/test.testbin.dir/test_image_data_layer.cpp.o
CMakeFiles/test.testbin.dir/test_infogain_loss_layer.cpp.o
CMakeFiles/test.tes
tbin.dir/test_inner_product_layer.cpp.o
CMakeFiles/test.testbin.dir/test_internal_thread.cpp.o
CMakeFiles/test.testbin.dir/test_io.cpp.o
CMakeFiles/test.testbin.dir/test_layer_factory.cpp.o
CMakeFiles/test.testbin.
dir/test_lrn_layer.cpp.o
CMakeFiles/test.testbin.dir/test_math_functions.cpp.o
CMakeFiles/test.testbin.dir/test_maxpool_dropout_layers.cpp.o
CMakeFiles/test.testbin.dir/test_memory_data_layer.cpp.o CMakeFiles/test.
testbin.dir/test_multinomial_logistic_loss_layer.cpp.o
CMakeFiles/test.testbin.dir/test_mvn_layer.cpp.o
CMakeFiles/test.testbin.dir/test_net.cpp.o
CMakeFiles/test.testbin.dir/test_neuron_layer.cpp.o CMakeFiles/test
.testbin.dir/test_platform.cpp.o
CMakeFiles/test.testbin.dir/test_pooling_layer.cpp.o
CMakeFiles/test.testbin.dir/test_power_layer.cpp.o
CMakeFiles/test.testbin.dir/test_protobuf.cpp.o
CMakeFiles/test.testbin.dir/t
est_random_number_generator.cpp.o
CMakeFiles/test.testbin.dir/test_reduction_layer.cpp.o
CMakeFiles/test.testbin.dir/test_reshape_layer.cpp.o
CMakeFiles/test.testbin.dir/test_scale_layer.cpp.o CMakeFiles/test.testb
in.dir/test_sigmoid_cross_entropy_loss_layer.cpp.o
CMakeFiles/test.testbin.dir/test_slice_layer.cpp.o
CMakeFiles/test.testbin.dir/test_softmax_layer.cpp.o
CMakeFiles/test.testbin.dir/test_softmax_with_loss_layer.cp
p.o CMakeFiles/test.testbin.dir/test_solver.cpp.o
CMakeFiles/test.testbin.dir/test_solver_factory.cpp.o
CMakeFiles/test.testbin.dir/test_split_layer.cpp.o
CMakeFiles/test.testbin.dir/test_spp_layer.cpp.o CMakeFiles
/test.testbin.dir/test_stochastic_pooling.cpp.o
CMakeFiles/test.testbin.dir/test_syncedmem.cpp.o
CMakeFiles/test.testbin.dir/test_tanh_layer.cpp.o
CMakeFiles/test.testbin.dir/test_threshold_layer.cpp.o CMakeFiles/t
est.testbin.dir/test_tile_layer.cpp.o
CMakeFiles/test.testbin.dir/test_upgrade_proto.cpp.o