dyld: Symbol not found: _util_malloc_

[DanielLenz]

Great paper presentation today, looking forward to diving into it and using the code! Will have to debug this first, though.

Just copying the issues from damonge/NaMaster#24, where several people have run into the following:

First executing the following, which works fine.

./configure
make
make install

Then running make check gives the following:

EST 16/53 nmt:he_beams [OK]
TEST 17/53 nmt:he_alm2cl dyld: lazy symbol binding failed: Symbol not found: _util_malloc_
  Referenced from: /Users/dlenz/software/pymodules/NaMaster/.libs/libnmt.0.dylib
  Expected in: flat namespace

dyld: Symbol not found: _util_malloc_
  Referenced from: /Users/dlenz/software/pymodules/NaMaster/.libs/libnmt.0.dylib
  Expected in: flat namespace

/bin/sh: line 1: 62678 Abort trap: 6           ${dir}$tst
FAIL: test/check_nmt

I get the very same error when trying to run the namaster executable.

It might make sense to add a travis build that uses a typical MacOS build, to catch this and similar errors.

[fjaviersanchez]

@DanielLenz I think this is a Mac specific issue. For some reason it is not finding one of the libraries. Does it work if you do export PATH=$PATH:/Users/dlenz/software/NaMaster/.libs?

Please, let us know of the outcome so we can add instructions to the README.

[DanielLenz]

Does it work if you do export PATH=$PATH:/Users/dlenz/software/NaMaster/.libs?

I'm afraid not.

[fjaviersanchez]

I'm afraid not.

Thanks! I'll look deeper into it.

[damonge]

@DanielLenz What compiler (and version) did you use to compile this? And was it the same that you used to install libsharp and any other libraries?

[damonge]

Also, could you check if the problem gets fixed if you do:

export LD_LIBRARY_PATH=/path/to/where/the/libsharp/libraries/are:$LD_LIBRARY_PATH

?

[DanielLenz]

Thank for helping me to debug this!

What compiler (and version) did you use to compile this?

NaMaster [master●] which gcc
gcc: aliased to /usr/local/bin/gcc-7
NaMaster [master●] gcc --version
gcc-7 (Homebrew GCC 7.2.0) 7.2.0

And was it the same that you used to install libsharp and any other libraries?

Good point, I might have used clang or gcc-6 to do that. I can fetch the latest libshard and fftw3 and fix that.
Is there a possibility to include these dependencies in NaMaster, as it is done e.g. in healpy (which includes e.g. healpix, cfitsio, and libsharp)?
Ironically, there are similar issues with the latest version, related to using different compilers: healpy/healpy#455

Also, could you check if the problem gets fixed if you do:
export LD_LIBRARY_PATH=/path/to/where/the/libsharp/libraries/are:$LD_LIBRARY_PATH

I'm afraid that didn't work.

[fjaviersanchez]

@DanielLenz I did a little bit of digging and it turns out that the c_utils library from libsharp wasn't being linked to libnmt.dylib (when linking libnmt.dylib for some reason the flags -lfftpack -lc_utils are missing). One solution consists on doing export LDFLAGS="-lfftpack -lc_utils" before running ./configure. Please, let me know if this works for you :)

[fjaviersanchez]

@damonge has just prepared a fix with which you don't even need to export these flags. Please, check it out at the dylib_fix branch.

[damonge]

Please let us know if it works, and we'll merge it

[DanielLenz]

Awesome, thanks for fixing this so rapidly!

One little issue popped up, everything else passed:

TEST 43/53 nmt:bins_flat_binning [FAIL]
  ERR: test/nmt_test_bins_flat.c:72  expected 0.000e+00, got 2.557e-04 (diff -2.557e-04, tol 1.000e-10)

Now looking into the python wrapper...

[DanielLenz]

Installing python didn't raise any issues, all the unittests I ran with python -m unittest -v passed as well. Thanks again!

[damonge]

Yeah, @fjaviersanchez mentioned that one. It's weird, because travis doesn't pick it up, and all tests have passed for me in the past. I think it may be a compiler-specific floating point thing related to the issue that caused the first problem to begin with. It's even wierder because python is supposed to be running the same test.

I'll have a quick look, but I wouldn't worry.

Many kudos to @fjaviersanchez for figuring this one out!

[damonge]

Tagging @anicola and @NicolettaK , since they had this problem too

[anicola]

Thank you very much @fjaviersanchez and @damonge! Now everything works for me as well. As Daniel, I also run into a one little issue:

TEST 41/53 nmt:bins_flat_binning [FAIL]
  ERR: test/nmt_test_bins_flat.c:72  expected 0.000e+00, got 2.652e+00 (diff -2.652e+00, tol 1.000e-10) 

Apart from that, everything works like a charm! Thanks again!