"make check" error when compiling Gawk-4.1.4 (for Linux From Scratch v8.0)

Microbob · 06-16-2017, 10:07 PM

Hello,
Recently I've started working on building a Linux based OS using the book Linux From Scratch (LFS 8.0). I'm currently working on section 5.23, compiling the Gawk-4.1.4 package. So far, I've ran the

Code:

./configure --prefix=/tools

and the

Code:

make

without any problems. When I run the

Code:

make check

command though, I get this error:

Code:

- gawk: mbstr1.awk:2: warning: Invalid multibyte data detected. There may be a mismatch between your data and your locale.

. This error is repeated multiple times in the "make check" verbose, and it ultimately terminates the command:

Code:

Makefile:1651: recipe for target 'check' failed
make[1]: *** [check] Error 1
make[1]: Leaving directory '/mnt/lfs/sources/gawk-4.1.4/test'
Makefile:742: recipe for target 'check-recursive' failed
make: *** [check-recursive] Error 1

.

What can I do to fix this issue? I've already searched around a bit online and found another instances of this issue, but I don't understand how to use the solution given.

Thank you in advance for help on my issue!

Luridis · 06-17-2017, 01:23 AM

echo $LC_ALL

If what it prints is not "C" then your host environment is likely polluting the build, or your build environment was setup wrong. Are you compiling on MacOS? It's hard to do any more speculation because you didn't include enough of the error to tell us anything more than something went wrong.

Microbob · 06-17-2017, 01:52 AM

Thank you for your input Luridis and I apologize for not including enough details in the beginning. Here's some data that may help:

My host OS is Peppermint OS 8.0 (based on Ubuntu 16.04.2)
My result for running "echo $LC_ALL" was "POSIX"
I have properly compiled the GCC and the glibc libraries into my system, and I have checked that they are working (which they are)
I initially posted this thread in the Linux Software thread group, and I was told that it was OK to just ignore that command and continue. Is that true?

Now that I've determined that my host is likely polluting my build, how can I fix this issue?

Thank you!

Microbob · 06-17-2017, 01:54 AM

Would it fix the problem if I just "export"ed "C" to "$LC_ALL"?

hazel · 06-17-2017, 02:03 AM

Why are you running "make check" anyway? The book recommends omitting checks in chapter 5 because of pollution by the host system. However the error you have reported does make me wonder if you are running this as the LFS user with the correct environment. It's worth checking this against section 4.4. If everything is as described, forget about the "make check" errors. Just don't do the checks until you reach chapter 6.

Microbob · 06-17-2017, 02:09 AM

I'm sure that I have the correct environment setup and that my "lfs" user is also setup properly. So, I'll just go ahead and skip the "make check" for now. Thank you for you input!

Luridis · 06-17-2017, 04:26 AM

Yea, it didn't click that you were in Chapter 5. There's no real point to running tests on tools for errors that will clearly announce themselves as soon as you're in the chroot (chapter 6). If any part of that chain is significantly broken, you won't manage to get the new system's toolchain standing.

jeremy · 06-18-2017, 01:46 PM

Please post your thread in only one forum. Posting a single thread in the most relevant forum will make it easier for members to help you and will keep the discussion in one place. This thread is being closed because it is a duplicate.