Installing the dev environment

[Jake Smith]

Hi all,

I've ran the gist from https://gist.github.com/makestuff/5edd5ed38c7f2138537fe99e177d559d (umdk-20170319) and it errors at what I think is the end.

The error at the end of the script is as follows—

mkdir -p lin.x64/rel/.build

gcc -O3 -DNDEBUG  -c -m64 -DBYTE_ORDER=1234 -Wall -Wextra -Wundef -pedantic -std=c99 -Wstrict-prototypes -Wno-missing-field-initializers -Wstrict-aliasing=3 -fstrict-aliasing -Warray-bounds  -I. -I/home/jake/20170319/makestuff/common  -I/home/jake/20170319/makestuff/libs/libfpgalink  -MMD -MP -MF lin.x64/rel/.build/sig.o.d -Wa,-adhlns=lin.x64/rel/.build/sig.o.lst sig.c -o lin.x64/rel/.build/sig.o

gcc -O3 -DNDEBUG  -c -m64 -DBYTE_ORDER=1234 -Wall -Wextra -Wundef -pedantic -std=c99 -Wstrict-prototypes -Wno-missing-field-initializers -Wstrict-aliasing=3 -fstrict-aliasing -Warray-bounds  -I. -I/home/jake/20170319/makestuff/common  -I/home/jake/20170319/makestuff/libs/libfpgalink  -MMD -MP -MF lin.x64/rel/.build/main.o.d -Wa,-adhlns=lin.x64/rel/.build/main.o.lst main.c -o lin.x64/rel/.build/main.o

gcc -O3 -DNDEBUG  -c -m64 -DBYTE_ORDER=1234 -Wall -Wextra -Wundef -pedantic -std=c99 -Wstrict-prototypes -Wno-missing-field-initializers -Wstrict-aliasing=3 -fstrict-aliasing -Warray-bounds  -I. -I/home/jake/20170319/makestuff/common  -I/home/jake/20170319/makestuff/libs/libfpgalink  -MMD -MP -MF lin.x64/rel/.build/args.o.d -Wa,-adhlns=lin.x64/rel/.build/args.o.lst args.c -o lin.x64/rel/.build/args.o

for i in /home/jake/20170319/makestuff/libs/libfpgalink/lin.x64/rel/libfx2loader.so /home/jake/20170319/makestuff/libs/libfpgalink/lin.x64/rel/libbuffer.so /home/jake/20170319/makestuff/libs/libfpgalink/lin.x64/rel/liberror.so /home/jake/20170319/makestuff/libs/libfpgalink/lin.x64/rel/libusbwrap.so /home/jake/20170319/makestuff/libs/libfpgalink/lin.x64/rel/libfpgalink.so ; do cp -rp $i lin.x64/rel; done

gcc -m64 -DBYTE_ORDER=1234 -Wl,--relax,--gc-sections,-Map=lin.x64/rel/.build/loader.map,--cref,-rpath,\$ORIGIN,-rpath-link,lin.x64/rel -o lin.x64/rel/loader lin.x64/rel/.build/sig.o lin.x64/rel/.build/main.o lin.x64/rel/.build/args.o   -L/home/jake/20170319/makestuff/libs/libfpgalink/lin.x64/rel -lfpgalink 

strip lin.x64/rel/loader

cp: cannot stat 'README-SW': No such file or directory

jake@jake-VirtualBox:~/Desktop$ tar zcf umdk-l${ARCH}-${RELEASE}.tar.gz umdkv2-bin

tar: umdkv2-bin: Cannot stat: No such file or directory

tar: Exiting with failure status due to previous errors

Ubuntu 16.04 LTS, pretty fresh install.

Any pointers?

regards

jake

[csmith]

Hi Jake,

can you try the same build process, clean it first but install and run script.

Then, start a script command as:

jake_umdk_build_date.txt

press Enter, execute the build, when it fails type exit.  You will now have a complete listing of all the actions that occur from the build.

I can then look further up and see what has happened.

Also, the original build scripts have some dependencies comment out so just open then up and look for hashes which are comments '#'.

Once you have that, compress it and send me a copy.

Cheers,

Chris.

--
You received this message because you are subscribed to the Google Groups "UMDKv2 Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to umdkv2-dev+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Chris McClelland]

I think I probably wrote that gist to be directly sourced from a file
called ${HOME}/README-SW, so you need to do something like this:

cd ${HOME}
rm -f README-SW
wget -qO README-SW
https://gist.githubusercontent.com/makestuff/5edd5ed38c7f2138537fe99e177d559d/raw/39fa4702c10b6ee0f7c5941c689d88075737e712/umdk-20170319
(. README-SW 2>&1) > build.log

...then the result of the build will end up in build.log, which is
just as good for troubleshooting as the script method Chris proposed.

The first inevitable question arises, however: why bother building
from source? Why not just use the binary download suggested in the
wiki?

And the second inevitable question: "jake@jake-VirtualBox" is a strong
hint that you're running on a VM, which is explicitly not supported
(first sentence of the wiki, in fact).

Chris

[Jake Smith]

Chris, you are completely correct. I wanted to step through on a virtual box before setting up a machine specifically.

As it's not supported, I withdraw my question. Thanks for reading and the suggestions tho.

[Chris McClelland]

Well, in practice many things *will* work on a VM (e.g loading ROMs,
the gdb connection, etc) but instruction-tracing will not work since
it is fairly sensitive to USB latency. So by all means test it on your
VM.

[csmith]

I have the latest Debian and that works ok.

Again, I would like us to have a list of MD hardware that we all have
for UMDK testing but it also might be a good idea for us to keep a
listing of the Linux versions and distributions that we have attempted
to build the tools for too.

If no one objects then I will add a page to zako-factory's Wiki.

On 14/05/17 23:33, Chris McClelland wrote:
> Well, in practice many things *will* work on a VM (e.g loading ROMs,
> the gdb connection, etc) but instruction-tracing will not work since
> it is fairly sensitive to USB latency. So by all means test it on your
> VM.
>
>
> On 14 May 2017 at 22:00, Jake Smith <jak...@gmail.com> wrote:
>> Chris, you are completely correct. I wanted to step through on a virtual box
>> before setting up a machine specifically.

This works fine, but if you definitely want to use a working Linux
distro then Debian works.

[Jake Smith]

The additions listed by Chris M allowed the script to run without issue (on a VM)