Open BaseHardware opened 8 months ago
I think specifying the tool that I used is also important. It was vivado.
I think there might be something funky with env vars in scripts. Have a vague memory of this. Unfortunately don't have time myself to look into that right now. Maybe set it before running FuseSoC as a workaround?
@olofk Thanks for your reply. I share my understanding of this problem: When I tested this problem with the tool 'ghdl', the problem disappeared and it seems the problem is related to the behavior of edalize.
Although I don't know the exact reason, let me say the apparent reason: If I use 'ghdl', the script is executed through the _run_scripts
of the edatool
class, which sets the environmental variables and any other configurations properly with the core file, while in 'vivado', the script is executed through the 'pre_build' section of the generated Makefile in the build directory, which does not perform such an initialization task for scripts.
Hello, I am a newbie of fusesoc. I declared the script entries in my core file and added an
env
section to the entries to designate some environmental variables when executing commands in the script entry. The entries in thescripts
section of my core file will be used in thepre_build
orpost_build
section of thesynth
target.However, I cannot see any environment variables declared in my core file. How can could it be? Am I wrong in writing the core file? The following codelet is a part of my core file:
print_environ.py
does not show the environmental variable oftest
. Is this an intended behavior? How can I put an environment variable to my script?