-
Notifications
You must be signed in to change notification settings - Fork 575
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
NVDLA Cmod library not found #277
Comments
@luisarocas please share log for "tools/bin/tmake -build cmod_top" |
The log file outdir/build.log contains: make: Entering directory |
Hi @luisarocas, from outdir/build.log the cmod building looks like incomplete. |
I notices that when doing sudo tools/bin/tmake -build cmod_top many errors appear. I still do not know what is wrong. Can someone help me?:
|
Isn't |
@luisarocas why are you using sudo? It will change the paths you have set. |
Hello @luisarocas , I have recently experienced a similar problem to yours. (issue nvdla/sw#134 , please if someone knows how to correctly link it). I got the same error while trying to compile the CMOD library, but in my case its due to java, perl and the rest of the programs being located on /usr/bin, and the make instruction cannot obtain permission to deal with that folder. |
Hi @luisarocas, did you follow http://nvdla.org/hw/v1/integration_guide.html or http://nvdla.org/vp.html to install the tools like g++, java? |
Hi @javiergcanada , could you check the setting in tree.make? CPP, GCC are defined in tree.make. |
Hi @fanqifei, as you can see in the screenshot i posted on nvdla/sw#134, I think I'm defining the location of CPP and GCC as /usr/bin correctly. I'm also not getting any error message on the next step, telling me that they were incorrectly located, it apparently just doesn't have permission to access those files. Thank you for your help. |
Maybe after executing
Or remove the project from disk and checkout from repository again. As for configuring GCC location... Keep in mind that CMOD may still use gcc installed with the OS. See #191 |
Hi @javiergcanada , in my tree.make, the setting are below: |
Please try with docker image (https://hub.docker.com/r/nvdla/vp) which has all dependencies install to build cmod and vp Step 1. Run docker image Step 2. Build CMOD Creating tree.make to setup your working environment and projects tools/bin/tmake -build cmod_top Step 3. Build VP cd .. git clone https://github.com/nvdla/vp.git cmake -DCMAKE_INSTALL_PREFIX=build -DSYSTEMC_PREFIX=/usr/local/systemc-2.3.0/ -DNVDLA_HW_PREFIX=<Path_to_hw_repo> -DNVDLA_HW_PROJECT=nv_small |
All the steps involved seemed to run successfully for me. But it said CMOD Library not found. The hardware project is named nv_full by default. Files were generated under NVDLA/hw/outdir/nv_full/cmod. So when we run the cmake in the vp directory, give -DNVDLA_HW_PROJECT=nv_full in place of -DNVDLA_HW_PROJECT=nv_small Check if this works for you. |
When following http://nvdla.org/vp.html, in step 2.4.1 I obtain the error: NVDLA Cmod library not found.
Everything has been followed as explained in nvdla.org/vp, the system requirements are all met. I have tried to repeat the process several times and I keep receiving the same error.
The paths of vp and hw are well defined. Can someone help me?
I think the problem is in: tools/bin/tmake -build cmod_top
After doing this command I do not have hw/outdir//cmod/release. The header files and library should have been generated in release folder but that directory has not been created.
The text was updated successfully, but these errors were encountered: