Re: [chipyard] Re: SOLVED: Issues in the set up

[Davide Giacomini]

After some hours, I managed to understand that the init-fpga didn't work well.

Thank you very much again for the help. I'm copying below (from the previous mail) the overview of what happened and the solutions.

For future helps, to make it work I did these steps:

- I created a file .sbtopts with this written inside it:

           -Dsbt.sourcemode=true

           -Dsbt.workspace=[/path/to/chipyard]/chipyard/tools

- I changed the SDK 14 in SDK 8 (I didn't try SDK 11, but accordingly to https://github.com/sbt/sbt/issues/5509 and https://github.com/sbt/sbt/issues/5093 it should equally work)

- If the version of verilator is different from the one indicated in the documentation (4.034) it may not work (it didn't work for me with a 3.092 version)

I have Ubuntu 20.04 LTS.

You all have a nice day.

Faithfully,

Davide Giacomini

Il Mer 14 Apr 2021, 13:28 Davide Giacomini <dadeg...@gmail.com> ha scritto:

Thank you! It worked!

These issues didn't come up searching for it, I am grateful you found it!

For future helps, to make it work I did these steps:

- I created a file .sbtopts with this written inside it:

           -Dsbt.sourcemode=true

           -Dsbt.workspace=[/path/to/chipyard]/chipyard/tools

- I changed the SDK 14 in SDK 8 (I didn't try SDK 11, but accordingly to https://github.com/sbt/sbt/issues/5509 and https://github.com/sbt/sbt/issues/5093 it should equally work)

- If the version of verilator is different from the one indicated in the documentation (4.034) it may not work (it didn't work for me with a 3.092 version)

I have Ubuntu 20.04 LTS.

Though, another error has just come up.

My aim was to run make SUB_PROJECT=vcu118 bitstream, and, after having initiated the fpga folder through the script init-fpga.sh, I ran the sub-project. Some syntax errors showed up and I understood that chipyard cannot import sifive.fpgashells*. That's a problem, because I also tried to search for those manually, but I couldn't find anything. I searched for it under the maven repository of the Berkeley university, but I couldn't find anything.

I report below the image:

Given that I don't know if it can be seen properly, I also attach it to the email.

Hoping this is the last step,

Thank you very much again and have a nice day.

Faithfully,

Davide Giacomini

Il giorno mer 14 apr 2021 alle ore 02:29 Abraham Gonzalez <abe.go...@berkeley.edu> ha scritto:

I've never seen that issue. A quick google search leads to this though: 

https://github.com/sbt/sbt/issues/5509
https://github.com/sbt/sbt/issues/5093

Have you tried what they recommend (switching Java/JDK versions)?

On Tue, Apr 13, 2021 at 12:22 AM Davide Giacomini <dadeg...@gmail.com> wrote:

I had deleted the whole stuff and started over, that's why now I have no more issues with env.sh and I have the env-riscv-tools.sh. Thank you very much!

I have another problem though. I'll explain it below. Should I post it in another conversation?

Precisely, now that the first step is done, I've another problem in starting the default example with verilator. If I run make under the sims/verilator directory, the error that I report below shows up. I've already tried all the versions of verilator that I found on the issues of GitHub. I h...ave the latest version of git and Ubuntu 20.04 LTS.

The error which shows up ends with:
...
...
[error] (hardfloat / Compile / compileIncremental) java.io.IOError: java.lang.RuntimeException: /packages cannot be represented as URI
[error] Total time: 4 s, completed Apr 12, 2021, 9:18:44 PM
make: *** [/home/davide/Projects/chipyard/common.mk:86: /home/davide/Projects/chipyard/sims/verilator/generated-src/chipyard.TestHarness.RocketConfig/sim_files.f] Error 1

The entire log is attached to this email.

Thank you very much again.

Faithfully,

Davide Giacomini

Il Mar 13 Apr 2021, 00:35 nat...@berkeley.edu <nat...@berkeley.edu> ha scritto:

The first error looks like QEMU's servers just had some problem, this happens from time to time. Running it again fixed it (as you saw). I don't see any obvious errors in your new log.

To clarify, are still having the issue with env.sh? If so, is there a chipyard/riscv-tools-install directory that has stuff in it? Is there a env-riscv-tools.sh? If you're still having problems, you might just try deleting the whole thing and starting over.

On Monday, April 12, 2021 at 12:14:11 PM UTC-7 Davide Giacomini wrote:

Forgot to attach the log.

Il giorno lunedì 12 aprile 2021 alle 21:13:32 UTC+2 Davide Giacomini ha scritto:

Update:

I set up again everything for the third time and it seems averything went smoothly. now the log of the erros is different, I upload it. The qemu submodules have been installed.

Anyway, I don't feel safe with this log...

Thank you very much.

Faithfully,

Davide Giacomini

Il giorno lunedì 12 aprile 2021 alle 17:35:52 UTC+2 Davide Giacomini ha scritto:

Hello everyone,

I am new to Chipyard and I am trying to get started with this tool. I tried two times to set up the tool for a risc-v environment, but after running buil-toolchains.sh, I don't have neither the /bin folder, nor the *tools-env.sh. 

I can only source env.sh, but it actually doesn't set up the RISCV environment. I have to set it up manually.

I also redirected the stderror to a log file, in order to understand if something goes wrong during the setup, but I can't understand what to look at among the error messages.

I attach it to this email, hoping that someone can better understand it. My root project directory is /home/davide/Projects/chipyard.

Thank you very much to all of you.

Faithfully,

Davide Giacomini

--
You received this message because you are subscribed to a topic in the Google Groups "Chipyard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/chipyard/nRij6ZHRKMY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to chipyard+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chipyard/7dacd160-b767-406b-a013-8c1311491e2dn%40googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Chipyard" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chipyard+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chipyard/CABnqHJLjy72FcY%3DiTkph40OL3A46qggPy9NBGL%3DJweFz2qZdPw%40mail.gmail.com.

--

Abraham J. Gonzalez

Electrical Engineering and Computer Science Ph.D. Student

The University of California at Berkeley

--
You received this message because you are subscribed to a topic in the Google Groups "Chipyard" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/chipyard/nRij6ZHRKMY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to chipyard+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chipyard/CAH9JGgp%3DiE%2BrohAMcRLwOgKdSiGQywx09%3Du5Afg9MS2Pk7TdFw%40mail.gmail.com.