On Wed, Sep 22, 2021 at 9:12 AM Jason Lowe-Power <ja...@lowepower.com> wrote:
> Hey Gabe, > > To solve this and the x86 test issue you've been having, I think there are > a couple of possibilities: > > 1. Can you use the docker images that kokoro uses? This will guarantee > that you are using the *exact* same "host" setup when running gem5. I think > this is the only way to have a consistent set of tests without something > like Bazel's reproducible build and test. > Yes, I think I might need to do that. Part of why that's a pain is that it uses a different tool chain than I do locally, and so I have to delete all my build directories or it chokes on the existing files. This is really a work around rather than an actual fix, of course. To be clear that's probably all we can do about it right now, but I think there's a flaw in the way our dynamic linking support works if it relies on things being installed on your host. I think you could argue it should *optionally*, or even by default, use things on your host just as a way to lower the initial barrier to getting things working, but I think it's a fundamental flaw if there's no possible optional way to use anything *but* what's on your host. That would mean if you wanted to use any dynamic library in your binary, you'd have to install it on your host, and you'd get possibly different or even incorrect behavior changing between a development workstation, a cluster to sweep parameters, a machine at home for tinkering, etc. > 2. We're open to modifying and improving the tests. If the tests don't > work for you, they probably don't work for others as well. Improving the > documentation so it's more clear how to use the tests or improving the > testing infrastructure or modifying the tests so that they work more easily > for more people would be a *very* welcome contribution. > Yes, this is already happening in the wild. I've seen reports from at least two people that have run into this x86 problem. I would be happy to fix up the testing infrastructure and would like to arrange for time to do a deep dive into it, but I can't for now at least.
_______________________________________________ gem5-dev mailing list -- gem5-dev@gem5.org To unsubscribe send an email to gem5-dev-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
