I built gem5-gcn3 following: https://www.gem5.org/documentation/general_docs/gpu_models/GCN3 <https://urldefense.com/v3/__https://www.gem5.org/documentation/general_docs/gpu_models/GCN3__;!!K-Hz7m0Vt54!mRTWMzzix5zgRepwL8DB5WQQlOgQ6HEtyWoa4AwRvVMuB3wwmxOp1fqFFYtTZlKsGiJ2OJcHMgBZh9n-1ldCVG0rUOI$>
I am currently trying the square test: https://gem5.googlesource.com/public/gem5-resources/+/refs/heads/stable/src/gpu/square/ <https://urldefense.com/v3/__https://gem5.googlesource.com/public/gem5-resources/*/refs/heads/stable/src/gpu/square/__;Kw!!K-Hz7m0Vt54!mRTWMzzix5zgRepwL8DB5WQQlOgQ6HEtyWoa4AwRvVMuB3wwmxOp1fqFFYtTZlKsGiJ2OJcHMgBZh9n-1ldCrIVMEao$> I was unable to run in docker; so I'm running on my host machine. When I run the following (for the square binary) <path to>gem5.opt <path to>apu_se.py -n 3 -c bin/square, I see at some point: build/GCN3_X86/gpu-compute/gpu_compute_driver.cc:926: fatal: 3223866242: bad ioctl <extra arg>% gem5: v22.1.0.0 gem5-resources: v22.0.0.2 Is this an error with the simulated driver in gem5? Or am I doing something wrong? This also happens with other binaries I tried from gem5-resources/src/gpu/...
_______________________________________________ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org
