Hi, I'm running the pytorch example (non-SGX, gramine-direct) following this tutorial. The example works, but I repeatedly see:
Emulating a raw system/supervisor call. This degrades performance, consider patching your application to use Gramine syscall API.
I found a similar issue for running Go applications (which doesn't use libc to invoke syscalls). However, for the pytorch example, I expected everything to go through Gramine's patched libc, so I'm trying to understand why the info is getting triggered here.
FYI, I am running this in a VM instance in GCP,
Machine typle: c4-standard-16 (16 vCPUs, 60 GB Memory)
CPU: Intel Emerald Rapids
OS: ubuntu-2204-jammy-v20250815
Thanks!
Hi, I'm running the pytorch example (non-SGX, gramine-direct) following this tutorial. The example works, but I repeatedly see:
I found a similar issue for running Go applications (which doesn't use libc to invoke syscalls). However, for the pytorch example, I expected everything to go through Gramine's patched libc, so I'm trying to understand why the info is getting triggered here.
FYI, I am running this in a VM instance in GCP,
Machine typle: c4-standard-16 (16 vCPUs, 60 GB Memory)
CPU: Intel Emerald Rapids
OS: ubuntu-2204-jammy-v20250815
Thanks!