gvisor
gvisor copied to clipboard
google
Application Kernel for Containers
Modify `make` rules to allow using a fully-local build cache. This includes the ability to force using local images only (do not check for updated manifests), and to explicitly mount...
![copybara-service[bot] avatar](https://avatars.githubusercontent.com/in/44061?v=4 "copybara-service[bot] avatar"){kind=avatar}
### Description In issue https://github.com/google/gvisor/issues/6664#issuecomment-935078809 it was discovered that the latest versions of libpcap have removed all non-mmapped packet capture support as of this commit https://github.com/the-tcpdump-group/libpcap/commit/457027ab43ceef03c578984653c0e59eedf5f46b. Which means latest versions...
Bumps the npm_and_yarn group with 2 updates in the /images/benchmarks/node directory: [express](https://github.com/expressjs/express) and [hbs](https://github.com/pillarjs/hbs). Updates `express` from 4.17.1 to 4.20.0 Release notes Sourced from express's releases. 4.20.0 What's Changed Important...
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot] avatar"){kind=avatar}
Add invalid error 58 in host_linux.go.
### Description The ioctl sniffer has a "compatibility enforcement mode" which verifies that a given GPU workload does not call an `ioctl` that is unknown to nvproxy. We should use...
### Description Our only GPU regression tests are doing inference. We should have a test that does some training too. ### Is this feature related to a specific bug? #10879...
### Description Currently, gVisor's NVIDIA GPU support feature (`nvproxy`) only supports CUDA-related commands (`ioctl`s, allocation classes, etc.). There have been multiple requests to expand this set to support non-CUDA GPU...
netstack: change sniffer LogPackets type to atomic.Bool This is a more appropriate type for what the values are.