Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update PSA-eBPF documentation #3320

Merged
merged 2 commits into from
May 16, 2022
Merged

Update PSA-eBPF documentation #3320

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
2228bb2
Update PSA-eBPF documentation
osinstom May 13, 2022
ffb9620
Use local clang include instead of installing libbpf globally
osinstom May 13, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
30 changes: 20 additions & 10 deletions backends/ebpf/psa/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -417,22 +417,36 @@ Follow standard steps for the P4 compiler to install the eBPF backend with the P
The PSA implemented for eBPF backend is verified to work with the kernel version 5.8+ and `x86-64` CPU architecture.
Moreover, make sure that the BPF filesystem is mounted under `/sys/fs/bpf`.

Also, make sure you have the following packages installed:

```bash
$ sudo apt install -y clang llvm libelf-dev
```

You should also install a static `libbpf` library. Run the following commands:

```bash
$ python3 backends/ebpf/build_libbpf
```

### Compilation

You can compile a P4-16 PSA program for eBPF in a single step using:

```bash
make -f backends/ebpf/runtime/kernel.mk BPFOBJ={output} ARGS="-DPSA_PORT_RECIRCULATE=<RECIRCULATE_PORT_IDX>" P4FILE=<P4-PROGRAM>.p4 P4C=p4c-ebpf psa
make -f backends/ebpf/runtime/kernel.mk BPFOBJ=out.o ARGS="-DPSA_PORT_RECIRCULATE=<RECIRCULATE_PORT_IDX>" P4FILE=<P4-PROGRAM>.p4 P4C=p4c-ebpf psa
```

You can also perform compilation step by step:

```
$ p4c-ebpf --arch psa --target kernel -o out.c <program>.p4
$ clang -O2 -g -c -emit-llvm -DBTF -DPSA_PORT_RECIRCULATE=<RECIRCULATE_PORT_IDX> -c -o out.bc out.c
$ llc -march=bpf -filetype=obj -o out.o out.bc
$ clang -Ibackends/ebpf/runtime -Ibackends/ebpf/runtime/usr/include -O2 -g -c -emit-llvm -DBTF -DPSA_PORT_RECIRCULATE=<RECIRCULATE_PORT_IDX> -o out.bc out.c
$ llc -march=bpf -mcpu=generic -filetype=obj -o out.o out.bc
```

Note that you can use `-mcpu` flag to define the eBPF instruction set. Visit [this blog post](https://pchaigno.github.io/bpf/2021/10/20/ebpf-instruction-sets.html) to learn more about eBPF instruction sets.

The above steps generate `out.o` BPF object file that can be loaded to the kernel.

### psabpf API and psabpf-ctl
Expand Down Expand Up @@ -505,8 +519,8 @@ Refer to [the bpftool guide](https://manpages.ubuntu.com/manpages/focal/man8/bpf
We list the known bugs/limitations below. Refer to the Roadmap section for features planned in the near future.

- Larger bit fields (e.g. IPv6 addresses) may not work properly.
- We noticed that `bpf_xdp_adjust_meta()` isn't implemented by some NIC drivers, so the PSA impl may not work
with some NICs. So far, we have verified the correct behavior with Intel 82599ES.
- We noticed that `bpf_xdp_adjust_meta()` isn't implemented by some NIC drivers, so the `meta` XDP2TC mode may not work
with some NICs. So far, we have verified the correct behavior with Intel 82599ES. If a NIC doesn't support the `meta` XDP2TC mode you can use `head` or `cpumap` modes.
- `lookahead()` with bit fields (e.g., `bit<16>`) doesn't work.
- `@atomic` operation is not supported yet.
- `psa_idle_timeout` is not supported yet.
Expand All @@ -528,11 +542,7 @@ All the below features are already implemented and will be contributed to the P4
- **XDP support.** The current version of P4-eBPF compiler leverages the BPF TC hook for P4-programmable packet processing.
The TC subsystem enables implementation of the full PSA specification, contrary to XDP, but offers lower throughput. We're going to
contribute the XDP-based version of P4-eBPF that is not fully spec-compliant, but provides higher throughput.
- **XDP2TC mode.** the generated eBPF programs use `bpf_xdp_adjust_meta()` to transfer original EtherType from XDP to TC. This mode
may not be supported by some NIC drivers. We will add two other modes that can be used alternatively.
- **ValueSet support.** We plan to extend implementation to support other match kinds and multiple fields in the `select()` expression.
- **All PSA externs.** We plan to contribute implementation of [all PSA externs defined by the PSA specification](https://p4.org/p4-spec/docs/PSA.html#sec-psa-externs).
- **Ternary matching.** The PSA implementation for eBPF backend currently supports `exact` and `lpm` only. We will add support for `ternary` match kind.
- **Extended ValueSet support.** We plan to extend implementation to support other match kinds and multiple fields in the `select()` expression.

## Long-term goals

Expand Down