[BPF] scale trampoline stride in policy programs #10602
Open
+88
−29
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
marvin-tigera
added
release-note-required
There was a problem hiding this comment.
Pull Request Overview
This PR adjusts the trampoline stride in policy programs to better handle long jumps when jit_harden is enabled, reducing the likelihood of jumps exceeding the allowed distance.
- Introduces an atomic field for caching the current trampoline stride and updates it dynamically when errors are encountered.
- Updates BPF policy program building and associated tests to use the new stride parameter.
- Adds error handling in the BPF system call code to detect potential JIT errors.
Reviewed Changes
Copilot reviewed 5 out of 5 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
| felix/dataplane/linux/bpf_ep_mgr.go | Introduces an atomic trampoline stride field and adjusts policy program loading with retry logic. |
| felix/bpf/polprog/pol_prog_builder.go | Adds support for passing and storing the trampoline stride to the builder. |
| felix/bpf/bpf_syscall.go | Updates error handling for loading BPF programs with potential JIT errors. |
| felix/bpf/asm/asm_test.go | Updates tests to use the new TrampolineStrideDefault constant. |
| felix/bpf/asm/asm.go | Replaces the hard-coded interval with the new trampoline stride and adds a setter with validation. |
tomastigera
added
docs-not-required
tomastigera
force-pushed
the
tomas-bpf-harden-fix
branch
from
adc8169 to
992cd65
Compare
When jit_harden is enabled, some long jump may get over the allowed distance of +/-15bit when rewritten by the kernel. When this happens, we may retry and recompile the policy programs with the trampolines that make the jumps shorter. If that still fails we may retry with even more dense trampolines. We reduce teh stride by 1/4 with eveery retry to coverge reasonably fast, but not to oveshoot to much. Reductions by about 1/2 (2 iterrations) may be good enough for most programs. We remember the newly found stride for the given node so that we do not need to keep figuring out for other policy programs. It is likely that many policies would face the same issue. Note that the hardedning already increases the amount of executed instructions and adding some extra trampolines have probably much smaller effect than in a non-hardened program.
tomastigera
force-pushed
the
tomas-bpf-harden-fix
branch
from
992cd65 to
720983c
Compare
|
LGTM to me. I feel it will be better to run the bpf UTs, FVs with jit_harden set to 2. WDYT? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Related issues/PRs
refs bottlerocket-os/bottlerocket#4567
Todos
Release Note
Reminder for the reviewer
Make sure that this PR has the correct labels and milestone set.
Every PR needs one
docs-*label.docs-pr-required: This change requires a change to the documentation that has not been completed yet.docs-completed: This change has all necessary documentation completed.docs-not-required: This change has no user-facing impact and requires no docs.Every PR needs one
release-note-*label.release-note-required: This PR has user-facing changes. Most PRs should have this label.release-note-not-required: This PR has no user-facing changes.Other optional labels:
cherry-pick-candidate: This PR should be cherry-picked to an earlier release. For bug fixes only.needs-operator-pr: This PR is related to install and requires a corresponding change to the operator.