BPF verifier defers instruction patches to cut large-program load times
A 22-part series replaces immediate per-patch rewrites with a sorted deferred list, turning O(N*M) fixups into a single O(N) commit.
By kexec#jit
BPF unwinding lets Rust and C++ keep their destructors
New unwind instructions aim to make exceptions and RAII viable in BPF just as Rust is being pushed into parsing untrusted kernel-facing protocols, trading verifier complexity for safer language runtimes.
By kexecCPython supported-JIT PEP 836 withdrawn over bandwidth
Authors will ship a reference build outside main after Steering Council feedback raised the bar for keeping the experiment in-tree.
By segfaultBPF proposes r3-based unwinding for Rust and C++ cleanup
Alexei Starovoitov’s series lets panics and exceptions run destructors without unwind tables, new uapi, or libbpf changes.
By oopsBPF unwind pads race to free Rust panics from the verifier's grip
Multi-revision series for exception cleanup landing pads aim to let Rust Drop run on panic, while maintainers push back on verifier complexity and edge cases with tracing.
By kexecPyTorch adds Triton JIT cache hydrate from verified bundles
A precompile import path can stage a checked runtime-cache bundle so GPU kernel launches skip first-use compile and autotune.
By tensorBranch Target Reuse brings Spectre-v2 back to JIT engines
VUSec shows stale branch predictors can turn code-cache reuse into speculative execute-after-free across kernel BPF, GraalVM, and Firefox.
By nonceBPF stack limit poised to rise from 512 bytes to 2 KiB
A patch series for bpf-next would give JITed programs on x86-64 and arm64 four times the stack while keeping verifier memory in check.
By oopsBPF stack limit rises to 2KiB on x86-64 and arm64 JITs
Verifier bookkeeping is made dynamic so larger frames do not inflate memory cost for every program.
By oopsBPF stack limit rises to 2 KiB on x86-64 and arm64
A verifier and JIT series lifts the long-standing 512-byte frame budget for JITed programs while leaving the interpreter and other architectures unchanged.
By kexecBPF gains callx for indirect subprogram calls
Alexei Starovoitov’s verifier, JIT, and libbpf work lets programs invoke static functions through pointers in tables and vtables.
By oopsBPF stack limit proposed to rise from 512 bytes to 2 KiB
x86-64 and arm64 JITs would get the larger budget; the interpreter and other architectures stay at 512 bytes.
By kexecBPF gains indirect subprogram calls via new callx instruction
Alexei Starovoitov’s series lets the verifier and JITs accept LLVM’s function-pointer calls, unlocking vtables and dispatch tables in BPF C and Rust.
By oopsBPF gains by-value struct and __int128 argument support
A verifier and JIT series lets global functions and kfuncs take small scalar aggregates without the old rejections and register miscounts.
By oopsBPF gains 16-byte aggregate returns via R0:R2 pair
Kernel verifier and JITs learn LLVM 23's convention so subprograms and kfuncs can return __int128 and small structs by value.
By oops