[bpf-next,v2,0/9] riscv: BPF JIT fix, optimizations and far jumps support

Message

Björn Töpel Dec. 16, 2019, 9:13 a.m. UTC

Hi!

This series contain one non-critical fix, support for far jumps. and
some optimizations for the BPF JIT.

Previously, the JIT only supported 12b branch targets for conditional
branches, and 21b for unconditional branches. Starting with this
series, 32b branching is supported.

As part of supporting far jumps, branch relaxation was introduced. The
idea is to start with a pessimistic jump (e.g. auipc/jalr) and for
each pass the JIT will have an opportunity to pick a better
instruction (e.g. jal) and shrink the image. Instead of two passes,
the JIT requires more passes. It typically converges after 3 passes.

The optimizations mentioned in the subject are for calls and tail
calls. In the tail call generation we can save one instruction by
using the offset in jalr. Calls are optimized by doing (auipc)/jal(r)
relative jumps instead of loading the entire absolute address and
doing jalr. This required that the JIT image allocator was made RISC-V
specific, so we can ensure that the JIT image and the kernel text are
in range (32b).

The last two patches of the series is not critical to the series, but
are two UAPI build issues for BPF events. A closer look from the
RV-folks would be much appreciated.

The test_bpf.ko module, selftests/bpf/test_verifier and
selftests/seccomp/seccomp_bpf pass all tests.

RISC-V is still missing proper kprobe and tracepoint support, so a lot
of BPF selftests cannot be run.


Thanks,
Björn

v1->v2: [1]
 * Removed unused function parameter from emit_branch()
 * Added patch to support far branch in tail call emit

[1] https://lore.kernel.org/bpf/[e-mail]/


Björn Töpel (9):
  riscv, bpf: fix broken BPF tail calls
  riscv, bpf: add support for far branching
  riscv, bpf: add support for far branching when emitting tail call
  riscv, bpf: add support for far jumps and exits
  riscv, bpf: optimize BPF tail calls
  riscv, bpf: provide RISC-V specific JIT image alloc/free
  riscv, bpf: optimize calls
  riscv, bpf: add missing uapi header for BPF_PROG_TYPE_PERF_EVENT
    programs
  riscv, perf: add arch specific perf_arch_bpf_user_pt_regs

 arch/riscv/include/asm/perf_event.h          |   4 +
 arch/riscv/include/asm/pgtable.h             |   4 +
 arch/riscv/include/uapi/asm/bpf_perf_event.h |   9 +
 arch/riscv/net/bpf_jit_comp.c                | 531 ++++++++++---------
 tools/include/uapi/asm/bpf_perf_event.h      |   2 +
 5 files changed, 312 insertions(+), 238 deletions(-)
 create mode 100644 arch/riscv/include/uapi/asm/bpf_perf_event.h

Comments

Daniel Borkmann Dec. 19, 2019, 3:07 p.m. UTC | #1

On Mon, Dec 16, 2019 at 10:13:34AM +0100, Björn Töpel wrote:
> 
> This series contain one non-critical fix, support for far jumps. and
> some optimizations for the BPF JIT.
> 
> Previously, the JIT only supported 12b branch targets for conditional
> branches, and 21b for unconditional branches. Starting with this
> series, 32b branching is supported.
> 
> As part of supporting far jumps, branch relaxation was introduced. The
> idea is to start with a pessimistic jump (e.g. auipc/jalr) and for
> each pass the JIT will have an opportunity to pick a better
> instruction (e.g. jal) and shrink the image. Instead of two passes,
> the JIT requires more passes. It typically converges after 3 passes.
> 
> The optimizations mentioned in the subject are for calls and tail
> calls. In the tail call generation we can save one instruction by
> using the offset in jalr. Calls are optimized by doing (auipc)/jal(r)
> relative jumps instead of loading the entire absolute address and
> doing jalr. This required that the JIT image allocator was made RISC-V
> specific, so we can ensure that the JIT image and the kernel text are
> in range (32b).
> 
> The last two patches of the series is not critical to the series, but
> are two UAPI build issues for BPF events. A closer look from the
> RV-folks would be much appreciated.
> 
> The test_bpf.ko module, selftests/bpf/test_verifier and
> selftests/seccomp/seccomp_bpf pass all tests.
> 
> RISC-V is still missing proper kprobe and tracepoint support, so a lot
> of BPF selftests cannot be run.

Applied, thanks!