[bpf-next,v4,0/3] bpf, x86: allow function arguments up to 12 for TRACING

  From: Menglong Dong <imagedong@tencent.com>

For now, the BPF program of type BPF_PROG_TYPE_TRACING can only be used
on the kernel functions whose arguments count less than 6. This is not
friendly at all, as too many functions have arguments count more than 6.
According to the current kernel version, below is a statistics of the
function arguments count:

argument count | function count
7              | 704
8              | 270
9              | 84
10             | 47
11             | 47
12             | 27
13             | 22
14             | 5
15             | 0
16             | 1

Therefore, let's enhance it by increasing the function arguments count
allowed in arch_prepare_bpf_trampoline(), for now, only x86_64.

In the 1st patch, we make arch_prepare_bpf_trampoline() support to copy
function arguments in stack for x86 arch. Therefore, the maximum
arguments can be up to MAX_BPF_FUNC_ARGS for FENTRY and FEXIT.

In the 2nd patch, we clean garbage value in upper bytes of the trampoline
when we store the arguments from regs or on-stack into stack.

And the 3rd patches are for the testcases of the 1st patch.

Changes since v3:
- try make the stack pointer 16-byte aligned. Not sure if I'm right :)
- introduce clean_garbage() to clean the grabage when argument count is 7
- use different data type in bpf_testmod_fentry_test{7,12}
- add testcase for grabage values in ctx

Changes since v2:
- keep MAX_BPF_FUNC_ARGS still
- clean garbage value in upper bytes in the 2nd patch
- move bpf_fentry_test{7,12} to bpf_testmod.c and rename them to
  bpf_testmod_fentry_test{7,12} meanwhile in the 3rd patch

Changes since v1:
- change the maximun function arguments to 14 from 12
- add testcases (Jiri Olsa)
- instead EMIT4 with EMIT3_off32 for "lea" to prevent overflow

Menglong Dong (3):
  bpf, x86: allow function arguments up to 12 for TRACING
  bpf, x86: clean garbage value in the stack of trampoline
  selftests/bpf: add testcase for FENTRY/FEXIT with 6+ arguments

 arch/x86/net/bpf_jit_comp.c                   | 145 ++++++++++++++++--
 .../selftests/bpf/bpf_testmod/bpf_testmod.c   |  19 ++-
 .../selftests/bpf/prog_tests/fentry_fexit.c   |   4 +-
 .../selftests/bpf/prog_tests/fentry_test.c    |   2 +
 .../selftests/bpf/prog_tests/fexit_test.c     |   2 +
 .../testing/selftests/bpf/progs/fentry_test.c |  33 ++++
 .../testing/selftests/bpf/progs/fexit_test.c  |  57 +++++++
 7 files changed, 245 insertions(+), 17 deletions(-)
  

On 6/9/23 2:56 AM, menglong8.dong@gmail.com wrote:
> From: Menglong Dong <imagedong@tencent.com>
> 
> For now, the BPF program of type BPF_PROG_TYPE_TRACING can only be used
> on the kernel functions whose arguments count less than 6. This is not
> friendly at all, as too many functions have arguments count more than 6.
> According to the current kernel version, below is a statistics of the
> function arguments count:
> 
> argument count | function count
> 7              | 704
> 8              | 270
> 9              | 84
> 10             | 47
> 11             | 47
> 12             | 27
> 13             | 22
> 14             | 5
> 15             | 0
> 16             | 1
> 
> Therefore, let's enhance it by increasing the function arguments count
> allowed in arch_prepare_bpf_trampoline(), for now, only x86_64.
> 
> In the 1st patch, we make arch_prepare_bpf_trampoline() support to copy
> function arguments in stack for x86 arch. Therefore, the maximum
> arguments can be up to MAX_BPF_FUNC_ARGS for FENTRY and FEXIT.
> 
> In the 2nd patch, we clean garbage value in upper bytes of the trampoline
> when we store the arguments from regs or on-stack into stack.
> 
> And the 3rd patches are for the testcases of the 1st patch.
> 
> Changes since v3:
> - try make the stack pointer 16-byte aligned. Not sure if I'm right :)
> - introduce clean_garbage() to clean the grabage when argument count is 7
> - use different data type in bpf_testmod_fentry_test{7,12}
> - add testcase for grabage values in ctx
> 
> Changes since v2:
> - keep MAX_BPF_FUNC_ARGS still
> - clean garbage value in upper bytes in the 2nd patch
> - move bpf_fentry_test{7,12} to bpf_testmod.c and rename them to
>    bpf_testmod_fentry_test{7,12} meanwhile in the 3rd patch
> 
> Changes since v1:
> - change the maximun function arguments to 14 from 12
> - add testcases (Jiri Olsa)
> - instead EMIT4 with EMIT3_off32 for "lea" to prevent overflow
> 
> Menglong Dong (3):
>    bpf, x86: allow function arguments up to 12 for TRACING
>    bpf, x86: clean garbage value in the stack of trampoline
>    selftests/bpf: add testcase for FENTRY/FEXIT with 6+ arguments
> 
>   arch/x86/net/bpf_jit_comp.c                   | 145 ++++++++++++++++--
>   .../selftests/bpf/bpf_testmod/bpf_testmod.c   |  19 ++-
>   .../selftests/bpf/prog_tests/fentry_fexit.c   |   4 +-
>   .../selftests/bpf/prog_tests/fentry_test.c    |   2 +
>   .../selftests/bpf/prog_tests/fexit_test.c     |   2 +
>   .../testing/selftests/bpf/progs/fentry_test.c |  33 ++++
>   .../testing/selftests/bpf/progs/fexit_test.c  |  57 +++++++
>   7 files changed, 245 insertions(+), 17 deletions(-)

Also replace rebase on top of bpf-next. Patch 3 cannot be applied.
  

On Sat, Jun 10, 2023 at 11:51 AM Yonghong Song <yhs@meta.com> wrote:
>
>
>
> On 6/9/23 2:56 AM, menglong8.dong@gmail.com wrote:
> > From: Menglong Dong <imagedong@tencent.com>
> >
> > For now, the BPF program of type BPF_PROG_TYPE_TRACING can only be used
> > on the kernel functions whose arguments count less than 6. This is not
> > friendly at all, as too many functions have arguments count more than 6.
> > According to the current kernel version, below is a statistics of the
> > function arguments count:
> >
> > argument count | function count
> > 7              | 704
> > 8              | 270
> > 9              | 84
> > 10             | 47
> > 11             | 47
> > 12             | 27
> > 13             | 22
> > 14             | 5
> > 15             | 0
> > 16             | 1
> >
> > Therefore, let's enhance it by increasing the function arguments count
> > allowed in arch_prepare_bpf_trampoline(), for now, only x86_64.
> >
> > In the 1st patch, we make arch_prepare_bpf_trampoline() support to copy
> > function arguments in stack for x86 arch. Therefore, the maximum
> > arguments can be up to MAX_BPF_FUNC_ARGS for FENTRY and FEXIT.
> >
> > In the 2nd patch, we clean garbage value in upper bytes of the trampoline
> > when we store the arguments from regs or on-stack into stack.
> >
> > And the 3rd patches are for the testcases of the 1st patch.
> >
> > Changes since v3:
> > - try make the stack pointer 16-byte aligned. Not sure if I'm right :)
> > - introduce clean_garbage() to clean the grabage when argument count is 7
> > - use different data type in bpf_testmod_fentry_test{7,12}
> > - add testcase for grabage values in ctx
> >
> > Changes since v2:
> > - keep MAX_BPF_FUNC_ARGS still
> > - clean garbage value in upper bytes in the 2nd patch
> > - move bpf_fentry_test{7,12} to bpf_testmod.c and rename them to
> >    bpf_testmod_fentry_test{7,12} meanwhile in the 3rd patch
> >
> > Changes since v1:
> > - change the maximun function arguments to 14 from 12
> > - add testcases (Jiri Olsa)
> > - instead EMIT4 with EMIT3_off32 for "lea" to prevent overflow
> >
> > Menglong Dong (3):
> >    bpf, x86: allow function arguments up to 12 for TRACING
> >    bpf, x86: clean garbage value in the stack of trampoline
> >    selftests/bpf: add testcase for FENTRY/FEXIT with 6+ arguments
> >
> >   arch/x86/net/bpf_jit_comp.c                   | 145 ++++++++++++++++--
> >   .../selftests/bpf/bpf_testmod/bpf_testmod.c   |  19 ++-
> >   .../selftests/bpf/prog_tests/fentry_fexit.c   |   4 +-
> >   .../selftests/bpf/prog_tests/fentry_test.c    |   2 +
> >   .../selftests/bpf/prog_tests/fexit_test.c     |   2 +
> >   .../testing/selftests/bpf/progs/fentry_test.c |  33 ++++
> >   .../testing/selftests/bpf/progs/fexit_test.c  |  57 +++++++
> >   7 files changed, 245 insertions(+), 17 deletions(-)
>
> Also replace rebase on top of bpf-next. Patch 3 cannot be applied.

Okay!