Re: [tip:x86/urgent] bpf: Fix ORC unwinding in non-JIT BPF code

Re: [tip:x86/urgent] bpf: Fix ORC unwinding in non-JIT BPF code

[Alexei Starovoitov]

On Tue, Jul 9, 2019 at 10:48 AM Josh Poimboeuf <jpoimboe@redhat.com> wrote:
>
> On Mon, Jul 08, 2019 at 04:16:25PM -0700, Alexei Starovoitov wrote:
> > total time is hard to compare.
> > Could you compare few tests?
> > like two that are called "tcpdump *"
> >
> > I think small regression is ok.
> > Folks that care about performance should be using JIT.
>
> I did each test 20 times and computed the averages:
>
> "tcpdump port 22":
>  default:       0.00743175s
>  -fno-gcse:     0.00709920s (~4.5% speedup)
>
> "tcpdump complex":
>  default:       0.00876715s
>  -fno-gcse:     0.00854895s (~2.5% speedup)
>
> So there does seem to be a small performance gain by disabling this
> optimization.

great. thanks for checking.

> We could change it for the whole file, by adjusting CFLAGS_core.o in the
> BPF makefile, or we could change it for the function only with something
> like the below patch.
>
> Thoughts?
>
> diff --git a/include/linux/compiler-gcc.h b/include/linux/compiler-gcc.h
> index e8579412ad21..d7ee4c6bad48 100644
> --- a/include/linux/compiler-gcc.h
> +++ b/include/linux/compiler-gcc.h
> @@ -170,3 +170,5 @@
>  #else
>  #define __diag_GCC_8(s)
>  #endif
> +
> +#define __no_fgcse __attribute__((optimize("-fno-gcse")))
> diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h
> index 095d55c3834d..599c27b56c29 100644
> --- a/include/linux/compiler_types.h
> +++ b/include/linux/compiler_types.h
> @@ -189,6 +189,10 @@ struct ftrace_likely_data {
>  #define asm_volatile_goto(x...) asm goto(x)
>  #endif
>
> +#ifndef __no_fgcse
> +# define __no_fgcse
> +#endif
> +
>  /* Are two types/vars the same type (ignoring qualifiers)? */
>  #define __same_type(a, b) __builtin_types_compatible_p(typeof(a), typeof(b))
>
> diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
> index 7e98f36a14e2..8191a7db2777 100644
> --- a/kernel/bpf/core.c
> +++ b/kernel/bpf/core.c
> @@ -1295,7 +1295,7 @@ bool bpf_opcode_in_insntable(u8 code)
>   *
>   * Decode and execute eBPF instructions.
>   */
> -static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack)
> +static u64 __no_fgcse ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack)

I prefer per-function flag.
If you want to route it via tip:
Acked-by: Alexei Starovoitov <ast@kerrnel.org>

or Daniel can take it into bpf tree while I'm traveling.

Re: [tip:x86/urgent] bpf: Fix ORC unwinding in non-JIT BPF code

[Josh Poimboeuf]

On Tue, Jul 09, 2019 at 11:02:40AM -0700, Alexei Starovoitov wrote:
> On Tue, Jul 9, 2019 at 10:48 AM Josh Poimboeuf <jpoimboe@redhat.com> wrote:
> >
> > On Mon, Jul 08, 2019 at 04:16:25PM -0700, Alexei Starovoitov wrote:
> > > total time is hard to compare.
> > > Could you compare few tests?
> > > like two that are called "tcpdump *"
> > >
> > > I think small regression is ok.
> > > Folks that care about performance should be using JIT.
> >
> > I did each test 20 times and computed the averages:
> >
> > "tcpdump port 22":
> >  default:       0.00743175s
> >  -fno-gcse:     0.00709920s (~4.5% speedup)
> >
> > "tcpdump complex":
> >  default:       0.00876715s
> >  -fno-gcse:     0.00854895s (~2.5% speedup)
> >
> > So there does seem to be a small performance gain by disabling this
> > optimization.
> 
> great. thanks for checking.
> 
> > We could change it for the whole file, by adjusting CFLAGS_core.o in the
> > BPF makefile, or we could change it for the function only with something
> > like the below patch.
> >
> > Thoughts?
> >
> > diff --git a/include/linux/compiler-gcc.h b/include/linux/compiler-gcc.h
> > index e8579412ad21..d7ee4c6bad48 100644
> > --- a/include/linux/compiler-gcc.h
> > +++ b/include/linux/compiler-gcc.h
> > @@ -170,3 +170,5 @@
> >  #else
> >  #define __diag_GCC_8(s)
> >  #endif
> > +
> > +#define __no_fgcse __attribute__((optimize("-fno-gcse")))
> > diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h
> > index 095d55c3834d..599c27b56c29 100644
> > --- a/include/linux/compiler_types.h
> > +++ b/include/linux/compiler_types.h
> > @@ -189,6 +189,10 @@ struct ftrace_likely_data {
> >  #define asm_volatile_goto(x...) asm goto(x)
> >  #endif
> >
> > +#ifndef __no_fgcse
> > +# define __no_fgcse
> > +#endif
> > +
> >  /* Are two types/vars the same type (ignoring qualifiers)? */
> >  #define __same_type(a, b) __builtin_types_compatible_p(typeof(a), typeof(b))
> >
> > diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
> > index 7e98f36a14e2..8191a7db2777 100644
> > --- a/kernel/bpf/core.c
> > +++ b/kernel/bpf/core.c
> > @@ -1295,7 +1295,7 @@ bool bpf_opcode_in_insntable(u8 code)
> >   *
> >   * Decode and execute eBPF instructions.
> >   */
> > -static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack)
> > +static u64 __no_fgcse ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack)
> 
> I prefer per-function flag.
> If you want to route it via tip:
> Acked-by: Alexei Starovoitov <ast@kerrnel.org>
> 
> or Daniel can take it into bpf tree while I'm traveling.

Thanks!  I''ll probably send it through the tip tree, along with an
objtool fix for the other optimization.

-- 
Josh

Re: [tip:x86/urgent] bpf: Fix ORC unwinding in non-JIT BPF code

[Daniel Borkmann]

On 07/09/2019 09:17 PM, Josh Poimboeuf wrote:
> On Tue, Jul 09, 2019 at 11:02:40AM -0700, Alexei Starovoitov wrote:
>> On Tue, Jul 9, 2019 at 10:48 AM Josh Poimboeuf <jpoimboe@redhat.com> wrote:
>>>
>>> On Mon, Jul 08, 2019 at 04:16:25PM -0700, Alexei Starovoitov wrote:
>>>> total time is hard to compare.
>>>> Could you compare few tests?
>>>> like two that are called "tcpdump *"
>>>>
>>>> I think small regression is ok.
>>>> Folks that care about performance should be using JIT.
>>>
>>> I did each test 20 times and computed the averages:
>>>
>>> "tcpdump port 22":
>>>  default:       0.00743175s
>>>  -fno-gcse:     0.00709920s (~4.5% speedup)
>>>
>>> "tcpdump complex":
>>>  default:       0.00876715s
>>>  -fno-gcse:     0.00854895s (~2.5% speedup)
>>>
>>> So there does seem to be a small performance gain by disabling this
>>> optimization.
>>
>> great. thanks for checking.
>>
>>> We could change it for the whole file, by adjusting CFLAGS_core.o in the
>>> BPF makefile, or we could change it for the function only with something
>>> like the below patch.
>>>
>>> Thoughts?
>>>
>>> diff --git a/include/linux/compiler-gcc.h b/include/linux/compiler-gcc.h
>>> index e8579412ad21..d7ee4c6bad48 100644
>>> --- a/include/linux/compiler-gcc.h
>>> +++ b/include/linux/compiler-gcc.h
>>> @@ -170,3 +170,5 @@
>>>  #else
>>>  #define __diag_GCC_8(s)
>>>  #endif
>>> +
>>> +#define __no_fgcse __attribute__((optimize("-fno-gcse")))
>>> diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h
>>> index 095d55c3834d..599c27b56c29 100644
>>> --- a/include/linux/compiler_types.h
>>> +++ b/include/linux/compiler_types.h
>>> @@ -189,6 +189,10 @@ struct ftrace_likely_data {
>>>  #define asm_volatile_goto(x...) asm goto(x)
>>>  #endif
>>>
>>> +#ifndef __no_fgcse
>>> +# define __no_fgcse
>>> +#endif
>>> +
>>>  /* Are two types/vars the same type (ignoring qualifiers)? */
>>>  #define __same_type(a, b) __builtin_types_compatible_p(typeof(a), typeof(b))
>>>
>>> diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
>>> index 7e98f36a14e2..8191a7db2777 100644
>>> --- a/kernel/bpf/core.c
>>> +++ b/kernel/bpf/core.c
>>> @@ -1295,7 +1295,7 @@ bool bpf_opcode_in_insntable(u8 code)
>>>   *
>>>   * Decode and execute eBPF instructions.
>>>   */
>>> -static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack)
>>> +static u64 __no_fgcse ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn, u64 *stack)
>>
>> I prefer per-function flag.

Same preference from my side.

>> If you want to route it via tip:
>> Acked-by: Alexei Starovoitov <ast@kerrnel.org>
>>
>> or Daniel can take it into bpf tree while I'm traveling.
> 
> Thanks!  I''ll probably send it through the tip tree, along with an
> objtool fix for the other optimization.

Ok, sounds good, thanks!