Re: [PATCH v17 06/16] tracing: Add ftrace_partial_regs() for converting ftrace_regs to pt_regs

[Google]

On Mon, 21 Oct 2024 17:46:19 +0100
Will Deacon <will@xxxxxxxxxx> wrote:

> On Wed, Oct 16, 2024 at 09:59:00AM +0900, Masami Hiramatsu (Google) wrote:
> > From: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
> > 
> > Add ftrace_partial_regs() which converts the ftrace_regs to pt_regs.
> > This is for the eBPF which needs this to keep the same pt_regs interface
> > to access registers.
> > Thus when replacing the pt_regs with ftrace_regs in fprobes (which is
> > used by kprobe_multi eBPF event), this will be used.
> > 
> > If the architecture defines its own ftrace_regs, this copies partial
> > registers to pt_regs and returns it. If not, ftrace_regs is the same as
> > pt_regs and ftrace_partial_regs() will return ftrace_regs::regs.
> > 
> > Signed-off-by: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
> > Acked-by: Florent Revest <revest@xxxxxxxxxxxx>
> > Cc: Steven Rostedt <rostedt@xxxxxxxxxxx>
> > Cc: Mark Rutland <mark.rutland@xxxxxxx>
> > Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
> > Cc: Will Deacon <will@xxxxxxxxxx>
> > Cc: Paul Walmsley <paul.walmsley@xxxxxxxxxx>
> > Cc: Palmer Dabbelt <palmer@xxxxxxxxxxx>
> > Cc: Albert Ou <aou@xxxxxxxxxxxxxxxxx>
> > 
> > ---
> >  Changes in v14:
> >   - Add riscv change.
> >  Changes in v8:
> >   - Add the reason why this required in changelog.
> >  Changes from previous series: NOTHING, just forward ported.
> > ---
> >  arch/arm64/include/asm/ftrace.h |   11 +++++++++++
> >  arch/riscv/include/asm/ftrace.h |   14 ++++++++++++++
> >  include/linux/ftrace.h          |   17 +++++++++++++++++
> >  3 files changed, 42 insertions(+)
> > 
> > diff --git a/arch/arm64/include/asm/ftrace.h b/arch/arm64/include/asm/ftrace.h
> > index b5fa57b61378..d344c69eb01e 100644
> > --- a/arch/arm64/include/asm/ftrace.h
> > +++ b/arch/arm64/include/asm/ftrace.h
> > @@ -135,6 +135,17 @@ ftrace_regs_get_frame_pointer(const struct ftrace_regs *fregs)
> >  	return arch_ftrace_regs(fregs)->fp;
> >  }
> >  
> > +static __always_inline struct pt_regs *
> > +ftrace_partial_regs(const struct ftrace_regs *fregs, struct pt_regs *regs)
> > +{
> > +	memcpy(regs->regs, arch_ftrace_regs(fregs)->regs, sizeof(u64) * 9);
> 
> Since ftrace_regs::regs is an 'unsigned long regs[9]' can we just use
> sizeof() on that instead of hard-coding the length of the array here?

Good catch! I will update it.

Thank you!

> 
> Will


-- 
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>