Re: [PATCH V9 3/4] riscv: ftrace: Add DYNAMIC_FTRACE_WITH_DIRECT_CALLS support

[Evgenii Shatokhin]

Hi,

On 11.05.2023 10:19, Björn Töpel wrote:
> Song Shuai <suagrfillet@xxxxxxxxx> writes:
>
> > This patch adds DYNAMIC_FTRACE_WITH_DIRECT_CALLS support for RISC-V.
> >
> > select the DYNAMIC_FTRACE_WITH_DIRECT_CALLS to provide the
> > register_ftrace_direct[_multi] interfaces allowing users to register
> > the customed trampoline (direct_caller) as the mcount for one or
> > more target functions. And modify_ftrace_direct[_multi] are also
> > provided for modifying direct_caller.
> >
> > To make the direct_caller and the other ftrace hooks (eg. function/fgraph
> > tracer, k[ret]probes) co-exist, a temporary register is nominated to
> > store the address of direct_caller in ftrace_regs_caller. After the
> > setting of the address direct_caller by direct_ops->func and the
> > RESTORE_REGS in ftrace_regs_caller, direct_caller will be jumped to
> > by the `jr` inst.
> >
> > Signed-off-by: Song Shuai <suagrfillet@xxxxxxxxx>
> > Tested-by: Guo Ren <guoren@xxxxxxxxxx>
> > Signed-off-by: Guo Ren <guoren@xxxxxxxxxx>
> > ---
> >   arch/riscv/Kconfig              |  1 +
> >   arch/riscv/include/asm/ftrace.h |  8 ++++++++
> >   arch/riscv/kernel/mcount-dyn.S  | 10 ++++++++++
> >   3 files changed, 19 insertions(+)
> >
> > diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
> > index e0632493482f..fdf0b219a02c 100644
> > --- a/arch/riscv/Kconfig
> > +++ b/arch/riscv/Kconfig
> > @@ -144,6 +144,7 @@ config RISCV
> >        select UACCESS_MEMCPY if !MMU
> >        select ZONE_DMA32 if 64BIT
> >        select HAVE_DYNAMIC_FTRACE if !XIP_KERNEL && MMU && (CLANG_SUPPORTS_DYNAMIC_FTRACE || GCC_SUPPORTS_DYNAMIC_FTRACE)
> > +     select HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
> >        select HAVE_DYNAMIC_FTRACE_WITH_REGS if HAVE_DYNAMIC_FTRACE
> >        select HAVE_FTRACE_MCOUNT_RECORD if !XIP_KERNEL
> >        select HAVE_FUNCTION_GRAPH_TRACER
> > diff --git a/arch/riscv/include/asm/ftrace.h b/arch/riscv/include/asm/ftrace.h
> > index 84f856a3286e..84904c1e4369 100644
> > --- a/arch/riscv/include/asm/ftrace.h
> > +++ b/arch/riscv/include/asm/ftrace.h
> > @@ -114,6 +114,14 @@ struct ftrace_regs;
> >   void ftrace_graph_func(unsigned long ip, unsigned long parent_ip,
> >                       struct ftrace_ops *op, struct ftrace_regs *fregs);
> >   #define ftrace_graph_func ftrace_graph_func
> > +
> > +static inline void
> > +__arch_ftrace_set_direct_caller(struct pt_regs *regs, unsigned long addr)
> > +{
> > +             regs->t1 = addr;
> > +}
> > +#define arch_ftrace_set_direct_caller(fregs, addr) \
> > +     __arch_ftrace_set_direct_caller(&(fregs)->regs, addr)
> >   #endif /* CONFIG_DYNAMIC_FTRACE_WITH_REGS */
> >
> >   #endif /* __ASSEMBLY__ */
> > diff --git a/arch/riscv/kernel/mcount-dyn.S b/arch/riscv/kernel/mcount-dyn.S
> > index f26e9f6e2fed..9d405baedb52 100644
> > --- a/arch/riscv/kernel/mcount-dyn.S
> > +++ b/arch/riscv/kernel/mcount-dyn.S
> > @@ -231,6 +231,7 @@ ENDPROC(ftrace_caller)
> >
> >   #else /* CONFIG_DYNAMIC_FTRACE_WITH_REGS */
> >   ENTRY(ftrace_regs_caller)
> > +     move    t1, zero
>
> Please use "mv", and not "move" [1].
>
> >        SAVE_ABI_REGS 1
> >        PREPARE_ARGS
> >
> > @@ -239,7 +240,10 @@ ftrace_regs_call:
> >        call    ftrace_stub
> >
> >        RESTORE_ABI_REGS 1
> > +     bnez    t1,.Ldirect
> >        jr t0
> > +.Ldirect:
> > +     jr t1
>
> Again, while you're doing changes here, please try to align op/operands.
>
> Wearing my BPF hat, I'm happy to finally get DIRECT_CALLS support!
>
> This does not take the WITH_CALL_OPS approach Mark suggested in the v7
> threads, but given that text patching story on RISC-V is still a bit sad
> (inconsistency in the RV tree, no specification, cannot work with
> preempt, ...) I'd say this approach is OK for now, and we can change to
> WITH_CALL_OPS later in a wider "let's improve RISC-V textpatching" work.
>
> Thoughts?

The WITH_CALL_OPS approach seems to need much more time and effort, so, 
yes, I'd also use this implementation of DIRECT_CALLS for now. Other 
improvements could wait for the future "big patching rework".

FWIW, the implementation of kprobes for RISC-V has been improving 
gradually too, not everything was done in the first very patchset, but 
it was usable nonetheless.

I have not tested this particular version of this 
DYNAMIC_FTRACE_WITH_DIRECT_CALLS series, only some previous ones - there 
were no outstanding problems there. The code looks sane to me.

> Björn
>
> [1] https://github.com/riscv-non-isa/riscv-asm-manual/blob/master/riscv-asm.md#-a-listing-of-standard-risc-v-pseudoinstructions

Regards,
Evgenii