Re: [PATCH v12 4/5] LoongArch: Mark some assembler symbols as non-kprobe-able

From: Huacai Chen
Date: Fri Jan 20 2023 - 08:23:38 EST


On Thu, Jan 19, 2023 at 11:32 PM Masami Hiramatsu <mhiramat@xxxxxxxxxx> wrote:
>
> On Wed, 18 Jan 2023 15:17:00 +0800
> Huacai Chen <chenhuacai@xxxxxxxxxx> wrote:
>
> > On Wed, Jan 18, 2023 at 2:24 PM Tiezhu Yang <yangtiezhu@xxxxxxxxxxx> wrote:
> > >
> > >
> > >
> > > On 01/18/2023 02:05 PM, Jinyang He wrote:
> > > >
> > > > On 2023-01-18 12:23, Tiezhu Yang wrote:
> > > >>
> > > >>
> > > >> On 01/18/2023 12:14 PM, Huacai Chen wrote:
> > > >>> If memcpy should be blacklisted, then what about memset and memmove?
> > > >>
> > > >> According to the test results, there are no problems to probe
> > > >> memset and memmove, so no need to blacklist them for now,
> > > >> blacklist memcpy is because it may cause recursive exceptions,
> > > >> there is a detailed discussion in the following link:
> > > >>
> > > >> https://lore.kernel.org/lkml/20230114143859.7ccc45c1c5d9ce302113ab0a@xxxxxxxxxx/
> > > >>
> > > >
> > > > Hi, Tiezhu,
> > > >
> > > > I cannot reproduce the results when kprobe memcpy. Could you please give
> > > > some details. Emm, I just replace "kernel_clone" with "memcpy" in
> > > > kprobe_example.c.
> > >
> > > Please remove the related "_ASM_NOKPROBE(memcpy)" code in
> > > arch/loongarch/lib/memcpy.S, and then compile and update kernel,
> > > execute the following cmd after reboot, I can reproduce the hang
> > > problem easily (it will take a few minutes).
> > >
> > > modprobe kprobe_example symbol="memcpy"
> > Then, why is handle_syscall different from other exception handlers?
>
> I need to check the loongarch implementation of handle_syscall() but
> I guess in that handler the register set is not completely set as
> kernel one. In that case, the software breakpoint handler may not
> possible to handle it correctly. So it is better to avoid probing such
> "border" function by kprobes.
Seems reasonable, handle_syscall() indeed doesn't save all registers.
But for memcpy(), I still think memmove() and memset() may have the
same problem.

Huacai
>
> Thank you,
>
> >
> > Huacai
> > >
> > > >
> > > > And for your call trace,
> > > >
> > > > handler_pre()
> > > > pr_info()
> > > > printk()
> > > > _printk()
> > > > vprintk()
> > > > vprintk_store()
> > > > memcpy()
> > > >
> > > > I think when we should skip this time kprobe which triggered in
> > > > handler_{pre, post}. That means this time kprobe will not call
> > > > handler_{pre, post} agian, and not cause recursion. I remember
> > > > your codes had done this skip action. So, that's so strange if
> > > > recursion in handler_{pre, post}.
> > > >
> > > >
> > > > Thanks,
> > > >
> > > > Jinyang
> > > >
> > > >
> > > >>
> > > >> Thanks,
> > > >> Tiezhu
> > > >>
> > > >>>
> > > >>> Huacai
> > > >>>
> > > >>> On Wed, Jan 18, 2023 at 10:01 AM Tiezhu Yang <yangtiezhu@xxxxxxxxxxx>
> > > >>> wrote:
> > > >>>>
> > > >>>> Some assembler symbols are not kprobe safe, such as handle_syscall
> > > >>>> (used as syscall exception handler), *memcpy* (may cause recursive
> > > >>>> exceptions), they can not be instrumented, just blacklist them for
> > > >>>> kprobing.
> > > >>>>
> > > >>>> Here is a related problem and discussion:
> > > >>>> Link:
> > > >>>> https://lore.kernel.org/lkml/20230114143859.7ccc45c1c5d9ce302113ab0a@xxxxxxxxxx/
> > > >>>>
> > > >>>>
> > > >>>> Signed-off-by: Tiezhu Yang <yangtiezhu@xxxxxxxxxxx>
> > > >>>> ---
> > > >>>> arch/loongarch/include/asm/asm.h | 10 ++++++++++
> > > >>>> arch/loongarch/kernel/entry.S | 1 +
> > > >>>> arch/loongarch/lib/memcpy.S | 3 +++
> > > >>>> 3 files changed, 14 insertions(+)
> > > >>>>
> > > >>>> diff --git a/arch/loongarch/include/asm/asm.h
> > > >>>> b/arch/loongarch/include/asm/asm.h
> > > >>>> index 40eea6a..f591b32 100644
> > > >>>> --- a/arch/loongarch/include/asm/asm.h
> > > >>>> +++ b/arch/loongarch/include/asm/asm.h
> > > >>>> @@ -188,4 +188,14 @@
> > > >>>> #define PTRLOG 3
> > > >>>> #endif
> > > >>>>
> > > >>>> +/* Annotate a function as being unsuitable for kprobes. */
> > > >>>> +#ifdef CONFIG_KPROBES
> > > >>>> +#define _ASM_NOKPROBE(name) \
> > > >>>> + .pushsection "_kprobe_blacklist", "aw"; \
> > > >>>> + .quad name; \
> > > >>>> + .popsection
> > > >>>> +#else
> > > >>>> +#define _ASM_NOKPROBE(name)
> > > >>>> +#endif
> > > >>>> +
> > > >>>> #endif /* __ASM_ASM_H */
> > > >>>> diff --git a/arch/loongarch/kernel/entry.S
> > > >>>> b/arch/loongarch/kernel/entry.S
> > > >>>> index d53b631..55e23b1 100644
> > > >>>> --- a/arch/loongarch/kernel/entry.S
> > > >>>> +++ b/arch/loongarch/kernel/entry.S
> > > >>>> @@ -67,6 +67,7 @@ SYM_FUNC_START(handle_syscall)
> > > >>>>
> > > >>>> RESTORE_ALL_AND_RET
> > > >>>> SYM_FUNC_END(handle_syscall)
> > > >>>> +_ASM_NOKPROBE(handle_syscall)
> > > >>>>
> > > >>>> SYM_CODE_START(ret_from_fork)
> > > >>>> bl schedule_tail # a0 = struct task_struct *prev
> > > >>>> diff --git a/arch/loongarch/lib/memcpy.S b/arch/loongarch/lib/memcpy.S
> > > >>>> index 7c07d59..3b7e1de 100644
> > > >>>> --- a/arch/loongarch/lib/memcpy.S
> > > >>>> +++ b/arch/loongarch/lib/memcpy.S
> > > >>>> @@ -17,6 +17,7 @@ SYM_FUNC_START(memcpy)
> > > >>>> ALTERNATIVE "b __memcpy_generic", \
> > > >>>> "b __memcpy_fast", CPU_FEATURE_UAL
> > > >>>> SYM_FUNC_END(memcpy)
> > > >>>> +_ASM_NOKPROBE(memcpy)
> > > >>>>
> > > >>>> EXPORT_SYMBOL(memcpy)
> > > >>>>
> > > >>>> @@ -41,6 +42,7 @@ SYM_FUNC_START(__memcpy_generic)
> > > >>>> 2: move a0, a3
> > > >>>> jr ra
> > > >>>> SYM_FUNC_END(__memcpy_generic)
> > > >>>> +_ASM_NOKPROBE(__memcpy_generic)
> > > >>>>
> > > >>>> /*
> > > >>>> * void *__memcpy_fast(void *dst, const void *src, size_t n)
> > > >>>> @@ -93,3 +95,4 @@ SYM_FUNC_START(__memcpy_fast)
> > > >>>> 3: move a0, a3
> > > >>>> jr ra
> > > >>>> SYM_FUNC_END(__memcpy_fast)
> > > >>>> +_ASM_NOKPROBE(__memcpy_fast)
> > > >>>> --
> > > >>>> 2.1.0
> > > >>>>
> > > >>
> > >
> > >
>
>
> --
> Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>