From: Ingo Molnar <mingo@kernel.org>
To: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>,
Josh Poimboeuf <jpoimboe@redhat.com>, X86 ML <x86@kernel.org>,
Daniel Xu <dxu@dxuuu.xyz>,
linux-kernel@vger.kernel.org, bpf@vger.kernel.org,
kuba@kernel.org, mingo@redhat.com, ast@kernel.org,
Thomas Gleixner <tglx@linutronix.de>,
Borislav Petkov <bp@alien8.de>,
Peter Zijlstra <peterz@infradead.org>,
kernel-team@fb.com, yhs@fb.com, linux-ia64@vger.kernel.org,
Abhishek Sagar <sagar.abhishek@gmail.com>,
Andrii Nakryiko <andrii.nakryiko@gmail.com>
Subject: Re: [PATCH -tip v8 05/13] x86/kprobes: Add UNWIND_HINT_FUNC on kretprobe_trampoline code
Date: Mon, 5 Jul 2021 10:02:47 +0200 [thread overview]
Message-ID: <YOK8pzp8B2V+1EaU@gmail.com> (raw)
In-Reply-To: <162399996966.506599.810050095040575221.stgit@devnote2>
* Masami Hiramatsu <mhiramat@kernel.org> wrote:
> From: Josh Poimboeuf <jpoimboe@redhat.com>
>
> Add UNWIND_HINT_FUNC on kretporbe_trampoline code so that ORC
> information is generated on the kretprobe_trampoline correctly.
What is a 'kretporbe'?
> Note that when the CONFIG_FRAME_POINTER=y, since the
> kretprobe_trampoline skips updating frame pointer, the stack frame
> of the kretprobe_trampoline seems non-standard. So this marks it
> is STACK_FRAME_NON_STANDARD() and undefine UNWIND_HINT_FUNC.
What does 'marks it is' mean?
'undefine' UNWIND_HINT_FUNC?
Doesn't the patch do the exact opposite:
> +#define UNWIND_HINT_FUNC \
> + UNWIND_HINT(ORC_REG_SP, 8, UNWIND_HINT_TYPE_FUNC, 0)
But it does undefine it in a specific spot:
> Anyway, with the frame pointer, FP unwinder can unwind the stack
> frame correctly without that hint.
>
> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
> Tested-by: Andrii Nakryik <andrii@kernel.org>
I have to say these changelogs are very careless.
> +#else
> +
In headers, in longer CPP blocks, please always mark the '#else' branch
with what it is the else branch of.
See the output of:
kepler:~/tip> git grep '#else' arch/x86/include/asm/ | head
> +#ifdef CONFIG_FRAME_POINTER
> +/*
> + * kretprobe_trampoline skips updating frame pointer. The frame pointer
> + * saved in trampoline_handler points to the real caller function's
> + * frame pointer. Thus the kretprobe_trampoline doesn't seems to have a
> + * standard stack frame with CONFIG_FRAME_POINTER=y.
> + * Let's mark it non-standard function. Anyway, FP unwinder can correctly
> + * unwind without the hint.
s/doesn't seems to have a standard stack frame
/doesn't have a standard stack frame
There's nothing 'seems' about the situation - it's a non-standard function
entry and stack frame situation, and the unwinder needs to know about it.
> +STACK_FRAME_NON_STANDARD(kretprobe_trampoline);
> +#undef UNWIND_HINT_FUNC
> +#define UNWIND_HINT_FUNC
> +#endif
> /*
> * When a retprobed function returns, this code saves registers and
> * calls trampoline_handler() runs, which calls the kretprobe's handler.
> @@ -1031,6 +1044,7 @@ asm(
> /* We don't bother saving the ss register */
> #ifdef CONFIG_X86_64
> " pushq %rsp\n"
> + UNWIND_HINT_FUNC
> " pushfq\n"
> SAVE_REGS_STRING
> " movq %rsp, %rdi\n"
> @@ -1041,6 +1055,7 @@ asm(
> " popfq\n"
> #else
> " pushl %esp\n"
> + UNWIND_HINT_FUNC
> " pushfl\n"
> SAVE_REGS_STRING
> " movl %esp, %eax\n"
Why not provide an appropriate annotation method in <asm/unwind_hints.h>,
so that other future code can use it too instead of reinventing the wheel?
Thanks,
Ingo
WARNING: multiple messages have this Message-ID (diff)
From: Ingo Molnar <mingo@kernel.org>
To: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>,
Josh Poimboeuf <jpoimboe@redhat.com>, X86 ML <x86@kernel.org>,
Daniel Xu <dxu@dxuuu.xyz>,
linux-kernel@vger.kernel.org, bpf@vger.kernel.org,
kuba@kernel.org, mingo@redhat.com, ast@kernel.org,
Thomas Gleixner <tglx@linutronix.de>,
Borislav Petkov <bp@alien8.de>,
Peter Zijlstra <peterz@infradead.org>,
kernel-team@fb.com, yhs@fb.com, linux-ia64@vger.kernel.org,
Abhishek Sagar <sagar.abhishek@gmail.com>,
Andrii Nakryiko <andrii.nakryiko@gmail.com>
Subject: Re: [PATCH -tip v8 05/13] x86/kprobes: Add UNWIND_HINT_FUNC on kretprobe_trampoline code
Date: Mon, 05 Jul 2021 08:02:47 +0000 [thread overview]
Message-ID: <YOK8pzp8B2V+1EaU@gmail.com> (raw)
In-Reply-To: <162399996966.506599.810050095040575221.stgit@devnote2>
* Masami Hiramatsu <mhiramat@kernel.org> wrote:
> From: Josh Poimboeuf <jpoimboe@redhat.com>
>
> Add UNWIND_HINT_FUNC on kretporbe_trampoline code so that ORC
> information is generated on the kretprobe_trampoline correctly.
What is a 'kretporbe'?
> Note that when the CONFIG_FRAME_POINTER=y, since the
> kretprobe_trampoline skips updating frame pointer, the stack frame
> of the kretprobe_trampoline seems non-standard. So this marks it
> is STACK_FRAME_NON_STANDARD() and undefine UNWIND_HINT_FUNC.
What does 'marks it is' mean?
'undefine' UNWIND_HINT_FUNC?
Doesn't the patch do the exact opposite:
> +#define UNWIND_HINT_FUNC \
> + UNWIND_HINT(ORC_REG_SP, 8, UNWIND_HINT_TYPE_FUNC, 0)
But it does undefine it in a specific spot:
> Anyway, with the frame pointer, FP unwinder can unwind the stack
> frame correctly without that hint.
>
> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
> Tested-by: Andrii Nakryik <andrii@kernel.org>
I have to say these changelogs are very careless.
> +#else
> +
In headers, in longer CPP blocks, please always mark the '#else' branch
with what it is the else branch of.
See the output of:
kepler:~/tip> git grep '#else' arch/x86/include/asm/ | head
> +#ifdef CONFIG_FRAME_POINTER
> +/*
> + * kretprobe_trampoline skips updating frame pointer. The frame pointer
> + * saved in trampoline_handler points to the real caller function's
> + * frame pointer. Thus the kretprobe_trampoline doesn't seems to have a
> + * standard stack frame with CONFIG_FRAME_POINTER=y.
> + * Let's mark it non-standard function. Anyway, FP unwinder can correctly
> + * unwind without the hint.
s/doesn't seems to have a standard stack frame
/doesn't have a standard stack frame
There's nothing 'seems' about the situation - it's a non-standard function
entry and stack frame situation, and the unwinder needs to know about it.
> +STACK_FRAME_NON_STANDARD(kretprobe_trampoline);
> +#undef UNWIND_HINT_FUNC
> +#define UNWIND_HINT_FUNC
> +#endif
> /*
> * When a retprobed function returns, this code saves registers and
> * calls trampoline_handler() runs, which calls the kretprobe's handler.
> @@ -1031,6 +1044,7 @@ asm(
> /* We don't bother saving the ss register */
> #ifdef CONFIG_X86_64
> " pushq %rsp\n"
> + UNWIND_HINT_FUNC
> " pushfq\n"
> SAVE_REGS_STRING
> " movq %rsp, %rdi\n"
> @@ -1041,6 +1055,7 @@ asm(
> " popfq\n"
> #else
> " pushl %esp\n"
> + UNWIND_HINT_FUNC
> " pushfl\n"
> SAVE_REGS_STRING
> " movl %esp, %eax\n"
Why not provide an appropriate annotation method in <asm/unwind_hints.h>,
so that other future code can use it too instead of reinventing the wheel?
Thanks,
Ingo
next prev parent reply other threads:[~2021-07-05 8:02 UTC|newest]
Thread overview: 114+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-18 7:05 [PATCH -tip v8 00/13] kprobes: Fix stacktrace with kretprobes on x86 Masami Hiramatsu
2021-06-18 7:05 ` Masami Hiramatsu
2021-06-18 7:05 ` [PATCH -tip v8 01/13] ia64: kprobes: Fix to pass correct trampoline address to the handler Masami Hiramatsu
2021-06-18 7:05 ` Masami Hiramatsu
2021-07-05 7:46 ` Ingo Molnar
2021-07-05 7:46 ` Ingo Molnar
2021-07-05 10:05 ` Masami Hiramatsu
2021-07-05 10:05 ` Masami Hiramatsu
2021-06-18 7:05 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbol_descriptor() Masami Hiramatsu
2021-06-18 7:05 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbol_de Masami Hiramatsu
2021-07-05 7:48 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbol_descriptor() Ingo Molnar
2021-07-05 7:48 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbo Ingo Molnar
2021-07-05 12:03 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbol_descriptor() Masami Hiramatsu
2021-07-05 12:03 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbo Masami Hiramatsu
2021-07-07 18:28 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbol_descriptor() Andrii Nakryiko
2021-07-07 18:28 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbo Andrii Nakryiko
2021-07-08 4:08 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbol_descriptor() Masami Hiramatsu
2021-07-08 4:08 ` [PATCH -tip v8 02/13] kprobes: treewide: Replace arch_deref_entry_point() with dereference_symbo Masami Hiramatsu
2021-06-18 7:05 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_handler() Masami Hiramatsu
2021-06-18 7:05 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_handler Masami Hiramatsu
2021-07-05 7:03 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_handler() Ingo Molnar
2021-07-05 7:03 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_han Ingo Molnar
2021-07-05 10:03 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_handler() Masami Hiramatsu
2021-07-05 10:03 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_han Masami Hiramatsu
2021-07-05 7:49 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_handler() Ingo Molnar
2021-07-05 7:49 ` [PATCH -tip v8 03/13] kprobes: treewide: Remove trampoline_address from kretprobe_trampoline_han Ingo Molnar
2021-06-18 7:05 ` [PATCH -tip v8 04/13] kprobes: Add kretprobe_find_ret_addr() for searching return address Masami Hiramatsu
2021-06-18 7:05 ` Masami Hiramatsu
2021-07-05 7:42 ` Ingo Molnar
2021-07-05 7:42 ` Ingo Molnar
2021-07-05 14:11 ` Masami Hiramatsu
2021-07-05 14:11 ` Masami Hiramatsu
2021-06-18 7:06 ` [PATCH -tip v8 05/13] x86/kprobes: Add UNWIND_HINT_FUNC on kretprobe_trampoline code Masami Hiramatsu
2021-06-18 7:06 ` Masami Hiramatsu
2021-07-05 8:02 ` Ingo Molnar [this message]
2021-07-05 8:02 ` Ingo Molnar
2021-07-09 15:31 ` Masami Hiramatsu
2021-07-09 15:31 ` Masami Hiramatsu
2021-07-10 1:41 ` Masami Hiramatsu
2021-07-10 1:41 ` Masami Hiramatsu
2021-07-10 19:01 ` Josh Poimboeuf
2021-07-10 19:01 ` Josh Poimboeuf
2021-07-10 19:24 ` [PATCH 1/2] objtool: Add frame-pointer-specific function ignore Josh Poimboeuf
2021-07-10 19:24 ` Josh Poimboeuf
2021-07-11 1:16 ` Masami Hiramatsu
2021-07-11 1:16 ` Masami Hiramatsu
2021-07-29 2:31 ` Masami Hiramatsu
2021-07-29 2:31 ` Masami Hiramatsu
2021-07-10 19:25 ` [PATCH 2/2] objtool: Ignore unwind hints for ignored functions Josh Poimboeuf
2021-07-10 19:25 ` Josh Poimboeuf
2021-07-11 2:07 ` Masami Hiramatsu
2021-07-11 2:07 ` Masami Hiramatsu
2021-06-18 7:06 ` [PATCH -tip v8 06/13] ARC: Add instruction_pointer_set() API Masami Hiramatsu
2021-06-18 7:06 ` Masami Hiramatsu
2021-06-18 7:06 ` [PATCH -tip v8 07/13] ia64: " Masami Hiramatsu
2021-06-18 7:06 ` Masami Hiramatsu
2021-06-18 7:06 ` [PATCH -tip v8 08/13] arm: kprobes: Make a space for regs->ARM_pc at kretprobe_trampoline Masami Hiramatsu
2021-06-18 7:06 ` Masami Hiramatsu
2021-07-05 8:04 ` Ingo Molnar
2021-07-05 8:04 ` Ingo Molnar
2021-07-05 14:40 ` Masami Hiramatsu
2021-07-05 14:40 ` Masami Hiramatsu
2021-06-18 7:06 ` [PATCH -tip v8 09/13] kprobes: Enable stacktrace from pt_regs in kretprobe handler Masami Hiramatsu
2021-06-18 7:06 ` Masami Hiramatsu
2021-06-18 14:04 ` Josh Poimboeuf
2021-06-18 14:04 ` Josh Poimboeuf
2021-06-18 7:06 ` [PATCH -tip v8 10/13] x86/kprobes: Push a fake return address at kretprobe_trampoline Masami Hiramatsu
2021-06-18 7:06 ` Masami Hiramatsu
2021-07-05 8:17 ` Ingo Molnar
2021-07-05 8:17 ` Ingo Molnar
2021-07-09 14:55 ` Masami Hiramatsu
2021-07-09 14:55 ` Masami Hiramatsu
2021-06-18 7:07 ` [PATCH -tip v8 11/13] x86/unwind: Recover kretprobe trampoline entry Masami Hiramatsu
2021-06-18 7:07 ` Masami Hiramatsu
2021-07-05 11:36 ` Peter Zijlstra
2021-07-05 11:36 ` Peter Zijlstra
2021-07-05 15:42 ` Masami Hiramatsu
2021-07-05 15:42 ` Masami Hiramatsu
2021-07-06 7:55 ` Peter Zijlstra
2021-07-06 7:55 ` Peter Zijlstra
2021-07-06 15:11 ` Steven Rostedt
2021-07-06 15:11 ` Steven Rostedt
2021-07-07 8:20 ` Peter Zijlstra
2021-07-07 8:20 ` Peter Zijlstra
2021-07-07 8:36 ` Peter Zijlstra
2021-07-07 8:36 ` Peter Zijlstra
2021-07-07 10:15 ` Masami Hiramatsu
2021-07-07 10:15 ` Masami Hiramatsu
2021-07-07 10:20 ` Peter Zijlstra
2021-07-07 10:20 ` Peter Zijlstra
2021-07-07 10:45 ` Masami Hiramatsu
2021-07-07 10:45 ` Masami Hiramatsu
2021-07-07 13:29 ` Masami Hiramatsu
2021-07-07 13:29 ` Masami Hiramatsu
2021-07-07 14:42 ` Matt Wu
2021-07-07 14:42 ` Matt Wu
2021-07-11 14:09 ` Masami Hiramatsu
2021-07-11 14:09 ` Masami Hiramatsu
2021-07-11 15:28 ` Matt Wu
2021-07-11 15:28 ` Matt Wu
2021-07-12 4:57 ` Masami Hiramatsu
2021-07-12 4:57 ` Masami Hiramatsu
2021-06-18 7:07 ` [PATCH -tip v8 12/13] tracing: Show kretprobe unknown indicator only for kretprobe_trampoline Masami Hiramatsu
2021-06-18 7:07 ` Masami Hiramatsu
2021-06-18 7:07 ` [PATCH -tip v8 13/13] x86/kprobes: Fixup return address in generic trampoline handler Masami Hiramatsu
2021-06-18 7:07 ` Masami Hiramatsu
2021-07-05 8:34 ` Ingo Molnar
2021-07-05 8:34 ` Ingo Molnar
2021-07-06 12:57 ` Masami Hiramatsu
2021-07-06 12:57 ` Masami Hiramatsu
2021-06-18 17:44 ` [PATCH -tip v8 00/13] kprobes: Fix stacktrace with kretprobes on x86 Andrii Nakryiko
2021-06-18 17:44 ` Andrii Nakryiko
2021-06-28 13:50 ` Masami Hiramatsu
2021-06-28 13:50 ` Masami Hiramatsu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=YOK8pzp8B2V+1EaU@gmail.com \
--to=mingo@kernel.org \
--cc=andrii.nakryiko@gmail.com \
--cc=ast@kernel.org \
--cc=bp@alien8.de \
--cc=bpf@vger.kernel.org \
--cc=dxu@dxuuu.xyz \
--cc=jpoimboe@redhat.com \
--cc=kernel-team@fb.com \
--cc=kuba@kernel.org \
--cc=linux-ia64@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mhiramat@kernel.org \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=sagar.abhishek@gmail.com \
--cc=tglx@linutronix.de \
--cc=x86@kernel.org \
--cc=yhs@fb.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.