mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Masami Hiramatsu (Google) <mhiramat@kernel.org>
To: Peter Zijlstra <peterz@infradead.org>
Cc: "Jiri Olsa" <jolsa@kernel.org>, "Oleg Nesterov" <oleg@redhat.com>,
	"Andrii Nakryiko" <andrii@kernel.org>,
	bpf@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-trace-kernel@vger.kernel.org, x86@kernel.org,
	"Song Liu" <songliubraving@fb.com>, "Yonghong Song" <yhs@fb.com>,
	"John Fastabend" <john.fastabend@gmail.com>,
	"Hao Luo" <haoluo@google.com>,
	"Steven Rostedt" <rostedt@goodmis.org>,
	"Alan Maguire" <alan.maguire@oracle.com>,
	"David Laight" <David.Laight@aculab.com>,
	"Thomas Weißschuh" <thomas@t-8ch.de>,
	"Ingo Molnar" <mingo@kernel.org>
Subject: Re: [PATCHv5 perf/core 09/22] uprobes/x86: Add uprobe syscall to speed up uprobe
Date: Mon, 14 Jul 2025 19:19:35 +0900	[thread overview]
Message-ID: <20250714191935.577ec7df5ae8a73282cddce7@kernel.org> (raw)
In-Reply-To: <20250714093903.GP905792@noisy.programming.kicks-ass.net>

On Mon, 14 Jul 2025 11:39:03 +0200
Peter Zijlstra <peterz@infradead.org> wrote:

> On Mon, Jul 14, 2025 at 05:39:15PM +0900, Masami Hiramatsu wrote:
> 
> > > +	/*
> > > +	 * Some of the uprobe consumers has changed sp, we can do nothing,
> > > +	 * just return via iret.
> > > +	 */
> > 
> > Do we allow consumers to change the `sp`? It seems dangerous
> > because consumer needs to know whether it is called from
> > breakpoint or syscall. Note that it has to set up ax, r11
> > and cx on the stack correctly only if it is called from syscall,
> > that is not compatible with breakpoint mode.
> > 
> > > +	if (regs->sp != sp)
> > > +		return regs->ax;
> > 
> > Shouldn't we recover regs->ip? Or in this case does consumer has
> > to change ip (== return address from trampline) too?
> > 
> > IMHO, it should not allow to change the `sp` and `ip` directly
> > in syscall mode. In case of kprobes, kprobe jump optimization
> > must be disabled explicitly (e.g. setting dummy post_handler)
> > if the handler changes `ip`.
> > 
> > Or, even if allowing to modify `sp` and `ip`, it should be helped
> > by this function, e.g. stack up the dummy regs->ax/r11/cx on the
> > new stack at the new `regs->sp`. This will allow modifying those
> > registries transparently as same as breakpoint mode.
> > In this case, I think we just need to remove above 2 lines.
> 
> There are two syscall return paths; the 'normal' is sysret and for that
> you need to undo all things just right.
> 
> The other is IRET. At which point we can have whatever state we want,
> including modified SP.
> 
> See arch/x86/entry/syscall_64.c:do_syscall_64() and
> arch/x86/entry/entry_64.S:entry_SYSCALL_64
> 
> The IRET path should return pt_regs as is from an interrupt/exception
> very much like INT3.

OK, so SYSRET case, we need to follow;

sys_uprobe -> do_syscall_64 -> entry_SYSCALL_64 -> trampoline -> retaddr

But using IRET to return, we can skip returning to trampoline,

sys_uprobe -> do_syscall_64 -> entry_SYSCALL_64 -> regs->ip

Thus we have to check the way, or in both cases use trampoline hack to
change return address.

Thank you,
-- 
Masami Hiramatsu (Google) <mhiramat@kernel.org>

  reply	other threads:[~2025-07-14 10:19 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-11  8:29 [PATCHv5 perf/core 00/22] uprobes: Add support to optimize usdt probes on x86_64 Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 01/22] uprobes: Remove breakpoint in unapply_uprobe under mmap_write_lock Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 02/22] uprobes: Rename arch_uretprobe_trampoline function Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 03/22] uprobes: Make copy_from_page global Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 04/22] uprobes: Add uprobe_write function Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 05/22] uprobes: Add nbytes argument to uprobe_write Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 06/22] uprobes: Add is_register argument to uprobe_write and uprobe_write_opcode Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 07/22] uprobes: Add do_ref_ctr argument to uprobe_write function Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 08/22] uprobes/x86: Add mapping for optimized uprobe trampolines Jiri Olsa
2025-07-11 17:46   ` Oleg Nesterov
2025-07-11 19:36     ` Jiri Olsa
2025-07-14  7:23   ` Masami Hiramatsu
2025-07-11  8:29 ` [PATCHv5 perf/core 09/22] uprobes/x86: Add uprobe syscall to speed up uprobe Jiri Olsa
2025-07-14  8:39   ` Masami Hiramatsu
2025-07-14  9:28     ` Peter Zijlstra
2025-07-14 21:29       ` Jiri Olsa
2025-07-14  9:39     ` Peter Zijlstra
2025-07-14 10:19       ` Masami Hiramatsu [this message]
2025-07-14 21:28         ` Jiri Olsa
2025-07-14 23:54           ` Masami Hiramatsu
2025-07-15 12:16             ` Jiri Olsa
2025-07-16  2:39               ` Masami Hiramatsu
2025-07-11  8:29 ` [PATCHv5 perf/core 10/22] uprobes/x86: Add support to optimize uprobes Jiri Olsa
2025-07-14  9:48   ` Peter Zijlstra
2025-07-14 21:29     ` Jiri Olsa
2025-07-17 15:29       ` Jiri Olsa
2025-07-14 10:13   ` Masami Hiramatsu
2025-07-14 21:29     ` Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 11/22] selftests/bpf: Import usdt.h from libbpf/usdt project Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 12/22] selftests/bpf: Reorg the uprobe_syscall test function Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 13/22] selftests/bpf: Rename uprobe_syscall_executed prog to test_uretprobe_multi Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 14/22] selftests/bpf: Add uprobe/usdt syscall tests Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 15/22] selftests/bpf: Add hit/attach/detach race optimized uprobe test Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 16/22] selftests/bpf: Add uprobe syscall sigill signal test Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 17/22] selftests/bpf: Add optimized usdt variant for basic usdt test Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 18/22] selftests/bpf: Add uprobe_regs_equal test Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 19/22] selftests/bpf: Change test_uretprobe_regs_change for uprobe and uretprobe Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 20/22] seccomp: passthrough uprobe systemcall without filtering Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 perf/core 21/22] selftests/seccomp: validate uprobe syscall passes through seccomp Jiri Olsa
2025-07-11  8:29 ` [PATCHv5 22/22] man2: Add uprobe syscall page Jiri Olsa
2025-07-14 14:04   ` Masami Hiramatsu
2025-07-11 17:17 ` [PATCHv5 perf/core 00/22] uprobes: Add support to optimize usdt probes on x86_64 Andrii Nakryiko

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=20250714191935.577ec7df5ae8a73282cddce7@kernel.org \
    --to=mhiramat@kernel.org \
    --cc=David.Laight@aculab.com \
    --cc=alan.maguire@oracle.com \
    --cc=andrii@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=haoluo@google.com \
    --cc=john.fastabend@gmail.com \
    --cc=jolsa@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-trace-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=oleg@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=songliubraving@fb.com \
    --cc=thomas@t-8ch.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®