Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1623411 > unrolled thread

[PATCH v3] ftrace/x86/32: fix triple fault with graph tracing and suspend-to-ram

Back to article view | Back to linux.kernel

Contents

  [PATCH v3] ftrace/x86/32: fix triple fault with graph tracing and suspend-to-ram Josh Poimboeuf <jpoimboe@redhat.com> - 2017-04-14 01:00 +0200
    Re: [PATCH v3] ftrace/x86/32: fix triple fault with graph tracing  and suspend-to-ram Steven Rostedt <rostedt@goodmis.org> - 2017-04-14 01:10 +0200
      Re: [PATCH v3] ftrace/x86/32: fix triple fault with graph tracing  and suspend-to-ram Josh Poimboeuf <jpoimboe@redhat.com> - 2017-04-14 04:20 +0200
    [tip:x86/urgent] ftrace/x86: Fix triple fault with graph tracing  and suspend-to-ram tip-bot for Josh Poimboeuf <tipbot@zytor.com> - 2017-04-14 12:00 +0200

#1623411 — [PATCH v3] ftrace/x86/32: fix triple fault with graph tracing and suspend-to-ram

On x86-32, with CONFIG_FIRMWARE and multiple CPUs, if you enable
function graph tracing and then suspend to RAM, it will triple fault and
reboot when it resumes.

The first fault happens when booting a secondary CPU:

startup_32_smp()
  load_ucode_ap()
    prepare_ftrace_return()
      ftrace_graph_is_dead()
        (accesses 'kill_ftrace_graph')

The early head_32.S code calls into load_ucode_ap(), which has an an
ftrace hook, so it calls prepare_ftrace_return(), which calls
ftrace_graph_is_dead(), which tries to access the global
'kill_ftrace_graph' variable with a virtual address, causing a fault
because the CPU is still in real mode.

The fix is to add a check in prepare_ftrace_return() to make sure it's
running in protected mode before continuing.  The check makes sure the
stack pointer is a virtual kernel address.  It's a bit of a hack, but
it's not very intrusive and it works well enough.

For reference, here are a few other (more difficult) ways this could
have potentially been fixed:

- Move startup_32_smp()'s call to load_ucode_ap() down to *after* paging
  is enabled.  (No idea what that would break.)

- Track down load_ucode_ap()'s entire callee tree and mark all the
  functions 'notrace'.  (Probably not realistic.)

- Pause graph tracing in ftrace_suspend_notifier_call() or bringup_cpu()
  or __cpu_up(), and ensure that the pause facility can be queried from
  real mode.

Reported-by: Paul Menzel <pmenzel@molgen.mpg.de>
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Cc: stable@kernel.org
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
---
v3: Added Tested-by: and Cc:stable tags

 arch/x86/kernel/ftrace.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
index 9dd546b..a3adf1f 100644
--- a/arch/x86/kernel/ftrace.c
+++ b/arch/x86/kernel/ftrace.c
@@ -995,6 +995,18 @@ void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent,
 	unsigned long return_hooker = (unsigned long)
 				&return_to_handler;
 
+	/*
+	 * When resuming from suspend-to-ram, this function can be indirectly
+	 * called from early CPU startup code while the CPU is in real mode,
+	 * which would fail miserably.  Make sure the stack pointer is a
+	 * virtual address.
+	 *
+	 * This check isn't as accurate as virt_addr_valid(), but it should be
+	 * good enough for this purpose, and it's fast.
+	 */
+	if (unlikely((long)__builtin_frame_address(0) >= 0))
+		return;
+
 	if (unlikely(ftrace_graph_is_dead()))
 		return;
 
-- 
2.7.4

[toc] | [next] | [standalone]

#1623414 — Re: [PATCH v3] ftrace/x86/32: fix triple fault with graph tracing and suspend-to-ram

On Thu, 13 Apr 2017 17:53:55 -0500
Josh Poimboeuf <jpoimboe@redhat.com> wrote:

> On x86-32, with CONFIG_FIRMWARE and multiple CPUs, if you enable
> function graph tracing and then suspend to RAM, it will triple fault and
> reboot when it resumes.
> 
> The first fault happens when booting a secondary CPU:
> 
> startup_32_smp()
>   load_ucode_ap()
>     prepare_ftrace_return()
>       ftrace_graph_is_dead()
>         (accesses 'kill_ftrace_graph')
> 
> The early head_32.S code calls into load_ucode_ap(), which has an an
> ftrace hook, so it calls prepare_ftrace_return(), which calls
> ftrace_graph_is_dead(), which tries to access the global
> 'kill_ftrace_graph' variable with a virtual address, causing a fault
> because the CPU is still in real mode.
> 
> The fix is to add a check in prepare_ftrace_return() to make sure it's
> running in protected mode before continuing.  The check makes sure the
> stack pointer is a virtual kernel address.  It's a bit of a hack, but
> it's not very intrusive and it works well enough.
> 
> For reference, here are a few other (more difficult) ways this could
> have potentially been fixed:
> 
> - Move startup_32_smp()'s call to load_ucode_ap() down to *after* paging
>   is enabled.  (No idea what that would break.)
> 
> - Track down load_ucode_ap()'s entire callee tree and mark all the
>   functions 'notrace'.  (Probably not realistic.)
> 
> - Pause graph tracing in ftrace_suspend_notifier_call() or bringup_cpu()
>   or __cpu_up(), and ensure that the pause facility can be queried from
>   real mode.
> 
> Reported-by: Paul Menzel <pmenzel@molgen.mpg.de>
> Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
> Cc: stable@kernel.org
> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>

This is pretty much the same thing we were talking about before, right?

If so, then:

Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>

if not, I need to spend more time reviewing it ;-)

-- Steve

> ---
> v3: Added Tested-by: and Cc:stable tags
> 
>  arch/x86/kernel/ftrace.c | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
> index 9dd546b..a3adf1f 100644
> --- a/arch/x86/kernel/ftrace.c
> +++ b/arch/x86/kernel/ftrace.c
> @@ -995,6 +995,18 @@ void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent,
>  	unsigned long return_hooker = (unsigned long)
>  				&return_to_handler;
>  
> +	/*
> +	 * When resuming from suspend-to-ram, this function can be indirectly
> +	 * called from early CPU startup code while the CPU is in real mode,
> +	 * which would fail miserably.  Make sure the stack pointer is a
> +	 * virtual address.
> +	 *
> +	 * This check isn't as accurate as virt_addr_valid(), but it should be
> +	 * good enough for this purpose, and it's fast.
> +	 */
> +	if (unlikely((long)__builtin_frame_address(0) >= 0))
> +		return;
> +
>  	if (unlikely(ftrace_graph_is_dead()))
>  		return;
>  

[toc] | [prev] | [next] | [standalone]

#1623477 — Re: [PATCH v3] ftrace/x86/32: fix triple fault with graph tracing and suspend-to-ram

On Thu, Apr 13, 2017 at 07:02:36PM -0400, Steven Rostedt wrote:
> On Thu, 13 Apr 2017 17:53:55 -0500
> Josh Poimboeuf <jpoimboe@redhat.com> wrote:
> 
> > On x86-32, with CONFIG_FIRMWARE and multiple CPUs, if you enable
> > function graph tracing and then suspend to RAM, it will triple fault and
> > reboot when it resumes.
> > 
> > The first fault happens when booting a secondary CPU:
> > 
> > startup_32_smp()
> >   load_ucode_ap()
> >     prepare_ftrace_return()
> >       ftrace_graph_is_dead()
> >         (accesses 'kill_ftrace_graph')
> > 
> > The early head_32.S code calls into load_ucode_ap(), which has an an
> > ftrace hook, so it calls prepare_ftrace_return(), which calls
> > ftrace_graph_is_dead(), which tries to access the global
> > 'kill_ftrace_graph' variable with a virtual address, causing a fault
> > because the CPU is still in real mode.
> > 
> > The fix is to add a check in prepare_ftrace_return() to make sure it's
> > running in protected mode before continuing.  The check makes sure the
> > stack pointer is a virtual kernel address.  It's a bit of a hack, but
> > it's not very intrusive and it works well enough.
> > 
> > For reference, here are a few other (more difficult) ways this could
> > have potentially been fixed:
> > 
> > - Move startup_32_smp()'s call to load_ucode_ap() down to *after* paging
> >   is enabled.  (No idea what that would break.)
> > 
> > - Track down load_ucode_ap()'s entire callee tree and mark all the
> >   functions 'notrace'.  (Probably not realistic.)
> > 
> > - Pause graph tracing in ftrace_suspend_notifier_call() or bringup_cpu()
> >   or __cpu_up(), and ensure that the pause facility can be queried from
> >   real mode.
> > 
> > Reported-by: Paul Menzel <pmenzel@molgen.mpg.de>
> > Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
> > Cc: stable@kernel.org
> > Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
> 
> This is pretty much the same thing we were talking about before, right?

Yeah, the same patch from before, now with more tags!

> If so, then:
> 
> Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>

Thanks!

-- 
Josh

[toc] | [prev] | [next] | [standalone]

#1623625 — [tip:x86/urgent] ftrace/x86: Fix triple fault with graph tracing and suspend-to-ram

Commit-ID:  34a477e5297cbaa6ecc6e17c042a866e1cbe80d6
Gitweb:     http://git.kernel.org/tip/34a477e5297cbaa6ecc6e17c042a866e1cbe80d6
Author:     Josh Poimboeuf <jpoimboe@redhat.com>
AuthorDate: Thu, 13 Apr 2017 17:53:55 -0500
Committer:  Thomas Gleixner <tglx@linutronix.de>
CommitDate: Fri, 14 Apr 2017 11:48:51 +0200

ftrace/x86: Fix triple fault with graph tracing and suspend-to-ram

On x86-32, with CONFIG_FIRMWARE and multiple CPUs, if you enable function
graph tracing and then suspend to RAM, it will triple fault and reboot when
it resumes.

The first fault happens when booting a secondary CPU:

startup_32_smp()
  load_ucode_ap()
    prepare_ftrace_return()
      ftrace_graph_is_dead()
        (accesses 'kill_ftrace_graph')

The early head_32.S code calls into load_ucode_ap(), which has an an
ftrace hook, so it calls prepare_ftrace_return(), which calls
ftrace_graph_is_dead(), which tries to access the global
'kill_ftrace_graph' variable with a virtual address, causing a fault
because the CPU is still in real mode.

The fix is to add a check in prepare_ftrace_return() to make sure it's
running in protected mode before continuing.  The check makes sure the
stack pointer is a virtual kernel address.  It's a bit of a hack, but
it's not very intrusive and it works well enough.

For reference, here are a few other (more difficult) ways this could
have potentially been fixed:

- Move startup_32_smp()'s call to load_ucode_ap() down to *after* paging
  is enabled.  (No idea what that would break.)

- Track down load_ucode_ap()'s entire callee tree and mark all the
  functions 'notrace'.  (Probably not realistic.)

- Pause graph tracing in ftrace_suspend_notifier_call() or bringup_cpu()
  or __cpu_up(), and ensure that the pause facility can be queried from
  real mode.

Reported-by: Paul Menzel <pmenzel@molgen.mpg.de>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Tested-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Cc: "Rafael J . Wysocki" <rjw@rjwysocki.net>
Cc: linux-acpi@vger.kernel.org
Cc: Borislav Petkov <bp@alien8.de>
Cc: stable@kernel.org
Cc: Len Brown <lenb@kernel.org>
Link: http://lkml.kernel.org/r/5c1272269a580660703ed2eccf44308e790c7a98.1492123841.git.jpoimboe@redhat.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

---
 arch/x86/kernel/ftrace.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
index cbd73eb..0e5ceac 100644
--- a/arch/x86/kernel/ftrace.c
+++ b/arch/x86/kernel/ftrace.c
@@ -989,6 +989,18 @@ void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent,
 	unsigned long return_hooker = (unsigned long)
 				&return_to_handler;
 
+	/*
+	 * When resuming from suspend-to-ram, this function can be indirectly
+	 * called from early CPU startup code while the CPU is in real mode,
+	 * which would fail miserably.  Make sure the stack pointer is a
+	 * virtual address.
+	 *
+	 * This check isn't as accurate as virt_addr_valid(), but it should be
+	 * good enough for this purpose, and it's fast.
+	 */
+	if (unlikely((long)__builtin_frame_address(0) >= 0))
+		return;
+
 	if (unlikely(ftrace_graph_is_dead()))
 		return;
 

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web