linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2] powerpc: Avoid nmi_enter/nmi_exit in real mode interrupt.
@ 2024-02-05  5:36 Mahesh Salgaonkar
  2024-02-12  8:06 ` Christophe Leroy
  0 siblings, 1 reply; 3+ messages in thread
From: Mahesh Salgaonkar @ 2024-02-05  5:36 UTC (permalink / raw)
  To: = linuxppc-dev; +Cc: = Aneesh Kumar K.V, = Ganesh Goudar, = Nicholas Piggin

nmi_enter()/nmi_exit() touches per cpu variables which can lead to kernel
crash when invoked during real mode interrupt handling (e.g. early HMI/MCE
interrupt handler) if percpu allocation comes from vmalloc area.

Early HMI/MCE handlers are called through DEFINE_INTERRUPT_HANDLER_NMI()
wrapper which invokes nmi_enter/nmi_exit calls. We don't see any issue when
percpu allocation is from the embedded first chunk. However with
CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK enabled there are chances where percpu
allocation can come from the vmalloc area.

With kernel command line "percpu_alloc=page" we can force percpu allocation
to come from vmalloc area and can see kernel crash in machine_check_early:

[    1.215714] NIP [c000000000e49eb4] rcu_nmi_enter+0x24/0x110
[    1.215717] LR [c0000000000461a0] machine_check_early+0xf0/0x2c0
[    1.215719] --- interrupt: 200
[    1.215720] [c000000fffd73180] [0000000000000000] 0x0 (unreliable)
[    1.215722] [c000000fffd731b0] [0000000000000000] 0x0
[    1.215724] [c000000fffd73210] [c000000000008364] machine_check_early_common+0x134/0x1f8

Fix this by avoiding use of nmi_enter()/nmi_exit() in real mode if percpu
first chunk is not embedded.

Signed-off-by: Mahesh Salgaonkar <mahesh@linux.ibm.com>
---
Changes in v2:
- Rebase to upstream master
- Use jump_labels, if CONFIG_JUMP_LABEL is enabled, to avoid redoing the
  test at each interrupt entry.
- v1 is at https://lore.kernel.org/linuxppc-dev/164578465828.74956.6065296024817333750.stgit@jupiter/
---
 arch/powerpc/include/asm/interrupt.h | 14 ++++++++++++++
 arch/powerpc/include/asm/percpu.h    | 11 +++++++++++
 arch/powerpc/kernel/setup_64.c       | 12 ++++++++++++
 3 files changed, 37 insertions(+)

diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
index a4196ab1d0167..3b4e17c23d9a9 100644
--- a/arch/powerpc/include/asm/interrupt.h
+++ b/arch/powerpc/include/asm/interrupt.h
@@ -336,6 +336,16 @@ static inline void interrupt_nmi_enter_prepare(struct pt_regs *regs, struct inte
 	if (IS_ENABLED(CONFIG_KASAN))
 		return;
 
+	/*
+	 * Likewise, do not use it in real mode if percpu first chunk is not
+	 * embedded. With CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK enabled there
+	 * are chances where percpu allocation can come from vmalloc area.
+	 */
+#ifdef CONFIG_PPC64
+	if (IS_ENABLED(CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK) && !is_embed_first_chunk)
+		return;
+#endif
+
 	/* Otherwise, it should be safe to call it */
 	nmi_enter();
 }
@@ -351,6 +361,10 @@ static inline void interrupt_nmi_exit_prepare(struct pt_regs *regs, struct inter
 		// no nmi_exit for a pseries hash guest taking a real mode exception
 	} else if (IS_ENABLED(CONFIG_KASAN)) {
 		// no nmi_exit for KASAN in real mode
+#ifdef CONFIG_PPC64
+	} else if (IS_ENABLED(CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK) && !is_embed_first_chunk) {
+		// no nmi_exit if percpu first chunk is not embedded
+#endif
 	} else {
 		nmi_exit();
 	}
diff --git a/arch/powerpc/include/asm/percpu.h b/arch/powerpc/include/asm/percpu.h
index 8e5b7d0b851c6..6b4dce4e78d5f 100644
--- a/arch/powerpc/include/asm/percpu.h
+++ b/arch/powerpc/include/asm/percpu.h
@@ -12,6 +12,17 @@
 
 #define __my_cpu_offset local_paca->data_offset
 
+#ifdef CONFIG_JUMP_LABEL
+DECLARE_STATIC_KEY_FALSE(__percpu_embed_first_chunk);
+
+#define is_embed_first_chunk	\
+		(static_key_enabled(&__percpu_embed_first_chunk.key))
+
+#else /* !CONFIG_JUMP_LABEL */
+extern bool __percpu_embed_first_chunk;
+#define is_embed_first_chunk	__percpu_embed_first_chunk
+
+#endif /* CONFIG_JUMP_LABEL */
 #endif /* CONFIG_SMP */
 #endif /* __powerpc64__ */
 
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 2f19d5e944852..674b6e1bebe9a 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -834,6 +834,11 @@ static __init int pcpu_cpu_to_node(int cpu)
 
 unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
 EXPORT_SYMBOL(__per_cpu_offset);
+#ifdef CONFIG_JUMP_LABEL
+DEFINE_STATIC_KEY_FALSE(__percpu_embed_first_chunk);
+#else
+bool __percpu_embed_first_chunk;
+#endif
 
 void __init setup_per_cpu_areas(void)
 {
@@ -869,6 +874,13 @@ void __init setup_per_cpu_areas(void)
 			pr_warn("PERCPU: %s allocator failed (%d), "
 				"falling back to page size\n",
 				pcpu_fc_names[pcpu_chosen_fc], rc);
+		else {
+#ifdef CONFIG_JUMP_LABEL
+			static_key_enable(&__percpu_embed_first_chunk.key);
+#else
+			__percpu_embed_first_chunk = true;
+#endif
+		}
 	}
 
 	if (rc < 0)
-- 
2.43.0


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH v2] powerpc: Avoid nmi_enter/nmi_exit in real mode interrupt.
  2024-02-05  5:36 [PATCH v2] powerpc: Avoid nmi_enter/nmi_exit in real mode interrupt Mahesh Salgaonkar
@ 2024-02-12  8:06 ` Christophe Leroy
  2024-02-12 15:32   ` Mahesh J Salgaonkar
  0 siblings, 1 reply; 3+ messages in thread
From: Christophe Leroy @ 2024-02-12  8:06 UTC (permalink / raw)
  To: Mahesh Salgaonkar, = linuxppc-dev
  Cc: = Aneesh Kumar K.V, = Ganesh Goudar, = Nicholas Piggin



Le 05/02/2024 à 06:36, Mahesh Salgaonkar a écrit :
> [Vous ne recevez pas souvent de courriers de mahesh@linux.ibm.com. Découvrez pourquoi ceci est important à https://aka.ms/LearnAboutSenderIdentification ]
> 
> nmi_enter()/nmi_exit() touches per cpu variables which can lead to kernel
> crash when invoked during real mode interrupt handling (e.g. early HMI/MCE
> interrupt handler) if percpu allocation comes from vmalloc area.
> 
> Early HMI/MCE handlers are called through DEFINE_INTERRUPT_HANDLER_NMI()
> wrapper which invokes nmi_enter/nmi_exit calls. We don't see any issue when
> percpu allocation is from the embedded first chunk. However with
> CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK enabled there are chances where percpu
> allocation can come from the vmalloc area.
> 
> With kernel command line "percpu_alloc=page" we can force percpu allocation
> to come from vmalloc area and can see kernel crash in machine_check_early:
> 
> [    1.215714] NIP [c000000000e49eb4] rcu_nmi_enter+0x24/0x110
> [    1.215717] LR [c0000000000461a0] machine_check_early+0xf0/0x2c0
> [    1.215719] --- interrupt: 200
> [    1.215720] [c000000fffd73180] [0000000000000000] 0x0 (unreliable)
> [    1.215722] [c000000fffd731b0] [0000000000000000] 0x0
> [    1.215724] [c000000fffd73210] [c000000000008364] machine_check_early_common+0x134/0x1f8
> 
> Fix this by avoiding use of nmi_enter()/nmi_exit() in real mode if percpu
> first chunk is not embedded.
> 
> Signed-off-by: Mahesh Salgaonkar <mahesh@linux.ibm.com>
> ---
> Changes in v2:
> - Rebase to upstream master
> - Use jump_labels, if CONFIG_JUMP_LABEL is enabled, to avoid redoing the
>    test at each interrupt entry.
> - v1 is at https://lore.kernel.org/linuxppc-dev/164578465828.74956.6065296024817333750.stgit@jupiter/
> ---
>   arch/powerpc/include/asm/interrupt.h | 14 ++++++++++++++
>   arch/powerpc/include/asm/percpu.h    | 11 +++++++++++
>   arch/powerpc/kernel/setup_64.c       | 12 ++++++++++++
>   3 files changed, 37 insertions(+)
> 
> diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
> index a4196ab1d0167..3b4e17c23d9a9 100644
> --- a/arch/powerpc/include/asm/interrupt.h
> +++ b/arch/powerpc/include/asm/interrupt.h
> @@ -336,6 +336,16 @@ static inline void interrupt_nmi_enter_prepare(struct pt_regs *regs, struct inte
>          if (IS_ENABLED(CONFIG_KASAN))
>                  return;
> 
> +       /*
> +        * Likewise, do not use it in real mode if percpu first chunk is not
> +        * embedded. With CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK enabled there
> +        * are chances where percpu allocation can come from vmalloc area.
> +        */
> +#ifdef CONFIG_PPC64

Instead of adding this #ifdef in middle of code, could you define 
is_embed_first_chunk as always 'true' when CONFIG_PPC64 is not defined ?

> +       if (IS_ENABLED(CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK) && !is_embed_first_chunk)
> +               return;
> +#endif
> +
>          /* Otherwise, it should be safe to call it */
>          nmi_enter();
>   }
> @@ -351,6 +361,10 @@ static inline void interrupt_nmi_exit_prepare(struct pt_regs *regs, struct inter
>                  // no nmi_exit for a pseries hash guest taking a real mode exception
>          } else if (IS_ENABLED(CONFIG_KASAN)) {
>                  // no nmi_exit for KASAN in real mode
> +#ifdef CONFIG_PPC64

Same

> +       } else if (IS_ENABLED(CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK) && !is_embed_first_chunk) {
> +               // no nmi_exit if percpu first chunk is not embedded
> +#endif
>          } else {
>                  nmi_exit();
>          }
> diff --git a/arch/powerpc/include/asm/percpu.h b/arch/powerpc/include/asm/percpu.h
> index 8e5b7d0b851c6..6b4dce4e78d5f 100644
> --- a/arch/powerpc/include/asm/percpu.h
> +++ b/arch/powerpc/include/asm/percpu.h
> @@ -12,6 +12,17 @@
> 
>   #define __my_cpu_offset local_paca->data_offset
> 
> +#ifdef CONFIG_JUMP_LABEL
> +DECLARE_STATIC_KEY_FALSE(__percpu_embed_first_chunk);
> +
> +#define is_embed_first_chunk   \
> +               (static_key_enabled(&__percpu_embed_first_chunk.key))
> +
> +#else /* !CONFIG_JUMP_LABEL */
> +extern bool __percpu_embed_first_chunk;
> +#define is_embed_first_chunk   __percpu_embed_first_chunk
> +
> +#endif /* CONFIG_JUMP_LABEL */
>   #endif /* CONFIG_SMP */
>   #endif /* __powerpc64__ */
> 
> diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
> index 2f19d5e944852..674b6e1bebe9a 100644
> --- a/arch/powerpc/kernel/setup_64.c
> +++ b/arch/powerpc/kernel/setup_64.c
> @@ -834,6 +834,11 @@ static __init int pcpu_cpu_to_node(int cpu)
> 
>   unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
>   EXPORT_SYMBOL(__per_cpu_offset);
> +#ifdef CONFIG_JUMP_LABEL

Why this ifdef ? Even when CONFIG_JUMP_LABEL is not selected all this 
should just work fine.

> +DEFINE_STATIC_KEY_FALSE(__percpu_embed_first_chunk);
> +#else
> +bool __percpu_embed_first_chunk;
> +#endif
> 
>   void __init setup_per_cpu_areas(void)
>   {
> @@ -869,6 +874,13 @@ void __init setup_per_cpu_areas(void)
>                          pr_warn("PERCPU: %s allocator failed (%d), "
>                                  "falling back to page size\n",
>                                  pcpu_fc_names[pcpu_chosen_fc], rc);
> +               else {
> +#ifdef CONFIG_JUMP_LABEL
> +                       static_key_enable(&__percpu_embed_first_chunk.key);
> +#else
> +                       __percpu_embed_first_chunk = true;
> +#endif
> +               }
>          }
> 
>          if (rc < 0)
> --
> 2.43.0
> 

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: Re: [PATCH v2] powerpc: Avoid nmi_enter/nmi_exit in real mode interrupt.
  2024-02-12  8:06 ` Christophe Leroy
@ 2024-02-12 15:32   ` Mahesh J Salgaonkar
  0 siblings, 0 replies; 3+ messages in thread
From: Mahesh J Salgaonkar @ 2024-02-12 15:32 UTC (permalink / raw)
  To: Christophe Leroy
  Cc: = linuxppc-dev, Ganesh Goudar, Nicholas Piggin, Aneesh Kumar K.V

On 2024-02-12 08:06:25 Mon, Christophe Leroy wrote:
> 
> 
> Le 05/02/2024 à 06:36, Mahesh Salgaonkar a écrit :
> > [Vous ne recevez pas souvent de courriers de mahesh@linux.ibm.com. Découvrez pourquoi ceci est important à https://aka.ms/LearnAboutSenderIdentification ]
> > 
> > nmi_enter()/nmi_exit() touches per cpu variables which can lead to kernel
> > crash when invoked during real mode interrupt handling (e.g. early HMI/MCE
> > interrupt handler) if percpu allocation comes from vmalloc area.
> > 
> > Early HMI/MCE handlers are called through DEFINE_INTERRUPT_HANDLER_NMI()
> > wrapper which invokes nmi_enter/nmi_exit calls. We don't see any issue when
> > percpu allocation is from the embedded first chunk. However with
> > CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK enabled there are chances where percpu
> > allocation can come from the vmalloc area.
> > 
> > With kernel command line "percpu_alloc=page" we can force percpu allocation
> > to come from vmalloc area and can see kernel crash in machine_check_early:
> > 
> > [    1.215714] NIP [c000000000e49eb4] rcu_nmi_enter+0x24/0x110
> > [    1.215717] LR [c0000000000461a0] machine_check_early+0xf0/0x2c0
> > [    1.215719] --- interrupt: 200
> > [    1.215720] [c000000fffd73180] [0000000000000000] 0x0 (unreliable)
> > [    1.215722] [c000000fffd731b0] [0000000000000000] 0x0
> > [    1.215724] [c000000fffd73210] [c000000000008364] machine_check_early_common+0x134/0x1f8
> > 
> > Fix this by avoiding use of nmi_enter()/nmi_exit() in real mode if percpu
> > first chunk is not embedded.
> > 
> > Signed-off-by: Mahesh Salgaonkar <mahesh@linux.ibm.com>
> > ---
> > Changes in v2:
> > - Rebase to upstream master
> > - Use jump_labels, if CONFIG_JUMP_LABEL is enabled, to avoid redoing the
> >    test at each interrupt entry.
> > - v1 is at https://lore.kernel.org/linuxppc-dev/164578465828.74956.6065296024817333750.stgit@jupiter/
> > ---
> >   arch/powerpc/include/asm/interrupt.h | 14 ++++++++++++++
> >   arch/powerpc/include/asm/percpu.h    | 11 +++++++++++
> >   arch/powerpc/kernel/setup_64.c       | 12 ++++++++++++
> >   3 files changed, 37 insertions(+)
> > 
> > diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
> > index a4196ab1d0167..3b4e17c23d9a9 100644
> > --- a/arch/powerpc/include/asm/interrupt.h
> > +++ b/arch/powerpc/include/asm/interrupt.h
> > @@ -336,6 +336,16 @@ static inline void interrupt_nmi_enter_prepare(struct pt_regs *regs, struct inte
> >          if (IS_ENABLED(CONFIG_KASAN))
> >                  return;
> > 
> > +       /*
> > +        * Likewise, do not use it in real mode if percpu first chunk is not
> > +        * embedded. With CONFIG_NEED_PER_CPU_PAGE_FIRST_CHUNK enabled there
> > +        * are chances where percpu allocation can come from vmalloc area.
> > +        */
> > +#ifdef CONFIG_PPC64
> 
> Instead of adding this #ifdef in middle of code, could you define 
> is_embed_first_chunk as always 'true' when CONFIG_PPC64 is not defined ?

Will fix this in v3.

[...]
> > diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
> > index 2f19d5e944852..674b6e1bebe9a 100644
> > --- a/arch/powerpc/kernel/setup_64.c
> > +++ b/arch/powerpc/kernel/setup_64.c
> > @@ -834,6 +834,11 @@ static __init int pcpu_cpu_to_node(int cpu)
> > 
> >   unsigned long __per_cpu_offset[NR_CPUS] __read_mostly;
> >   EXPORT_SYMBOL(__per_cpu_offset);
> > +#ifdef CONFIG_JUMP_LABEL
> 
> Why this ifdef ? Even when CONFIG_JUMP_LABEL is not selected all this 
> should just work fine.

Yes you are right. I overlooked this. Will fix it in next revision.

Thanks for your review.

-- 
Mahesh J Salgaonkar

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2024-02-12 15:33 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-02-05  5:36 [PATCH v2] powerpc: Avoid nmi_enter/nmi_exit in real mode interrupt Mahesh Salgaonkar
2024-02-12  8:06 ` Christophe Leroy
2024-02-12 15:32   ` Mahesh J Salgaonkar

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).