Mailing List Archive

[PATCH v2 3/4] arm64: Make debug exception handlers visible from RCU
Make debug exceptions visible from RCU so that synchronize_rcu()
correctly track the debug exception handler.

This also introduces sanity checks for user-mode exceptions as same
as x86's ist_enter()/ist_exit().

The debug exception can interrupt in idle task. For example, it warns
if we put a kprobe on a function called from idle task as below.
The warning message showed that the rcu_read_lock() caused this
problem. But actually, this means the RCU is lost the context which
is already in NMI/IRQ.

/sys/kernel/debug/tracing # echo p default_idle_call >> kprobe_events
/sys/kernel/debug/tracing # echo 1 > events/kprobes/enable
/sys/kernel/debug/tracing # [ 135.122237]
[ 135.125035] =============================
[ 135.125310] WARNING: suspicious RCU usage
[ 135.125581] 5.2.0-08445-g9187c508bdc7 #20 Not tainted
[ 135.125904] -----------------------------
[ 135.126205] include/linux/rcupdate.h:594 rcu_read_lock() used illegally while idle!
[ 135.126839]
[ 135.126839] other info that might help us debug this:
[ 135.126839]
[ 135.127410]
[ 135.127410] RCU used illegally from idle CPU!
[ 135.127410] rcu_scheduler_active = 2, debug_locks = 1
[ 135.128114] RCU used illegally from extended quiescent state!
[ 135.128555] 1 lock held by swapper/0/0:
[ 135.128944] #0: (____ptrval____) (rcu_read_lock){....}, at: call_break_hook+0x0/0x178
[ 135.130499]
[ 135.130499] stack backtrace:
[ 135.131192] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.2.0-08445-g9187c508bdc7 #20
[ 135.131841] Hardware name: linux,dummy-virt (DT)
[ 135.132224] Call trace:
[ 135.132491] dump_backtrace+0x0/0x140
[ 135.132806] show_stack+0x24/0x30
[ 135.133133] dump_stack+0xc4/0x10c
[ 135.133726] lockdep_rcu_suspicious+0xf8/0x108
[ 135.134171] call_break_hook+0x170/0x178
[ 135.134486] brk_handler+0x28/0x68
[ 135.134792] do_debug_exception+0x90/0x150
[ 135.135051] el1_dbg+0x18/0x8c
[ 135.135260] default_idle_call+0x0/0x44
[ 135.135516] cpu_startup_entry+0x2c/0x30
[ 135.135815] rest_init+0x1b0/0x280
[ 135.136044] arch_call_rest_init+0x14/0x1c
[ 135.136305] start_kernel+0x4d4/0x500
[ 135.136597]

So make debug exception visible to RCU can fix this warning.

Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Cc: Paul E. McKenney <paulmck@linux.ibm.com>
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
---
arch/arm64/mm/fault.c | 40 ++++++++++++++++++++++++++++++++++++++++
1 file changed, 40 insertions(+)

diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
index 9568c116ac7f..a6b244240db6 100644
--- a/arch/arm64/mm/fault.c
+++ b/arch/arm64/mm/fault.c
@@ -777,6 +777,42 @@ void __init hook_debug_fault_code(int nr,
debug_fault_info[nr].name = name;
}

+/*
+ * In debug exception context, we explicitly disable preemption.
+ * This serves two purposes: it makes it much less likely that we would
+ * accidentally schedule in exception context and it will force a warning
+ * if we somehow manage to schedule by accident.
+ */
+static void debug_exception_enter(struct pt_regs *regs)
+{
+ if (user_mode(regs)) {
+ RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU");
+ } else {
+ /*
+ * We might have interrupted pretty much anything. In
+ * fact, if we're a debug exception, we can even interrupt
+ * NMI processing. We don't want in_nmi() to return true,
+ * but we need to notify RCU.
+ */
+ rcu_nmi_enter();
+ }
+
+ preempt_disable();
+
+ /* This code is a bit fragile. Test it. */
+ RCU_LOCKDEP_WARN(!rcu_is_watching(), "exception_enter didn't work");
+}
+NOKPROBE_SYMBOL(debug_exception_enter);
+
+static void debug_exception_exit(struct pt_regs *regs)
+{
+ preempt_enable_no_resched();
+
+ if (!user_mode(regs))
+ rcu_nmi_exit();
+}
+NOKPROBE_SYMBOL(debug_exception_exit);
+
#ifdef CONFIG_ARM64_ERRATUM_1463225
DECLARE_PER_CPU(int, __in_cortex_a76_erratum_1463225_wa);

@@ -824,6 +860,8 @@ asmlinkage void __exception do_debug_exception(unsigned long addr_if_watchpoint,
if (interrupts_enabled(regs))
trace_hardirqs_off();

+ debug_exception_enter(regs);
+
if (user_mode(regs) && !is_ttbr0_addr(pc))
arm64_apply_bp_hardening();

@@ -832,6 +870,8 @@ asmlinkage void __exception do_debug_exception(unsigned long addr_if_watchpoint,
inf->sig, inf->code, (void __user *)pc, esr);
}

+ debug_exception_exit(regs);
+
if (interrupts_enabled(regs))
trace_hardirqs_on();
}
Re: [PATCH v2 3/4] arm64: Make debug exception handlers visible from RCU [ In reply to ]
On Mon, Jul 22, 2019 at 04:48:58PM +0900, Masami Hiramatsu wrote:
> Make debug exceptions visible from RCU so that synchronize_rcu()
> correctly track the debug exception handler.
>
> This also introduces sanity checks for user-mode exceptions as same
> as x86's ist_enter()/ist_exit().
>
> The debug exception can interrupt in idle task. For example, it warns
> if we put a kprobe on a function called from idle task as below.
> The warning message showed that the rcu_read_lock() caused this
> problem. But actually, this means the RCU is lost the context which
> is already in NMI/IRQ.
>
> /sys/kernel/debug/tracing # echo p default_idle_call >> kprobe_events
> /sys/kernel/debug/tracing # echo 1 > events/kprobes/enable
> /sys/kernel/debug/tracing # [ 135.122237]
> [ 135.125035] =============================
> [ 135.125310] WARNING: suspicious RCU usage
> [ 135.125581] 5.2.0-08445-g9187c508bdc7 #20 Not tainted
> [ 135.125904] -----------------------------
> [ 135.126205] include/linux/rcupdate.h:594 rcu_read_lock() used illegally while idle!
> [ 135.126839]
> [ 135.126839] other info that might help us debug this:
> [ 135.126839]
> [ 135.127410]
> [ 135.127410] RCU used illegally from idle CPU!
> [ 135.127410] rcu_scheduler_active = 2, debug_locks = 1
> [ 135.128114] RCU used illegally from extended quiescent state!
> [ 135.128555] 1 lock held by swapper/0/0:
> [ 135.128944] #0: (____ptrval____) (rcu_read_lock){....}, at: call_break_hook+0x0/0x178
> [ 135.130499]
> [ 135.130499] stack backtrace:
> [ 135.131192] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.2.0-08445-g9187c508bdc7 #20
> [ 135.131841] Hardware name: linux,dummy-virt (DT)
> [ 135.132224] Call trace:
> [ 135.132491] dump_backtrace+0x0/0x140
> [ 135.132806] show_stack+0x24/0x30
> [ 135.133133] dump_stack+0xc4/0x10c
> [ 135.133726] lockdep_rcu_suspicious+0xf8/0x108
> [ 135.134171] call_break_hook+0x170/0x178
> [ 135.134486] brk_handler+0x28/0x68
> [ 135.134792] do_debug_exception+0x90/0x150
> [ 135.135051] el1_dbg+0x18/0x8c
> [ 135.135260] default_idle_call+0x0/0x44
> [ 135.135516] cpu_startup_entry+0x2c/0x30
> [ 135.135815] rest_init+0x1b0/0x280
> [ 135.136044] arch_call_rest_init+0x14/0x1c
> [ 135.136305] start_kernel+0x4d4/0x500
> [ 135.136597]
>
> So make debug exception visible to RCU can fix this warning.
>
> Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
> Cc: Paul E. McKenney <paulmck@linux.ibm.com>

From an RCU viewpoint:

Acked-by: Paul E. McKenney <paulmck@linux.ibm.com>

> Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
> ---
> arch/arm64/mm/fault.c | 40 ++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 40 insertions(+)
>
> diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
> index 9568c116ac7f..a6b244240db6 100644
> --- a/arch/arm64/mm/fault.c
> +++ b/arch/arm64/mm/fault.c
> @@ -777,6 +777,42 @@ void __init hook_debug_fault_code(int nr,
> debug_fault_info[nr].name = name;
> }
>
> +/*
> + * In debug exception context, we explicitly disable preemption.
> + * This serves two purposes: it makes it much less likely that we would
> + * accidentally schedule in exception context and it will force a warning
> + * if we somehow manage to schedule by accident.
> + */
> +static void debug_exception_enter(struct pt_regs *regs)
> +{
> + if (user_mode(regs)) {
> + RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU");
> + } else {
> + /*
> + * We might have interrupted pretty much anything. In
> + * fact, if we're a debug exception, we can even interrupt
> + * NMI processing. We don't want in_nmi() to return true,
> + * but we need to notify RCU.
> + */
> + rcu_nmi_enter();
> + }
> +
> + preempt_disable();
> +
> + /* This code is a bit fragile. Test it. */
> + RCU_LOCKDEP_WARN(!rcu_is_watching(), "exception_enter didn't work");
> +}
> +NOKPROBE_SYMBOL(debug_exception_enter);
> +
> +static void debug_exception_exit(struct pt_regs *regs)
> +{
> + preempt_enable_no_resched();
> +
> + if (!user_mode(regs))
> + rcu_nmi_exit();
> +}
> +NOKPROBE_SYMBOL(debug_exception_exit);
> +
> #ifdef CONFIG_ARM64_ERRATUM_1463225
> DECLARE_PER_CPU(int, __in_cortex_a76_erratum_1463225_wa);
>
> @@ -824,6 +860,8 @@ asmlinkage void __exception do_debug_exception(unsigned long addr_if_watchpoint,
> if (interrupts_enabled(regs))
> trace_hardirqs_off();
>
> + debug_exception_enter(regs);
> +
> if (user_mode(regs) && !is_ttbr0_addr(pc))
> arm64_apply_bp_hardening();
>
> @@ -832,6 +870,8 @@ asmlinkage void __exception do_debug_exception(unsigned long addr_if_watchpoint,
> inf->sig, inf->code, (void __user *)pc, esr);
> }
>
> + debug_exception_exit(regs);
> +
> if (interrupts_enabled(regs))
> trace_hardirqs_on();
> }
>
Re: [PATCH v2 3/4] arm64: Make debug exception handlers visible from RCU [ In reply to ]
Hi,

On 22/07/2019 08:48, Masami Hiramatsu wrote:
> Make debug exceptions visible from RCU so that synchronize_rcu()
> correctly track the debug exception handler.
>
> This also introduces sanity checks for user-mode exceptions as same
> as x86's ist_enter()/ist_exit().
>
> The debug exception can interrupt in idle task. For example, it warns
> if we put a kprobe on a function called from idle task as below.
> The warning message showed that the rcu_read_lock() caused this
> problem. But actually, this means the RCU is lost the context which
> is already in NMI/IRQ.

> So make debug exception visible to RCU can fix this warning.

> diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
> index 9568c116ac7f..a6b244240db6 100644
> --- a/arch/arm64/mm/fault.c
> +++ b/arch/arm64/mm/fault.c
> @@ -777,6 +777,42 @@ void __init hook_debug_fault_code(int nr,
> debug_fault_info[nr].name = name;
> }
>
> +/*
> + * In debug exception context, we explicitly disable preemption.
> + * This serves two purposes: it makes it much less likely that we would
> + * accidentally schedule in exception context and it will force a warning
> + * if we somehow manage to schedule by accident.
> + */
> +static void debug_exception_enter(struct pt_regs *regs)
> +{
> + if (user_mode(regs)) {
> + RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU");

Would moving entry.S's context_tracking_user_exit() call to be before do_debug_exception()
also fix this?

I don't know the reason its done 'after' debug exception handling. Its always been like
this: commit 6c81fe7925cc4c42 ("arm64: enable context tracking").


> + } else {
> + /*
> + * We might have interrupted pretty much anything. In
> + * fact, if we're a debug exception, we can even interrupt
> + * NMI processing.

> + * We don't want in_nmi() to return true,
> + * but we need to notify RCU.

How come? If you interrupted an SError or pseudo-nmi, it already is. Those paths should
all be painted no-kprobe, but I'm sure there are gaps. The hw-breakpoints can almost
certainly hook them.


> + */
> + rcu_nmi_enter();

Can we interrupt printk()? Do we need printk_nmi_enter()? ... What about ftrace?

Because SError and pseudo-nmi can interrupt interrupt-masked code, we describe them as
NMI. The only difference here is these exceptions are synchronous.


I suspect we should make these debug exceptions nmi for EL1. We can then use this for the
kprobe-re-entrance stuff so the pre/post hooks don't get run if they interrupted something
also described as NMI.


> + }
> +
> + preempt_disable();
> +
> + /* This code is a bit fragile. Test it. */
> + RCU_LOCKDEP_WARN(!rcu_is_watching(), "exception_enter didn't work");
> +}
> +NOKPROBE_SYMBOL(debug_exception_enter);


Thanks,

James
Re: [PATCH v2 3/4] arm64: Make debug exception handlers visible from RCU [ In reply to ]
On Tue, 23 Jul 2019 18:07:56 +0100
James Morse <james.morse@arm.com> wrote:

> Hi,
>
> On 22/07/2019 08:48, Masami Hiramatsu wrote:
> > Make debug exceptions visible from RCU so that synchronize_rcu()
> > correctly track the debug exception handler.
> >
> > This also introduces sanity checks for user-mode exceptions as same
> > as x86's ist_enter()/ist_exit().
> >
> > The debug exception can interrupt in idle task. For example, it warns
> > if we put a kprobe on a function called from idle task as below.
> > The warning message showed that the rcu_read_lock() caused this
> > problem. But actually, this means the RCU is lost the context which
> > is already in NMI/IRQ.
>
> > So make debug exception visible to RCU can fix this warning.
>
> > diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
> > index 9568c116ac7f..a6b244240db6 100644
> > --- a/arch/arm64/mm/fault.c
> > +++ b/arch/arm64/mm/fault.c
> > @@ -777,6 +777,42 @@ void __init hook_debug_fault_code(int nr,
> > debug_fault_info[nr].name = name;
> > }
> >
> > +/*
> > + * In debug exception context, we explicitly disable preemption.
> > + * This serves two purposes: it makes it much less likely that we would
> > + * accidentally schedule in exception context and it will force a warning
> > + * if we somehow manage to schedule by accident.
> > + */
> > +static void debug_exception_enter(struct pt_regs *regs)
> > +{
> > + if (user_mode(regs)) {
> > + RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU");
>
> Would moving entry.S's context_tracking_user_exit() call to be before do_debug_exception()
> also fix this?

It sounds like treating only user context, correct?
This part is just adding assertion, not fixing the problem which Naresh reported.

>
> I don't know the reason its done 'after' debug exception handling. Its always been like
> this: commit 6c81fe7925cc4c42 ("arm64: enable context tracking").
>
>
> > + } else {
> > + /*
> > + * We might have interrupted pretty much anything. In
> > + * fact, if we're a debug exception, we can even interrupt
> > + * NMI processing.
>
> > + * We don't want in_nmi() to return true,
> > + * but we need to notify RCU.
>
> How come? If you interrupted an SError or pseudo-nmi, it already is. Those paths should
> all be painted no-kprobe, but I'm sure there are gaps. The hw-breakpoints can almost
> certainly hook them.

I think that sentense means "we don't want that this code makes in_nmi() to return true"
So, if the breakpoint interrupts pNMI/SError context, it is OK that in_nmi() returns true.

>
>
> > + */
> > + rcu_nmi_enter();
>
> Can we interrupt printk()? Do we need printk_nmi_enter()? ... What about ftrace?

Good point! As far as I know, we don't use it because ftrace doesn't use printk.
But indeed, kprobes user can use printk and they have to call printk_nmi_enter()/exit(),
that must be commented in the documentation. Anyway, basically it is user's choice.

>
> Because SError and pseudo-nmi can interrupt interrupt-masked code, we describe them as
> NMI. The only difference here is these exceptions are synchronous.
>
>
> I suspect we should make these debug exceptions nmi for EL1. We can then use this for the
> kprobe-re-entrance stuff so the pre/post hooks don't get run if they interrupted something
> also described as NMI.

I'm not sure how it can prevent... anyway because we have to run a single-stepping for
recovery, and kprobe already check the reentered kprobes and skip user-handlers in
such case.

Thank you,

>
>
> > + }
> > +
> > + preempt_disable();
> > +
> > + /* This code is a bit fragile. Test it. */
> > + RCU_LOCKDEP_WARN(!rcu_is_watching(), "exception_enter didn't work");
> > +}
> > +NOKPROBE_SYMBOL(debug_exception_enter);
>
>
> Thanks,
>
> James


--
Masami Hiramatsu <mhiramat@kernel.org>