From: "Shi, Yang" <yang.shi@linaro.org>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: catalin.marinas@arm.com, will.deacon@arm.com,
dave.long@linaro.org, panand@redhat.com,
linux-kernel@vger.kernel.org, linux-rt-users@vger.kernel.org,
linux-arm-kernel@lists.infradead.org,
linaro-kernel@lists.linaro.org,
"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Subject: Re: [v2 PATCH] arm64: replace read_lock to rcu lock in call_break_hook
Date: Thu, 01 Oct 2015 13:53:51 -0700 [thread overview]
Message-ID: <560D9D5F.70403@linaro.org> (raw)
In-Reply-To: <20151001130819.5080bc24@gandalf.local.home>
On 10/1/2015 10:08 AM, Steven Rostedt wrote:
> On Thu, 1 Oct 2015 09:37:37 -0700
> Yang Shi <yang.shi@linaro.org> wrote:
>
>> BUG: sleeping function called from invalid context at kernel/locking/rtmutex.c:917
>> in_atomic(): 0, irqs_disabled(): 128, pid: 342, name: perf
>> 1 lock held by perf/342:
>> #0: (break_hook_lock){+.+...}, at: [<ffffffc0000851ac>] call_break_hook+0x34/0xd0
>> irq event stamp: 62224
>> hardirqs last enabled at (62223): [<ffffffc00010b7bc>] __call_rcu.constprop.59+0x104/0x270
>> hardirqs last disabled at (62224): [<ffffffc0000fbe20>] vprintk_emit+0x68/0x640
>> softirqs last enabled at (0): [<ffffffc000097928>] copy_process.part.8+0x428/0x17f8
>> softirqs last disabled at (0): [< (null)>] (null)
>> CPU: 0 PID: 342 Comm: perf Not tainted 4.1.6-rt5 #4
>> Hardware name: linux,dummy-virt (DT)
>> Call trace:
>> [<ffffffc000089968>] dump_backtrace+0x0/0x128
>> [<ffffffc000089ab0>] show_stack+0x20/0x30
>> [<ffffffc0007030d0>] dump_stack+0x7c/0xa0
>> [<ffffffc0000c878c>] ___might_sleep+0x174/0x260
>> [<ffffffc000708ac8>] __rt_spin_lock+0x28/0x40
>> [<ffffffc000708db0>] rt_read_lock+0x60/0x80
>> [<ffffffc0000851a8>] call_break_hook+0x30/0xd0
>> [<ffffffc000085a70>] brk_handler+0x30/0x98
>> [<ffffffc000082248>] do_debug_exception+0x50/0xb8
>> Exception stack(0xffffffc00514fe30 to 0xffffffc00514ff50)
>> fe20: 00000000 00000000 c1594680 0000007f
>> fe40: ffffffff ffffffff 92063940 0000007f 0550dcd8 ffffffc0 00000000 00000000
>> fe60: 0514fe70 ffffffc0 000be1f8 ffffffc0 0514feb0 ffffffc0 0008948c ffffffc0
>> fe80: 00000004 00000000 0514fed0 ffffffc0 ffffffff ffffffff 9282a948 0000007f
>> fea0: 00000000 00000000 9282b708 0000007f c1592820 0000007f 00083914 ffffffc0
>> fec0: 00000000 00000000 00000010 00000000 00000064 00000000 00000001 00000000
>> fee0: 005101e0 00000000 c1594680 0000007f c1594740 0000007f ffffffd8 ffffff80
>> ff00: 00000000 00000000 00000000 00000000 c1594770 0000007f c1594770 0000007f
>> ff20: 00665e10 00000000 7f7f7f7f 7f7f7f7f 01010101 01010101 00000000 00000000
>> ff40: 928e4cc0 0000007f 91ff11e8 0000007f
>>
>> call_break_hook is called in atomic context (hard irq disabled), so replace
>> the sleepable lock to rcu lock and replace relevant list operations to rcu
>> version.
>>
>> Signed-off-by: Yang Shi <yang.shi@linaro.org>
>> ---
>> v1-> v2
>> Replace list operations to rcu version.
>>
>> arch/arm64/kernel/debug-monitors.c | 10 +++++-----
>> 1 file changed, 5 insertions(+), 5 deletions(-)
>>
>> diff --git a/arch/arm64/kernel/debug-monitors.c b/arch/arm64/kernel/debug-monitors.c
>> index cebf786..cf0e4fc 100644
>> --- a/arch/arm64/kernel/debug-monitors.c
>> +++ b/arch/arm64/kernel/debug-monitors.c
>> @@ -276,14 +276,14 @@ static DEFINE_RWLOCK(break_hook_lock);
>> void register_break_hook(struct break_hook *hook)
>> {
>> write_lock(&break_hook_lock);
>> - list_add(&hook->node, &break_hook);
>> + list_add_rcu(&hook->node, &break_hook);
>> write_unlock(&break_hook_lock);
>> }
>>
>> void unregister_break_hook(struct break_hook *hook)
>> {
>> write_lock(&break_hook_lock);
>> - list_del(&hook->node);
>> + list_del_rcu(&hook->node);
>> write_unlock(&break_hook_lock);
>> }
>
> Shouldn't there be a synchronize_rcu() somewhere?
So far kgdb is the only user of unregister_break_hook in mainline kernel.
Just read Documentation/RCU/checklist.txt, it says:
Note that synchronize_rcu() -only- guarantees to wait until
all currently executing rcu_read_lock()-protected RCU read-side critical
sections complete.
For kgdb, the unregister is just called in kgdb_arch_exit by
kgdb_unregister_io_module, which is called when rmmod kgdb module.
The break point handler is done synchronously. So, it sounds should be
not a problem without calling synchronize_rcu().
Yang
> -- Steve
>
>>
>> @@ -292,11 +292,11 @@ static int call_break_hook(struct pt_regs *regs, unsigned int esr)
>> struct break_hook *hook;
>> int (*fn)(struct pt_regs *regs, unsigned int esr) = NULL;
>>
>> - read_lock(&break_hook_lock);
>> - list_for_each_entry(hook, &break_hook, node)
>> + rcu_read_lock();
>> + list_for_each_entry_rcu(hook, &break_hook, node)
>> if ((esr & hook->esr_mask) == hook->esr_val)
>> fn = hook->fn;
>> - read_unlock(&break_hook_lock);
>> + rcu_read_unlock();
>>
>> return fn ? fn(regs, esr) : DBG_HOOK_ERROR;
>> }
>
WARNING: multiple messages have this Message-ID (diff)
From: yang.shi@linaro.org (Shi, Yang)
To: linux-arm-kernel@lists.infradead.org
Subject: [v2 PATCH] arm64: replace read_lock to rcu lock in call_break_hook
Date: Thu, 01 Oct 2015 13:53:51 -0700 [thread overview]
Message-ID: <560D9D5F.70403@linaro.org> (raw)
In-Reply-To: <20151001130819.5080bc24@gandalf.local.home>
On 10/1/2015 10:08 AM, Steven Rostedt wrote:
> On Thu, 1 Oct 2015 09:37:37 -0700
> Yang Shi <yang.shi@linaro.org> wrote:
>
>> BUG: sleeping function called from invalid context at kernel/locking/rtmutex.c:917
>> in_atomic(): 0, irqs_disabled(): 128, pid: 342, name: perf
>> 1 lock held by perf/342:
>> #0: (break_hook_lock){+.+...}, at: [<ffffffc0000851ac>] call_break_hook+0x34/0xd0
>> irq event stamp: 62224
>> hardirqs last enabled at (62223): [<ffffffc00010b7bc>] __call_rcu.constprop.59+0x104/0x270
>> hardirqs last disabled at (62224): [<ffffffc0000fbe20>] vprintk_emit+0x68/0x640
>> softirqs last enabled at (0): [<ffffffc000097928>] copy_process.part.8+0x428/0x17f8
>> softirqs last disabled at (0): [< (null)>] (null)
>> CPU: 0 PID: 342 Comm: perf Not tainted 4.1.6-rt5 #4
>> Hardware name: linux,dummy-virt (DT)
>> Call trace:
>> [<ffffffc000089968>] dump_backtrace+0x0/0x128
>> [<ffffffc000089ab0>] show_stack+0x20/0x30
>> [<ffffffc0007030d0>] dump_stack+0x7c/0xa0
>> [<ffffffc0000c878c>] ___might_sleep+0x174/0x260
>> [<ffffffc000708ac8>] __rt_spin_lock+0x28/0x40
>> [<ffffffc000708db0>] rt_read_lock+0x60/0x80
>> [<ffffffc0000851a8>] call_break_hook+0x30/0xd0
>> [<ffffffc000085a70>] brk_handler+0x30/0x98
>> [<ffffffc000082248>] do_debug_exception+0x50/0xb8
>> Exception stack(0xffffffc00514fe30 to 0xffffffc00514ff50)
>> fe20: 00000000 00000000 c1594680 0000007f
>> fe40: ffffffff ffffffff 92063940 0000007f 0550dcd8 ffffffc0 00000000 00000000
>> fe60: 0514fe70 ffffffc0 000be1f8 ffffffc0 0514feb0 ffffffc0 0008948c ffffffc0
>> fe80: 00000004 00000000 0514fed0 ffffffc0 ffffffff ffffffff 9282a948 0000007f
>> fea0: 00000000 00000000 9282b708 0000007f c1592820 0000007f 00083914 ffffffc0
>> fec0: 00000000 00000000 00000010 00000000 00000064 00000000 00000001 00000000
>> fee0: 005101e0 00000000 c1594680 0000007f c1594740 0000007f ffffffd8 ffffff80
>> ff00: 00000000 00000000 00000000 00000000 c1594770 0000007f c1594770 0000007f
>> ff20: 00665e10 00000000 7f7f7f7f 7f7f7f7f 01010101 01010101 00000000 00000000
>> ff40: 928e4cc0 0000007f 91ff11e8 0000007f
>>
>> call_break_hook is called in atomic context (hard irq disabled), so replace
>> the sleepable lock to rcu lock and replace relevant list operations to rcu
>> version.
>>
>> Signed-off-by: Yang Shi <yang.shi@linaro.org>
>> ---
>> v1-> v2
>> Replace list operations to rcu version.
>>
>> arch/arm64/kernel/debug-monitors.c | 10 +++++-----
>> 1 file changed, 5 insertions(+), 5 deletions(-)
>>
>> diff --git a/arch/arm64/kernel/debug-monitors.c b/arch/arm64/kernel/debug-monitors.c
>> index cebf786..cf0e4fc 100644
>> --- a/arch/arm64/kernel/debug-monitors.c
>> +++ b/arch/arm64/kernel/debug-monitors.c
>> @@ -276,14 +276,14 @@ static DEFINE_RWLOCK(break_hook_lock);
>> void register_break_hook(struct break_hook *hook)
>> {
>> write_lock(&break_hook_lock);
>> - list_add(&hook->node, &break_hook);
>> + list_add_rcu(&hook->node, &break_hook);
>> write_unlock(&break_hook_lock);
>> }
>>
>> void unregister_break_hook(struct break_hook *hook)
>> {
>> write_lock(&break_hook_lock);
>> - list_del(&hook->node);
>> + list_del_rcu(&hook->node);
>> write_unlock(&break_hook_lock);
>> }
>
> Shouldn't there be a synchronize_rcu() somewhere?
So far kgdb is the only user of unregister_break_hook in mainline kernel.
Just read Documentation/RCU/checklist.txt, it says:
Note that synchronize_rcu() -only- guarantees to wait until
all currently executing rcu_read_lock()-protected RCU read-side critical
sections complete.
For kgdb, the unregister is just called in kgdb_arch_exit by
kgdb_unregister_io_module, which is called when rmmod kgdb module.
The break point handler is done synchronously. So, it sounds should be
not a problem without calling synchronize_rcu().
Yang
> -- Steve
>
>>
>> @@ -292,11 +292,11 @@ static int call_break_hook(struct pt_regs *regs, unsigned int esr)
>> struct break_hook *hook;
>> int (*fn)(struct pt_regs *regs, unsigned int esr) = NULL;
>>
>> - read_lock(&break_hook_lock);
>> - list_for_each_entry(hook, &break_hook, node)
>> + rcu_read_lock();
>> + list_for_each_entry_rcu(hook, &break_hook, node)
>> if ((esr & hook->esr_mask) == hook->esr_val)
>> fn = hook->fn;
>> - read_unlock(&break_hook_lock);
>> + rcu_read_unlock();
>>
>> return fn ? fn(regs, esr) : DBG_HOOK_ERROR;
>> }
>
next prev parent reply other threads:[~2015-10-01 20:53 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-01 16:37 [v2 PATCH] arm64: replace read_lock to rcu lock in call_break_hook Yang Shi
2015-10-01 16:37 ` Yang Shi
2015-10-01 17:08 ` Steven Rostedt
2015-10-01 17:08 ` Steven Rostedt
2015-10-01 20:53 ` Shi, Yang [this message]
2015-10-01 20:53 ` Shi, Yang
2015-10-01 21:27 ` Paul E. McKenney
2015-10-01 21:27 ` Paul E. McKenney
2015-10-01 22:15 ` Shi, Yang
2015-10-01 22:15 ` Shi, Yang
2015-10-05 20:08 ` Shi, Yang
2015-10-05 20:08 ` Shi, Yang
2015-10-06 17:09 ` Paul E. McKenney
2015-10-06 17:09 ` Paul E. McKenney
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=560D9D5F.70403@linaro.org \
--to=yang.shi@linaro.org \
--cc=catalin.marinas@arm.com \
--cc=dave.long@linaro.org \
--cc=linaro-kernel@lists.linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rt-users@vger.kernel.org \
--cc=panand@redhat.com \
--cc=paulmck@linux.vnet.ibm.com \
--cc=rostedt@goodmis.org \
--cc=will.deacon@arm.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.