Re: [PATCH rcu 04/11] kernel/smp: Provide boot-time timeout for CSD lock diagnostics

[Juergen Gross <jgross@suse.com>]

[-- Attachment #1.1.1: Type: text/plain, Size: 3429 bytes --]

On 19.04.22 00:53, Paul E. McKenney wrote:
> Debugging of problems involving insanely long-running SMI handlers
> proceeds better if the CSD-lock timeout can be adjusted.  This commit
> therefore provides a new smp.csd_lock_timeout kernel boot parameter
> that specifies the timeout in milliseconds.  The default remains at the
> previously hard-coded value of five seconds.
> 
> Cc: Rik van Riel <riel@surriel.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Ingo Molnar <mingo@kernel.org>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> Cc: Juergen Gross <jgross@suse.com>
> Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
> ---
>   Documentation/admin-guide/kernel-parameters.txt | 11 +++++++++++
>   kernel/smp.c                                    |  7 +++++--
>   2 files changed, 16 insertions(+), 2 deletions(-)
> 
> diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
> index 3f1cc5e317ed..645c4c001b16 100644
> --- a/Documentation/admin-guide/kernel-parameters.txt
> +++ b/Documentation/admin-guide/kernel-parameters.txt
> @@ -5377,6 +5377,17 @@
>   	smart2=		[HW]
>   			Format: <io1>[,<io2>[,...,<io8>]]
>   
> +	smp.csd_lock_timeout= [KNL]
> +			Specify the period of time in milliseconds
> +			that smp_call_function() and friends will wait
> +			for a CPU to release the CSD lock.  This is
> +			useful when diagnosing bugs involving CPUs
> +			disabling interrupts for extended periods
> +			of time.  Defaults to 5,000 milliseconds, and
> +			setting a value of zero disables this feature.
> +			This feature may be more efficiently disabled
> +			using the csdlock_debug- kernel parameter.
> +
>   	smsc-ircc2.nopnp	[HW] Don't use PNP to discover SMC devices
>   	smsc-ircc2.ircc_cfg=	[HW] Device configuration I/O port
>   	smsc-ircc2.ircc_sir=	[HW] SIR base I/O port
> diff --git a/kernel/smp.c b/kernel/smp.c
> index 01a7c1706a58..d82439bac401 100644
> --- a/kernel/smp.c
> +++ b/kernel/smp.c
> @@ -183,7 +183,9 @@ static DEFINE_PER_CPU(smp_call_func_t, cur_csd_func);
>   static DEFINE_PER_CPU(void *, cur_csd_info);
>   static DEFINE_PER_CPU(struct cfd_seq_local, cfd_seq_local);
>   
> -#define CSD_LOCK_TIMEOUT (5ULL * NSEC_PER_SEC)
> +static ulong csd_lock_timeout = 5000;  /* CSD lock timeout in milliseconds. */
> +module_param(csd_lock_timeout, ulong, 0444);
> +
>   static atomic_t csd_bug_count = ATOMIC_INIT(0);
>   static u64 cfd_seq;
>   
> @@ -329,6 +331,7 @@ static bool csd_lock_wait_toolong(struct __call_single_data *csd, u64 ts0, u64 *
>   	u64 ts2, ts_delta;
>   	call_single_data_t *cpu_cur_csd;
>   	unsigned int flags = READ_ONCE(csd->node.u_flags);
> +	unsigned long long csd_lock_timeout_ns = csd_lock_timeout * NSEC_PER_MSEC;
>   
>   	if (!(flags & CSD_FLAG_LOCK)) {
>   		if (!unlikely(*bug_id))
> @@ -341,7 +344,7 @@ static bool csd_lock_wait_toolong(struct __call_single_data *csd, u64 ts0, u64 *
>   
>   	ts2 = sched_clock();
>   	ts_delta = ts2 - *ts1;
> -	if (likely(ts_delta <= CSD_LOCK_TIMEOUT))
> +	if (likely(ts_delta <= csd_lock_timeout_ns || csd_lock_timeout_ns <= 0))

csd_lock_timeout_ns is unsigned, so "csd_lock_timeout_ns <= 0" should be rather
"csd_lock_timeout_ns == 0".

With that fixed you can add my

Reviewed-by: Juergen Gross <jgross@suse.com>


Juergen

[-- Attachment #1.1.2: OpenPGP public key --]
[-- Type: application/pgp-keys, Size: 3149 bytes --]

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 495 bytes --]