Received-SPF: pass (google.com: domain of linux-kernel+bounces-102960-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Message-ID: <94DB0209EA3D1C5F+b646f31e-15a1-4753-8309-687b5860863e@shingroup.cn>
Date: Thu, 14 Mar 2024 16:09:42 +0800
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH 02/10] perf: Add capability for common event support
To: Robin Murphy <robin.murphy@arm.com>, Peter Zijlstra
 <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>,
 Arnaldo Carvalho de Melo <acme@kernel.org>,
 Namhyung Kim <namhyung@kernel.org>, Mark Rutland <mark.rutland@arm.com>,
 Will Deacon <will@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>,
 Jiri Olsa <jolsa@kernel.org>, Ian Rogers <irogers@google.com>,
 Adrian Hunter <adrian.hunter@intel.com>, linux-kernel@vger.kernel.org,
 linux-arm-kernel@lists.infradead.org, x86@kernel.org,
 linux-perf-users@vger.kernel.org
References: <cover.1710257512.git.robin.murphy@arm.com>
 <feb10873fe9e4e10b5ffbbe8e296c8a45632e3c2.1710257512.git.robin.murphy@arm.com>
Content-Language: en-US
From: =?UTF-8?B?WWFuZyBKaWFsb25nIOadqOS9s+m+mQ==?= <jialong.yang@shingroup.cn>
In-Reply-To: <feb10873fe9e4e10b5ffbbe8e296c8a45632e3c2.1710257512.git.robin.murphy@arm.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Feedback-ID: bizesmtp:shingroup.cn:qybglogicsvrgz:qybglogicsvrgz6a-1


在 2024/3/13 1:34, Robin Murphy 写道:
> Many PMUs do not support common hardware/cache/etc. events and only
> handle their own PMU-specific events. Since this only depends on
> matching the event and PMU types, it's a prime candidate for a core
> capability to save more event_init boilerplate in drivers.
> 
> Signed-off-by: Robin Murphy <robin.murphy@arm.com>
> ---
>   include/linux/perf_event.h | 1 +
>   kernel/events/core.c       | 5 +++++
>   2 files changed, 6 insertions(+)
> 
> diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
> index d2a15c0c6f8a..983201f21dd2 100644
> --- a/include/linux/perf_event.h
> +++ b/include/linux/perf_event.h
> @@ -291,6 +291,7 @@ struct perf_event_pmu_context;
>   #define PERF_PMU_CAP_NO_EXCLUDE			0x0040
>   #define PERF_PMU_CAP_AUX_OUTPUT			0x0080
>   #define PERF_PMU_CAP_EXTENDED_HW_TYPE		0x0100
> +#define PERF_PMU_CAP_NO_COMMON_EVENTS		0x0200
>   
>   struct perf_output_handle;
>   
> diff --git a/kernel/events/core.c b/kernel/events/core.c
> index f0f0f71213a1..7ad80826c218 100644
> --- a/kernel/events/core.c
> +++ b/kernel/events/core.c
> @@ -11649,6 +11649,11 @@ static int perf_try_init_event(struct pmu *pmu, struct perf_event *event)
>   	struct perf_event_context *ctx = NULL;
>   	int ret;
>   
> +	/* Short-circuit if we know the PMU won't want this event */
> +	if (pmu->capabilities & PERF_PMU_CAP_NO_COMMON_EVENTS &&
> +	    event->attr.type != pmu->type)
> +		return -ENOENT;
> +

         /*
          * PERF_TYPE_HARDWARE and PERF_TYPE_HW_CACHE
          * are often aliases for PERF_TYPE_RAW.
          */
         type = event->attr.type;
         if (type == PERF_TYPE_HARDWARE || type == PERF_TYPE_HW_CACHE) {
                 type = event->attr.config >> PERF_PMU_TYPE_SHIFT;
                 if (!type) {
                         type = PERF_TYPE_RAW;
                 } else {
                         extended_type = true;
                         event->attr.config &= PERF_HW_EVENT_MASK;
                 }
         }

again:
         rcu_read_lock();
         pmu = idr_find(&pmu_idr, type);
         rcu_read_unlock();
         if (pmu) {
Above code tells me it's possible that 'pmu->type != event->attr.type' 
is true when event->attr.type equals to PERF_TYPE_HARDWARE or 
PERF_TYPE_HW_CACHE, and pmu->type should equal to event->attr.config >> 
PERF_PMU_TYPE_SHIFT.

We find the target pmu by event->attr.config >> PERF_PMU_TYPE_SHIFT.

Code added discard this option.

And code tells me that no try. Target PMU is doubtless.


>   	if (!try_module_get(pmu->module))
>   		return -ENODEV;
>