Re: [PATCH v7 3/4] drivers/perf: add DesignWare PCIe PMU driver

[Shuai Xue]

On 2023/10/14 00:30, Bjorn Helgaas wrote:
> On Fri, Oct 13, 2023 at 11:46:44AM +0800, Shuai Xue wrote:
>>
>>
>> On 2023/10/13 00:25, Bjorn Helgaas wrote:
>>> On Thu, Oct 12, 2023 at 11:28:55AM +0800, Shuai Xue wrote:
>>>> This commit adds the PCIe Performance Monitoring Unit (PMU) driver support
>>>> for T-Head Yitian SoC chip. Yitian is based on the Synopsys PCI Express
>>>> Core controller IP which provides statistics feature. The PMU is not a PCIe
>>>> Root Complex integrated End Point(RCiEP) device but only register counters
>>>> provided by each PCIe Root Port.
> 
> IIUC, the PMU is directly integrated into the Root Port: it's
> discovered and operated via the Root Port config space.  If so, I
> wouldn't bother mentioning RCiEP because there's no need to list all
> the things it's *not*.

I see, will not mention RCiEP next version.

> 
>>>> To facilitate collection of statistics the controller provides the
>>>> following two features for each Root Port:
>>>>
>>>> - Time Based Analysis (RX/TX data throughput and time spent in each
>>>>   low-power LTSSM state)
>>>> - Event counters (Error and Non-Error for lanes)
>>>>
>>>> Note, only one counter for each type and does not overflow interrupt.
>>>
>>> Not sure what "does not overflow interrupt" means.  Does it mean
>>> there's no interrupt generated when the counter overflows?
>>
>> Yes, exactly. The rootport does NOT generate interrupt when the
>> couter overflows.  I think the assumption hidden in this design is
>> 64-bit counter will not overflow within observable time.
>>
>> PCIe 5.0 slots can now reach anywhere between ~4GB/sec for a x1 slot
>> up to ~64GB/sec for a x16 slot. The unit of counter is 16 byte.
>>
>> 	2^64/(64/16*10^9)/60/60/24/365=146 years
>>
>> so, the counter will not overflow within 146 years.
> 
> Certainly a reasonable assumption :)
> 
> But I'm confused about how many counters there are.  Clearly there are
> two features ((1) time-based analysis and (2) event counters).
> 
> "One counter for each type" suggests there's one counter for
> time-based analysis and a second counter for event counting, 

You are right, two counters, TIME_BASED_ANAL_DATA_REG for time-based
analysis and EVENT_COUNTER_DATA_REG for event counting. And both of them
do not support generate interrupt when the counter overflows.

> but from
> dwc_pcie_pmu_event_add(), it looks like each Root Port might have a
> single counter, and you can decide whether that counter is used for
> time-based analysis or event counting, but you can't do both at the
> same time? 

dwc_pcie_pmu_event_add() now limit the PMU usage to stat event for either
time-based analysis or event counting.

I will extend to support stat them at the same time.

@@ -447,10 +447,10 @@ static int dwc_pcie_pmu_event_add(struct perf_event *event, int flags)
 	u32 ctrl;

 	/* Only one counter and it is in use */
-	if (pcie_pmu->event)
+	if (pcie_pmu->event[type])
 		return -ENOSPC;

-	pcie_pmu->event = event;
+	pcie_pmu->event[type] = event;
 	hwc->state = PERF_HES_STOPPED | PERF_HES_UPTODATE;

 	if (type == DWC_PCIE_LANE_EVENT) {
@@ -486,10 +486,11 @@ static int dwc_pcie_pmu_event_add(struct perf_event *event, int flags)
 static void dwc_pcie_pmu_event_del(struct perf_event *event, int flags)
 {
 	struct dwc_pcie_pmu *pcie_pmu = to_dwc_pcie_pmu(event->pmu);
+	enum dwc_pcie_event_type type = DWC_PCIE_EVENT_TYPE(event);

 	dwc_pcie_pmu_event_stop(event, flags | PERF_EF_UPDATE);
 	perf_event_update_userpage(event);
-	pcie_pmu->event = NULL;
+	pcie_pmu->event[type] = NULL;
 }

> And the event counting is for a single lane, not for the
> link as a whole?

Yes.

> 
> If so, I might word this as:
> 
>   Each Root Port contains one counter that can be used for either:
> 
>     - Time-Based Analysis (RX/TX data throughput and time spent in
>       each low-power LTSSM state) or
> 
>     - Event counting (error and non-error events for a specified lane)
> 
>   There is no interrupt for counter overflow.

Based on above, I change the word to:

	To facilitate collection of statistics the controller provides the
	following two features for each Root Port:

	- one 64-bit counter for Time Based Analysis (RX/TX data throughput and
	  time spent in each low-power LTSSM state) and
	- one 32-bit counter for Event counting (error and non-error events for
	  a specified lane)

	Note: There is no interrupt for counter overflow.


> 
>>>> +	  Enable perf support for Synopsys DesignWare PCIe PMU Performance
>>>> +	  monitoring event on platform including the Yitian 710.
>>>
>>> Should this mention Alibaba or T-Head?  I don't know how
>>> Alibaba/T-Head/Yitian are all related.
>>
>> The server chips, named Yitian 710, are custom-built by Alibaba Group's chip
>> development business, T-Head.
>>
>> 	  Enable perf support for Synopsys DesignWare PCIe PMU Performance
>> 	  monitoring event on platform including the Alibaba Yitian 710.
>>
>> Is this okay?
> 
> Perfect :)


Thank you for valuable comments.

Best Regards,
Shuai