[Intel-xe] [PATCH v5 3/3] drm/xe/pmu: Enable PMU interface
Belgaumkar, Vinay
vinay.belgaumkar at intel.com
Thu Aug 31 22:21:41 UTC 2023
On 8/31/2023 3:11 PM, Aravind Iddamsetty wrote:
> On 31/08/23 22:28, Dixit, Ashutosh wrote:
> HI Ashutosh,
>
>> On Thu, 31 Aug 2023 03:29:11 -0700, Aravind Iddamsetty wrote:
>> Hi Aravind,
>>
>> Hmm, what happened to the email formatting here?
> not sure, some how my email client is showing proper or I messed up.
>>> On Tue, 29 Aug 2023 22:15:44 -0700, Aravind Iddamsetty wrote:
>>>
>>> diff --git a/drivers/gpu/drm/xe/xe_pmu.c b/drivers/gpu/drm/xe/xe_pmu.c
>>> new file mode 100644
>>> index 000000000000..41dd422812ff
>>> --- /dev/null
>>> +++ b/drivers/gpu/drm/xe/xe_pmu.c
>>> @@ -0,0 +1,679 @@
>>> +// SPDX-License-Identifier: MIT
>>> +/*
>>> + * Copyright © 2023 Intel Corporation
>>> + */
>>> +
>>> +#include <drm/drm_drv.h>
>>> +#include <drm/drm_managed.h>
>>> +#include <drm/xe_drm.h>
>>> +
>>> +#include "regs/xe_gt_regs.h"
>>> +#include "xe_device.h"
>>> +#include "xe_gt_clock.h"
>>> +#include "xe_mmio.h"
>>> +
>>> +static cpumask_t xe_pmu_cpumask;
>>> +static unsigned int xe_pmu_target_cpu = -1;
>>> +
>>> +static unsigned int config_gt_id(const u64 config)
>>> +{
>>> + return config >> __XE_PMU_GT_SHIFT;
>>> +}
>>> +
>>> +static u64 config_counter(const u64 config)
>>> +{
>>> + return config & ~(~0ULL << __XE_PMU_GT_SHIFT);
>>> +}
>>> +
>>> +static int engine_busyness_sample_type(u64 config)
>>> +{
>>> + int type = 0;
>>>
>>>
>>> Why initialize? The switch statement should have a default with a BUG/WARN_ON
>>> below? Also see the comment below.
>>>
>>> +
>>> + switch (config) {
>>> + case XE_PMU_RENDER_GROUP_BUSY(0):
>>> + type = __XE_SAMPLE_RENDER_GROUP_BUSY;
>>> + break;
>>> + case XE_PMU_COPY_GROUP_BUSY(0):
>>> + type = __XE_SAMPLE_COPY_GROUP_BUSY;
>>> + break;
>>> + case XE_PMU_MEDIA_GROUP_BUSY(0):
>>> + type = __XE_SAMPLE_MEDIA_GROUP_BUSY;
>>> + break;
>>> + case XE_PMU_ANY_ENGINE_GROUP_BUSY(0):
>>> + type = __XE_SAMPLE_ANY_ENGINE_GROUP_BUSY;
>>> + break;
>>> + }
>>> +
>>> + return type;
>>> +}
>>>
>>>
>>> I am thinking this function is not really needed. We can just do:
>>>
>>> int sample_type = config - 1;
>>>
>>> or
>>>
>>> int sample_type = config_counter(config) - 1;
>>>
>>> It might not always be true in future, the configs can start from any range.
>> Disagree. This is uapi. Once it is exposed it cannot change. I am talking
>> about this:
>>
>> +#define XE_PMU_INTERRUPTS(gt) ___XE_PMU_OTHER(gt, 0)
>> +#define XE_PMU_RENDER_GROUP_BUSY(gt) ___XE_PMU_OTHER(gt, 1)
>> +#define XE_PMU_COPY_GROUP_BUSY(gt) ___XE_PMU_OTHER(gt, 2)
>> +#define XE_PMU_MEDIA_GROUP_BUSY(gt) ___XE_PMU_OTHER(gt, 3)
>> +#define XE_PMU_ANY_ENGINE_GROUP_BUSY(gt) ___XE_PMU_OTHER(gt, 4)
>>
>> How can this "start from any range"? We can only add new counters after
>> these, not before these, correct?
> I didn't mean to say that these particular one's would change but any
> future new events that might fall into these categories might start
> from a different range. sorry for the confusion.
>
> Your suggestion makes it looks simple but somehow i wanted to tie this to
> the enums we defined in sample array, so ya will check one more time if
> it doesn't really makes any sense will clean it up.
>
>
>>> in engine_group_busyness_read? See comment at __xe_pmu_event_read below.
>>>
>>> +
>>> +static void xe_pmu_event_destroy(struct perf_event *event)
>>> +{
>>> + struct xe_device *xe =
>>> + container_of(event->pmu, typeof(*xe), pmu.base);
>>> +
>>> + drm_WARN_ON(&xe->drm, event->parent);
>>> +
>>> + drm_dev_put(&xe->drm);
>>> +}
>>> +
>>> +static u64 __engine_group_busyness_read(struct xe_gt *gt, int sample_type)
>>> +{
>>> + u64 val = 0;
>> No need to initialize here I think. We are not really expecting to drop
>> into the default case, which should be caught much before we enter this
>> function.
>>
>>> +
>>> + switch (sample_type) {
>>> + case __XE_SAMPLE_RENDER_GROUP_BUSY:
>>> + val = xe_mmio_read32(gt, XE_OAG_RENDER_BUSY_FREE);
>>> + break;
>>> + case __XE_SAMPLE_COPY_GROUP_BUSY:
>>> + val = xe_mmio_read32(gt, XE_OAG_BLT_BUSY_FREE);
>>> + break;
>>> + case __XE_SAMPLE_MEDIA_GROUP_BUSY:
>>> + val = xe_mmio_read32(gt, XE_OAG_ANY_MEDIA_FF_BUSY_FREE);
>>> + break;
>>> + case __XE_SAMPLE_ANY_ENGINE_GROUP_BUSY:
>>> + val = xe_mmio_read32(gt, XE_OAG_RC0_ANY_ENGINE_BUSY_FREE);
>>> + break;
>>> + default:
>>> + drm_warn(>->tile->xe->drm, "unknown pmu event\n");
>>> + }
>>> +
>>> + return xe_gt_clock_cycles_to_ns(gt, val * 16);
>>> +}
>>> +
>>> +static u64 engine_group_busyness_read(struct xe_gt *gt, u64 config)
>>> +{
>>> + int sample_type = engine_busyness_sample_type(config);
>>>
>>>
>>> If config is event->attr.config, this can just be 'config_counter(config) - 1'.
>>> See comment at __xe_pmu_event_read below.
>>>
>>> + const unsigned int gt_id = gt->info.id;
>>> + struct xe_device *xe = gt->tile->xe;
>>> + struct xe_pmu *pmu = &xe->pmu;
>>> + unsigned long flags;
>>> + bool device_awake;
>>> + u64 val;
>>> +
>>> + device_awake = xe_device_mem_access_get_if_ongoing(xe);
>>> + if (device_awake) {
>>> + XE_WARN_ON(xe_force_wake_get(gt_to_fw(gt), XE_FW_GT));
>>> + val = __engine_group_busyness_read(gt, sample_type);
>>> + XE_WARN_ON(xe_force_wake_put(gt_to_fw(gt), XE_FW_GT));
>>> + xe_device_mem_access_put(xe);
>>> + }
>>> +
>>> + spin_lock_irqsave(&pmu->lock, flags);
>>> +
>>> + if (device_awake)
>>> + pmu->sample[gt_id][sample_type] = val;
>>> + else
>>> + val = pmu->sample[gt_id][sample_type];
>>> +
>>> + spin_unlock_irqrestore(&pmu->lock, flags);
>>> +
>>> + return val;
>>> +}
>>> +
>>> +static void engine_group_busyness_store(struct xe_gt *gt)
>>> +{
>>> + struct xe_pmu *pmu = >->tile->xe->pmu;
>>> + unsigned int gt_id = gt->info.id;
>>> + unsigned long flags;
>>> + int i;
>>> +
>>> + spin_lock_irqsave(&pmu->lock, flags);
>>> +
>>> + for (i = __XE_SAMPLE_RENDER_GROUP_BUSY; i <= __XE_SAMPLE_ANY_ENGINE_GROUP_BUSY; i++) {
>>> + pmu->sample[gt_id][i] = __engine_group_busyness_read(gt, i);
>>>
>>>
>>> This is not quite right. At the minimum we need to take forcewake
>>> here. Also since this is called in both suspend and runtime_suspend code
>>> paths we might also need to the take the runtime_pm reference.
>>>
>>> The pm reference and forcewake are already taken in suspend paths hence
>>> didn't add here again as this is called only from those paths.
>>>
>>> check xe_gt_suspend.
>> Sorry, you are right, I missed it. So this is fine.
>>
>>
>>>
>>> I think the simplest might be to just construct 'config'
>>> (event->attr.config) here and call engine_group_busyness_read? Something
>>> like:
>>>
>>> for (i = __XE_SAMPLE_RENDER_GROUP_BUSY; i <= __XE_SAMPLE_ANY_ENGINE_GROUP_BUSY; i++) {
>>> config = ; // Construct config using gt_id and i
>>> engine_group_busyness_read(gt, i);
>>> }
>>>
>>> This will automatically save the read values in pmu->sample[][] if the
>>> device is awake. Thoughts?
>>>
>>> I think this is best kept separate from usual read paths(which are
>>> atomic) didn't want to club them. Also because forcewakes and pm
>>> reference are taken separately in suspend path.
>> Sure, no changes needed here. Just get rid of the braces to keep checkpatch
>> happy.
>>
>>
>>>
>>> + }
>>> +
>>> + spin_unlock_irqrestore(&pmu->lock, flags);
>>> +}
>>> +
>>> +static int
>>> +config_status(struct xe_device *xe, u64 config)
>>> +{
>>> + unsigned int max_gt_id = xe->info.gt_count > 1 ? 1 : 0;
>>>
>>>
>>> What is this for? See below.
>>>
>>> reminiscent of my previous code, will clean it up.
>>>
>>>
>>>
>>> + unsigned int gt_id = config_gt_id(config);
>>> + struct xe_gt *gt = xe_device_get_gt(xe, gt_id);
>>> +
>>> + if (gt_id > max_gt_id)
>>>
>>>
>>> Maybe this can just be:
>>>
>>> if (gt_id >= XE_PMU_MAX_GT)
>>>
>>> + return -ENOENT;
>>> +
>>> + switch (config_counter(config)) {
>>> + case XE_PMU_INTERRUPTS(0):
>>> + if (gt_id)
>>> + return -ENOENT;
>>> + break;
>>> + case XE_PMU_RENDER_GROUP_BUSY(0):
>>> + case XE_PMU_COPY_GROUP_BUSY(0):
>>> + case XE_PMU_ANY_ENGINE_GROUP_BUSY(0):
>>> + if (gt->info.type == XE_GT_TYPE_MEDIA)
>>> + return -ENOENT;
>>> + break;
>>> + case XE_PMU_MEDIA_GROUP_BUSY(0):
>>> + if (!(gt->info.engine_mask & (BIT(XE_HW_ENGINE_VCS0) | BIT(XE_HW_ENGINE_VECS0))))
>>> + return -ENOENT;
>>> + break;
>>> + default:
>>> + return -ENOENT;
>>> + }
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int xe_pmu_event_init(struct perf_event *event)
>>> +{
>>> + struct xe_device *xe =
>>> + container_of(event->pmu, typeof(*xe), pmu.base);
>>> + struct xe_pmu *pmu = &xe->pmu;
>>> + int ret;
>>> +
>>> + if (pmu->closed)
>>> + return -ENODEV;
>>> +
>>> + if (event->attr.type != event->pmu->type)
>>> + return -ENOENT;
>>> +
>>> + /* unsupported modes and filters */
>>> + if (event->attr.sample_period) /* no sampling */
>>> + return -EINVAL;
>>> +
>>> + if (has_branch_stack(event))
>>> + return -EOPNOTSUPP;
>>> +
>>> + if (event->cpu < 0)
>>> + return -EINVAL;
>>> +
>>> + /* only allow running on one cpu at a time */
>>> + if (!cpumask_test_cpu(event->cpu, &xe_pmu_cpumask))
>>> + return -EINVAL;
>>> +
>>> + ret = config_status(xe, event->attr.config);
>>> + if (ret)
>>> + return ret;
>>> +
>>> + if (!event->parent) {
>>> + drm_dev_get(&xe->drm);
>>> + event->destroy = xe_pmu_event_destroy;
>>> + }
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static u64 __xe_pmu_event_read(struct perf_event *event)
>>> +{
>>> + struct xe_device *xe =
>>> + container_of(event->pmu, typeof(*xe), pmu.base);
>>> + const unsigned int gt_id = config_gt_id(event->attr.config);
>>> + const u64 config = config_counter(event->attr.config);
>>>
>>>
>>> Probably nit but this config being different from event->attr.config is
>>> confusing. Let's use 'event->attr.config' throughout as argument to
>>> functions and use config_counter() to get rid of gt_id. So get rid of this
>>> config variable.
>>>
>>> + struct xe_gt *gt = xe_device_get_gt(xe, gt_id);
>>> + struct xe_pmu *pmu = &xe->pmu;
>>> + u64 val = 0;
>>> +
>>> + switch (config) {
>>>
>>>
>>> switch (config_counter(event->attr.config))
>>>
>>> + case XE_PMU_INTERRUPTS(0):
>>> + val = READ_ONCE(pmu->irq_count);
>>> + break;
>>> + case XE_PMU_RENDER_GROUP_BUSY(0):
>>> + case XE_PMU_COPY_GROUP_BUSY(0):
>>> + case XE_PMU_ANY_ENGINE_GROUP_BUSY(0):
>>> + case XE_PMU_MEDIA_GROUP_BUSY(0):
>>> + val = engine_group_busyness_read(gt, config);
>>>
>>>
>>> engine_group_busyness_read(gt, event->attr.config);
>>>
>>> hmmm ok.
>>>
>>>
>>>
>>> Also, need a default case.
>>>
>>> + }
>>> +
>>> + return val;
>>> +}
>>> +
>>> +static void xe_pmu_event_read(struct perf_event *event)
>>> +{
>>> + struct xe_device *xe =
>>> + container_of(event->pmu, typeof(*xe), pmu.base);
>>> + struct hw_perf_event *hwc = &event->hw;
>>> + struct xe_pmu *pmu = &xe->pmu;
>>> + u64 prev, new;
>>> +
>>> + if (pmu->closed) {
>>> + event->hw.state = PERF_HES_STOPPED;
>>> + return;
>>> + }
>>> +again:
>>> + prev = local64_read(&hwc->prev_count);
>>> + new = __xe_pmu_event_read(event);
>>> +
>>> + if (local64_cmpxchg(&hwc->prev_count, prev, new) != prev)
>>> + goto again;
>>> +
>>> + local64_add(new - prev, &event->count);
>>> +}
>>> +
>>> +static void xe_pmu_enable(struct perf_event *event)
>>> +{
The i915_pmu code checks which event is requested here and accordingly
sets pmu->enable (which doesn't seem to be defined here yet). Any reason
we are not doing this yet?
Thanks,
Vinay.
>>> + /*
>>> + * Store the current counter value so we can report the correct delta
>>> + * for all listeners. Even when the event was already enabled and has
>>> + * an existing non-zero value.
>>> + */
>>> + local64_set(&event->hw.prev_count, __xe_pmu_event_read(event));
>>>
>>>
>>> Right now nothing is being enabled here (unlike i915) so the function name
>>> xe_pmu_enable looks weird. Not sure, maybe leave as is for when things get
>>> added in the future?
>>>
>>> +static int xe_pmu_cpu_online(unsigned int cpu, struct hlist_node *node)
>>> +{
>>> + struct xe_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), cpuhp.node);
>>> +
>>> + BUG_ON(!pmu->base.event_init);
>>> +
>>> + /* Select the first online CPU as a designated reader. */
>>> + if (cpumask_empty(&xe_pmu_cpumask))
>>> + cpumask_set_cpu(cpu, &xe_pmu_cpumask);
>>> +
>>> + return 0;
>>> +}
>>> +
>>> +static int xe_pmu_cpu_offline(unsigned int cpu, struct hlist_node *node)
>>> +{
>>> + struct xe_pmu *pmu = hlist_entry_safe(node, typeof(*pmu), cpuhp.node);
>>> + unsigned int target = xe_pmu_target_cpu;
>>> +
>>> + BUG_ON(!pmu->base.event_init);
>>>
>>>
>>> nit but wondering if we should remove these two BUG_ON's (and save a couple
>>> of checkpatch warnings even though the BUG_ON's are in i915) and just do
>>> something like:
>>>
>>> if (!pmu->base.event_init)
>>> return 0;
>>>
>>> The reason for the BUG_ON's seems to be that these functions can be called
>>> after module_init but before probe.
>>>
>>> xe_pmu_cpu_online() doesn't depend on pmu at all so looks like the BUG_ON
>>> can just be dropped?
>>>
>>> the xe_pmu_cpu_online/offline are not invoked when they are registered with
>>> cpuhp_setup_state_multi, but rather when cpuhp_state_add_instance() is called
>>> which is done post the PMU is initialized hence the check for BUG_ON.
>> cpuhp_setup_state_multi is called at module_init
>> time. cpuhp_state_add_instance is called from xe_pmu_register, i.e. during
>> device probe when pmu->base.event_init is already initialized. Therefore
>> seems even less reason to have the BUG_ON's.
> that is true even, so will remove the BUG_ON.
>> Just a few minor issues left now so I am hoping we can wrap this marathon
>> review up soon :)
> ya i'm waiting for the same :)
>
> Thanks,
> Aravind.
>> Thanks.
>> --
>> Ashutosh
More information about the Intel-xe
mailing list