Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-12-04 Thread John Harrison
On 11/29/2017 11:55 PM, Sagar Arun Kamble wrote: On 11/30/2017 12:45 PM, John Harrison wrote: On 11/29/2017 10:19 PM, Sagar Arun Kamble wrote: On 11/30/2017 8:34 AM, John Harrison wrote: On 11/24/2017 6:12 AM, Chris Wilson wrote: Quoting Michał Winiarski (2017-11-24 12:37:56) Since we see

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-29 Thread Sagar Arun Kamble
On 11/30/2017 12:45 PM, John Harrison wrote: On 11/29/2017 10:19 PM, Sagar Arun Kamble wrote: On 11/30/2017 8:34 AM, John Harrison wrote: On 11/24/2017 6:12 AM, Chris Wilson wrote: Quoting Michał Winiarski (2017-11-24 12:37:56) Since we see the effects for GuC preeption, let's gather some

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-29 Thread John Harrison
On 11/29/2017 10:19 PM, Sagar Arun Kamble wrote: On 11/30/2017 8:34 AM, John Harrison wrote: On 11/24/2017 6:12 AM, Chris Wilson wrote: Quoting Michał Winiarski (2017-11-24 12:37:56) Since we see the effects for GuC preeption, let's gather some evidence. (SKL) intel_guc_send_mmio latency:

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-29 Thread Sagar Arun Kamble
On 11/30/2017 8:34 AM, John Harrison wrote: On 11/24/2017 6:12 AM, Chris Wilson wrote: Quoting Michał Winiarski (2017-11-24 12:37:56) Since we see the effects for GuC preeption, let's gather some evidence. (SKL) intel_guc_send_mmio latency: 100 rounds of gem_exec_latency --r '*-preemption'

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-29 Thread John Harrison
On 11/24/2017 6:12 AM, Chris Wilson wrote: Quoting Michał Winiarski (2017-11-24 12:37:56) Since we see the effects for GuC preeption, let's gather some evidence. (SKL) intel_guc_send_mmio latency: 100 rounds of gem_exec_latency --r '*-preemption' drm-tip: usecs : count

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-24 Thread Chris Wilson
Quoting Michał Winiarski (2017-11-24 12:37:56) > Since we see the effects for GuC preeption, let's gather some evidence. > > (SKL) > intel_guc_send_mmio latency: 100 rounds of gem_exec_latency --r '*-preemption' > > drm-tip: > usecs : count distribution > 0 -> 1

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-24 Thread Michał Winiarski
On Tue, Nov 21, 2017 at 08:59:51PM +, Chris Wilson wrote: > Instead of sleeping for a fixed 1ms (roughly, depending on timer slack), > start with a small sleep and exponentially increase the sleep on each > cycle. > > A good example of a beneficiary is the guc mmio communication channel. >

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-22 Thread Sagar Arun Kamble
On 11/22/2017 3:06 PM, Chris Wilson wrote: Quoting Sagar Arun Kamble (2017-11-22 07:41:02) On 11/22/2017 2:29 AM, Chris Wilson wrote: Instead of sleeping for a fixed 1ms (roughly, depending on timer slack), start with a small sleep and exponentially increase the sleep on each cycle. A good

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-22 Thread Michal Wajdeczko
On Wed, 22 Nov 2017 10:36:15 +0100, Chris Wilson wrote: Quoting Sagar Arun Kamble (2017-11-22 07:41:02) On 11/22/2017 2:29 AM, Chris Wilson wrote: > Instead of sleeping for a fixed 1ms (roughly, depending on timer slack), > start with a small sleep and

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-22 Thread Chris Wilson
Quoting Sagar Arun Kamble (2017-11-22 07:41:02) > > > On 11/22/2017 2:29 AM, Chris Wilson wrote: > > Instead of sleeping for a fixed 1ms (roughly, depending on timer slack), > > start with a small sleep and exponentially increase the sleep on each > > cycle. > > > > A good example of a

Re: [Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-21 Thread Sagar Arun Kamble
On 11/22/2017 2:29 AM, Chris Wilson wrote: Instead of sleeping for a fixed 1ms (roughly, depending on timer slack), start with a small sleep and exponentially increase the sleep on each cycle. A good example of a beneficiary is the guc mmio communication channel. As Tvrtko said, for the

[Intel-gfx] [PATCH v3] drm/i915: Use exponential backoff for wait_for()

2017-11-21 Thread Chris Wilson
Instead of sleeping for a fixed 1ms (roughly, depending on timer slack), start with a small sleep and exponentially increase the sleep on each cycle. A good example of a beneficiary is the guc mmio communication channel. Typically we expect (and so spin) for 10us for a quick response, but this