From: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
To: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [Intel-gfx] [PATCH] drm/i915/perf: Enable application triggered OA reports
Date: Thu, 2 Apr 2020 09:07:56 +0300 [thread overview]
Message-ID: <2f5b1a30-96a5-c42e-86da-7d300d45fddd@intel.com> (raw)
In-Reply-To: <20200401184342.GA4430@orsosgc001.amr.corp.intel.com>
On 01/04/2020 21:43, Umesh Nerlige Ramappa wrote:
> On Tue, Mar 31, 2020 at 02:46:46PM +0300, Lionel Landwerlin wrote:
>> Gen12 brought an important redesign of the OA unit, splitting it in 2
>> with a per context part (OAR) and a global part (OAG).
>>
>> OAR deals with per context counters and implements the
>> MI_REPORT_PERF_COUNT command.
>>
>> OAG deals with global counters and the OA buffer.
>>
>> Unfortunately some of the counters available in OAG are not available
>> in OAR, for instance counters that would report global caches
>> utilization.
>>
>> Since applications making use of this want to access those additional
>> OAG counters we can enable them to generate a report from their
>> command buffer into the OA buffer. This is somewhat equivalent to
>> having them doing their own MI_REPORT_PERF_COUNT. The application then
>> parse the OA buffer as they were doing previously, only looking for a
>> begin/end OA report with the appropriate reason field in the OA buffer
>> instead of using MI_REPORT_PERF_COUNT generated reports for begin/end.
>>
>> Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
>> ---
>> drivers/gpu/drm/i915/gt/intel_workarounds.c | 18 ++++++++++++++++++
>> drivers/gpu/drm/i915/i915_perf.c | 10 +++++++---
>> drivers/gpu/drm/i915/i915_reg.h | 2 ++
>> 3 files changed, 27 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c
>> b/drivers/gpu/drm/i915/gt/intel_workarounds.c
>> index e96cc7fa0936..552eadaa6f9a 100644
>> --- a/drivers/gpu/drm/i915/gt/intel_workarounds.c
>> +++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c
>> @@ -1127,6 +1127,10 @@ static void gen9_whitelist_build(struct
>> i915_wa_list *w)
>>
>> /* WaSendPushConstantsFromMMIO:skl,bxt */
>> whitelist_reg(w, COMMON_SLICE_CHICKEN2);
>> +
>> + /* Allow userspace trigger OA report generation in OA buffer. */
>> + whitelist_reg(w, OAREPORTTRIG2);
>> + whitelist_reg(w, OAREPORTTRIG6);
>> }
>>
>> static void skl_whitelist_build(struct intel_engine_cs *engine)
>> @@ -1208,6 +1212,10 @@ static void cnl_whitelist_build(struct
>> intel_engine_cs *engine)
>>
>> /* WaEnablePreemptionGranularityControlByUMD:cnl */
>> whitelist_reg(w, GEN8_CS_CHICKEN1);
>> +
>> + /* Allow userspace trigger OA report generation in OA buffer. */
>> + whitelist_reg(w, OAREPORTTRIG2);
>> + whitelist_reg(w, OAREPORTTRIG6);
>> }
>>
>> static void icl_whitelist_build(struct intel_engine_cs *engine)
>> @@ -1237,6 +1245,12 @@ static void icl_whitelist_build(struct
>> intel_engine_cs *engine)
>> whitelist_reg_ext(w, PS_INVOCATION_COUNT,
>> RING_FORCE_TO_NONPRIV_ACCESS_RD |
>> RING_FORCE_TO_NONPRIV_RANGE_4);
>> +
>> + /*
>> + * Allow userspace trigger OA report generation in OA buffer.
>> + */
>> + whitelist_reg(w, OAREPORTTRIG2);
>> + whitelist_reg(w, OAREPORTTRIG6);
>> break;
>>
>> case VIDEO_DECODE_CLASS:
>> @@ -1281,6 +1295,10 @@ static void tgl_whitelist_build(struct
>> intel_engine_cs *engine)
>>
>> /* Wa_1806527549:tgl */
>> whitelist_reg(w, HIZ_CHICKEN);
>> +
>> + /* Allow userspace trigger OA report generation in OA
>> buffer. */
>> + whitelist_reg(w, GEN12_OAG_OAREPORTTRIG2);
>> + whitelist_reg(w, GEN12_OAG_OAREPORTTRIG6);
>> break;
>> default:
>> break;
>> diff --git a/drivers/gpu/drm/i915/i915_perf.c
>> b/drivers/gpu/drm/i915/i915_perf.c
>> index 28e3d76fa2e6..ae935b1b1ae3 100644
>> --- a/drivers/gpu/drm/i915/i915_perf.c
>> +++ b/drivers/gpu/drm/i915/i915_perf.c
>> @@ -1450,7 +1450,8 @@ static void gen8_init_oa_buffer(struct
>> i915_perf_stream *stream)
>> * bit."
>> */
>> intel_uncore_write(uncore, GEN8_OABUFFER, gtt_offset |
>> - OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT);
>> + OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT |
>> + GEN8_OABUFFER_EDGE_TRIGGER);
>> intel_uncore_write(uncore, GEN8_OATAILPTR, gtt_offset &
>> GEN8_OATAILPTR_MASK);
>>
>> /* Mark that we need updated tail pointers to read from... */
>> @@ -1503,7 +1504,8 @@ static void gen12_init_oa_buffer(struct
>> i915_perf_stream *stream)
>> * bit."
>> */
>> intel_uncore_write(uncore, GEN12_OAG_OABUFFER, gtt_offset |
>> - OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT);
>> + OABUFFER_SIZE_16M | GEN8_OABUFFER_MEM_SELECT_GGTT |
>> + GEN12_OAG_OABUFFER_EDGE_TRIGGER);
>> intel_uncore_write(uncore, GEN12_OAG_OATAILPTR,
>> gtt_offset & GEN12_OAG_OATAILPTR_MASK);
>>
>> @@ -4481,8 +4483,10 @@ int i915_perf_ioctl_version(void)
>> *
>> * 5: Add DRM_I915_PERF_PROP_POLL_OA_PERIOD parameter that
>> controls the
>> * interval for the hrtimer used to check for OA data.
>> + *
>> + * 6. Add edge trigger report generation support.
>> */
>> - return 5;
>> + return 6;
>
> Do you think we should be adding a comment in uapi for revision 6?
>
> If not, this patch looks good and is:
>
> Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
>
> Thanks,
> Umesh
Not sure what to do to be fair. The command parser version details are
only in the source file, nothing in the header for instance.
-Lionel
>
>> }
>>
>> #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
>> diff --git a/drivers/gpu/drm/i915/i915_reg.h
>> b/drivers/gpu/drm/i915/i915_reg.h
>> index 17484345cb80..813ae77ffeb7 100644
>> --- a/drivers/gpu/drm/i915/i915_reg.h
>> +++ b/drivers/gpu/drm/i915/i915_reg.h
>> @@ -661,6 +661,7 @@ static inline bool i915_mmio_reg_valid(i915_reg_t
>> reg)
>>
>> #define GEN8_OABUFFER_UDW _MMIO(0x23b4)
>> #define GEN8_OABUFFER _MMIO(0x2b14)
>> +#define GEN8_OABUFFER_EDGE_TRIGGER (1 << 2)
>> #define GEN8_OABUFFER_MEM_SELECT_GGTT (1 << 0) /* 0: PPGTT, 1:
>> GGTT */
>>
>> #define GEN7_OASTATUS1 _MMIO(0x2364)
>> @@ -712,6 +713,7 @@ static inline bool i915_mmio_reg_valid(i915_reg_t
>> reg)
>> #define GEN12_OAG_OABUFFER _MMIO(0xdb08)
>> #define GEN12_OAG_OABUFFER_BUFFER_SIZE_MASK (0x7)
>> #define GEN12_OAG_OABUFFER_BUFFER_SIZE_SHIFT (3)
>> +#define GEN12_OAG_OABUFFER_EDGE_TRIGGER (1 << 2)
>> #define GEN12_OAG_OABUFFER_MEMORY_SELECT (1 << 0) /* 0: PPGTT,
>> 1: GGTT */
>>
>> #define GEN12_OAG_OAGLBCTXCTRL _MMIO(0x2b28)
>> --
>> 2.26.0
>>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
prev parent reply other threads:[~2020-04-02 6:08 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-31 11:46 [Intel-gfx] [PATCH] drm/i915/perf: Enable application triggered OA reports Lionel Landwerlin
2020-03-31 18:44 ` [Intel-gfx] ✗ Fi.CI.BAT: failure for " Patchwork
2020-04-01 18:43 ` [Intel-gfx] [PATCH] " Umesh Nerlige Ramappa
2020-04-02 6:07 ` Lionel Landwerlin [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2f5b1a30-96a5-c42e-86da-7d300d45fddd@intel.com \
--to=lionel.g.landwerlin@intel.com \
--cc=intel-gfx@lists.freedesktop.org \
--cc=umesh.nerlige.ramappa@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox