public inbox for igt-dev@lists.freedesktop.org
 help / color / mirror / Atom feed
From: "Dixit, Ashutosh" <ashutosh.dixit@intel.com>
To: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Cc: <igt-dev@lists.freedesktop.org>, <jose.souza@intel.com>,
	<sushma.venkatesh.reddy@intel.com>
Subject: Re: [PATCH i-g-t 1/2] lib/xe/oa: Add CRI metrics XML
Date: Mon, 30 Mar 2026 14:56:18 -0700	[thread overview]
Message-ID: <87y0j9x9q5.wl-ashutosh.dixit@intel.com> (raw)
In-Reply-To: <acrU97JYHHwRvOXh@soc-5CG1426VCC.clients.intel.com>

On Mon, 30 Mar 2026 12:54:31 -0700, Umesh Nerlige Ramappa wrote:
>

Hi Umesh,

> > diff --git a/lib/xe/oa-configs/oa-metricset-codegen.py b/lib/xe/oa-configs/oa-metricset-codegen.py
> > index 922fae58cb..c5c7a2c8d4 100644
> > --- a/lib/xe/oa-configs/oa-metricset-codegen.py
> > +++ b/lib/xe/oa-configs/oa-metricset-codegen.py
> > @@ -138,9 +138,9 @@ def generate_metric_sets(args, gen):
> >                     metric_set->c_offset = metric_set->b_offset + 8;
> >                     metric_set->perfcnt_offset = metric_set->c_offset + 8;
> >                 """))
> > -        elif gen.chipset == "lnl" or gen.chipset == "bmg" or gen.chipset == "ptl":
> > +        elif gen.chipset == "lnl" or gen.chipset == "bmg" or gen.chipset == "ptl" or gen.chipset == "cri":
> >             # See intel_xe_perf_accumulate_reports for the offsets
> > -            if set.oa_format == "128B_MPEC8_NOA16":
> > +            if set.oa_format == "128B_MPEC8_NOA16" or set.oa_format == "128B_MERT_PEC8":
>
> Since you are handling multiple platforms with this if block, I would
> change it to:
>
> if set.oa_format == "128B_MPEC8_NOA16" or
>   (set.oa_format == "128B_MERT_PEC8" and gen.chipset == "cri"):
>
> Rest looks good.
>
> >                 c(textwrap.dedent("""\
> >                     metric_set->perf_oa_format = XE_OAM_FORMAT_MPEC8u32_B8_C8;


I think this is fine as is. First, MERT may be present on other platforms
too (other than cri). So I am thinking the format is specific to MERT,
rather than to any particular platform. So we just say here that both
"128B_MERT_PEC8" and "128B_MPEC8_NOA16" formats in the XML's get mapped
into XE_OAM_FORMAT_MPEC8u32_B8_C8 in the code. So the formats themselves
are not platform specific. So I would like this to be format related,
rather than platform related.

So I think it is ok as is. Let me know what you think.

Thanks.
--
Ashutosh




> >
> > diff --git a/lib/xe/oa-configs/oa_guid_registry.py b/lib/xe/oa-configs/oa_guid_registry.py
> > index 55c4968161..75939cfa2b 100644
> > --- a/lib/xe/oa-configs/oa_guid_registry.py
> > +++ b/lib/xe/oa-configs/oa_guid_registry.py
> > @@ -104,6 +104,7 @@ class Registry:
> >                            'LNL',
> >                            'BMG',
> >                            'PTL',
> > +                           'CRI',
> >                           )
> >         if name in known_chipsets:
> >             return name
> > diff --git a/lib/xe/oa-configs/update-guids.py b/lib/xe/oa-configs/update-guids.py
> > index c3f5a37c6e..d55958a0c0 100755
> > --- a/lib/xe/oa-configs/update-guids.py
> > +++ b/lib/xe/oa-configs/update-guids.py
> > @@ -168,6 +168,7 @@ chipsets = [ 'hsw',
> >              'lnl',
> >              'bmg',
> >              'ptl',
> > +             'cri',
> > ]
> >
> > for chipset in chipsets:
> > --
> > 2.48.1
> >

  reply	other threads:[~2026-03-30 21:56 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-02-24  0:41 [PATCH i-g-t 0/2] lib/xe/oa: Add CRI metrics XML Ashutosh Dixit
2026-02-24  0:41 ` [PATCH i-g-t 1/2] " Ashutosh Dixit
2026-03-30 19:54   ` Umesh Nerlige Ramappa
2026-03-30 21:56     ` Dixit, Ashutosh [this message]
2026-04-01 20:59       ` Umesh Nerlige Ramappa
2026-02-24  0:41 ` [PATCH i-g-t 2/2] tests/intel/xe_oa: more CRI changes Ashutosh Dixit
2026-02-24  1:34 ` ✓ Xe.CI.BAT: success for lib/xe/oa: Add CRI metrics XML Patchwork
2026-02-24  1:51 ` ✗ i915.CI.BAT: failure " Patchwork
2026-02-24 16:45 ` ✗ Xe.CI.FULL: " Patchwork

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=87y0j9x9q5.wl-ashutosh.dixit@intel.com \
    --to=ashutosh.dixit@intel.com \
    --cc=igt-dev@lists.freedesktop.org \
    --cc=jose.souza@intel.com \
    --cc=sushma.venkatesh.reddy@intel.com \
    --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