linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: kan.liang@linux.intel.com
To: peterz@infradead.org, mingo@redhat.com,
	gregkh@linuxfoundation.org, acme@kernel.org,
	linux-kernel@vger.kernel.org
Cc: eranian@google.com, namhyung@kernel.org, jolsa@redhat.com,
	ak@linux.intel.com, yao.jin@linux.intel.com,
	Kan Liang <kan.liang@linux.intel.com>
Subject: [PATCH V3 0/6] perf: Add Sapphire Rapids server uncore support
Date: Tue, 29 Jun 2021 11:13:57 -0700	[thread overview]
Message-ID: <1624990443-168533-1-git-send-email-kan.liang@linux.intel.com> (raw)

From: Kan Liang <kan.liang@linux.intel.com>

Changes since V2:
- Use sysfs_emit() for the new sysfs attribute "alias" in patch 2.
- Use ATTRIBUTE_GROUPS() for uncore_alias in patch 2

Changes since V1:
- Drop the symlink between two PMU names. Add a new attribute "alias"
  to indicate the old numeric name.
- Add document and perf tool support for the new attribute "alias".
(I once planned to use "caps/pmu_name" as the new attribute name in V1's
discussion.
https://lore.kernel.org/lkml/1edc6666-c4b3-ebf5-964a-e5ffc97d8563@linux.intel.com/
The name has a different meaning in the CPU PMU, which may causes
confusion. So the "alias" is used in V2.)
- The number of the free-running counter units should be the same as
  the max index of the relative units plus one. Correct the number in
  the patch 4.

Intel Sapphire Rapids supports a discovery mechanism, that allows an
uncore driver to discover the different components ("boxes") of the
chip.

All the generic information of the uncore boxes should be retrieved from
the discovery tables. This has been enabled with the commit edae1f06c2cd
("perf/x86/intel/uncore: Parse uncore discovery tables"). The uncore
driver doesn't need to hard code the generic information for each uncore
box. But we still need to enable various functionality that cannot be
directly discovered. This is done in the patchset.

Without this platform-specific enabling patch set, perf uses a type ID
plus a box ID, e.g., uncore_type_0_0 to name an uncore PMU. With the
patch set, perf has the mapping information from a type ID to a specific
uncore unit. Just like the previous platforms, the uncore PMU can be
named by the real PMU name, e.g., uncore_cha_0. Add a new attribute
"alias" to indicate the old numeric name. The userspace perf tool is
modified in the patch-set to handle both names.

The uncore spec of Sapphire Rapids server can be found at
https://cdrdv2.intel.com/v1/dl/getContent/642245

Kan Liang (6):
  perf/x86/intel/uncore: Add Sapphire Rapids server support
  perf/x86/intel/uncore: Add alias PMU name
  perf/x86/intel/uncore: Factor out snr_uncore_mmio_map()
  perf/x86/intel/uncore: Support free-running counters on Sapphire
    Rapids server
  perf/x86/intel/uncore: Fix invalid unit check
  perf pmu: Add PMU alias support

 arch/x86/events/intel/uncore.c           |  45 ++-
 arch/x86/events/intel/uncore.h           |   4 +
 arch/x86/events/intel/uncore_discovery.c |  42 +--
 arch/x86/events/intel/uncore_discovery.h |  23 +-
 arch/x86/events/intel/uncore_snbep.c     | 543 ++++++++++++++++++++++++++++++-
 tools/perf/arch/x86/util/pmu.c           | 129 +++++++-
 tools/perf/util/parse-events.y           |   4 +-
 tools/perf/util/pmu.c                    |  23 +-
 tools/perf/util/pmu.h                    |   5 +
 9 files changed, 770 insertions(+), 48 deletions(-)

-- 
2.7.4


             reply	other threads:[~2021-06-29 18:28 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-29 18:13 kan.liang [this message]
2021-06-29 18:13 ` [PATCH V3 1/6] perf/x86/intel/uncore: Add Sapphire Rapids server support kan.liang
2021-06-30  9:36   ` Greg KH
2021-06-30 12:56     ` Liang, Kan
2021-06-30 13:05       ` Greg KH
2021-06-30  9:39   ` Greg KH
2021-06-30 13:06     ` Liang, Kan
2021-06-29 18:13 ` [PATCH V3 2/6] perf/x86/intel/uncore: Add alias PMU name kan.liang
2021-06-30  9:39   ` Greg KH
2021-06-30 13:06     ` Liang, Kan
2021-06-29 18:14 ` [PATCH V3 3/6] perf/x86/intel/uncore: Factor out snr_uncore_mmio_map() kan.liang
2021-06-29 18:14 ` [PATCH V3 4/6] perf/x86/intel/uncore: Support free-running counters on Sapphire Rapids server kan.liang
2021-06-29 18:14 ` [PATCH V3 5/6] perf/x86/intel/uncore: Fix invalid unit check kan.liang
2021-06-30  9:36   ` Greg KH
2021-06-30 12:54     ` Liang, Kan
2021-06-29 18:14 ` [PATCH V3 6/6] perf pmu: Add PMU alias support kan.liang

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=1624990443-168533-1-git-send-email-kan.liang@linux.intel.com \
    --to=kan.liang@linux.intel.com \
    --cc=acme@kernel.org \
    --cc=ak@linux.intel.com \
    --cc=eranian@google.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=jolsa@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=peterz@infradead.org \
    --cc=yao.jin@linux.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;
as well as URLs for NNTP newsgroup(s).