public inbox for igt-dev@lists.freedesktop.org
 help / color / mirror / Atom feed
* [PATCH i-g-t v10 0/2] tests/intel/xe_pm_residency: Add ASPM Link residency test
@ 2026-01-21 14:57 Karthik Poosa
  2026-01-21 14:57 ` [PATCH i-g-t v10 1/2] lib/igt_device: Add API to get pci device upstream port Karthik Poosa
                   ` (5 more replies)
  0 siblings, 6 replies; 15+ messages in thread
From: Karthik Poosa @ 2026-01-21 14:57 UTC (permalink / raw)
  To: igt-dev
  Cc: anshuman.gupta, badal.nilawar, riana.tauro, rodrigo.vivi,
	kamil.konieczny, Karthik Poosa

Add subtest aspm_link_residency to xe_pm_residency test to check ASPM Link state residency.
This uses the debugfs /sys/kernel/debug/dri/0/dgfx_pcie_link_residencies to read the
residency counters.
Add an API igt_device_get_pci_usp() to get pcie device's upstream port.

v2:
 - Avoid using igt_require() in igt_device_get_pci_device. (Kamil)
 - Add description about PCIe ASPM in commit message and code. (Kamil)
 - Resolve couple of compilation warnings about using variable as format string to sscanf.

v3:
 - Rename igt_device_get_pci_usp() to igt_device_get_pci_upstream_port(). (Kamil)

v4:
 - Refactor and enhance readability. (Badal)
 - Move save and restore of link states to separate functions. (Badal)

v5:
 - Use device type from pci express capability of pci config space to
   identify upstream port. (Badal)

v6:
 - Search for upstream port only two levels up of the endpoint as we want to
   verify ASPM only on PCIe link on which GPU card is present.
 - Skip test for integrated platforms.

v7:
 - Address below review comments from Riana.
 - Use igt_sysfs_has_attr() instead of faccess().
 - Remove unnecessary spaces, debug logs, if checks.
 - Wrap line length to 100 chars.
 - Use spinner instead of mmap as workload.

v8:
 - Address minor review comments from Kamil.

v9:
 - Add exit handler for aspm_link_residency test. (Riana)
 - Remove unused and unnecessary variables.

Karthik Poosa (2):
  lib/igt_device: Add API to get pci device upstream port
  tests/intel/xe_pm_residency: Add subtest for ASPM Link state residency

 lib/igt_device.c              |  43 ++++++++
 lib/igt_device.h              |   1 +
 lib/igt_pci.h                 |   2 +
 tests/intel/xe_pm_residency.c | 196 ++++++++++++++++++++++++++++++++++
 4 files changed, 242 insertions(+)

-- 
2.25.1


^ permalink raw reply	[flat|nested] 15+ messages in thread
* [PATCH i-g-t v10 0/2] tests/intel/xe_pm_residency: Add ASPM Link residency test
@ 2026-01-06 13:51 Karthik Poosa
  2026-01-06 13:51 ` [PATCH i-g-t v10 2/2] tests/intel/xe_pm_residency: Add subtest for ASPM Link state residency Karthik Poosa
  0 siblings, 1 reply; 15+ messages in thread
From: Karthik Poosa @ 2026-01-06 13:51 UTC (permalink / raw)
  To: igt-dev
  Cc: anshuman.gupta, badal.nilawar, riana.tauro, rodrigo.vivi,
	kamil.konieczny, Karthik Poosa

Add subtest aspm_link_residency to xe_pm_residency test to check ASPM Link state residency.
This uses the debugfs /sys/kernel/debug/dri/0/dgfx_pcie_link_residencies to read the
residency counters.
Add an API igt_device_get_pci_usp() to get pcie device's upstream port.

v2:
 - Avoid using igt_require() in igt_device_get_pci_device. (Kamil)
 - Add description about PCIe ASPM in commit message and code. (Kamil)
 - Resolve couple of compilation warnings about using variable as format string to sscanf.

v3:
 - Rename igt_device_get_pci_usp() to igt_device_get_pci_upstream_port(). (Kamil)

v4:
 - Refactor and enhance readability. (Badal)
 - Move save and restore of link states to separate functions. (Badal)

v5:
 - Use device type from pci express capability of pci config space to
   identify upstream port. (Badal)

v6:
 - Search for upstream port only two levels up of the endpoint as we want to
   verify ASPM only on PCIe link on which GPU card is present.
 - Skip test for integrated platforms.

v7:
 - Address below review comments from Riana.
 - Use igt_sysfs_has_attr() instead of faccess().
 - Remove unnecessary spaces, debug logs, if checks.
 - Wrap line length to 100 chars.
 - Use spinner instead of mmap as workload.

v8:
 - Address minor review comments from Kamil.

Karthik Poosa (2):
  lib/igt_device: Add API to get pci device upstream port
  tests/intel/xe_pm_residency: Add subtest for ASPM Link state residency

 lib/igt_device.c              |  43 +++++++++
 lib/igt_device.h              |   1 +
 lib/igt_pci.h                 |   2 +
 tests/intel/xe_pm_residency.c | 176 ++++++++++++++++++++++++++++++++++
 4 files changed, 222 insertions(+)

-- 
2.25.1


^ permalink raw reply	[flat|nested] 15+ messages in thread

end of thread, other threads:[~2026-02-04  5:47 UTC | newest]

Thread overview: 15+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-01-21 14:57 [PATCH i-g-t v10 0/2] tests/intel/xe_pm_residency: Add ASPM Link residency test Karthik Poosa
2026-01-21 14:57 ` [PATCH i-g-t v10 1/2] lib/igt_device: Add API to get pci device upstream port Karthik Poosa
2026-01-21 14:57 ` [PATCH i-g-t v10 2/2] tests/intel/xe_pm_residency: Add subtest for ASPM Link state residency Karthik Poosa
2026-01-22 15:00   ` Kamil Konieczny
2026-01-23 10:34     ` Poosa, Karthik
2026-01-23 12:07       ` Kamil Konieczny
2026-01-23 12:35         ` Poosa, Karthik
2026-01-23 14:06           ` Kamil Konieczny
2026-02-04  5:46             ` Poosa, Karthik
2026-01-21 16:55 ` ✗ Xe.CI.BAT: failure for tests/intel/xe_pm_residency: Add ASPM Link residency test Patchwork
2026-01-21 17:15 ` ✓ i915.CI.BAT: success " Patchwork
2026-01-22  2:05 ` ✗ Xe.CI.Full: failure " Patchwork
2026-01-22  3:52 ` ✗ i915.CI.Full: " Patchwork
  -- strict thread matches above, loose matches on Subject: below --
2026-01-06 13:51 [PATCH i-g-t v10 0/2] " Karthik Poosa
2026-01-06 13:51 ` [PATCH i-g-t v10 2/2] tests/intel/xe_pm_residency: Add subtest for ASPM Link state residency Karthik Poosa
2026-01-07  6:24   ` Riana Tauro

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox