Tags

Linux 3.18.37

Linux 4.1.28

This is the 4.4.15 stable release

This is the 4.6.4 stable release

- select igt testing depencies for CONFIG_DRM_I915_DEBUG (Chris)
- track outputs in crtc state and clean up all our ad-hoc connector/encoder
  walking in modest code (Ville)
- demidlayer drm_device/drm_i915_private (Chris Wilson)
- thundering herd fix from Chris Wilson, with lots of help from Tvrtko Ursulin
- piles of assorted clean and fallout from the thundering herd fix
- documentation and more tuning for waitboosting (Chris)
- pooled EU support on bxt (Arun Siluvery)
- bxt support is no longer considered prelimary!
- ring/engine vfunc cleanup from Tvrtko
- introduce intel_wait_for_register helper (Chris)
- opregion updates (Jani Nukla)
- tuning and fixes for wait_for macros (Tvrkto&Imre)
- more kabylake pci ids (Rodrigo)
- pps cleanup and fixes for bxt (Imre)
- move sink crc support over to atomic state (Maarten)
- fix up async fbdev init ordering (Chris)
- fbc fixes from Paulo and Chris

Linux 4.7-rc7

IOMMU Fixes for Linux v4.7-rc6

Two Fixes:

	* Intel VT-d fix for a suspend/resume issue, introduced with the
	  scalability improvements in this cycle.

	* AMD IOMMU fix for systems that have unity mappings defined. There was
	  a race where translation got enabled before the unity mappings were
	  in place. This issue was seen on some HP servers.

Qualcomm ARM Based Driver Updates for v4.8 Part 2

* Fix probe order issue in SCM
* Add missing qcom_scm_is_available() API

Linux 4.7-rc6

IOMMU Fixes for Linux v4.7-rc5

Three fixes:

	* Fix use of smp_processor_id() in preemptible code in the IOVA
	  allocation code. This got introduced with the scalability
	  improvements in this release cycle.

	* A VT-d fix for out-of-bounds access of the iommu->domains
	  array. The bug showed during suspend/resume.

	* AMD IOMMU fix to print the correct device id in the ACPI
	  parsing code.