1. 04 Feb, 2019 1 commit
    • Thierry Reding's avatar
      gpu: host1x: Set up stream ID table · 6841482b
      Thierry Reding authored
      In order to enable the MMIO path stream ID protection provided by the
      incarnation of host1x found in Tegra186 and later, the host1x must be
      provided with the list of stream ID register offsets for each of its
      clients. Some clients (such as VIC) have multiple stream ID registers
      that are assumed to be contiguous. The host1x is programmed with the
      base offset and a limit which provide the range of registers that the
      host1x needs to monitor for writes.
      Signed-off-by: default avatarThierry Reding <treding@nvidia.com>
  2. 18 May, 2018 1 commit
  3. 21 Dec, 2017 1 commit
    • Thierry Reding's avatar
      gpu: host1x: Use IOMMU groups · 41c3068c
      Thierry Reding authored
      Use IOMMU groups to attach the host1x device to its IOMMU domain. This
      is not strictly necessary because the domain isn't shared with any other
      device, but it makes the code consistent with how IOMMU is handled in
      other drivers and provides an easy way to detect when no IOMMU has been
      attached via device tree.
      Signed-off-by: default avatarThierry Reding <treding@nvidia.com>
  4. 20 Oct, 2017 2 commits
  5. 15 Jun, 2017 1 commit
  6. 05 Apr, 2017 3 commits
  7. 11 Nov, 2016 1 commit
  8. 23 Jun, 2016 3 commits
    • Thierry Reding's avatar
      gpu: host1x: Use unsigned int consistently for IDs · 5c0d8d38
      Thierry Reding authored
      IDs can never be negative so use unsigned int. In some instances an
      explicitly sized type (such as u32) was used for no particular reason,
      so turn those into unsigned int as well for consistency.
      Signed-off-by: default avatarThierry Reding <treding@nvidia.com>
    • Thierry Reding's avatar
      gpu: host1x: Consistently use unsigned int for counts · 14c95fc8
      Thierry Reding authored
      The number of channels, syncpoints, bases and mlocks can never be
      negative, so use unsigned int instead of int. Also make loop variables
      the same type for consistency.
      Signed-off-by: default avatarThierry Reding <treding@nvidia.com>
    • Bhaktipriya Shridhar's avatar
      gpu: host1x: hw: intr_hw: Remove create_workqueue · 57574bd7
      Bhaktipriya Shridhar authored
      System workqueues have been able to handle high level of concurrency
      for a long time now and there's no reason to use dedicated workqueues
      just to gain concurrency. Since the workqueue host->intr_wq is involved
      in sync point interrupts, and sync point wait and is not being used on
      a memory reclaim path, dedicated host->intr_wq has been replaced with the
      use of system_wq.
      Unlike a dedicated per-cpu workqueue created with create_workqueue(),
      system_wq allows multiple work items to overlap executions even on
      the same CPU; however, a per-cpu workqueue doesn't have any CPU
      locality or global ordering guarantees unless the target CPU is
      explicitly specified and thus the increase of local concurrency
      shouldn't make any difference.
      cancel_work_sync() has been used  in _host1x_free_syncpt_irq() to ensure
      that no work is pending by the time exit path runs.
      Signed-off-by: default avatarBhaktipriya Shridhar <bhaktipriya96@gmail.com>
      Acked-by: default avatarTejun Heo <tj@kernel.org>
      Signed-off-by: default avatarThierry Reding <treding@nvidia.com>
  9. 04 Mar, 2016 1 commit
  10. 19 Dec, 2013 1 commit
  11. 31 Oct, 2013 2 commits
  12. 03 Sep, 2013 1 commit
  13. 22 Jun, 2013 1 commit
  14. 22 Apr, 2013 5 commits