1. 24 Jan, 2012 1 commit
  2. 06 Jan, 2012 1 commit
    • Jerome Glisse's avatar
      drm/radeon: GPU virtual memory support v22 · 721604a1
      Jerome Glisse authored
      
      
      Virtual address space are per drm client (opener of /dev/drm).
      Client are in charge of virtual address space, they need to
      map bo into it by calling DRM_RADEON_GEM_VA ioctl.
      
      First 16M of virtual address space is reserved by the kernel.
      
      Once using 2 level page table we should be able to have a small
      vram memory footprint for each pt (there would be one pt for all
      gart, one for all vram and then one first level for each virtual
      address space).
      
      Plan include using the sub allocator for a common vm page table
      area and using memcpy to copy vm page table in & out. Or use
      a gart object and copy things in & out using dma.
      
      v2: agd5f fixes:
      - Add vram base offset for vram pages.  The GPU physical address of a
      vram page is FB_OFFSET + page offset.  FB_OFFSET is 0 on discrete
      cards and the physical bus address of the stolen memory on
      integrated chips.
      - VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR covers all vmid's >= 1
      
      v3: agd5f:
      - integrate with the semaphore/multi-ring stuff
      
      v4:
      - rebase on top ttm dma & multi-ring stuff
      - userspace is now in charge of the address space
      - no more specific cs vm ioctl, instead cs ioctl has a new
        chunk
      
      v5:
      - properly handle mem == NULL case from move_notify callback
      - fix the vm cleanup path
      
      v6:
      - fix update of page table to only happen on valid mem placement
      
      v7:
      - add tlb flush for each vm context
      - add flags to define mapping property (readable, writeable, snooped)
      - make ring id implicit from ib->fence->ring, up to each asic callback
        to then do ring specific scheduling if vm ib scheduling function
      
      v8:
      - add query for ib limit and kernel reserved virtual space
      - rename vm->size to max_pfn (maximum number of page)
      - update gem_va ioctl to also allow unmap operation
      - bump kernel version to allow userspace to query for vm support
      
      v9:
      - rebuild page table only when bind and incrementaly depending
        on bo referenced by cs and that have been moved
      - allow virtual address space to grow
      - use sa allocator for vram page table
      - return invalid when querying vm limit on non cayman GPU
      - dump vm fault register on lockup
      
      v10: agd5f:
      - Move the vm schedule_ib callback to a standalone function, remove
        the callback and use the existing ib_execute callback for VM IBs.
      
      v11:
      - rebase on top of lastest Linus
      
      v12: agd5f:
      - remove spurious backslash
      - set IB vm_id to 0 in radeon_ib_get()
      
      v13: agd5f:
      - fix handling of RADEON_CHUNK_ID_FLAGS
      
      v14:
      - fix va destruction
      - fix suspend resume
      - forbid bo to have several different va in same vm
      
      v15:
      - rebase
      
      v16:
      - cleanup left over of vm init/fini
      
      v17: agd5f:
      - cs checker
      
      v18: agd5f:
      - reworks the CS ioctl to better support multiple rings and
      VM.  Rather than adding a new chunk id for VM, just re-use the
      IB chunk id and add a new flags for VM mode.  Also define additional
      dwords for the flags chunk id to define the what ring we want to use
      (gfx, compute, uvd, etc.) and the priority.
      
      v19:
      - fix cs fini in weird case of no ib
      - semi working flush fix for ni
      - rebase on top of sa allocator changes
      
      v20: agd5f:
      - further CS ioctl cleanups from Christian's comments
      
      v21: agd5f:
      - integrate CS checker improvements
      
      v22: agd5f:
      - final cleanups for release, only allow VM CS on cayman
      Signed-off-by: default avatarJerome Glisse <jglisse@redhat.com>
      Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
      Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
      721604a1
  3. 05 Jan, 2012 2 commits
  4. 20 Dec, 2011 6 commits
  5. 06 Dec, 2011 1 commit
  6. 11 Nov, 2011 1 commit
  7. 18 Oct, 2011 1 commit
  8. 10 Oct, 2011 2 commits
    • Alex Deucher's avatar
      drm/radeon/kms: set DMA mask properly on newer PCI asics · 005a83f1
      Alex Deucher authored
      
      
      If a card wasn't PCIE, we always set the DMA mask to 32 bits.
      This is only applies to the old rage128/r1xx gart block on
      early radeon asics (~r1xx-r4xx).  Newer PCI and IGP cards
      can handle 40 bits just fine.
      Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
      Cc: Chen Jie <chenj@lemote.com>
      Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
      005a83f1
    • Michael Witten's avatar
      DRM: bug: RADEON_DEBUGFS_MAX_{NUM_FILES => COMPONENTS} · c245cb9e
      Michael Witten authored
      
      
      The value of RADEON_DEBUGFS_MAX_NUM_FILES has been used to
      specify the size of an array, each element of which looks
      like this:
      
        struct radeon_debugfs {
                struct drm_info_list    *files;
                unsigned                num_files;
        };
      
      Consequently, the number of debugfs files may be much greater
      than RADEON_DEBUGFS_MAX_NUM_FILES, something that the current
      code ignores:
      
        if ((_radeon_debugfs_count + nfiles) > RADEON_DEBUGFS_MAX_NUM_FILES) {
                DRM_ERROR("Reached maximum number of debugfs files.\n");
                DRM_ERROR("Report so we increase RADEON_DEBUGFS_MAX_NUM_FILES.\n");
                return -EINVAL;
        }
      
      This commit fixes this make, and accordingly renames:
      
        RADEON_DEBUGFS_MAX_NUM_FILES
      
      to:
      
        RADEON_DEBUGFS_MAX_COMPONENTS
      Signed-off-by: default avatarMichael Witten <mfwitten@gmail.com>
      Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
      c245cb9e
  9. 22 Aug, 2011 1 commit
  10. 12 Aug, 2011 1 commit
  11. 04 Aug, 2011 1 commit
  12. 16 Jun, 2011 1 commit
  13. 08 Jun, 2011 1 commit
  14. 01 Jun, 2011 1 commit
  15. 23 May, 2011 1 commit
  16. 31 Mar, 2011 1 commit
  17. 02 Mar, 2011 1 commit
  18. 22 Feb, 2011 2 commits
  19. 13 Feb, 2011 1 commit
    • Dave Airlie's avatar
      drm/radeon: fix race between GPU reset and TTM delayed delete thread. · 8fd1b84c
      Dave Airlie authored
      
      
      My evergreen has been in a remote PC for week and reset has never once
      saved me from certain doom, I finally relocated to the box with a
      serial cable and noticed an oops when the GPU resets, and the TTM
      delayed delete thread tries to remove something from the GTT.
      
      This stops the delayed delete thread from executing across the GPU
      reset handler, and woot I can GPU reset now.
      Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
      8fd1b84c
  20. 25 Jan, 2011 1 commit
    • Torben Hohn's avatar
      console: rename acquire/release_console_sem() to console_lock/unlock() · ac751efa
      Torben Hohn authored
      
      
      The -rt patches change the console_semaphore to console_mutex.  As a
      result, a quite large chunk of the patches changes all
      acquire/release_console_sem() to acquire/release_console_mutex()
      
      This commit makes things use more neutral function names which dont make
      implications about the underlying lock.
      
      The only real change is the return value of console_trylock which is
      inverted from try_acquire_console_sem()
      
      This patch also paves the way to switching console_sem from a semaphore to
      a mutex.
      
      [akpm@linux-foundation.org: coding-style fixes]
      [akpm@linux-foundation.org: make console_trylock return 1 on success, per Geert]
      Signed-off-by: default avatarTorben Hohn <torbenh@gmx.de>
      Cc: Thomas Gleixner <tglx@tglx.de>
      Cc: Greg KH <gregkh@suse.de>
      Cc: Ingo Molnar <mingo@elte.hu>
      Cc: Geert Uytterhoeven <geert@linux-m68k.org>
      Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
      Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      ac751efa
  21. 06 Jan, 2011 2 commits
  22. 05 Jan, 2011 1 commit
    • Tejun Heo's avatar
      drm/radeon: use system_wq instead of dev_priv->wq · 32c87fca
      Tejun Heo authored
      
      
      With cmwq, there's no reason for radeon to use a dedicated workqueue.
      Drop dev_priv->wq and use system_wq instead.
      
      Because radeon_driver_irq_uninstall_kms() may be called from
      unsleepable context, the work items can't be flushed from there.
      Instead, init and flush from radeon_irq_kms_init/fini().
      
      While at it, simplify canceling/flushing of rdev->pm.dynpm_idle_work.
      Always initialize and sync cancel instead of being unnecessarily smart
      about it.
      Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      Acked-by: default avatarAlex Deucher <alexdeucher@gmail.com>
      Cc: dri-devel@lists.freedesktop.org
      Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
      32c87fca
  23. 04 Jan, 2011 2 commits
  24. 20 Dec, 2010 1 commit
  25. 15 Dec, 2010 1 commit
  26. 07 Dec, 2010 1 commit
  27. 22 Nov, 2010 1 commit
  28. 17 Nov, 2010 1 commit
  29. 05 Oct, 2010 2 commits
    • Alex Deucher's avatar
      drm/radeon/kms/r6xx+: use new style fencing (v3) · d0f8a854
      Alex Deucher authored
      On r6xx+ a newer fence mechanism was implemented to replace
      the old wait_until plus scratch regs setup.  A single EOP event
      will flush the destination caches, write a fence value, and generate
      an interrupt.  This is the recommended fence mechanism on r6xx+ asics.
      
      This requires my previous writeback patch.
      
      v2: fix typo that enabled event fence checking on all asics
      rather than just r6xx+.
      
      v3: properly enable EOP interrupts
      Should fix:
      https://bugs.freedesktop.org/show_bug.cgi?id=29972
      
      Signed-off-by: default avatarAlex Deucher <alexdeucher@gmail.com>
      Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
      d0f8a854
    • Alex Deucher's avatar
      drm/radeon/kms: enable writeback (v2) · 724c80e1
      Alex Deucher authored
      
      
      When writeback is enabled, the GPU shadows writes to certain
      registers into a buffer in memory.  The driver can then read
      the values from the shadow rather than reading back from the
      register across the bus.  Writeback can be disabled by setting
      the no_wb module param to 1.
      
      On r6xx/r7xx/evergreen, the following registers are shadowed:
      - CP scratch registers
      - CP read pointer
      - IH write pointer
      On r1xx-rr5xx, the following registers are shadowed:
      - CP scratch registers
      - CP read pointer
      
      v2:
      - Combine wb patches for r6xx-evergreen and r1xx-r5xx
      - Writeback is disabled on AGP boards since it tends to be
      unreliable on AGP using the gart.
      - Check radeon_wb_init return values properly.
      Signed-off-by: default avatarAlex Deucher <alexdeucher@gmail.com>
      Signed-off-by: default avatarDave Airlie <airlied@redhat.com>
      724c80e1