Merge branch 'drm-next' of git://people.freedesktop.org/~airlied/linux

Pull drm updates from Dave Airlie:
 "This is the main drm pull request for v4.2.

  I've one other new driver from freescale on my radar, it's been posted
  and reviewed, I'd just like to get someone to give it a last look, so
  maybe I'll send it or maybe I'll leave it.

  There is no major nouveau changes in here, Ben was working on
  something big, and we agreed it was a bit late, there wasn't anything
  else he considered urgent to merge.

  There might be another msm pull for some bits that are waiting on
  arm-soc, I'll see how we time it.

  This touches some "of" stuff, acks are in place except for the fixes
  to the build in various configs,t hat I just applied.

  Summary:

  New drivers:
      - virtio-gpu:
                KMS only pieces of driver for virtio-gpu in qemu.
                This is just the first part of this driver, enough to run
                unaccelerated userspace on. As qemu merges more we'll start
                adding the 3D features for the virgl 3d work.
      - amdgpu:
                a new driver from AMD to driver their newer GPUs. (VI+)
                It contains a new cleaner userspace API, and is a clean
                break from radeon moving forward, that AMD are going to
                concentrate on. It also contains a set of register headers
                auto generated from AMD internal database.

  core:
      - atomic modesetting API completed, enabled by default now.
      - Add support for mode_id blob to atomic ioctl to complete interface.
      - bunch of Displayport MST fixes
      - lots of misc fixes.

  panel:
      - new simple panels
      - fix some long-standing build issues with bridge drivers

  radeon:
      - VCE1 support
      - add a GPU reset counter for userspace
      - lots of fixes.

  amdkfd:
      - H/W debugger support module
      - static user-mode queues
      - support killing all the waves when a process terminates
      - use standard DECLARE_BITMAP

  i915:
      - Add Broxton support
      - S3, rotation support for Skylake
      - RPS booting tuning
      - CPT modeset sequence fixes
      - ns2501 dither support
      - enable cmd parser on haswell
      - cdclk handling fixes
      - gen8 dynamic pte allocation
      - lots of atomic conversion work

  exynos:
      - Add atomic modesetting support
      - Add iommu support
      - Consolidate drm driver initialization
      - and MIC, DECON and MIPI-DSI support for exynos5433

  omapdrm:
      - atomic modesetting support (fixes lots of things in rewrite)

  tegra:
      - DP aux transaction fixes
      - iommu support fix

  msm:
      - adreno a306 support
      - various dsi bits
      - various 64-bit fixes
      - NV12MT support

  rcar-du:
      - atomic and misc fixes

  sti:
      - fix HDMI timing complaince

  tilcdc:
      - use drm component API to access tda998x driver
      - fix module unloading

  qxl:
      - stability fixes"

* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (872 commits)
  drm/nouveau: Pause between setting gpu to D3hot and cutting the power
  drm/dp/mst: close deadlock in connector destruction.
  drm: Always enable atomic API
  drm/vgem: Set unique to "vgem"
  of: fix a build error to of_graph_get_endpoint_by_regs function
  drm/dp/mst: take lock around looking up the branch device on hpd irq
  drm/dp/mst: make sure mst_primary mstb is valid in work function
  of: add EXPORT_SYMBOL for of_graph_get_endpoint_by_regs
  ARM: dts: rename the clock of MIPI DSI 'pll_clk' to 'sclk_mipi'
  drm/atomic: Don't set crtc_state->enable manually
  drm/exynos: dsi: do not set TE GPIO direction by input
  drm/exynos: dsi: add support for MIC driver as a bridge
  drm/exynos: dsi: add support for Exynos5433
  drm/exynos: dsi: make use of array for clock access
  drm/exynos: dsi: make use of driver data for static values
  drm/exynos: dsi: add macros for register access
  drm/exynos: dsi: rename pll_clk to sclk_clk
  drm/exynos: mic: add MIC driver
  of: add helper for getting endpoint node of specific identifiers
  drm/exynos: add Exynos5433 decon driver
  ...

1 files changed, 2337 insertions, 0 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
new file mode 100644
index 000000000000..22866d1c3d69
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -0,0 +1,2337 @@
+/*
+ * Copyright 2008 Advanced Micro Devices, Inc.
+ * Copyright 2008 Red Hat Inc.
+ * Copyright 2009 Jerome Glisse.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Dave Airlie
+ *          Alex Deucher
+ *          Jerome Glisse
+ */
+#ifndef __AMDGPU_H__
+#define __AMDGPU_H__
+
+#include <linux/atomic.h>
+#include <linux/wait.h>
+#include <linux/list.h>
+#include <linux/kref.h>
+#include <linux/interval_tree.h>
+#include <linux/hashtable.h>
+#include <linux/fence.h>
+
+#include <ttm/ttm_bo_api.h>
+#include <ttm/ttm_bo_driver.h>
+#include <ttm/ttm_placement.h>
+#include <ttm/ttm_module.h>
+#include <ttm/ttm_execbuf_util.h>
+
+#include <drm/drm_gem.h>
+#include <drm/amdgpu_drm.h>
+
+#include "amd_shared.h"
+#include "amdgpu_family.h"
+#include "amdgpu_mode.h"
+#include "amdgpu_ih.h"
+#include "amdgpu_irq.h"
+#include "amdgpu_ucode.h"
+#include "amdgpu_gds.h"
+
+/*
+ * Modules parameters.
+ */
+extern int amdgpu_modeset;
+extern int amdgpu_vram_limit;
+extern int amdgpu_gart_size;
+extern int amdgpu_benchmarking;
+extern int amdgpu_testing;
+extern int amdgpu_audio;
+extern int amdgpu_disp_priority;
+extern int amdgpu_hw_i2c;
+extern int amdgpu_pcie_gen2;
+extern int amdgpu_msi;
+extern int amdgpu_lockup_timeout;
+extern int amdgpu_dpm;
+extern int amdgpu_smc_load_fw;
+extern int amdgpu_aspm;
+extern int amdgpu_runtime_pm;
+extern int amdgpu_hard_reset;
+extern unsigned amdgpu_ip_block_mask;
+extern int amdgpu_bapm;
+extern int amdgpu_deep_color;
+extern int amdgpu_vm_size;
+extern int amdgpu_vm_block_size;
+
+#define AMDGPU_MAX_USEC_TIMEOUT			100000	/* 100 ms */
+#define AMDGPU_FENCE_JIFFIES_TIMEOUT		(HZ / 2)
+/* AMDGPU_IB_POOL_SIZE must be a power of 2 */
+#define AMDGPU_IB_POOL_SIZE			16
+#define AMDGPU_DEBUGFS_MAX_COMPONENTS		32
+#define AMDGPUFB_CONN_LIMIT			4
+#define AMDGPU_BIOS_NUM_SCRATCH			8
+
+/* max number of rings */
+#define AMDGPU_MAX_RINGS			16
+#define AMDGPU_MAX_GFX_RINGS			1
+#define AMDGPU_MAX_COMPUTE_RINGS		8
+#define AMDGPU_MAX_VCE_RINGS			2
+
+/* number of hw syncs before falling back on blocking */
+#define AMDGPU_NUM_SYNCS			4
+
+/* hardcode that limit for now */
+#define AMDGPU_VA_RESERVED_SIZE			(8 << 20)
+
+/* hard reset data */
+#define AMDGPU_ASIC_RESET_DATA                  0x39d5e86b
+
+/* reset flags */
+#define AMDGPU_RESET_GFX			(1 << 0)
+#define AMDGPU_RESET_COMPUTE			(1 << 1)
+#define AMDGPU_RESET_DMA			(1 << 2)
+#define AMDGPU_RESET_CP				(1 << 3)
+#define AMDGPU_RESET_GRBM			(1 << 4)
+#define AMDGPU_RESET_DMA1			(1 << 5)
+#define AMDGPU_RESET_RLC			(1 << 6)
+#define AMDGPU_RESET_SEM			(1 << 7)
+#define AMDGPU_RESET_IH				(1 << 8)
+#define AMDGPU_RESET_VMC			(1 << 9)
+#define AMDGPU_RESET_MC				(1 << 10)
+#define AMDGPU_RESET_DISPLAY			(1 << 11)
+#define AMDGPU_RESET_UVD			(1 << 12)
+#define AMDGPU_RESET_VCE			(1 << 13)
+#define AMDGPU_RESET_VCE1			(1 << 14)
+
+/* CG block flags */
+#define AMDGPU_CG_BLOCK_GFX			(1 << 0)
+#define AMDGPU_CG_BLOCK_MC			(1 << 1)
+#define AMDGPU_CG_BLOCK_SDMA			(1 << 2)
+#define AMDGPU_CG_BLOCK_UVD			(1 << 3)
+#define AMDGPU_CG_BLOCK_VCE			(1 << 4)
+#define AMDGPU_CG_BLOCK_HDP			(1 << 5)
+#define AMDGPU_CG_BLOCK_BIF			(1 << 6)
+
+/* CG flags */
+#define AMDGPU_CG_SUPPORT_GFX_MGCG		(1 << 0)
+#define AMDGPU_CG_SUPPORT_GFX_MGLS		(1 << 1)
+#define AMDGPU_CG_SUPPORT_GFX_CGCG		(1 << 2)
+#define AMDGPU_CG_SUPPORT_GFX_CGLS		(1 << 3)
+#define AMDGPU_CG_SUPPORT_GFX_CGTS		(1 << 4)
+#define AMDGPU_CG_SUPPORT_GFX_CGTS_LS		(1 << 5)
+#define AMDGPU_CG_SUPPORT_GFX_CP_LS		(1 << 6)
+#define AMDGPU_CG_SUPPORT_GFX_RLC_LS		(1 << 7)
+#define AMDGPU_CG_SUPPORT_MC_LS			(1 << 8)
+#define AMDGPU_CG_SUPPORT_MC_MGCG		(1 << 9)
+#define AMDGPU_CG_SUPPORT_SDMA_LS		(1 << 10)
+#define AMDGPU_CG_SUPPORT_SDMA_MGCG		(1 << 11)
+#define AMDGPU_CG_SUPPORT_BIF_LS		(1 << 12)
+#define AMDGPU_CG_SUPPORT_UVD_MGCG		(1 << 13)
+#define AMDGPU_CG_SUPPORT_VCE_MGCG		(1 << 14)
+#define AMDGPU_CG_SUPPORT_HDP_LS		(1 << 15)
+#define AMDGPU_CG_SUPPORT_HDP_MGCG		(1 << 16)
+
+/* PG flags */
+#define AMDGPU_PG_SUPPORT_GFX_PG		(1 << 0)
+#define AMDGPU_PG_SUPPORT_GFX_SMG		(1 << 1)
+#define AMDGPU_PG_SUPPORT_GFX_DMG		(1 << 2)
+#define AMDGPU_PG_SUPPORT_UVD			(1 << 3)
+#define AMDGPU_PG_SUPPORT_VCE			(1 << 4)
+#define AMDGPU_PG_SUPPORT_CP			(1 << 5)
+#define AMDGPU_PG_SUPPORT_GDS			(1 << 6)
+#define AMDGPU_PG_SUPPORT_RLC_SMU_HS		(1 << 7)
+#define AMDGPU_PG_SUPPORT_SDMA			(1 << 8)
+#define AMDGPU_PG_SUPPORT_ACP			(1 << 9)
+#define AMDGPU_PG_SUPPORT_SAMU			(1 << 10)
+
+/* GFX current status */
+#define AMDGPU_GFX_NORMAL_MODE			0x00000000L
+#define AMDGPU_GFX_SAFE_MODE			0x00000001L
+#define AMDGPU_GFX_PG_DISABLED_MODE		0x00000002L
+#define AMDGPU_GFX_CG_DISABLED_MODE		0x00000004L
+#define AMDGPU_GFX_LBPW_DISABLED_MODE		0x00000008L
+
+/* max cursor sizes (in pixels) */
+#define CIK_CURSOR_WIDTH 128
+#define CIK_CURSOR_HEIGHT 128
+
+struct amdgpu_device;
+struct amdgpu_fence;
+struct amdgpu_ib;
+struct amdgpu_vm;
+struct amdgpu_ring;
+struct amdgpu_semaphore;
+struct amdgpu_cs_parser;
+struct amdgpu_irq_src;
+
+enum amdgpu_cp_irq {
+	AMDGPU_CP_IRQ_GFX_EOP = 0,
+	AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE0_EOP,
+	AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE1_EOP,
+	AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE2_EOP,
+	AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE3_EOP,
+	AMDGPU_CP_IRQ_COMPUTE_MEC2_PIPE0_EOP,
+	AMDGPU_CP_IRQ_COMPUTE_MEC2_PIPE1_EOP,
+	AMDGPU_CP_IRQ_COMPUTE_MEC2_PIPE2_EOP,
+	AMDGPU_CP_IRQ_COMPUTE_MEC2_PIPE3_EOP,
+
+	AMDGPU_CP_IRQ_LAST
+};
+
+enum amdgpu_sdma_irq {
+	AMDGPU_SDMA_IRQ_TRAP0 = 0,
+	AMDGPU_SDMA_IRQ_TRAP1,
+
+	AMDGPU_SDMA_IRQ_LAST
+};
+
+enum amdgpu_thermal_irq {
+	AMDGPU_THERMAL_IRQ_LOW_TO_HIGH = 0,
+	AMDGPU_THERMAL_IRQ_HIGH_TO_LOW,
+
+	AMDGPU_THERMAL_IRQ_LAST
+};
+
+int amdgpu_set_clockgating_state(struct amdgpu_device *adev,
+				  enum amd_ip_block_type block_type,
+				  enum amd_clockgating_state state);
+int amdgpu_set_powergating_state(struct amdgpu_device *adev,
+				  enum amd_ip_block_type block_type,
+				  enum amd_powergating_state state);
+
+struct amdgpu_ip_block_version {
+	enum amd_ip_block_type type;
+	u32 major;
+	u32 minor;
+	u32 rev;
+	const struct amd_ip_funcs *funcs;
+};
+
+int amdgpu_ip_block_version_cmp(struct amdgpu_device *adev,
+				enum amd_ip_block_type type,
+				u32 major, u32 minor);
+
+const struct amdgpu_ip_block_version * amdgpu_get_ip_block(
+					struct amdgpu_device *adev,
+					enum amd_ip_block_type type);
+
+/* provided by hw blocks that can move/clear data.  e.g., gfx or sdma */
+struct amdgpu_buffer_funcs {
+	/* maximum bytes in a single operation */
+	uint32_t	copy_max_bytes;
+
+	/* number of dw to reserve per operation */
+	unsigned	copy_num_dw;
+
+	/* used for buffer migration */
+	void (*emit_copy_buffer)(struct amdgpu_ring *ring,
+				 /* src addr in bytes */
+				 uint64_t src_offset,
+				 /* dst addr in bytes */
+				 uint64_t dst_offset,
+				 /* number of byte to transfer */
+				 uint32_t byte_count);
+
+	/* maximum bytes in a single operation */
+	uint32_t	fill_max_bytes;
+
+	/* number of dw to reserve per operation */
+	unsigned	fill_num_dw;
+
+	/* used for buffer clearing */
+	void (*emit_fill_buffer)(struct amdgpu_ring *ring,
+				 /* value to write to memory */
+				 uint32_t src_data,
+				 /* dst addr in bytes */
+				 uint64_t dst_offset,
+				 /* number of byte to fill */
+				 uint32_t byte_count);
+};
+
+/* provided by hw blocks that can write ptes, e.g., sdma */
+struct amdgpu_vm_pte_funcs {
+	/* copy pte entries from GART */
+	void (*copy_pte)(struct amdgpu_ib *ib,
+			 uint64_t pe, uint64_t src,
+			 unsigned count);
+	/* write pte one entry at a time with addr mapping */
+	void (*write_pte)(struct amdgpu_ib *ib,
+			  uint64_t pe,
+			  uint64_t addr, unsigned count,
+			  uint32_t incr, uint32_t flags);
+	/* for linear pte/pde updates without addr mapping */
+	void (*set_pte_pde)(struct amdgpu_ib *ib,
+			    uint64_t pe,
+			    uint64_t addr, unsigned count,
+			    uint32_t incr, uint32_t flags);
+	/* pad the indirect buffer to the necessary number of dw */
+	void (*pad_ib)(struct amdgpu_ib *ib);
+};
+
+/* provided by the gmc block */
+struct amdgpu_gart_funcs {
+	/* flush the vm tlb via mmio */
+	void (*flush_gpu_tlb)(struct amdgpu_device *adev,
+			      uint32_t vmid);
+	/* write pte/pde updates using the cpu */
+	int (*set_pte_pde)(struct amdgpu_device *adev,
+			   void *cpu_pt_addr, /* cpu addr of page table */
+			   uint32_t gpu_page_idx, /* pte/pde to update */
+			   uint64_t addr, /* addr to write into pte/pde */
+			   uint32_t flags); /* access flags */
+};
+
+/* provided by the ih block */
+struct amdgpu_ih_funcs {
+	/* ring read/write ptr handling, called from interrupt context */
+	u32 (*get_wptr)(struct amdgpu_device *adev);
+	void (*decode_iv)(struct amdgpu_device *adev,
+			  struct amdgpu_iv_entry *entry);
+	void (*set_rptr)(struct amdgpu_device *adev);
+};
+
+/* provided by hw blocks that expose a ring buffer for commands */
+struct amdgpu_ring_funcs {
+	/* ring read/write ptr handling */
+	u32 (*get_rptr)(struct amdgpu_ring *ring);
+	u32 (*get_wptr)(struct amdgpu_ring *ring);
+	void (*set_wptr)(struct amdgpu_ring *ring);
+	/* validating and patching of IBs */
+	int (*parse_cs)(struct amdgpu_cs_parser *p, uint32_t ib_idx);
+	/* command emit functions */
+	void (*emit_ib)(struct amdgpu_ring *ring,
+			struct amdgpu_ib *ib);
+	void (*emit_fence)(struct amdgpu_ring *ring, uint64_t addr,
+			   uint64_t seq, unsigned flags);
+	bool (*emit_semaphore)(struct amdgpu_ring *ring,
+			       struct amdgpu_semaphore *semaphore,
+			       bool emit_wait);
+	void (*emit_vm_flush)(struct amdgpu_ring *ring, unsigned vm_id,
+			      uint64_t pd_addr);
+	void (*emit_hdp_flush)(struct amdgpu_ring *ring);
+	void (*emit_gds_switch)(struct amdgpu_ring *ring, uint32_t vmid,
+				uint32_t gds_base, uint32_t gds_size,
+				uint32_t gws_base, uint32_t gws_size,
+				uint32_t oa_base, uint32_t oa_size);
+	/* testing functions */
+	int (*test_ring)(struct amdgpu_ring *ring);
+	int (*test_ib)(struct amdgpu_ring *ring);
+	bool (*is_lockup)(struct amdgpu_ring *ring);
+};
+
+/*
+ * BIOS.
+ */
+bool amdgpu_get_bios(struct amdgpu_device *adev);
+bool amdgpu_read_bios(struct amdgpu_device *adev);
+
+/*
+ * Dummy page
+ */
+struct amdgpu_dummy_page {
+	struct page	*page;
+	dma_addr_t	addr;
+};
+int amdgpu_dummy_page_init(struct amdgpu_device *adev);
+void amdgpu_dummy_page_fini(struct amdgpu_device *adev);
+
+
+/*
+ * Clocks
+ */
+
+#define AMDGPU_MAX_PPLL 3
+
+struct amdgpu_clock {
+	struct amdgpu_pll ppll[AMDGPU_MAX_PPLL];
+	struct amdgpu_pll spll;
+	struct amdgpu_pll mpll;
+	/* 10 Khz units */
+	uint32_t default_mclk;
+	uint32_t default_sclk;
+	uint32_t default_dispclk;
+	uint32_t current_dispclk;
+	uint32_t dp_extclk;
+	uint32_t max_pixel_clock;
+};
+
+/*
+ * Fences.
+ */
+struct amdgpu_fence_driver {
+	struct amdgpu_ring		*ring;
+	uint64_t			gpu_addr;
+	volatile uint32_t		*cpu_addr;
+	/* sync_seq is protected by ring emission lock */
+	uint64_t			sync_seq[AMDGPU_MAX_RINGS];
+	atomic64_t			last_seq;
+	bool				initialized;
+	bool				delayed_irq;
+	struct amdgpu_irq_src		*irq_src;
+	unsigned			irq_type;
+	struct delayed_work             lockup_work;
+};
+
+/* some special values for the owner field */
+#define AMDGPU_FENCE_OWNER_UNDEFINED	((void*)0ul)
+#define AMDGPU_FENCE_OWNER_VM		((void*)1ul)
+#define AMDGPU_FENCE_OWNER_MOVE		((void*)2ul)
+
+#define AMDGPU_FENCE_FLAG_64BIT         (1 << 0)
+#define AMDGPU_FENCE_FLAG_INT           (1 << 1)
+
+struct amdgpu_fence {
+	struct fence base;
+
+	/* RB, DMA, etc. */
+	struct amdgpu_ring		*ring;
+	uint64_t			seq;
+
+	/* filp or special value for fence creator */
+	void				*owner;
+
+	wait_queue_t			fence_wake;
+};
+
+struct amdgpu_user_fence {
+	/* write-back bo */
+	struct amdgpu_bo 	*bo;
+	/* write-back address offset to bo start */
+	uint32_t                offset;
+};
+
+int amdgpu_fence_driver_init(struct amdgpu_device *adev);
+void amdgpu_fence_driver_fini(struct amdgpu_device *adev);
+void amdgpu_fence_driver_force_completion(struct amdgpu_device *adev);
+
+void amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring);
+int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
+				   struct amdgpu_irq_src *irq_src,
+				   unsigned irq_type);
+int amdgpu_fence_emit(struct amdgpu_ring *ring, void *owner,
+		      struct amdgpu_fence **fence);
+void amdgpu_fence_process(struct amdgpu_ring *ring);
+int amdgpu_fence_wait_next(struct amdgpu_ring *ring);
+int amdgpu_fence_wait_empty(struct amdgpu_ring *ring);
+unsigned amdgpu_fence_count_emitted(struct amdgpu_ring *ring);
+
+bool amdgpu_fence_signaled(struct amdgpu_fence *fence);
+int amdgpu_fence_wait(struct amdgpu_fence *fence, bool interruptible);
+int amdgpu_fence_wait_any(struct amdgpu_device *adev,
+			  struct amdgpu_fence **fences,
+			  bool intr);
+long amdgpu_fence_wait_seq_timeout(struct amdgpu_device *adev,
+				   u64 *target_seq, bool intr,
+				   long timeout);
+struct amdgpu_fence *amdgpu_fence_ref(struct amdgpu_fence *fence);
+void amdgpu_fence_unref(struct amdgpu_fence **fence);
+
+bool amdgpu_fence_need_sync(struct amdgpu_fence *fence,
+			    struct amdgpu_ring *ring);
+void amdgpu_fence_note_sync(struct amdgpu_fence *fence,
+			    struct amdgpu_ring *ring);
+
+static inline struct amdgpu_fence *amdgpu_fence_later(struct amdgpu_fence *a,
+						      struct amdgpu_fence *b)
+{
+	if (!a) {
+		return b;
+	}
+
+	if (!b) {
+		return a;
+	}
+
+	BUG_ON(a->ring != b->ring);
+
+	if (a->seq > b->seq) {
+		return a;
+	} else {
+		return b;
+	}
+}
+
+static inline bool amdgpu_fence_is_earlier(struct amdgpu_fence *a,
+					   struct amdgpu_fence *b)
+{
+	if (!a) {
+		return false;
+	}
+
+	if (!b) {
+		return true;
+	}
+
+	BUG_ON(a->ring != b->ring);
+
+	return a->seq < b->seq;
+}
+
+int amdgpu_user_fence_emit(struct amdgpu_ring *ring, struct amdgpu_user_fence *user, 
+			   void *owner, struct amdgpu_fence **fence);
+
+/*
+ * TTM.
+ */
+struct amdgpu_mman {
+	struct ttm_bo_global_ref        bo_global_ref;
+	struct drm_global_reference	mem_global_ref;
+	struct ttm_bo_device		bdev;
+	bool				mem_global_referenced;
+	bool				initialized;
+
+#if defined(CONFIG_DEBUG_FS)
+	struct dentry			*vram;
+	struct dentry			*gtt;
+#endif
+
+	/* buffer handling */
+	const struct amdgpu_buffer_funcs	*buffer_funcs;
+	struct amdgpu_ring			*buffer_funcs_ring;
+};
+
+int amdgpu_copy_buffer(struct amdgpu_ring *ring,
+		       uint64_t src_offset,
+		       uint64_t dst_offset,
+		       uint32_t byte_count,
+		       struct reservation_object *resv,
+		       struct amdgpu_fence **fence);
+int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma);
+
+struct amdgpu_bo_list_entry {
+	struct amdgpu_bo		*robj;
+	struct ttm_validate_buffer	tv;
+	struct amdgpu_bo_va		*bo_va;
+	unsigned			prefered_domains;
+	unsigned			allowed_domains;
+	uint32_t			priority;
+};
+
+struct amdgpu_bo_va_mapping {
+	struct list_head		list;
+	struct interval_tree_node	it;
+	uint64_t			offset;
+	uint32_t			flags;
+};
+
+/* bo virtual addresses in a specific vm */
+struct amdgpu_bo_va {
+	/* protected by bo being reserved */
+	struct list_head		bo_list;
+	uint64_t			addr;
+	struct amdgpu_fence		*last_pt_update;
+	unsigned			ref_count;
+
+	/* protected by vm mutex */
+	struct list_head		mappings;
+	struct list_head		vm_status;
+
+	/* constant after initialization */
+	struct amdgpu_vm		*vm;
+	struct amdgpu_bo		*bo;
+};
+
+#define AMDGPU_GEM_DOMAIN_MAX		0x3
+
+struct amdgpu_bo {
+	/* Protected by gem.mutex */
+	struct list_head		list;
+	/* Protected by tbo.reserved */
+	u32				initial_domain;
+	struct ttm_place		placements[AMDGPU_GEM_DOMAIN_MAX + 1];
+	struct ttm_placement		placement;
+	struct ttm_buffer_object	tbo;
+	struct ttm_bo_kmap_obj		kmap;
+	u64				flags;
+	unsigned			pin_count;
+	void				*kptr;
+	u64				tiling_flags;
+	u64				metadata_flags;
+	void				*metadata;
+	u32				metadata_size;
+	/* list of all virtual address to which this bo
+	 * is associated to
+	 */
+	struct list_head		va;
+	/* Constant after initialization */
+	struct amdgpu_device		*adev;
+	struct drm_gem_object		gem_base;
+
+	struct ttm_bo_kmap_obj		dma_buf_vmap;
+	pid_t				pid;
+	struct amdgpu_mn		*mn;
+	struct list_head		mn_list;
+};
+#define gem_to_amdgpu_bo(gobj) container_of((gobj), struct amdgpu_bo, gem_base)
+
+void amdgpu_gem_object_free(struct drm_gem_object *obj);
+int amdgpu_gem_object_open(struct drm_gem_object *obj,
+				struct drm_file *file_priv);
+void amdgpu_gem_object_close(struct drm_gem_object *obj,
+				struct drm_file *file_priv);
+unsigned long amdgpu_gem_timeout(uint64_t timeout_ns);
+struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj);
+struct drm_gem_object *amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
+							struct dma_buf_attachment *attach,
+							struct sg_table *sg);
+struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev,
+					struct drm_gem_object *gobj,
+					int flags);
+int amdgpu_gem_prime_pin(struct drm_gem_object *obj);
+void amdgpu_gem_prime_unpin(struct drm_gem_object *obj);
+struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *);
+void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj);
+void amdgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
+int amdgpu_gem_debugfs_init(struct amdgpu_device *adev);
+
+/* sub-allocation manager, it has to be protected by another lock.
+ * By conception this is an helper for other part of the driver
+ * like the indirect buffer or semaphore, which both have their
+ * locking.
+ *
+ * Principe is simple, we keep a list of sub allocation in offset
+ * order (first entry has offset == 0, last entry has the highest
+ * offset).
+ *
+ * When allocating new object we first check if there is room at
+ * the end total_size - (last_object_offset + last_object_size) >=
+ * alloc_size. If so we allocate new object there.
+ *
+ * When there is not enough room at the end, we start waiting for
+ * each sub object until we reach object_offset+object_size >=
+ * alloc_size, this object then become the sub object we return.
+ *
+ * Alignment can't be bigger than page size.
+ *
+ * Hole are not considered for allocation to keep things simple.
+ * Assumption is that there won't be hole (all object on same
+ * alignment).
+ */
+struct amdgpu_sa_manager {
+	wait_queue_head_t	wq;
+	struct amdgpu_bo	*bo;
+	struct list_head	*hole;
+	struct list_head	flist[AMDGPU_MAX_RINGS];
+	struct list_head	olist;
+	unsigned		size;
+	uint64_t		gpu_addr;
+	void			*cpu_ptr;
+	uint32_t		domain;
+	uint32_t		align;
+};
+
+struct amdgpu_sa_bo;
+
+/* sub-allocation buffer */
+struct amdgpu_sa_bo {
+	struct list_head		olist;
+	struct list_head		flist;
+	struct amdgpu_sa_manager	*manager;
+	unsigned			soffset;
+	unsigned			eoffset;
+	struct amdgpu_fence		*fence;
+};
+
+/*
+ * GEM objects.
+ */
+struct amdgpu_gem {
+	struct mutex		mutex;
+	struct list_head	objects;
+};
+
+int amdgpu_gem_init(struct amdgpu_device *adev);
+void amdgpu_gem_fini(struct amdgpu_device *adev);
+int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
+				int alignment, u32 initial_domain,
+				u64 flags, bool kernel,
+				struct drm_gem_object **obj);
+
+int amdgpu_mode_dumb_create(struct drm_file *file_priv,
+			    struct drm_device *dev,
+			    struct drm_mode_create_dumb *args);
+int amdgpu_mode_dumb_mmap(struct drm_file *filp,
+			  struct drm_device *dev,
+			  uint32_t handle, uint64_t *offset_p);
+
+/*
+ * Semaphores.
+ */
+struct amdgpu_semaphore {
+	struct amdgpu_sa_bo	*sa_bo;
+	signed			waiters;
+	uint64_t		gpu_addr;
+};
+
+int amdgpu_semaphore_create(struct amdgpu_device *adev,
+			    struct amdgpu_semaphore **semaphore);
+bool amdgpu_semaphore_emit_signal(struct amdgpu_ring *ring,
+				  struct amdgpu_semaphore *semaphore);
+bool amdgpu_semaphore_emit_wait(struct amdgpu_ring *ring,
+				struct amdgpu_semaphore *semaphore);
+void amdgpu_semaphore_free(struct amdgpu_device *adev,
+			   struct amdgpu_semaphore **semaphore,
+			   struct amdgpu_fence *fence);
+
+/*
+ * Synchronization
+ */
+struct amdgpu_sync {
+	struct amdgpu_semaphore *semaphores[AMDGPU_NUM_SYNCS];
+	struct amdgpu_fence	*sync_to[AMDGPU_MAX_RINGS];
+	struct amdgpu_fence	*last_vm_update;
+};
+
+void amdgpu_sync_create(struct amdgpu_sync *sync);
+void amdgpu_sync_fence(struct amdgpu_sync *sync,
+		       struct amdgpu_fence *fence);
+int amdgpu_sync_resv(struct amdgpu_device *adev,
+		     struct amdgpu_sync *sync,
+		     struct reservation_object *resv,
+		     void *owner);
+int amdgpu_sync_rings(struct amdgpu_sync *sync,
+		      struct amdgpu_ring *ring);
+void amdgpu_sync_free(struct amdgpu_device *adev, struct amdgpu_sync *sync,
+		      struct amdgpu_fence *fence);
+
+/*
+ * GART structures, functions & helpers
+ */
+struct amdgpu_mc;
+
+#define AMDGPU_GPU_PAGE_SIZE 4096
+#define AMDGPU_GPU_PAGE_MASK (AMDGPU_GPU_PAGE_SIZE - 1)
+#define AMDGPU_GPU_PAGE_SHIFT 12
+#define AMDGPU_GPU_PAGE_ALIGN(a) (((a) + AMDGPU_GPU_PAGE_MASK) & ~AMDGPU_GPU_PAGE_MASK)
+
+struct amdgpu_gart {
+	dma_addr_t			table_addr;
+	struct amdgpu_bo		*robj;
+	void				*ptr;
+	unsigned			num_gpu_pages;
+	unsigned			num_cpu_pages;
+	unsigned			table_size;
+	struct page			**pages;
+	dma_addr_t			*pages_addr;
+	bool				ready;
+	const struct amdgpu_gart_funcs *gart_funcs;
+};
+
+int amdgpu_gart_table_ram_alloc(struct amdgpu_device *adev);
+void amdgpu_gart_table_ram_free(struct amdgpu_device *adev);
+int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev);
+void amdgpu_gart_table_vram_free(struct amdgpu_device *adev);
+int amdgpu_gart_table_vram_pin(struct amdgpu_device *adev);
+void amdgpu_gart_table_vram_unpin(struct amdgpu_device *adev);
+int amdgpu_gart_init(struct amdgpu_device *adev);
+void amdgpu_gart_fini(struct amdgpu_device *adev);
+void amdgpu_gart_unbind(struct amdgpu_device *adev, unsigned offset,
+			int pages);
+int amdgpu_gart_bind(struct amdgpu_device *adev, unsigned offset,
+		     int pages, struct page **pagelist,
+		     dma_addr_t *dma_addr, uint32_t flags);
+
+/*
+ * GPU MC structures, functions & helpers
+ */
+struct amdgpu_mc {
+	resource_size_t		aper_size;
+	resource_size_t		aper_base;
+	resource_size_t		agp_base;
+	/* for some chips with <= 32MB we need to lie
+	 * about vram size near mc fb location */
+	u64			mc_vram_size;
+	u64			visible_vram_size;
+	u64			gtt_size;
+	u64			gtt_start;
+	u64			gtt_end;
+	u64			vram_start;
+	u64			vram_end;
+	unsigned		vram_width;
+	u64			real_vram_size;
+	int			vram_mtrr;
+	u64                     gtt_base_align;
+	u64                     mc_mask;
+	const struct firmware   *fw;	/* MC firmware */
+	uint32_t                fw_version;
+	struct amdgpu_irq_src	vm_fault;
+	uint32_t		vram_type;
+};
+
+/*
+ * GPU doorbell structures, functions & helpers
+ */
+typedef enum _AMDGPU_DOORBELL_ASSIGNMENT
+{
+	AMDGPU_DOORBELL_KIQ                     = 0x000,
+	AMDGPU_DOORBELL_HIQ                     = 0x001,
+	AMDGPU_DOORBELL_DIQ                     = 0x002,
+	AMDGPU_DOORBELL_MEC_RING0               = 0x010,
+	AMDGPU_DOORBELL_MEC_RING1               = 0x011,
+	AMDGPU_DOORBELL_MEC_RING2               = 0x012,
+	AMDGPU_DOORBELL_MEC_RING3               = 0x013,
+	AMDGPU_DOORBELL_MEC_RING4               = 0x014,
+	AMDGPU_DOORBELL_MEC_RING5               = 0x015,
+	AMDGPU_DOORBELL_MEC_RING6               = 0x016,
+	AMDGPU_DOORBELL_MEC_RING7               = 0x017,
+	AMDGPU_DOORBELL_GFX_RING0               = 0x020,
+	AMDGPU_DOORBELL_sDMA_ENGINE0            = 0x1E0,
+	AMDGPU_DOORBELL_sDMA_ENGINE1            = 0x1E1,
+	AMDGPU_DOORBELL_IH                      = 0x1E8,
+	AMDGPU_DOORBELL_MAX_ASSIGNMENT          = 0x3FF,
+	AMDGPU_DOORBELL_INVALID                 = 0xFFFF
+} AMDGPU_DOORBELL_ASSIGNMENT;
+
+struct amdgpu_doorbell {
+	/* doorbell mmio */
+	resource_size_t		base;
+	resource_size_t		size;
+	u32 __iomem		*ptr;
+	u32			num_doorbells;	/* Number of doorbells actually reserved for amdgpu. */
+};
+
+void amdgpu_doorbell_get_kfd_info(struct amdgpu_device *adev,
+				phys_addr_t *aperture_base,
+				size_t *aperture_size,
+				size_t *start_offset);
+
+/*
+ * IRQS.
+ */
+
+struct amdgpu_flip_work {
+	struct work_struct		flip_work;
+	struct work_struct		unpin_work;
+	struct amdgpu_device		*adev;
+	int				crtc_id;
+	uint64_t			base;
+	struct drm_pending_vblank_event *event;
+	struct amdgpu_bo		*old_rbo;
+	struct fence			*fence;
+};
+
+
+/*
+ * CP & rings.
+ */
+
+struct amdgpu_ib {
+	struct amdgpu_sa_bo		*sa_bo;
+	uint32_t			length_dw;
+	uint64_t			gpu_addr;
+	uint32_t			*ptr;
+	struct amdgpu_ring		*ring;
+	struct amdgpu_fence		*fence;
+	struct amdgpu_user_fence        *user;
+	struct amdgpu_vm		*vm;
+	struct amdgpu_ctx		*ctx;
+	struct amdgpu_sync		sync;
+	uint32_t			gds_base, gds_size;
+	uint32_t			gws_base, gws_size;
+	uint32_t			oa_base, oa_size;
+	uint32_t			flags;
+};
+
+enum amdgpu_ring_type {
+	AMDGPU_RING_TYPE_GFX,
+	AMDGPU_RING_TYPE_COMPUTE,
+	AMDGPU_RING_TYPE_SDMA,
+	AMDGPU_RING_TYPE_UVD,
+	AMDGPU_RING_TYPE_VCE
+};
+
+struct amdgpu_ring {
+	struct amdgpu_device		*adev;
+	const struct amdgpu_ring_funcs	*funcs;
+	struct amdgpu_fence_driver	fence_drv;
+
+	struct mutex		*ring_lock;
+	struct amdgpu_bo	*ring_obj;
+	volatile uint32_t	*ring;
+	unsigned		rptr_offs;
+	u64			next_rptr_gpu_addr;
+	volatile u32		*next_rptr_cpu_addr;
+	unsigned		wptr;
+	unsigned		wptr_old;
+	unsigned		ring_size;
+	unsigned		ring_free_dw;
+	int			count_dw;
+	atomic_t		last_rptr;
+	atomic64_t		last_activity;
+	uint64_t		gpu_addr;
+	uint32_t		align_mask;
+	uint32_t		ptr_mask;
+	bool			ready;
+	u32			nop;
+	u32			idx;
+	u64			last_semaphore_signal_addr;
+	u64			last_semaphore_wait_addr;
+	u32			me;
+	u32			pipe;
+	u32			queue;
+	struct amdgpu_bo	*mqd_obj;
+	u32			doorbell_index;
+	bool			use_doorbell;
+	unsigned		wptr_offs;
+	unsigned		next_rptr_offs;
+	unsigned		fence_offs;
+	struct amdgpu_ctx	*current_ctx;
+	enum amdgpu_ring_type	type;
+	char			name[16];
+};
+
+/*
+ * VM
+ */
+
+/* maximum number of VMIDs */
+#define AMDGPU_NUM_VM	16
+
+/* number of entries in page table */
+#define AMDGPU_VM_PTE_COUNT (1 << amdgpu_vm_block_size)
+
+/* PTBs (Page Table Blocks) need to be aligned to 32K */
+#define AMDGPU_VM_PTB_ALIGN_SIZE   32768
+#define AMDGPU_VM_PTB_ALIGN_MASK (AMDGPU_VM_PTB_ALIGN_SIZE - 1)
+#define AMDGPU_VM_PTB_ALIGN(a) (((a) + AMDGPU_VM_PTB_ALIGN_MASK) & ~AMDGPU_VM_PTB_ALIGN_MASK)
+
+#define AMDGPU_PTE_VALID	(1 << 0)
+#define AMDGPU_PTE_SYSTEM	(1 << 1)
+#define AMDGPU_PTE_SNOOPED	(1 << 2)
+
+/* VI only */
+#define AMDGPU_PTE_EXECUTABLE	(1 << 4)
+
+#define AMDGPU_PTE_READABLE	(1 << 5)
+#define AMDGPU_PTE_WRITEABLE	(1 << 6)
+
+/* PTE (Page Table Entry) fragment field for different page sizes */
+#define AMDGPU_PTE_FRAG_4KB	(0 << 7)
+#define AMDGPU_PTE_FRAG_64KB	(4 << 7)
+#define AMDGPU_LOG2_PAGES_PER_FRAG 4
+
+struct amdgpu_vm_pt {
+	struct amdgpu_bo		*bo;
+	uint64_t			addr;
+};
+
+struct amdgpu_vm_id {
+	unsigned		id;
+	uint64_t		pd_gpu_addr;
+	/* last flushed PD/PT update */
+	struct amdgpu_fence	*flushed_updates;
+	/* last use of vmid */
+	struct amdgpu_fence	*last_id_use;
+};
+
+struct amdgpu_vm {
+	struct mutex		mutex;
+
+	struct rb_root		va;
+
+	/* protecting invalidated and freed */
+	spinlock_t		status_lock;
+
+	/* BOs moved, but not yet updated in the PT */
+	struct list_head	invalidated;
+
+	/* BOs freed, but not yet updated in the PT */
+	struct list_head	freed;
+
+	/* contains the page directory */
+	struct amdgpu_bo	*page_directory;
+	unsigned		max_pde_used;
+
+	/* array of page tables, one for each page directory entry */
+	struct amdgpu_vm_pt	*page_tables;
+
+	/* for id and flush management per ring */
+	struct amdgpu_vm_id	ids[AMDGPU_MAX_RINGS];
+};
+
+struct amdgpu_vm_manager {
+	struct amdgpu_fence		*active[AMDGPU_NUM_VM];
+	uint32_t			max_pfn;
+	/* number of VMIDs */
+	unsigned			nvm;
+	/* vram base address for page table entry  */
+	u64				vram_base_offset;
+	/* is vm enabled? */
+	bool				enabled;
+	/* for hw to save the PD addr on suspend/resume */
+	uint32_t			saved_table_addr[AMDGPU_NUM_VM];
+	/* vm pte handling */
+	const struct amdgpu_vm_pte_funcs        *vm_pte_funcs;
+	struct amdgpu_ring                      *vm_pte_funcs_ring;
+};
+
+/*
+ * context related structures
+ */