drm/radeon: Add CP init for CIK (v7)

Sets up the GFX ring and loads ucode for GFX and Compute.

Todo:
- handle compute queue setup.

v2: add documentation
v3: integrate with latest reset changes
v4: additional init fixes
v5: scratch reg write back no longer supported on CIK
v6: properly set CP_RB0_BASE_HI
v7: rebase

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Alex Deucher 2012-12-18 21:47:44 -05:00
parent bc8273fe97
commit 841cf442fd
6 changed files with 899 additions and 2 deletions

View file

@ -76,7 +76,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
evergreen.o evergreen_cs.o evergreen_blit_shaders.o evergreen_blit_kms.o \
evergreen_hdmi.o radeon_trace_points.o ni.o cayman_blit_shaders.o \
atombios_encoders.o radeon_semaphore.o radeon_sa.o atombios_i2c.o si.o \
si_blit_shaders.o radeon_prime.o radeon_uvd.o cik.o
si_blit_shaders.o radeon_prime.o radeon_uvd.o cik.o cik_blit_shaders.o
radeon-$(CONFIG_COMPAT) += radeon_ioc32.o
radeon-$(CONFIG_VGA_SWITCHEROO) += radeon_atpx_handler.o

View file

@ -30,6 +30,7 @@
#include "radeon_asic.h"
#include "cikd.h"
#include "atom.h"
#include "cik_blit_shaders.h"
/* GFX */
#define CIK_PFP_UCODE_SIZE 2144
@ -1491,6 +1492,400 @@ static void cik_gpu_init(struct radeon_device *rdev)
udelay(50);
}
/*
* CP.
* On CIK, gfx and compute now have independant command processors.
*
* GFX
* Gfx consists of a single ring and can process both gfx jobs and
* compute jobs. The gfx CP consists of three microengines (ME):
* PFP - Pre-Fetch Parser
* ME - Micro Engine
* CE - Constant Engine
* The PFP and ME make up what is considered the Drawing Engine (DE).
* The CE is an asynchronous engine used for updating buffer desciptors
* used by the DE so that they can be loaded into cache in parallel
* while the DE is processing state update packets.
*
* Compute
* The compute CP consists of two microengines (ME):
* MEC1 - Compute MicroEngine 1
* MEC2 - Compute MicroEngine 2
* Each MEC supports 4 compute pipes and each pipe supports 8 queues.
* The queues are exposed to userspace and are programmed directly
* by the compute runtime.
*/
/**
* cik_cp_gfx_enable - enable/disable the gfx CP MEs
*
* @rdev: radeon_device pointer
* @enable: enable or disable the MEs
*
* Halts or unhalts the gfx MEs.
*/
static void cik_cp_gfx_enable(struct radeon_device *rdev, bool enable)
{
if (enable)
WREG32(CP_ME_CNTL, 0);
else {
WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT));
rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
}
udelay(50);
}
/**
* cik_cp_gfx_load_microcode - load the gfx CP ME ucode
*
* @rdev: radeon_device pointer
*
* Loads the gfx PFP, ME, and CE ucode.
* Returns 0 for success, -EINVAL if the ucode is not available.
*/
static int cik_cp_gfx_load_microcode(struct radeon_device *rdev)
{
const __be32 *fw_data;
int i;
if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw)
return -EINVAL;
cik_cp_gfx_enable(rdev, false);
/* PFP */
fw_data = (const __be32 *)rdev->pfp_fw->data;
WREG32(CP_PFP_UCODE_ADDR, 0);
for (i = 0; i < CIK_PFP_UCODE_SIZE; i++)
WREG32(CP_PFP_UCODE_DATA, be32_to_cpup(fw_data++));
WREG32(CP_PFP_UCODE_ADDR, 0);
/* CE */
fw_data = (const __be32 *)rdev->ce_fw->data;
WREG32(CP_CE_UCODE_ADDR, 0);
for (i = 0; i < CIK_CE_UCODE_SIZE; i++)
WREG32(CP_CE_UCODE_DATA, be32_to_cpup(fw_data++));
WREG32(CP_CE_UCODE_ADDR, 0);
/* ME */
fw_data = (const __be32 *)rdev->me_fw->data;
WREG32(CP_ME_RAM_WADDR, 0);
for (i = 0; i < CIK_ME_UCODE_SIZE; i++)
WREG32(CP_ME_RAM_DATA, be32_to_cpup(fw_data++));
WREG32(CP_ME_RAM_WADDR, 0);
WREG32(CP_PFP_UCODE_ADDR, 0);
WREG32(CP_CE_UCODE_ADDR, 0);
WREG32(CP_ME_RAM_WADDR, 0);
WREG32(CP_ME_RAM_RADDR, 0);
return 0;
}
/**
* cik_cp_gfx_start - start the gfx ring
*
* @rdev: radeon_device pointer
*
* Enables the ring and loads the clear state context and other
* packets required to init the ring.
* Returns 0 for success, error for failure.
*/
static int cik_cp_gfx_start(struct radeon_device *rdev)
{
struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
int r, i;
/* init the CP */
WREG32(CP_MAX_CONTEXT, rdev->config.cik.max_hw_contexts - 1);
WREG32(CP_ENDIAN_SWAP, 0);
WREG32(CP_DEVICE_ID, 1);
cik_cp_gfx_enable(rdev, true);
r = radeon_ring_lock(rdev, ring, cik_default_size + 17);
if (r) {
DRM_ERROR("radeon: cp failed to lock ring (%d).\n", r);
return r;
}
/* init the CE partitions. CE only used for gfx on CIK */
radeon_ring_write(ring, PACKET3(PACKET3_SET_BASE, 2));
radeon_ring_write(ring, PACKET3_BASE_INDEX(CE_PARTITION_BASE));
radeon_ring_write(ring, 0xc000);
radeon_ring_write(ring, 0xc000);
/* setup clear context state */
radeon_ring_write(ring, PACKET3(PACKET3_PREAMBLE_CNTL, 0));
radeon_ring_write(ring, PACKET3_PREAMBLE_BEGIN_CLEAR_STATE);
radeon_ring_write(ring, PACKET3(PACKET3_CONTEXT_CONTROL, 1));
radeon_ring_write(ring, 0x80000000);
radeon_ring_write(ring, 0x80000000);
for (i = 0; i < cik_default_size; i++)
radeon_ring_write(ring, cik_default_state[i]);
radeon_ring_write(ring, PACKET3(PACKET3_PREAMBLE_CNTL, 0));
radeon_ring_write(ring, PACKET3_PREAMBLE_END_CLEAR_STATE);
/* set clear context state */
radeon_ring_write(ring, PACKET3(PACKET3_CLEAR_STATE, 0));
radeon_ring_write(ring, 0);
radeon_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2));
radeon_ring_write(ring, 0x00000316);
radeon_ring_write(ring, 0x0000000e); /* VGT_VERTEX_REUSE_BLOCK_CNTL */
radeon_ring_write(ring, 0x00000010); /* VGT_OUT_DEALLOC_CNTL */
radeon_ring_unlock_commit(rdev, ring);
return 0;
}
/**
* cik_cp_gfx_fini - stop the gfx ring
*
* @rdev: radeon_device pointer
*
* Stop the gfx ring and tear down the driver ring
* info.
*/
static void cik_cp_gfx_fini(struct radeon_device *rdev)
{
cik_cp_gfx_enable(rdev, false);
radeon_ring_fini(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]);
}
/**
* cik_cp_gfx_resume - setup the gfx ring buffer registers
*
* @rdev: radeon_device pointer
*
* Program the location and size of the gfx ring buffer
* and test it to make sure it's working.
* Returns 0 for success, error for failure.
*/
static int cik_cp_gfx_resume(struct radeon_device *rdev)
{
struct radeon_ring *ring;
u32 tmp;
u32 rb_bufsz;
u64 rb_addr;
int r;
WREG32(CP_SEM_WAIT_TIMER, 0x0);
WREG32(CP_SEM_INCOMPLETE_TIMER_CNTL, 0x0);
/* Set the write pointer delay */
WREG32(CP_RB_WPTR_DELAY, 0);
/* set the RB to use vmid 0 */
WREG32(CP_RB_VMID, 0);
WREG32(SCRATCH_ADDR, ((rdev->wb.gpu_addr + RADEON_WB_SCRATCH_OFFSET) >> 8) & 0xFFFFFFFF);
/* ring 0 - compute and gfx */
/* Set ring buffer size */
ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
rb_bufsz = drm_order(ring->ring_size / 8);
tmp = (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz;
#ifdef __BIG_ENDIAN
tmp |= BUF_SWAP_32BIT;
#endif
WREG32(CP_RB0_CNTL, tmp);
/* Initialize the ring buffer's read and write pointers */
WREG32(CP_RB0_CNTL, tmp | RB_RPTR_WR_ENA);
ring->wptr = 0;
WREG32(CP_RB0_WPTR, ring->wptr);
/* set the wb address wether it's enabled or not */
WREG32(CP_RB0_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFFFFFFFC);
WREG32(CP_RB0_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFF);
/* scratch register shadowing is no longer supported */
WREG32(SCRATCH_UMSK, 0);
if (!rdev->wb.enabled)
tmp |= RB_NO_UPDATE;
mdelay(1);
WREG32(CP_RB0_CNTL, tmp);
rb_addr = ring->gpu_addr >> 8;
WREG32(CP_RB0_BASE, rb_addr);
WREG32(CP_RB0_BASE_HI, upper_32_bits(rb_addr));
ring->rptr = RREG32(CP_RB0_RPTR);
/* start the ring */
cik_cp_gfx_start(rdev);
rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = true;
r = radeon_ring_test(rdev, RADEON_RING_TYPE_GFX_INDEX, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]);
if (r) {
rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
return r;
}
return 0;
}
/**
* cik_cp_compute_enable - enable/disable the compute CP MEs
*
* @rdev: radeon_device pointer
* @enable: enable or disable the MEs
*
* Halts or unhalts the compute MEs.
*/
static void cik_cp_compute_enable(struct radeon_device *rdev, bool enable)
{
if (enable)
WREG32(CP_MEC_CNTL, 0);
else
WREG32(CP_MEC_CNTL, (MEC_ME1_HALT | MEC_ME2_HALT));
udelay(50);
}
/**
* cik_cp_compute_load_microcode - load the compute CP ME ucode
*
* @rdev: radeon_device pointer
*
* Loads the compute MEC1&2 ucode.
* Returns 0 for success, -EINVAL if the ucode is not available.
*/
static int cik_cp_compute_load_microcode(struct radeon_device *rdev)
{
const __be32 *fw_data;
int i;
if (!rdev->mec_fw)
return -EINVAL;
cik_cp_compute_enable(rdev, false);
/* MEC1 */
fw_data = (const __be32 *)rdev->mec_fw->data;
WREG32(CP_MEC_ME1_UCODE_ADDR, 0);
for (i = 0; i < CIK_MEC_UCODE_SIZE; i++)
WREG32(CP_MEC_ME1_UCODE_DATA, be32_to_cpup(fw_data++));
WREG32(CP_MEC_ME1_UCODE_ADDR, 0);
if (rdev->family == CHIP_KAVERI) {
/* MEC2 */
fw_data = (const __be32 *)rdev->mec_fw->data;
WREG32(CP_MEC_ME2_UCODE_ADDR, 0);
for (i = 0; i < CIK_MEC_UCODE_SIZE; i++)
WREG32(CP_MEC_ME2_UCODE_DATA, be32_to_cpup(fw_data++));
WREG32(CP_MEC_ME2_UCODE_ADDR, 0);
}
return 0;
}
/**
* cik_cp_compute_start - start the compute queues
*
* @rdev: radeon_device pointer
*
* Enable the compute queues.
* Returns 0 for success, error for failure.
*/
static int cik_cp_compute_start(struct radeon_device *rdev)
{
//todo
return 0;
}
/**
* cik_cp_compute_fini - stop the compute queues
*
* @rdev: radeon_device pointer
*
* Stop the compute queues and tear down the driver queue
* info.
*/
static void cik_cp_compute_fini(struct radeon_device *rdev)
{
cik_cp_compute_enable(rdev, false);
//todo
}
/**
* cik_cp_compute_resume - setup the compute queue registers
*
* @rdev: radeon_device pointer
*
* Program the compute queues and test them to make sure they
* are working.
* Returns 0 for success, error for failure.
*/
static int cik_cp_compute_resume(struct radeon_device *rdev)
{
int r;
//todo
r = cik_cp_compute_start(rdev);
if (r)
return r;
return 0;
}
/* XXX temporary wrappers to handle both compute and gfx */
/* XXX */
static void cik_cp_enable(struct radeon_device *rdev, bool enable)
{
cik_cp_gfx_enable(rdev, enable);
cik_cp_compute_enable(rdev, enable);
}
/* XXX */
static int cik_cp_load_microcode(struct radeon_device *rdev)
{
int r;
r = cik_cp_gfx_load_microcode(rdev);
if (r)
return r;
r = cik_cp_compute_load_microcode(rdev);
if (r)
return r;
return 0;
}
/* XXX */
static void cik_cp_fini(struct radeon_device *rdev)
{
cik_cp_gfx_fini(rdev);
cik_cp_compute_fini(rdev);
}
/* XXX */
static int cik_cp_resume(struct radeon_device *rdev)
{
int r;
/* Reset all cp blocks */
WREG32(GRBM_SOFT_RESET, SOFT_RESET_CP);
RREG32(GRBM_SOFT_RESET);
mdelay(15);
WREG32(GRBM_SOFT_RESET, 0);
RREG32(GRBM_SOFT_RESET);
r = cik_cp_load_microcode(rdev);
if (r)
return r;
r = cik_cp_gfx_resume(rdev);
if (r)
return r;
r = cik_cp_compute_resume(rdev);
if (r)
return r;
return 0;
}
/**
* cik_gpu_is_lockup - check if the 3D engine is locked up
*

View file

@ -0,0 +1,246 @@
/*
* Copyright 2012 Advanced Micro Devices, Inc.
*
* 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 (including the next
* paragraph) 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) AND/OR ITS SUPPLIERS 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:
* Alex Deucher <alexander.deucher@amd.com>
*/
#include <linux/types.h>
#include <linux/bug.h>
#include <linux/kernel.h>
const u32 cik_default_state[] =
{
0xc0066900,
0x00000000,
0x00000060, /* DB_RENDER_CONTROL */
0x00000000, /* DB_COUNT_CONTROL */
0x00000000, /* DB_DEPTH_VIEW */
0x0000002a, /* DB_RENDER_OVERRIDE */
0x00000000, /* DB_RENDER_OVERRIDE2 */
0x00000000, /* DB_HTILE_DATA_BASE */
0xc0046900,
0x00000008,
0x00000000, /* DB_DEPTH_BOUNDS_MIN */
0x00000000, /* DB_DEPTH_BOUNDS_MAX */
0x00000000, /* DB_STENCIL_CLEAR */
0x00000000, /* DB_DEPTH_CLEAR */
0xc0036900,
0x0000000f,
0x00000000, /* DB_DEPTH_INFO */
0x00000000, /* DB_Z_INFO */
0x00000000, /* DB_STENCIL_INFO */
0xc0016900,
0x00000080,
0x00000000, /* PA_SC_WINDOW_OFFSET */
0xc00d6900,
0x00000083,
0x0000ffff, /* PA_SC_CLIPRECT_RULE */
0x00000000, /* PA_SC_CLIPRECT_0_TL */
0x20002000, /* PA_SC_CLIPRECT_0_BR */
0x00000000,
0x20002000,
0x00000000,
0x20002000,
0x00000000,
0x20002000,
0xaaaaaaaa, /* PA_SC_EDGERULE */
0x00000000, /* PA_SU_HARDWARE_SCREEN_OFFSET */
0x0000000f, /* CB_TARGET_MASK */
0x0000000f, /* CB_SHADER_MASK */
0xc0226900,
0x00000094,
0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */
0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x80000000,
0x20002000,
0x00000000, /* PA_SC_VPORT_ZMIN_0 */
0x3f800000, /* PA_SC_VPORT_ZMAX_0 */
0xc0046900,
0x00000100,
0xffffffff, /* VGT_MAX_VTX_INDX */
0x00000000, /* VGT_MIN_VTX_INDX */
0x00000000, /* VGT_INDX_OFFSET */
0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */
0xc0046900,
0x00000105,
0x00000000, /* CB_BLEND_RED */
0x00000000, /* CB_BLEND_GREEN */
0x00000000, /* CB_BLEND_BLUE */
0x00000000, /* CB_BLEND_ALPHA */
0xc0016900,
0x000001e0,
0x00000000, /* CB_BLEND0_CONTROL */
0xc00c6900,
0x00000200,
0x00000000, /* DB_DEPTH_CONTROL */
0x00000000, /* DB_EQAA */
0x00cc0010, /* CB_COLOR_CONTROL */
0x00000210, /* DB_SHADER_CONTROL */
0x00010000, /* PA_CL_CLIP_CNTL */
0x00000004, /* PA_SU_SC_MODE_CNTL */
0x00000100, /* PA_CL_VTE_CNTL */
0x00000000, /* PA_CL_VS_OUT_CNTL */
0x00000000, /* PA_CL_NANINF_CNTL */
0x00000000, /* PA_SU_LINE_STIPPLE_CNTL */
0x00000000, /* PA_SU_LINE_STIPPLE_SCALE */
0x00000000, /* PA_SU_PRIM_FILTER_CNTL */
0xc0116900,
0x00000280,
0x00000000, /* PA_SU_POINT_SIZE */
0x00000000, /* PA_SU_POINT_MINMAX */
0x00000008, /* PA_SU_LINE_CNTL */
0x00000000, /* PA_SC_LINE_STIPPLE */
0x00000000, /* VGT_OUTPUT_PATH_CNTL */
0x00000000, /* VGT_HOS_CNTL */
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000, /* VGT_GS_MODE */
0xc0026900,
0x00000292,
0x00000000, /* PA_SC_MODE_CNTL_0 */
0x00000000, /* PA_SC_MODE_CNTL_1 */
0xc0016900,
0x000002a1,
0x00000000, /* VGT_PRIMITIVEID_EN */
0xc0016900,
0x000002a5,
0x00000000, /* VGT_MULTI_PRIM_IB_RESET_EN */
0xc0026900,
0x000002a8,
0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */
0x00000000,
0xc0026900,
0x000002ad,
0x00000000, /* VGT_REUSE_OFF */
0x00000000,
0xc0016900,
0x000002d5,
0x00000000, /* VGT_SHADER_STAGES_EN */
0xc0016900,
0x000002dc,
0x0000aa00, /* DB_ALPHA_TO_MASK */
0xc0066900,
0x000002de,
0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0xc0026900,
0x000002e5,
0x00000000, /* VGT_STRMOUT_CONFIG */
0x00000000,
0xc01b6900,
0x000002f5,
0x76543210, /* PA_SC_CENTROID_PRIORITY_0 */
0xfedcba98, /* PA_SC_CENTROID_PRIORITY_1 */
0x00000000, /* PA_SC_LINE_CNTL */
0x00000000, /* PA_SC_AA_CONFIG */
0x00000005, /* PA_SU_VTX_CNTL */
0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */
0x3f800000, /* PA_CL_GB_VERT_DISC_ADJ */
0x3f800000, /* PA_CL_GB_HORZ_CLIP_ADJ */
0x3f800000, /* PA_CL_GB_HORZ_DISC_ADJ */
0x00000000, /* PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0 */
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0x00000000,
0xffffffff, /* PA_SC_AA_MASK_X0Y0_X1Y0 */
0xffffffff,
0xc0026900,
0x00000316,
0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */
0x00000010, /* */
};
const u32 cik_default_size = ARRAY_SIZE(cik_default_state);

View file

@ -0,0 +1,32 @@
/*
* Copyright 2012 Advanced Micro Devices, Inc.
*
* 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 (including the next
* paragraph) 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) AND/OR ITS SUPPLIERS 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.
*
*/
#ifndef CIK_BLIT_SHADERS_H
#define CIK_BLIT_SHADERS_H
extern const u32 cik_default_state[];
extern const u32 cik_default_size;
#endif

View file

@ -263,11 +263,18 @@
#define MEC_ME2_HALT (1 << 28)
#define MEC_ME1_HALT (1 << 30)
#define CP_MEC_CNTL 0x8234
#define MEC_ME2_HALT (1 << 28)
#define MEC_ME1_HALT (1 << 30)
#define CP_ME_CNTL 0x86D8
#define CP_CE_HALT (1 << 24)
#define CP_PFP_HALT (1 << 26)
#define CP_ME_HALT (1 << 28)
#define CP_RB0_RPTR 0x8700
#define CP_RB_WPTR_DELAY 0x8704
#define CP_MEQ_THRESHOLDS 0x8764
#define MEQ1_START(x) ((x) << 0)
#define MEQ2_START(x) ((x) << 8)
@ -445,12 +452,62 @@
#define TC_CFG_L1_VOLATILE 0xAC88
#define TC_CFG_L2_VOLATILE 0xAC8C
#define CP_RB0_BASE 0xC100
#define CP_RB0_CNTL 0xC104
#define RB_BUFSZ(x) ((x) << 0)
#define RB_BLKSZ(x) ((x) << 8)
#define BUF_SWAP_32BIT (2 << 16)
#define RB_NO_UPDATE (1 << 27)
#define RB_RPTR_WR_ENA (1 << 31)
#define CP_RB0_RPTR_ADDR 0xC10C
#define RB_RPTR_SWAP_32BIT (2 << 0)
#define CP_RB0_RPTR_ADDR_HI 0xC110
#define CP_RB0_WPTR 0xC114
#define CP_DEVICE_ID 0xC12C
#define CP_ENDIAN_SWAP 0xC140
#define CP_RB_VMID 0xC144
#define CP_PFP_UCODE_ADDR 0xC150
#define CP_PFP_UCODE_DATA 0xC154
#define CP_ME_RAM_RADDR 0xC158
#define CP_ME_RAM_WADDR 0xC15C
#define CP_ME_RAM_DATA 0xC160
#define CP_CE_UCODE_ADDR 0xC168
#define CP_CE_UCODE_DATA 0xC16C
#define CP_MEC_ME1_UCODE_ADDR 0xC170
#define CP_MEC_ME1_UCODE_DATA 0xC174
#define CP_MEC_ME2_UCODE_ADDR 0xC178
#define CP_MEC_ME2_UCODE_DATA 0xC17C
#define CP_MAX_CONTEXT 0xC2B8
#define CP_RB0_BASE_HI 0xC2C4
#define PA_SC_RASTER_CONFIG 0x28350
# define RASTER_CONFIG_RB_MAP_0 0
# define RASTER_CONFIG_RB_MAP_1 1
# define RASTER_CONFIG_RB_MAP_2 2
# define RASTER_CONFIG_RB_MAP_3 3
#define SCRATCH_REG0 0x30100
#define SCRATCH_REG1 0x30104
#define SCRATCH_REG2 0x30108
#define SCRATCH_REG3 0x3010C
#define SCRATCH_REG4 0x30110
#define SCRATCH_REG5 0x30114
#define SCRATCH_REG6 0x30118
#define SCRATCH_REG7 0x3011C
#define SCRATCH_UMSK 0x30140
#define SCRATCH_ADDR 0x30144
#define CP_SEM_WAIT_TIMER 0x301BC
#define CP_SEM_INCOMPLETE_TIMER_CNTL 0x301C8
#define GRBM_GFX_INDEX 0x30800
#define INSTANCE_INDEX(x) ((x) << 0)
#define SH_INDEX(x) ((x) << 8)
@ -482,4 +539,169 @@
#define TCC_DISABLE_MASK 0xFFFF0000
#define TCC_DISABLE_SHIFT 16
/*
* PM4
*/
#define PACKET_TYPE0 0
#define PACKET_TYPE1 1
#define PACKET_TYPE2 2
#define PACKET_TYPE3 3
#define CP_PACKET_GET_TYPE(h) (((h) >> 30) & 3)
#define CP_PACKET_GET_COUNT(h) (((h) >> 16) & 0x3FFF)
#define CP_PACKET0_GET_REG(h) (((h) & 0xFFFF) << 2)
#define CP_PACKET3_GET_OPCODE(h) (((h) >> 8) & 0xFF)
#define PACKET0(reg, n) ((PACKET_TYPE0 << 30) | \
(((reg) >> 2) & 0xFFFF) | \
((n) & 0x3FFF) << 16)
#define CP_PACKET2 0x80000000
#define PACKET2_PAD_SHIFT 0
#define PACKET2_PAD_MASK (0x3fffffff << 0)
#define PACKET2(v) (CP_PACKET2 | REG_SET(PACKET2_PAD, (v)))
#define PACKET3(op, n) ((PACKET_TYPE3 << 30) | \
(((op) & 0xFF) << 8) | \
((n) & 0x3FFF) << 16)
#define PACKET3_COMPUTE(op, n) (PACKET3(op, n) | 1 << 1)
/* Packet 3 types */
#define PACKET3_NOP 0x10
#define PACKET3_SET_BASE 0x11
#define PACKET3_BASE_INDEX(x) ((x) << 0)
#define CE_PARTITION_BASE 3
#define PACKET3_CLEAR_STATE 0x12
#define PACKET3_INDEX_BUFFER_SIZE 0x13
#define PACKET3_DISPATCH_DIRECT 0x15
#define PACKET3_DISPATCH_INDIRECT 0x16
#define PACKET3_ATOMIC_GDS 0x1D
#define PACKET3_ATOMIC_MEM 0x1E
#define PACKET3_OCCLUSION_QUERY 0x1F
#define PACKET3_SET_PREDICATION 0x20
#define PACKET3_REG_RMW 0x21
#define PACKET3_COND_EXEC 0x22
#define PACKET3_PRED_EXEC 0x23
#define PACKET3_DRAW_INDIRECT 0x24
#define PACKET3_DRAW_INDEX_INDIRECT 0x25
#define PACKET3_INDEX_BASE 0x26
#define PACKET3_DRAW_INDEX_2 0x27
#define PACKET3_CONTEXT_CONTROL 0x28
#define PACKET3_INDEX_TYPE 0x2A
#define PACKET3_DRAW_INDIRECT_MULTI 0x2C
#define PACKET3_DRAW_INDEX_AUTO 0x2D
#define PACKET3_NUM_INSTANCES 0x2F
#define PACKET3_DRAW_INDEX_MULTI_AUTO 0x30
#define PACKET3_INDIRECT_BUFFER_CONST 0x33
#define PACKET3_STRMOUT_BUFFER_UPDATE 0x34
#define PACKET3_DRAW_INDEX_OFFSET_2 0x35
#define PACKET3_DRAW_PREAMBLE 0x36
#define PACKET3_WRITE_DATA 0x37
#define PACKET3_DRAW_INDEX_INDIRECT_MULTI 0x38
#define PACKET3_MEM_SEMAPHORE 0x39
#define PACKET3_COPY_DW 0x3B
#define PACKET3_WAIT_REG_MEM 0x3C
#define PACKET3_INDIRECT_BUFFER 0x3F
#define PACKET3_COPY_DATA 0x40
#define PACKET3_PFP_SYNC_ME 0x42
#define PACKET3_SURFACE_SYNC 0x43
# define PACKET3_DEST_BASE_0_ENA (1 << 0)
# define PACKET3_DEST_BASE_1_ENA (1 << 1)
# define PACKET3_CB0_DEST_BASE_ENA (1 << 6)
# define PACKET3_CB1_DEST_BASE_ENA (1 << 7)
# define PACKET3_CB2_DEST_BASE_ENA (1 << 8)
# define PACKET3_CB3_DEST_BASE_ENA (1 << 9)
# define PACKET3_CB4_DEST_BASE_ENA (1 << 10)
# define PACKET3_CB5_DEST_BASE_ENA (1 << 11)
# define PACKET3_CB6_DEST_BASE_ENA (1 << 12)
# define PACKET3_CB7_DEST_BASE_ENA (1 << 13)
# define PACKET3_DB_DEST_BASE_ENA (1 << 14)
# define PACKET3_TCL1_VOL_ACTION_ENA (1 << 15)
# define PACKET3_TC_VOL_ACTION_ENA (1 << 16) /* L2 */
# define PACKET3_TC_WB_ACTION_ENA (1 << 18) /* L2 */
# define PACKET3_DEST_BASE_2_ENA (1 << 19)
# define PACKET3_DEST_BASE_3_ENA (1 << 21)
# define PACKET3_TCL1_ACTION_ENA (1 << 22)
# define PACKET3_TC_ACTION_ENA (1 << 23) /* L2 */
# define PACKET3_CB_ACTION_ENA (1 << 25)
# define PACKET3_DB_ACTION_ENA (1 << 26)
# define PACKET3_SH_KCACHE_ACTION_ENA (1 << 27)
# define PACKET3_SH_KCACHE_VOL_ACTION_ENA (1 << 28)
# define PACKET3_SH_ICACHE_ACTION_ENA (1 << 29)
#define PACKET3_COND_WRITE 0x45
#define PACKET3_EVENT_WRITE 0x46
#define EVENT_TYPE(x) ((x) << 0)
#define EVENT_INDEX(x) ((x) << 8)
/* 0 - any non-TS event
* 1 - ZPASS_DONE, PIXEL_PIPE_STAT_*
* 2 - SAMPLE_PIPELINESTAT
* 3 - SAMPLE_STREAMOUTSTAT*
* 4 - *S_PARTIAL_FLUSH
* 5 - EOP events
* 6 - EOS events
*/
#define PACKET3_EVENT_WRITE_EOP 0x47
#define EOP_TCL1_VOL_ACTION_EN (1 << 12)
#define EOP_TC_VOL_ACTION_EN (1 << 13) /* L2 */
#define EOP_TC_WB_ACTION_EN (1 << 15) /* L2 */
#define EOP_TCL1_ACTION_EN (1 << 16)
#define EOP_TC_ACTION_EN (1 << 17) /* L2 */
#define CACHE_POLICY(x) ((x) << 25)
/* 0 - LRU
* 1 - Stream
* 2 - Bypass
*/
#define TCL2_VOLATILE (1 << 27)
#define DATA_SEL(x) ((x) << 29)
/* 0 - discard
* 1 - send low 32bit data
* 2 - send 64bit data
* 3 - send 64bit GPU counter value
* 4 - send 64bit sys counter value
*/
#define INT_SEL(x) ((x) << 24)
/* 0 - none
* 1 - interrupt only (DATA_SEL = 0)
* 2 - interrupt when data write is confirmed
*/
#define DST_SEL(x) ((x) << 16)
/* 0 - MC
* 1 - TC/L2
*/
#define PACKET3_EVENT_WRITE_EOS 0x48
#define PACKET3_RELEASE_MEM 0x49
#define PACKET3_PREAMBLE_CNTL 0x4A
# define PACKET3_PREAMBLE_BEGIN_CLEAR_STATE (2 << 28)
# define PACKET3_PREAMBLE_END_CLEAR_STATE (3 << 28)
#define PACKET3_DMA_DATA 0x50
#define PACKET3_AQUIRE_MEM 0x58
#define PACKET3_REWIND 0x59
#define PACKET3_LOAD_UCONFIG_REG 0x5E
#define PACKET3_LOAD_SH_REG 0x5F
#define PACKET3_LOAD_CONFIG_REG 0x60
#define PACKET3_LOAD_CONTEXT_REG 0x61
#define PACKET3_SET_CONFIG_REG 0x68
#define PACKET3_SET_CONFIG_REG_START 0x00008000
#define PACKET3_SET_CONFIG_REG_END 0x0000b000
#define PACKET3_SET_CONTEXT_REG 0x69
#define PACKET3_SET_CONTEXT_REG_START 0x00028000
#define PACKET3_SET_CONTEXT_REG_END 0x00029000
#define PACKET3_SET_CONTEXT_REG_INDIRECT 0x73
#define PACKET3_SET_SH_REG 0x76
#define PACKET3_SET_SH_REG_START 0x0000b000
#define PACKET3_SET_SH_REG_END 0x0000c000
#define PACKET3_SET_SH_REG_OFFSET 0x77
#define PACKET3_SET_QUEUE_REG 0x78
#define PACKET3_SET_UCONFIG_REG 0x79
#define PACKET3_SCRATCH_RAM_WRITE 0x7D
#define PACKET3_SCRATCH_RAM_READ 0x7E
#define PACKET3_LOAD_CONST_RAM 0x80
#define PACKET3_WRITE_CONST_RAM 0x81
#define PACKET3_DUMP_CONST_RAM 0x83
#define PACKET3_INCREMENT_CE_COUNTER 0x84
#define PACKET3_INCREMENT_DE_COUNTER 0x85
#define PACKET3_WAIT_ON_CE_COUNTER 0x86
#define PACKET3_WAIT_ON_DE_COUNTER_DIFF 0x88
#endif

View file

@ -121,7 +121,9 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority
p->ring = RADEON_RING_TYPE_GFX_INDEX;
break;
case RADEON_CS_RING_COMPUTE:
if (p->rdev->family >= CHIP_TAHITI) {
if (p->rdev->family >= CHIP_BONAIRE)
p->ring = RADEON_RING_TYPE_GFX_INDEX;
else if (p->rdev->family >= CHIP_TAHITI) {
if (p->priority > 0)
p->ring = CAYMAN_RING_TYPE_CP1_INDEX;
else