xenocara/lib/libdrm/tests/amdgpu/vcn_tests.c

/*
 * Copyright 2017 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 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.
 *
*/

#include <stdio.h>
#include <inttypes.h>

#include "CUnit/Basic.h"

#include "util_math.h"

#include "amdgpu_test.h"
#include "amdgpu_drm.h"
#include "amdgpu_internal.h"
#include "decode_messages.h"

#define IB_SIZE		4096
#define MAX_RESOURCES	16

struct amdgpu_vcn_bo {
	amdgpu_bo_handle handle;
	amdgpu_va_handle va_handle;
	uint64_t addr;
	uint64_t size;
	uint8_t *ptr;
};

struct amdgpu_vcn_reg {
	uint32_t data0;
	uint32_t data1;
	uint32_t cmd;
	uint32_t nop;
	uint32_t cntl;
};

static amdgpu_device_handle device_handle;
static uint32_t major_version;
static uint32_t minor_version;
static uint32_t family_id;

static amdgpu_context_handle context_handle;
static amdgpu_bo_handle ib_handle;
static amdgpu_va_handle ib_va_handle;
static uint64_t ib_mc_address;
static uint32_t *ib_cpu;

static amdgpu_bo_handle resources[MAX_RESOURCES];
static unsigned num_resources;
static struct amdgpu_vcn_reg reg;

static void amdgpu_cs_vcn_dec_create(void);
static void amdgpu_cs_vcn_dec_decode(void);
static void amdgpu_cs_vcn_dec_destroy(void);

static void amdgpu_cs_vcn_enc_create(void);
static void amdgpu_cs_vcn_enc_encode(void);
static void amdgpu_cs_vcn_enc_destroy(void);

CU_TestInfo vcn_tests[] = {

	{ "VCN DEC create",  amdgpu_cs_vcn_dec_create },
	{ "VCN DEC decode",  amdgpu_cs_vcn_dec_decode },
	{ "VCN DEC destroy",  amdgpu_cs_vcn_dec_destroy },

	{ "VCN ENC create",  amdgpu_cs_vcn_enc_create },
	{ "VCN ENC decode",  amdgpu_cs_vcn_enc_encode },
	{ "VCN ENC destroy",  amdgpu_cs_vcn_enc_destroy },
	CU_TEST_INFO_NULL,
};

CU_BOOL suite_vcn_tests_enable(void)
{

	if (amdgpu_device_initialize(drm_amdgpu[0], &major_version,
				   &minor_version, &device_handle))
		return CU_FALSE;

	family_id = device_handle->info.family_id;

	if (amdgpu_device_deinitialize(device_handle))
			return CU_FALSE;


	if (family_id < AMDGPU_FAMILY_RV) {
		printf("\n\nThe ASIC NOT support VCN, suite disabled\n");
		return CU_FALSE;
	}

	if (family_id == AMDGPU_FAMILY_RV) {
		reg.data0 = 0x81c4;
		reg.data1 = 0x81c5;
		reg.cmd = 0x81c3;
		reg.nop = 0x81ff;
		reg.cntl = 0x81c6;
	} else if (family_id == AMDGPU_FAMILY_NV) {
		reg.data0 = 0x504;
		reg.data1 = 0x505;
		reg.cmd = 0x503;
		reg.nop = 0x53f;
		reg.cntl = 0x506;
	} else
		return CU_FALSE;

	return CU_TRUE;
}

int suite_vcn_tests_init(void)
{
	int r;

	r = amdgpu_device_initialize(drm_amdgpu[0], &major_version,
				     &minor_version, &device_handle);
	if (r)
		return CUE_SINIT_FAILED;

	family_id = device_handle->info.family_id;

	r = amdgpu_cs_ctx_create(device_handle, &context_handle);
	if (r)
		return CUE_SINIT_FAILED;

	r = amdgpu_bo_alloc_and_map(device_handle, IB_SIZE, 4096,
				    AMDGPU_GEM_DOMAIN_GTT, 0,
				    &ib_handle, (void**)&ib_cpu,
				    &ib_mc_address, &ib_va_handle);
	if (r)
		return CUE_SINIT_FAILED;

	return CUE_SUCCESS;
}

int suite_vcn_tests_clean(void)
{
	int r;

	r = amdgpu_bo_unmap_and_free(ib_handle, ib_va_handle,
			     ib_mc_address, IB_SIZE);
	if (r)
		return CUE_SCLEAN_FAILED;

	r = amdgpu_cs_ctx_free(context_handle);
	if (r)
		return CUE_SCLEAN_FAILED;

	r = amdgpu_device_deinitialize(device_handle);
	if (r)
		return CUE_SCLEAN_FAILED;

	return CUE_SUCCESS;
}

static int submit(unsigned ndw, unsigned ip)
{
	struct amdgpu_cs_request ibs_request = {0};
	struct amdgpu_cs_ib_info ib_info = {0};
	struct amdgpu_cs_fence fence_status = {0};
	uint32_t expired;
	int r;

	ib_info.ib_mc_address = ib_mc_address;
	ib_info.size = ndw;

	ibs_request.ip_type = ip;

	r = amdgpu_bo_list_create(device_handle, num_resources, resources,
				  NULL, &ibs_request.resources);
	if (r)
		return r;

	ibs_request.number_of_ibs = 1;
	ibs_request.ibs = &ib_info;
	ibs_request.fence_info.handle = NULL;

	r = amdgpu_cs_submit(context_handle, 0, &ibs_request, 1);
	if (r)
		return r;

	r = amdgpu_bo_list_destroy(ibs_request.resources);
	if (r)
		return r;

	fence_status.context = context_handle;
	fence_status.ip_type = ip;
	fence_status.fence = ibs_request.seq_no;

	r = amdgpu_cs_query_fence_status(&fence_status,
					 AMDGPU_TIMEOUT_INFINITE,
					 0, &expired);
	if (r)
		return r;

	return 0;
}

static void alloc_resource(struct amdgpu_vcn_bo *vcn_bo,
			unsigned size, unsigned domain)
{
	struct amdgpu_bo_alloc_request req = {0};
	amdgpu_bo_handle buf_handle;
	amdgpu_va_handle va_handle;
	uint64_t va = 0;
	int r;

	req.alloc_size = ALIGN(size, 4096);
	req.preferred_heap = domain;
	r = amdgpu_bo_alloc(device_handle, &req, &buf_handle);
	CU_ASSERT_EQUAL(r, 0);
	r = amdgpu_va_range_alloc(device_handle,
				  amdgpu_gpu_va_range_general,
				  req.alloc_size, 1, 0, &va,
				  &va_handle, 0);
	CU_ASSERT_EQUAL(r, 0);
	r = amdgpu_bo_va_op(buf_handle, 0, req.alloc_size, va, 0,
			    AMDGPU_VA_OP_MAP);
	CU_ASSERT_EQUAL(r, 0);
	vcn_bo->addr = va;
	vcn_bo->handle = buf_handle;
	vcn_bo->size = req.alloc_size;
	vcn_bo->va_handle = va_handle;
	r = amdgpu_bo_cpu_map(vcn_bo->handle, (void **)&vcn_bo->ptr);
	CU_ASSERT_EQUAL(r, 0);
	memset(vcn_bo->ptr, 0, size);
	r = amdgpu_bo_cpu_unmap(vcn_bo->handle);
	CU_ASSERT_EQUAL(r, 0);
}

static void free_resource(struct amdgpu_vcn_bo *vcn_bo)
{
	int r;

	r = amdgpu_bo_va_op(vcn_bo->handle, 0, vcn_bo->size,
			    vcn_bo->addr, 0, AMDGPU_VA_OP_UNMAP);
	CU_ASSERT_EQUAL(r, 0);

	r = amdgpu_va_range_free(vcn_bo->va_handle);
	CU_ASSERT_EQUAL(r, 0);

	r = amdgpu_bo_free(vcn_bo->handle);
	CU_ASSERT_EQUAL(r, 0);
	memset(vcn_bo, 0, sizeof(*vcn_bo));
}

static void vcn_dec_cmd(uint64_t addr, unsigned cmd, int *idx)
{
	ib_cpu[(*idx)++] = reg.data0;
	ib_cpu[(*idx)++] = addr;
	ib_cpu[(*idx)++] = reg.data1;
	ib_cpu[(*idx)++] = addr >> 32;
	ib_cpu[(*idx)++] = reg.cmd;
	ib_cpu[(*idx)++] = cmd << 1;
}

static void amdgpu_cs_vcn_dec_create(void)
{
	struct amdgpu_vcn_bo msg_buf;
	int len, r;

	num_resources  = 0;
	alloc_resource(&msg_buf, 4096, AMDGPU_GEM_DOMAIN_GTT);
	resources[num_resources++] = msg_buf.handle;
	resources[num_resources++] = ib_handle;

	r = amdgpu_bo_cpu_map(msg_buf.handle, (void **)&msg_buf.ptr);
	CU_ASSERT_EQUAL(r, 0);

	memset(msg_buf.ptr, 0, 4096);
	memcpy(msg_buf.ptr, vcn_dec_create_msg, sizeof(vcn_dec_create_msg));

	len = 0;
	ib_cpu[len++] = reg.data0;
	ib_cpu[len++] = msg_buf.addr;
	ib_cpu[len++] = reg.data1;
	ib_cpu[len++] = msg_buf.addr >> 32;
	ib_cpu[len++] = reg.cmd;
	ib_cpu[len++] = 0;
	for (; len % 16; ) {
		ib_cpu[len++] = reg.nop;
		ib_cpu[len++] = 0;
	}

	r = submit(len, AMDGPU_HW_IP_VCN_DEC);
	CU_ASSERT_EQUAL(r, 0);

	free_resource(&msg_buf);
}

static void amdgpu_cs_vcn_dec_decode(void)
{
	const unsigned dpb_size = 15923584, dt_size = 737280;
	uint64_t msg_addr, fb_addr, bs_addr, dpb_addr, ctx_addr, dt_addr, it_addr, sum;
	struct amdgpu_vcn_bo dec_buf;
	int size, len, i, r;
	uint8_t *dec;

	size = 4*1024; /* msg */
	size += 4*1024; /* fb */
	size += 4096; /*it_scaling_table*/
	size += ALIGN(sizeof(uvd_bitstream), 4*1024);
	size += ALIGN(dpb_size, 4*1024);
	size += ALIGN(dt_size, 4*1024);

	num_resources  = 0;
	alloc_resource(&dec_buf, size, AMDGPU_GEM_DOMAIN_GTT);
	resources[num_resources++] = dec_buf.handle;
	resources[num_resources++] = ib_handle;

	r = amdgpu_bo_cpu_map(dec_buf.handle, (void **)&dec_buf.ptr);
	dec = dec_buf.ptr;

	CU_ASSERT_EQUAL(r, 0);
	memset(dec_buf.ptr, 0, size);
	memcpy(dec_buf.ptr, vcn_dec_decode_msg, sizeof(vcn_dec_decode_msg));
	memcpy(dec_buf.ptr + sizeof(vcn_dec_decode_msg),
			avc_decode_msg, sizeof(avc_decode_msg));

	dec += 4*1024;
	dec += 4*1024;
	memcpy(dec, uvd_it_scaling_table, sizeof(uvd_it_scaling_table));

	dec += 4*1024;
	memcpy(dec, uvd_bitstream, sizeof(uvd_bitstream));

	dec += ALIGN(sizeof(uvd_bitstream), 4*1024);

	dec += ALIGN(dpb_size, 4*1024);

	msg_addr = dec_buf.addr;
	fb_addr = msg_addr + 4*1024;
	it_addr = fb_addr + 4*1024;
	bs_addr = it_addr + 4*1024;
	dpb_addr = ALIGN(bs_addr + sizeof(uvd_bitstream), 4*1024);
	ctx_addr = ALIGN(dpb_addr + 0x006B9400, 4*1024);
	dt_addr = ALIGN(dpb_addr + dpb_size, 4*1024);

	len = 0;
	vcn_dec_cmd(msg_addr, 0x0, &len);
	vcn_dec_cmd(dpb_addr, 0x1, &len);
	vcn_dec_cmd(dt_addr, 0x2, &len);
	vcn_dec_cmd(fb_addr, 0x3, &len);
	vcn_dec_cmd(bs_addr, 0x100, &len);
	vcn_dec_cmd(it_addr, 0x204, &len);
	vcn_dec_cmd(ctx_addr, 0x206, &len);

	ib_cpu[len++] = reg.cntl;
	ib_cpu[len++] = 0x1;
	for (; len % 16; ) {
		ib_cpu[len++] = reg.nop;
		ib_cpu[len++] = 0;
	}

	r = submit(len, AMDGPU_HW_IP_VCN_DEC);
	CU_ASSERT_EQUAL(r, 0);

	for (i = 0, sum = 0; i < dt_size; ++i)
		sum += dec[i];

	CU_ASSERT_EQUAL(sum, SUM_DECODE);

	free_resource(&dec_buf);
}

static void amdgpu_cs_vcn_dec_destroy(void)
{
	struct amdgpu_vcn_bo msg_buf;
	int len, r;

	num_resources  = 0;
	alloc_resource(&msg_buf, 1024, AMDGPU_GEM_DOMAIN_GTT);
	resources[num_resources++] = msg_buf.handle;
	resources[num_resources++] = ib_handle;

	r = amdgpu_bo_cpu_map(msg_buf.handle, (void **)&msg_buf.ptr);
	CU_ASSERT_EQUAL(r, 0);

	memset(msg_buf.ptr, 0, 1024);
	memcpy(msg_buf.ptr, vcn_dec_destroy_msg, sizeof(vcn_dec_destroy_msg));

	len = 0;
	ib_cpu[len++] = reg.data0;
	ib_cpu[len++] = msg_buf.addr;
	ib_cpu[len++] = reg.data1;
	ib_cpu[len++] = msg_buf.addr >> 32;
	ib_cpu[len++] = reg.cmd;
	ib_cpu[len++] = 0;
	for (; len % 16; ) {
		ib_cpu[len++] = reg.nop;
		ib_cpu[len++] = 0;
	}

	r = submit(len, AMDGPU_HW_IP_VCN_DEC);
	CU_ASSERT_EQUAL(r, 0);

	free_resource(&msg_buf);
}

static void amdgpu_cs_vcn_enc_create(void)
{
	/* TODO */
}

static void amdgpu_cs_vcn_enc_encode(void)
{
	/* TODO */
}

static void amdgpu_cs_vcn_enc_destroy(void)
{
	/* TODO */
}