amdgpu_ucode.c 12.0 KB
Newer Older
A
Alex Deucher 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
/*
 * Copyright 2014 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 <linux/firmware.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <drm/drmP.h>
#include "amdgpu.h"
#include "amdgpu_ucode.h"

static void amdgpu_ucode_print_common_hdr(const struct common_firmware_header *hdr)
{
	DRM_DEBUG("size_bytes: %u\n", le32_to_cpu(hdr->size_bytes));
	DRM_DEBUG("header_size_bytes: %u\n", le32_to_cpu(hdr->header_size_bytes));
	DRM_DEBUG("header_version_major: %u\n", le16_to_cpu(hdr->header_version_major));
	DRM_DEBUG("header_version_minor: %u\n", le16_to_cpu(hdr->header_version_minor));
	DRM_DEBUG("ip_version_major: %u\n", le16_to_cpu(hdr->ip_version_major));
	DRM_DEBUG("ip_version_minor: %u\n", le16_to_cpu(hdr->ip_version_minor));
	DRM_DEBUG("ucode_version: 0x%08x\n", le32_to_cpu(hdr->ucode_version));
	DRM_DEBUG("ucode_size_bytes: %u\n", le32_to_cpu(hdr->ucode_size_bytes));
	DRM_DEBUG("ucode_array_offset_bytes: %u\n",
		  le32_to_cpu(hdr->ucode_array_offset_bytes));
	DRM_DEBUG("crc32: 0x%08x\n", le32_to_cpu(hdr->crc32));
}

void amdgpu_ucode_print_mc_hdr(const struct common_firmware_header *hdr)
{
	uint16_t version_major = le16_to_cpu(hdr->header_version_major);
	uint16_t version_minor = le16_to_cpu(hdr->header_version_minor);

	DRM_DEBUG("MC\n");
	amdgpu_ucode_print_common_hdr(hdr);

	if (version_major == 1) {
		const struct mc_firmware_header_v1_0 *mc_hdr =
			container_of(hdr, struct mc_firmware_header_v1_0, header);

		DRM_DEBUG("io_debug_size_bytes: %u\n",
			  le32_to_cpu(mc_hdr->io_debug_size_bytes));
		DRM_DEBUG("io_debug_array_offset_bytes: %u\n",
			  le32_to_cpu(mc_hdr->io_debug_array_offset_bytes));
	} else {
		DRM_ERROR("Unknown MC ucode version: %u.%u\n", version_major, version_minor);
	}
}

void amdgpu_ucode_print_smc_hdr(const struct common_firmware_header *hdr)
{
	uint16_t version_major = le16_to_cpu(hdr->header_version_major);
	uint16_t version_minor = le16_to_cpu(hdr->header_version_minor);

	DRM_DEBUG("SMC\n");
	amdgpu_ucode_print_common_hdr(hdr);

	if (version_major == 1) {
		const struct smc_firmware_header_v1_0 *smc_hdr =
			container_of(hdr, struct smc_firmware_header_v1_0, header);

		DRM_DEBUG("ucode_start_addr: %u\n", le32_to_cpu(smc_hdr->ucode_start_addr));
	} else {
		DRM_ERROR("Unknown SMC ucode version: %u.%u\n", version_major, version_minor);
	}
}

void amdgpu_ucode_print_gfx_hdr(const struct common_firmware_header *hdr)
{
	uint16_t version_major = le16_to_cpu(hdr->header_version_major);
	uint16_t version_minor = le16_to_cpu(hdr->header_version_minor);

	DRM_DEBUG("GFX\n");
	amdgpu_ucode_print_common_hdr(hdr);

	if (version_major == 1) {
		const struct gfx_firmware_header_v1_0 *gfx_hdr =
			container_of(hdr, struct gfx_firmware_header_v1_0, header);

		DRM_DEBUG("ucode_feature_version: %u\n",
			  le32_to_cpu(gfx_hdr->ucode_feature_version));
		DRM_DEBUG("jt_offset: %u\n", le32_to_cpu(gfx_hdr->jt_offset));
		DRM_DEBUG("jt_size: %u\n", le32_to_cpu(gfx_hdr->jt_size));
	} else {
		DRM_ERROR("Unknown GFX ucode version: %u.%u\n", version_major, version_minor);
	}
}

void amdgpu_ucode_print_rlc_hdr(const struct common_firmware_header *hdr)
{
	uint16_t version_major = le16_to_cpu(hdr->header_version_major);
	uint16_t version_minor = le16_to_cpu(hdr->header_version_minor);

	DRM_DEBUG("RLC\n");
	amdgpu_ucode_print_common_hdr(hdr);

	if (version_major == 1) {
		const struct rlc_firmware_header_v1_0 *rlc_hdr =
			container_of(hdr, struct rlc_firmware_header_v1_0, header);

		DRM_DEBUG("ucode_feature_version: %u\n",
			  le32_to_cpu(rlc_hdr->ucode_feature_version));
		DRM_DEBUG("save_and_restore_offset: %u\n",
			  le32_to_cpu(rlc_hdr->save_and_restore_offset));
		DRM_DEBUG("clear_state_descriptor_offset: %u\n",
			  le32_to_cpu(rlc_hdr->clear_state_descriptor_offset));
		DRM_DEBUG("avail_scratch_ram_locations: %u\n",
			  le32_to_cpu(rlc_hdr->avail_scratch_ram_locations));
		DRM_DEBUG("master_pkt_description_offset: %u\n",
			  le32_to_cpu(rlc_hdr->master_pkt_description_offset));
	} else if (version_major == 2) {
		const struct rlc_firmware_header_v2_0 *rlc_hdr =
			container_of(hdr, struct rlc_firmware_header_v2_0, header);

		DRM_DEBUG("ucode_feature_version: %u\n",
			  le32_to_cpu(rlc_hdr->ucode_feature_version));
		DRM_DEBUG("jt_offset: %u\n", le32_to_cpu(rlc_hdr->jt_offset));
		DRM_DEBUG("jt_size: %u\n", le32_to_cpu(rlc_hdr->jt_size));
		DRM_DEBUG("save_and_restore_offset: %u\n",
			  le32_to_cpu(rlc_hdr->save_and_restore_offset));
		DRM_DEBUG("clear_state_descriptor_offset: %u\n",
			  le32_to_cpu(rlc_hdr->clear_state_descriptor_offset));
		DRM_DEBUG("avail_scratch_ram_locations: %u\n",
			  le32_to_cpu(rlc_hdr->avail_scratch_ram_locations));
		DRM_DEBUG("reg_restore_list_size: %u\n",
			  le32_to_cpu(rlc_hdr->reg_restore_list_size));
		DRM_DEBUG("reg_list_format_start: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_format_start));
		DRM_DEBUG("reg_list_format_separate_start: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_format_separate_start));
		DRM_DEBUG("starting_offsets_start: %u\n",
			  le32_to_cpu(rlc_hdr->starting_offsets_start));
		DRM_DEBUG("reg_list_format_size_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_format_size_bytes));
		DRM_DEBUG("reg_list_format_array_offset_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_format_array_offset_bytes));
		DRM_DEBUG("reg_list_size_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_size_bytes));
		DRM_DEBUG("reg_list_array_offset_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_array_offset_bytes));
		DRM_DEBUG("reg_list_format_separate_size_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_format_separate_size_bytes));
		DRM_DEBUG("reg_list_format_separate_array_offset_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_format_separate_array_offset_bytes));
		DRM_DEBUG("reg_list_separate_size_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_separate_size_bytes));
		DRM_DEBUG("reg_list_separate_size_bytes: %u\n",
			  le32_to_cpu(rlc_hdr->reg_list_separate_size_bytes));
	} else {
		DRM_ERROR("Unknown RLC ucode version: %u.%u\n", version_major, version_minor);
	}
}

void amdgpu_ucode_print_sdma_hdr(const struct common_firmware_header *hdr)
{
	uint16_t version_major = le16_to_cpu(hdr->header_version_major);
	uint16_t version_minor = le16_to_cpu(hdr->header_version_minor);

	DRM_DEBUG("SDMA\n");
	amdgpu_ucode_print_common_hdr(hdr);

	if (version_major == 1) {
		const struct sdma_firmware_header_v1_0 *sdma_hdr =
			container_of(hdr, struct sdma_firmware_header_v1_0, header);

		DRM_DEBUG("ucode_feature_version: %u\n",
			  le32_to_cpu(sdma_hdr->ucode_feature_version));
		DRM_DEBUG("ucode_change_version: %u\n",
			  le32_to_cpu(sdma_hdr->ucode_change_version));
		DRM_DEBUG("jt_offset: %u\n", le32_to_cpu(sdma_hdr->jt_offset));
		DRM_DEBUG("jt_size: %u\n", le32_to_cpu(sdma_hdr->jt_size));
		if (version_minor >= 1) {
			const struct sdma_firmware_header_v1_1 *sdma_v1_1_hdr =
				container_of(sdma_hdr, struct sdma_firmware_header_v1_1, v1_0);
			DRM_DEBUG("digest_size: %u\n", le32_to_cpu(sdma_v1_1_hdr->digest_size));
		}
	} else {
		DRM_ERROR("Unknown SDMA ucode version: %u.%u\n",
			  version_major, version_minor);
	}
}

int amdgpu_ucode_validate(const struct firmware *fw)
{
	const struct common_firmware_header *hdr =
		(const struct common_firmware_header *)fw->data;

	if (fw->size == le32_to_cpu(hdr->size_bytes))
		return 0;

	return -EINVAL;
}

bool amdgpu_ucode_hdr_version(union amdgpu_firmware_header *hdr,
				uint16_t hdr_major, uint16_t hdr_minor)
{
	if ((hdr->common.header_version_major == hdr_major) &&
		(hdr->common.header_version_minor == hdr_minor))
		return false;
	return true;
}

static int amdgpu_ucode_init_single_fw(struct amdgpu_firmware_info *ucode,
				uint64_t mc_addr, void *kptr)
{
	const struct common_firmware_header *header = NULL;

	if (NULL == ucode->fw)
		return 0;

	ucode->mc_addr = mc_addr;
	ucode->kaddr = kptr;

231 232 233
	if (ucode->ucode_id == AMDGPU_UCODE_ID_STORAGE)
		return 0;

A
Alex Deucher 已提交
234 235 236 237 238 239 240 241
	header = (const struct common_firmware_header *)ucode->fw->data;
	memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
		le32_to_cpu(header->ucode_array_offset_bytes)),
		le32_to_cpu(header->ucode_size_bytes));

	return 0;
}

242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266
static int amdgpu_ucode_patch_jt(struct amdgpu_firmware_info *ucode,
				uint64_t mc_addr, void *kptr)
{
	const struct gfx_firmware_header_v1_0 *header = NULL;
	const struct common_firmware_header *comm_hdr = NULL;
	uint8_t* src_addr = NULL;
	uint8_t* dst_addr = NULL;

	if (NULL == ucode->fw)
		return 0;

	comm_hdr = (const struct common_firmware_header *)ucode->fw->data;
	header = (const struct gfx_firmware_header_v1_0 *)ucode->fw->data;
	dst_addr = ucode->kaddr +
			   ALIGN(le32_to_cpu(comm_hdr->ucode_size_bytes),
			   PAGE_SIZE);
	src_addr = (uint8_t *)ucode->fw->data +
			   le32_to_cpu(comm_hdr->ucode_array_offset_bytes) +
			   (le32_to_cpu(header->jt_offset) * 4);
	memcpy(dst_addr, src_addr, le32_to_cpu(header->jt_size) * 4);

	return 0;
}


A
Alex Deucher 已提交
267 268 269 270 271 272 273 274 275 276 277
int amdgpu_ucode_init_bo(struct amdgpu_device *adev)
{
	struct amdgpu_bo **bo = &adev->firmware.fw_buf;
	uint64_t fw_mc_addr;
	void *fw_buf_ptr = NULL;
	uint64_t fw_offset = 0;
	int i, err;
	struct amdgpu_firmware_info *ucode = NULL;
	const struct common_firmware_header *header = NULL;

	err = amdgpu_bo_create(adev, adev->firmware.fw_size, PAGE_SIZE, true,
278 279
				amdgpu_sriov_vf(adev) ? AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT,
				0, NULL, NULL, bo);
A
Alex Deucher 已提交
280 281 282 283 284 285 286 287
	if (err) {
		dev_err(adev->dev, "(%d) Firmware buffer allocate failed\n", err);
		goto failed;
	}

	err = amdgpu_bo_reserve(*bo, false);
	if (err) {
		dev_err(adev->dev, "(%d) Firmware buffer reserve failed\n", err);
288
		goto failed_reserve;
A
Alex Deucher 已提交
289 290
	}

291 292
	err = amdgpu_bo_pin(*bo, amdgpu_sriov_vf(adev) ? AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT,
				&fw_mc_addr);
A
Alex Deucher 已提交
293 294
	if (err) {
		dev_err(adev->dev, "(%d) Firmware buffer pin failed\n", err);
295
		goto failed_pin;
A
Alex Deucher 已提交
296 297 298 299 300
	}

	err = amdgpu_bo_kmap(*bo, &fw_buf_ptr);
	if (err) {
		dev_err(adev->dev, "(%d) Firmware buffer kmap failed\n", err);
301
		goto failed_kmap;
A
Alex Deucher 已提交
302 303 304 305 306 307 308 309 310 311
	}

	amdgpu_bo_unreserve(*bo);

	for (i = 0; i < AMDGPU_UCODE_ID_MAXIMUM; i++) {
		ucode = &adev->firmware.ucode[i];
		if (ucode->fw) {
			header = (const struct common_firmware_header *)ucode->fw->data;
			amdgpu_ucode_init_single_fw(ucode, fw_mc_addr + fw_offset,
						    fw_buf_ptr + fw_offset);
312 313 314 315 316 317 318
			if (i == AMDGPU_UCODE_ID_CP_MEC1) {
				const struct gfx_firmware_header_v1_0 *cp_hdr;
				cp_hdr = (const struct gfx_firmware_header_v1_0 *)ucode->fw->data;
				amdgpu_ucode_patch_jt(ucode, fw_mc_addr + fw_offset,
						    fw_buf_ptr + fw_offset);
				fw_offset += ALIGN(le32_to_cpu(cp_hdr->jt_size) << 2, PAGE_SIZE);
			}
A
Alex Deucher 已提交
319 320 321
			fw_offset += ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE);
		}
	}
322
	return 0;
A
Alex Deucher 已提交
323

324 325 326 327 328 329
failed_kmap:
	amdgpu_bo_unpin(*bo);
failed_pin:
	amdgpu_bo_unreserve(*bo);
failed_reserve:
	amdgpu_bo_unref(bo);
A
Alex Deucher 已提交
330
failed:
331
	adev->firmware.smu_load = false;
A
Alex Deucher 已提交
332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352

	return err;
}

int amdgpu_ucode_fini_bo(struct amdgpu_device *adev)
{
	int i;
	struct amdgpu_firmware_info *ucode = NULL;

	for (i = 0; i < AMDGPU_UCODE_ID_MAXIMUM; i++) {
		ucode = &adev->firmware.ucode[i];
		if (ucode->fw) {
			ucode->mc_addr = 0;
			ucode->kaddr = NULL;
		}
	}
	amdgpu_bo_unref(&adev->firmware.fw_buf);
	adev->firmware.fw_buf = NULL;

	return 0;
}