tree: https://gitlab.freedesktop.org/agd5f/linux.git drm-next head: e90c92dc6b09574f7fe2e7e5774a3af30e36c632 commit: 86541dd5d1690a823cd5e6505522825e832f83e4 [356/428] drm/amdkfd: add aldebaran kfd2kgd callbacks to kfd device (v2) config: powerpc-randconfig-r035-20210309 (attached as .config) compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project cd9a69289c7825d54450cb6829fef2c8e0f1963a) reproduce (this is a W=1 build): wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross # install powerpc cross compiling tool for clang build # apt-get install binutils-powerpc-linux-gnu git remote add agd5f https://gitlab.freedesktop.org/agd5f/linux.git git fetch --no-tags agd5f drm-next git checkout 86541dd5d1690a823cd5e6505522825e832f83e4 # save the attached .config to linux build tree COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=powerpc If you fix the issue, kindly add following tag as appropriate Reported-by: kernel test robot <lkp(a)intel.com&gt; All warnings (new ones prefixed by >>): int kgd_arcturus_hqd_sdma_load(struct kgd_dev *kgd, void *mqd, ^ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c:125:1: note: declare 'static' if the function is not intended to be used outside of this translation unit int kgd_arcturus_hqd_sdma_load(struct kgd_dev *kgd, void *mqd, ^ static int kgd_arcturus_hqd_sdma_dump(struct kgd_dev *kgd, ^ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c:195:1: note: declare 'static' if the function is not intended to be used outside of this translation unit int kgd_arcturus_hqd_sdma_dump(struct kgd_dev *kgd, ^ static bool kgd_arcturus_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd) ^ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c:227:1: note: declare 'static' if the function is not intended to be used outside of this translation unit bool kgd_arcturus_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd) ^ static int kgd_arcturus_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, ^ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c:246:1: note: declare 'static' if the function is not intended to be used outside of this translation unit int kgd_arcturus_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, ^ static 4 warnings generated. vim +/kgd_arcturus_hqd_sdma_load +125 drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c 124 126 uint32_t __user *wptr, struct mm_struct *mm) 127 { 128 struct amdgpu_device *adev = get_amdgpu_device(kgd); 129 struct v9_sdma_mqd *m; 130 uint32_t sdma_rlc_reg_offset; 131 unsigned long end_jiffies; 132 uint32_t data; 133 uint64_t data64; 134 uint64_t __user *wptr64 = (uint64_t __user *)wptr; 135 136 m = get_sdma_mqd(mqd); 137 sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id, 138 m->sdma_queue_id); 139 140 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, 141 m->sdmax_rlcx_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)); 142 143 end_jiffies = msecs_to_jiffies(2000) + jiffies; 144 while (true) { 145 data = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS); 146 if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK) 147 break; 148 if (time_after(jiffies, end_jiffies)) { 149 pr_err("SDMA RLC not idle in %s\n", __func__); 150 return -ETIME; 151 } 152 usleep_range(500, 1000); 153 } 154 155 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL_OFFSET, 156 m->sdmax_rlcx_doorbell_offset); 157 158 data = REG_SET_FIELD(m->sdmax_rlcx_doorbell, SDMA0_RLC0_DOORBELL, 159 ENABLE, 1); 160 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, data); 161 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR, 162 m->sdmax_rlcx_rb_rptr); 163 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI, 164 m->sdmax_rlcx_rb_rptr_hi); 165 166 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1); 167 if (read_user_wptr(mm, wptr64, data64)) { 168 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR, 169 lower_32_bits(data64)); 170 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI, 171 upper_32_bits(data64)); 172 } else { 173 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR, 174 m->sdmax_rlcx_rb_rptr); 175 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_WPTR_HI, 176 m->sdmax_rlcx_rb_rptr_hi); 177 } 178 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0); 179 180 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base); 181 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_BASE_HI, 182 m->sdmax_rlcx_rb_base_hi); 183 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_LO, 184 m->sdmax_rlcx_rb_rptr_addr_lo); 185 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_ADDR_HI, 186 m->sdmax_rlcx_rb_rptr_addr_hi); 187 188 data = REG_SET_FIELD(m->sdmax_rlcx_rb_cntl, SDMA0_RLC0_RB_CNTL, 189 RB_ENABLE, 1); 190 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, data); 191 192 return 0; 193 } 194 196 uint32_t engine_id, uint32_t queue_id, 197 uint32_t (**dump)[2], uint32_t *n_regs) 198 { 199 struct amdgpu_device *adev = get_amdgpu_device(kgd); 200 uint32_t sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, 201 engine_id, queue_id); 202 uint32_t i = 0, reg; 203 #undef HQD_N_REGS 204 #define HQD_N_REGS (19+6+7+10) 205 206 *dump = kmalloc_array(HQD_N_REGS * 2, sizeof(uint32_t), GFP_KERNEL); 207 if (*dump == NULL) 208 return -ENOMEM; 209 210 for (reg = mmSDMA0_RLC0_RB_CNTL; reg <= mmSDMA0_RLC0_DOORBELL; reg++) 211 DUMP_REG(sdma_rlc_reg_offset + reg); 212 for (reg = mmSDMA0_RLC0_STATUS; reg <= mmSDMA0_RLC0_CSA_ADDR_HI; reg++) 213 DUMP_REG(sdma_rlc_reg_offset + reg); 214 for (reg = mmSDMA0_RLC0_IB_SUB_REMAIN; 215 reg <= mmSDMA0_RLC0_MINOR_PTR_UPDATE; reg++) 216 DUMP_REG(sdma_rlc_reg_offset + reg); 217 for (reg = mmSDMA0_RLC0_MIDCMD_DATA0; 218 reg <= mmSDMA0_RLC0_MIDCMD_CNTL; reg++) 219 DUMP_REG(sdma_rlc_reg_offset + reg); 220 221 WARN_ON_ONCE(i != HQD_N_REGS); 222 *n_regs = i; 223 224 return 0; 225 } 226 228 { 229 struct amdgpu_device *adev = get_amdgpu_device(kgd); 230 struct v9_sdma_mqd *m; 231 uint32_t sdma_rlc_reg_offset; 232 uint32_t sdma_rlc_rb_cntl; 233 234 m = get_sdma_mqd(mqd); 235 sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id, 236 m->sdma_queue_id); 237 238 sdma_rlc_rb_cntl = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL); 239 240 if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK) 241 return true; 242 243 return false; 244 } 245 247 unsigned int utimeout) 248 { 249 struct amdgpu_device *adev = get_amdgpu_device(kgd); 250 struct v9_sdma_mqd *m; 251 uint32_t sdma_rlc_reg_offset; 252 uint32_t temp; 253 unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies; 254 255 m = get_sdma_mqd(mqd); 256 sdma_rlc_reg_offset = get_sdma_rlc_reg_offset(adev, m->sdma_engine_id, 257 m->sdma_queue_id); 258 259 temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL); 260 temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK; 261 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, temp); 262 263 while (true) { 264 temp = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_CONTEXT_STATUS); 265 if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK) 266 break; 267 if (time_after(jiffies, end_jiffies)) { 268 pr_err("SDMA RLC not idle in %s\n", __func__); 269 return -ETIME; 270 } 271 usleep_range(500, 1000); 272 } 273 274 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_DOORBELL, 0); 275 WREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL, 276 RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_CNTL) | 277 SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK); 278 279 m->sdmax_rlcx_rb_rptr = RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR); 280 m->sdmax_rlcx_rb_rptr_hi = 281 RREG32(sdma_rlc_reg_offset + mmSDMA0_RLC0_RB_RPTR_HI); 282 283 return 0; 284 } 285 --- 0-DAY CI Kernel Test Service, Intel Corporation https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org