diff options
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c | 196 |
1 files changed, 166 insertions, 30 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c index 985fe704203e..ecf8ceb53311 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c @@ -46,6 +46,7 @@ #include "clearstate_gfx11.h" #include "v11_structs.h" #include "gfx_v11_0.h" +#include "gfx_v11_0_3.h" #include "nbio_v4_3.h" #include "mes_v11_0.h" @@ -753,8 +754,8 @@ static void gfx_v11_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, * zero here */ WARN_ON(simd != 0); - /* type 2 wave data */ - dst[(*no_fields)++] = 2; + /* type 3 wave data */ + dst[(*no_fields)++] = 3; dst[(*no_fields)++] = wave_read_ind(adev, wave, ixSQ_WAVE_STATUS); dst[(*no_fields)++] = wave_read_ind(adev, wave, ixSQ_WAVE_PC_LO); dst[(*no_fields)++] = wave_read_ind(adev, wave, ixSQ_WAVE_PC_HI); @@ -815,7 +816,14 @@ static int gfx_v11_0_gpu_early_init(struct amdgpu_device *adev) switch (adev->ip_versions[GC_HWIP][0]) { case IP_VERSION(11, 0, 0): case IP_VERSION(11, 0, 2): + adev->gfx.config.max_hw_contexts = 8; + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x100; + adev->gfx.config.sc_hiz_tile_fifo_size = 0; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x4C0; + break; case IP_VERSION(11, 0, 3): + adev->gfx.ras = &gfx_v11_0_3_ras; adev->gfx.config.max_hw_contexts = 8; adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; adev->gfx.config.sc_prim_fifo_size_backend = 0x100; @@ -1251,10 +1259,8 @@ static int gfx_v11_0_sw_init(void *handle) switch (adev->ip_versions[GC_HWIP][0]) { case IP_VERSION(11, 0, 0): - case IP_VERSION(11, 0, 1): case IP_VERSION(11, 0, 2): case IP_VERSION(11, 0, 3): - case IP_VERSION(11, 0, 4): adev->gfx.me.num_me = 1; adev->gfx.me.num_pipe_per_me = 1; adev->gfx.me.num_queue_per_pipe = 1; @@ -1262,6 +1268,15 @@ static int gfx_v11_0_sw_init(void *handle) adev->gfx.mec.num_pipe_per_mec = 4; adev->gfx.mec.num_queue_per_pipe = 4; break; + case IP_VERSION(11, 0, 1): + case IP_VERSION(11, 0, 4): + adev->gfx.me.num_me = 1; + adev->gfx.me.num_pipe_per_me = 1; + adev->gfx.me.num_queue_per_pipe = 1; + adev->gfx.mec.num_mec = 1; + adev->gfx.mec.num_pipe_per_mec = 4; + adev->gfx.mec.num_queue_per_pipe = 4; + break; default: adev->gfx.me.num_me = 1; adev->gfx.me.num_pipe_per_me = 1; @@ -1272,6 +1287,11 @@ static int gfx_v11_0_sw_init(void *handle) break; } + /* Enable CG flag in one VF mode for enabling RLC safe mode enter/exit */ + if (adev->ip_versions[GC_HWIP][0] == IP_VERSION(11, 0, 3) && + amdgpu_sriov_is_pp_one_vf(adev)) + adev->cg_flags = AMD_CG_SUPPORT_GFX_CGCG; + /* EOP Event */ r = amdgpu_irq_add_id(adev, SOC21_IH_CLIENTID_GRBM_CP, GFX_11_0_0__SRCID__CP_EOP_INTERRUPT, @@ -1293,6 +1313,20 @@ static int gfx_v11_0_sw_init(void *handle) if (r) return r; + /* ECC error */ + r = amdgpu_irq_add_id(adev, SOC21_IH_CLIENTID_GRBM_CP, + GFX_11_0_0__SRCID__CP_ECC_ERROR, + &adev->gfx.cp_ecc_error_irq); + if (r) + return r; + + /* FED error */ + r = amdgpu_irq_add_id(adev, SOC21_IH_CLIENTID_GFX, + GFX_11_0_0__SRCID__RLC_GC_FED_INTERRUPT, + &adev->gfx.rlc_gc_fed_irq); + if (r) + return r; + adev->gfx.gfx_current_status = AMDGPU_GFX_NORMAL_MODE; if (adev->gfx.imu.funcs) { @@ -1380,6 +1414,11 @@ static int gfx_v11_0_sw_init(void *handle) if (r) return r; + if (amdgpu_gfx_ras_sw_init(adev)) { + dev_err(adev->dev, "Failed to initialize gfx ras block!\n"); + return -EINVAL; + } + return 0; } @@ -1469,44 +1508,70 @@ static void gfx_v11_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, WREG32_SOC15(GC, 0, regGRBM_GFX_INDEX, data); } -static u32 gfx_v11_0_get_rb_active_bitmap(struct amdgpu_device *adev) +static u32 gfx_v11_0_get_sa_active_bitmap(struct amdgpu_device *adev) { - u32 data, mask; + u32 gc_disabled_sa_mask, gc_user_disabled_sa_mask, sa_mask; - data = RREG32_SOC15(GC, 0, regCC_RB_BACKEND_DISABLE); - data |= RREG32_SOC15(GC, 0, regGC_USER_RB_BACKEND_DISABLE); + gc_disabled_sa_mask = RREG32_SOC15(GC, 0, regCC_GC_SA_UNIT_DISABLE); + gc_disabled_sa_mask = REG_GET_FIELD(gc_disabled_sa_mask, + CC_GC_SA_UNIT_DISABLE, + SA_DISABLE); + gc_user_disabled_sa_mask = RREG32_SOC15(GC, 0, regGC_USER_SA_UNIT_DISABLE); + gc_user_disabled_sa_mask = REG_GET_FIELD(gc_user_disabled_sa_mask, + GC_USER_SA_UNIT_DISABLE, + SA_DISABLE); + sa_mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_sh_per_se * + adev->gfx.config.max_shader_engines); - data &= CC_RB_BACKEND_DISABLE__BACKEND_DISABLE_MASK; - data >>= GC_USER_RB_BACKEND_DISABLE__BACKEND_DISABLE__SHIFT; + return sa_mask & (~(gc_disabled_sa_mask | gc_user_disabled_sa_mask)); +} - mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_backends_per_se / - adev->gfx.config.max_sh_per_se); +static u32 gfx_v11_0_get_rb_active_bitmap(struct amdgpu_device *adev) +{ + u32 gc_disabled_rb_mask, gc_user_disabled_rb_mask; + u32 rb_mask; - return (~data) & mask; + gc_disabled_rb_mask = RREG32_SOC15(GC, 0, regCC_RB_BACKEND_DISABLE); + gc_disabled_rb_mask = REG_GET_FIELD(gc_disabled_rb_mask, + CC_RB_BACKEND_DISABLE, + BACKEND_DISABLE); + gc_user_disabled_rb_mask = RREG32_SOC15(GC, 0, regGC_USER_RB_BACKEND_DISABLE); + gc_user_disabled_rb_mask = REG_GET_FIELD(gc_user_disabled_rb_mask, + GC_USER_RB_BACKEND_DISABLE, + BACKEND_DISABLE); + rb_mask = amdgpu_gfx_create_bitmask(adev->gfx.config.max_backends_per_se * + adev->gfx.config.max_shader_engines); + + return rb_mask & (~(gc_disabled_rb_mask | gc_user_disabled_rb_mask)); } static void gfx_v11_0_setup_rb(struct amdgpu_device *adev) { - int i, j; - u32 data; - u32 active_rbs = 0; - u32 rb_bitmap_width_per_sh = adev->gfx.config.max_backends_per_se / - adev->gfx.config.max_sh_per_se; + u32 rb_bitmap_width_per_sa; + u32 max_sa; + u32 active_sa_bitmap; + u32 global_active_rb_bitmap; + u32 active_rb_bitmap = 0; + u32 i; - mutex_lock(&adev->grbm_idx_mutex); - for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { - for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { - gfx_v11_0_select_se_sh(adev, i, j, 0xffffffff); - data = gfx_v11_0_get_rb_active_bitmap(adev); - active_rbs |= data << ((i * adev->gfx.config.max_sh_per_se + j) * - rb_bitmap_width_per_sh); - } + /* query sa bitmap from SA_UNIT_DISABLE registers */ + active_sa_bitmap = gfx_v11_0_get_sa_active_bitmap(adev); + /* query rb bitmap from RB_BACKEND_DISABLE registers */ + global_active_rb_bitmap = gfx_v11_0_get_rb_active_bitmap(adev); + + /* generate active rb bitmap according to active sa bitmap */ + max_sa = adev->gfx.config.max_shader_engines * + adev->gfx.config.max_sh_per_se; + rb_bitmap_width_per_sa = adev->gfx.config.max_backends_per_se / + adev->gfx.config.max_sh_per_se; + for (i = 0; i < max_sa; i++) { + if (active_sa_bitmap & (1 << i)) + active_rb_bitmap |= (0x3 << (i * rb_bitmap_width_per_sa)); } - gfx_v11_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); - mutex_unlock(&adev->grbm_idx_mutex); - adev->gfx.config.backend_enable_mask = active_rbs; - adev->gfx.config.num_rbs = hweight32(active_rbs); + active_rb_bitmap |= global_active_rb_bitmap; + adev->gfx.config.backend_enable_mask = active_rb_bitmap; + adev->gfx.config.num_rbs = hweight32(active_rb_bitmap); } #define DEFAULT_SH_MEM_BASES (0x6000) @@ -1599,6 +1664,11 @@ static void gfx_v11_0_constants_init(struct amdgpu_device *adev) gfx_v11_0_get_tcc_info(adev); adev->gfx.config.pa_sc_tile_steering_override = 0; + /* Set whether texture coordinate truncation is conformant. */ + tmp = RREG32_SOC15(GC, 0, regTA_CNTL2); + adev->gfx.config.ta_cntl2_truncate_coord_mode = + REG_GET_FIELD(tmp, TA_CNTL2, TRUNCATE_COORD_MODE); + /* XXX SH_MEM regs */ /* where to put LDS, scratch, GPUVM in FSA64 space */ mutex_lock(&adev->srbm_mutex); @@ -4372,6 +4442,7 @@ static int gfx_v11_0_hw_fini(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; int r; + amdgpu_irq_put(adev, &adev->gfx.cp_ecc_error_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0); amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0); @@ -4589,6 +4660,14 @@ static bool gfx_v11_0_check_soft_reset(void *handle) return false; } +static int gfx_v11_0_post_soft_reset(void *handle) +{ + /** + * GFX soft reset will impact MES, need resume MES when do GFX soft reset + */ + return amdgpu_mes_resume((struct amdgpu_device *)handle); +} + static uint64_t gfx_v11_0_get_gpu_clock_counter(struct amdgpu_device *adev) { uint64_t clock; @@ -5803,6 +5882,36 @@ static void gfx_v11_0_set_compute_eop_interrupt_state(struct amdgpu_device *adev } } +#define CP_ME1_PIPE_INST_ADDR_INTERVAL 0x1 +#define SET_ECC_ME_PIPE_STATE(reg_addr, state) \ + do { \ + uint32_t tmp = RREG32_SOC15_IP(GC, reg_addr); \ + tmp = REG_SET_FIELD(tmp, CP_ME1_PIPE0_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, state); \ + WREG32_SOC15_IP(GC, reg_addr, tmp); \ + } while (0) + +static int gfx_v11_0_set_cp_ecc_error_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned type, + enum amdgpu_interrupt_state state) +{ + uint32_t ecc_irq_state = 0; + uint32_t pipe0_int_cntl_addr = 0; + int i = 0; + + ecc_irq_state = (state == AMDGPU_IRQ_STATE_ENABLE) ? 1 : 0; + + pipe0_int_cntl_addr = SOC15_REG_OFFSET(GC, 0, regCP_ME1_PIPE0_INT_CNTL); + + WREG32_FIELD15_PREREG(GC, 0, CP_INT_CNTL_RING0, CP_ECC_ERROR_INT_ENABLE, ecc_irq_state); + + for (i = 0; i < adev->gfx.mec.num_pipe_per_mec; i++) + SET_ECC_ME_PIPE_STATE(pipe0_int_cntl_addr + i * CP_ME1_PIPE_INST_ADDR_INTERVAL, + ecc_irq_state); + + return 0; +} + static int gfx_v11_0_set_eop_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type, @@ -5979,6 +6088,16 @@ static int gfx_v11_0_priv_inst_irq(struct amdgpu_device *adev, return 0; } +static int gfx_v11_0_rlc_gc_fed_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + if (adev->gfx.ras && adev->gfx.ras->rlc_gc_fed_irq) + return adev->gfx.ras->rlc_gc_fed_irq(adev, source, entry); + + return 0; +} + #if 0 static int gfx_v11_0_kiq_set_interrupt_state(struct amdgpu_device *adev, struct amdgpu_irq_src *src, @@ -6060,6 +6179,7 @@ static const struct amd_ip_funcs gfx_v11_0_ip_funcs = { .wait_for_idle = gfx_v11_0_wait_for_idle, .soft_reset = gfx_v11_0_soft_reset, .check_soft_reset = gfx_v11_0_check_soft_reset, + .post_soft_reset = gfx_v11_0_post_soft_reset, .set_clockgating_state = gfx_v11_0_set_clockgating_state, .set_powergating_state = gfx_v11_0_set_powergating_state, .get_clockgating_state = gfx_v11_0_get_clockgating_state, @@ -6209,6 +6329,15 @@ static const struct amdgpu_irq_src_funcs gfx_v11_0_priv_inst_irq_funcs = { .process = gfx_v11_0_priv_inst_irq, }; +static const struct amdgpu_irq_src_funcs gfx_v11_0_cp_ecc_error_irq_funcs = { + .set = gfx_v11_0_set_cp_ecc_error_state, + .process = amdgpu_gfx_cp_ecc_error_irq, +}; + +static const struct amdgpu_irq_src_funcs gfx_v11_0_rlc_gc_fed_irq_funcs = { + .process = gfx_v11_0_rlc_gc_fed_irq, +}; + static void gfx_v11_0_set_irq_funcs(struct amdgpu_device *adev) { adev->gfx.eop_irq.num_types = AMDGPU_CP_IRQ_LAST; @@ -6219,6 +6348,13 @@ static void gfx_v11_0_set_irq_funcs(struct amdgpu_device *adev) adev->gfx.priv_inst_irq.num_types = 1; adev->gfx.priv_inst_irq.funcs = &gfx_v11_0_priv_inst_irq_funcs; + + adev->gfx.cp_ecc_error_irq.num_types = 1; /* CP ECC error */ + adev->gfx.cp_ecc_error_irq.funcs = &gfx_v11_0_cp_ecc_error_irq_funcs; + + adev->gfx.rlc_gc_fed_irq.num_types = 1; /* 0x80 FED error */ + adev->gfx.rlc_gc_fed_irq.funcs = &gfx_v11_0_rlc_gc_fed_irq_funcs; + } static void gfx_v11_0_set_imu_funcs(struct amdgpu_device *adev) |