diff options
Diffstat (limited to 'drivers/gpu/drm/radeon')
55 files changed, 1406 insertions, 503 deletions
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index 967d193d36d..30d242b2507 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -270,8 +270,6 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode) switch (mode) { case DRM_MODE_DPMS_ON: radeon_crtc->enabled = true; - /* adjust pm to dpms changes BEFORE enabling crtcs */ - radeon_pm_compute_clocks(rdev); atombios_enable_crtc(crtc, ATOM_ENABLE); if (ASIC_IS_DCE3(rdev) && !ASIC_IS_DCE6(rdev)) atombios_enable_crtc_memreq(crtc, ATOM_ENABLE); @@ -289,10 +287,10 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode) atombios_enable_crtc_memreq(crtc, ATOM_DISABLE); atombios_enable_crtc(crtc, ATOM_DISABLE); radeon_crtc->enabled = false; - /* adjust pm to dpms changes AFTER disabling crtcs */ - radeon_pm_compute_clocks(rdev); break; } + /* adjust pm to dpms */ + radeon_pm_compute_clocks(rdev); } static void @@ -964,6 +962,9 @@ static bool atombios_crtc_prepare_pll(struct drm_crtc *crtc, struct drm_display_ struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv; int dp_clock; + + /* Assign mode clock for hdmi deep color max clock limit check */ + radeon_connector->pixelclock_for_modeset = mode->clock; radeon_crtc->bpc = radeon_get_monitor_bpc(connector); switch (encoder_mode) { @@ -1051,7 +1052,7 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode int encoder_mode = atombios_get_encoder_mode(radeon_crtc->encoder); /* pass the actual clock to atombios_crtc_program_pll for DCE5,6 for HDMI */ - if (ASIC_IS_DCE5(rdev) && !ASIC_IS_DCE8(rdev) && + if (ASIC_IS_DCE5(rdev) && (encoder_mode == ATOM_ENCODER_MODE_HDMI) && (radeon_crtc->bpc > 8)) clock = radeon_crtc->adjusted_clock; @@ -1135,6 +1136,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc, u32 fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_NONE); u32 tmp, viewport_w, viewport_h; int r; + bool bypass_lut = false; /* no fb bound */ if (!atomic && !crtc->primary->fb) { @@ -1173,33 +1175,73 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc, radeon_bo_get_tiling_flags(rbo, &tiling_flags, NULL); radeon_bo_unreserve(rbo); - switch (target_fb->bits_per_pixel) { - case 8: + switch (target_fb->pixel_format) { + case DRM_FORMAT_C8: fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_8BPP) | EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_INDEXED)); break; - case 15: + case DRM_FORMAT_XRGB4444: + case DRM_FORMAT_ARGB4444: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB4444)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); +#endif + break; + case DRM_FORMAT_XRGB1555: + case DRM_FORMAT_ARGB1555: fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB1555)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); +#endif + break; + case DRM_FORMAT_BGRX5551: + case DRM_FORMAT_BGRA5551: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA5551)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); +#endif break; - case 16: + case DRM_FORMAT_RGB565: fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB565)); #ifdef __BIG_ENDIAN fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); #endif break; - case 24: - case 32: + case DRM_FORMAT_XRGB8888: + case DRM_FORMAT_ARGB8888: fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) | EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB8888)); #ifdef __BIG_ENDIAN fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32); #endif break; + case DRM_FORMAT_XRGB2101010: + case DRM_FORMAT_ARGB2101010: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB2101010)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32); +#endif + /* Greater 8 bpc fb needs to bypass hw-lut to retain precision */ + bypass_lut = true; + break; + case DRM_FORMAT_BGRX1010102: + case DRM_FORMAT_BGRA1010102: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA1010102)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32); +#endif + /* Greater 8 bpc fb needs to bypass hw-lut to retain precision */ + bypass_lut = true; + break; default: - DRM_ERROR("Unsupported screen depth %d\n", - target_fb->bits_per_pixel); + DRM_ERROR("Unsupported screen format %s\n", + drm_get_format_name(target_fb->pixel_format)); return -EINVAL; } @@ -1208,27 +1250,43 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc, /* Set NUM_BANKS. */ if (rdev->family >= CHIP_TAHITI) { - unsigned tileb, index, num_banks, tile_split_bytes; + unsigned index, num_banks; - /* Calculate the macrotile mode index. */ - tile_split_bytes = 64 << tile_split; - tileb = 8 * 8 * target_fb->bits_per_pixel / 8; - tileb = min(tile_split_bytes, tileb); + if (rdev->family >= CHIP_BONAIRE) { + unsigned tileb, tile_split_bytes; - for (index = 0; tileb > 64; index++) { - tileb >>= 1; - } + /* Calculate the macrotile mode index. */ + tile_split_bytes = 64 << tile_split; + tileb = 8 * 8 * target_fb->bits_per_pixel / 8; + tileb = min(tile_split_bytes, tileb); - if (index >= 16) { - DRM_ERROR("Wrong screen bpp (%u) or tile split (%u)\n", - target_fb->bits_per_pixel, tile_split); - return -EINVAL; - } + for (index = 0; tileb > 64; index++) + tileb >>= 1; + + if (index >= 16) { + DRM_ERROR("Wrong screen bpp (%u) or tile split (%u)\n", + target_fb->bits_per_pixel, tile_split); + return -EINVAL; + } - if (rdev->family >= CHIP_BONAIRE) num_banks = (rdev->config.cik.macrotile_mode_array[index] >> 6) & 0x3; - else + } else { + switch (target_fb->bits_per_pixel) { + case 8: + index = 10; + break; + case 16: + index = SI_TILE_MODE_COLOR_2D_SCANOUT_16BPP; + break; + default: + case 32: + index = SI_TILE_MODE_COLOR_2D_SCANOUT_32BPP; + break; + } + num_banks = (rdev->config.si.tile_mode_array[index] >> 20) & 0x3; + } + fb_format |= EVERGREEN_GRPH_NUM_BANKS(num_banks); } else { /* NI and older. */ @@ -1312,6 +1370,18 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc, WREG32(EVERGREEN_GRPH_CONTROL + radeon_crtc->crtc_offset, fb_format); WREG32(EVERGREEN_GRPH_SWAP_CONTROL + radeon_crtc->crtc_offset, fb_swap); + /* + * The LUT only has 256 slots for indexing by a 8 bpc fb. Bypass the LUT + * for > 8 bpc scanout to avoid truncation of fb indices to 8 msb's, to + * retain the full precision throughout the pipeline. + */ + WREG32_P(EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL + radeon_crtc->crtc_offset, + (bypass_lut ? EVERGREEN_LUT_10BIT_BYPASS_EN : 0), + ~EVERGREEN_LUT_10BIT_BYPASS_EN); + + if (bypass_lut) + DRM_DEBUG_KMS("Bypassing hardware LUT due to 10 bit fb scanout.\n"); + WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_X + radeon_crtc->crtc_offset, 0); WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_Y + radeon_crtc->crtc_offset, 0); WREG32(EVERGREEN_GRPH_X_START + radeon_crtc->crtc_offset, 0); @@ -1344,8 +1414,8 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc, tmp &= ~EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN; WREG32(EVERGREEN_GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, tmp); - /* set pageflip to happen anywhere in vblank interval */ - WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0); + /* set pageflip to happen only at start of vblank interval (front porch) */ + WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3); if (!atomic && fb && fb != crtc->primary->fb) { radeon_fb = to_radeon_framebuffer(fb); @@ -1379,6 +1449,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, u32 fb_swap = R600_D1GRPH_SWAP_ENDIAN_NONE; u32 tmp, viewport_w, viewport_h; int r; + bool bypass_lut = false; /* no fb bound */ if (!atomic && !crtc->primary->fb) { @@ -1416,18 +1487,30 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, radeon_bo_get_tiling_flags(rbo, &tiling_flags, NULL); radeon_bo_unreserve(rbo); - switch (target_fb->bits_per_pixel) { - case 8: + switch (target_fb->pixel_format) { + case DRM_FORMAT_C8: fb_format = AVIVO_D1GRPH_CONTROL_DEPTH_8BPP | AVIVO_D1GRPH_CONTROL_8BPP_INDEXED; break; - case 15: + case DRM_FORMAT_XRGB4444: + case DRM_FORMAT_ARGB4444: + fb_format = + AVIVO_D1GRPH_CONTROL_DEPTH_16BPP | + AVIVO_D1GRPH_CONTROL_16BPP_ARGB4444; +#ifdef __BIG_ENDIAN + fb_swap = R600_D1GRPH_SWAP_ENDIAN_16BIT; +#endif + break; + case DRM_FORMAT_XRGB1555: fb_format = AVIVO_D1GRPH_CONTROL_DEPTH_16BPP | AVIVO_D1GRPH_CONTROL_16BPP_ARGB1555; +#ifdef __BIG_ENDIAN + fb_swap = R600_D1GRPH_SWAP_ENDIAN_16BIT; +#endif break; - case 16: + case DRM_FORMAT_RGB565: fb_format = AVIVO_D1GRPH_CONTROL_DEPTH_16BPP | AVIVO_D1GRPH_CONTROL_16BPP_RGB565; @@ -1435,8 +1518,8 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, fb_swap = R600_D1GRPH_SWAP_ENDIAN_16BIT; #endif break; - case 24: - case 32: + case DRM_FORMAT_XRGB8888: + case DRM_FORMAT_ARGB8888: fb_format = AVIVO_D1GRPH_CONTROL_DEPTH_32BPP | AVIVO_D1GRPH_CONTROL_32BPP_ARGB8888; @@ -1444,9 +1527,20 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, fb_swap = R600_D1GRPH_SWAP_ENDIAN_32BIT; #endif break; + case DRM_FORMAT_XRGB2101010: + case DRM_FORMAT_ARGB2101010: + fb_format = + AVIVO_D1GRPH_CONTROL_DEPTH_32BPP | + AVIVO_D1GRPH_CONTROL_32BPP_ARGB2101010; +#ifdef __BIG_ENDIAN + fb_swap = R600_D1GRPH_SWAP_ENDIAN_32BIT; +#endif + /* Greater 8 bpc fb needs to bypass hw-lut to retain precision */ + bypass_lut = true; + break; default: - DRM_ERROR("Unsupported screen depth %d\n", - target_fb->bits_per_pixel); + DRM_ERROR("Unsupported screen format %s\n", + drm_get_format_name(target_fb->pixel_format)); return -EINVAL; } @@ -1485,6 +1579,13 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, if (rdev->family >= CHIP_R600) WREG32(R600_D1GRPH_SWAP_CONTROL + radeon_crtc->crtc_offset, fb_swap); + /* LUT only has 256 slots for 8 bpc fb. Bypass for > 8 bpc scanout for precision */ + WREG32_P(AVIVO_D1GRPH_LUT_SEL + radeon_crtc->crtc_offset, + (bypass_lut ? AVIVO_LUT_10BIT_BYPASS_EN : 0), ~AVIVO_LUT_10BIT_BYPASS_EN); + + if (bypass_lut) + DRM_DEBUG_KMS("Bypassing hardware LUT due to 10 bit fb scanout.\n"); + WREG32(AVIVO_D1GRPH_SURFACE_OFFSET_X + radeon_crtc->crtc_offset, 0); WREG32(AVIVO_D1GRPH_SURFACE_OFFSET_Y + radeon_crtc->crtc_offset, 0); WREG32(AVIVO_D1GRPH_X_START + radeon_crtc->crtc_offset, 0); @@ -1513,8 +1614,8 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, tmp &= ~AVIVO_D1GRPH_SURFACE_UPDATE_H_RETRACE_EN; WREG32(AVIVO_D1GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, tmp); - /* set pageflip to happen anywhere in vblank interval */ - WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0); + /* set pageflip to happen only at start of vblank interval (front porch) */ + WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3); if (!atomic && fb && fb != crtc->primary->fb) { radeon_fb = to_radeon_framebuffer(fb); @@ -1751,8 +1852,9 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc) } /* otherwise, pick one of the plls */ if ((rdev->family == CHIP_KAVERI) || - (rdev->family == CHIP_KABINI)) { - /* KB/KV has PPLL1 and PPLL2 */ + (rdev->family == CHIP_KABINI) || + (rdev->family == CHIP_MULLINS)) { + /* KB/KV/ML has PPLL1 and PPLL2 */ pll_in_use = radeon_get_pll_use_mask(crtc); if (!(pll_in_use & (1 << ATOM_PPLL2))) return ATOM_PPLL2; @@ -1916,6 +2018,9 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc, (ATOM_DEVICE_TV_SUPPORT | ATOM_DEVICE_CV_SUPPORT)) is_tvcv = true; + if (!radeon_crtc->adjusted_clock) + return -EINVAL; + atombios_crtc_set_pll(crtc, adjusted_mode); if (ASIC_IS_DCE4(rdev)) diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index 225f6c66eff..b1e11f8434e 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -127,7 +127,7 @@ static int radeon_process_aux_ch(struct radeon_i2c_chan *chan, /* flags not zero */ if (args.v1.ucReplyStatus == 2) { DRM_DEBUG_KMS("dp_aux_ch flags not zero\n"); - r = -EBUSY; + r = -EIO; goto done; } @@ -222,11 +222,12 @@ void radeon_dp_aux_init(struct radeon_connector *radeon_connector) radeon_connector->ddc_bus->rec.hpd = radeon_connector->hpd.hpd; radeon_connector->ddc_bus->aux.dev = radeon_connector->base.kdev; radeon_connector->ddc_bus->aux.transfer = radeon_dp_aux_transfer; - ret = drm_dp_aux_register_i2c_bus(&radeon_connector->ddc_bus->aux); + + ret = drm_dp_aux_register(&radeon_connector->ddc_bus->aux); if (!ret) radeon_connector->ddc_bus->has_aux = true; - WARN(ret, "drm_dp_aux_register_i2c_bus() failed with error %d\n", ret); + WARN(ret, "drm_dp_aux_register() failed with error %d\n", ret); } /***** general DP utility functions *****/ @@ -389,11 +390,11 @@ static void radeon_dp_probe_oui(struct radeon_connector *radeon_connector) if (!(dig_connector->dpcd[DP_DOWN_STREAM_PORT_COUNT] & DP_OUI_SUPPORT)) return; - if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_SINK_OUI, buf, 3)) + if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_SINK_OUI, buf, 3) == 3) DRM_DEBUG_KMS("Sink OUI: %02hx%02hx%02hx\n", buf[0], buf[1], buf[2]); - if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_BRANCH_OUI, buf, 3)) + if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_BRANCH_OUI, buf, 3) == 3) DRM_DEBUG_KMS("Branch OUI: %02hx%02hx%02hx\n", buf[0], buf[1], buf[2]); } @@ -402,16 +403,18 @@ bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector) { struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv; u8 msg[DP_DPCD_SIZE]; - int ret, i; + int ret; + + char dpcd_hex_dump[DP_DPCD_SIZE * 3]; ret = drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_DPCD_REV, msg, DP_DPCD_SIZE); if (ret > 0) { memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE); - DRM_DEBUG_KMS("DPCD: "); - for (i = 0; i < DP_DPCD_SIZE; i++) - DRM_DEBUG_KMS("%02x ", msg[i]); - DRM_DEBUG_KMS("\n"); + + hex_dump_to_buffer(dig_connector->dpcd, sizeof(dig_connector->dpcd), + 32, 1, dpcd_hex_dump, sizeof(dpcd_hex_dump), false); + DRM_DEBUG_KMS("DPCD: %s\n", dpcd_hex_dump); radeon_dp_probe_oui(radeon_connector); @@ -442,21 +445,23 @@ int radeon_dp_get_panel_mode(struct drm_encoder *encoder, if (dp_bridge != ENCODER_OBJECT_ID_NONE) { /* DP bridge chips */ - drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, - DP_EDP_CONFIGURATION_CAP, &tmp); - if (tmp & 1) - panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE; - else if ((dp_bridge == ENCODER_OBJECT_ID_NUTMEG) || - (dp_bridge == ENCODER_OBJECT_ID_TRAVIS)) - panel_mode = DP_PANEL_MODE_INTERNAL_DP1_MODE; - else - panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE; + if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, + DP_EDP_CONFIGURATION_CAP, &tmp) == 1) { + if (tmp & 1) + panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE; + else if ((dp_bridge == ENCODER_OBJECT_ID_NUTMEG) || + (dp_bridge == ENCODER_OBJECT_ID_TRAVIS)) + panel_mode = DP_PANEL_MODE_INTERNAL_DP1_MODE; + else + panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE; + } } else if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) { /* eDP */ - drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, - DP_EDP_CONFIGURATION_CAP, &tmp); - if (tmp & 1) - panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE; + if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, + DP_EDP_CONFIGURATION_CAP, &tmp) == 1) { + if (tmp & 1) + panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE; + } } return panel_mode; @@ -832,11 +837,15 @@ void radeon_dp_link_train(struct drm_encoder *encoder, else dp_info.enc_id |= ATOM_DP_CONFIG_LINK_A; - drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, DP_MAX_LANE_COUNT, &tmp); - if (ASIC_IS_DCE5(rdev) && (tmp & DP_TPS3_SUPPORTED)) - dp_info.tp3_supported = true; - else + if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, DP_MAX_LANE_COUNT, &tmp) + == 1) { + if (ASIC_IS_DCE5(rdev) && (tmp & DP_TPS3_SUPPORTED)) + dp_info.tp3_supported = true; + else + dp_info.tp3_supported = false; + } else { dp_info.tp3_supported = false; + } memcpy(dp_info.dpcd, dig_connector->dpcd, DP_RECEIVER_CAP_SIZE); dp_info.rdev = rdev; diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index 2b290844064..7d68203a373 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -183,7 +183,6 @@ void radeon_atom_backlight_init(struct radeon_encoder *radeon_encoder, struct backlight_properties props; struct radeon_backlight_privdata *pdata; struct radeon_encoder_atom_dig *dig; - u8 backlight_level; char bl_name[16]; /* Mac laptops with multiple GPUs use the gmux driver for backlight @@ -222,12 +221,17 @@ void radeon_atom_backlight_init(struct radeon_encoder *radeon_encoder, pdata->encoder = radeon_encoder; - backlight_level = radeon_atom_get_backlight_level_from_reg(rdev); - dig = radeon_encoder->enc_priv; dig->bl_dev = bd; bd->props.brightness = radeon_atom_backlight_get_brightness(bd); + /* Set a reasonable default here if the level is 0 otherwise + * fbdev will attempt to turn the backlight on after console + * unblanking and it will try and restore 0 which turns the backlight + * off again. + */ + if (bd->props.brightness == 0) + bd->props.brightness = RADEON_MAX_BL_LEVEL; bd->props.power = FB_BLANK_UNBLANK; backlight_update_status(bd); diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c index 10dae4106c0..584090ac3eb 100644 --- a/drivers/gpu/drm/radeon/ci_dpm.c +++ b/drivers/gpu/drm/radeon/ci_dpm.c @@ -1179,7 +1179,7 @@ static int ci_stop_dpm(struct radeon_device *rdev) tmp &= ~GLOBAL_PWRMGT_EN; WREG32_SMC(GENERAL_PWRMGT, tmp); - tmp = RREG32(SCLK_PWRMGT_CNTL); + tmp = RREG32_SMC(SCLK_PWRMGT_CNTL); tmp &= ~DYNAMIC_PM_EN; WREG32_SMC(SCLK_PWRMGT_CNTL, tmp); diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c index a5181404f13..c0ea66192fe 100644 --- a/drivers/gpu/drm/radeon/cik.c +++ b/drivers/gpu/drm/radeon/cik.c @@ -63,6 +63,12 @@ MODULE_FIRMWARE("radeon/KABINI_ce.bin"); MODULE_FIRMWARE("radeon/KABINI_mec.bin"); MODULE_FIRMWARE("radeon/KABINI_rlc.bin"); MODULE_FIRMWARE("radeon/KABINI_sdma.bin"); +MODULE_FIRMWARE("radeon/MULLINS_pfp.bin"); +MODULE_FIRMWARE("radeon/MULLINS_me.bin"); +MODULE_FIRMWARE("radeon/MULLINS_ce.bin"); +MODULE_FIRMWARE("radeon/MULLINS_mec.bin"); +MODULE_FIRMWARE("radeon/MULLINS_rlc.bin"); +MODULE_FIRMWARE("radeon/MULLINS_sdma.bin"); extern int r600_ih_ring_alloc(struct radeon_device *rdev); extern void r600_ih_ring_fini(struct radeon_device *rdev); @@ -74,6 +80,7 @@ extern int sumo_rlc_init(struct radeon_device *rdev); extern void si_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc); extern void si_rlc_reset(struct radeon_device *rdev); extern void si_init_uvd_internal_cg(struct radeon_device *rdev); +static u32 cik_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh); extern int cik_sdma_resume(struct radeon_device *rdev); extern void cik_sdma_enable(struct radeon_device *rdev, bool enable); extern void cik_sdma_fini(struct radeon_device *rdev); @@ -1473,6 +1480,43 @@ static const u32 hawaii_mgcg_cgcg_init[] = 0xd80c, 0xff000ff0, 0x00000100 }; +static const u32 godavari_golden_registers[] = +{ + 0x55e4, 0xff607fff, 0xfc000100, + 0x6ed8, 0x00010101, 0x00010000, + 0x9830, 0xffffffff, 0x00000000, + 0x98302, 0xf00fffff, 0x00000400, + 0x6130, 0xffffffff, 0x00010000, + 0x5bb0, 0x000000f0, 0x00000070, + 0x5bc0, 0xf0311fff, 0x80300000, + 0x98f8, 0x73773777, 0x12010001, + 0x98fc, 0xffffffff, 0x00000010, + 0x8030, 0x00001f0f, 0x0000100a, + 0x2f48, 0x73773777, 0x12010001, + 0x2408, 0x000fffff, 0x000c007f, + 0x8a14, 0xf000003f, 0x00000007, + 0x8b24, 0xffffffff, 0x00ff0fff, + 0x30a04, 0x0000ff0f, 0x00000000, + 0x28a4c, 0x07ffffff, 0x06000000, + 0x4d8, 0x00000fff, 0x00000100, + 0xd014, 0x00010000, 0x00810001, + 0xd814, 0x00010000, 0x00810001, + 0x3e78, 0x00000001, 0x00000002, + 0xc768, 0x00000008, 0x00000008, + 0xc770, 0x00000f00, 0x00000800, + 0xc774, 0x00000f00, 0x00000800, + 0xc798, 0x00ffffff, 0x00ff7fbf, + 0xc79c, 0x00ffffff, 0x00ff7faf, + 0x8c00, 0x000000ff, 0x00000001, + 0x214f8, 0x01ff01ff, 0x00000002, + 0x21498, 0x007ff800, 0x00200000, + 0x2015c, 0xffffffff, 0x00000f40, + 0x88c4, 0x001f3ae3, 0x00000082, + 0x88d4, 0x0000001f, 0x00000010, + 0x30934, 0xffffffff, 0x00000000 +}; + + static void cik_init_golden_registers(struct radeon_device *rdev) { switch (rdev->family) { @@ -1504,6 +1548,20 @@ static void cik_init_golden_registers(struct radeon_device *rdev) kalindi_golden_spm_registers, (const u32)ARRAY_SIZE(kalindi_golden_spm_registers)); break; + case CHIP_MULLINS: + radeon_program_register_sequence(rdev, + kalindi_mgcg_cgcg_init, + (const u32)ARRAY_SIZE(kalindi_mgcg_cgcg_init)); + radeon_program_register_sequence(rdev, + godavari_golden_registers, + (const u32)ARRAY_SIZE(godavari_golden_registers)); + radeon_program_register_sequence(rdev, + kalindi_golden_common_registers, + (const u32)ARRAY_SIZE(kalindi_golden_common_registers)); + radeon_program_register_sequence(rdev, + kalindi_golden_spm_registers, + (const u32)ARRAY_SIZE(kalindi_golden_spm_registers)); + break; case CHIP_KAVERI: radeon_program_register_sequence(rdev, spectre_mgcg_cgcg_init, @@ -1834,6 +1892,15 @@ static int cik_init_microcode(struct radeon_device *rdev) rlc_req_size = KB_RLC_UCODE_SIZE * 4; sdma_req_size = CIK_SDMA_UCODE_SIZE * 4; break; + case CHIP_MULLINS: + chip_name = "MULLINS"; + pfp_req_size = CIK_PFP_UCODE_SIZE * 4; + me_req_size = CIK_ME_UCODE_SIZE * 4; + ce_req_size = CIK_CE_UCODE_SIZE * 4; + mec_req_size = CIK_MEC_UCODE_SIZE * 4; + rlc_req_size = ML_RLC_UCODE_SIZE * 4; + sdma_req_size = CIK_SDMA_UCODE_SIZE * 4; + break; default: BUG(); } @@ -2224,6 +2291,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev) gb_tile_moden = 0; break; } + rdev->config.cik.macrotile_mode_array[reg_offset] = gb_tile_moden; WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), gb_tile_moden); } } else if (num_pipe_configs == 8) { @@ -3191,7 +3259,7 @@ static void cik_gpu_init(struct radeon_device *rdev) u32 mc_shared_chmap, mc_arb_ramcfg; u32 hdp_host_path_cntl; u32 tmp; - int i, j; + int i, j, k; switch (rdev->family) { case CHIP_BONAIRE: @@ -3272,6 +3340,7 @@ static void cik_gpu_init(struct radeon_device *rdev) gb_addr_config = BONAIRE_GB_ADDR_CONFIG_GOLDEN; break; case CHIP_KABINI: + case CHIP_MULLINS: default: rdev->config.cik.max_shader_engines = 1; rdev->config.cik.max_tile_pipes = 2; @@ -3379,6 +3448,15 @@ static void cik_gpu_init(struct radeon_device *rdev) rdev->config.cik.max_sh_per_se, rdev->config.cik.max_backends_per_se); + for (i = 0; i < rdev->config.cik.max_shader_engines; i++) { + for (j = 0; j < rdev->config.cik.max_sh_per_se; j++) { + for (k = 0; k < rdev->config.cik.max_cu_per_sh; k++) { + rdev->config.cik.active_cus += + hweight32(cik_get_cu_active_bitmap(rdev, i, j)); + } + } + } + /* set HW defaults for 3D engine */ WREG32(CP_MEQ_THRESHOLDS, MEQ1_START(0x30) | MEQ2_START(0x60)); @@ -3631,7 +3709,7 @@ bool cik_semaphore_ring_emit(struct radeon_device *rdev, unsigned sel = emit_wait ? PACKET3_SEM_SEL_WAIT : PACKET3_SEM_SEL_SIGNAL; radeon_ring_write(ring, PACKET3(PACKET3_MEM_SEMAPHORE, 1)); - radeon_ring_write(ring, addr & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(addr)); radeon_ring_write(ring, (upper_32_bits(addr) & 0xffff) | sel); return true; @@ -3702,6 +3780,7 @@ int cik_copy_cpdma(struct radeon_device *rdev, r = radeon_fence_emit(rdev, fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); + radeon_semaphore_free(rdev, &sem, NULL); return r; } @@ -3750,7 +3829,7 @@ void cik_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib) radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); radeon_ring_write(ring, WRITE_DATA_DST_SEL(1)); radeon_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); + radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr)); radeon_ring_write(ring, next_rptr); } @@ -5378,7 +5457,7 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev) (u32)(rdev->dummy_page.addr >> 12)); WREG32(VM_CONTEXT1_CNTL2, 4); WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) | - PAGE_TABLE_BLOCK_SIZE(RADEON_VM_BLOCK_SIZE - 9) | + PAGE_TABLE_BLOCK_SIZE(radeon_vm_block_size - 9) | RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT | RANGE_PROTECTION_FAULT_ENABLE_DEFAULT | DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT | @@ -5803,6 +5882,9 @@ static int cik_rlc_resume(struct radeon_device *rdev) case CHIP_KABINI: size = KB_RLC_UCODE_SIZE; break; + case CHIP_MULLINS: + size = ML_RLC_UCODE_SIZE; + break; } cik_rlc_stop(rdev); @@ -6551,6 +6633,7 @@ void cik_get_csb_buffer(struct radeon_device *rdev, volatile u32 *buffer) buffer[count++] = cpu_to_le32(0x00000000); break; case CHIP_KABINI: + case CHIP_MULLINS: buffer[count++] = cpu_to_le32(0x00000000); /* XXX */ buffer[count++] = cpu_to_le32(0x00000000); break; @@ -6696,6 +6779,19 @@ static void cik_disable_interrupt_state(struct radeon_device *rdev) WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, 0); WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, 0); } + /* pflip */ + if (rdev->num_crtc >= 2) { + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, 0); + } + if (rdev->num_crtc >= 4) { + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, 0); + } + if (rdev->num_crtc >= 6) { + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, 0); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0); + } /* dac hotplug */ WREG32(DAC_AUTODETECT_INT_CONTROL, 0); @@ -7052,6 +7148,25 @@ int cik_irq_set(struct radeon_device *rdev) WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6); } + if (rdev->num_crtc >= 2) { + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + } + if (rdev->num_crtc >= 4) { + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + } + if (rdev->num_crtc >= 6) { + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + } + WREG32(DC_HPD1_INT_CONTROL, hpd1); WREG32(DC_HPD2_INT_CONTROL, hpd2); WREG32(DC_HPD3_INT_CONTROL, hpd3); @@ -7088,6 +7203,29 @@ static inline void cik_irq_ack(struct radeon_device *rdev) rdev->irq.stat_regs.cik.disp_int_cont5 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE5); rdev->irq.stat_regs.cik.disp_int_cont6 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE6); + rdev->irq.stat_regs.cik.d1grph_int = RREG32(GRPH_INT_STATUS + + EVERGREEN_CRTC0_REGISTER_OFFSET); + rdev->irq.stat_regs.cik.d2grph_int = RREG32(GRPH_INT_STATUS + + EVERGREEN_CRTC1_REGISTER_OFFSET); + if (rdev->num_crtc >= 4) { + rdev->irq.stat_regs.cik.d3grph_int = RREG32(GRPH_INT_STATUS + + EVERGREEN_CRTC2_REGISTER_OFFSET); + rdev->irq.stat_regs.cik.d4grph_int = RREG32(GRPH_INT_STATUS + + EVERGREEN_CRTC3_REGISTER_OFFSET); + } + if (rdev->num_crtc >= 6) { + rdev->irq.stat_regs.cik.d5grph_int = RREG32(GRPH_INT_STATUS + + EVERGREEN_CRTC4_REGISTER_OFFSET); + rdev->irq.stat_regs.cik.d6grph_int = RREG32(GRPH_INT_STATUS + + EVERGREEN_CRTC5_REGISTER_OFFSET); + } + + if (rdev->irq.stat_regs.cik.d1grph_int & GRPH_PFLIP_INT_OCCURRED) + WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, + GRPH_PFLIP_INT_CLEAR); + if (rdev->irq.stat_regs.cik.d2grph_int & GRPH_PFLIP_INT_OCCURRED) + WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, + GRPH_PFLIP_INT_CLEAR); if (rdev->irq.stat_regs.cik.disp_int & LB_D1_VBLANK_INTERRUPT) WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VBLANK_ACK); if (rdev->irq.stat_regs.cik.disp_int & LB_D1_VLINE_INTERRUPT) @@ -7098,6 +7236,12 @@ static inline void cik_irq_ack(struct radeon_device *rdev) WREG32(LB_VLINE_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VLINE_ACK); if (rdev->num_crtc >= 4) { + if (rdev->irq.stat_regs.cik.d3grph_int & GRPH_PFLIP_INT_OCCURRED) + WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, + GRPH_PFLIP_INT_CLEAR); + if (rdev->irq.stat_regs.cik.d4grph_int & GRPH_PFLIP_INT_OCCURRED) + WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET, + GRPH_PFLIP_INT_CLEAR); if (rdev->irq.stat_regs.cik.disp_int_cont2 & LB_D3_VBLANK_INTERRUPT) WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VBLANK_ACK); if (rdev->irq.stat_regs.cik.disp_int_cont2 & LB_D3_VLINE_INTERRUPT) @@ -7109,6 +7253,12 @@ static inline void cik_irq_ack(struct radeon_device *rdev) } if (rdev->num_crtc >= 6) { + if (rdev->irq.stat_regs.cik.d5grph_int & GRPH_PFLIP_INT_OCCURRED) + WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, + GRPH_PFLIP_INT_CLEAR); + if (rdev->irq.stat_regs.cik.d6grph_int & GRPH_PFLIP_INT_OCCURRED) + WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET, + GRPH_PFLIP_INT_CLEAR); if (rdev->irq.stat_regs.cik.disp_int_cont4 & LB_D5_VBLANK_INTERRUPT) WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VBLANK_ACK); if (rdev->irq.stat_regs.cik.disp_int_cont4 & LB_D5_VLINE_INTERRUPT) @@ -7227,6 +7377,7 @@ static inline u32 cik_get_ih_wptr(struct radeon_device *rdev) tmp = RREG32(IH_RB_CNTL); tmp |= IH_WPTR_OVERFLOW_CLEAR; WREG32(IH_RB_CNTL, tmp); + wptr &= ~RB_OVERFLOW; } return (wptr & rdev->ih.ptr_mask); } @@ -7460,6 +7611,15 @@ restart_ih: break; } break; + case 8: /* D1 page flip */ + case 10: /* D2 page flip */ + case 12: /* D3 page flip */ + case 14: /* D4 page flip */ + case 16: /* D5 page flip */ + case 18: /* D6 page flip */ + DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1); + radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1); + break; case 42: /* HPD hotplug */ switch (src_data) { case 0: @@ -7518,14 +7678,16 @@ restart_ih: addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR); status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS); mc_client = RREG32(VM_CONTEXT1_PROTECTION_FAULT_MCCLIENT); + /* reset addr and status */ + WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1); + if (addr == 0x0 && status == 0x0) + break; dev_err(rdev->dev, "GPU fault detected: %d 0x%08x\n", src_id, src_data); dev_err(rdev->dev, " VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x%08X\n", addr); dev_err(rdev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n", status); cik_vm_decode_fault(rdev, status, addr, mc_client); - /* reset addr and status */ - WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1); break; case 167: /* VCE */ DRM_DEBUG("IH: VCE int: 0x%08x\n", src_data); diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c index 3c2407bad1f..8e9d0f1d858 100644 --- a/drivers/gpu/drm/radeon/cik_sdma.c +++ b/drivers/gpu/drm/radeon/cik_sdma.c @@ -141,7 +141,7 @@ void cik_sdma_ring_ib_execute(struct radeon_device *rdev, next_rptr += 4; radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0)); radeon_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); + radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr)); radeon_ring_write(ring, 1); /* number of DWs to follow */ radeon_ring_write(ring, next_rptr); } @@ -151,7 +151,7 @@ void cik_sdma_ring_ib_execute(struct radeon_device *rdev, radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_NOP, 0, 0)); radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_INDIRECT_BUFFER, 0, extra_bits)); radeon_ring_write(ring, ib->gpu_addr & 0xffffffe0); /* base must be 32 byte aligned */ - radeon_ring_write(ring, upper_32_bits(ib->gpu_addr) & 0xffffffff); + radeon_ring_write(ring, upper_32_bits(ib->gpu_addr)); radeon_ring_write(ring, ib->length_dw); } @@ -203,8 +203,8 @@ void cik_sdma_fence_ring_emit(struct radeon_device *rdev, /* write the fence */ radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_FENCE, 0, 0)); - radeon_ring_write(ring, addr & 0xffffffff); - radeon_ring_write(ring, upper_32_bits(addr) & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(addr)); + radeon_ring_write(ring, upper_32_bits(addr)); radeon_ring_write(ring, fence->seq); /* generate an interrupt */ radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_TRAP, 0, 0)); @@ -233,7 +233,7 @@ bool cik_sdma_semaphore_ring_emit(struct radeon_device *rdev, radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SEMAPHORE, 0, extra_bits)); radeon_ring_write(ring, addr & 0xfffffff8); - radeon_ring_write(ring, upper_32_bits(addr) & 0xffffffff); + radeon_ring_write(ring, upper_32_bits(addr)); return true; } @@ -551,10 +551,10 @@ int cik_copy_dma(struct radeon_device *rdev, radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_COPY, SDMA_COPY_SUB_OPCODE_LINEAR, 0)); radeon_ring_write(ring, cur_size_in_bytes); radeon_ring_write(ring, 0); /* src/dst endian swap */ - radeon_ring_write(ring, src_offset & 0xffffffff); - radeon_ring_write(ring, upper_32_bits(src_offset) & 0xffffffff); - radeon_ring_write(ring, dst_offset & 0xffffffff); - radeon_ring_write(ring, upper_32_bits(dst_offset) & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(src_offset)); + radeon_ring_write(ring, upper_32_bits(src_offset)); + radeon_ring_write(ring, lower_32_bits(dst_offset)); + radeon_ring_write(ring, upper_32_bits(dst_offset)); src_offset += cur_size_in_bytes; dst_offset += cur_size_in_bytes; } @@ -562,6 +562,7 @@ int cik_copy_dma(struct radeon_device *rdev, r = radeon_fence_emit(rdev, fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); + radeon_semaphore_free(rdev, &sem, NULL); return r; } @@ -604,7 +605,7 @@ int cik_sdma_ring_test(struct radeon_device *rdev, } radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0)); radeon_ring_write(ring, rdev->vram_scratch.gpu_addr & 0xfffffffc); - radeon_ring_write(ring, upper_32_bits(rdev->vram_scratch.gpu_addr) & 0xffffffff); + radeon_ring_write(ring, upper_32_bits(rdev->vram_scratch.gpu_addr)); radeon_ring_write(ring, 1); /* number of DWs to follow */ radeon_ring_write(ring, 0xDEADBEEF); radeon_ring_unlock_commit(rdev, ring); @@ -659,7 +660,7 @@ int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring) ib.ptr[0] = SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0); ib.ptr[1] = rdev->vram_scratch.gpu_addr & 0xfffffffc; - ib.ptr[2] = upper_32_bits(rdev->vram_scratch.gpu_addr) & 0xffffffff; + ib.ptr[2] = upper_32_bits(rdev->vram_scratch.gpu_addr); ib.ptr[3] = 1; ib.ptr[4] = 0xDEADBEEF; ib.length_dw = 5; @@ -751,9 +752,9 @@ void cik_sdma_vm_set_page(struct radeon_device *rdev, ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_COPY, SDMA_WRITE_SUB_OPCODE_LINEAR, 0); ib->ptr[ib->length_dw++] = bytes; ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */ - ib->ptr[ib->length_dw++] = src & 0xffffffff; + ib->ptr[ib->length_dw++] = lower_32_bits(src); ib->ptr[ib->length_dw++] = upper_32_bits(src); - ib->ptr[ib->length_dw++] = pe & 0xffffffff; + ib->ptr[ib->length_dw++] = lower_32_bits(pe); ib->ptr[ib->length_dw++] = upper_32_bits(pe); pe += bytes; diff --git a/drivers/gpu/drm/radeon/cikd.h b/drivers/gpu/drm/radeon/cikd.h index 0b27ea08c29..0c6e1b55d96 100644 --- a/drivers/gpu/drm/radeon/cikd.h +++ b/drivers/gpu/drm/radeon/cikd.h @@ -889,6 +889,15 @@ # define DC_HPD6_RX_INTERRUPT (1 << 18) #define DISP_INTERRUPT_STATUS_CONTINUE6 0x6780 +/* 0x6858, 0x7458, 0x10058, 0x10c58, 0x11858, 0x12458 */ +#define GRPH_INT_STATUS 0x6858 +# define GRPH_PFLIP_INT_OCCURRED (1 << 0) +# define GRPH_PFLIP_INT_CLEAR (1 << 8) +/* 0x685c, 0x745c, 0x1005c, 0x10c5c, 0x1185c, 0x1245c */ +#define GRPH_INT_CONTROL 0x685c +# define GRPH_PFLIP_INT_MASK (1 << 0) +# define GRPH_PFLIP_INT_TYPE (1 << 8) + #define DAC_AUTODETECT_INT_CONTROL 0x67c8 #define DC_HPD1_INT_STATUS 0x601c @@ -1743,12 +1752,12 @@ #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 EOP_TCL2_VOLATILE (1 << 24) #define EOP_CACHE_POLICY(x) ((x) << 25) /* 0 - LRU * 1 - Stream * 2 - Bypass */ -#define EOP_TCL2_VOLATILE (1 << 27) #define DATA_SEL(x) ((x) << 29) /* 0 - discard * 1 - send low 32bit data diff --git a/drivers/gpu/drm/radeon/cypress_dpm.c b/drivers/gpu/drm/radeon/cypress_dpm.c index 5a9a5f4d788..47d31e91575 100644 --- a/drivers/gpu/drm/radeon/cypress_dpm.c +++ b/drivers/gpu/drm/radeon/cypress_dpm.c @@ -1551,7 +1551,7 @@ int cypress_populate_smc_voltage_tables(struct radeon_device *rdev, table->voltageMaskTable.highMask[RV770_SMC_VOLTAGEMASK_VDDCI] = 0; table->voltageMaskTable.lowMask[RV770_SMC_VOLTAGEMASK_VDDCI] = - cpu_to_be32(eg_pi->vddc_voltage_table.mask_low); + cpu_to_be32(eg_pi->vddci_voltage_table.mask_low); } return 0; diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 0318230ef27..15e4f28015e 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -189,7 +189,7 @@ static const u32 evergreen_golden_registers[] = 0x8c1c, 0xffffffff, 0x00001010, 0x28350, 0xffffffff, 0x00000000, 0xa008, 0xffffffff, 0x00010000, - 0x5cc, 0xffffffff, 0x00000001, + 0x5c4, 0xffffffff, 0x00000001, 0x9508, 0xffffffff, 0x00000002, 0x913c, 0x0000000f, 0x0000000a }; @@ -476,7 +476,7 @@ static const u32 cedar_golden_registers[] = 0x8c1c, 0xffffffff, 0x00001010, 0x28350, 0xffffffff, 0x00000000, 0xa008, 0xffffffff, 0x00010000, - 0x5cc, 0xffffffff, 0x00000001, + 0x5c4, 0xffffffff, 0x00000001, 0x9508, 0xffffffff, 0x00000002 }; @@ -635,7 +635,7 @@ static const u32 juniper_mgcg_init[] = static const u32 supersumo_golden_registers[] = { 0x5eb4, 0xffffffff, 0x00000002, - 0x5cc, 0xffffffff, 0x00000001, + 0x5c4, 0xffffffff, 0x00000001, 0x7030, 0xffffffff, 0x00000011, 0x7c30, 0xffffffff, 0x00000011, 0x6104, 0x01000300, 0x00000000, @@ -719,7 +719,7 @@ static const u32 sumo_golden_registers[] = static const u32 wrestler_golden_registers[] = { 0x5eb4, 0xffffffff, 0x00000002, - 0x5cc, 0xffffffff, 0x00000001, + 0x5c4, 0xffffffff, 0x00000001, 0x7030, 0xffffffff, 0x00000011, 0x7c30, 0xffffffff, 0x00000011, 0x6104, 0x01000300, 0x00000000, @@ -2642,8 +2642,9 @@ void evergreen_mc_resume(struct radeon_device *rdev, struct evergreen_mc_save *s for (i = 0; i < rdev->num_crtc; i++) { if (save->crtc_enabled[i]) { tmp = RREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i]); - if ((tmp & 0x3) != 0) { - tmp &= ~0x3; + if ((tmp & 0x7) != 3) { + tmp &= ~0x7; + tmp |= 0x3; WREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i], tmp); } tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]); @@ -3337,6 +3338,18 @@ static void evergreen_gpu_init(struct radeon_device *rdev) disabled_rb_mask &= ~(1 << i); } + for (i = 0; i < rdev->config.evergreen.num_ses; i++) { + u32 simd_disable_bitmap; + + WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i)); + WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i)); + simd_disable_bitmap = (RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0xffff0000) >> 16; + simd_disable_bitmap |= 0xffffffff << rdev->config.evergreen.max_simds; + tmp <<= 16; + tmp |= simd_disable_bitmap; + } + rdev->config.evergreen.active_simds = hweight32(~tmp); + WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES); WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES); @@ -4355,7 +4368,6 @@ int evergreen_irq_set(struct radeon_device *rdev) u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0; u32 hpd1, hpd2, hpd3, hpd4, hpd5, hpd6; u32 grbm_int_cntl = 0; - u32 grph1 = 0, grph2 = 0, grph3 = 0, grph4 = 0, grph5 = 0, grph6 = 0; u32 afmt1 = 0, afmt2 = 0, afmt3 = 0, afmt4 = 0, afmt5 = 0, afmt6 = 0; u32 dma_cntl, dma_cntl1 = 0; u32 thermal_int = 0; @@ -4538,15 +4550,21 @@ int evergreen_irq_set(struct radeon_device *rdev) WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6); } - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, grph1); - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, grph2); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); if (rdev->num_crtc >= 4) { - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, grph3); - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, grph4); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); } if (rdev->num_crtc >= 6) { - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, grph5); - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, grph6); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); } WREG32(DC_HPD1_INT_CONTROL, hpd1); @@ -4738,6 +4756,7 @@ static u32 evergreen_get_ih_wptr(struct radeon_device *rdev) tmp = RREG32(IH_RB_CNTL); tmp |= IH_WPTR_OVERFLOW_CLEAR; WREG32(IH_RB_CNTL, tmp); + wptr &= ~RB_OVERFLOW; } return (wptr & rdev->ih.ptr_mask); } @@ -4935,6 +4954,15 @@ restart_ih: break; } break; + case 8: /* D1 page flip */ + case 10: /* D2 page flip */ + case 12: /* D3 page flip */ + case 14: /* D4 page flip */ + case 16: /* D5 page flip */ + case 18: /* D6 page flip */ + DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1); + radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1); + break; case 42: /* HPD hotplug */ switch (src_data) { case 0: @@ -5040,14 +5068,16 @@ restart_ih: case 147: addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR); status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS); + /* reset addr and status */ + WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1); + if (addr == 0x0 && status == 0x0) + break; dev_err(rdev->dev, "GPU fault detected: %d 0x%08x\n", src_id, src_data); dev_err(rdev->dev, " VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x%08X\n", addr); dev_err(rdev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n", status); cayman_vm_decode_fault(rdev, status, addr); - /* reset addr and status */ - WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1); break; case 176: /* CP_INT in ring buffer */ case 177: /* CP_INT in IB1 */ diff --git a/drivers/gpu/drm/radeon/evergreen_dma.c b/drivers/gpu/drm/radeon/evergreen_dma.c index 287fe966d7d..478caefe0fe 100644 --- a/drivers/gpu/drm/radeon/evergreen_dma.c +++ b/drivers/gpu/drm/radeon/evergreen_dma.c @@ -151,6 +151,7 @@ int evergreen_copy_dma(struct radeon_device *rdev, r = radeon_fence_emit(rdev, fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); + radeon_semaphore_free(rdev, &sem, NULL); return r; } diff --git a/drivers/gpu/drm/radeon/evergreen_reg.h b/drivers/gpu/drm/radeon/evergreen_reg.h index a0f63ff5a5e..23bff590fb6 100644 --- a/drivers/gpu/drm/radeon/evergreen_reg.h +++ b/drivers/gpu/drm/radeon/evergreen_reg.h @@ -116,6 +116,8 @@ # define EVERGREEN_GRPH_ARRAY_LINEAR_ALIGNED 1 # define EVERGREEN_GRPH_ARRAY_1D_TILED_THIN1 2 # define EVERGREEN_GRPH_ARRAY_2D_TILED_THIN1 4 +#define EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL 0x6808 +# define EVERGREEN_LUT_10BIT_BYPASS_EN (1 << 8) #define EVERGREEN_GRPH_SWAP_CONTROL 0x680c # define EVERGREEN_GRPH_ENDIAN_SWAP(x) (((x) & 0x3) << 0) # define EVERGREEN_GRPH_ENDIAN_NONE 0 @@ -237,7 +239,6 @@ # define EVERGREEN_CRTC_V_BLANK (1 << 0) #define EVERGREEN_CRTC_STATUS_POSITION 0x6e90 #define EVERGREEN_CRTC_STATUS_HV_COUNT 0x6ea0 -#define EVERGREEN_MASTER_UPDATE_MODE 0x6ef8 #define EVERGREEN_CRTC_UPDATE_LOCK 0x6ed4 #define EVERGREEN_MASTER_UPDATE_LOCK 0x6ef4 #define EVERGREEN_MASTER_UPDATE_MODE 0x6ef8 diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c index 16ec9d56a23..9ef8c38f2d6 100644 --- a/drivers/gpu/drm/radeon/kv_dpm.c +++ b/drivers/gpu/drm/radeon/kv_dpm.c @@ -546,6 +546,52 @@ static int kv_set_divider_value(struct radeon_device *rdev, return 0; } +static u32 kv_convert_vid2_to_vid7(struct radeon_device *rdev, + struct sumo_vid_mapping_table *vid_mapping_table, + u32 vid_2bit) +{ + struct radeon_clock_voltage_dependency_table *vddc_sclk_table = + &rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk; + u32 i; + + if (vddc_sclk_table && vddc_sclk_table->count) { + if (vid_2bit < vddc_sclk_table->count) + return vddc_sclk_table->entries[vid_2bit].v; + else + return vddc_sclk_table->entries[vddc_sclk_table->count - 1].v; + } else { + for (i = 0; i < vid_mapping_table->num_entries; i++) { + if (vid_mapping_table->entries[i].vid_2bit == vid_2bit) + return vid_mapping_table->entries[i].vid_7bit; + } + return vid_mapping_table->entries[vid_mapping_table->num_entries - 1].vid_7bit; + } +} + +static u32 kv_convert_vid7_to_vid2(struct radeon_device *rdev, + struct sumo_vid_mapping_table *vid_mapping_table, + u32 vid_7bit) +{ + struct radeon_clock_voltage_dependency_table *vddc_sclk_table = + &rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk; + u32 i; + + if (vddc_sclk_table && vddc_sclk_table->count) { + for (i = 0; i < vddc_sclk_table->count; i++) { + if (vddc_sclk_table->entries[i].v == vid_7bit) + return i; + } + return vddc_sclk_table->count - 1; + } else { + for (i = 0; i < vid_mapping_table->num_entries; i++) { + if (vid_mapping_table->entries[i].vid_7bit == vid_7bit) + return vid_mapping_table->entries[i].vid_2bit; + } + + return vid_mapping_table->entries[vid_mapping_table->num_entries - 1].vid_2bit; + } +} + static u16 kv_convert_8bit_index_to_voltage(struct radeon_device *rdev, u16 voltage) { @@ -556,9 +602,9 @@ static u16 kv_convert_2bit_index_to_voltage(struct radeon_device *rdev, u32 vid_2bit) { struct kv_power_info *pi = kv_get_pi(rdev); - u32 vid_8bit = sumo_convert_vid2_to_vid7(rdev, - &pi->sys_info.vid_mapping_table, - vid_2bit); + u32 vid_8bit = kv_convert_vid2_to_vid7(rdev, + &pi->sys_info.vid_mapping_table, + vid_2bit); return kv_convert_8bit_index_to_voltage(rdev, (u16)vid_8bit); } @@ -639,7 +685,7 @@ static int kv_force_lowest_valid(struct radeon_device *rdev) static int kv_unforce_levels(struct radeon_device *rdev) { - if (rdev->family == CHIP_KABINI) + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) return kv_notify_message_to_smu(rdev, PPSMC_MSG_NoForcedLevel); else return kv_set_enabled_levels(rdev); @@ -1362,13 +1408,20 @@ static int kv_update_uvd_dpm(struct radeon_device *rdev, bool gate) struct radeon_uvd_clock_voltage_dependency_table *table = &rdev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table; int ret; + u32 mask; if (!gate) { - if (!pi->caps_uvd_dpm || table->count || pi->caps_stable_p_state) + if (table->count) pi->uvd_boot_level = table->count - 1; else pi->uvd_boot_level = 0; + if (!pi->caps_uvd_dpm || pi->caps_stable_p_state) { + mask = 1 << pi->uvd_boot_level; + } else { + mask = 0x1f; + } + ret = kv_copy_bytes_to_smc(rdev, pi->dpm_table_start + offsetof(SMU7_Fusion_DpmTable, UvdBootLevel), @@ -1377,11 +1430,9 @@ static int kv_update_uvd_dpm(struct radeon_device *rdev, bool gate) if (ret) return ret; - if (!pi->caps_uvd_dpm || - pi->caps_stable_p_state) - kv_send_msg_to_smc_with_parameter(rdev, - PPSMC_MSG_UVDDPM_SetEnabledMask, - (1 << pi->uvd_boot_level)); + kv_send_msg_to_smc_with_parameter(rdev, + PPSMC_MSG_UVDDPM_SetEnabledMask, + mask); } return kv_enable_uvd_dpm(rdev, !gate); @@ -1617,7 +1668,7 @@ static void kv_dpm_powergate_acp(struct radeon_device *rdev, bool gate) if (pi->acp_power_gated == gate) return; - if (rdev->family == CHIP_KABINI) + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) return; pi->acp_power_gated = gate; @@ -1786,7 +1837,7 @@ int kv_dpm_set_power_state(struct radeon_device *rdev) } } - if (rdev->family == CHIP_KABINI) { + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) { if (pi->enable_dpm) { kv_set_valid_clock_range(rdev, new_ps); kv_update_dfs_bypass_settings(rdev, new_ps); @@ -1812,6 +1863,8 @@ int kv_dpm_set_power_state(struct radeon_device *rdev) return ret; } kv_update_sclk_t(rdev); + if (rdev->family == CHIP_MULLINS) + kv_enable_nb_dpm(rdev); } } else { if (pi->enable_dpm) { @@ -1862,7 +1915,7 @@ void kv_dpm_reset_asic(struct radeon_device *rdev) { struct kv_power_info *pi = kv_get_pi(rdev); - if (rdev->family == CHIP_KABINI) { + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) { kv_force_lowest_valid(rdev); kv_init_graphics_levels(rdev); kv_program_bootup_state(rdev); @@ -1901,14 +1954,41 @@ static void kv_construct_max_power_limits_table(struct radeon_device *rdev, static void kv_patch_voltage_values(struct radeon_device *rdev) { int i; - struct radeon_uvd_clock_voltage_dependency_table *table = + struct radeon_uvd_clock_voltage_dependency_table *uvd_table = &rdev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table; + struct radeon_vce_clock_voltage_dependency_table *vce_table = + &rdev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table; + struct radeon_clock_voltage_dependency_table *samu_table = + &rdev->pm.dpm.dyn_state.samu_clock_voltage_dependency_table; + struct radeon_clock_voltage_dependency_table *acp_table = + &rdev->pm.dpm.dyn_state.acp_clock_voltage_dependency_table; - if (table->count) { - for (i = 0; i < table->count; i++) - table->entries[i].v = + if (uvd_table->count) { + for (i = 0; i < uvd_table->count; i++) + uvd_table->entries[i].v = kv_convert_8bit_index_to_voltage(rdev, - table->entries[i].v); + uvd_table->entries[i].v); + } + + if (vce_table->count) { + for (i = 0; i < vce_table->count; i++) + vce_table->entries[i].v = + kv_convert_8bit_index_to_voltage(rdev, + vce_table->entries[i].v); + } + + if (samu_table->count) { + for (i = 0; i < samu_table->count; i++) + samu_table->entries[i].v = + kv_convert_8bit_index_to_voltage(rdev, + samu_table->entries[i].v); + } + + if (acp_table->count) { + for (i = 0; i < acp_table->count; i++) + acp_table->entries[i].v = + kv_convert_8bit_index_to_voltage(rdev, + acp_table->entries[i].v); } } @@ -1941,7 +2021,7 @@ static int kv_force_dpm_highest(struct radeon_device *rdev) break; } - if (rdev->family == CHIP_KABINI) + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) return kv_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_DPM_ForceState, i); else return kv_set_enabled_level(rdev, i); @@ -1961,7 +2041,7 @@ static int kv_force_dpm_lowest(struct radeon_device *rdev) break; } - if (rdev->family == CHIP_KABINI) + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) return kv_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_DPM_ForceState, i); else return kv_set_enabled_level(rdev, i); @@ -2118,7 +2198,7 @@ static void kv_apply_state_adjust_rules(struct radeon_device *rdev, else pi->battery_state = false; - if (rdev->family == CHIP_KABINI) { + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) { ps->dpm0_pg_nb_ps_lo = 0x1; ps->dpm0_pg_nb_ps_hi = 0x0; ps->dpmx_nb_ps_lo = 0x1; @@ -2179,7 +2259,7 @@ static int kv_calculate_nbps_level_settings(struct radeon_device *rdev) if (pi->lowest_valid > pi->highest_valid) return -EINVAL; - if (rdev->family == CHIP_KABINI) { + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) { for (i = pi->lowest_valid; i <= pi->highest_valid; i++) { pi->graphics_level[i].GnbSlow = 1; pi->graphics_level[i].ForceNbPs1 = 0; @@ -2253,9 +2333,9 @@ static void kv_init_graphics_levels(struct radeon_device *rdev) break; kv_set_divider_value(rdev, i, table->entries[i].clk); - vid_2bit = sumo_convert_vid7_to_vid2(rdev, - &pi->sys_info.vid_mapping_table, - table->entries[i].v); + vid_2bit = kv_convert_vid7_to_vid2(rdev, + &pi->sys_info.vid_mapping_table, + table->entries[i].v); kv_set_vid(rdev, i, vid_2bit); kv_set_at(rdev, i, pi->at[i]); kv_dpm_power_level_enabled_for_throttle(rdev, i, true); @@ -2324,7 +2404,7 @@ static void kv_program_nbps_index_settings(struct radeon_device *rdev, struct kv_power_info *pi = kv_get_pi(rdev); u32 nbdpmconfig1; - if (rdev->family == CHIP_KABINI) + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) return; if (pi->sys_info.nb_dpm_enable) { @@ -2631,9 +2711,6 @@ int kv_dpm_init(struct radeon_device *rdev) pi->sram_end = SMC_RAM_END; - if (rdev->family == CHIP_KABINI) - pi->high_voltage_t = 4001; - pi->enable_nb_dpm = true; pi->caps_power_containment = true; @@ -2649,7 +2726,7 @@ int kv_dpm_init(struct radeon_device *rdev) pi->caps_sclk_ds = true; pi->enable_auto_thermal_throttling = true; pi->disable_nb_ps3_in_battery = false; - pi->bapm_enable = false; + pi->bapm_enable = true; pi->voltage_drop_t = 0; pi->caps_sclk_throttle_low_notification = false; pi->caps_fps = false; /* true? */ diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c index 1d3209ffbbd..5a33ca68186 100644 --- a/drivers/gpu/drm/radeon/ni.c +++ b/drivers/gpu/drm/radeon/ni.c @@ -1057,6 +1057,18 @@ static void cayman_gpu_init(struct radeon_device *rdev) disabled_rb_mask &= ~(1 << i); } + for (i = 0; i < rdev->config.cayman.max_shader_engines; i++) { + u32 simd_disable_bitmap; + + WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i)); + WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i)); + simd_disable_bitmap = (RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0xffff0000) >> 16; + simd_disable_bitmap |= 0xffffffff << rdev->config.cayman.max_simds_per_se; + tmp <<= 16; + tmp |= simd_disable_bitmap; + } + rdev->config.cayman.active_simds = hweight32(~tmp); + WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES); WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES); @@ -1268,7 +1280,7 @@ static int cayman_pcie_gart_enable(struct radeon_device *rdev) (u32)(rdev->dummy_page.addr >> 12)); WREG32(VM_CONTEXT1_CNTL2, 4); WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) | - PAGE_TABLE_BLOCK_SIZE(RADEON_VM_BLOCK_SIZE - 9) | + PAGE_TABLE_BLOCK_SIZE(radeon_vm_block_size - 9) | RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT | RANGE_PROTECTION_FAULT_ENABLE_DEFAULT | DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT | @@ -1346,7 +1358,7 @@ void cayman_fence_ring_emit(struct radeon_device *rdev, /* EVENT_WRITE_EOP - flush caches, send int */ radeon_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4)); radeon_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_EVENT_TS) | EVENT_INDEX(5)); - radeon_ring_write(ring, addr & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(addr)); radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | DATA_SEL(1) | INT_SEL(2)); radeon_ring_write(ring, fence->seq); radeon_ring_write(ring, 0); diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c index 004c931606c..01fc4888e6f 100644 --- a/drivers/gpu/drm/radeon/ni_dpm.c +++ b/drivers/gpu/drm/radeon/ni_dpm.c @@ -1315,7 +1315,7 @@ static void ni_populate_smc_voltage_tables(struct radeon_device *rdev, table->voltageMaskTable.highMask[NISLANDS_SMC_VOLTAGEMASK_VDDCI] = 0; table->voltageMaskTable.lowMask[NISLANDS_SMC_VOLTAGEMASK_VDDCI] = - cpu_to_be32(eg_pi->vddc_voltage_table.mask_low); + cpu_to_be32(eg_pi->vddci_voltage_table.mask_low); } } diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index ad99813cfa8..1544efcf1c3 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -682,15 +682,11 @@ void r100_pci_gart_disable(struct radeon_device *rdev) WREG32(RADEON_AIC_HI_ADDR, 0); } -int r100_pci_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr) +void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t addr) { u32 *gtt = rdev->gart.ptr; - - if (i < 0 || i > rdev->gart.num_gpu_pages) { - return -EINVAL; - } gtt[i] = cpu_to_le32(lower_32_bits(addr)); - return 0; } void r100_pci_gart_fini(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c index 206caf9700b..3c21d77a483 100644 --- a/drivers/gpu/drm/radeon/r300.c +++ b/drivers/gpu/drm/radeon/r300.c @@ -72,13 +72,11 @@ void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev) #define R300_PTE_WRITEABLE (1 << 2) #define R300_PTE_READABLE (1 << 3) -int rv370_pcie_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr) +void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t addr) { void __iomem *ptr = rdev->gart.ptr; - if (i < 0 || i > rdev->gart.num_gpu_pages) { - return -EINVAL; - } addr = (lower_32_bits(addr) >> 8) | ((upper_32_bits(addr) & 0xff) << 24) | R300_PTE_WRITEABLE | R300_PTE_READABLE; @@ -86,7 +84,6 @@ int rv370_pcie_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr) * on powerpc without HW swappers, it'll get swapped on way * into VRAM - so no need for cpu_to_le32 on VRAM tables */ writel(addr, ((void __iomem *)ptr) + (i * 4)); - return 0; } int rv370_pcie_gart_init(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/r500_reg.h b/drivers/gpu/drm/radeon/r500_reg.h index 1dd0d32993d..136b7bc7cd2 100644 --- a/drivers/gpu/drm/radeon/r500_reg.h +++ b/drivers/gpu/drm/radeon/r500_reg.h @@ -402,6 +402,7 @@ * block and vice versa. This applies to GRPH, CUR, etc. */ #define AVIVO_D1GRPH_LUT_SEL 0x6108 +# define AVIVO_LUT_10BIT_BYPASS_EN (1 << 8) #define AVIVO_D1GRPH_PRIMARY_SURFACE_ADDRESS 0x6110 #define R700_D1GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x6914 #define R700_D2GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x6114 diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index 436e55092e9..3c69f58e46e 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -1958,6 +1958,9 @@ static void r600_gpu_init(struct radeon_device *rdev) if (tmp < rdev->config.r600.max_simds) { rdev->config.r600.max_simds = tmp; } + tmp = rdev->config.r600.max_simds - + r600_count_pipe_bits((cc_gc_shader_pipe_config >> 16) & R6XX_MAX_SIMDS_MASK); + rdev->config.r600.active_simds = tmp; disabled_rb_mask = (RREG32(CC_RB_BACKEND_DISABLE) >> 16) & R6XX_MAX_BACKENDS_MASK; tmp = (tiling_config & PIPE_TILING__MASK) >> PIPE_TILING__SHIFT; @@ -2724,7 +2727,7 @@ void r600_fence_ring_emit(struct radeon_device *rdev, /* EVENT_WRITE_EOP - flush caches, send int */ radeon_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4)); radeon_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_EVENT_TS) | EVENT_INDEX(5)); - radeon_ring_write(ring, addr & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(addr)); radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | DATA_SEL(1) | INT_SEL(2)); radeon_ring_write(ring, fence->seq); radeon_ring_write(ring, 0); @@ -2763,7 +2766,7 @@ bool r600_semaphore_ring_emit(struct radeon_device *rdev, sel |= PACKET3_SEM_WAIT_ON_SIGNAL; radeon_ring_write(ring, PACKET3(PACKET3_MEM_SEMAPHORE, 1)); - radeon_ring_write(ring, addr & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(addr)); radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | sel); return true; @@ -2824,9 +2827,9 @@ int r600_copy_cpdma(struct radeon_device *rdev, if (size_in_bytes == 0) tmp |= PACKET3_CP_DMA_CP_SYNC; radeon_ring_write(ring, PACKET3(PACKET3_CP_DMA, 4)); - radeon_ring_write(ring, src_offset & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(src_offset)); radeon_ring_write(ring, tmp); - radeon_ring_write(ring, dst_offset & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(dst_offset)); radeon_ring_write(ring, upper_32_bits(dst_offset) & 0xff); radeon_ring_write(ring, cur_size_in_bytes); src_offset += cur_size_in_bytes; @@ -2839,6 +2842,7 @@ int r600_copy_cpdma(struct radeon_device *rdev, r = radeon_fence_emit(rdev, fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); + radeon_semaphore_free(rdev, &sem, NULL); return r; } @@ -3505,7 +3509,6 @@ int r600_irq_set(struct radeon_device *rdev) u32 hpd1, hpd2, hpd3, hpd4 = 0, hpd5 = 0, hpd6 = 0; u32 grbm_int_cntl = 0; u32 hdmi0, hdmi1; - u32 d1grph = 0, d2grph = 0; u32 dma_cntl; u32 thermal_int = 0; @@ -3614,8 +3617,8 @@ int r600_irq_set(struct radeon_device *rdev) WREG32(CP_INT_CNTL, cp_int_cntl); WREG32(DMA_CNTL, dma_cntl); WREG32(DxMODE_INT_MASK, mode_int); - WREG32(D1GRPH_INTERRUPT_CONTROL, d1grph); - WREG32(D2GRPH_INTERRUPT_CONTROL, d2grph); + WREG32(D1GRPH_INTERRUPT_CONTROL, DxGRPH_PFLIP_INT_MASK); + WREG32(D2GRPH_INTERRUPT_CONTROL, DxGRPH_PFLIP_INT_MASK); WREG32(GRBM_INT_CNTL, grbm_int_cntl); if (ASIC_IS_DCE3(rdev)) { WREG32(DC_HPD1_INT_CONTROL, hpd1); @@ -3792,6 +3795,7 @@ static u32 r600_get_ih_wptr(struct radeon_device *rdev) tmp = RREG32(IH_RB_CNTL); tmp |= IH_WPTR_OVERFLOW_CLEAR; WREG32(IH_RB_CNTL, tmp); + wptr &= ~RB_OVERFLOW; } return (wptr & rdev->ih.ptr_mask); } @@ -3918,6 +3922,14 @@ restart_ih: break; } break; + case 9: /* D1 pflip */ + DRM_DEBUG("IH: D1 flip\n"); + radeon_crtc_handle_flip(rdev, 0); + break; + case 11: /* D2 pflip */ + DRM_DEBUG("IH: D2 flip\n"); + radeon_crtc_handle_flip(rdev, 1); + break; case 19: /* HPD/DAC hotplug */ switch (src_data) { case 0: diff --git a/drivers/gpu/drm/radeon/r600_dma.c b/drivers/gpu/drm/radeon/r600_dma.c index 53fcb28f557..4969cef44a1 100644 --- a/drivers/gpu/drm/radeon/r600_dma.c +++ b/drivers/gpu/drm/radeon/r600_dma.c @@ -489,6 +489,7 @@ int r600_copy_dma(struct radeon_device *rdev, r = radeon_fence_emit(rdev, fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); + radeon_semaphore_free(rdev, &sem, NULL); return r; } diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index dd4da88b3ab..60c47f82912 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -100,6 +100,9 @@ extern int radeon_dpm; extern int radeon_aspm; extern int radeon_runtime_pm; extern int radeon_hard_reset; +extern int radeon_vm_size; +extern int radeon_vm_block_size; +extern int radeon_deep_color; /* * Copy from radeon_drv.h so we don't have to include both and have conflicting @@ -446,6 +449,7 @@ struct radeon_bo_va { /* protected by vm mutex */ struct list_head vm_list; + struct list_head vm_status; /* constant after initialization */ struct radeon_vm *vm; @@ -681,10 +685,9 @@ struct radeon_flip_work { struct work_struct unpin_work; struct radeon_device *rdev; int crtc_id; - struct drm_framebuffer *fb; + uint64_t base; struct drm_pending_vblank_event *event; struct radeon_bo *old_rbo; - struct radeon_bo *new_rbo; struct radeon_fence *fence; }; @@ -732,6 +735,12 @@ struct cik_irq_stat_regs { u32 disp_int_cont4; u32 disp_int_cont5; u32 disp_int_cont6; + u32 d1grph_int; + u32 d2grph_int; + u32 d3grph_int; + u32 d4grph_int; + u32 d5grph_int; + u32 d6grph_int; }; union radeon_irq_stat_regs { @@ -741,10 +750,6 @@ union radeon_irq_stat_regs { struct cik_irq_stat_regs cik; }; -#define RADEON_MAX_HPD_PINS 7 -#define RADEON_MAX_CRTCS 6 -#define RADEON_MAX_AFMT_BLOCKS 7 - struct radeon_irq { bool installed; spinlock_t lock; @@ -831,13 +836,8 @@ struct radeon_mec { /* maximum number of VMIDs */ #define RADEON_NUM_VM 16 -/* defines number of bits in page table versus page directory, - * a page is 4KB so we have 12 bits offset, 9 bits in the page - * table and the remaining 19 bits are in the page directory */ -#define RADEON_VM_BLOCK_SIZE 9 - /* number of entries in page table */ -#define RADEON_VM_PTE_COUNT (1 << RADEON_VM_BLOCK_SIZE) +#define RADEON_VM_PTE_COUNT (1 << radeon_vm_block_size) /* PTBs (Page Table Blocks) need to be aligned to 32K */ #define RADEON_VM_PTB_ALIGN_SIZE 32768 @@ -868,6 +868,9 @@ struct radeon_vm { struct list_head va; unsigned id; + /* BOs freed, but not yet updated in the PT */ + struct list_head freed; + /* contains the page directory */ struct radeon_bo *page_directory; uint64_t pd_gpu_addr; @@ -876,6 +879,8 @@ struct radeon_vm { /* array of page tables, one for each page directory entry */ struct radeon_vm_pt *page_tables; + struct radeon_bo_va *ib_bo_va; + struct mutex mutex; /* last fence for cs using this vm */ struct radeon_fence *fence; @@ -991,8 +996,8 @@ struct radeon_cs_reloc { struct radeon_bo *robj; struct ttm_validate_buffer tv; uint64_t gpu_offset; - unsigned domain; - unsigned alt_domain; + unsigned prefered_domains; + unsigned allowed_domains; uint32_t tiling_flags; uint32_t handle; }; @@ -1647,6 +1652,7 @@ struct radeon_vce { unsigned fb_version; atomic_t handles[RADEON_MAX_VCE_HANDLES]; struct drm_file *filp[RADEON_MAX_VCE_HANDLES]; + unsigned img_size[RADEON_MAX_VCE_HANDLES]; struct delayed_work idle_work; }; @@ -1660,7 +1666,7 @@ int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring, uint32_t handle, struct radeon_fence **fence); void radeon_vce_free_handles(struct radeon_device *rdev, struct drm_file *filp); void radeon_vce_note_usage(struct radeon_device *rdev); -int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi); +int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi, unsigned size); int radeon_vce_cs_parse(struct radeon_cs_parser *p); bool radeon_vce_semaphore_emit(struct radeon_device *rdev, struct radeon_ring *ring, @@ -1775,7 +1781,8 @@ struct radeon_asic { /* gart */ struct { void (*tlb_flush)(struct radeon_device *rdev); - int (*set_page)(struct radeon_device *rdev, int i, uint64_t addr); + void (*set_page)(struct radeon_device *rdev, unsigned i, + uint64_t addr); } gart; struct { int (*init)(struct radeon_device *rdev); @@ -1927,6 +1934,7 @@ struct r600_asic { unsigned tiling_group_size; unsigned tile_config; unsigned backend_map; + unsigned active_simds; }; struct rv770_asic { @@ -1952,6 +1960,7 @@ struct rv770_asic { unsigned tiling_group_size; unsigned tile_config; unsigned backend_map; + unsigned active_simds; }; struct evergreen_asic { @@ -1978,6 +1987,7 @@ struct evergreen_asic { unsigned tiling_group_size; unsigned tile_config; unsigned backend_map; + unsigned active_simds; }; struct cayman_asic { @@ -2016,6 +2026,7 @@ struct cayman_asic { unsigned multi_gpu_tile_size; unsigned tile_config; + unsigned active_simds; }; struct si_asic { @@ -2046,6 +2057,7 @@ struct si_asic { unsigned tile_config; uint32_t tile_mode_array[32]; + uint32_t active_cus; }; struct cik_asic { @@ -2077,6 +2089,7 @@ struct cik_asic { unsigned tile_config; uint32_t tile_mode_array[32]; uint32_t macrotile_mode_array[16]; + uint32_t active_cus; }; union radeon_asic_config { @@ -2644,7 +2657,8 @@ void r100_pll_errata_after_index(struct radeon_device *rdev); #define ASIC_IS_DCE8(rdev) ((rdev->family >= CHIP_BONAIRE)) #define ASIC_IS_DCE81(rdev) ((rdev->family == CHIP_KAVERI)) #define ASIC_IS_DCE82(rdev) ((rdev->family == CHIP_BONAIRE)) -#define ASIC_IS_DCE83(rdev) ((rdev->family == CHIP_KABINI)) +#define ASIC_IS_DCE83(rdev) ((rdev->family == CHIP_KABINI) || \ + (rdev->family == CHIP_MULLINS)) #define ASIC_IS_LOMBOK(rdev) ((rdev->ddev->pdev->device == 0x6849) || \ (rdev->ddev->pdev->device == 0x6850) || \ @@ -2824,9 +2838,10 @@ void radeon_vm_fence(struct radeon_device *rdev, uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr); int radeon_vm_update_page_directory(struct radeon_device *rdev, struct radeon_vm *vm); +int radeon_vm_clear_freed(struct radeon_device *rdev, + struct radeon_vm *vm); int radeon_vm_bo_update(struct radeon_device *rdev, - struct radeon_vm *vm, - struct radeon_bo *bo, + struct radeon_bo_va *bo_va, struct ttm_mem_reg *mem); void radeon_vm_bo_invalidate(struct radeon_device *rdev, struct radeon_bo *bo); @@ -2839,8 +2854,8 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev, struct radeon_bo_va *bo_va, uint64_t offset, uint32_t flags); -int radeon_vm_bo_rmv(struct radeon_device *rdev, - struct radeon_bo_va *bo_va); +void radeon_vm_bo_rmv(struct radeon_device *rdev, + struct radeon_bo_va *bo_va); /* audio */ void r600_audio_update_hdmi(struct work_struct *work); diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index d8e1587d89c..34b9aa9e3c0 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -2494,6 +2494,7 @@ int radeon_asic_init(struct radeon_device *rdev) break; case CHIP_KAVERI: case CHIP_KABINI: + case CHIP_MULLINS: rdev->asic = &kv_asic; /* set num crtcs */ if (rdev->family == CHIP_KAVERI) { diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h index 0eab015b2cf..01e7c0ad8f0 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.h +++ b/drivers/gpu/drm/radeon/radeon_asic.h @@ -67,7 +67,8 @@ bool r100_gpu_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp); int r100_asic_reset(struct radeon_device *rdev); u32 r100_get_vblank_counter(struct radeon_device *rdev, int crtc); void r100_pci_gart_tlb_flush(struct radeon_device *rdev); -int r100_pci_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr); +void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t addr); void r100_ring_start(struct radeon_device *rdev, struct radeon_ring *ring); int r100_irq_set(struct radeon_device *rdev); int r100_irq_process(struct radeon_device *rdev); @@ -171,7 +172,8 @@ extern void r300_fence_ring_emit(struct radeon_device *rdev, struct radeon_fence *fence); extern int r300_cs_parse(struct radeon_cs_parser *p); extern void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev); -extern int rv370_pcie_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr); +extern void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t addr); extern void rv370_set_pcie_lanes(struct radeon_device *rdev, int lanes); extern int rv370_get_pcie_lanes(struct radeon_device *rdev); extern void r300_set_reg_safe(struct radeon_device *rdev); @@ -206,7 +208,8 @@ extern void rs400_fini(struct radeon_device *rdev); extern int rs400_suspend(struct radeon_device *rdev); extern int rs400_resume(struct radeon_device *rdev); void rs400_gart_tlb_flush(struct radeon_device *rdev); -int rs400_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr); +void rs400_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t addr); uint32_t rs400_mc_rreg(struct radeon_device *rdev, uint32_t reg); void rs400_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); int rs400_gart_init(struct radeon_device *rdev); @@ -229,7 +232,8 @@ int rs600_irq_process(struct radeon_device *rdev); void rs600_irq_disable(struct radeon_device *rdev); u32 rs600_get_vblank_counter(struct radeon_device *rdev, int crtc); void rs600_gart_tlb_flush(struct radeon_device *rdev); -int rs600_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr); +void rs600_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t addr); uint32_t rs600_mc_rreg(struct radeon_device *rdev, uint32_t reg); void rs600_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); void rs600_bandwidth_update(struct radeon_device *rdev); diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index 30844814c25..173f378428a 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -1227,11 +1227,19 @@ bool radeon_atom_get_clock_info(struct drm_device *dev) rdev->clock.default_dispclk = le32_to_cpu(firmware_info->info_21.ulDefaultDispEngineClkFreq); if (rdev->clock.default_dispclk == 0) { - if (ASIC_IS_DCE5(rdev)) + if (ASIC_IS_DCE6(rdev)) + rdev->clock.default_dispclk = 60000; /* 600 Mhz */ + else if (ASIC_IS_DCE5(rdev)) rdev->clock.default_dispclk = 54000; /* 540 Mhz */ else rdev->clock.default_dispclk = 60000; /* 600 Mhz */ } + /* set a reasonable default for DP */ + if (ASIC_IS_DCE6(rdev) && (rdev->clock.default_dispclk < 53900)) { + DRM_INFO("Changing default dispclk from %dMhz to 600Mhz\n", + rdev->clock.default_dispclk / 100); + rdev->clock.default_dispclk = 60000; + } rdev->clock.dp_extclk = le16_to_cpu(firmware_info->info_21.usUniphyDPModeExtClkFreq); rdev->clock.current_dispclk = rdev->clock.default_dispclk; diff --git a/drivers/gpu/drm/radeon/radeon_bios.c b/drivers/gpu/drm/radeon/radeon_bios.c index b3633d9a531..6a03624fada 100644 --- a/drivers/gpu/drm/radeon/radeon_bios.c +++ b/drivers/gpu/drm/radeon/radeon_bios.c @@ -196,6 +196,20 @@ static bool radeon_atrm_get_bios(struct radeon_device *rdev) } } + if (!found) { + while ((pdev = pci_get_class(PCI_CLASS_DISPLAY_OTHER << 8, pdev)) != NULL) { + dhandle = ACPI_HANDLE(&pdev->dev); + if (!dhandle) + continue; + + status = acpi_get_handle(dhandle, "ATRM", &atrm_handle); + if (!ACPI_FAILURE(status)) { + found = true; + break; + } + } + } + if (!found) return false; @@ -612,7 +626,7 @@ static bool radeon_acpi_vfct_bios(struct radeon_device *rdev) vhdr->DeviceID != rdev->pdev->device) { DRM_INFO("ACPI VFCT table is not for this card\n"); goto out_unmap; - }; + } if (vfct->VBIOSImageOffset + sizeof(VFCT_IMAGE_HEADER) + vhdr->ImageLength > tbl_size) { DRM_ERROR("ACPI VFCT image truncated\n"); diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index 4522f7dce65..44831197e82 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -101,6 +101,7 @@ int radeon_get_monitor_bpc(struct drm_connector *connector) struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct radeon_connector_atom_dig *dig_connector; int bpc = 8; + int mode_clock, max_tmds_clock; switch (connector->connector_type) { case DRM_MODE_CONNECTOR_DVII: @@ -166,8 +167,41 @@ int radeon_get_monitor_bpc(struct drm_connector *connector) connector->name, bpc); bpc = 12; } + + /* Any defined maximum tmds clock limit we must not exceed? */ + if (connector->max_tmds_clock > 0) { + /* mode_clock is clock in kHz for mode to be modeset on this connector */ + mode_clock = radeon_connector->pixelclock_for_modeset; + + /* Maximum allowable input clock in kHz */ + max_tmds_clock = connector->max_tmds_clock * 1000; + + DRM_DEBUG("%s: hdmi mode dotclock %d kHz, max tmds input clock %d kHz.\n", + connector->name, mode_clock, max_tmds_clock); + + /* Check if bpc is within clock limit. Try to degrade gracefully otherwise */ + if ((bpc == 12) && (mode_clock * 3/2 > max_tmds_clock)) { + if ((connector->display_info.edid_hdmi_dc_modes & DRM_EDID_HDMI_DC_30) && + (mode_clock * 5/4 <= max_tmds_clock)) + bpc = 10; + else + bpc = 8; + + DRM_DEBUG("%s: HDMI deep color 12 bpc exceeds max tmds clock. Using %d bpc.\n", + connector->name, bpc); + } + + if ((bpc == 10) && (mode_clock * 5/4 > max_tmds_clock)) { + bpc = 8; + DRM_DEBUG("%s: HDMI deep color 10 bpc exceeds max tmds clock. Using %d bpc.\n", + connector->name, bpc); + } + } } + if ((radeon_deep_color == 0) && (bpc > 8)) + bpc = 8; + DRM_DEBUG("%s: Display bpc=%d, returned bpc=%d\n", connector->name, connector->display_info.bpc, bpc); @@ -1257,17 +1291,15 @@ static int radeon_dvi_mode_valid(struct drm_connector *connector, (radeon_connector->connector_object_id == CONNECTOR_OBJECT_ID_DUAL_LINK_DVI_D) || (radeon_connector->connector_object_id == CONNECTOR_OBJECT_ID_HDMI_TYPE_B)) return MODE_OK; - else if (radeon_connector->connector_object_id == CONNECTOR_OBJECT_ID_HDMI_TYPE_A) { - if (ASIC_IS_DCE6(rdev)) { - /* HDMI 1.3+ supports max clock of 340 Mhz */ - if (mode->clock > 340000) - return MODE_CLOCK_HIGH; - else - return MODE_OK; - } else + else if (ASIC_IS_DCE6(rdev) && drm_detect_hdmi_monitor(radeon_connector->edid)) { + /* HDMI 1.3+ supports max clock of 340 Mhz */ + if (mode->clock > 340000) return MODE_CLOCK_HIGH; - } else + else + return MODE_OK; + } else { return MODE_CLOCK_HIGH; + } } /* check against the max pixel clock */ @@ -1518,6 +1550,8 @@ out: static int radeon_dp_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) { + struct drm_device *dev = connector->dev; + struct radeon_device *rdev = dev->dev_private; struct radeon_connector *radeon_connector = to_radeon_connector(connector); struct radeon_connector_atom_dig *radeon_dig_connector = radeon_connector->con_priv; @@ -1548,14 +1582,23 @@ static int radeon_dp_mode_valid(struct drm_connector *connector, return MODE_PANEL; } } - return MODE_OK; } else { if ((radeon_dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT) || - (radeon_dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_eDP)) + (radeon_dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_eDP)) { return radeon_dp_mode_valid_helper(connector, mode); - else - return MODE_OK; + } else { + if (ASIC_IS_DCE6(rdev) && drm_detect_hdmi_monitor(radeon_connector->edid)) { + /* HDMI 1.3+ supports max clock of 340 Mhz */ + if (mode->clock > 340000) + return MODE_CLOCK_HIGH; + } else { + if (mode->clock > 165000) + return MODE_CLOCK_HIGH; + } + } } + + return MODE_OK; } static const struct drm_connector_helper_funcs radeon_dp_connector_helper_funcs = { diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c index 2b6e0ebcc13..ae763f60c8a 100644 --- a/drivers/gpu/drm/radeon/radeon_cs.c +++ b/drivers/gpu/drm/radeon/radeon_cs.c @@ -140,10 +140,10 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) if (p->ring == R600_RING_TYPE_UVD_INDEX && (i == 0 || drm_pci_device_is_agp(p->rdev->ddev))) { /* TODO: is this still needed for NI+ ? */ - p->relocs[i].domain = + p->relocs[i].prefered_domains = RADEON_GEM_DOMAIN_VRAM; - p->relocs[i].alt_domain = + p->relocs[i].allowed_domains = RADEON_GEM_DOMAIN_VRAM; /* prioritize this over any other relocation */ @@ -152,10 +152,16 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) uint32_t domain = r->write_domain ? r->write_domain : r->read_domains; - p->relocs[i].domain = domain; + if (domain & RADEON_GEM_DOMAIN_CPU) { + DRM_ERROR("RADEON_GEM_DOMAIN_CPU is not valid " + "for command submission\n"); + return -EINVAL; + } + + p->relocs[i].prefered_domains = domain; if (domain == RADEON_GEM_DOMAIN_VRAM) domain |= RADEON_GEM_DOMAIN_GTT; - p->relocs[i].alt_domain = domain; + p->relocs[i].allowed_domains = domain; } p->relocs[i].tv.bo = &p->relocs[i].robj->tbo; @@ -342,10 +348,17 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data) return -EINVAL; /* we only support VM on some SI+ rings */ - if ((p->rdev->asic->ring[p->ring]->cs_parse == NULL) && - ((p->cs_flags & RADEON_CS_USE_VM) == 0)) { - DRM_ERROR("Ring %d requires VM!\n", p->ring); - return -EINVAL; + if ((p->cs_flags & RADEON_CS_USE_VM) == 0) { + if (p->rdev->asic->ring[p->ring]->cs_parse == NULL) { + DRM_ERROR("Ring %d requires VM!\n", p->ring); + return -EINVAL; + } + } else { + if (p->rdev->asic->ring[p->ring]->ib_parse == NULL) { + DRM_ERROR("VM not supported on ring %d!\n", + p->ring); + return -EINVAL; + } } } @@ -448,13 +461,23 @@ static int radeon_bo_vm_update_pte(struct radeon_cs_parser *p, struct radeon_vm *vm) { struct radeon_device *rdev = p->rdev; + struct radeon_bo_va *bo_va; int i, r; r = radeon_vm_update_page_directory(rdev, vm); if (r) return r; - r = radeon_vm_bo_update(rdev, vm, rdev->ring_tmp_bo.bo, + r = radeon_vm_clear_freed(rdev, vm); + if (r) + return r; + + if (vm->ib_bo_va == NULL) { + DRM_ERROR("Tmp BO not in VM!\n"); + return -EINVAL; + } + + r = radeon_vm_bo_update(rdev, vm->ib_bo_va, &rdev->ring_tmp_bo.bo->tbo.mem); if (r) return r; @@ -467,7 +490,13 @@ static int radeon_bo_vm_update_pte(struct radeon_cs_parser *p, continue; bo = p->relocs[i].robj; - r = radeon_vm_bo_update(rdev, vm, bo, &bo->tbo.mem); + bo_va = radeon_vm_bo_find(vm, bo); + if (bo_va == NULL) { + dev_err(rdev->dev, "bo %p not in vm %p\n", bo, vm); + return -EINVAL; + } + + r = radeon_vm_bo_update(rdev, bo_va, &bo->tbo.mem); if (r) return r; } diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 9aa1afd1786..697add2cd4e 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -99,6 +99,7 @@ static const char radeon_family_name[][16] = { "KAVERI", "KABINI", "HAWAII", + "MULLINS", "LAST", }; @@ -1051,6 +1052,43 @@ static void radeon_check_arguments(struct radeon_device *rdev) radeon_agpmode = 0; break; } + + if (!radeon_check_pot_argument(radeon_vm_size)) { + dev_warn(rdev->dev, "VM size (%d) must be a power of 2\n", + radeon_vm_size); + radeon_vm_size = 4; + } + + if (radeon_vm_size < 1) { + dev_warn(rdev->dev, "VM size (%d) to small, min is 1GB\n", + radeon_vm_size); + radeon_vm_size = 4; + } + + /* + * Max GPUVM size for Cayman, SI and CI are 40 bits. + */ + if (radeon_vm_size > 1024) { + dev_warn(rdev->dev, "VM size (%d) too large, max is 1TB\n", + radeon_vm_size); + radeon_vm_size = 4; + } + + /* defines number of bits in page table versus page directory, + * a page is 4KB so we have 12 bits offset, minimum 9 bits in the + * page table and the remaining bits are in the page directory */ + if (radeon_vm_block_size < 9) { + dev_warn(rdev->dev, "VM page table size (%d) too small\n", + radeon_vm_block_size); + radeon_vm_block_size = 9; + } + + if (radeon_vm_block_size > 24 || + (radeon_vm_size * 1024) < (1ull << radeon_vm_block_size)) { + dev_warn(rdev->dev, "VM page table size (%d) too large\n", + radeon_vm_block_size); + radeon_vm_block_size = 9; + } } /** @@ -1196,17 +1234,16 @@ int radeon_device_init(struct radeon_device *rdev, if (r) return r; + radeon_check_arguments(rdev); /* Adjust VM size here. - * Currently set to 4GB ((1 << 20) 4k pages). - * Max GPUVM size for cayman and SI is 40 bits. + * Max GPUVM size for cayman+ is 40 bits. */ - rdev->vm_manager.max_pfn = 1 << 20; + rdev->vm_manager.max_pfn = radeon_vm_size << 18; /* Set asic functions */ r = radeon_asic_init(rdev); if (r) return r; - radeon_check_arguments(rdev); /* all of the newer IGP chips have an internal gart * However some rs4xx report as AGP, so remove that here. @@ -1533,11 +1570,6 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon) radeon_restore_bios_scratch_regs(rdev); - if (fbcon) { - radeon_fbdev_set_suspend(rdev, 0); - console_unlock(); - } - /* init dig PHYs, disp eng pll */ if (rdev->is_atom_bios) { radeon_atom_encoder_init(rdev); @@ -1562,6 +1594,16 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon) } drm_kms_helper_poll_enable(dev); + + /* set the power state here in case we are a PX system or headless */ + if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) + radeon_pm_compute_clocks(rdev); + + if (fbcon) { + radeon_fbdev_set_suspend(rdev, 0); + console_unlock(); + } + return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index e330e762c36..bf25061c8ac 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -66,7 +66,8 @@ static void avivo_crtc_load_lut(struct drm_crtc *crtc) (radeon_crtc->lut_b[i] << 0)); } - WREG32(AVIVO_D1GRPH_LUT_SEL + radeon_crtc->crtc_offset, radeon_crtc->crtc_id); + /* Only change bit 0 of LUT_SEL, other bits are set elsewhere */ + WREG32_P(AVIVO_D1GRPH_LUT_SEL + radeon_crtc->crtc_offset, radeon_crtc->crtc_id, ~1); } static void dce4_crtc_load_lut(struct drm_crtc *crtc) @@ -284,14 +285,20 @@ static void radeon_unpin_work_func(struct work_struct *__work) void radeon_crtc_handle_vblank(struct radeon_device *rdev, int crtc_id) { struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id]; - struct radeon_flip_work *work; unsigned long flags; u32 update_pending; int vpos, hpos; + /* can happen during initialization */ + if (radeon_crtc == NULL) + return; + spin_lock_irqsave(&rdev->ddev->event_lock, flags); - work = radeon_crtc->flip_work; - if (work == NULL) { + if (radeon_crtc->flip_status != RADEON_FLIP_SUBMITTED) { + DRM_DEBUG_DRIVER("radeon_crtc->flip_status = %d != " + "RADEON_FLIP_SUBMITTED(%d)\n", + radeon_crtc->flip_status, + RADEON_FLIP_SUBMITTED); spin_unlock_irqrestore(&rdev->ddev->event_lock, flags); return; } @@ -339,12 +346,17 @@ void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id) spin_lock_irqsave(&rdev->ddev->event_lock, flags); work = radeon_crtc->flip_work; - if (work == NULL) { + if (radeon_crtc->flip_status != RADEON_FLIP_SUBMITTED) { + DRM_DEBUG_DRIVER("radeon_crtc->flip_status = %d != " + "RADEON_FLIP_SUBMITTED(%d)\n", + radeon_crtc->flip_status, + RADEON_FLIP_SUBMITTED); spin_unlock_irqrestore(&rdev->ddev->event_lock, flags); return; } /* Pageflip completed. Clean up. */ + radeon_crtc->flip_status = RADEON_FLIP_NONE; radeon_crtc->flip_work = NULL; /* wakeup userspace */ @@ -353,8 +365,8 @@ void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id) spin_unlock_irqrestore(&rdev->ddev->event_lock, flags); - radeon_fence_unref(&work->fence); - radeon_irq_kms_pflip_irq_get(rdev, work->crtc_id); + drm_vblank_put(rdev->ddev, radeon_crtc->crtc_id); + radeon_irq_kms_pflip_irq_put(rdev, work->crtc_id); queue_work(radeon_crtc->flip_queue, &work->unpin_work); } @@ -373,51 +385,108 @@ static void radeon_flip_work_func(struct work_struct *__work) struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[work->crtc_id]; struct drm_crtc *crtc = &radeon_crtc->base; - struct drm_framebuffer *fb = work->fb; - - uint32_t tiling_flags, pitch_pixels; - uint64_t base; - unsigned long flags; int r; down_read(&rdev->exclusive_lock); - while (work->fence) { + if (work->fence) { r = radeon_fence_wait(work->fence, false); if (r == -EDEADLK) { up_read(&rdev->exclusive_lock); r = radeon_gpu_reset(rdev); down_read(&rdev->exclusive_lock); } + if (r) + DRM_ERROR("failed to wait on page flip fence (%d)!\n", r); - if (r) { - DRM_ERROR("failed to wait on page flip fence (%d)!\n", - r); - goto cleanup; - } else - radeon_fence_unref(&work->fence); + /* We continue with the page flip even if we failed to wait on + * the fence, otherwise the DRM core and userspace will be + * confused about which BO the CRTC is scanning out + */ + + radeon_fence_unref(&work->fence); } + /* We borrow the event spin lock for protecting flip_status */ + spin_lock_irqsave(&crtc->dev->event_lock, flags); + + /* set the proper interrupt */ + radeon_irq_kms_pflip_irq_get(rdev, radeon_crtc->crtc_id); + + /* do the flip (mmio) */ + radeon_page_flip(rdev, radeon_crtc->crtc_id, work->base); + + radeon_crtc->flip_status = RADEON_FLIP_SUBMITTED; + spin_unlock_irqrestore(&crtc->dev->event_lock, flags); + up_read(&rdev->exclusive_lock); +} + +static int radeon_crtc_page_flip(struct drm_crtc *crtc, + struct drm_framebuffer *fb, + struct drm_pending_vblank_event *event, + uint32_t page_flip_flags) +{ + struct drm_device *dev = crtc->dev; + struct radeon_device *rdev = dev->dev_private; + struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); + struct radeon_framebuffer *old_radeon_fb; + struct radeon_framebuffer *new_radeon_fb; + struct drm_gem_object *obj; + struct radeon_flip_work *work; + struct radeon_bo *new_rbo; + uint32_t tiling_flags, pitch_pixels; + uint64_t base; + unsigned long flags; + int r; + + work = kzalloc(sizeof *work, GFP_KERNEL); + if (work == NULL) + return -ENOMEM; + + INIT_WORK(&work->flip_work, radeon_flip_work_func); + INIT_WORK(&work->unpin_work, radeon_unpin_work_func); + + work->rdev = rdev; + work->crtc_id = radeon_crtc->crtc_id; + work->event = event; + + /* schedule unpin of the old buffer */ + old_radeon_fb = to_radeon_framebuffer(crtc->primary->fb); + obj = old_radeon_fb->obj; + + /* take a reference to the old object */ + drm_gem_object_reference(obj); + work->old_rbo = gem_to_radeon_bo(obj); + + new_radeon_fb = to_radeon_framebuffer(fb); + obj = new_radeon_fb->obj; + new_rbo = gem_to_radeon_bo(obj); + + spin_lock(&new_rbo->tbo.bdev->fence_lock); + if (new_rbo->tbo.sync_obj) + work->fence = radeon_fence_ref(new_rbo->tbo.sync_obj); + spin_unlock(&new_rbo->tbo.bdev->fence_lock); + /* pin the new buffer */ - DRM_DEBUG_DRIVER("flip-ioctl() cur_fbo = %p, cur_bbo = %p\n", - work->old_rbo, work->new_rbo); + DRM_DEBUG_DRIVER("flip-ioctl() cur_rbo = %p, new_rbo = %p\n", + work->old_rbo, new_rbo); - r = radeon_bo_reserve(work->new_rbo, false); + r = radeon_bo_reserve(new_rbo, false); if (unlikely(r != 0)) { DRM_ERROR("failed to reserve new rbo buffer before flip\n"); goto cleanup; } /* Only 27 bit offset for legacy CRTC */ - r = radeon_bo_pin_restricted(work->new_rbo, RADEON_GEM_DOMAIN_VRAM, + r = radeon_bo_pin_restricted(new_rbo, RADEON_GEM_DOMAIN_VRAM, ASIC_IS_AVIVO(rdev) ? 0 : 1 << 27, &base); if (unlikely(r != 0)) { - radeon_bo_unreserve(work->new_rbo); + radeon_bo_unreserve(new_rbo); r = -EINVAL; DRM_ERROR("failed to pin new rbo buffer before flip\n"); goto cleanup; } - radeon_bo_get_tiling_flags(work->new_rbo, &tiling_flags, NULL); - radeon_bo_unreserve(work->new_rbo); + radeon_bo_get_tiling_flags(new_rbo, &tiling_flags, NULL); + radeon_bo_unreserve(new_rbo); if (!ASIC_IS_AVIVO(rdev)) { /* crtc offset is from display base addr not FB location */ @@ -454,92 +523,53 @@ static void radeon_flip_work_func(struct work_struct *__work) } base &= ~7; } + work->base = base; + + r = drm_vblank_get(crtc->dev, radeon_crtc->crtc_id); + if (r) { + DRM_ERROR("failed to get vblank before flip\n"); + goto pflip_cleanup; + } /* We borrow the event spin lock for protecting flip_work */ spin_lock_irqsave(&crtc->dev->event_lock, flags); + if (radeon_crtc->flip_status != RADEON_FLIP_NONE) { + DRM_DEBUG_DRIVER("flip queue: crtc already busy\n"); + spin_unlock_irqrestore(&crtc->dev->event_lock, flags); + r = -EBUSY; + goto vblank_cleanup; + } + radeon_crtc->flip_status = RADEON_FLIP_PENDING; + radeon_crtc->flip_work = work; + /* update crtc fb */ crtc->primary->fb = fb; - /* set the proper interrupt */ - radeon_irq_kms_pflip_irq_get(rdev, radeon_crtc->crtc_id); + spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - /* do the flip (mmio) */ - radeon_page_flip(rdev, radeon_crtc->crtc_id, base); + queue_work(radeon_crtc->flip_queue, &work->flip_work); + return 0; - spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - up_read(&rdev->exclusive_lock); +vblank_cleanup: + drm_vblank_put(crtc->dev, radeon_crtc->crtc_id); - return; +pflip_cleanup: + if (unlikely(radeon_bo_reserve(new_rbo, false) != 0)) { + DRM_ERROR("failed to reserve new rbo in error path\n"); + goto cleanup; + } + if (unlikely(radeon_bo_unpin(new_rbo) != 0)) { + DRM_ERROR("failed to unpin new rbo in error path\n"); + } + radeon_bo_unreserve(new_rbo); cleanup: drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base); radeon_fence_unref(&work->fence); kfree(work); - up_read(&rdev->exclusive_lock); -} - -static int radeon_crtc_page_flip(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct drm_pending_vblank_event *event, - uint32_t page_flip_flags) -{ - struct drm_device *dev = crtc->dev; - struct radeon_device *rdev = dev->dev_private; - struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); - struct radeon_framebuffer *old_radeon_fb; - struct radeon_framebuffer *new_radeon_fb; - struct drm_gem_object *obj; - struct radeon_flip_work *work; - unsigned long flags; - - work = kzalloc(sizeof *work, GFP_KERNEL); - if (work == NULL) - return -ENOMEM; - - INIT_WORK(&work->flip_work, radeon_flip_work_func); - INIT_WORK(&work->unpin_work, radeon_unpin_work_func); - - work->rdev = rdev; - work->crtc_id = radeon_crtc->crtc_id; - work->fb = fb; - work->event = event; - - /* schedule unpin of the old buffer */ - old_radeon_fb = to_radeon_framebuffer(crtc->primary->fb); - obj = old_radeon_fb->obj; - /* take a reference to the old object */ - drm_gem_object_reference(obj); - work->old_rbo = gem_to_radeon_bo(obj); - - new_radeon_fb = to_radeon_framebuffer(fb); - obj = new_radeon_fb->obj; - work->new_rbo = gem_to_radeon_bo(obj); - - spin_lock(&work->new_rbo->tbo.bdev->fence_lock); - if (work->new_rbo->tbo.sync_obj) - work->fence = radeon_fence_ref(work->new_rbo->tbo.sync_obj); - spin_unlock(&work->new_rbo->tbo.bdev->fence_lock); - - /* We borrow the event spin lock for protecting flip_work */ - spin_lock_irqsave(&crtc->dev->event_lock, flags); - - if (radeon_crtc->flip_work) { - DRM_DEBUG_DRIVER("flip queue: crtc already busy\n"); - spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base); - radeon_fence_unref(&work->fence); - kfree(work); - return -EBUSY; - } - radeon_crtc->flip_work = work; - - spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - - queue_work(radeon_crtc->flip_queue, &work->flip_work); - - return 0; + return r; } static int @@ -799,6 +829,10 @@ int radeon_ddc_get_modes(struct radeon_connector *radeon_connector) struct radeon_device *rdev = dev->dev_private; int ret = 0; + /* don't leak the edid if we already fetched it in detect() */ + if (radeon_connector->edid) + goto got_edid; + /* on hw with routers, select right port */ if (radeon_connector->router.ddc_valid) radeon_router_select_ddc_port(radeon_connector); @@ -837,6 +871,7 @@ int radeon_ddc_get_modes(struct radeon_connector *radeon_connector) radeon_connector->edid = radeon_bios_get_hardcoded_edid(rdev); } if (radeon_connector->edid) { +got_edid: drm_mode_connector_update_edid_property(&radeon_connector->base, radeon_connector->edid); ret = drm_add_edid_modes(&radeon_connector->base, radeon_connector->edid); drm_edid_to_eld(&radeon_connector->base, radeon_connector->edid); @@ -872,14 +907,14 @@ static void avivo_reduce_ratio(unsigned *nom, unsigned *den, /* make sure nominator is large enough */ if (*nom < nom_min) { - tmp = (nom_min + *nom - 1) / *nom; + tmp = DIV_ROUND_UP(nom_min, *nom); *nom *= tmp; *den *= tmp; } /* make sure the denominator is large enough */ if (*den < den_min) { - tmp = (den_min + *den - 1) / *den; + tmp = DIV_ROUND_UP(den_min, *den); *nom *= tmp; *den *= tmp; } @@ -904,7 +939,7 @@ static void avivo_get_fb_ref_div(unsigned nom, unsigned den, unsigned post_div, unsigned *fb_div, unsigned *ref_div) { /* limit reference * post divider to a maximum */ - ref_div_max = min(210 / post_div, ref_div_max); + ref_div_max = max(min(100 / post_div, ref_div_max), 1u); /* get matching reference and feedback divider */ *ref_div = min(max(DIV_ROUND_CLOSEST(den, post_div), 1u), ref_div_max); @@ -1039,6 +1074,16 @@ void radeon_compute_pll_avivo(struct radeon_pll *pll, /* this also makes sure that the reference divider is large enough */ avivo_reduce_ratio(&fb_div, &ref_div, fb_div_min, ref_div_min); + /* avoid high jitter with small fractional dividers */ + if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV && (fb_div % 10)) { + fb_div_min = max(fb_div_min, (9 - (fb_div % 10)) * 20 + 50); + if (fb_div < fb_div_min) { + unsigned tmp = DIV_ROUND_UP(fb_div_min, fb_div); + fb_div *= tmp; + ref_div *= tmp; + } + } + /* and finally save the result */ if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV) { *fb_div_p = fb_div / 10; diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index 15447a4119f..e9e36108424 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -81,9 +81,10 @@ * 2.37.0 - allow GS ring setup on r6xx/r7xx * 2.38.0 - RADEON_GEM_OP (GET_INITIAL_DOMAIN, SET_INITIAL_DOMAIN), * CIK: 1D and linear tiling modes contain valid PIPE_CONFIG + * 2.39.0 - Add INFO query for number of active CUs */ #define KMS_DRIVER_MAJOR 2 -#define KMS_DRIVER_MINOR 38 +#define KMS_DRIVER_MINOR 39 #define KMS_DRIVER_PATCHLEVEL 0 int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags); int radeon_driver_unload_kms(struct drm_device *dev); @@ -172,6 +173,9 @@ int radeon_dpm = -1; int radeon_aspm = -1; int radeon_runtime_pm = -1; int radeon_hard_reset = 0; +int radeon_vm_size = 4; +int radeon_vm_block_size = 9; +int radeon_deep_color = 0; MODULE_PARM_DESC(no_wb, "Disable AGP writeback for scratch registers"); module_param_named(no_wb, radeon_no_wb, int, 0444); @@ -239,6 +243,15 @@ module_param_named(runpm, radeon_runtime_pm, int, 0444); MODULE_PARM_DESC(hard_reset, "PCI config reset (1 = force enable, 0 = disable (default))"); module_param_named(hard_reset, radeon_hard_reset, int, 0444); +MODULE_PARM_DESC(vm_size, "VM address space size in gigabytes (default 4GB)"); +module_param_named(vm_size, radeon_vm_size, int, 0444); + +MODULE_PARM_DESC(vm_block_size, "VM page table size in bits (default 9)"); +module_param_named(vm_block_size, radeon_vm_block_size, int, 0444); + +MODULE_PARM_DESC(deep_color, "Deep Color support (1 = enable, 0 = disable (default))"); +module_param_named(deep_color, radeon_deep_color, int, 0444); + static struct pci_device_id pciidlist[] = { radeon_PCI_IDS }; diff --git a/drivers/gpu/drm/radeon/radeon_family.h b/drivers/gpu/drm/radeon/radeon_family.h index 9da5da4ffd1..4b7b87f71a6 100644 --- a/drivers/gpu/drm/radeon/radeon_family.h +++ b/drivers/gpu/drm/radeon/radeon_family.h @@ -97,6 +97,7 @@ enum radeon_family { CHIP_KAVERI, CHIP_KABINI, CHIP_HAWAII, + CHIP_MULLINS, CHIP_LAST, }; diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index a77b1c13ea4..913787085df 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -819,15 +819,35 @@ static int radeon_debugfs_fence_info(struct seq_file *m, void *data) return 0; } +/** + * radeon_debugfs_gpu_reset - manually trigger a gpu reset + * + * Manually trigger a gpu reset at the next fence wait. + */ +static int radeon_debugfs_gpu_reset(struct seq_file *m, void *data) +{ + struct drm_info_node *node = (struct drm_info_node *) m->private; + struct drm_device *dev = node->minor->dev; + struct radeon_device *rdev = dev->dev_private; + + down_read(&rdev->exclusive_lock); + seq_printf(m, "%d\n", rdev->needs_reset); + rdev->needs_reset = true; + up_read(&rdev->exclusive_lock); + + return 0; +} + static struct drm_info_list radeon_debugfs_fence_list[] = { {"radeon_fence_info", &radeon_debugfs_fence_info, 0, NULL}, + {"radeon_gpu_reset", &radeon_debugfs_gpu_reset, 0, NULL} }; #endif int radeon_debugfs_fence_init(struct radeon_device *rdev) { #if defined(CONFIG_DEBUG_FS) - return radeon_debugfs_add_files(rdev, radeon_debugfs_fence_list, 1); + return radeon_debugfs_add_files(rdev, radeon_debugfs_fence_list, 2); #else return 0; #endif diff --git a/drivers/gpu/drm/radeon/radeon_i2c.c b/drivers/gpu/drm/radeon/radeon_i2c.c index 427ee4d6d0b..add62200840 100644 --- a/drivers/gpu/drm/radeon/radeon_i2c.c +++ b/drivers/gpu/drm/radeon/radeon_i2c.c @@ -988,7 +988,7 @@ void radeon_i2c_destroy(struct radeon_i2c_chan *i2c) return; i2c_del_adapter(&i2c->adapter); if (i2c->has_aux) - drm_dp_aux_unregister_i2c_bus(&i2c->aux); + drm_dp_aux_unregister(&i2c->aux); kfree(i2c); } diff --git a/drivers/gpu/drm/radeon/radeon_ioc32.c b/drivers/gpu/drm/radeon/radeon_ioc32.c index bdb0f93e73b..0b98ea13457 100644 --- a/drivers/gpu/drm/radeon/radeon_ioc32.c +++ b/drivers/gpu/drm/radeon/radeon_ioc32.c @@ -399,7 +399,7 @@ long radeon_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) if (nr < DRM_COMMAND_BASE) return drm_compat_ioctl(filp, cmd, arg); - if (nr < DRM_COMMAND_BASE + DRM_ARRAY_SIZE(radeon_compat_ioctls)) + if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(radeon_compat_ioctls)) fn = radeon_compat_ioctls[nr - DRM_COMMAND_BASE]; if (fn != NULL) diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index 0cc47f12d99..d25ae6acfd5 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -513,6 +513,22 @@ static int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file value_size = sizeof(uint64_t); value64 = atomic64_read(&rdev->gtt_usage); break; + case RADEON_INFO_ACTIVE_CU_COUNT: + if (rdev->family >= CHIP_BONAIRE) + *value = rdev->config.cik.active_cus; + else if (rdev->family >= CHIP_TAHITI) + *value = rdev->config.si.active_cus; + else if (rdev->family >= CHIP_CAYMAN) + *value = rdev->config.cayman.active_simds; + else if (rdev->family >= CHIP_CEDAR) + *value = rdev->config.evergreen.active_simds; + else if (rdev->family >= CHIP_RV770) + *value = rdev->config.rv770.active_simds; + else if (rdev->family >= CHIP_R600) + *value = rdev->config.r600.active_simds; + else + *value = 1; + break; default: DRM_DEBUG_KMS("Invalid request %d\n", info->request); return -EINVAL; @@ -563,7 +579,7 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) /* new gpu have virtual address space support */ if (rdev->family >= CHIP_CAYMAN) { struct radeon_fpriv *fpriv; - struct radeon_bo_va *bo_va; + struct radeon_vm *vm; int r; fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL); @@ -571,34 +587,37 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) return -ENOMEM; } - r = radeon_vm_init(rdev, &fpriv->vm); + vm = &fpriv->vm; + r = radeon_vm_init(rdev, vm); if (r) { kfree(fpriv); return r; } - r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false); - if (r) { - radeon_vm_fini(rdev, &fpriv->vm); - kfree(fpriv); - return r; - } + if (rdev->accel_working) { + r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false); + if (r) { + radeon_vm_fini(rdev, vm); + kfree(fpriv); + return r; + } - /* map the ib pool buffer read only into - * virtual address space */ - bo_va = radeon_vm_bo_add(rdev, &fpriv->vm, - rdev->ring_tmp_bo.bo); - r = radeon_vm_bo_set_addr(rdev, bo_va, RADEON_VA_IB_OFFSET, - RADEON_VM_PAGE_READABLE | - RADEON_VM_PAGE_SNOOPED); + /* map the ib pool buffer read only into + * virtual address space */ + vm->ib_bo_va = radeon_vm_bo_add(rdev, vm, + rdev->ring_tmp_bo.bo); + r = radeon_vm_bo_set_addr(rdev, vm->ib_bo_va, + RADEON_VA_IB_OFFSET, + RADEON_VM_PAGE_READABLE | + RADEON_VM_PAGE_SNOOPED); - radeon_bo_unreserve(rdev->ring_tmp_bo.bo); - if (r) { - radeon_vm_fini(rdev, &fpriv->vm); - kfree(fpriv); - return r; + radeon_bo_unreserve(rdev->ring_tmp_bo.bo); + if (r) { + radeon_vm_fini(rdev, vm); + kfree(fpriv); + return r; + } } - file_priv->driver_priv = fpriv; } @@ -623,19 +642,19 @@ void radeon_driver_postclose_kms(struct drm_device *dev, /* new gpu have virtual address space support */ if (rdev->family >= CHIP_CAYMAN && file_priv->driver_priv) { struct radeon_fpriv *fpriv = file_priv->driver_priv; - struct radeon_bo_va *bo_va; + struct radeon_vm *vm = &fpriv->vm; int r; - r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false); - if (!r) { - bo_va = radeon_vm_bo_find(&fpriv->vm, - rdev->ring_tmp_bo.bo); - if (bo_va) - radeon_vm_bo_rmv(rdev, bo_va); - radeon_bo_unreserve(rdev->ring_tmp_bo.bo); + if (rdev->accel_working) { + r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false); + if (!r) { + if (vm->ib_bo_va) + radeon_vm_bo_rmv(rdev, vm->ib_bo_va); + radeon_bo_unreserve(rdev->ring_tmp_bo.bo); + } } - radeon_vm_fini(rdev, &fpriv->vm); + radeon_vm_fini(rdev, vm); kfree(fpriv); file_priv->driver_priv = NULL; } @@ -856,4 +875,4 @@ const struct drm_ioctl_desc radeon_ioctls_kms[] = { DRM_IOCTL_DEF_DRV(RADEON_GEM_VA, radeon_gem_va_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(RADEON_GEM_OP, radeon_gem_op_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW), }; -int radeon_max_kms_ioctl = DRM_ARRAY_SIZE(radeon_ioctls_kms); +int radeon_max_kms_ioctl = ARRAY_SIZE(radeon_ioctls_kms); diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index ea72ad889a1..0592ddb0904 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -46,6 +46,10 @@ struct radeon_device; #define to_radeon_encoder(x) container_of(x, struct radeon_encoder, base) #define to_radeon_framebuffer(x) container_of(x, struct radeon_framebuffer, base) +#define RADEON_MAX_HPD_PINS 7 +#define RADEON_MAX_CRTCS 6 +#define RADEON_MAX_AFMT_BLOCKS 7 + enum radeon_rmx_type { RMX_OFF, RMX_FULL, @@ -233,8 +237,8 @@ struct radeon_mode_info { struct card_info *atom_card_info; enum radeon_connector_table connector_table; bool mode_config_initialized; - struct radeon_crtc *crtcs[6]; - struct radeon_afmt *afmt[7]; + struct radeon_crtc *crtcs[RADEON_MAX_CRTCS]; + struct radeon_afmt *afmt[RADEON_MAX_AFMT_BLOCKS]; /* DVI-I properties */ struct drm_property *coherent_mode_property; /* DAC enable load detect */ @@ -302,6 +306,12 @@ struct radeon_atom_ss { uint16_t amount; }; +enum radeon_flip_status { + RADEON_FLIP_NONE, + RADEON_FLIP_PENDING, + RADEON_FLIP_SUBMITTED +}; + struct radeon_crtc { struct drm_crtc base; int crtc_id; @@ -327,6 +337,7 @@ struct radeon_crtc { /* page flipping */ struct workqueue_struct *flip_queue; struct radeon_flip_work *flip_work; + enum radeon_flip_status flip_status; /* pll sharing */ struct radeon_atom_ss ss; bool ss_enabled; @@ -506,6 +517,7 @@ struct radeon_connector { struct radeon_i2c_chan *router_bus; enum radeon_connector_audio audio; enum radeon_connector_dither dither; + int pixelclock_for_modeset; }; struct radeon_framebuffer { diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 95197aa4de4..6c717b257d6 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -446,7 +446,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev, list_for_each_entry(lobj, head, tv.head) { bo = lobj->robj; if (!bo->pin_count) { - u32 domain = lobj->domain; + u32 domain = lobj->prefered_domains; u32 current_domain = radeon_mem_type_to_domain(bo->tbo.mem.mem_type); @@ -458,7 +458,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev, * into account. We don't want to disallow buffer moves * completely. */ - if (current_domain != RADEON_GEM_DOMAIN_CPU && + if ((lobj->allowed_domains & current_domain) != 0 && (domain & current_domain) == 0 && /* will be moved */ bytes_moved > bytes_moved_threshold) { /* don't move it */ @@ -476,8 +476,9 @@ int radeon_bo_list_validate(struct radeon_device *rdev, initial_bytes_moved; if (unlikely(r)) { - if (r != -ERESTARTSYS && domain != lobj->alt_domain) { - domain = lobj->alt_domain; + if (r != -ERESTARTSYS && + domain != lobj->allowed_domains) { + domain = lobj->allowed_domains; goto retry; } ttm_eu_backoff_reservation(ticket, head); @@ -699,22 +700,30 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo) rbo = container_of(bo, struct radeon_bo, tbo); radeon_bo_check_tiling(rbo, 0, 0); rdev = rbo->rdev; - if (bo->mem.mem_type == TTM_PL_VRAM) { - size = bo->mem.num_pages << PAGE_SHIFT; - offset = bo->mem.start << PAGE_SHIFT; - if ((offset + size) > rdev->mc.visible_vram_size) { - /* hurrah the memory is not visible ! */ - radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM); - rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT; - r = ttm_bo_validate(bo, &rbo->placement, false, false); - if (unlikely(r != 0)) - return r; - offset = bo->mem.start << PAGE_SHIFT; - /* this should not happen */ - if ((offset + size) > rdev->mc.visible_vram_size) - return -EINVAL; - } + if (bo->mem.mem_type != TTM_PL_VRAM) + return 0; + + size = bo->mem.num_pages << PAGE_SHIFT; + offset = bo->mem.start << PAGE_SHIFT; + if ((offset + size) <= rdev->mc.visible_vram_size) + return 0; + + /* hurrah the memory is not visible ! */ + radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM); + rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT; + r = ttm_bo_validate(bo, &rbo->placement, false, false); + if (unlikely(r == -ENOMEM)) { + radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT); + return ttm_bo_validate(bo, &rbo->placement, false, false); + } else if (unlikely(r != 0)) { + return r; } + + offset = bo->mem.start << PAGE_SHIFT; + /* this should never happen */ + if ((offset + size) > rdev->mc.visible_vram_size) + return -EINVAL; + return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index 6fac8efe834..e447e390d09 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -73,8 +73,10 @@ void radeon_pm_acpi_event_handler(struct radeon_device *rdev) rdev->pm.dpm.ac_power = true; else rdev->pm.dpm.ac_power = false; - if (rdev->asic->dpm.enable_bapm) - radeon_dpm_enable_bapm(rdev, rdev->pm.dpm.ac_power); + if (rdev->family == CHIP_ARUBA) { + if (rdev->asic->dpm.enable_bapm) + radeon_dpm_enable_bapm(rdev, rdev->pm.dpm.ac_power); + } mutex_unlock(&rdev->pm.mutex); } else if (rdev->pm.pm_method == PM_METHOD_PROFILE) { if (rdev->pm.profile == PM_PROFILE_AUTO) { @@ -361,6 +363,11 @@ static ssize_t radeon_set_pm_profile(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct radeon_device *rdev = ddev->dev_private; + /* Can't set profile when the card is off */ + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + mutex_lock(&rdev->pm.mutex); if (rdev->pm.pm_method == PM_METHOD_PROFILE) { if (strncmp("default", buf, strlen("default")) == 0) @@ -409,6 +416,13 @@ static ssize_t radeon_set_pm_method(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct radeon_device *rdev = ddev->dev_private; + /* Can't set method when the card is off */ + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) { + count = -EINVAL; + goto fail; + } + /* we don't support the legacy modes with dpm */ if (rdev->pm.pm_method == PM_METHOD_DPM) { count = -EINVAL; @@ -446,6 +460,10 @@ static ssize_t radeon_get_dpm_state(struct device *dev, struct radeon_device *rdev = ddev->dev_private; enum radeon_pm_state_type pm = rdev->pm.dpm.user_state; + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return snprintf(buf, PAGE_SIZE, "off\n"); + return snprintf(buf, PAGE_SIZE, "%s\n", (pm == POWER_STATE_TYPE_BATTERY) ? "battery" : (pm == POWER_STATE_TYPE_BALANCED) ? "balanced" : "performance"); @@ -459,6 +477,11 @@ static ssize_t radeon_set_dpm_state(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct radeon_device *rdev = ddev->dev_private; + /* Can't set dpm state when the card is off */ + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + mutex_lock(&rdev->pm.mutex); if (strncmp("battery", buf, strlen("battery")) == 0) rdev->pm.dpm.user_state = POWER_STATE_TYPE_BATTERY; @@ -485,6 +508,10 @@ static ssize_t radeon_get_dpm_forced_performance_level(struct device *dev, struct radeon_device *rdev = ddev->dev_private; enum radeon_dpm_forced_level level = rdev->pm.dpm.forced_level; + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return snprintf(buf, PAGE_SIZE, "off\n"); + return snprintf(buf, PAGE_SIZE, "%s\n", (level == RADEON_DPM_FORCED_LEVEL_AUTO) ? "auto" : (level == RADEON_DPM_FORCED_LEVEL_LOW) ? "low" : "high"); @@ -500,6 +527,11 @@ static ssize_t radeon_set_dpm_forced_performance_level(struct device *dev, enum radeon_dpm_forced_level level; int ret = 0; + /* Can't force performance level when the card is off */ + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + mutex_lock(&rdev->pm.mutex); if (strncmp("low", buf, strlen("low")) == 0) { level = RADEON_DPM_FORCED_LEVEL_LOW; @@ -538,8 +570,14 @@ static ssize_t radeon_hwmon_show_temp(struct device *dev, char *buf) { struct radeon_device *rdev = dev_get_drvdata(dev); + struct drm_device *ddev = rdev->ddev; int temp; + /* Can't get temperature when the card is off */ + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) + return -EINVAL; + if (rdev->asic->pm.get_temperature) temp = radeon_get_temperature(rdev); else @@ -948,6 +986,8 @@ void radeon_dpm_enable_uvd(struct radeon_device *rdev, bool enable) if (enable) { mutex_lock(&rdev->pm.mutex); rdev->pm.dpm.uvd_active = true; + /* disable this for now */ +#if 0 if ((rdev->pm.dpm.sd == 1) && (rdev->pm.dpm.hd == 0)) dpm_state = POWER_STATE_TYPE_INTERNAL_UVD_SD; else if ((rdev->pm.dpm.sd == 2) && (rdev->pm.dpm.hd == 0)) @@ -957,6 +997,7 @@ void radeon_dpm_enable_uvd(struct radeon_device *rdev, bool enable) else if ((rdev->pm.dpm.sd == 0) && (rdev->pm.dpm.hd == 2)) dpm_state = POWER_STATE_TYPE_INTERNAL_UVD_HD2; else +#endif dpm_state = POWER_STATE_TYPE_INTERNAL_UVD; rdev->pm.dpm.state = dpm_state; mutex_unlock(&rdev->pm.mutex); @@ -1068,7 +1109,6 @@ static void radeon_pm_resume_dpm(struct radeon_device *rdev) if (ret) goto dpm_resume_fail; rdev->pm.dpm_enabled = true; - radeon_pm_compute_clocks(rdev); return; dpm_resume_fail: @@ -1300,6 +1340,7 @@ int radeon_pm_init(struct radeon_device *rdev) case CHIP_KABINI: case CHIP_KAVERI: case CHIP_HAWAII: + case CHIP_MULLINS: /* DPM requires the RLC, RV770+ dGPU requires SMC */ if (!rdev->rlc_fw) rdev->pm.pm_method = PM_METHOD_PROFILE; @@ -1613,8 +1654,12 @@ static int radeon_debugfs_pm_info(struct seq_file *m, void *data) struct drm_info_node *node = (struct drm_info_node *) m->private; struct drm_device *dev = node->minor->dev; struct radeon_device *rdev = dev->dev_private; + struct drm_device *ddev = rdev->ddev; - if (rdev->pm.dpm_enabled) { + if ((rdev->flags & RADEON_IS_PX) && + (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) { + seq_printf(m, "PX asic powered off\n"); + } else if (rdev->pm.dpm_enabled) { mutex_lock(&rdev->pm.mutex); if (rdev->asic->dpm.debugfs_print_current_performance_level) radeon_dpm_debugfs_print_current_performance_level(rdev, m); diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c index b576549fc78..23bb64fd775 100644 --- a/drivers/gpu/drm/radeon/radeon_state.c +++ b/drivers/gpu/drm/radeon/radeon_state.c @@ -3258,4 +3258,4 @@ struct drm_ioctl_desc radeon_ioctls[] = { DRM_IOCTL_DEF_DRV(RADEON_CS, r600_cs_legacy_ioctl, DRM_AUTH) }; -int radeon_max_ioctl = DRM_ARRAY_SIZE(radeon_ioctls); +int radeon_max_ioctl = ARRAY_SIZE(radeon_ioctls); diff --git a/drivers/gpu/drm/radeon/radeon_ucode.h b/drivers/gpu/drm/radeon/radeon_ucode.h index 58d12938c0b..4e7c3269b18 100644 --- a/drivers/gpu/drm/radeon/radeon_ucode.h +++ b/drivers/gpu/drm/radeon/radeon_ucode.h @@ -52,6 +52,7 @@ #define BONAIRE_RLC_UCODE_SIZE 2048 #define KB_RLC_UCODE_SIZE 2560 #define KV_RLC_UCODE_SIZE 2560 +#define ML_RLC_UCODE_SIZE 2560 /* MC */ #define BTC_MC_UCODE_SIZE 6024 diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c index 5748bdaeacc..a4ad270e826 100644 --- a/drivers/gpu/drm/radeon/radeon_uvd.c +++ b/drivers/gpu/drm/radeon/radeon_uvd.c @@ -99,6 +99,7 @@ int radeon_uvd_init(struct radeon_device *rdev) case CHIP_KABINI: case CHIP_KAVERI: case CHIP_HAWAII: + case CHIP_MULLINS: fw_name = FIRMWARE_BONAIRE; break; @@ -465,6 +466,10 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p, cmd = radeon_get_ib_value(p, p->idx) >> 1; if (cmd < 0x4) { + if (end <= start) { + DRM_ERROR("invalid reloc offset %X!\n", offset); + return -EINVAL; + } if ((end - start) < buf_sizes[cmd]) { DRM_ERROR("buffer (%d) to small (%d / %d)!\n", cmd, (unsigned)(end - start), buf_sizes[cmd]); @@ -807,7 +812,8 @@ void radeon_uvd_note_usage(struct radeon_device *rdev) (rdev->pm.dpm.hd != hd)) { rdev->pm.dpm.sd = sd; rdev->pm.dpm.hd = hd; - streams_changed = true; + /* disable this for now */ + /*streams_changed = true;*/ } } diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c index ced53dd03e7..aa21c31a846 100644 --- a/drivers/gpu/drm/radeon/radeon_vce.c +++ b/drivers/gpu/drm/radeon/radeon_vce.c @@ -66,6 +66,8 @@ int radeon_vce_init(struct radeon_device *rdev) case CHIP_BONAIRE: case CHIP_KAVERI: case CHIP_KABINI: + case CHIP_HAWAII: + case CHIP_MULLINS: fw_name = FIRMWARE_BONAIRE; break; @@ -442,13 +444,16 @@ int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring, * @p: parser context * @lo: address of lower dword * @hi: address of higher dword + * @size: size of checker for relocation buffer * * Patch relocation inside command stream with real buffer address */ -int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi) +int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi, + unsigned size) { struct radeon_cs_chunk *relocs_chunk; - uint64_t offset; + struct radeon_cs_reloc *reloc; + uint64_t start, end, offset; unsigned idx; relocs_chunk = &p->chunks[p->chunk_relocs_idx]; @@ -461,15 +466,60 @@ int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi) return -EINVAL; } - offset += p->relocs_ptr[(idx / 4)]->gpu_offset; + reloc = p->relocs_ptr[(idx / 4)]; + start = reloc->gpu_offset; + end = start + radeon_bo_size(reloc->robj); + start += offset; - p->ib.ptr[lo] = offset & 0xFFFFFFFF; - p->ib.ptr[hi] = offset >> 32; + p->ib.ptr[lo] = start & 0xFFFFFFFF; + p->ib.ptr[hi] = start >> 32; + + if (end <= start) { + DRM_ERROR("invalid reloc offset %llX!\n", offset); + return -EINVAL; + } + if ((end - start) < size) { + DRM_ERROR("buffer to small (%d / %d)!\n", + (unsigned)(end - start), size); + return -EINVAL; + } return 0; } /** + * radeon_vce_validate_handle - validate stream handle + * + * @p: parser context + * @handle: handle to validate + * + * Validates the handle and return the found session index or -EINVAL + * we we don't have another free session index. + */ +int radeon_vce_validate_handle(struct radeon_cs_parser *p, uint32_t handle) +{ + unsigned i; + + /* validate the handle */ + for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) { + if (atomic_read(&p->rdev->vce.handles[i]) == handle) + return i; + } + + /* handle not found try to alloc a new one */ + for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) { + if (!atomic_cmpxchg(&p->rdev->vce.handles[i], 0, handle)) { + p->rdev->vce.filp[i] = p->filp; + p->rdev->vce.img_size[i] = 0; + return i; + } + } + + DRM_ERROR("No more free VCE handles!\n"); + return -EINVAL; +} + +/** * radeon_vce_cs_parse - parse and validate the command stream * * @p: parser context @@ -477,8 +527,10 @@ int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi) */ int radeon_vce_cs_parse(struct radeon_cs_parser *p) { - uint32_t handle = 0; - bool destroy = false; + int session_idx = -1; + bool destroyed = false; + uint32_t tmp, handle = 0; + uint32_t *size = &tmp; int i, r; while (p->idx < p->chunks[p->chunk_ib_idx].length_dw) { @@ -490,13 +542,29 @@ int radeon_vce_cs_parse(struct radeon_cs_parser *p) return -EINVAL; } + if (destroyed) { + DRM_ERROR("No other command allowed after destroy!\n"); + return -EINVAL; + } + switch (cmd) { case 0x00000001: // session handle = radeon_get_ib_value(p, p->idx + 2); + session_idx = radeon_vce_validate_handle(p, handle); + if (session_idx < 0) + return session_idx; + size = &p->rdev->vce.img_size[session_idx]; break; case 0x00000002: // task info + break; + case 0x01000001: // create + *size = radeon_get_ib_value(p, p->idx + 8) * + radeon_get_ib_value(p, p->idx + 10) * + 8 * 3 / 2; + break; + case 0x04000001: // config extension case 0x04000002: // pic control case 0x04000005: // rate control @@ -505,23 +573,39 @@ int radeon_vce_cs_parse(struct radeon_cs_parser *p) break; case 0x03000001: // encode - r = radeon_vce_cs_reloc(p, p->idx + 10, p->idx + 9); + r = radeon_vce_cs_reloc(p, p->idx + 10, p->idx + 9, + *size); if (r) return r; - r = radeon_vce_cs_reloc(p, p->idx + 12, p->idx + 11); + r = radeon_vce_cs_reloc(p, p->idx + 12, p->idx + 11, + *size / 3); if (r) return r; break; case 0x02000001: // destroy - destroy = true; + destroyed = true; break; case 0x05000001: // context buffer + r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2, + *size * 2); + if (r) + return r; + break; + case 0x05000004: // video bitstream buffer + tmp = radeon_get_ib_value(p, p->idx + 4); + r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2, + tmp); + if (r) + return r; + break; + case 0x05000005: // feedback buffer - r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2); + r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2, + 4096); if (r) return r; break; @@ -531,33 +615,21 @@ int radeon_vce_cs_parse(struct radeon_cs_parser *p) return -EINVAL; } + if (session_idx == -1) { + DRM_ERROR("no session command at start of IB\n"); + return -EINVAL; + } + p->idx += len / 4; } - if (destroy) { + if (destroyed) { /* IB contains a destroy msg, free the handle */ for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) atomic_cmpxchg(&p->rdev->vce.handles[i], handle, 0); - - return 0; - } - - /* create or encode, validate the handle */ - for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) { - if (atomic_read(&p->rdev->vce.handles[i]) == handle) - return 0; } - /* handle not found try to alloc a new one */ - for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) { - if (!atomic_cmpxchg(&p->rdev->vce.handles[i], 0, handle)) { - p->rdev->vce.filp[i] = p->filp; - return 0; - } - } - - DRM_ERROR("No more free VCE handles!\n"); - return -EINVAL; + return 0; } /** diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c index a128a4fd64b..725d3669014 100644 --- a/drivers/gpu/drm/radeon/radeon_vm.c +++ b/drivers/gpu/drm/radeon/radeon_vm.c @@ -59,7 +59,7 @@ */ static unsigned radeon_vm_num_pdes(struct radeon_device *rdev) { - return rdev->vm_manager.max_pfn >> RADEON_VM_BLOCK_SIZE; + return rdev->vm_manager.max_pfn >> radeon_vm_block_size; } /** @@ -130,18 +130,18 @@ struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev, struct list_head *head) { struct radeon_cs_reloc *list; - unsigned i, idx, size; + unsigned i, idx; - size = (radeon_vm_num_pdes(rdev) + 1) * sizeof(struct radeon_cs_reloc); - list = kmalloc(size, GFP_KERNEL); + list = kmalloc_array(vm->max_pde_used + 2, + sizeof(struct radeon_cs_reloc), GFP_KERNEL); if (!list) return NULL; /* add the vm page table to the list */ list[0].gobj = NULL; list[0].robj = vm->page_directory; - list[0].domain = RADEON_GEM_DOMAIN_VRAM; - list[0].alt_domain = RADEON_GEM_DOMAIN_VRAM; + list[0].prefered_domains = RADEON_GEM_DOMAIN_VRAM; + list[0].allowed_domains = RADEON_GEM_DOMAIN_VRAM; list[0].tv.bo = &vm->page_directory->tbo; list[0].tiling_flags = 0; list[0].handle = 0; @@ -153,8 +153,8 @@ struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev, list[idx].gobj = NULL; list[idx].robj = vm->page_tables[i].bo; - list[idx].domain = RADEON_GEM_DOMAIN_VRAM; - list[idx].alt_domain = RADEON_GEM_DOMAIN_VRAM; + list[idx].prefered_domains = RADEON_GEM_DOMAIN_VRAM; + list[idx].allowed_domains = RADEON_GEM_DOMAIN_VRAM; list[idx].tv.bo = &list[idx].robj->tbo; list[idx].tiling_flags = 0; list[idx].handle = 0; @@ -332,6 +332,7 @@ struct radeon_bo_va *radeon_vm_bo_add(struct radeon_device *rdev, bo_va->ref_count = 1; INIT_LIST_HEAD(&bo_va->bo_list); INIT_LIST_HEAD(&bo_va->vm_list); + INIT_LIST_HEAD(&bo_va->vm_status); mutex_lock(&vm->mutex); list_add(&bo_va->vm_list, &vm->va); @@ -468,14 +469,29 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev, head = &tmp->vm_list; } + if (bo_va->soffset) { + /* add a clone of the bo_va to clear the old address */ + tmp = kzalloc(sizeof(struct radeon_bo_va), GFP_KERNEL); + if (!tmp) { + mutex_unlock(&vm->mutex); + return -ENOMEM; + } + tmp->soffset = bo_va->soffset; + tmp->eoffset = bo_va->eoffset; + tmp->vm = vm; + list_add(&tmp->vm_status, &vm->freed); + } + bo_va->soffset = soffset; bo_va->eoffset = eoffset; bo_va->flags = flags; bo_va->valid = false; list_move(&bo_va->vm_list, head); - soffset = (soffset / RADEON_GPU_PAGE_SIZE) >> RADEON_VM_BLOCK_SIZE; - eoffset = (eoffset / RADEON_GPU_PAGE_SIZE) >> RADEON_VM_BLOCK_SIZE; + soffset = (soffset / RADEON_GPU_PAGE_SIZE) >> radeon_vm_block_size; + eoffset = (eoffset / RADEON_GPU_PAGE_SIZE) >> radeon_vm_block_size; + + BUG_ON(eoffset >= radeon_vm_num_pdes(rdev)); if (eoffset > vm->max_pde_used) vm->max_pde_used = eoffset; @@ -493,7 +509,7 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev, mutex_unlock(&vm->mutex); r = radeon_bo_create(rdev, RADEON_VM_PTE_COUNT * 8, - RADEON_GPU_PAGE_SIZE, false, + RADEON_GPU_PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, NULL, &pt); if (r) return r; @@ -583,9 +599,9 @@ static uint32_t radeon_vm_page_flags(uint32_t flags) int radeon_vm_update_page_directory(struct radeon_device *rdev, struct radeon_vm *vm) { - static const uint32_t incr = RADEON_VM_PTE_COUNT * 8; - - uint64_t pd_addr = radeon_bo_gpu_offset(vm->page_directory); + struct radeon_bo *pd = vm->page_directory; + uint64_t pd_addr = radeon_bo_gpu_offset(pd); + uint32_t incr = RADEON_VM_PTE_COUNT * 8; uint64_t last_pde = ~0, last_pt = ~0; unsigned count = 0, pt_idx, ndw; struct radeon_ib ib; @@ -595,7 +611,7 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev, ndw = 64; /* assume the worst case */ - ndw += vm->max_pde_used * 12; + ndw += vm->max_pde_used * 16; /* update too big for an IB */ if (ndw > 0xfffff) @@ -642,6 +658,7 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev, incr, R600_PTE_VALID); if (ib.length_dw != 0) { + radeon_semaphore_sync_to(ib.semaphore, pd->tbo.sync_obj); radeon_semaphore_sync_to(ib.semaphore, vm->last_id_use); r = radeon_ib_schedule(rdev, &ib, NULL); if (r) { @@ -755,8 +772,7 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev, uint64_t start, uint64_t end, uint64_t dst, uint32_t flags) { - static const uint64_t mask = RADEON_VM_PTE_COUNT - 1; - + uint64_t mask = RADEON_VM_PTE_COUNT - 1; uint64_t last_pte = ~0, last_dst = ~0; unsigned count = 0; uint64_t addr; @@ -766,16 +782,19 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev, /* walk over the address space and update the page tables */ for (addr = start; addr < end; ) { - uint64_t pt_idx = addr >> RADEON_VM_BLOCK_SIZE; + uint64_t pt_idx = addr >> radeon_vm_block_size; + struct radeon_bo *pt = vm->page_tables[pt_idx].bo; unsigned nptes; uint64_t pte; + radeon_semaphore_sync_to(ib->semaphore, pt->tbo.sync_obj); + if ((addr & ~mask) == (end & ~mask)) nptes = end - addr; else nptes = RADEON_VM_PTE_COUNT - (addr & mask); - pte = radeon_bo_gpu_offset(vm->page_tables[pt_idx].bo); + pte = radeon_bo_gpu_offset(pt); pte += (addr & mask) * 8; if ((last_pte + 8 * count) != pte) { @@ -818,25 +837,19 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev, * Object have to be reserved and mutex must be locked! */ int radeon_vm_bo_update(struct radeon_device *rdev, - struct radeon_vm *vm, - struct radeon_bo *bo, + struct radeon_bo_va *bo_va, struct ttm_mem_reg *mem) { + struct radeon_vm *vm = bo_va->vm; struct radeon_ib ib; - struct radeon_bo_va *bo_va; unsigned nptes, ndw; uint64_t addr; int r; - bo_va = radeon_vm_bo_find(vm, bo); - if (bo_va == NULL) { - dev_err(rdev->dev, "bo %p not in vm %p\n", bo, vm); - return -EINVAL; - } if (!bo_va->soffset) { dev_err(rdev->dev, "bo %p don't has a mapping in vm %p\n", - bo, vm); + bo_va->bo, vm); return -EINVAL; } @@ -863,18 +876,18 @@ int radeon_vm_bo_update(struct radeon_device *rdev, trace_radeon_vm_bo_update(bo_va); - nptes = radeon_bo_ngpu_pages(bo); + nptes = (bo_va->eoffset - bo_va->soffset) / RADEON_GPU_PAGE_SIZE; /* padding, etc. */ ndw = 64; - if (RADEON_VM_BLOCK_SIZE > 11) + if (radeon_vm_block_size > 11) /* reserve space for one header for every 2k dwords */ ndw += (nptes >> 11) * 4; else /* reserve space for one header for every (1 << BLOCK_SIZE) entries */ - ndw += (nptes >> RADEON_VM_BLOCK_SIZE) * 4; + ndw += (nptes >> radeon_vm_block_size) * 4; /* reserve space for pte addresses */ ndw += nptes * 2; @@ -906,33 +919,61 @@ int radeon_vm_bo_update(struct radeon_device *rdev, } /** + * radeon_vm_clear_freed - clear freed BOs in the PT + * + * @rdev: radeon_device pointer + * @vm: requested vm + * + * Make sure all freed BOs are cleared in the PT. + * Returns 0 for success. + * + * PTs have to be reserved and mutex must be locked! + */ +int radeon_vm_clear_freed(struct radeon_device *rdev, + struct radeon_vm *vm) +{ + struct radeon_bo_va *bo_va, *tmp; + int r; + + list_for_each_entry_safe(bo_va, tmp, &vm->freed, vm_status) { + list_del(&bo_va->vm_status); + r = radeon_vm_bo_update(rdev, bo_va, NULL); + kfree(bo_va); + if (r) + return r; + } + return 0; + +} + +/** * radeon_vm_bo_rmv - remove a bo to a specific vm * * @rdev: radeon_device pointer * @bo_va: requested bo_va * * Remove @bo_va->bo from the requested vm (cayman+). - * Remove @bo_va->bo from the list of bos associated with the bo_va->vm and - * remove the ptes for @bo_va in the page table. - * Returns 0 for success. * * Object have to be reserved! */ -int radeon_vm_bo_rmv(struct radeon_device *rdev, - struct radeon_bo_va *bo_va) +void radeon_vm_bo_rmv(struct radeon_device *rdev, + struct radeon_bo_va *bo_va) { - int r = 0; + struct radeon_vm *vm = bo_va->vm; - mutex_lock(&bo_va->vm->mutex); - if (bo_va->soffset) - r = radeon_vm_bo_update(rdev, bo_va->vm, bo_va->bo, NULL); + list_del(&bo_va->bo_list); + mutex_lock(&vm->mutex); list_del(&bo_va->vm_list); - mutex_unlock(&bo_va->vm->mutex); - list_del(&bo_va->bo_list); - kfree(bo_va); - return r; + if (bo_va->soffset) { + bo_va->bo = NULL; + list_add(&bo_va->vm_status, &vm->freed); + } else { + kfree(bo_va); + } + + mutex_unlock(&vm->mutex); } /** @@ -970,11 +1011,13 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm) int r; vm->id = 0; + vm->ib_bo_va = NULL; vm->fence = NULL; vm->last_flush = NULL; vm->last_id_use = NULL; mutex_init(&vm->mutex); INIT_LIST_HEAD(&vm->va); + INIT_LIST_HEAD(&vm->freed); pd_size = radeon_vm_directory_size(rdev); pd_entries = radeon_vm_num_pdes(rdev); @@ -987,7 +1030,7 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm) return -ENOMEM; } - r = radeon_bo_create(rdev, pd_size, align, false, + r = radeon_bo_create(rdev, pd_size, align, true, RADEON_GEM_DOMAIN_VRAM, NULL, &vm->page_directory); if (r) @@ -1029,7 +1072,8 @@ void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm) kfree(bo_va); } } - + list_for_each_entry_safe(bo_va, tmp, &vm->freed, vm_status) + kfree(bo_va); for (i = 0; i < radeon_vm_num_pdes(rdev); i++) radeon_bo_unref(&vm->page_tables[i].bo); diff --git a/drivers/gpu/drm/radeon/rs400.c b/drivers/gpu/drm/radeon/rs400.c index 130d5cc50d4..a0f96decece 100644 --- a/drivers/gpu/drm/radeon/rs400.c +++ b/drivers/gpu/drm/radeon/rs400.c @@ -212,21 +212,16 @@ void rs400_gart_fini(struct radeon_device *rdev) #define RS400_PTE_WRITEABLE (1 << 2) #define RS400_PTE_READABLE (1 << 3) -int rs400_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr) +void rs400_gart_set_page(struct radeon_device *rdev, unsigned i, uint64_t addr) { uint32_t entry; u32 *gtt = rdev->gart.ptr; - if (i < 0 || i > rdev->gart.num_gpu_pages) { - return -EINVAL; - } - entry = (lower_32_bits(addr) & PAGE_MASK) | ((upper_32_bits(addr) & 0xff) << 4) | RS400_PTE_WRITEABLE | RS400_PTE_READABLE; entry = cpu_to_le32(entry); gtt[i] = entry; - return 0; } int rs400_mc_wait_for_idle(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c index 0a8be63926d..d1a35cb1c91 100644 --- a/drivers/gpu/drm/radeon/rs600.c +++ b/drivers/gpu/drm/radeon/rs600.c @@ -626,17 +626,16 @@ static void rs600_gart_fini(struct radeon_device *rdev) radeon_gart_table_vram_free(rdev); } -int rs600_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr) +void rs600_gart_set_page(struct radeon_device *rdev, unsigned i, uint64_t addr) { void __iomem *ptr = (void *)rdev->gart.ptr; - if (i < 0 || i > rdev->gart.num_gpu_pages) { - return -EINVAL; - } addr = addr & 0xFFFFFFFFFFFFF000ULL; - addr |= R600_PTE_GART; + if (addr == rdev->dummy_page.addr) + addr |= R600_PTE_SYSTEM | R600_PTE_SNOOPED; + else + addr |= R600_PTE_GART; writeq(addr, ptr + (i * 8)); - return 0; } int rs600_irq_set(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/rv515.c b/drivers/gpu/drm/radeon/rv515.c index 237dd29d9f1..3e21e869015 100644 --- a/drivers/gpu/drm/radeon/rv515.c +++ b/drivers/gpu/drm/radeon/rv515.c @@ -406,8 +406,9 @@ void rv515_mc_resume(struct radeon_device *rdev, struct rv515_mc_save *save) for (i = 0; i < rdev->num_crtc; i++) { if (save->crtc_enabled[i]) { tmp = RREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + crtc_offsets[i]); - if ((tmp & 0x3) != 0) { - tmp &= ~0x3; + if ((tmp & 0x7) != 3) { + tmp &= ~0x7; + tmp |= 0x3; WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + crtc_offsets[i], tmp); } tmp = RREG32(AVIVO_D1GRPH_UPDATE + crtc_offsets[i]); diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index 97b776666b7..da8703d8d45 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -1327,6 +1327,9 @@ static void rv770_gpu_init(struct radeon_device *rdev) if (tmp < rdev->config.rv770.max_simds) { rdev->config.rv770.max_simds = tmp; } + tmp = rdev->config.rv770.max_simds - + r600_count_pipe_bits((cc_gc_shader_pipe_config >> 16) & R7XX_MAX_SIMDS_MASK); + rdev->config.rv770.active_simds = tmp; switch (rdev->config.rv770.max_tile_pipes) { case 1: diff --git a/drivers/gpu/drm/radeon/rv770_dma.c b/drivers/gpu/drm/radeon/rv770_dma.c index aca8cbe8a33..bbf2e076ee4 100644 --- a/drivers/gpu/drm/radeon/rv770_dma.c +++ b/drivers/gpu/drm/radeon/rv770_dma.c @@ -86,6 +86,7 @@ int rv770_copy_dma(struct radeon_device *rdev, r = radeon_fence_emit(rdev, fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); + radeon_semaphore_free(rdev, &sem, NULL); return r; } diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c index da041a43d82..3c76e1dcdf0 100644 --- a/drivers/gpu/drm/radeon/rv770_dpm.c +++ b/drivers/gpu/drm/radeon/rv770_dpm.c @@ -2329,12 +2329,6 @@ void rv770_get_engine_memory_ss(struct radeon_device *rdev) pi->mclk_ss = radeon_atombios_get_asic_ss_info(rdev, &ss, ASIC_INTERNAL_MEMORY_SS, 0); - /* disable ss, causes hangs on some cayman boards */ - if (rdev->family == CHIP_CAYMAN) { - pi->sclk_ss = false; - pi->mclk_ss = false; - } - if (pi->sclk_ss || pi->mclk_ss) pi->dynamic_ss = true; else diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c index 5c1c0c795e9..9e854fd016d 100644 --- a/drivers/gpu/drm/radeon/si.c +++ b/drivers/gpu/drm/radeon/si.c @@ -71,6 +71,7 @@ MODULE_FIRMWARE("radeon/HAINAN_mc2.bin"); MODULE_FIRMWARE("radeon/HAINAN_rlc.bin"); MODULE_FIRMWARE("radeon/HAINAN_smc.bin"); +static u32 si_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh); static void si_pcie_gen3_enable(struct radeon_device *rdev); static void si_program_aspm(struct radeon_device *rdev); extern void sumo_rlc_fini(struct radeon_device *rdev); @@ -2900,7 +2901,7 @@ static void si_gpu_init(struct radeon_device *rdev) u32 sx_debug_1; u32 hdp_host_path_cntl; u32 tmp; - int i, j; + int i, j, k; switch (rdev->family) { case CHIP_TAHITI: @@ -3098,6 +3099,14 @@ static void si_gpu_init(struct radeon_device *rdev) rdev->config.si.max_sh_per_se, rdev->config.si.max_cu_per_sh); + for (i = 0; i < rdev->config.si.max_shader_engines; i++) { + for (j = 0; j < rdev->config.si.max_sh_per_se; j++) { + for (k = 0; k < rdev->config.si.max_cu_per_sh; k++) { + rdev->config.si.active_cus += + hweight32(si_get_cu_active_bitmap(rdev, i, j)); + } + } + } /* set HW defaults for 3D engine */ WREG32(CP_QUEUE_THRESHOLDS, (ROQ_IB1_START(0x16) | @@ -3186,7 +3195,7 @@ void si_fence_ring_emit(struct radeon_device *rdev, /* EVENT_WRITE_EOP - flush caches, send int */ radeon_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4)); radeon_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | EVENT_INDEX(5)); - radeon_ring_write(ring, addr & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(addr)); radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | DATA_SEL(1) | INT_SEL(2)); radeon_ring_write(ring, fence->seq); radeon_ring_write(ring, 0); @@ -3219,7 +3228,7 @@ void si_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib) radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); radeon_ring_write(ring, (1 << 8)); radeon_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); + radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr)); radeon_ring_write(ring, next_rptr); } @@ -4095,7 +4104,7 @@ static int si_pcie_gart_enable(struct radeon_device *rdev) (u32)(rdev->dummy_page.addr >> 12)); WREG32(VM_CONTEXT1_CNTL2, 4); WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) | - PAGE_TABLE_BLOCK_SIZE(RADEON_VM_BLOCK_SIZE - 9) | + PAGE_TABLE_BLOCK_SIZE(radeon_vm_block_size - 9) | RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT | RANGE_PROTECTION_FAULT_ENABLE_DEFAULT | DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT | @@ -5784,7 +5793,6 @@ int si_irq_set(struct radeon_device *rdev) u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0; u32 hpd1 = 0, hpd2 = 0, hpd3 = 0, hpd4 = 0, hpd5 = 0, hpd6 = 0; u32 grbm_int_cntl = 0; - u32 grph1 = 0, grph2 = 0, grph3 = 0, grph4 = 0, grph5 = 0, grph6 = 0; u32 dma_cntl, dma_cntl1; u32 thermal_int = 0; @@ -5923,16 +5931,22 @@ int si_irq_set(struct radeon_device *rdev) } if (rdev->num_crtc >= 2) { - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, grph1); - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, grph2); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); } if (rdev->num_crtc >= 4) { - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, grph3); - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, grph4); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); } if (rdev->num_crtc >= 6) { - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, grph5); - WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, grph6); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); + WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, + GRPH_PFLIP_INT_MASK); } if (!ASIC_IS_NODCE(rdev)) { @@ -6089,6 +6103,7 @@ static inline u32 si_get_ih_wptr(struct radeon_device *rdev) tmp = RREG32(IH_RB_CNTL); tmp |= IH_WPTR_OVERFLOW_CLEAR; WREG32(IH_RB_CNTL, tmp); + wptr &= ~RB_OVERFLOW; } return (wptr & rdev->ih.ptr_mask); } @@ -6296,6 +6311,15 @@ restart_ih: break; } break; + case 8: /* D1 page flip */ + case 10: /* D2 page flip */ + case 12: /* D3 page flip */ + case 14: /* D4 page flip */ + case 16: /* D5 page flip */ + case 18: /* D6 page flip */ + DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1); + radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1); + break; case 42: /* HPD hotplug */ switch (src_data) { case 0: @@ -6353,14 +6377,16 @@ restart_ih: case 147: addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR); status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS); + /* reset addr and status */ + WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1); + if (addr == 0x0 && status == 0x0) + break; dev_err(rdev->dev, "GPU fault detected: %d 0x%08x\n", src_id, src_data); dev_err(rdev->dev, " VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x%08X\n", addr); dev_err(rdev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n", status); si_vm_decode_fault(rdev, status, addr); - /* reset addr and status */ - WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1); break; case 176: /* RINGID0 CP_INT */ radeon_fence_process(rdev, RADEON_RING_TYPE_GFX_INDEX); diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c index 952166930fb..e24c94b6d14 100644 --- a/drivers/gpu/drm/radeon/si_dma.c +++ b/drivers/gpu/drm/radeon/si_dma.c @@ -88,8 +88,8 @@ void si_dma_vm_set_page(struct radeon_device *rdev, ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY, 1, 0, 0, bytes); - ib->ptr[ib->length_dw++] = pe & 0xffffffff; - ib->ptr[ib->length_dw++] = src & 0xffffffff; + ib->ptr[ib->length_dw++] = lower_32_bits(pe); + ib->ptr[ib->length_dw++] = lower_32_bits(src); ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff; ib->ptr[ib->length_dw++] = upper_32_bits(src) & 0xff; @@ -220,8 +220,8 @@ int si_copy_dma(struct radeon_device *rdev, cur_size_in_bytes = 0xFFFFF; size_in_bytes -= cur_size_in_bytes; radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_COPY, 1, 0, 0, cur_size_in_bytes)); - radeon_ring_write(ring, dst_offset & 0xffffffff); - radeon_ring_write(ring, src_offset & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(dst_offset)); + radeon_ring_write(ring, lower_32_bits(src_offset)); radeon_ring_write(ring, upper_32_bits(dst_offset) & 0xff); radeon_ring_write(ring, upper_32_bits(src_offset) & 0xff); src_offset += cur_size_in_bytes; @@ -231,6 +231,7 @@ int si_copy_dma(struct radeon_device *rdev, r = radeon_fence_emit(rdev, fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); + radeon_semaphore_free(rdev, &sem, NULL); return r; } diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c index 9a3567bedaa..58918868f89 100644 --- a/drivers/gpu/drm/radeon/si_dpm.c +++ b/drivers/gpu/drm/radeon/si_dpm.c @@ -1948,6 +1948,10 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev) si_pi->cac_weights = cac_weights_cape_verde_pro; si_pi->dte_data = dte_data_cape_verde; break; + case 0x682C: + si_pi->cac_weights = cac_weights_cape_verde_pro; + si_pi->dte_data = dte_data_sun_xt; + break; case 0x6825: case 0x6827: si_pi->cac_weights = cac_weights_heathrow; @@ -1971,10 +1975,9 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev) si_pi->dte_data = dte_data_venus_xt; break; case 0x6823: - si_pi->cac_weights = cac_weights_chelsea_pro; - si_pi->dte_data = dte_data_venus_pro; - break; case 0x682B: + case 0x6822: + case 0x682A: si_pi->cac_weights = cac_weights_chelsea_pro; si_pi->dte_data = dte_data_venus_pro; break; @@ -1988,6 +1991,7 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev) case 0x6601: case 0x6621: case 0x6603: + case 0x6605: si_pi->cac_weights = cac_weights_mars_pro; si_pi->lcac_config = lcac_mars_pro; si_pi->cac_override = cac_override_oland; @@ -1998,6 +2002,7 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev) case 0x6600: case 0x6606: case 0x6620: + case 0x6604: si_pi->cac_weights = cac_weights_mars_xt; si_pi->lcac_config = lcac_mars_pro; si_pi->cac_override = cac_override_oland; @@ -2006,6 +2011,8 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev) update_dte_from_pl2 = true; break; case 0x6611: + case 0x6613: + case 0x6608: si_pi->cac_weights = cac_weights_oland_pro; si_pi->lcac_config = lcac_mars_pro; si_pi->cac_override = cac_override_oland; diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h index da8f8674a55..fd414d34d88 100644 --- a/drivers/gpu/drm/radeon/sid.h +++ b/drivers/gpu/drm/radeon/sid.h @@ -107,8 +107,8 @@ #define SPLL_CHG_STATUS (1 << 1) #define SPLL_CNTL_MODE 0x618 #define SPLL_SW_DIR_CONTROL (1 << 0) -# define SPLL_REFCLK_SEL(x) ((x) << 8) -# define SPLL_REFCLK_SEL_MASK 0xFF00 +# define SPLL_REFCLK_SEL(x) ((x) << 26) +# define SPLL_REFCLK_SEL_MASK (3 << 26) #define CG_SPLL_SPREAD_SPECTRUM 0x620 #define SSEN (1 << 0) diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c index 2a2822c0332..32e50be9c4a 100644 --- a/drivers/gpu/drm/radeon/trinity_dpm.c +++ b/drivers/gpu/drm/radeon/trinity_dpm.c @@ -1874,7 +1874,16 @@ int trinity_dpm_init(struct radeon_device *rdev) for (i = 0; i < SUMO_MAX_HARDWARE_POWERLEVELS; i++) pi->at[i] = TRINITY_AT_DFLT; - pi->enable_bapm = false; + /* There are stability issues reported on with + * bapm enabled when switching between AC and battery + * power. At the same time, some MSI boards hang + * if it's not enabled and dpm is enabled. Just enable + * it for MSI boards right now. + */ + if (rdev->pdev->subsystem_vendor == 0x1462) + pi->enable_bapm = true; + else + pi->enable_bapm = false; pi->enable_nbps_policy = true; pi->enable_sclk_ds = true; pi->enable_gfx_power_gating = true; diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c index 0a243f0e5d6..be42c812520 100644 --- a/drivers/gpu/drm/radeon/uvd_v1_0.c +++ b/drivers/gpu/drm/radeon/uvd_v1_0.c @@ -83,7 +83,10 @@ int uvd_v1_0_init(struct radeon_device *rdev) int r; /* raise clocks while booting up the VCPU */ - radeon_set_uvd_clocks(rdev, 53300, 40000); + if (rdev->family < CHIP_RV740) + radeon_set_uvd_clocks(rdev, 10000, 10000); + else + radeon_set_uvd_clocks(rdev, 53300, 40000); r = uvd_v1_0_start(rdev); if (r) @@ -407,7 +410,10 @@ int uvd_v1_0_ib_test(struct radeon_device *rdev, struct radeon_ring *ring) struct radeon_fence *fence = NULL; int r; - r = radeon_set_uvd_clocks(rdev, 53300, 40000); + if (rdev->family < CHIP_RV740) + r = radeon_set_uvd_clocks(rdev, 10000, 10000); + else + r = radeon_set_uvd_clocks(rdev, 53300, 40000); if (r) { DRM_ERROR("radeon: failed to raise UVD clocks (%d).\n", r); return r; diff --git a/drivers/gpu/drm/radeon/uvd_v2_2.c b/drivers/gpu/drm/radeon/uvd_v2_2.c index d1771004cb5..8bfdadd5659 100644 --- a/drivers/gpu/drm/radeon/uvd_v2_2.c +++ b/drivers/gpu/drm/radeon/uvd_v2_2.c @@ -45,7 +45,7 @@ void uvd_v2_2_fence_emit(struct radeon_device *rdev, radeon_ring_write(ring, PACKET0(UVD_CONTEXT_ID, 0)); radeon_ring_write(ring, fence->seq); radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0)); - radeon_ring_write(ring, addr & 0xffffffff); + radeon_ring_write(ring, lower_32_bits(addr)); radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0)); radeon_ring_write(ring, upper_32_bits(addr) & 0xff); radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0)); |