diff options
Diffstat (limited to 'drivers/gpu/drm/radeon')
63 files changed, 3835 insertions, 623 deletions
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index e47eecf..3896ef8 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -36,6 +36,9 @@ $(obj)/r600_reg_safe.h: $(src)/reg_srcs/r600 $(obj)/mkregtable $(obj)/evergreen_reg_safe.h: $(src)/reg_srcs/evergreen $(obj)/mkregtable $(call if_changed,mkregtable) +$(obj)/cayman_reg_safe.h: $(src)/reg_srcs/cayman $(obj)/mkregtable + $(call if_changed,mkregtable) + $(obj)/r100.o: $(obj)/r100_reg_safe.h $(obj)/rn50_reg_safe.h $(obj)/r200.o: $(obj)/r200_reg_safe.h @@ -50,7 +53,7 @@ $(obj)/rs600.o: $(obj)/rs600_reg_safe.h $(obj)/r600_cs.o: $(obj)/r600_reg_safe.h -$(obj)/evergreen_cs.o: $(obj)/evergreen_reg_safe.h +$(obj)/evergreen_cs.o: $(obj)/evergreen_reg_safe.h $(obj)/cayman_reg_safe.h radeon-y := radeon_drv.o radeon_cp.o radeon_state.o radeon_mem.o \ radeon_irq.o r300_cmdbuf.o r600_cp.o @@ -66,7 +69,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ r200.o radeon_legacy_tv.o r600_cs.o r600_blit.o r600_blit_shaders.o \ r600_blit_kms.o radeon_pm.o atombios_dp.o r600_audio.o r600_hdmi.o \ evergreen.o evergreen_cs.o evergreen_blit_shaders.o evergreen_blit_kms.o \ - radeon_trace_points.o ni.o + radeon_trace_points.o ni.o cayman_blit_shaders.o radeon-$(CONFIG_COMPAT) += radeon_ioc32.o radeon-$(CONFIG_VGA_SWITCHEROO) += radeon_atpx_handler.o diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index b153700..3cd3234 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -48,29 +48,29 @@ static void atombios_overscan_setup(struct drm_crtc *crtc, switch (radeon_crtc->rmx_type) { case RMX_CENTER: - args.usOverscanTop = (adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2; - args.usOverscanBottom = (adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2; - args.usOverscanLeft = (adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2; - args.usOverscanRight = (adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2; + args.usOverscanTop = cpu_to_le16((adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2); + args.usOverscanBottom = cpu_to_le16((adjusted_mode->crtc_vdisplay - mode->crtc_vdisplay) / 2); + args.usOverscanLeft = cpu_to_le16((adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2); + args.usOverscanRight = cpu_to_le16((adjusted_mode->crtc_hdisplay - mode->crtc_hdisplay) / 2); break; case RMX_ASPECT: a1 = mode->crtc_vdisplay * adjusted_mode->crtc_hdisplay; a2 = adjusted_mode->crtc_vdisplay * mode->crtc_hdisplay; if (a1 > a2) { - args.usOverscanLeft = (adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2; - args.usOverscanRight = (adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2; + args.usOverscanLeft = cpu_to_le16((adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2); + args.usOverscanRight = cpu_to_le16((adjusted_mode->crtc_hdisplay - (a2 / mode->crtc_vdisplay)) / 2); } else if (a2 > a1) { - args.usOverscanLeft = (adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2; - args.usOverscanRight = (adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2; + args.usOverscanTop = cpu_to_le16((adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2); + args.usOverscanBottom = cpu_to_le16((adjusted_mode->crtc_vdisplay - (a1 / mode->crtc_hdisplay)) / 2); } break; case RMX_FULL: default: - args.usOverscanRight = radeon_crtc->h_border; - args.usOverscanLeft = radeon_crtc->h_border; - args.usOverscanBottom = radeon_crtc->v_border; - args.usOverscanTop = radeon_crtc->v_border; + args.usOverscanRight = cpu_to_le16(radeon_crtc->h_border); + args.usOverscanLeft = cpu_to_le16(radeon_crtc->h_border); + args.usOverscanBottom = cpu_to_le16(radeon_crtc->v_border); + args.usOverscanTop = cpu_to_le16(radeon_crtc->v_border); break; } atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); @@ -419,23 +419,23 @@ static void atombios_crtc_program_ss(struct drm_crtc *crtc, memset(&args, 0, sizeof(args)); if (ASIC_IS_DCE5(rdev)) { - args.v3.usSpreadSpectrumAmountFrac = 0; + args.v3.usSpreadSpectrumAmountFrac = cpu_to_le16(0); args.v3.ucSpreadSpectrumType = ss->type; switch (pll_id) { case ATOM_PPLL1: args.v3.ucSpreadSpectrumType |= ATOM_PPLL_SS_TYPE_V3_P1PLL; - args.v3.usSpreadSpectrumAmount = ss->amount; - args.v3.usSpreadSpectrumStep = ss->step; + args.v3.usSpreadSpectrumAmount = cpu_to_le16(ss->amount); + args.v3.usSpreadSpectrumStep = cpu_to_le16(ss->step); break; case ATOM_PPLL2: args.v3.ucSpreadSpectrumType |= ATOM_PPLL_SS_TYPE_V3_P2PLL; - args.v3.usSpreadSpectrumAmount = ss->amount; - args.v3.usSpreadSpectrumStep = ss->step; + args.v3.usSpreadSpectrumAmount = cpu_to_le16(ss->amount); + args.v3.usSpreadSpectrumStep = cpu_to_le16(ss->step); break; case ATOM_DCPLL: args.v3.ucSpreadSpectrumType |= ATOM_PPLL_SS_TYPE_V3_DCPLL; - args.v3.usSpreadSpectrumAmount = 0; - args.v3.usSpreadSpectrumStep = 0; + args.v3.usSpreadSpectrumAmount = cpu_to_le16(0); + args.v3.usSpreadSpectrumStep = cpu_to_le16(0); break; case ATOM_PPLL_INVALID: return; @@ -447,18 +447,18 @@ static void atombios_crtc_program_ss(struct drm_crtc *crtc, switch (pll_id) { case ATOM_PPLL1: args.v2.ucSpreadSpectrumType |= ATOM_PPLL_SS_TYPE_V2_P1PLL; - args.v2.usSpreadSpectrumAmount = ss->amount; - args.v2.usSpreadSpectrumStep = ss->step; + args.v2.usSpreadSpectrumAmount = cpu_to_le16(ss->amount); + args.v2.usSpreadSpectrumStep = cpu_to_le16(ss->step); break; case ATOM_PPLL2: args.v2.ucSpreadSpectrumType |= ATOM_PPLL_SS_TYPE_V2_P2PLL; - args.v2.usSpreadSpectrumAmount = ss->amount; - args.v2.usSpreadSpectrumStep = ss->step; + args.v2.usSpreadSpectrumAmount = cpu_to_le16(ss->amount); + args.v2.usSpreadSpectrumStep = cpu_to_le16(ss->step); break; case ATOM_DCPLL: args.v2.ucSpreadSpectrumType |= ATOM_PPLL_SS_TYPE_V2_DCPLL; - args.v2.usSpreadSpectrumAmount = 0; - args.v2.usSpreadSpectrumStep = 0; + args.v2.usSpreadSpectrumAmount = cpu_to_le16(0); + args.v2.usSpreadSpectrumStep = cpu_to_le16(0); break; case ATOM_PPLL_INVALID: return; @@ -538,7 +538,6 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, pll->flags |= RADEON_PLL_PREFER_HIGH_FB_DIV; else pll->flags |= RADEON_PLL_PREFER_LOW_REF_DIV; - } list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { @@ -555,29 +554,28 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, dp_clock = dig_connector->dp_clock; } } -/* this might work properly with the new pll algo */ -#if 0 /* doesn't work properly on some laptops */ + /* use recommended ref_div for ss */ if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) { if (ss_enabled) { if (ss->refdiv) { + pll->flags |= RADEON_PLL_PREFER_MINM_OVER_MAXP; pll->flags |= RADEON_PLL_USE_REF_DIV; pll->reference_div = ss->refdiv; + if (ASIC_IS_AVIVO(rdev)) + pll->flags |= RADEON_PLL_USE_FRAC_FB_DIV; } } } -#endif + if (ASIC_IS_AVIVO(rdev)) { /* DVO wants 2x pixel clock if the DVO chip is in 12 bit mode */ if (radeon_encoder->encoder_id == ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1) adjusted_clock = mode->clock * 2; if (radeon_encoder->active_device & (ATOM_DEVICE_TV_SUPPORT)) pll->flags |= RADEON_PLL_PREFER_CLOSEST_LOWER; - /* rv515 needs more testing with this option */ - if (rdev->family != CHIP_RV515) { - if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) - pll->flags |= RADEON_PLL_IS_LCD; - } + if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) + pll->flags |= RADEON_PLL_IS_LCD; } else { if (encoder->encoder_type != DRM_MODE_ENCODER_DAC) pll->flags |= RADEON_PLL_NO_ODD_POST_DIV; @@ -664,10 +662,12 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, index, (uint32_t *)&args); adjusted_clock = le32_to_cpu(args.v3.sOutput.ulDispPllFreq) * 10; if (args.v3.sOutput.ucRefDiv) { + pll->flags |= RADEON_PLL_USE_FRAC_FB_DIV; pll->flags |= RADEON_PLL_USE_REF_DIV; pll->reference_div = args.v3.sOutput.ucRefDiv; } if (args.v3.sOutput.ucPostDiv) { + pll->flags |= RADEON_PLL_USE_FRAC_FB_DIV; pll->flags |= RADEON_PLL_USE_POST_DIV; pll->post_div = args.v3.sOutput.ucPostDiv; } @@ -721,14 +721,14 @@ static void atombios_crtc_set_dcpll(struct drm_crtc *crtc, * SetPixelClock provides the dividers */ args.v5.ucCRTC = ATOM_CRTC_INVALID; - args.v5.usPixelClock = dispclk; + args.v5.usPixelClock = cpu_to_le16(dispclk); args.v5.ucPpll = ATOM_DCPLL; break; case 6: /* if the default dcpll clock is specified, * SetPixelClock provides the dividers */ - args.v6.ulDispEngClkFreq = dispclk; + args.v6.ulDispEngClkFreq = cpu_to_le32(dispclk); args.v6.ucPpll = ATOM_DCPLL; break; default: @@ -957,11 +957,7 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode /* adjust pixel clock as needed */ adjusted_clock = atombios_adjust_pll(crtc, mode, pll, ss_enabled, &ss); - /* rv515 seems happier with the old algo */ - if (rdev->family == CHIP_RV515) - radeon_compute_pll_legacy(pll, adjusted_clock, &pll_clock, &fb_div, &frac_fb_div, - &ref_div, &post_div); - else if (ASIC_IS_AVIVO(rdev)) + if (ASIC_IS_AVIVO(rdev)) radeon_compute_pll_avivo(pll, adjusted_clock, &pll_clock, &fb_div, &frac_fb_div, &ref_div, &post_div); else @@ -995,9 +991,9 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode } } -static int evergreen_crtc_do_set_base(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - int x, int y, int atomic) +static int dce4_crtc_do_set_base(struct drm_crtc *crtc, + struct drm_framebuffer *fb, + int x, int y, int atomic) { struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); struct drm_device *dev = crtc->dev; @@ -1030,7 +1026,7 @@ static int evergreen_crtc_do_set_base(struct drm_crtc *crtc, * just update base pointers */ obj = radeon_fb->obj; - rbo = obj->driver_private; + rbo = gem_to_radeon_bo(obj); r = radeon_bo_reserve(rbo, false); if (unlikely(r != 0)) return r; @@ -1137,15 +1133,9 @@ static int evergreen_crtc_do_set_base(struct drm_crtc *crtc, WREG32(EVERGREEN_VIEWPORT_SIZE + radeon_crtc->crtc_offset, (crtc->mode.hdisplay << 16) | crtc->mode.vdisplay); - if (crtc->mode.flags & DRM_MODE_FLAG_INTERLACE) - WREG32(EVERGREEN_DATA_FORMAT + radeon_crtc->crtc_offset, - EVERGREEN_INTERLEAVE_EN); - else - WREG32(EVERGREEN_DATA_FORMAT + radeon_crtc->crtc_offset, 0); - if (!atomic && fb && fb != crtc->fb) { radeon_fb = to_radeon_framebuffer(fb); - rbo = radeon_fb->obj->driver_private; + rbo = gem_to_radeon_bo(radeon_fb->obj); r = radeon_bo_reserve(rbo, false); if (unlikely(r != 0)) return r; @@ -1191,7 +1181,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, } obj = radeon_fb->obj; - rbo = obj->driver_private; + rbo = gem_to_radeon_bo(obj); r = radeon_bo_reserve(rbo, false); if (unlikely(r != 0)) return r; @@ -1300,15 +1290,9 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc, WREG32(AVIVO_D1MODE_VIEWPORT_SIZE + radeon_crtc->crtc_offset, (crtc->mode.hdisplay << 16) | crtc->mode.vdisplay); - if (crtc->mode.flags & DRM_MODE_FLAG_INTERLACE) - WREG32(AVIVO_D1MODE_DATA_FORMAT + radeon_crtc->crtc_offset, - AVIVO_D1MODE_INTERLEAVE_EN); - else - WREG32(AVIVO_D1MODE_DATA_FORMAT + radeon_crtc->crtc_offset, 0); - if (!atomic && fb && fb != crtc->fb) { radeon_fb = to_radeon_framebuffer(fb); - rbo = radeon_fb->obj->driver_private; + rbo = gem_to_radeon_bo(radeon_fb->obj); r = radeon_bo_reserve(rbo, false); if (unlikely(r != 0)) return r; @@ -1329,7 +1313,7 @@ int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y, struct radeon_device *rdev = dev->dev_private; if (ASIC_IS_DCE4(rdev)) - return evergreen_crtc_do_set_base(crtc, old_fb, x, y, 0); + return dce4_crtc_do_set_base(crtc, old_fb, x, y, 0); else if (ASIC_IS_AVIVO(rdev)) return avivo_crtc_do_set_base(crtc, old_fb, x, y, 0); else @@ -1344,7 +1328,7 @@ int atombios_crtc_set_base_atomic(struct drm_crtc *crtc, struct radeon_device *rdev = dev->dev_private; if (ASIC_IS_DCE4(rdev)) - return evergreen_crtc_do_set_base(crtc, fb, x, y, 1); + return dce4_crtc_do_set_base(crtc, fb, x, y, 1); else if (ASIC_IS_AVIVO(rdev)) return avivo_crtc_do_set_base(crtc, fb, x, y, 1); else diff --git a/drivers/gpu/drm/radeon/cayman_blit_shaders.c b/drivers/gpu/drm/radeon/cayman_blit_shaders.c new file mode 100644 index 0000000..e148ab0 --- /dev/null +++ b/drivers/gpu/drm/radeon/cayman_blit_shaders.c @@ -0,0 +1,55 @@ +/* + * Copyright 2010 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + * Authors: + * Alex Deucher <alexander.deucher@amd.com> + */ + +#include <linux/types.h> +#include <linux/kernel.h> + +/* + * evergreen cards need to use the 3D engine to blit data which requires + * quite a bit of hw state setup. Rather than pull the whole 3D driver + * (which normally generates the 3D state) into the DRM, we opt to use + * statically generated state tables. The regsiter state and shaders + * were hand generated to support blitting functionality. See the 3D + * driver or documentation for descriptions of the registers and + * shader instructions. + */ + +const u32 cayman_default_state[] = +{ + /* XXX fill in additional blit state */ + + 0xc0026900, + 0x00000316, + 0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */ + 0x00000010, /* */ + + 0xc0026900, + 0x000000d9, + 0x00000000, /* CP_RINGID */ + 0x00000000, /* CP_VMID */ +}; + +const u32 cayman_default_size = ARRAY_SIZE(cayman_default_state); diff --git a/drivers/gpu/drm/radeon/cayman_blit_shaders.h b/drivers/gpu/drm/radeon/cayman_blit_shaders.h new file mode 100644 index 0000000..33b75e5d --- /dev/null +++ b/drivers/gpu/drm/radeon/cayman_blit_shaders.h @@ -0,0 +1,32 @@ +/* + * Copyright 2010 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef CAYMAN_BLIT_SHADERS_H +#define CAYMAN_BLIT_SHADERS_H + +extern const u32 cayman_default_state[]; + +extern const u32 cayman_default_size; + +#endif diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 9e0760b..941080a 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -804,7 +804,7 @@ void evergreen_bandwidth_update(struct radeon_device *rdev) } } -static int evergreen_mc_wait_for_idle(struct radeon_device *rdev) +int evergreen_mc_wait_for_idle(struct radeon_device *rdev) { unsigned i; u32 tmp; @@ -957,7 +957,7 @@ void evergreen_agp_enable(struct radeon_device *rdev) WREG32(VM_CONTEXT1_CNTL, 0); } -static void evergreen_mc_stop(struct radeon_device *rdev, struct evergreen_mc_save *save) +void evergreen_mc_stop(struct radeon_device *rdev, struct evergreen_mc_save *save) { save->vga_control[0] = RREG32(D1VGA_CONTROL); save->vga_control[1] = RREG32(D2VGA_CONTROL); @@ -1011,7 +1011,7 @@ static void evergreen_mc_stop(struct radeon_device *rdev, struct evergreen_mc_sa WREG32(EVERGREEN_D6VGA_CONTROL, 0); } -static void evergreen_mc_resume(struct radeon_device *rdev, struct evergreen_mc_save *save) +void evergreen_mc_resume(struct radeon_device *rdev, struct evergreen_mc_save *save) { WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + EVERGREEN_CRTC0_REGISTER_OFFSET, upper_32_bits(rdev->mc.vram_start)); @@ -1108,7 +1108,7 @@ static void evergreen_mc_resume(struct radeon_device *rdev, struct evergreen_mc_ WREG32(VGA_RENDER_CONTROL, save->vga_render_control); } -static void evergreen_mc_program(struct radeon_device *rdev) +void evergreen_mc_program(struct radeon_device *rdev) { struct evergreen_mc_save save; u32 tmp; @@ -1192,7 +1192,11 @@ void evergreen_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib) radeon_ring_write(rdev, 1); /* FIXME: implement */ radeon_ring_write(rdev, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); - radeon_ring_write(rdev, ib->gpu_addr & 0xFFFFFFFC); + radeon_ring_write(rdev, +#ifdef __BIG_ENDIAN + (2 << 0) | +#endif + (ib->gpu_addr & 0xFFFFFFFC)); radeon_ring_write(rdev, upper_32_bits(ib->gpu_addr) & 0xFF); radeon_ring_write(rdev, ib->length_dw); } @@ -1207,7 +1211,11 @@ static int evergreen_cp_load_microcode(struct radeon_device *rdev) return -EINVAL; r700_cp_stop(rdev); - WREG32(CP_RB_CNTL, RB_NO_UPDATE | (15 << 8) | (3 << 0)); + WREG32(CP_RB_CNTL, +#ifdef __BIG_ENDIAN + BUF_SWAP_32BIT | +#endif + RB_NO_UPDATE | RB_BLKSZ(15) | RB_BUFSZ(3)); fw_data = (const __be32 *)rdev->pfp_fw->data; WREG32(CP_PFP_UCODE_ADDR, 0); @@ -1326,7 +1334,11 @@ int evergreen_cp_resume(struct radeon_device *rdev) WREG32(CP_RB_WPTR, 0); /* set the wb address wether it's enabled or not */ - WREG32(CP_RB_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFFFFFFFC); + WREG32(CP_RB_RPTR_ADDR, +#ifdef __BIG_ENDIAN + RB_RPTR_SWAP(2) | +#endif + ((rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFFFFFFFC)); WREG32(CP_RB_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFF); WREG32(SCRATCH_ADDR, ((rdev->wb.gpu_addr + RADEON_WB_SCRATCH_OFFSET) >> 8) & 0xFFFFFFFF); @@ -2182,7 +2194,6 @@ int evergreen_mc_init(struct radeon_device *rdev) rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE) * 1024 * 1024; } rdev->mc.visible_vram_size = rdev->mc.aper_size; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r700_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); @@ -2565,7 +2576,7 @@ void evergreen_irq_disable(struct radeon_device *rdev) evergreen_disable_interrupt_state(rdev); } -static void evergreen_irq_suspend(struct radeon_device *rdev) +void evergreen_irq_suspend(struct radeon_device *rdev) { evergreen_irq_disable(rdev); r600_rlc_stop(rdev); @@ -2627,8 +2638,8 @@ restart_ih: while (rptr != wptr) { /* wptr/rptr are in bytes! */ ring_index = rptr / 4; - src_id = rdev->ih.ring[ring_index] & 0xff; - src_data = rdev->ih.ring[ring_index + 1] & 0xfffffff; + src_id = le32_to_cpu(rdev->ih.ring[ring_index]) & 0xff; + src_data = le32_to_cpu(rdev->ih.ring[ring_index + 1]) & 0xfffffff; switch (src_id) { case 1: /* D1 vblank/vline */ @@ -2888,7 +2899,7 @@ static int evergreen_startup(struct radeon_device *rdev) return r; } } - r = btc_mc_load_microcode(rdev); + r = ni_mc_load_microcode(rdev); if (r) { DRM_ERROR("Failed to load MC firmware!\n"); return r; @@ -2922,7 +2933,7 @@ static int evergreen_startup(struct radeon_device *rdev) /* XXX: ontario has problems blitting to gart at the moment */ if (rdev->family == CHIP_PALM) { rdev->asic->copy = NULL; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); } /* allocate wb buffer */ @@ -2970,7 +2981,7 @@ int evergreen_resume(struct radeon_device *rdev) r = evergreen_startup(rdev); if (r) { - DRM_ERROR("r600 startup failed on resume\n"); + DRM_ERROR("evergreen startup failed on resume\n"); return r; } @@ -3050,7 +3061,7 @@ int evergreen_init(struct radeon_device *rdev) } /* Must be an ATOMBIOS */ if (!rdev->is_atom_bios) { - dev_err(rdev->dev, "Expecting atombios for R600 GPU\n"); + dev_err(rdev->dev, "Expecting atombios for evergreen GPU\n"); return -EINVAL; } r = radeon_atombios_init(rdev); diff --git a/drivers/gpu/drm/radeon/evergreen_blit_kms.c b/drivers/gpu/drm/radeon/evergreen_blit_kms.c index a1ba4b3..ba06a69 100644 --- a/drivers/gpu/drm/radeon/evergreen_blit_kms.c +++ b/drivers/gpu/drm/radeon/evergreen_blit_kms.c @@ -55,7 +55,7 @@ set_render_target(struct radeon_device *rdev, int format, if (h < 8) h = 8; - cb_color_info = ((format << 2) | (1 << 24)); + cb_color_info = ((format << 2) | (1 << 24) | (1 << 8)); pitch = (w / 8) - 1; slice = ((w * h) / 64) - 1; @@ -133,6 +133,9 @@ set_vtx_resource(struct radeon_device *rdev, u64 gpu_addr) /* high addr, stride */ sq_vtx_constant_word2 = ((upper_32_bits(gpu_addr) & 0xff) | (16 << 8)); +#ifdef __BIG_ENDIAN + sq_vtx_constant_word2 |= (2 << 30); +#endif /* xyzw swizzles */ sq_vtx_constant_word3 = (0 << 3) | (1 << 6) | (2 << 9) | (3 << 12); @@ -173,7 +176,7 @@ set_tex_resource(struct radeon_device *rdev, sq_tex_resource_word0 = (1 << 0); /* 2D */ sq_tex_resource_word0 |= ((((pitch >> 3) - 1) << 6) | ((w - 1) << 18)); - sq_tex_resource_word1 = ((h - 1) << 0); + sq_tex_resource_word1 = ((h - 1) << 0) | (1 << 28); /* xyzw swizzles */ sq_tex_resource_word4 = (0 << 16) | (1 << 19) | (2 << 22) | (3 << 25); @@ -221,7 +224,11 @@ draw_auto(struct radeon_device *rdev) radeon_ring_write(rdev, DI_PT_RECTLIST); radeon_ring_write(rdev, PACKET3(PACKET3_INDEX_TYPE, 0)); - radeon_ring_write(rdev, DI_INDEX_SIZE_16_BIT); + radeon_ring_write(rdev, +#ifdef __BIG_ENDIAN + (2 << 2) | +#endif + DI_INDEX_SIZE_16_BIT); radeon_ring_write(rdev, PACKET3(PACKET3_NUM_INSTANCES, 0)); radeon_ring_write(rdev, 1); @@ -541,7 +548,7 @@ static inline uint32_t i2f(uint32_t input) int evergreen_blit_init(struct radeon_device *rdev) { u32 obj_size; - int r, dwords; + int i, r, dwords; void *ptr; u32 packet2s[16]; int num_packet2s = 0; @@ -557,7 +564,7 @@ int evergreen_blit_init(struct radeon_device *rdev) dwords = rdev->r600_blit.state_len; while (dwords & 0xf) { - packet2s[num_packet2s++] = PACKET2(0); + packet2s[num_packet2s++] = cpu_to_le32(PACKET2(0)); dwords++; } @@ -572,7 +579,7 @@ int evergreen_blit_init(struct radeon_device *rdev) obj_size += evergreen_ps_size * 4; obj_size = ALIGN(obj_size, 256); - r = radeon_bo_create(rdev, NULL, obj_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, + r = radeon_bo_create(rdev, obj_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, &rdev->r600_blit.shader_obj); if (r) { DRM_ERROR("evergreen failed to allocate shader\n"); @@ -598,8 +605,10 @@ int evergreen_blit_init(struct radeon_device *rdev) if (num_packet2s) memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), packet2s, num_packet2s * 4); - memcpy(ptr + rdev->r600_blit.vs_offset, evergreen_vs, evergreen_vs_size * 4); - memcpy(ptr + rdev->r600_blit.ps_offset, evergreen_ps, evergreen_ps_size * 4); + for (i = 0; i < evergreen_vs_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(evergreen_vs[i]); + for (i = 0; i < evergreen_ps_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(evergreen_ps[i]); radeon_bo_kunmap(rdev->r600_blit.shader_obj); radeon_bo_unreserve(rdev->r600_blit.shader_obj); @@ -614,7 +623,7 @@ done: dev_err(rdev->dev, "(%d) pin blit object failed\n", r); return r; } - rdev->mc.active_vram_size = rdev->mc.real_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size); return 0; } @@ -622,7 +631,7 @@ void evergreen_blit_fini(struct radeon_device *rdev) { int r; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); if (rdev->r600_blit.shader_obj == NULL) return; /* If we can't reserve the bo, unref should be enough to destroy diff --git a/drivers/gpu/drm/radeon/evergreen_blit_shaders.c b/drivers/gpu/drm/radeon/evergreen_blit_shaders.c index ef1d28c..3a10399 100644 --- a/drivers/gpu/drm/radeon/evergreen_blit_shaders.c +++ b/drivers/gpu/drm/radeon/evergreen_blit_shaders.c @@ -311,11 +311,19 @@ const u32 evergreen_vs[] = 0x00000000, 0x3c000000, 0x67961001, +#ifdef __BIG_ENDIAN + 0x000a0000, +#else 0x00080000, +#endif 0x00000000, 0x1c000000, 0x67961000, +#ifdef __BIG_ENDIAN + 0x00020008, +#else 0x00000008, +#endif 0x00000000, }; diff --git a/drivers/gpu/drm/radeon/evergreen_cs.c b/drivers/gpu/drm/radeon/evergreen_cs.c index 345a75a..edde90b 100644 --- a/drivers/gpu/drm/radeon/evergreen_cs.c +++ b/drivers/gpu/drm/radeon/evergreen_cs.c @@ -29,6 +29,7 @@ #include "radeon.h" #include "evergreend.h" #include "evergreen_reg_safe.h" +#include "cayman_reg_safe.h" static int evergreen_cs_packet_next_reloc(struct radeon_cs_parser *p, struct radeon_cs_reloc **cs_reloc); @@ -292,33 +293,28 @@ static int evergreen_cs_packet_parse_vline(struct radeon_cs_parser *p) if (wait_reg_mem.type != PACKET_TYPE3 || wait_reg_mem.opcode != PACKET3_WAIT_REG_MEM) { DRM_ERROR("vline wait missing WAIT_REG_MEM segment\n"); - r = -EINVAL; - return r; + return -EINVAL; } wait_reg_mem_info = radeon_get_ib_value(p, wait_reg_mem.idx + 1); /* bit 4 is reg (0) or mem (1) */ if (wait_reg_mem_info & 0x10) { DRM_ERROR("vline WAIT_REG_MEM waiting on MEM rather than REG\n"); - r = -EINVAL; - return r; + return -EINVAL; } /* waiting for value to be equal */ if ((wait_reg_mem_info & 0x7) != 0x3) { DRM_ERROR("vline WAIT_REG_MEM function not equal\n"); - r = -EINVAL; - return r; + return -EINVAL; } if ((radeon_get_ib_value(p, wait_reg_mem.idx + 2) << 2) != EVERGREEN_VLINE_STATUS) { DRM_ERROR("vline WAIT_REG_MEM bad reg\n"); - r = -EINVAL; - return r; + return -EINVAL; } if (radeon_get_ib_value(p, wait_reg_mem.idx + 5) != EVERGREEN_VLINE_STAT) { DRM_ERROR("vline WAIT_REG_MEM bad bit mask\n"); - r = -EINVAL; - return r; + return -EINVAL; } /* jump over the NOP */ @@ -336,8 +332,7 @@ static int evergreen_cs_packet_parse_vline(struct radeon_cs_parser *p) obj = drm_mode_object_find(p->rdev->ddev, crtc_id, DRM_MODE_OBJECT_CRTC); if (!obj) { DRM_ERROR("cannot find crtc %d\n", crtc_id); - r = -EINVAL; - goto out; + return -EINVAL; } crtc = obj_to_crtc(obj); radeon_crtc = to_radeon_crtc(crtc); @@ -362,12 +357,10 @@ static int evergreen_cs_packet_parse_vline(struct radeon_cs_parser *p) break; default: DRM_ERROR("unknown crtc reloc\n"); - r = -EINVAL; - goto out; + return -EINVAL; } } -out: - return r; + return 0; } static int evergreen_packet0_check(struct radeon_cs_parser *p, @@ -425,18 +418,28 @@ static inline int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u3 { struct evergreen_cs_track *track = (struct evergreen_cs_track *)p->track; struct radeon_cs_reloc *reloc; - u32 last_reg = ARRAY_SIZE(evergreen_reg_safe_bm); + u32 last_reg; u32 m, i, tmp, *ib; int r; + if (p->rdev->family >= CHIP_CAYMAN) + last_reg = ARRAY_SIZE(cayman_reg_safe_bm); + else + last_reg = ARRAY_SIZE(evergreen_reg_safe_bm); + i = (reg >> 7); if (i > last_reg) { dev_warn(p->dev, "forbidden register 0x%08x at %d\n", reg, idx); return -EINVAL; } m = 1 << ((reg >> 2) & 31); - if (!(evergreen_reg_safe_bm[i] & m)) - return 0; + if (p->rdev->family >= CHIP_CAYMAN) { + if (!(cayman_reg_safe_bm[i] & m)) + return 0; + } else { + if (!(evergreen_reg_safe_bm[i] & m)) + return 0; + } ib = p->ib->ptr; switch (reg) { /* force following reg to 0 in an attemp to disable out buffer @@ -468,12 +471,42 @@ static inline int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u3 case SQ_VSTMP_RING_ITEMSIZE: case VGT_TF_RING_SIZE: /* get value to populate the IB don't remove */ - tmp =radeon_get_ib_value(p, idx); - ib[idx] = 0; + /*tmp =radeon_get_ib_value(p, idx); + ib[idx] = 0;*/ + break; + case SQ_ESGS_RING_BASE: + case SQ_GSVS_RING_BASE: + case SQ_ESTMP_RING_BASE: + case SQ_GSTMP_RING_BASE: + case SQ_HSTMP_RING_BASE: + case SQ_LSTMP_RING_BASE: + case SQ_PSTMP_RING_BASE: + case SQ_VSTMP_RING_BASE: + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff); break; case DB_DEPTH_CONTROL: track->db_depth_control = radeon_get_ib_value(p, idx); break; + case CAYMAN_DB_EQAA: + if (p->rdev->family < CHIP_CAYMAN) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + break; + case CAYMAN_DB_DEPTH_INFO: + if (p->rdev->family < CHIP_CAYMAN) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + break; case DB_Z_INFO: r = evergreen_cs_packet_next_reloc(p, &reloc); if (r) { @@ -559,9 +592,23 @@ static inline int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u3 track->cb_shader_mask = radeon_get_ib_value(p, idx); break; case PA_SC_AA_CONFIG: + if (p->rdev->family >= CHIP_CAYMAN) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } tmp = radeon_get_ib_value(p, idx) & MSAA_NUM_SAMPLES_MASK; track->nsamples = 1 << tmp; break; + case CAYMAN_PA_SC_AA_CONFIG: + if (p->rdev->family < CHIP_CAYMAN) { + dev_warn(p->dev, "bad SET_CONTEXT_REG " + "0x%04X\n", reg); + return -EINVAL; + } + tmp = radeon_get_ib_value(p, idx) & CAYMAN_MSAA_NUM_SAMPLES_MASK; + track->nsamples = 1 << tmp; + break; case CB_COLOR0_VIEW: case CB_COLOR1_VIEW: case CB_COLOR2_VIEW: @@ -942,6 +989,37 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p, idx_value = radeon_get_ib_value(p, idx); switch (pkt->opcode) { + case PACKET3_SET_PREDICATION: + { + int pred_op; + int tmp; + if (pkt->count != 1) { + DRM_ERROR("bad SET PREDICATION\n"); + return -EINVAL; + } + + tmp = radeon_get_ib_value(p, idx + 1); + pred_op = (tmp >> 16) & 0x7; + + /* for the clear predicate operation */ + if (pred_op == 0) + return 0; + + if (pred_op > 2) { + DRM_ERROR("bad SET PREDICATION operation %d\n", pred_op); + return -EINVAL; + } + + r = evergreen_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad SET PREDICATION\n"); + return -EINVAL; + } + + ib[idx + 0] = idx_value + (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx + 1] = tmp + (upper_32_bits(reloc->lobj.gpu_offset) & 0xff); + } + break; case PACKET3_CONTEXT_CONTROL: if (pkt->count != 1) { DRM_ERROR("bad CONTEXT_CONTROL\n"); @@ -956,6 +1034,16 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p, return -EINVAL; } break; + case CAYMAN_PACKET3_DEALLOC_STATE: + if (p->rdev->family < CHIP_CAYMAN) { + DRM_ERROR("bad PACKET3_DEALLOC_STATE\n"); + return -EINVAL; + } + if (pkt->count) { + DRM_ERROR("bad INDEX_TYPE/NUM_INSTANCES/CLEAR_STATE\n"); + return -EINVAL; + } + break; case PACKET3_INDEX_BASE: if (pkt->count != 1) { DRM_ERROR("bad INDEX_BASE\n"); diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h index afec1ac..9aaa3f0 100644 --- a/drivers/gpu/drm/radeon/evergreend.h +++ b/drivers/gpu/drm/radeon/evergreend.h @@ -98,6 +98,7 @@ #define BUF_SWAP_32BIT (2 << 16) #define CP_RB_RPTR 0x8700 #define CP_RB_RPTR_ADDR 0xC10C +#define RB_RPTR_SWAP(x) ((x) << 0) #define CP_RB_RPTR_ADDR_HI 0xC110 #define CP_RB_RPTR_WR 0xC108 #define CP_RB_WPTR 0xC114 @@ -754,13 +755,21 @@ #define SQ_CONST_MEM_BASE 0x8df8 +#define SQ_ESGS_RING_BASE 0x8c40 #define SQ_ESGS_RING_SIZE 0x8c44 +#define SQ_GSVS_RING_BASE 0x8c48 #define SQ_GSVS_RING_SIZE 0x8c4c +#define SQ_ESTMP_RING_BASE 0x8c50 #define SQ_ESTMP_RING_SIZE 0x8c54 +#define SQ_GSTMP_RING_BASE 0x8c58 #define SQ_GSTMP_RING_SIZE 0x8c5c +#define SQ_VSTMP_RING_BASE 0x8c60 #define SQ_VSTMP_RING_SIZE 0x8c64 +#define SQ_PSTMP_RING_BASE 0x8c68 #define SQ_PSTMP_RING_SIZE 0x8c6c +#define SQ_LSTMP_RING_BASE 0x8e10 #define SQ_LSTMP_RING_SIZE 0x8e14 +#define SQ_HSTMP_RING_BASE 0x8e18 #define SQ_HSTMP_RING_SIZE 0x8e1c #define VGT_TF_RING_SIZE 0x8988 @@ -1092,5 +1101,14 @@ #define SQ_TEX_RESOURCE_WORD6_0 0x30018 #define SQ_TEX_RESOURCE_WORD7_0 0x3001c +/* cayman 3D regs */ +#define CAYMAN_VGT_OFFCHIP_LDS_BASE 0x89B0 +#define CAYMAN_DB_EQAA 0x28804 +#define CAYMAN_DB_DEPTH_INFO 0x2803C +#define CAYMAN_PA_SC_AA_CONFIG 0x28BE0 +#define CAYMAN_MSAA_NUM_SAMPLES_SHIFT 0 +#define CAYMAN_MSAA_NUM_SAMPLES_MASK 0x7 +/* cayman packet3 addition */ +#define CAYMAN_PACKET3_DEALLOC_STATE 0x14 #endif diff --git a/drivers/gpu/drm/radeon/mkregtable.c b/drivers/gpu/drm/radeon/mkregtable.c index 607241c..5a82b6b 100644 --- a/drivers/gpu/drm/radeon/mkregtable.c +++ b/drivers/gpu/drm/radeon/mkregtable.c @@ -673,8 +673,10 @@ static int parser_auth(struct table *t, const char *filename) last_reg = strtol(last_reg_s, NULL, 16); do { - if (fgets(buf, 1024, file) == NULL) + if (fgets(buf, 1024, file) == NULL) { + fclose(file); return -1; + } len = strlen(buf); if (ftell(file) == end) done = 1; @@ -685,6 +687,7 @@ static int parser_auth(struct table *t, const char *filename) fprintf(stderr, "Error matching regular expression %d in %s\n", r, filename); + fclose(file); return -1; } else { buf[match[0].rm_eo] = 0; diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c index 5e0bef8..7aade20 100644 --- a/drivers/gpu/drm/radeon/ni.c +++ b/drivers/gpu/drm/radeon/ni.c @@ -31,12 +31,25 @@ #include "nid.h" #include "atom.h" #include "ni_reg.h" +#include "cayman_blit_shaders.h" + +extern void evergreen_mc_stop(struct radeon_device *rdev, struct evergreen_mc_save *save); +extern void evergreen_mc_resume(struct radeon_device *rdev, struct evergreen_mc_save *save); +extern int evergreen_mc_wait_for_idle(struct radeon_device *rdev); +extern void evergreen_mc_program(struct radeon_device *rdev); +extern void evergreen_irq_suspend(struct radeon_device *rdev); +extern int evergreen_mc_init(struct radeon_device *rdev); #define EVERGREEN_PFP_UCODE_SIZE 1120 #define EVERGREEN_PM4_UCODE_SIZE 1376 #define EVERGREEN_RLC_UCODE_SIZE 768 #define BTC_MC_UCODE_SIZE 6024 +#define CAYMAN_PFP_UCODE_SIZE 2176 +#define CAYMAN_PM4_UCODE_SIZE 2176 +#define CAYMAN_RLC_UCODE_SIZE 1024 +#define CAYMAN_MC_UCODE_SIZE 6037 + /* Firmware Names */ MODULE_FIRMWARE("radeon/BARTS_pfp.bin"); MODULE_FIRMWARE("radeon/BARTS_me.bin"); @@ -48,6 +61,10 @@ MODULE_FIRMWARE("radeon/TURKS_mc.bin"); MODULE_FIRMWARE("radeon/CAICOS_pfp.bin"); MODULE_FIRMWARE("radeon/CAICOS_me.bin"); MODULE_FIRMWARE("radeon/CAICOS_mc.bin"); +MODULE_FIRMWARE("radeon/CAYMAN_pfp.bin"); +MODULE_FIRMWARE("radeon/CAYMAN_me.bin"); +MODULE_FIRMWARE("radeon/CAYMAN_mc.bin"); +MODULE_FIRMWARE("radeon/CAYMAN_rlc.bin"); #define BTC_IO_MC_REGS_SIZE 29 @@ -147,12 +164,44 @@ static const u32 caicos_io_mc_regs[BTC_IO_MC_REGS_SIZE][2] = { {0x0000009f, 0x00916a00} }; -int btc_mc_load_microcode(struct radeon_device *rdev) +static const u32 cayman_io_mc_regs[BTC_IO_MC_REGS_SIZE][2] = { + {0x00000077, 0xff010100}, + {0x00000078, 0x00000000}, + {0x00000079, 0x00001434}, + {0x0000007a, 0xcc08ec08}, + {0x0000007b, 0x00040000}, + {0x0000007c, 0x000080c0}, + {0x0000007d, 0x09000000}, + {0x0000007e, 0x00210404}, + {0x00000081, 0x08a8e800}, + {0x00000082, 0x00030444}, + {0x00000083, 0x00000000}, + {0x00000085, 0x00000001}, + {0x00000086, 0x00000002}, + {0x00000087, 0x48490000}, + {0x00000088, 0x20244647}, + {0x00000089, 0x00000005}, + {0x0000008b, 0x66030000}, + {0x0000008c, 0x00006603}, + {0x0000008d, 0x00000100}, + {0x0000008f, 0x00001c0a}, + {0x00000090, 0xff000001}, + {0x00000094, 0x00101101}, + {0x00000095, 0x00000fff}, + {0x00000096, 0x00116fff}, + {0x00000097, 0x60010000}, + {0x00000098, 0x10010000}, + {0x00000099, 0x00006000}, + {0x0000009a, 0x00001000}, + {0x0000009f, 0x00976b00} +}; + +int ni_mc_load_microcode(struct radeon_device *rdev) { const __be32 *fw_data; u32 mem_type, running, blackout = 0; u32 *io_mc_regs; - int i; + int i, ucode_size, regs_size; if (!rdev->mc_fw) return -EINVAL; @@ -160,13 +209,24 @@ int btc_mc_load_microcode(struct radeon_device *rdev) switch (rdev->family) { case CHIP_BARTS: io_mc_regs = (u32 *)&barts_io_mc_regs; + ucode_size = BTC_MC_UCODE_SIZE; + regs_size = BTC_IO_MC_REGS_SIZE; break; case CHIP_TURKS: io_mc_regs = (u32 *)&turks_io_mc_regs; + ucode_size = BTC_MC_UCODE_SIZE; + regs_size = BTC_IO_MC_REGS_SIZE; break; case CHIP_CAICOS: default: io_mc_regs = (u32 *)&caicos_io_mc_regs; + ucode_size = BTC_MC_UCODE_SIZE; + regs_size = BTC_IO_MC_REGS_SIZE; + break; + case CHIP_CAYMAN: + io_mc_regs = (u32 *)&cayman_io_mc_regs; + ucode_size = CAYMAN_MC_UCODE_SIZE; + regs_size = BTC_IO_MC_REGS_SIZE; break; } @@ -184,13 +244,13 @@ int btc_mc_load_microcode(struct radeon_device *rdev) WREG32(MC_SEQ_SUP_CNTL, 0x00000010); /* load mc io regs */ - for (i = 0; i < BTC_IO_MC_REGS_SIZE; i++) { + for (i = 0; i < regs_size; i++) { WREG32(MC_SEQ_IO_DEBUG_INDEX, io_mc_regs[(i << 1)]); WREG32(MC_SEQ_IO_DEBUG_DATA, io_mc_regs[(i << 1) + 1]); } /* load the MC ucode */ fw_data = (const __be32 *)rdev->mc_fw->data; - for (i = 0; i < BTC_MC_UCODE_SIZE; i++) + for (i = 0; i < ucode_size; i++) WREG32(MC_SEQ_SUP_PGM, be32_to_cpup(fw_data++)); /* put the engine back into the active state */ @@ -231,23 +291,38 @@ int ni_init_microcode(struct radeon_device *rdev) case CHIP_BARTS: chip_name = "BARTS"; rlc_chip_name = "BTC"; + pfp_req_size = EVERGREEN_PFP_UCODE_SIZE * 4; + me_req_size = EVERGREEN_PM4_UCODE_SIZE * 4; + rlc_req_size = EVERGREEN_RLC_UCODE_SIZE * 4; + mc_req_size = BTC_MC_UCODE_SIZE * 4; break; case CHIP_TURKS: chip_name = "TURKS"; rlc_chip_name = "BTC"; + pfp_req_size = EVERGREEN_PFP_UCODE_SIZE * 4; + me_req_size = EVERGREEN_PM4_UCODE_SIZE * 4; + rlc_req_size = EVERGREEN_RLC_UCODE_SIZE * 4; + mc_req_size = BTC_MC_UCODE_SIZE * 4; break; case CHIP_CAICOS: chip_name = "CAICOS"; rlc_chip_name = "BTC"; + pfp_req_size = EVERGREEN_PFP_UCODE_SIZE * 4; + me_req_size = EVERGREEN_PM4_UCODE_SIZE * 4; + rlc_req_size = EVERGREEN_RLC_UCODE_SIZE * 4; + mc_req_size = BTC_MC_UCODE_SIZE * 4; + break; + case CHIP_CAYMAN: + chip_name = "CAYMAN"; + rlc_chip_name = "CAYMAN"; + pfp_req_size = CAYMAN_PFP_UCODE_SIZE * 4; + me_req_size = CAYMAN_PM4_UCODE_SIZE * 4; + rlc_req_size = CAYMAN_RLC_UCODE_SIZE * 4; + mc_req_size = CAYMAN_MC_UCODE_SIZE * 4; break; default: BUG(); } - pfp_req_size = EVERGREEN_PFP_UCODE_SIZE * 4; - me_req_size = EVERGREEN_PM4_UCODE_SIZE * 4; - rlc_req_size = EVERGREEN_RLC_UCODE_SIZE * 4; - mc_req_size = BTC_MC_UCODE_SIZE * 4; - DRM_INFO("Loading %s Microcode\n", chip_name); snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", chip_name); @@ -314,3 +389,1204 @@ out: return err; } +/* + * Core functions + */ +static u32 cayman_get_tile_pipe_to_backend_map(struct radeon_device *rdev, + u32 num_tile_pipes, + u32 num_backends_per_asic, + u32 *backend_disable_mask_per_asic, + u32 num_shader_engines) +{ + u32 backend_map = 0; + u32 enabled_backends_mask = 0; + u32 enabled_backends_count = 0; + u32 num_backends_per_se; + u32 cur_pipe; + u32 swizzle_pipe[CAYMAN_MAX_PIPES]; + u32 cur_backend = 0; + u32 i; + bool force_no_swizzle; + + /* force legal values */ + if (num_tile_pipes < 1) + num_tile_pipes = 1; + if (num_tile_pipes > rdev->config.cayman.max_tile_pipes) + num_tile_pipes = rdev->config.cayman.max_tile_pipes; + if (num_shader_engines < 1) + num_shader_engines = 1; + if (num_shader_engines > rdev->config.cayman.max_shader_engines) + num_shader_engines = rdev->config.cayman.max_shader_engines; + if (num_backends_per_asic > num_shader_engines) + num_backends_per_asic = num_shader_engines; + if (num_backends_per_asic > (rdev->config.cayman.max_backends_per_se * num_shader_engines)) + num_backends_per_asic = rdev->config.cayman.max_backends_per_se * num_shader_engines; + + /* make sure we have the same number of backends per se */ + num_backends_per_asic = ALIGN(num_backends_per_asic, num_shader_engines); + /* set up the number of backends per se */ + num_backends_per_se = num_backends_per_asic / num_shader_engines; + if (num_backends_per_se > rdev->config.cayman.max_backends_per_se) { + num_backends_per_se = rdev->config.cayman.max_backends_per_se; + num_backends_per_asic = num_backends_per_se * num_shader_engines; + } + + /* create enable mask and count for enabled backends */ + for (i = 0; i < CAYMAN_MAX_BACKENDS; ++i) { + if (((*backend_disable_mask_per_asic >> i) & 1) == 0) { + enabled_backends_mask |= (1 << i); + ++enabled_backends_count; + } + if (enabled_backends_count == num_backends_per_asic) + break; + } + + /* force the backends mask to match the current number of backends */ + if (enabled_backends_count != num_backends_per_asic) { + u32 this_backend_enabled; + u32 shader_engine; + u32 backend_per_se; + + enabled_backends_mask = 0; + enabled_backends_count = 0; + *backend_disable_mask_per_asic = CAYMAN_MAX_BACKENDS_MASK; + for (i = 0; i < CAYMAN_MAX_BACKENDS; ++i) { + /* calc the current se */ + shader_engine = i / rdev->config.cayman.max_backends_per_se; + /* calc the backend per se */ + backend_per_se = i % rdev->config.cayman.max_backends_per_se; + /* default to not enabled */ + this_backend_enabled = 0; + if ((shader_engine < num_shader_engines) && + (backend_per_se < num_backends_per_se)) + this_backend_enabled = 1; + if (this_backend_enabled) { + enabled_backends_mask |= (1 << i); + *backend_disable_mask_per_asic &= ~(1 << i); + ++enabled_backends_count; + } + } + } + + + memset((uint8_t *)&swizzle_pipe[0], 0, sizeof(u32) * CAYMAN_MAX_PIPES); + switch (rdev->family) { + case CHIP_CAYMAN: + force_no_swizzle = true; + break; + default: + force_no_swizzle = false; + break; + } + if (force_no_swizzle) { + bool last_backend_enabled = false; + + force_no_swizzle = false; + for (i = 0; i < CAYMAN_MAX_BACKENDS; ++i) { + if (((enabled_backends_mask >> i) & 1) == 1) { + if (last_backend_enabled) + force_no_swizzle = true; + last_backend_enabled = true; + } else + last_backend_enabled = false; + } + } + + switch (num_tile_pipes) { + case 1: + case 3: + case 5: + case 7: + DRM_ERROR("odd number of pipes!\n"); + break; + case 2: + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + break; + case 4: + if (force_no_swizzle) { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + swizzle_pipe[2] = 2; + swizzle_pipe[3] = 3; + } else { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 2; + swizzle_pipe[2] = 1; + swizzle_pipe[3] = 3; + } + break; + case 6: + if (force_no_swizzle) { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + swizzle_pipe[2] = 2; + swizzle_pipe[3] = 3; + swizzle_pipe[4] = 4; + swizzle_pipe[5] = 5; + } else { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 2; + swizzle_pipe[2] = 4; + swizzle_pipe[3] = 1; + swizzle_pipe[4] = 3; + swizzle_pipe[5] = 5; + } + break; + case 8: + if (force_no_swizzle) { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 1; + swizzle_pipe[2] = 2; + swizzle_pipe[3] = 3; + swizzle_pipe[4] = 4; + swizzle_pipe[5] = 5; + swizzle_pipe[6] = 6; + swizzle_pipe[7] = 7; + } else { + swizzle_pipe[0] = 0; + swizzle_pipe[1] = 2; + swizzle_pipe[2] = 4; + swizzle_pipe[3] = 6; + swizzle_pipe[4] = 1; + swizzle_pipe[5] = 3; + swizzle_pipe[6] = 5; + swizzle_pipe[7] = 7; + } + break; + } + + for (cur_pipe = 0; cur_pipe < num_tile_pipes; ++cur_pipe) { + while (((1 << cur_backend) & enabled_backends_mask) == 0) + cur_backend = (cur_backend + 1) % CAYMAN_MAX_BACKENDS; + + backend_map |= (((cur_backend & 0xf) << (swizzle_pipe[cur_pipe] * 4))); + + cur_backend = (cur_backend + 1) % CAYMAN_MAX_BACKENDS; + } + + return backend_map; +} + +static void cayman_program_channel_remap(struct radeon_device *rdev) +{ + u32 tcp_chan_steer_lo, tcp_chan_steer_hi, mc_shared_chremap, tmp; + + tmp = RREG32(MC_SHARED_CHMAP); + switch ((tmp & NOOFCHAN_MASK) >> NOOFCHAN_SHIFT) { + case 0: + case 1: + case 2: + case 3: + default: + /* default mapping */ + mc_shared_chremap = 0x00fac688; + break; + } + + switch (rdev->family) { + case CHIP_CAYMAN: + default: + //tcp_chan_steer_lo = 0x54763210 + tcp_chan_steer_lo = 0x76543210; + tcp_chan_steer_hi = 0x0000ba98; + break; + } + + WREG32(TCP_CHAN_STEER_LO, tcp_chan_steer_lo); + WREG32(TCP_CHAN_STEER_HI, tcp_chan_steer_hi); + WREG32(MC_SHARED_CHREMAP, mc_shared_chremap); +} + +static u32 cayman_get_disable_mask_per_asic(struct radeon_device *rdev, + u32 disable_mask_per_se, + u32 max_disable_mask_per_se, + u32 num_shader_engines) +{ + u32 disable_field_width_per_se = r600_count_pipe_bits(disable_mask_per_se); + u32 disable_mask_per_asic = disable_mask_per_se & max_disable_mask_per_se; + + if (num_shader_engines == 1) + return disable_mask_per_asic; + else if (num_shader_engines == 2) + return disable_mask_per_asic | (disable_mask_per_asic << disable_field_width_per_se); + else + return 0xffffffff; +} + +static void cayman_gpu_init(struct radeon_device *rdev) +{ + u32 cc_rb_backend_disable = 0; + u32 cc_gc_shader_pipe_config; + u32 gb_addr_config = 0; + u32 mc_shared_chmap, mc_arb_ramcfg; + u32 gb_backend_map; + u32 cgts_tcc_disable; + u32 sx_debug_1; + u32 smx_dc_ctl0; + u32 gc_user_shader_pipe_config; + u32 gc_user_rb_backend_disable; + u32 cgts_user_tcc_disable; + u32 cgts_sm_ctrl_reg; + u32 hdp_host_path_cntl; + u32 tmp; + int i, j; + + switch (rdev->family) { + case CHIP_CAYMAN: + default: + rdev->config.cayman.max_shader_engines = 2; + rdev->config.cayman.max_pipes_per_simd = 4; + rdev->config.cayman.max_tile_pipes = 8; + rdev->config.cayman.max_simds_per_se = 12; + rdev->config.cayman.max_backends_per_se = 4; + rdev->config.cayman.max_texture_channel_caches = 8; + rdev->config.cayman.max_gprs = 256; + rdev->config.cayman.max_threads = 256; + rdev->config.cayman.max_gs_threads = 32; + rdev->config.cayman.max_stack_entries = 512; + rdev->config.cayman.sx_num_of_sets = 8; + rdev->config.cayman.sx_max_export_size = 256; + rdev->config.cayman.sx_max_export_pos_size = 64; + rdev->config.cayman.sx_max_export_smx_size = 192; + rdev->config.cayman.max_hw_contexts = 8; + rdev->config.cayman.sq_num_cf_insts = 2; + + rdev->config.cayman.sc_prim_fifo_size = 0x100; + rdev->config.cayman.sc_hiz_tile_fifo_size = 0x30; + rdev->config.cayman.sc_earlyz_tile_fifo_size = 0x130; + break; + } + + /* Initialize HDP */ + for (i = 0, j = 0; i < 32; i++, j += 0x18) { + WREG32((0x2c14 + j), 0x00000000); + WREG32((0x2c18 + j), 0x00000000); + WREG32((0x2c1c + j), 0x00000000); + WREG32((0x2c20 + j), 0x00000000); + WREG32((0x2c24 + j), 0x00000000); + } + + WREG32(GRBM_CNTL, GRBM_READ_TIMEOUT(0xff)); + + mc_shared_chmap = RREG32(MC_SHARED_CHMAP); + mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG); + + cc_rb_backend_disable = RREG32(CC_RB_BACKEND_DISABLE); + cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG); + cgts_tcc_disable = RREG32(CGTS_TCC_DISABLE); + gc_user_rb_backend_disable = RREG32(GC_USER_RB_BACKEND_DISABLE); + gc_user_shader_pipe_config = RREG32(GC_USER_SHADER_PIPE_CONFIG); + cgts_user_tcc_disable = RREG32(CGTS_USER_TCC_DISABLE); + + rdev->config.cayman.num_shader_engines = rdev->config.cayman.max_shader_engines; + tmp = ((~gc_user_shader_pipe_config) & INACTIVE_QD_PIPES_MASK) >> INACTIVE_QD_PIPES_SHIFT; + rdev->config.cayman.num_shader_pipes_per_simd = r600_count_pipe_bits(tmp); + rdev->config.cayman.num_tile_pipes = rdev->config.cayman.max_tile_pipes; + tmp = ((~gc_user_shader_pipe_config) & INACTIVE_SIMDS_MASK) >> INACTIVE_SIMDS_SHIFT; + rdev->config.cayman.num_simds_per_se = r600_count_pipe_bits(tmp); + tmp = ((~gc_user_rb_backend_disable) & BACKEND_DISABLE_MASK) >> BACKEND_DISABLE_SHIFT; + rdev->config.cayman.num_backends_per_se = r600_count_pipe_bits(tmp); + tmp = (gc_user_rb_backend_disable & BACKEND_DISABLE_MASK) >> BACKEND_DISABLE_SHIFT; + rdev->config.cayman.backend_disable_mask_per_asic = + cayman_get_disable_mask_per_asic(rdev, tmp, CAYMAN_MAX_BACKENDS_PER_SE_MASK, + rdev->config.cayman.num_shader_engines); + rdev->config.cayman.backend_map = + cayman_get_tile_pipe_to_backend_map(rdev, rdev->config.cayman.num_tile_pipes, + rdev->config.cayman.num_backends_per_se * + rdev->config.cayman.num_shader_engines, + &rdev->config.cayman.backend_disable_mask_per_asic, + rdev->config.cayman.num_shader_engines); + tmp = ((~cgts_user_tcc_disable) & TCC_DISABLE_MASK) >> TCC_DISABLE_SHIFT; + rdev->config.cayman.num_texture_channel_caches = r600_count_pipe_bits(tmp); + tmp = (mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT; + rdev->config.cayman.mem_max_burst_length_bytes = (tmp + 1) * 256; + if (rdev->config.cayman.mem_max_burst_length_bytes > 512) + rdev->config.cayman.mem_max_burst_length_bytes = 512; + tmp = (mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT; + rdev->config.cayman.mem_row_size_in_kb = (4 * (1 << (8 + tmp))) / 1024; + if (rdev->config.cayman.mem_row_size_in_kb > 4) + rdev->config.cayman.mem_row_size_in_kb = 4; + /* XXX use MC settings? */ + rdev->config.cayman.shader_engine_tile_size = 32; + rdev->config.cayman.num_gpus = 1; + rdev->config.cayman.multi_gpu_tile_size = 64; + + //gb_addr_config = 0x02011003 +#if 0 + gb_addr_config = RREG32(GB_ADDR_CONFIG); +#else + gb_addr_config = 0; + switch (rdev->config.cayman.num_tile_pipes) { + case 1: + default: + gb_addr_config |= NUM_PIPES(0); + break; + case 2: + gb_addr_config |= NUM_PIPES(1); + break; + case 4: + gb_addr_config |= NUM_PIPES(2); + break; + case 8: + gb_addr_config |= NUM_PIPES(3); + break; + } + + tmp = (rdev->config.cayman.mem_max_burst_length_bytes / 256) - 1; + gb_addr_config |= PIPE_INTERLEAVE_SIZE(tmp); + gb_addr_config |= NUM_SHADER_ENGINES(rdev->config.cayman.num_shader_engines - 1); + tmp = (rdev->config.cayman.shader_engine_tile_size / 16) - 1; + gb_addr_config |= SHADER_ENGINE_TILE_SIZE(tmp); + switch (rdev->config.cayman.num_gpus) { + case 1: + default: + gb_addr_config |= NUM_GPUS(0); + break; + case 2: + gb_addr_config |= NUM_GPUS(1); + break; + case 4: + gb_addr_config |= NUM_GPUS(2); + break; + } + switch (rdev->config.cayman.multi_gpu_tile_size) { + case 16: + gb_addr_config |= MULTI_GPU_TILE_SIZE(0); + break; + case 32: + default: + gb_addr_config |= MULTI_GPU_TILE_SIZE(1); + break; + case 64: + gb_addr_config |= MULTI_GPU_TILE_SIZE(2); + break; + case 128: + gb_addr_config |= MULTI_GPU_TILE_SIZE(3); + break; + } + switch (rdev->config.cayman.mem_row_size_in_kb) { + case 1: + default: + gb_addr_config |= ROW_SIZE(0); + break; + case 2: + gb_addr_config |= ROW_SIZE(1); + break; + case 4: + gb_addr_config |= ROW_SIZE(2); + break; + } +#endif + + tmp = (gb_addr_config & NUM_PIPES_MASK) >> NUM_PIPES_SHIFT; + rdev->config.cayman.num_tile_pipes = (1 << tmp); + tmp = (gb_addr_config & PIPE_INTERLEAVE_SIZE_MASK) >> PIPE_INTERLEAVE_SIZE_SHIFT; + rdev->config.cayman.mem_max_burst_length_bytes = (tmp + 1) * 256; + tmp = (gb_addr_config & NUM_SHADER_ENGINES_MASK) >> NUM_SHADER_ENGINES_SHIFT; + rdev->config.cayman.num_shader_engines = tmp + 1; + tmp = (gb_addr_config & NUM_GPUS_MASK) >> NUM_GPUS_SHIFT; + rdev->config.cayman.num_gpus = tmp + 1; + tmp = (gb_addr_config & MULTI_GPU_TILE_SIZE_MASK) >> MULTI_GPU_TILE_SIZE_SHIFT; + rdev->config.cayman.multi_gpu_tile_size = 1 << tmp; + tmp = (gb_addr_config & ROW_SIZE_MASK) >> ROW_SIZE_SHIFT; + rdev->config.cayman.mem_row_size_in_kb = 1 << tmp; + + //gb_backend_map = 0x76541032; +#if 0 + gb_backend_map = RREG32(GB_BACKEND_MAP); +#else + gb_backend_map = + cayman_get_tile_pipe_to_backend_map(rdev, rdev->config.cayman.num_tile_pipes, + rdev->config.cayman.num_backends_per_se * + rdev->config.cayman.num_shader_engines, + &rdev->config.cayman.backend_disable_mask_per_asic, + rdev->config.cayman.num_shader_engines); +#endif + /* setup tiling info dword. gb_addr_config is not adequate since it does + * not have bank info, so create a custom tiling dword. + * bits 3:0 num_pipes + * bits 7:4 num_banks + * bits 11:8 group_size + * bits 15:12 row_size + */ + rdev->config.cayman.tile_config = 0; + switch (rdev->config.cayman.num_tile_pipes) { + case 1: + default: + rdev->config.cayman.tile_config |= (0 << 0); + break; + case 2: + rdev->config.cayman.tile_config |= (1 << 0); + break; + case 4: + rdev->config.cayman.tile_config |= (2 << 0); + break; + case 8: + rdev->config.cayman.tile_config |= (3 << 0); + break; + } + rdev->config.cayman.tile_config |= + ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT) << 4; + rdev->config.cayman.tile_config |= + (gb_addr_config & PIPE_INTERLEAVE_SIZE_MASK) >> PIPE_INTERLEAVE_SIZE_SHIFT; + rdev->config.cayman.tile_config |= + ((gb_addr_config & ROW_SIZE_MASK) >> ROW_SIZE_SHIFT) << 12; + + WREG32(GB_BACKEND_MAP, gb_backend_map); + WREG32(GB_ADDR_CONFIG, gb_addr_config); + WREG32(DMIF_ADDR_CONFIG, gb_addr_config); + WREG32(HDP_ADDR_CONFIG, gb_addr_config); + + cayman_program_channel_remap(rdev); + + /* primary versions */ + WREG32(CC_RB_BACKEND_DISABLE, cc_rb_backend_disable); + WREG32(CC_SYS_RB_BACKEND_DISABLE, cc_rb_backend_disable); + WREG32(CC_GC_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config); + + WREG32(CGTS_TCC_DISABLE, cgts_tcc_disable); + WREG32(CGTS_SYS_TCC_DISABLE, cgts_tcc_disable); + + /* user versions */ + WREG32(GC_USER_RB_BACKEND_DISABLE, cc_rb_backend_disable); + WREG32(GC_USER_SYS_RB_BACKEND_DISABLE, cc_rb_backend_disable); + WREG32(GC_USER_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config); + + WREG32(CGTS_USER_SYS_TCC_DISABLE, cgts_tcc_disable); + WREG32(CGTS_USER_TCC_DISABLE, cgts_tcc_disable); + + /* reprogram the shader complex */ + cgts_sm_ctrl_reg = RREG32(CGTS_SM_CTRL_REG); + for (i = 0; i < 16; i++) + WREG32(CGTS_SM_CTRL_REG, OVERRIDE); + WREG32(CGTS_SM_CTRL_REG, cgts_sm_ctrl_reg); + + /* set HW defaults for 3D engine */ + WREG32(CP_MEQ_THRESHOLDS, MEQ1_START(0x30) | MEQ2_START(0x60)); + + sx_debug_1 = RREG32(SX_DEBUG_1); + sx_debug_1 |= ENABLE_NEW_SMX_ADDRESS; + WREG32(SX_DEBUG_1, sx_debug_1); + + smx_dc_ctl0 = RREG32(SMX_DC_CTL0); + smx_dc_ctl0 &= ~NUMBER_OF_SETS(0x1ff); + smx_dc_ctl0 |= NUMBER_OF_SETS(rdev->config.evergreen.sx_num_of_sets); + WREG32(SMX_DC_CTL0, smx_dc_ctl0); + + WREG32(SPI_CONFIG_CNTL_1, VTX_DONE_DELAY(4) | CRC_SIMD_ID_WADDR_DISABLE); + + /* need to be explicitly zero-ed */ + WREG32(VGT_OFFCHIP_LDS_BASE, 0); + WREG32(SQ_LSTMP_RING_BASE, 0); + WREG32(SQ_HSTMP_RING_BASE, 0); + WREG32(SQ_ESTMP_RING_BASE, 0); + WREG32(SQ_GSTMP_RING_BASE, 0); + WREG32(SQ_VSTMP_RING_BASE, 0); + WREG32(SQ_PSTMP_RING_BASE, 0); + + WREG32(TA_CNTL_AUX, DISABLE_CUBE_ANISO); + + WREG32(SX_EXPORT_BUFFER_SIZES, (COLOR_BUFFER_SIZE((rdev->config.evergreen.sx_max_export_size / 4) - 1) | + POSITION_BUFFER_SIZE((rdev->config.evergreen.sx_max_export_pos_size / 4) - 1) | + SMX_BUFFER_SIZE((rdev->config.evergreen.sx_max_export_smx_size / 4) - 1))); + + WREG32(PA_SC_FIFO_SIZE, (SC_PRIM_FIFO_SIZE(rdev->config.evergreen.sc_prim_fifo_size) | + SC_HIZ_TILE_FIFO_SIZE(rdev->config.evergreen.sc_hiz_tile_fifo_size) | + SC_EARLYZ_TILE_FIFO_SIZE(rdev->config.evergreen.sc_earlyz_tile_fifo_size))); + + + WREG32(VGT_NUM_INSTANCES, 1); + + WREG32(CP_PERFMON_CNTL, 0); + + WREG32(SQ_MS_FIFO_SIZES, (CACHE_FIFO_SIZE(16 * rdev->config.evergreen.sq_num_cf_insts) | + FETCH_FIFO_HIWATER(0x4) | + DONE_FIFO_HIWATER(0xe0) | + ALU_UPDATE_FIFO_HIWATER(0x8))); + + WREG32(SQ_GPR_RESOURCE_MGMT_1, NUM_CLAUSE_TEMP_GPRS(4)); + WREG32(SQ_CONFIG, (VC_ENABLE | + EXPORT_SRC_C | + GFX_PRIO(0) | + CS1_PRIO(0) | + CS2_PRIO(1))); + WREG32(SQ_DYN_GPR_CNTL_PS_FLUSH_REQ, DYN_GPR_ENABLE); + + WREG32(PA_SC_FORCE_EOV_MAX_CNTS, (FORCE_EOV_MAX_CLK_CNT(4095) | + FORCE_EOV_MAX_REZ_CNT(255))); + + WREG32(VGT_CACHE_INVALIDATION, CACHE_INVALIDATION(VC_AND_TC) | + AUTO_INVLD_EN(ES_AND_GS_AUTO)); + + WREG32(VGT_GS_VERTEX_REUSE, 16); + WREG32(PA_SC_LINE_STIPPLE_STATE, 0); + + WREG32(CB_PERF_CTR0_SEL_0, 0); + WREG32(CB_PERF_CTR0_SEL_1, 0); + WREG32(CB_PERF_CTR1_SEL_0, 0); + WREG32(CB_PERF_CTR1_SEL_1, 0); + WREG32(CB_PERF_CTR2_SEL_0, 0); + WREG32(CB_PERF_CTR2_SEL_1, 0); + WREG32(CB_PERF_CTR3_SEL_0, 0); + WREG32(CB_PERF_CTR3_SEL_1, 0); + + hdp_host_path_cntl = RREG32(HDP_HOST_PATH_CNTL); + WREG32(HDP_HOST_PATH_CNTL, hdp_host_path_cntl); + + WREG32(PA_CL_ENHANCE, CLIP_VTX_REORDER_ENA | NUM_CLIP_SEQ(3)); + + udelay(50); +} + +/* + * GART + */ +void cayman_pcie_gart_tlb_flush(struct radeon_device *rdev) +{ + /* flush hdp cache */ + WREG32(HDP_MEM_COHERENCY_FLUSH_CNTL, 0x1); + + /* bits 0-7 are the VM contexts0-7 */ + WREG32(VM_INVALIDATE_REQUEST, 1); +} + +int cayman_pcie_gart_enable(struct radeon_device *rdev) +{ + int r; + + if (rdev->gart.table.vram.robj == NULL) { + dev_err(rdev->dev, "No VRAM object for PCIE GART.\n"); + return -EINVAL; + } + r = radeon_gart_table_vram_pin(rdev); + if (r) + return r; + radeon_gart_restore(rdev); + /* Setup TLB control */ + WREG32(MC_VM_MX_L1_TLB_CNTL, ENABLE_L1_TLB | + ENABLE_L1_FRAGMENT_PROCESSING | + SYSTEM_ACCESS_MODE_NOT_IN_SYS | + SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU); + /* Setup L2 cache */ + WREG32(VM_L2_CNTL, ENABLE_L2_CACHE | + ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE | + ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE | + EFFECTIVE_L2_QUEUE_SIZE(7) | + CONTEXT1_IDENTITY_ACCESS_MODE(1)); + WREG32(VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS | INVALIDATE_L2_CACHE); + WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY | + L2_CACHE_BIGK_FRAGMENT_SIZE(6)); + /* setup context0 */ + WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, rdev->mc.gtt_start >> 12); + WREG32(VM_CONTEXT0_PAGE_TABLE_END_ADDR, rdev->mc.gtt_end >> 12); + WREG32(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR, rdev->gart.table_addr >> 12); + WREG32(VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, + (u32)(rdev->dummy_page.addr >> 12)); + WREG32(VM_CONTEXT0_CNTL2, 0); + WREG32(VM_CONTEXT0_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(0) | + RANGE_PROTECTION_FAULT_ENABLE_DEFAULT); + /* disable context1-7 */ + WREG32(VM_CONTEXT1_CNTL2, 0); + WREG32(VM_CONTEXT1_CNTL, 0); + + cayman_pcie_gart_tlb_flush(rdev); + rdev->gart.ready = true; + return 0; +} + +void cayman_pcie_gart_disable(struct radeon_device *rdev) +{ + int r; + + /* Disable all tables */ + WREG32(VM_CONTEXT0_CNTL, 0); + WREG32(VM_CONTEXT1_CNTL, 0); + /* Setup TLB control */ + WREG32(MC_VM_MX_L1_TLB_CNTL, ENABLE_L1_FRAGMENT_PROCESSING | + SYSTEM_ACCESS_MODE_NOT_IN_SYS | + SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU); + /* Setup L2 cache */ + WREG32(VM_L2_CNTL, ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE | + ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE | + EFFECTIVE_L2_QUEUE_SIZE(7) | + CONTEXT1_IDENTITY_ACCESS_MODE(1)); + WREG32(VM_L2_CNTL2, 0); + WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY | + L2_CACHE_BIGK_FRAGMENT_SIZE(6)); + if (rdev->gart.table.vram.robj) { + r = radeon_bo_reserve(rdev->gart.table.vram.robj, false); + if (likely(r == 0)) { + radeon_bo_kunmap(rdev->gart.table.vram.robj); + radeon_bo_unpin(rdev->gart.table.vram.robj); + radeon_bo_unreserve(rdev->gart.table.vram.robj); + } + } +} + +void cayman_pcie_gart_fini(struct radeon_device *rdev) +{ + cayman_pcie_gart_disable(rdev); + radeon_gart_table_vram_free(rdev); + radeon_gart_fini(rdev); +} + +/* + * CP. + */ +static void cayman_cp_enable(struct radeon_device *rdev, bool enable) +{ + if (enable) + WREG32(CP_ME_CNTL, 0); + else { + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); + WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT)); + WREG32(SCRATCH_UMSK, 0); + } +} + +static int cayman_cp_load_microcode(struct radeon_device *rdev) +{ + const __be32 *fw_data; + int i; + + if (!rdev->me_fw || !rdev->pfp_fw) + return -EINVAL; + + cayman_cp_enable(rdev, false); + + fw_data = (const __be32 *)rdev->pfp_fw->data; + WREG32(CP_PFP_UCODE_ADDR, 0); + for (i = 0; i < CAYMAN_PFP_UCODE_SIZE; i++) + WREG32(CP_PFP_UCODE_DATA, be32_to_cpup(fw_data++)); + WREG32(CP_PFP_UCODE_ADDR, 0); + + fw_data = (const __be32 *)rdev->me_fw->data; + WREG32(CP_ME_RAM_WADDR, 0); + for (i = 0; i < CAYMAN_PM4_UCODE_SIZE; i++) + WREG32(CP_ME_RAM_DATA, be32_to_cpup(fw_data++)); + + WREG32(CP_PFP_UCODE_ADDR, 0); + WREG32(CP_ME_RAM_WADDR, 0); + WREG32(CP_ME_RAM_RADDR, 0); + return 0; +} + +static int cayman_cp_start(struct radeon_device *rdev) +{ + int r, i; + + r = radeon_ring_lock(rdev, 7); + if (r) { + DRM_ERROR("radeon: cp failed to lock ring (%d).\n", r); + return r; + } + radeon_ring_write(rdev, PACKET3(PACKET3_ME_INITIALIZE, 5)); + radeon_ring_write(rdev, 0x1); + radeon_ring_write(rdev, 0x0); + radeon_ring_write(rdev, rdev->config.cayman.max_hw_contexts - 1); + radeon_ring_write(rdev, PACKET3_ME_INITIALIZE_DEVICE_ID(1)); + radeon_ring_write(rdev, 0); + radeon_ring_write(rdev, 0); + radeon_ring_unlock_commit(rdev); + + cayman_cp_enable(rdev, true); + + r = radeon_ring_lock(rdev, cayman_default_size + 19); + if (r) { + DRM_ERROR("radeon: cp failed to lock ring (%d).\n", r); + return r; + } + + /* setup clear context state */ + radeon_ring_write(rdev, PACKET3(PACKET3_PREAMBLE_CNTL, 0)); + radeon_ring_write(rdev, PACKET3_PREAMBLE_BEGIN_CLEAR_STATE); + + for (i = 0; i < cayman_default_size; i++) + radeon_ring_write(rdev, cayman_default_state[i]); + + radeon_ring_write(rdev, PACKET3(PACKET3_PREAMBLE_CNTL, 0)); + radeon_ring_write(rdev, PACKET3_PREAMBLE_END_CLEAR_STATE); + + /* set clear context state */ + radeon_ring_write(rdev, PACKET3(PACKET3_CLEAR_STATE, 0)); + radeon_ring_write(rdev, 0); + + /* SQ_VTX_BASE_VTX_LOC */ + radeon_ring_write(rdev, 0xc0026f00); + radeon_ring_write(rdev, 0x00000000); + radeon_ring_write(rdev, 0x00000000); + radeon_ring_write(rdev, 0x00000000); + + /* Clear consts */ + radeon_ring_write(rdev, 0xc0036f00); + radeon_ring_write(rdev, 0x00000bc4); + radeon_ring_write(rdev, 0xffffffff); + radeon_ring_write(rdev, 0xffffffff); + radeon_ring_write(rdev, 0xffffffff); + + radeon_ring_write(rdev, 0xc0026900); + radeon_ring_write(rdev, 0x00000316); + radeon_ring_write(rdev, 0x0000000e); /* VGT_VERTEX_REUSE_BLOCK_CNTL */ + radeon_ring_write(rdev, 0x00000010); /* */ + + radeon_ring_unlock_commit(rdev); + + /* XXX init other rings */ + + return 0; +} + +static void cayman_cp_fini(struct radeon_device *rdev) +{ + cayman_cp_enable(rdev, false); + radeon_ring_fini(rdev); +} + +int cayman_cp_resume(struct radeon_device *rdev) +{ + u32 tmp; + u32 rb_bufsz; + int r; + + /* Reset cp; if cp is reset, then PA, SH, VGT also need to be reset */ + WREG32(GRBM_SOFT_RESET, (SOFT_RESET_CP | + SOFT_RESET_PA | + SOFT_RESET_SH | + SOFT_RESET_VGT | + SOFT_RESET_SX)); + RREG32(GRBM_SOFT_RESET); + mdelay(15); + WREG32(GRBM_SOFT_RESET, 0); + RREG32(GRBM_SOFT_RESET); + + WREG32(CP_SEM_WAIT_TIMER, 0x4); + + /* Set the write pointer delay */ + WREG32(CP_RB_WPTR_DELAY, 0); + + WREG32(CP_DEBUG, (1 << 27)); + + /* ring 0 - compute and gfx */ + /* Set ring buffer size */ + rb_bufsz = drm_order(rdev->cp.ring_size / 8); + tmp = (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz; +#ifdef __BIG_ENDIAN + tmp |= BUF_SWAP_32BIT; +#endif + WREG32(CP_RB0_CNTL, tmp); + + /* Initialize the ring buffer's read and write pointers */ + WREG32(CP_RB0_CNTL, tmp | RB_RPTR_WR_ENA); + WREG32(CP_RB0_WPTR, 0); + + /* set the wb address wether it's enabled or not */ + WREG32(CP_RB0_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFFFFFFFC); + WREG32(CP_RB0_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFF); + WREG32(SCRATCH_ADDR, ((rdev->wb.gpu_addr + RADEON_WB_SCRATCH_OFFSET) >> 8) & 0xFFFFFFFF); + + if (rdev->wb.enabled) + WREG32(SCRATCH_UMSK, 0xff); + else { + tmp |= RB_NO_UPDATE; + WREG32(SCRATCH_UMSK, 0); + } + + mdelay(1); + WREG32(CP_RB0_CNTL, tmp); + + WREG32(CP_RB0_BASE, rdev->cp.gpu_addr >> 8); + + rdev->cp.rptr = RREG32(CP_RB0_RPTR); + rdev->cp.wptr = RREG32(CP_RB0_WPTR); + + /* ring1 - compute only */ + /* Set ring buffer size */ + rb_bufsz = drm_order(rdev->cp1.ring_size / 8); + tmp = (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz; +#ifdef __BIG_ENDIAN + tmp |= BUF_SWAP_32BIT; +#endif + WREG32(CP_RB1_CNTL, tmp); + + /* Initialize the ring buffer's read and write pointers */ + WREG32(CP_RB1_CNTL, tmp | RB_RPTR_WR_ENA); + WREG32(CP_RB1_WPTR, 0); + + /* set the wb address wether it's enabled or not */ + WREG32(CP_RB1_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP1_RPTR_OFFSET) & 0xFFFFFFFC); + WREG32(CP_RB1_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP1_RPTR_OFFSET) & 0xFF); + + mdelay(1); + WREG32(CP_RB1_CNTL, tmp); + + WREG32(CP_RB1_BASE, rdev->cp1.gpu_addr >> 8); + + rdev->cp1.rptr = RREG32(CP_RB1_RPTR); + rdev->cp1.wptr = RREG32(CP_RB1_WPTR); + + /* ring2 - compute only */ + /* Set ring buffer size */ + rb_bufsz = drm_order(rdev->cp2.ring_size / 8); + tmp = (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz; +#ifdef __BIG_ENDIAN + tmp |= BUF_SWAP_32BIT; +#endif + WREG32(CP_RB2_CNTL, tmp); + + /* Initialize the ring buffer's read and write pointers */ + WREG32(CP_RB2_CNTL, tmp | RB_RPTR_WR_ENA); + WREG32(CP_RB2_WPTR, 0); + + /* set the wb address wether it's enabled or not */ + WREG32(CP_RB2_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP2_RPTR_OFFSET) & 0xFFFFFFFC); + WREG32(CP_RB2_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP2_RPTR_OFFSET) & 0xFF); + + mdelay(1); + WREG32(CP_RB2_CNTL, tmp); + + WREG32(CP_RB2_BASE, rdev->cp2.gpu_addr >> 8); + + rdev->cp2.rptr = RREG32(CP_RB2_RPTR); + rdev->cp2.wptr = RREG32(CP_RB2_WPTR); + + /* start the rings */ + cayman_cp_start(rdev); + rdev->cp.ready = true; + rdev->cp1.ready = true; + rdev->cp2.ready = true; + /* this only test cp0 */ + r = radeon_ring_test(rdev); + if (r) { + rdev->cp.ready = false; + rdev->cp1.ready = false; + rdev->cp2.ready = false; + return r; + } + + return 0; +} + +bool cayman_gpu_is_lockup(struct radeon_device *rdev) +{ + u32 srbm_status; + u32 grbm_status; + u32 grbm_status_se0, grbm_status_se1; + struct r100_gpu_lockup *lockup = &rdev->config.cayman.lockup; + int r; + + srbm_status = RREG32(SRBM_STATUS); + grbm_status = RREG32(GRBM_STATUS); + grbm_status_se0 = RREG32(GRBM_STATUS_SE0); + grbm_status_se1 = RREG32(GRBM_STATUS_SE1); + if (!(grbm_status & GUI_ACTIVE)) { + r100_gpu_lockup_update(lockup, &rdev->cp); + return false; + } + /* force CP activities */ + r = radeon_ring_lock(rdev, 2); + if (!r) { + /* PACKET2 NOP */ + radeon_ring_write(rdev, 0x80000000); + radeon_ring_write(rdev, 0x80000000); + radeon_ring_unlock_commit(rdev); + } + /* XXX deal with CP0,1,2 */ + rdev->cp.rptr = RREG32(CP_RB0_RPTR); + return r100_gpu_cp_is_lockup(rdev, lockup, &rdev->cp); +} + +static int cayman_gpu_soft_reset(struct radeon_device *rdev) +{ + struct evergreen_mc_save save; + u32 grbm_reset = 0; + + if (!(RREG32(GRBM_STATUS) & GUI_ACTIVE)) + return 0; + + dev_info(rdev->dev, "GPU softreset \n"); + dev_info(rdev->dev, " GRBM_STATUS=0x%08X\n", + RREG32(GRBM_STATUS)); + dev_info(rdev->dev, " GRBM_STATUS_SE0=0x%08X\n", + RREG32(GRBM_STATUS_SE0)); + dev_info(rdev->dev, " GRBM_STATUS_SE1=0x%08X\n", + RREG32(GRBM_STATUS_SE1)); + dev_info(rdev->dev, " SRBM_STATUS=0x%08X\n", + RREG32(SRBM_STATUS)); + evergreen_mc_stop(rdev, &save); + if (evergreen_mc_wait_for_idle(rdev)) { + dev_warn(rdev->dev, "Wait for MC idle timedout !\n"); + } + /* Disable CP parsing/prefetching */ + WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT); + + /* reset all the gfx blocks */ + grbm_reset = (SOFT_RESET_CP | + SOFT_RESET_CB | + SOFT_RESET_DB | + SOFT_RESET_GDS | + SOFT_RESET_PA | + SOFT_RESET_SC | + SOFT_RESET_SPI | + SOFT_RESET_SH | + SOFT_RESET_SX | + SOFT_RESET_TC | + SOFT_RESET_TA | + SOFT_RESET_VGT | + SOFT_RESET_IA); + + dev_info(rdev->dev, " GRBM_SOFT_RESET=0x%08X\n", grbm_reset); + WREG32(GRBM_SOFT_RESET, grbm_reset); + (void)RREG32(GRBM_SOFT_RESET); + udelay(50); + WREG32(GRBM_SOFT_RESET, 0); + (void)RREG32(GRBM_SOFT_RESET); + /* Wait a little for things to settle down */ + udelay(50); + dev_info(rdev->dev, " GRBM_STATUS=0x%08X\n", + RREG32(GRBM_STATUS)); + dev_info(rdev->dev, " GRBM_STATUS_SE0=0x%08X\n", + RREG32(GRBM_STATUS_SE0)); + dev_info(rdev->dev, " GRBM_STATUS_SE1=0x%08X\n", + RREG32(GRBM_STATUS_SE1)); + dev_info(rdev->dev, " SRBM_STATUS=0x%08X\n", + RREG32(SRBM_STATUS)); + evergreen_mc_resume(rdev, &save); + return 0; +} + +int cayman_asic_reset(struct radeon_device *rdev) +{ + return cayman_gpu_soft_reset(rdev); +} + +static int cayman_startup(struct radeon_device *rdev) +{ + int r; + + if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) { + r = ni_init_microcode(rdev); + if (r) { + DRM_ERROR("Failed to load firmware!\n"); + return r; + } + } + r = ni_mc_load_microcode(rdev); + if (r) { + DRM_ERROR("Failed to load MC firmware!\n"); + return r; + } + + evergreen_mc_program(rdev); + r = cayman_pcie_gart_enable(rdev); + if (r) + return r; + cayman_gpu_init(rdev); + +#if 0 + r = cayman_blit_init(rdev); + if (r) { + cayman_blit_fini(rdev); + rdev->asic->copy = NULL; + dev_warn(rdev->dev, "failed blitter (%d) falling back to memcpy\n", r); + } +#endif + + /* allocate wb buffer */ + r = radeon_wb_init(rdev); + if (r) + return r; + + /* Enable IRQ */ + r = r600_irq_init(rdev); + if (r) { + DRM_ERROR("radeon: IH init failed (%d).\n", r); + radeon_irq_kms_fini(rdev); + return r; + } + evergreen_irq_set(rdev); + + r = radeon_ring_init(rdev, rdev->cp.ring_size); + if (r) + return r; + r = cayman_cp_load_microcode(rdev); + if (r) + return r; + r = cayman_cp_resume(rdev); + if (r) + return r; + + return 0; +} + +int cayman_resume(struct radeon_device *rdev) +{ + int r; + + /* Do not reset GPU before posting, on rv770 hw unlike on r500 hw, + * posting will perform necessary task to bring back GPU into good + * shape. + */ + /* post card */ + atom_asic_init(rdev->mode_info.atom_context); + + r = cayman_startup(rdev); + if (r) { + DRM_ERROR("cayman startup failed on resume\n"); + return r; + } + + r = r600_ib_test(rdev); + if (r) { + DRM_ERROR("radeon: failled testing IB (%d).\n", r); + return r; + } + + return r; + +} + +int cayman_suspend(struct radeon_device *rdev) +{ + /* int r; */ + + /* FIXME: we should wait for ring to be empty */ + cayman_cp_enable(rdev, false); + rdev->cp.ready = false; + evergreen_irq_suspend(rdev); + radeon_wb_disable(rdev); + cayman_pcie_gart_disable(rdev); + +#if 0 + /* unpin shaders bo */ + r = radeon_bo_reserve(rdev->r600_blit.shader_obj, false); + if (likely(r == 0)) { + radeon_bo_unpin(rdev->r600_blit.shader_obj); + radeon_bo_unreserve(rdev->r600_blit.shader_obj); + } +#endif + return 0; +} + +/* Plan is to move initialization in that function and use + * helper function so that radeon_device_init pretty much + * do nothing more than calling asic specific function. This + * should also allow to remove a bunch of callback function + * like vram_info. + */ +int cayman_init(struct radeon_device *rdev) +{ + int r; + + /* This don't do much */ + r = radeon_gem_init(rdev); + if (r) + return r; + /* Read BIOS */ + if (!radeon_get_bios(rdev)) { + if (ASIC_IS_AVIVO(rdev)) + return -EINVAL; + } + /* Must be an ATOMBIOS */ + if (!rdev->is_atom_bios) { + dev_err(rdev->dev, "Expecting atombios for cayman GPU\n"); + return -EINVAL; + } + r = radeon_atombios_init(rdev); + if (r) + return r; + + /* Post card if necessary */ + if (!radeon_card_posted(rdev)) { + if (!rdev->bios) { + dev_err(rdev->dev, "Card not posted and no BIOS - ignoring\n"); + return -EINVAL; + } + DRM_INFO("GPU not posted. posting now...\n"); + atom_asic_init(rdev->mode_info.atom_context); + } + /* Initialize scratch registers */ + r600_scratch_init(rdev); + /* Initialize surface registers */ + radeon_surface_init(rdev); + /* Initialize clocks */ + radeon_get_clock_info(rdev->ddev); + /* Fence driver */ + r = radeon_fence_driver_init(rdev); + if (r) + return r; + /* initialize memory controller */ + r = evergreen_mc_init(rdev); + if (r) + return r; + /* Memory manager */ + r = radeon_bo_init(rdev); + if (r) + return r; + + r = radeon_irq_kms_init(rdev); + if (r) + return r; + + rdev->cp.ring_obj = NULL; + r600_ring_init(rdev, 1024 * 1024); + + rdev->ih.ring_obj = NULL; + r600_ih_ring_init(rdev, 64 * 1024); + + r = r600_pcie_gart_init(rdev); + if (r) + return r; + + rdev->accel_working = true; + r = cayman_startup(rdev); + if (r) { + dev_err(rdev->dev, "disabling GPU acceleration\n"); + cayman_cp_fini(rdev); + r600_irq_fini(rdev); + radeon_wb_fini(rdev); + radeon_irq_kms_fini(rdev); + cayman_pcie_gart_fini(rdev); + rdev->accel_working = false; + } + if (rdev->accel_working) { + r = radeon_ib_pool_init(rdev); + if (r) { + DRM_ERROR("radeon: failed initializing IB pool (%d).\n", r); + rdev->accel_working = false; + } + r = r600_ib_test(rdev); + if (r) { + DRM_ERROR("radeon: failed testing IB (%d).\n", r); + rdev->accel_working = false; + } + } + + /* Don't start up if the MC ucode is missing. + * The default clocks and voltages before the MC ucode + * is loaded are not suffient for advanced operations. + */ + if (!rdev->mc_fw) { + DRM_ERROR("radeon: MC ucode required for NI+.\n"); + return -EINVAL; + } + + return 0; +} + +void cayman_fini(struct radeon_device *rdev) +{ + /* cayman_blit_fini(rdev); */ + cayman_cp_fini(rdev); + r600_irq_fini(rdev); + radeon_wb_fini(rdev); + radeon_irq_kms_fini(rdev); + cayman_pcie_gart_fini(rdev); + radeon_gem_fini(rdev); + radeon_fence_driver_fini(rdev); + radeon_bo_fini(rdev); + radeon_atombios_fini(rdev); + kfree(rdev->bios); + rdev->bios = NULL; +} + diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h index f7b4453..0f9a08b 100644 --- a/drivers/gpu/drm/radeon/nid.h +++ b/drivers/gpu/drm/radeon/nid.h @@ -24,7 +24,101 @@ #ifndef NI_H #define NI_H +#define CAYMAN_MAX_SH_GPRS 256 +#define CAYMAN_MAX_TEMP_GPRS 16 +#define CAYMAN_MAX_SH_THREADS 256 +#define CAYMAN_MAX_SH_STACK_ENTRIES 4096 +#define CAYMAN_MAX_FRC_EOV_CNT 16384 +#define CAYMAN_MAX_BACKENDS 8 +#define CAYMAN_MAX_BACKENDS_MASK 0xFF +#define CAYMAN_MAX_BACKENDS_PER_SE_MASK 0xF +#define CAYMAN_MAX_SIMDS 16 +#define CAYMAN_MAX_SIMDS_MASK 0xFFFF +#define CAYMAN_MAX_SIMDS_PER_SE_MASK 0xFFF +#define CAYMAN_MAX_PIPES 8 +#define CAYMAN_MAX_PIPES_MASK 0xFF +#define CAYMAN_MAX_LDS_NUM 0xFFFF +#define CAYMAN_MAX_TCC 16 +#define CAYMAN_MAX_TCC_MASK 0xFF + +#define DMIF_ADDR_CONFIG 0xBD4 +#define SRBM_STATUS 0x0E50 + +#define VM_CONTEXT0_REQUEST_RESPONSE 0x1470 +#define REQUEST_TYPE(x) (((x) & 0xf) << 0) +#define RESPONSE_TYPE_MASK 0x000000F0 +#define RESPONSE_TYPE_SHIFT 4 +#define VM_L2_CNTL 0x1400 +#define ENABLE_L2_CACHE (1 << 0) +#define ENABLE_L2_FRAGMENT_PROCESSING (1 << 1) +#define ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE (1 << 9) +#define ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE (1 << 10) +#define EFFECTIVE_L2_QUEUE_SIZE(x) (((x) & 7) << 14) +#define CONTEXT1_IDENTITY_ACCESS_MODE(x) (((x) & 3) << 18) +/* CONTEXT1_IDENTITY_ACCESS_MODE + * 0 physical = logical + * 1 logical via context1 page table + * 2 inside identity aperture use translation, outside physical = logical + * 3 inside identity aperture physical = logical, outside use translation + */ +#define VM_L2_CNTL2 0x1404 +#define INVALIDATE_ALL_L1_TLBS (1 << 0) +#define INVALIDATE_L2_CACHE (1 << 1) +#define VM_L2_CNTL3 0x1408 +#define BANK_SELECT(x) ((x) << 0) +#define CACHE_UPDATE_MODE(x) ((x) << 6) +#define L2_CACHE_BIGK_ASSOCIATIVITY (1 << 20) +#define L2_CACHE_BIGK_FRAGMENT_SIZE(x) ((x) << 15) +#define VM_L2_STATUS 0x140C +#define L2_BUSY (1 << 0) +#define VM_CONTEXT0_CNTL 0x1410 +#define ENABLE_CONTEXT (1 << 0) +#define PAGE_TABLE_DEPTH(x) (((x) & 3) << 1) +#define RANGE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 4) +#define VM_CONTEXT1_CNTL 0x1414 +#define VM_CONTEXT0_CNTL2 0x1430 +#define VM_CONTEXT1_CNTL2 0x1434 +#define VM_INVALIDATE_REQUEST 0x1478 +#define VM_INVALIDATE_RESPONSE 0x147c +#define VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR 0x1518 +#define VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR 0x151c +#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR 0x153C +#define VM_CONTEXT0_PAGE_TABLE_START_ADDR 0x155C +#define VM_CONTEXT0_PAGE_TABLE_END_ADDR 0x157C + +#define MC_SHARED_CHMAP 0x2004 +#define NOOFCHAN_SHIFT 12 +#define NOOFCHAN_MASK 0x00003000 +#define MC_SHARED_CHREMAP 0x2008 + +#define MC_VM_SYSTEM_APERTURE_LOW_ADDR 0x2034 +#define MC_VM_SYSTEM_APERTURE_HIGH_ADDR 0x2038 +#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR 0x203C +#define MC_VM_MX_L1_TLB_CNTL 0x2064 +#define ENABLE_L1_TLB (1 << 0) +#define ENABLE_L1_FRAGMENT_PROCESSING (1 << 1) +#define SYSTEM_ACCESS_MODE_PA_ONLY (0 << 3) +#define SYSTEM_ACCESS_MODE_USE_SYS_MAP (1 << 3) +#define SYSTEM_ACCESS_MODE_IN_SYS (2 << 3) +#define SYSTEM_ACCESS_MODE_NOT_IN_SYS (3 << 3) +#define SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU (0 << 5) +#define ENABLE_ADVANCED_DRIVER_MODEL (1 << 6) + #define MC_SHARED_BLACKOUT_CNTL 0x20ac +#define MC_ARB_RAMCFG 0x2760 +#define NOOFBANK_SHIFT 0 +#define NOOFBANK_MASK 0x00000003 +#define NOOFRANK_SHIFT 2 +#define NOOFRANK_MASK 0x00000004 +#define NOOFROWS_SHIFT 3 +#define NOOFROWS_MASK 0x00000038 +#define NOOFCOLS_SHIFT 6 +#define NOOFCOLS_MASK 0x000000C0 +#define CHANSIZE_SHIFT 8 +#define CHANSIZE_MASK 0x00000100 +#define BURSTLENGTH_SHIFT 9 +#define BURSTLENGTH_MASK 0x00000200 +#define CHANSIZE_OVERRIDE (1 << 11) #define MC_SEQ_SUP_CNTL 0x28c8 #define RUN_MASK (1 << 0) #define MC_SEQ_SUP_PGM 0x28cc @@ -37,5 +131,406 @@ #define MC_SEQ_IO_DEBUG_INDEX 0x2a44 #define MC_SEQ_IO_DEBUG_DATA 0x2a48 +#define HDP_HOST_PATH_CNTL 0x2C00 +#define HDP_NONSURFACE_BASE 0x2C04 +#define HDP_NONSURFACE_INFO 0x2C08 +#define HDP_NONSURFACE_SIZE 0x2C0C +#define HDP_ADDR_CONFIG 0x2F48 + +#define CC_SYS_RB_BACKEND_DISABLE 0x3F88 +#define GC_USER_SYS_RB_BACKEND_DISABLE 0x3F8C +#define CGTS_SYS_TCC_DISABLE 0x3F90 +#define CGTS_USER_SYS_TCC_DISABLE 0x3F94 + +#define CONFIG_MEMSIZE 0x5428 + +#define HDP_MEM_COHERENCY_FLUSH_CNTL 0x5480 +#define HDP_REG_COHERENCY_FLUSH_CNTL 0x54A0 + +#define GRBM_CNTL 0x8000 +#define GRBM_READ_TIMEOUT(x) ((x) << 0) +#define GRBM_STATUS 0x8010 +#define CMDFIFO_AVAIL_MASK 0x0000000F +#define RING2_RQ_PENDING (1 << 4) +#define SRBM_RQ_PENDING (1 << 5) +#define RING1_RQ_PENDING (1 << 6) +#define CF_RQ_PENDING (1 << 7) +#define PF_RQ_PENDING (1 << 8) +#define GDS_DMA_RQ_PENDING (1 << 9) +#define GRBM_EE_BUSY (1 << 10) +#define SX_CLEAN (1 << 11) +#define DB_CLEAN (1 << 12) +#define CB_CLEAN (1 << 13) +#define TA_BUSY (1 << 14) +#define GDS_BUSY (1 << 15) +#define VGT_BUSY_NO_DMA (1 << 16) +#define VGT_BUSY (1 << 17) +#define IA_BUSY_NO_DMA (1 << 18) +#define IA_BUSY (1 << 19) +#define SX_BUSY (1 << 20) +#define SH_BUSY (1 << 21) +#define SPI_BUSY (1 << 22) +#define SC_BUSY (1 << 24) +#define PA_BUSY (1 << 25) +#define DB_BUSY (1 << 26) +#define CP_COHERENCY_BUSY (1 << 28) +#define CP_BUSY (1 << 29) +#define CB_BUSY (1 << 30) +#define GUI_ACTIVE (1 << 31) +#define GRBM_STATUS_SE0 0x8014 +#define GRBM_STATUS_SE1 0x8018 +#define SE_SX_CLEAN (1 << 0) +#define SE_DB_CLEAN (1 << 1) +#define SE_CB_CLEAN (1 << 2) +#define SE_VGT_BUSY (1 << 23) +#define SE_PA_BUSY (1 << 24) +#define SE_TA_BUSY (1 << 25) +#define SE_SX_BUSY (1 << 26) +#define SE_SPI_BUSY (1 << 27) +#define SE_SH_BUSY (1 << 28) +#define SE_SC_BUSY (1 << 29) +#define SE_DB_BUSY (1 << 30) +#define SE_CB_BUSY (1 << 31) +#define GRBM_SOFT_RESET 0x8020 +#define SOFT_RESET_CP (1 << 0) +#define SOFT_RESET_CB (1 << 1) +#define SOFT_RESET_DB (1 << 3) +#define SOFT_RESET_GDS (1 << 4) +#define SOFT_RESET_PA (1 << 5) +#define SOFT_RESET_SC (1 << 6) +#define SOFT_RESET_SPI (1 << 8) +#define SOFT_RESET_SH (1 << 9) +#define SOFT_RESET_SX (1 << 10) +#define SOFT_RESET_TC (1 << 11) +#define SOFT_RESET_TA (1 << 12) +#define SOFT_RESET_VGT (1 << 14) +#define SOFT_RESET_IA (1 << 15) + +#define SCRATCH_REG0 0x8500 +#define SCRATCH_REG1 0x8504 +#define SCRATCH_REG2 0x8508 +#define SCRATCH_REG3 0x850C +#define SCRATCH_REG4 0x8510 +#define SCRATCH_REG5 0x8514 +#define SCRATCH_REG6 0x8518 +#define SCRATCH_REG7 0x851C +#define SCRATCH_UMSK 0x8540 +#define SCRATCH_ADDR 0x8544 +#define CP_SEM_WAIT_TIMER 0x85BC +#define CP_ME_CNTL 0x86D8 +#define CP_ME_HALT (1 << 28) +#define CP_PFP_HALT (1 << 26) +#define CP_RB2_RPTR 0x86f8 +#define CP_RB1_RPTR 0x86fc +#define CP_RB0_RPTR 0x8700 +#define CP_RB_WPTR_DELAY 0x8704 +#define CP_MEQ_THRESHOLDS 0x8764 +#define MEQ1_START(x) ((x) << 0) +#define MEQ2_START(x) ((x) << 8) +#define CP_PERFMON_CNTL 0x87FC + +#define VGT_CACHE_INVALIDATION 0x88C4 +#define CACHE_INVALIDATION(x) ((x) << 0) +#define VC_ONLY 0 +#define TC_ONLY 1 +#define VC_AND_TC 2 +#define AUTO_INVLD_EN(x) ((x) << 6) +#define NO_AUTO 0 +#define ES_AUTO 1 +#define GS_AUTO 2 +#define ES_AND_GS_AUTO 3 +#define VGT_GS_VERTEX_REUSE 0x88D4 + +#define CC_GC_SHADER_PIPE_CONFIG 0x8950 +#define GC_USER_SHADER_PIPE_CONFIG 0x8954 +#define INACTIVE_QD_PIPES(x) ((x) << 8) +#define INACTIVE_QD_PIPES_MASK 0x0000FF00 +#define INACTIVE_QD_PIPES_SHIFT 8 +#define INACTIVE_SIMDS(x) ((x) << 16) +#define INACTIVE_SIMDS_MASK 0xFFFF0000 +#define INACTIVE_SIMDS_SHIFT 16 + +#define VGT_PRIMITIVE_TYPE 0x8958 +#define VGT_NUM_INSTANCES 0x8974 +#define VGT_TF_RING_SIZE 0x8988 +#define VGT_OFFCHIP_LDS_BASE 0x89b4 + +#define PA_SC_LINE_STIPPLE_STATE 0x8B10 +#define PA_CL_ENHANCE 0x8A14 +#define CLIP_VTX_REORDER_ENA (1 << 0) +#define NUM_CLIP_SEQ(x) ((x) << 1) +#define PA_SC_FIFO_SIZE 0x8BCC +#define SC_PRIM_FIFO_SIZE(x) ((x) << 0) +#define SC_HIZ_TILE_FIFO_SIZE(x) ((x) << 12) +#define SC_EARLYZ_TILE_FIFO_SIZE(x) ((x) << 20) +#define PA_SC_FORCE_EOV_MAX_CNTS 0x8B24 +#define FORCE_EOV_MAX_CLK_CNT(x) ((x) << 0) +#define FORCE_EOV_MAX_REZ_CNT(x) ((x) << 16) + +#define SQ_CONFIG 0x8C00 +#define VC_ENABLE (1 << 0) +#define EXPORT_SRC_C (1 << 1) +#define GFX_PRIO(x) ((x) << 2) +#define CS1_PRIO(x) ((x) << 4) +#define CS2_PRIO(x) ((x) << 6) +#define SQ_GPR_RESOURCE_MGMT_1 0x8C04 +#define NUM_PS_GPRS(x) ((x) << 0) +#define NUM_VS_GPRS(x) ((x) << 16) +#define NUM_CLAUSE_TEMP_GPRS(x) ((x) << 28) +#define SQ_ESGS_RING_SIZE 0x8c44 +#define SQ_GSVS_RING_SIZE 0x8c4c +#define SQ_ESTMP_RING_BASE 0x8c50 +#define SQ_ESTMP_RING_SIZE 0x8c54 +#define SQ_GSTMP_RING_BASE 0x8c58 +#define SQ_GSTMP_RING_SIZE 0x8c5c +#define SQ_VSTMP_RING_BASE 0x8c60 +#define SQ_VSTMP_RING_SIZE 0x8c64 +#define SQ_PSTMP_RING_BASE 0x8c68 +#define SQ_PSTMP_RING_SIZE 0x8c6c +#define SQ_MS_FIFO_SIZES 0x8CF0 +#define CACHE_FIFO_SIZE(x) ((x) << 0) +#define FETCH_FIFO_HIWATER(x) ((x) << 8) +#define DONE_FIFO_HIWATER(x) ((x) << 16) +#define ALU_UPDATE_FIFO_HIWATER(x) ((x) << 24) +#define SQ_LSTMP_RING_BASE 0x8e10 +#define SQ_LSTMP_RING_SIZE 0x8e14 +#define SQ_HSTMP_RING_BASE 0x8e18 +#define SQ_HSTMP_RING_SIZE 0x8e1c +#define SQ_DYN_GPR_CNTL_PS_FLUSH_REQ 0x8D8C +#define DYN_GPR_ENABLE (1 << 8) +#define SQ_CONST_MEM_BASE 0x8df8 + +#define SX_EXPORT_BUFFER_SIZES 0x900C +#define COLOR_BUFFER_SIZE(x) ((x) << 0) +#define POSITION_BUFFER_SIZE(x) ((x) << 8) +#define SMX_BUFFER_SIZE(x) ((x) << 16) +#define SX_DEBUG_1 0x9058 +#define ENABLE_NEW_SMX_ADDRESS (1 << 16) + +#define SPI_CONFIG_CNTL 0x9100 +#define GPR_WRITE_PRIORITY(x) ((x) << 0) +#define SPI_CONFIG_CNTL_1 0x913C +#define VTX_DONE_DELAY(x) ((x) << 0) +#define INTERP_ONE_PRIM_PER_ROW (1 << 4) +#define CRC_SIMD_ID_WADDR_DISABLE (1 << 8) + +#define CGTS_TCC_DISABLE 0x9148 +#define CGTS_USER_TCC_DISABLE 0x914C +#define TCC_DISABLE_MASK 0xFFFF0000 +#define TCC_DISABLE_SHIFT 16 +#define CGTS_SM_CTRL_REG 0x915C +#define OVERRIDE (1 << 21) + +#define TA_CNTL_AUX 0x9508 +#define DISABLE_CUBE_WRAP (1 << 0) +#define DISABLE_CUBE_ANISO (1 << 1) + +#define TCP_CHAN_STEER_LO 0x960c +#define TCP_CHAN_STEER_HI 0x9610 + +#define CC_RB_BACKEND_DISABLE 0x98F4 +#define BACKEND_DISABLE(x) ((x) << 16) +#define GB_ADDR_CONFIG 0x98F8 +#define NUM_PIPES(x) ((x) << 0) +#define NUM_PIPES_MASK 0x00000007 +#define NUM_PIPES_SHIFT 0 +#define PIPE_INTERLEAVE_SIZE(x) ((x) << 4) +#define PIPE_INTERLEAVE_SIZE_MASK 0x00000070 +#define PIPE_INTERLEAVE_SIZE_SHIFT 4 +#define BANK_INTERLEAVE_SIZE(x) ((x) << 8) +#define NUM_SHADER_ENGINES(x) ((x) << 12) +#define NUM_SHADER_ENGINES_MASK 0x00003000 +#define NUM_SHADER_ENGINES_SHIFT 12 +#define SHADER_ENGINE_TILE_SIZE(x) ((x) << 16) +#define SHADER_ENGINE_TILE_SIZE_MASK 0x00070000 +#define SHADER_ENGINE_TILE_SIZE_SHIFT 16 +#define NUM_GPUS(x) ((x) << 20) +#define NUM_GPUS_MASK 0x00700000 +#define NUM_GPUS_SHIFT 20 +#define MULTI_GPU_TILE_SIZE(x) ((x) << 24) +#define MULTI_GPU_TILE_SIZE_MASK 0x03000000 +#define MULTI_GPU_TILE_SIZE_SHIFT 24 +#define ROW_SIZE(x) ((x) << 28) +#define ROW_SIZE_MASK 0x30000007 +#define ROW_SIZE_SHIFT 28 +#define NUM_LOWER_PIPES(x) ((x) << 30) +#define NUM_LOWER_PIPES_MASK 0x40000000 +#define NUM_LOWER_PIPES_SHIFT 30 +#define GB_BACKEND_MAP 0x98FC + +#define CB_PERF_CTR0_SEL_0 0x9A20 +#define CB_PERF_CTR0_SEL_1 0x9A24 +#define CB_PERF_CTR1_SEL_0 0x9A28 +#define CB_PERF_CTR1_SEL_1 0x9A2C +#define CB_PERF_CTR2_SEL_0 0x9A30 +#define CB_PERF_CTR2_SEL_1 0x9A34 +#define CB_PERF_CTR3_SEL_0 0x9A38 +#define CB_PERF_CTR3_SEL_1 0x9A3C + +#define GC_USER_RB_BACKEND_DISABLE 0x9B7C +#define BACKEND_DISABLE_MASK 0x00FF0000 +#define BACKEND_DISABLE_SHIFT 16 + +#define SMX_DC_CTL0 0xA020 +#define USE_HASH_FUNCTION (1 << 0) +#define NUMBER_OF_SETS(x) ((x) << 1) +#define FLUSH_ALL_ON_EVENT (1 << 10) +#define STALL_ON_EVENT (1 << 11) +#define SMX_EVENT_CTL 0xA02C +#define ES_FLUSH_CTL(x) ((x) << 0) +#define GS_FLUSH_CTL(x) ((x) << 3) +#define ACK_FLUSH_CTL(x) ((x) << 6) +#define SYNC_FLUSH_CTL (1 << 8) + +#define CP_RB0_BASE 0xC100 +#define CP_RB0_CNTL 0xC104 +#define RB_BUFSZ(x) ((x) << 0) +#define RB_BLKSZ(x) ((x) << 8) +#define RB_NO_UPDATE (1 << 27) +#define RB_RPTR_WR_ENA (1 << 31) +#define BUF_SWAP_32BIT (2 << 16) +#define CP_RB0_RPTR_ADDR 0xC10C +#define CP_RB0_RPTR_ADDR_HI 0xC110 +#define CP_RB0_WPTR 0xC114 +#define CP_RB1_BASE 0xC180 +#define CP_RB1_CNTL 0xC184 +#define CP_RB1_RPTR_ADDR 0xC188 +#define CP_RB1_RPTR_ADDR_HI 0xC18C +#define CP_RB1_WPTR 0xC190 +#define CP_RB2_BASE 0xC194 +#define CP_RB2_CNTL 0xC198 +#define CP_RB2_RPTR_ADDR 0xC19C +#define CP_RB2_RPTR_ADDR_HI 0xC1A0 +#define CP_RB2_WPTR 0xC1A4 +#define CP_PFP_UCODE_ADDR 0xC150 +#define CP_PFP_UCODE_DATA 0xC154 +#define CP_ME_RAM_RADDR 0xC158 +#define CP_ME_RAM_WADDR 0xC15C +#define CP_ME_RAM_DATA 0xC160 +#define CP_DEBUG 0xC1FC + +/* + * PM4 + */ +#define PACKET_TYPE0 0 +#define PACKET_TYPE1 1 +#define PACKET_TYPE2 2 +#define PACKET_TYPE3 3 + +#define CP_PACKET_GET_TYPE(h) (((h) >> 30) & 3) +#define CP_PACKET_GET_COUNT(h) (((h) >> 16) & 0x3FFF) +#define CP_PACKET0_GET_REG(h) (((h) & 0xFFFF) << 2) +#define CP_PACKET3_GET_OPCODE(h) (((h) >> 8) & 0xFF) +#define PACKET0(reg, n) ((PACKET_TYPE0 << 30) | \ + (((reg) >> 2) & 0xFFFF) | \ + ((n) & 0x3FFF) << 16) +#define CP_PACKET2 0x80000000 +#define PACKET2_PAD_SHIFT 0 +#define PACKET2_PAD_MASK (0x3fffffff << 0) + +#define PACKET2(v) (CP_PACKET2 | REG_SET(PACKET2_PAD, (v))) + +#define PACKET3(op, n) ((PACKET_TYPE3 << 30) | \ + (((op) & 0xFF) << 8) | \ + ((n) & 0x3FFF) << 16) + +/* Packet 3 types */ +#define PACKET3_NOP 0x10 +#define PACKET3_SET_BASE 0x11 +#define PACKET3_CLEAR_STATE 0x12 +#define PACKET3_INDEX_BUFFER_SIZE 0x13 +#define PACKET3_DEALLOC_STATE 0x14 +#define PACKET3_DISPATCH_DIRECT 0x15 +#define PACKET3_DISPATCH_INDIRECT 0x16 +#define PACKET3_INDIRECT_BUFFER_END 0x17 +#define PACKET3_SET_PREDICATION 0x20 +#define PACKET3_REG_RMW 0x21 +#define PACKET3_COND_EXEC 0x22 +#define PACKET3_PRED_EXEC 0x23 +#define PACKET3_DRAW_INDIRECT 0x24 +#define PACKET3_DRAW_INDEX_INDIRECT 0x25 +#define PACKET3_INDEX_BASE 0x26 +#define PACKET3_DRAW_INDEX_2 0x27 +#define PACKET3_CONTEXT_CONTROL 0x28 +#define PACKET3_DRAW_INDEX_OFFSET 0x29 +#define PACKET3_INDEX_TYPE 0x2A +#define PACKET3_DRAW_INDEX 0x2B +#define PACKET3_DRAW_INDEX_AUTO 0x2D +#define PACKET3_DRAW_INDEX_IMMD 0x2E +#define PACKET3_NUM_INSTANCES 0x2F +#define PACKET3_DRAW_INDEX_MULTI_AUTO 0x30 +#define PACKET3_INDIRECT_BUFFER 0x32 +#define PACKET3_STRMOUT_BUFFER_UPDATE 0x34 +#define PACKET3_DRAW_INDEX_OFFSET_2 0x35 +#define PACKET3_DRAW_INDEX_MULTI_ELEMENT 0x36 +#define PACKET3_WRITE_DATA 0x37 +#define PACKET3_MEM_SEMAPHORE 0x39 +#define PACKET3_MPEG_INDEX 0x3A +#define PACKET3_WAIT_REG_MEM 0x3C +#define PACKET3_MEM_WRITE 0x3D +#define PACKET3_SURFACE_SYNC 0x43 +# define PACKET3_CB0_DEST_BASE_ENA (1 << 6) +# define PACKET3_CB1_DEST_BASE_ENA (1 << 7) +# define PACKET3_CB2_DEST_BASE_ENA (1 << 8) +# define PACKET3_CB3_DEST_BASE_ENA (1 << 9) +# define PACKET3_CB4_DEST_BASE_ENA (1 << 10) +# define PACKET3_CB5_DEST_BASE_ENA (1 << 11) +# define PACKET3_CB6_DEST_BASE_ENA (1 << 12) +# define PACKET3_CB7_DEST_BASE_ENA (1 << 13) +# define PACKET3_DB_DEST_BASE_ENA (1 << 14) +# define PACKET3_CB8_DEST_BASE_ENA (1 << 15) +# define PACKET3_CB9_DEST_BASE_ENA (1 << 16) +# define PACKET3_CB10_DEST_BASE_ENA (1 << 17) +# define PACKET3_CB11_DEST_BASE_ENA (1 << 18) +# define PACKET3_FULL_CACHE_ENA (1 << 20) +# define PACKET3_TC_ACTION_ENA (1 << 23) +# define PACKET3_CB_ACTION_ENA (1 << 25) +# define PACKET3_DB_ACTION_ENA (1 << 26) +# define PACKET3_SH_ACTION_ENA (1 << 27) +# define PACKET3_SX_ACTION_ENA (1 << 28) +#define PACKET3_ME_INITIALIZE 0x44 +#define PACKET3_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16) +#define PACKET3_COND_WRITE 0x45 +#define PACKET3_EVENT_WRITE 0x46 +#define PACKET3_EVENT_WRITE_EOP 0x47 +#define PACKET3_EVENT_WRITE_EOS 0x48 +#define PACKET3_PREAMBLE_CNTL 0x4A +# define PACKET3_PREAMBLE_BEGIN_CLEAR_STATE (2 << 28) +# define PACKET3_PREAMBLE_END_CLEAR_STATE (3 << 28) +#define PACKET3_ALU_PS_CONST_BUFFER_COPY 0x4C +#define PACKET3_ALU_VS_CONST_BUFFER_COPY 0x4D +#define PACKET3_ALU_PS_CONST_UPDATE 0x4E +#define PACKET3_ALU_VS_CONST_UPDATE 0x4F +#define PACKET3_ONE_REG_WRITE 0x57 +#define PACKET3_SET_CONFIG_REG 0x68 +#define PACKET3_SET_CONFIG_REG_START 0x00008000 +#define PACKET3_SET_CONFIG_REG_END 0x0000ac00 +#define PACKET3_SET_CONTEXT_REG 0x69 +#define PACKET3_SET_CONTEXT_REG_START 0x00028000 +#define PACKET3_SET_CONTEXT_REG_END 0x00029000 +#define PACKET3_SET_ALU_CONST 0x6A +/* alu const buffers only; no reg file */ +#define PACKET3_SET_BOOL_CONST 0x6B +#define PACKET3_SET_BOOL_CONST_START 0x0003a500 +#define PACKET3_SET_BOOL_CONST_END 0x0003a518 +#define PACKET3_SET_LOOP_CONST 0x6C +#define PACKET3_SET_LOOP_CONST_START 0x0003a200 +#define PACKET3_SET_LOOP_CONST_END 0x0003a500 +#define PACKET3_SET_RESOURCE 0x6D +#define PACKET3_SET_RESOURCE_START 0x00030000 +#define PACKET3_SET_RESOURCE_END 0x00038000 +#define PACKET3_SET_SAMPLER 0x6E +#define PACKET3_SET_SAMPLER_START 0x0003c000 +#define PACKET3_SET_SAMPLER_END 0x0003c600 +#define PACKET3_SET_CTL_CONST 0x6F +#define PACKET3_SET_CTL_CONST_START 0x0003cff0 +#define PACKET3_SET_CTL_CONST_END 0x0003ff0c +#define PACKET3_SET_RESOURCE_OFFSET 0x70 +#define PACKET3_SET_ALU_CONST_VS 0x71 +#define PACKET3_SET_ALU_CONST_DI 0x72 +#define PACKET3_SET_CONTEXT_REG_INDIRECT 0x73 +#define PACKET3_SET_RESOURCE_INDIRECT 0x74 +#define PACKET3_SET_APPEND_CNT 0x75 + #endif diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index a78d09f..f2204cb 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -70,23 +70,6 @@ MODULE_FIRMWARE(FIRMWARE_R520); void r100_pre_page_flip(struct radeon_device *rdev, int crtc) { - struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc]; - u32 tmp; - - /* make sure flip is at vb rather than hb */ - tmp = RREG32(RADEON_CRTC_OFFSET_CNTL + radeon_crtc->crtc_offset); - tmp &= ~RADEON_CRTC_OFFSET_FLIP_CNTL; - /* make sure pending bit is asserted */ - tmp |= RADEON_CRTC_GUI_TRIG_OFFSET_LEFT_EN; - WREG32(RADEON_CRTC_OFFSET_CNTL + radeon_crtc->crtc_offset, tmp); - - /* set pageflip to happen as late as possible in the vblank interval. - * same field for crtc1/2 - */ - tmp = RREG32(RADEON_CRTC_GEN_CNTL); - tmp &= ~RADEON_CRTC_VSTAT_MODE_MASK; - WREG32(RADEON_CRTC_GEN_CNTL, tmp); - /* enable the pflip int */ radeon_irq_kms_pflip_irq_get(rdev, crtc); } @@ -1041,7 +1024,7 @@ int r100_cp_init(struct radeon_device *rdev, unsigned ring_size) return r; } rdev->cp.ready = true; - rdev->mc.active_vram_size = rdev->mc.real_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size); return 0; } @@ -1059,7 +1042,7 @@ void r100_cp_fini(struct radeon_device *rdev) void r100_cp_disable(struct radeon_device *rdev) { /* Disable ring */ - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); rdev->cp.ready = false; WREG32(RADEON_CP_CSQ_MODE, 0); WREG32(RADEON_CP_CSQ_CNTL, 0); @@ -1222,14 +1205,12 @@ int r100_cs_packet_parse_vline(struct radeon_cs_parser *p) if (waitreloc.reg != RADEON_WAIT_UNTIL || waitreloc.count != 0) { DRM_ERROR("vline wait had illegal wait until segment\n"); - r = -EINVAL; - return r; + return -EINVAL; } if (radeon_get_ib_value(p, waitreloc.idx + 1) != RADEON_WAIT_CRTC_VLINE) { DRM_ERROR("vline wait had illegal wait until\n"); - r = -EINVAL; - return r; + return -EINVAL; } /* jump over the NOP */ @@ -1247,8 +1228,7 @@ int r100_cs_packet_parse_vline(struct radeon_cs_parser *p) obj = drm_mode_object_find(p->rdev->ddev, crtc_id, DRM_MODE_OBJECT_CRTC); if (!obj) { DRM_ERROR("cannot find crtc %d\n", crtc_id); - r = -EINVAL; - goto out; + return -EINVAL; } crtc = obj_to_crtc(obj); radeon_crtc = to_radeon_crtc(crtc); @@ -1270,14 +1250,13 @@ int r100_cs_packet_parse_vline(struct radeon_cs_parser *p) break; default: DRM_ERROR("unknown crtc reloc\n"); - r = -EINVAL; - goto out; + return -EINVAL; } ib[h_idx] = header; ib[h_idx + 3] |= RADEON_ENG_DISPLAY_SELECT_CRTC1; } -out: - return r; + + return 0; } /** @@ -1427,6 +1406,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, } track->zb.robj = reloc->robj; track->zb.offset = idx_value; + track->zb_dirty = true; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); break; case RADEON_RB3D_COLOROFFSET: @@ -1439,6 +1419,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, } track->cb[0].robj = reloc->robj; track->cb[0].offset = idx_value; + track->cb_dirty = true; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); break; case RADEON_PP_TXOFFSET_0: @@ -1454,6 +1435,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, } ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); track->textures[i].robj = reloc->robj; + track->tex_dirty = true; break; case RADEON_PP_CUBIC_OFFSET_T0_0: case RADEON_PP_CUBIC_OFFSET_T0_1: @@ -1471,6 +1453,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, track->textures[0].cube_info[i].offset = idx_value; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); track->textures[0].cube_info[i].robj = reloc->robj; + track->tex_dirty = true; break; case RADEON_PP_CUBIC_OFFSET_T1_0: case RADEON_PP_CUBIC_OFFSET_T1_1: @@ -1488,6 +1471,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, track->textures[1].cube_info[i].offset = idx_value; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); track->textures[1].cube_info[i].robj = reloc->robj; + track->tex_dirty = true; break; case RADEON_PP_CUBIC_OFFSET_T2_0: case RADEON_PP_CUBIC_OFFSET_T2_1: @@ -1505,9 +1489,12 @@ static int r100_packet0_check(struct radeon_cs_parser *p, track->textures[2].cube_info[i].offset = idx_value; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); track->textures[2].cube_info[i].robj = reloc->robj; + track->tex_dirty = true; break; case RADEON_RE_WIDTH_HEIGHT: track->maxy = ((idx_value >> 16) & 0x7FF); + track->cb_dirty = true; + track->zb_dirty = true; break; case RADEON_RB3D_COLORPITCH: r = r100_cs_packet_next_reloc(p, &reloc); @@ -1528,9 +1515,11 @@ static int r100_packet0_check(struct radeon_cs_parser *p, ib[idx] = tmp; track->cb[0].pitch = idx_value & RADEON_COLORPITCH_MASK; + track->cb_dirty = true; break; case RADEON_RB3D_DEPTHPITCH: track->zb.pitch = idx_value & RADEON_DEPTHPITCH_MASK; + track->zb_dirty = true; break; case RADEON_RB3D_CNTL: switch ((idx_value >> RADEON_RB3D_COLOR_FORMAT_SHIFT) & 0x1f) { @@ -1555,6 +1544,8 @@ static int r100_packet0_check(struct radeon_cs_parser *p, return -EINVAL; } track->z_enabled = !!(idx_value & RADEON_Z_ENABLE); + track->cb_dirty = true; + track->zb_dirty = true; break; case RADEON_RB3D_ZSTENCILCNTL: switch (idx_value & 0xf) { @@ -1572,6 +1563,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, default: break; } + track->zb_dirty = true; break; case RADEON_RB3D_ZPASS_ADDR: r = r100_cs_packet_next_reloc(p, &reloc); @@ -1588,6 +1580,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, uint32_t temp = idx_value >> 4; for (i = 0; i < track->num_texture; i++) track->textures[i].enabled = !!(temp & (1 << i)); + track->tex_dirty = true; } break; case RADEON_SE_VF_CNTL: @@ -1602,12 +1595,14 @@ static int r100_packet0_check(struct radeon_cs_parser *p, i = (reg - RADEON_PP_TEX_SIZE_0) / 8; track->textures[i].width = (idx_value & RADEON_TEX_USIZE_MASK) + 1; track->textures[i].height = ((idx_value & RADEON_TEX_VSIZE_MASK) >> RADEON_TEX_VSIZE_SHIFT) + 1; + track->tex_dirty = true; break; case RADEON_PP_TEX_PITCH_0: case RADEON_PP_TEX_PITCH_1: case RADEON_PP_TEX_PITCH_2: i = (reg - RADEON_PP_TEX_PITCH_0) / 8; track->textures[i].pitch = idx_value + 32; + track->tex_dirty = true; break; case RADEON_PP_TXFILTER_0: case RADEON_PP_TXFILTER_1: @@ -1621,6 +1616,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, tmp = (idx_value >> 27) & 0x7; if (tmp == 2 || tmp == 6) track->textures[i].roundup_h = false; + track->tex_dirty = true; break; case RADEON_PP_TXFORMAT_0: case RADEON_PP_TXFORMAT_1: @@ -1673,6 +1669,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, } track->textures[i].cube_info[4].width = 1 << ((idx_value >> 16) & 0xf); track->textures[i].cube_info[4].height = 1 << ((idx_value >> 20) & 0xf); + track->tex_dirty = true; break; case RADEON_PP_CUBIC_FACES_0: case RADEON_PP_CUBIC_FACES_1: @@ -1683,6 +1680,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p, track->textures[i].cube_info[face].width = 1 << ((tmp >> (face * 8)) & 0xf); track->textures[i].cube_info[face].height = 1 << ((tmp >> ((face * 8) + 4)) & 0xf); } + track->tex_dirty = true; break; default: printk(KERN_ERR "Forbidden register 0x%04X in cs at %d\n", @@ -2310,7 +2308,6 @@ void r100_vram_init_sizes(struct radeon_device *rdev) /* FIXME we don't use the second aperture yet when we could use it */ if (rdev->mc.visible_vram_size > rdev->mc.aper_size) rdev->mc.visible_vram_size = rdev->mc.aper_size; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; config_aper_size = RREG32(RADEON_CONFIG_APER_SIZE); if (rdev->flags & RADEON_IS_IGP) { uint32_t tom; @@ -3318,9 +3315,9 @@ int r100_cs_track_check(struct radeon_device *rdev, struct r100_cs_track *track) unsigned long size; unsigned prim_walk; unsigned nverts; - unsigned num_cb = track->num_cb; + unsigned num_cb = track->cb_dirty ? track->num_cb : 0; - if (!track->zb_cb_clear && !track->color_channel_mask && + if (num_cb && !track->zb_cb_clear && !track->color_channel_mask && !track->blend_read_enable) num_cb = 0; @@ -3341,7 +3338,9 @@ int r100_cs_track_check(struct radeon_device *rdev, struct r100_cs_track *track) return -EINVAL; } } - if (track->z_enabled) { + track->cb_dirty = false; + + if (track->zb_dirty && track->z_enabled) { if (track->zb.robj == NULL) { DRM_ERROR("[drm] No buffer for z buffer !\n"); return -EINVAL; @@ -3358,6 +3357,28 @@ int r100_cs_track_check(struct radeon_device *rdev, struct r100_cs_track *track) return -EINVAL; } } + track->zb_dirty = false; + + if (track->aa_dirty && track->aaresolve) { + if (track->aa.robj == NULL) { + DRM_ERROR("[drm] No buffer for AA resolve buffer %d !\n", i); + return -EINVAL; + } + /* I believe the format comes from colorbuffer0. */ + size = track->aa.pitch * track->cb[0].cpp * track->maxy; + size += track->aa.offset; + if (size > radeon_bo_size(track->aa.robj)) { + DRM_ERROR("[drm] Buffer too small for AA resolve buffer %d " + "(need %lu have %lu) !\n", i, size, + radeon_bo_size(track->aa.robj)); + DRM_ERROR("[drm] AA resolve buffer %d (%u %u %u %u)\n", + i, track->aa.pitch, track->cb[0].cpp, + track->aa.offset, track->maxy); + return -EINVAL; + } + } + track->aa_dirty = false; + prim_walk = (track->vap_vf_cntl >> 4) & 0x3; if (track->vap_vf_cntl & (1 << 14)) { nverts = track->vap_alt_nverts; @@ -3417,13 +3438,23 @@ int r100_cs_track_check(struct radeon_device *rdev, struct r100_cs_track *track) prim_walk); return -EINVAL; } - return r100_cs_track_texture_check(rdev, track); + + if (track->tex_dirty) { + track->tex_dirty = false; + return r100_cs_track_texture_check(rdev, track); + } + return 0; } void r100_cs_track_clear(struct radeon_device *rdev, struct r100_cs_track *track) { unsigned i, face; + track->cb_dirty = true; + track->zb_dirty = true; + track->tex_dirty = true; + track->aa_dirty = true; + if (rdev->family < CHIP_R300) { track->num_cb = 1; if (rdev->family <= CHIP_RS200) @@ -3437,6 +3468,8 @@ void r100_cs_track_clear(struct radeon_device *rdev, struct r100_cs_track *track track->num_texture = 16; track->maxy = 4096; track->separate_cube = 0; + track->aaresolve = false; + track->aa.robj = NULL; } for (i = 0; i < track->num_cb; i++) { @@ -3746,8 +3779,6 @@ static int r100_startup(struct radeon_device *rdev) r100_mc_program(rdev); /* Resume clock */ r100_clock_startup(rdev); - /* Initialize GPU configuration (# pipes, ...) */ -// r100_gpu_init(rdev); /* Initialize GART (initialize after TTM so we can allocate * memory through TTM but finalize after TTM) */ r100_enable_bm(rdev); diff --git a/drivers/gpu/drm/radeon/r100_track.h b/drivers/gpu/drm/radeon/r100_track.h index af65600..2fef9de 100644 --- a/drivers/gpu/drm/radeon/r100_track.h +++ b/drivers/gpu/drm/radeon/r100_track.h @@ -52,14 +52,7 @@ struct r100_cs_track_texture { unsigned compress_format; }; -struct r100_cs_track_limits { - unsigned num_cb; - unsigned num_texture; - unsigned max_levels; -}; - struct r100_cs_track { - struct radeon_device *rdev; unsigned num_cb; unsigned num_texture; unsigned maxy; @@ -73,11 +66,17 @@ struct r100_cs_track { struct r100_cs_track_array arrays[11]; struct r100_cs_track_cb cb[R300_MAX_CB]; struct r100_cs_track_cb zb; + struct r100_cs_track_cb aa; struct r100_cs_track_texture textures[R300_TRACK_MAX_TEXTURE]; bool z_enabled; bool separate_cube; bool zb_cb_clear; bool blend_read_enable; + bool cb_dirty; + bool zb_dirty; + bool tex_dirty; + bool aa_dirty; + bool aaresolve; }; int r100_cs_track_check(struct radeon_device *rdev, struct r100_cs_track *track); diff --git a/drivers/gpu/drm/radeon/r200.c b/drivers/gpu/drm/radeon/r200.c index d2408c3..f240583 100644 --- a/drivers/gpu/drm/radeon/r200.c +++ b/drivers/gpu/drm/radeon/r200.c @@ -184,6 +184,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, } track->zb.robj = reloc->robj; track->zb.offset = idx_value; + track->zb_dirty = true; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); break; case RADEON_RB3D_COLOROFFSET: @@ -196,6 +197,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, } track->cb[0].robj = reloc->robj; track->cb[0].offset = idx_value; + track->cb_dirty = true; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); break; case R200_PP_TXOFFSET_0: @@ -214,6 +216,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, } ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); track->textures[i].robj = reloc->robj; + track->tex_dirty = true; break; case R200_PP_CUBIC_OFFSET_F1_0: case R200_PP_CUBIC_OFFSET_F2_0: @@ -257,9 +260,12 @@ int r200_packet0_check(struct radeon_cs_parser *p, track->textures[i].cube_info[face - 1].offset = idx_value; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); track->textures[i].cube_info[face - 1].robj = reloc->robj; + track->tex_dirty = true; break; case RADEON_RE_WIDTH_HEIGHT: track->maxy = ((idx_value >> 16) & 0x7FF); + track->cb_dirty = true; + track->zb_dirty = true; break; case RADEON_RB3D_COLORPITCH: r = r100_cs_packet_next_reloc(p, &reloc); @@ -280,9 +286,11 @@ int r200_packet0_check(struct radeon_cs_parser *p, ib[idx] = tmp; track->cb[0].pitch = idx_value & RADEON_COLORPITCH_MASK; + track->cb_dirty = true; break; case RADEON_RB3D_DEPTHPITCH: track->zb.pitch = idx_value & RADEON_DEPTHPITCH_MASK; + track->zb_dirty = true; break; case RADEON_RB3D_CNTL: switch ((idx_value >> RADEON_RB3D_COLOR_FORMAT_SHIFT) & 0x1f) { @@ -312,6 +320,8 @@ int r200_packet0_check(struct radeon_cs_parser *p, } track->z_enabled = !!(idx_value & RADEON_Z_ENABLE); + track->cb_dirty = true; + track->zb_dirty = true; break; case RADEON_RB3D_ZSTENCILCNTL: switch (idx_value & 0xf) { @@ -329,6 +339,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, default: break; } + track->zb_dirty = true; break; case RADEON_RB3D_ZPASS_ADDR: r = r100_cs_packet_next_reloc(p, &reloc); @@ -345,6 +356,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, uint32_t temp = idx_value >> 4; for (i = 0; i < track->num_texture; i++) track->textures[i].enabled = !!(temp & (1 << i)); + track->tex_dirty = true; } break; case RADEON_SE_VF_CNTL: @@ -369,6 +381,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, i = (reg - R200_PP_TXSIZE_0) / 32; track->textures[i].width = (idx_value & RADEON_TEX_USIZE_MASK) + 1; track->textures[i].height = ((idx_value & RADEON_TEX_VSIZE_MASK) >> RADEON_TEX_VSIZE_SHIFT) + 1; + track->tex_dirty = true; break; case R200_PP_TXPITCH_0: case R200_PP_TXPITCH_1: @@ -378,6 +391,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, case R200_PP_TXPITCH_5: i = (reg - R200_PP_TXPITCH_0) / 32; track->textures[i].pitch = idx_value + 32; + track->tex_dirty = true; break; case R200_PP_TXFILTER_0: case R200_PP_TXFILTER_1: @@ -394,6 +408,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, tmp = (idx_value >> 27) & 0x7; if (tmp == 2 || tmp == 6) track->textures[i].roundup_h = false; + track->tex_dirty = true; break; case R200_PP_TXMULTI_CTL_0: case R200_PP_TXMULTI_CTL_1: @@ -432,6 +447,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, track->textures[i].tex_coord_type = 1; break; } + track->tex_dirty = true; break; case R200_PP_TXFORMAT_0: case R200_PP_TXFORMAT_1: @@ -488,6 +504,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, } track->textures[i].cube_info[4].width = 1 << ((idx_value >> 16) & 0xf); track->textures[i].cube_info[4].height = 1 << ((idx_value >> 20) & 0xf); + track->tex_dirty = true; break; case R200_PP_CUBIC_FACES_0: case R200_PP_CUBIC_FACES_1: @@ -501,6 +518,7 @@ int r200_packet0_check(struct radeon_cs_parser *p, track->textures[i].cube_info[face].width = 1 << ((tmp >> (face * 8)) & 0xf); track->textures[i].cube_info[face].height = 1 << ((tmp >> ((face * 8) + 4)) & 0xf); } + track->tex_dirty = true; break; default: printk(KERN_ERR "Forbidden register 0x%04X in cs at %d\n", diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c index 07dfb54..8713731 100644 --- a/drivers/gpu/drm/radeon/r300.c +++ b/drivers/gpu/drm/radeon/r300.c @@ -667,6 +667,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, } track->cb[i].robj = reloc->robj; track->cb[i].offset = idx_value; + track->cb_dirty = true; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); break; case R300_ZB_DEPTHOFFSET: @@ -679,6 +680,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, } track->zb.robj = reloc->robj; track->zb.offset = idx_value; + track->zb_dirty = true; ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); break; case R300_TX_OFFSET_0: @@ -717,6 +719,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, tmp |= tile_flags; ib[idx] = tmp; track->textures[i].robj = reloc->robj; + track->tex_dirty = true; break; /* Tracked registers */ case 0x2084: @@ -743,6 +746,8 @@ static int r300_packet0_check(struct radeon_cs_parser *p, if (p->rdev->family < CHIP_RV515) { track->maxy -= 1440; } + track->cb_dirty = true; + track->zb_dirty = true; break; case 0x4E00: /* RB3D_CCTL */ @@ -752,6 +757,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, return -EINVAL; } track->num_cb = ((idx_value >> 5) & 0x3) + 1; + track->cb_dirty = true; break; case 0x4E38: case 0x4E3C: @@ -814,6 +820,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, ((idx_value >> 21) & 0xF)); return -EINVAL; } + track->cb_dirty = true; break; case 0x4F00: /* ZB_CNTL */ @@ -822,6 +829,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, } else { track->z_enabled = false; } + track->zb_dirty = true; break; case 0x4F10: /* ZB_FORMAT */ @@ -838,6 +846,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, (idx_value & 0xF)); return -EINVAL; } + track->zb_dirty = true; break; case 0x4F24: /* ZB_DEPTHPITCH */ @@ -861,14 +870,17 @@ static int r300_packet0_check(struct radeon_cs_parser *p, ib[idx] = tmp; track->zb.pitch = idx_value & 0x3FFC; + track->zb_dirty = true; break; case 0x4104: + /* TX_ENABLE */ for (i = 0; i < 16; i++) { bool enabled; enabled = !!(idx_value & (1 << i)); track->textures[i].enabled = enabled; } + track->tex_dirty = true; break; case 0x44C0: case 0x44C4: @@ -898,6 +910,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R300_TX_FORMAT_X16: + case R300_TX_FORMAT_FL_I16: case R300_TX_FORMAT_Y8X8: case R300_TX_FORMAT_Z5Y6X5: case R300_TX_FORMAT_Z6Y5X5: @@ -910,6 +923,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, track->textures[i].compress_format = R100_TRACK_COMP_NONE; break; case R300_TX_FORMAT_Y16X16: + case R300_TX_FORMAT_FL_I16A16: case R300_TX_FORMAT_Z11Y11X10: case R300_TX_FORMAT_Z10Y11X11: case R300_TX_FORMAT_W8Z8Y8X8: @@ -951,8 +965,8 @@ static int r300_packet0_check(struct radeon_cs_parser *p, DRM_ERROR("Invalid texture format %u\n", (idx_value & 0x1F)); return -EINVAL; - break; } + track->tex_dirty = true; break; case 0x4400: case 0x4404: @@ -980,6 +994,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, if (tmp == 2 || tmp == 4 || tmp == 6) { track->textures[i].roundup_h = false; } + track->tex_dirty = true; break; case 0x4500: case 0x4504: @@ -1017,6 +1032,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, DRM_ERROR("Forbidden bit TXFORMAT_MSB\n"); return -EINVAL; } + track->tex_dirty = true; break; case 0x4480: case 0x4484: @@ -1046,6 +1062,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, track->textures[i].use_pitch = !!tmp; tmp = (idx_value >> 22) & 0xF; track->textures[i].txdepth = tmp; + track->tex_dirty = true; break; case R300_ZB_ZPASS_ADDR: r = r100_cs_packet_next_reloc(p, &reloc); @@ -1060,6 +1077,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p, case 0x4e0c: /* RB3D_COLOR_CHANNEL_MASK */ track->color_channel_mask = idx_value; + track->cb_dirty = true; break; case 0x43a4: /* SC_HYPERZ_EN */ @@ -1073,6 +1091,8 @@ static int r300_packet0_check(struct radeon_cs_parser *p, case 0x4f1c: /* ZB_BW_CNTL */ track->zb_cb_clear = !!(idx_value & (1 << 5)); + track->cb_dirty = true; + track->zb_dirty = true; if (p->rdev->hyperz_filp != p->filp) { if (idx_value & (R300_HIZ_ENABLE | R300_RD_COMP_ENABLE | @@ -1084,8 +1104,28 @@ static int r300_packet0_check(struct radeon_cs_parser *p, case 0x4e04: /* RB3D_BLENDCNTL */ track->blend_read_enable = !!(idx_value & (1 << 2)); + track->cb_dirty = true; + break; + case R300_RB3D_AARESOLVE_OFFSET: + r = r100_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("No reloc for ib[%d]=0x%04X\n", + idx, reg); + r100_cs_dump_packet(p, pkt); + return r; + } + track->aa.robj = reloc->robj; + track->aa.offset = idx_value; + track->aa_dirty = true; + ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset); + break; + case R300_RB3D_AARESOLVE_PITCH: + track->aa.pitch = idx_value & 0x3FFE; + track->aa_dirty = true; break; - case 0x4f28: /* ZB_DEPTHCLEARVALUE */ + case R300_RB3D_AARESOLVE_CTL: + track->aaresolve = idx_value & 0x1; + track->aa_dirty = true; break; case 0x4f30: /* ZB_MASK_OFFSET */ case 0x4f34: /* ZB_ZMASK_PITCH */ diff --git a/drivers/gpu/drm/radeon/r300_reg.h b/drivers/gpu/drm/radeon/r300_reg.h index 1a0d536..f0bce39 100644 --- a/drivers/gpu/drm/radeon/r300_reg.h +++ b/drivers/gpu/drm/radeon/r300_reg.h @@ -1371,6 +1371,8 @@ #define R300_RB3D_COLORPITCH2 0x4E40 /* GUESS */ #define R300_RB3D_COLORPITCH3 0x4E44 /* GUESS */ +#define R300_RB3D_AARESOLVE_OFFSET 0x4E80 +#define R300_RB3D_AARESOLVE_PITCH 0x4E84 #define R300_RB3D_AARESOLVE_CTL 0x4E88 /* gap */ diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index b8f15bf..be271c4 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -47,6 +47,7 @@ #define EVERGREEN_PFP_UCODE_SIZE 1120 #define EVERGREEN_PM4_UCODE_SIZE 1376 #define EVERGREEN_RLC_UCODE_SIZE 768 +#define CAYMAN_RLC_UCODE_SIZE 1024 /* Firmware Names */ MODULE_FIRMWARE("radeon/R600_pfp.bin"); @@ -1255,7 +1256,6 @@ int r600_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r600_vram_gtt_location(rdev, &rdev->mc); if (rdev->flags & RADEON_IS_IGP) { @@ -1937,7 +1937,7 @@ void r600_pciep_wreg(struct radeon_device *rdev, u32 reg, u32 v) */ void r600_cp_stop(struct radeon_device *rdev) { - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1)); WREG32(SCRATCH_UMSK, 0); } @@ -2105,7 +2105,11 @@ static int r600_cp_load_microcode(struct radeon_device *rdev) r600_cp_stop(rdev); - WREG32(CP_RB_CNTL, RB_NO_UPDATE | RB_BLKSZ(15) | RB_BUFSZ(3)); + WREG32(CP_RB_CNTL, +#ifdef __BIG_ENDIAN + BUF_SWAP_32BIT | +#endif + RB_NO_UPDATE | RB_BLKSZ(15) | RB_BUFSZ(3)); /* Reset cp */ WREG32(GRBM_SOFT_RESET, SOFT_RESET_CP); @@ -2192,7 +2196,11 @@ int r600_cp_resume(struct radeon_device *rdev) WREG32(CP_RB_WPTR, 0); /* set the wb address whether it's enabled or not */ - WREG32(CP_RB_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFFFFFFFC); + WREG32(CP_RB_RPTR_ADDR, +#ifdef __BIG_ENDIAN + RB_RPTR_SWAP(2) | +#endif + ((rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFFFFFFFC)); WREG32(CP_RB_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFF); WREG32(SCRATCH_ADDR, ((rdev->wb.gpu_addr + RADEON_WB_SCRATCH_OFFSET) >> 8) & 0xFFFFFFFF); @@ -2628,7 +2636,11 @@ void r600_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib) { /* FIXME: implement */ radeon_ring_write(rdev, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); - radeon_ring_write(rdev, ib->gpu_addr & 0xFFFFFFFC); + radeon_ring_write(rdev, +#ifdef __BIG_ENDIAN + (2 << 0) | +#endif + (ib->gpu_addr & 0xFFFFFFFC)); radeon_ring_write(rdev, upper_32_bits(ib->gpu_addr) & 0xFF); radeon_ring_write(rdev, ib->length_dw); } @@ -2728,7 +2740,7 @@ static int r600_ih_ring_alloc(struct radeon_device *rdev) /* Allocate ring buffer */ if (rdev->ih.ring_obj == NULL) { - r = radeon_bo_create(rdev, NULL, rdev->ih.ring_size, + r = radeon_bo_create(rdev, rdev->ih.ring_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT, &rdev->ih.ring_obj); @@ -2809,13 +2821,20 @@ static int r600_rlc_init(struct radeon_device *rdev) WREG32(RLC_HB_CNTL, 0); WREG32(RLC_HB_RPTR, 0); WREG32(RLC_HB_WPTR, 0); - WREG32(RLC_HB_WPTR_LSB_ADDR, 0); - WREG32(RLC_HB_WPTR_MSB_ADDR, 0); + if (rdev->family <= CHIP_CAICOS) { + WREG32(RLC_HB_WPTR_LSB_ADDR, 0); + WREG32(RLC_HB_WPTR_MSB_ADDR, 0); + } WREG32(RLC_MC_CNTL, 0); WREG32(RLC_UCODE_CNTL, 0); fw_data = (const __be32 *)rdev->rlc_fw->data; - if (rdev->family >= CHIP_CEDAR) { + if (rdev->family >= CHIP_CAYMAN) { + for (i = 0; i < CAYMAN_RLC_UCODE_SIZE; i++) { + WREG32(RLC_UCODE_ADDR, i); + WREG32(RLC_UCODE_DATA, be32_to_cpup(fw_data++)); + } + } else if (rdev->family >= CHIP_CEDAR) { for (i = 0; i < EVERGREEN_RLC_UCODE_SIZE; i++) { WREG32(RLC_UCODE_ADDR, i); WREG32(RLC_UCODE_DATA, be32_to_cpup(fw_data++)); @@ -3297,8 +3316,8 @@ restart_ih: while (rptr != wptr) { /* wptr/rptr are in bytes! */ ring_index = rptr / 4; - src_id = rdev->ih.ring[ring_index] & 0xff; - src_data = rdev->ih.ring[ring_index + 1] & 0xfffffff; + src_id = le32_to_cpu(rdev->ih.ring[ring_index]) & 0xff; + src_data = le32_to_cpu(rdev->ih.ring[ring_index + 1]) & 0xfffffff; switch (src_id) { case 1: /* D1 vblank/vline */ diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c index b5443fe..846fae5 100644 --- a/drivers/gpu/drm/radeon/r600_audio.c +++ b/drivers/gpu/drm/radeon/r600_audio.c @@ -26,6 +26,7 @@ #include "drmP.h" #include "radeon.h" #include "radeon_reg.h" +#include "radeon_asic.h" #include "atom.h" #define AUDIO_TIMER_INTERVALL 100 /* 1/10 sekund should be enough */ diff --git a/drivers/gpu/drm/radeon/r600_blit.c b/drivers/gpu/drm/radeon/r600_blit.c index ca5c29f..7f10434 100644 --- a/drivers/gpu/drm/radeon/r600_blit.c +++ b/drivers/gpu/drm/radeon/r600_blit.c @@ -137,9 +137,9 @@ set_shaders(struct drm_device *dev) ps = (u32 *) ((char *)dev->agp_buffer_map->handle + dev_priv->blit_vb->offset + 256); for (i = 0; i < r6xx_vs_size; i++) - vs[i] = r6xx_vs[i]; + vs[i] = cpu_to_le32(r6xx_vs[i]); for (i = 0; i < r6xx_ps_size; i++) - ps[i] = r6xx_ps[i]; + ps[i] = cpu_to_le32(r6xx_ps[i]); dev_priv->blit_vb->used = 512; @@ -192,6 +192,9 @@ set_vtx_resource(drm_radeon_private_t *dev_priv, u64 gpu_addr) DRM_DEBUG("\n"); sq_vtx_constant_word2 = (((gpu_addr >> 32) & 0xff) | (16 << 8)); +#ifdef __BIG_ENDIAN + sq_vtx_constant_word2 |= (2 << 30); +#endif BEGIN_RING(9); OUT_RING(CP_PACKET3(R600_IT_SET_RESOURCE, 7)); @@ -291,7 +294,11 @@ draw_auto(drm_radeon_private_t *dev_priv) OUT_RING(DI_PT_RECTLIST); OUT_RING(CP_PACKET3(R600_IT_INDEX_TYPE, 0)); +#ifdef __BIG_ENDIAN + OUT_RING((2 << 2) | DI_INDEX_SIZE_16_BIT); +#else OUT_RING(DI_INDEX_SIZE_16_BIT); +#endif OUT_RING(CP_PACKET3(R600_IT_NUM_INSTANCES, 0)); OUT_RING(1); diff --git a/drivers/gpu/drm/radeon/r600_blit_kms.c b/drivers/gpu/drm/radeon/r600_blit_kms.c index 86e5aa0..9aa74c3f 100644 --- a/drivers/gpu/drm/radeon/r600_blit_kms.c +++ b/drivers/gpu/drm/radeon/r600_blit_kms.c @@ -54,7 +54,7 @@ set_render_target(struct radeon_device *rdev, int format, if (h < 8) h = 8; - cb_color_info = ((format << 2) | (1 << 27)); + cb_color_info = ((format << 2) | (1 << 27) | (1 << 8)); pitch = (w / 8) - 1; slice = ((w * h) / 64) - 1; @@ -165,6 +165,9 @@ set_vtx_resource(struct radeon_device *rdev, u64 gpu_addr) u32 sq_vtx_constant_word2; sq_vtx_constant_word2 = ((upper_32_bits(gpu_addr) & 0xff) | (16 << 8)); +#ifdef __BIG_ENDIAN + sq_vtx_constant_word2 |= (2 << 30); +#endif radeon_ring_write(rdev, PACKET3(PACKET3_SET_RESOURCE, 7)); radeon_ring_write(rdev, 0x460); @@ -199,7 +202,7 @@ set_tex_resource(struct radeon_device *rdev, if (h < 1) h = 1; - sq_tex_resource_word0 = (1 << 0); + sq_tex_resource_word0 = (1 << 0) | (1 << 3); sq_tex_resource_word0 |= ((((pitch >> 3) - 1) << 8) | ((w - 1) << 19)); @@ -253,7 +256,11 @@ draw_auto(struct radeon_device *rdev) radeon_ring_write(rdev, DI_PT_RECTLIST); radeon_ring_write(rdev, PACKET3(PACKET3_INDEX_TYPE, 0)); - radeon_ring_write(rdev, DI_INDEX_SIZE_16_BIT); + radeon_ring_write(rdev, +#ifdef __BIG_ENDIAN + (2 << 2) | +#endif + DI_INDEX_SIZE_16_BIT); radeon_ring_write(rdev, PACKET3(PACKET3_NUM_INSTANCES, 0)); radeon_ring_write(rdev, 1); @@ -424,7 +431,11 @@ set_default_state(struct radeon_device *rdev) dwords = ALIGN(rdev->r600_blit.state_len, 0x10); gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.state_offset; radeon_ring_write(rdev, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); - radeon_ring_write(rdev, gpu_addr & 0xFFFFFFFC); + radeon_ring_write(rdev, +#ifdef __BIG_ENDIAN + (2 << 0) | +#endif + (gpu_addr & 0xFFFFFFFC)); radeon_ring_write(rdev, upper_32_bits(gpu_addr) & 0xFF); radeon_ring_write(rdev, dwords); @@ -467,7 +478,7 @@ static inline uint32_t i2f(uint32_t input) int r600_blit_init(struct radeon_device *rdev) { u32 obj_size; - int r, dwords; + int i, r, dwords; void *ptr; u32 packet2s[16]; int num_packet2s = 0; @@ -486,7 +497,7 @@ int r600_blit_init(struct radeon_device *rdev) dwords = rdev->r600_blit.state_len; while (dwords & 0xf) { - packet2s[num_packet2s++] = PACKET2(0); + packet2s[num_packet2s++] = cpu_to_le32(PACKET2(0)); dwords++; } @@ -501,7 +512,7 @@ int r600_blit_init(struct radeon_device *rdev) obj_size += r6xx_ps_size * 4; obj_size = ALIGN(obj_size, 256); - r = radeon_bo_create(rdev, NULL, obj_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, + r = radeon_bo_create(rdev, obj_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, &rdev->r600_blit.shader_obj); if (r) { DRM_ERROR("r600 failed to allocate shader\n"); @@ -529,8 +540,10 @@ int r600_blit_init(struct radeon_device *rdev) if (num_packet2s) memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4), packet2s, num_packet2s * 4); - memcpy(ptr + rdev->r600_blit.vs_offset, r6xx_vs, r6xx_vs_size * 4); - memcpy(ptr + rdev->r600_blit.ps_offset, r6xx_ps, r6xx_ps_size * 4); + for (i = 0; i < r6xx_vs_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.vs_offset + i * 4) = cpu_to_le32(r6xx_vs[i]); + for (i = 0; i < r6xx_ps_size; i++) + *(u32 *)((unsigned long)ptr + rdev->r600_blit.ps_offset + i * 4) = cpu_to_le32(r6xx_ps[i]); radeon_bo_kunmap(rdev->r600_blit.shader_obj); radeon_bo_unreserve(rdev->r600_blit.shader_obj); @@ -545,7 +558,7 @@ done: dev_err(rdev->dev, "(%d) pin blit object failed\n", r); return r; } - rdev->mc.active_vram_size = rdev->mc.real_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size); return 0; } @@ -553,7 +566,7 @@ void r600_blit_fini(struct radeon_device *rdev) { int r; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); if (rdev->r600_blit.shader_obj == NULL) return; /* If we can't reserve the bo, unref should be enough to destroy diff --git a/drivers/gpu/drm/radeon/r600_blit_shaders.c b/drivers/gpu/drm/radeon/r600_blit_shaders.c index e8151c1..2d1f6c5 100644 --- a/drivers/gpu/drm/radeon/r600_blit_shaders.c +++ b/drivers/gpu/drm/radeon/r600_blit_shaders.c @@ -684,7 +684,11 @@ const u32 r6xx_vs[] = 0x00000000, 0x3c000000, 0x68cd1000, +#ifdef __BIG_ENDIAN + 0x000a0000, +#else 0x00080000, +#endif 0x00000000, }; diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c index 4f4cd8b..c3ab959 100644 --- a/drivers/gpu/drm/radeon/r600_cp.c +++ b/drivers/gpu/drm/radeon/r600_cp.c @@ -396,6 +396,9 @@ static void r600_cp_load_microcode(drm_radeon_private_t *dev_priv) r600_do_cp_stop(dev_priv); RADEON_WRITE(R600_CP_RB_CNTL, +#ifdef __BIG_ENDIAN + R600_BUF_SWAP_32BIT | +#endif R600_RB_NO_UPDATE | R600_RB_BLKSZ(15) | R600_RB_BUFSZ(3)); @@ -486,9 +489,12 @@ static void r700_cp_load_microcode(drm_radeon_private_t *dev_priv) r600_do_cp_stop(dev_priv); RADEON_WRITE(R600_CP_RB_CNTL, +#ifdef __BIG_ENDIAN + R600_BUF_SWAP_32BIT | +#endif R600_RB_NO_UPDATE | - (15 << 8) | - (3 << 0)); + R600_RB_BLKSZ(15) | + R600_RB_BUFSZ(3)); RADEON_WRITE(R600_GRBM_SOFT_RESET, R600_SOFT_RESET_CP); RADEON_READ(R600_GRBM_SOFT_RESET); @@ -550,8 +556,12 @@ static void r600_test_writeback(drm_radeon_private_t *dev_priv) if (!dev_priv->writeback_works) { /* Disable writeback to avoid unnecessary bus master transfer */ - RADEON_WRITE(R600_CP_RB_CNTL, RADEON_READ(R600_CP_RB_CNTL) | - RADEON_RB_NO_UPDATE); + RADEON_WRITE(R600_CP_RB_CNTL, +#ifdef __BIG_ENDIAN + R600_BUF_SWAP_32BIT | +#endif + RADEON_READ(R600_CP_RB_CNTL) | + R600_RB_NO_UPDATE); RADEON_WRITE(R600_SCRATCH_UMSK, 0); } } @@ -575,7 +585,11 @@ int r600_do_engine_reset(struct drm_device *dev) RADEON_WRITE(R600_CP_RB_WPTR_DELAY, 0); cp_rb_cntl = RADEON_READ(R600_CP_RB_CNTL); - RADEON_WRITE(R600_CP_RB_CNTL, R600_RB_RPTR_WR_ENA); + RADEON_WRITE(R600_CP_RB_CNTL, +#ifdef __BIG_ENDIAN + R600_BUF_SWAP_32BIT | +#endif + R600_RB_RPTR_WR_ENA); RADEON_WRITE(R600_CP_RB_RPTR_WR, cp_ptr); RADEON_WRITE(R600_CP_RB_WPTR, cp_ptr); @@ -1838,7 +1852,10 @@ static void r600_cp_init_ring_buffer(struct drm_device *dev, + dev_priv->gart_vm_start; } RADEON_WRITE(R600_CP_RB_RPTR_ADDR, - rptr_addr & 0xffffffff); +#ifdef __BIG_ENDIAN + (2 << 0) | +#endif + (rptr_addr & 0xfffffffc)); RADEON_WRITE(R600_CP_RB_RPTR_ADDR_HI, upper_32_bits(rptr_addr)); @@ -1889,7 +1906,7 @@ static void r600_cp_init_ring_buffer(struct drm_device *dev, { u64 scratch_addr; - scratch_addr = RADEON_READ(R600_CP_RB_RPTR_ADDR); + scratch_addr = RADEON_READ(R600_CP_RB_RPTR_ADDR) & 0xFFFFFFFC; scratch_addr |= ((u64)RADEON_READ(R600_CP_RB_RPTR_ADDR_HI)) << 32; scratch_addr += R600_SCRATCH_REG_OFFSET; scratch_addr >>= 8; diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c index 7831e08..3324620 100644 --- a/drivers/gpu/drm/radeon/r600_cs.c +++ b/drivers/gpu/drm/radeon/r600_cs.c @@ -71,75 +71,167 @@ struct r600_cs_track { u64 db_bo_mc; }; +#define FMT_8_BIT(fmt, vc) [fmt] = { 1, 1, 1, vc } +#define FMT_16_BIT(fmt, vc) [fmt] = { 1, 1, 2, vc } +#define FMT_24_BIT(fmt) [fmt] = { 1, 1, 3, 0 } +#define FMT_32_BIT(fmt, vc) [fmt] = { 1, 1, 4, vc } +#define FMT_48_BIT(fmt) [fmt] = { 1, 1, 6, 0 } +#define FMT_64_BIT(fmt, vc) [fmt] = { 1, 1, 8, vc } +#define FMT_96_BIT(fmt) [fmt] = { 1, 1, 12, 0 } +#define FMT_128_BIT(fmt, vc) [fmt] = { 1, 1, 16, vc } + +struct gpu_formats { + unsigned blockwidth; + unsigned blockheight; + unsigned blocksize; + unsigned valid_color; +}; + +static const struct gpu_formats color_formats_table[] = { + /* 8 bit */ + FMT_8_BIT(V_038004_COLOR_8, 1), + FMT_8_BIT(V_038004_COLOR_4_4, 1), + FMT_8_BIT(V_038004_COLOR_3_3_2, 1), + FMT_8_BIT(V_038004_FMT_1, 0), + + /* 16-bit */ + FMT_16_BIT(V_038004_COLOR_16, 1), + FMT_16_BIT(V_038004_COLOR_16_FLOAT, 1), + FMT_16_BIT(V_038004_COLOR_8_8, 1), + FMT_16_BIT(V_038004_COLOR_5_6_5, 1), + FMT_16_BIT(V_038004_COLOR_6_5_5, 1), + FMT_16_BIT(V_038004_COLOR_1_5_5_5, 1), + FMT_16_BIT(V_038004_COLOR_4_4_4_4, 1), + FMT_16_BIT(V_038004_COLOR_5_5_5_1, 1), + + /* 24-bit */ + FMT_24_BIT(V_038004_FMT_8_8_8), + + /* 32-bit */ + FMT_32_BIT(V_038004_COLOR_32, 1), + FMT_32_BIT(V_038004_COLOR_32_FLOAT, 1), + FMT_32_BIT(V_038004_COLOR_16_16, 1), + FMT_32_BIT(V_038004_COLOR_16_16_FLOAT, 1), + FMT_32_BIT(V_038004_COLOR_8_24, 1), + FMT_32_BIT(V_038004_COLOR_8_24_FLOAT, 1), + FMT_32_BIT(V_038004_COLOR_24_8, 1), + FMT_32_BIT(V_038004_COLOR_24_8_FLOAT, 1), + FMT_32_BIT(V_038004_COLOR_10_11_11, 1), + FMT_32_BIT(V_038004_COLOR_10_11_11_FLOAT, 1), + FMT_32_BIT(V_038004_COLOR_11_11_10, 1), + FMT_32_BIT(V_038004_COLOR_11_11_10_FLOAT, 1), + FMT_32_BIT(V_038004_COLOR_2_10_10_10, 1), + FMT_32_BIT(V_038004_COLOR_8_8_8_8, 1), + FMT_32_BIT(V_038004_COLOR_10_10_10_2, 1), + FMT_32_BIT(V_038004_FMT_5_9_9_9_SHAREDEXP, 0), + FMT_32_BIT(V_038004_FMT_32_AS_8, 0), + FMT_32_BIT(V_038004_FMT_32_AS_8_8, 0), + + /* 48-bit */ + FMT_48_BIT(V_038004_FMT_16_16_16), + FMT_48_BIT(V_038004_FMT_16_16_16_FLOAT), + + /* 64-bit */ + FMT_64_BIT(V_038004_COLOR_X24_8_32_FLOAT, 1), + FMT_64_BIT(V_038004_COLOR_32_32, 1), + FMT_64_BIT(V_038004_COLOR_32_32_FLOAT, 1), + FMT_64_BIT(V_038004_COLOR_16_16_16_16, 1), + FMT_64_BIT(V_038004_COLOR_16_16_16_16_FLOAT, 1), + + FMT_96_BIT(V_038004_FMT_32_32_32), + FMT_96_BIT(V_038004_FMT_32_32_32_FLOAT), + + /* 128-bit */ + FMT_128_BIT(V_038004_COLOR_32_32_32_32, 1), + FMT_128_BIT(V_038004_COLOR_32_32_32_32_FLOAT, 1), + + [V_038004_FMT_GB_GR] = { 2, 1, 4, 0 }, + [V_038004_FMT_BG_RG] = { 2, 1, 4, 0 }, + + /* block compressed formats */ + [V_038004_FMT_BC1] = { 4, 4, 8, 0 }, + [V_038004_FMT_BC2] = { 4, 4, 16, 0 }, + [V_038004_FMT_BC3] = { 4, 4, 16, 0 }, + [V_038004_FMT_BC4] = { 4, 4, 8, 0 }, + [V_038004_FMT_BC5] = { 4, 4, 16, 0}, + +}; + +static inline bool fmt_is_valid_color(u32 format) +{ + if (format >= ARRAY_SIZE(color_formats_table)) + return false; + + if (color_formats_table[format].valid_color) + return true; + + return false; +} + +static inline bool fmt_is_valid_texture(u32 format) +{ + if (format >= ARRAY_SIZE(color_formats_table)) + return false; + + if (color_formats_table[format].blockwidth > 0) + return true; + + return false; +} + +static inline int fmt_get_blocksize(u32 format) +{ + if (format >= ARRAY_SIZE(color_formats_table)) + return 0; + + return color_formats_table[format].blocksize; +} + +static inline int fmt_get_nblocksx(u32 format, u32 w) +{ + unsigned bw; + + if (format >= ARRAY_SIZE(color_formats_table)) + return 0; + + bw = color_formats_table[format].blockwidth; + if (bw == 0) + return 0; + + return (w + bw - 1) / bw; +} + +static inline int fmt_get_nblocksy(u32 format, u32 h) +{ + unsigned bh; + + if (format >= ARRAY_SIZE(color_formats_table)) + return 0; + + bh = color_formats_table[format].blockheight; + if (bh == 0) + return 0; + + return (h + bh - 1) / bh; +} + static inline int r600_bpe_from_format(u32 *bpe, u32 format) { - switch (format) { - case V_038004_COLOR_8: - case V_038004_COLOR_4_4: - case V_038004_COLOR_3_3_2: - case V_038004_FMT_1: - *bpe = 1; - break; - case V_038004_COLOR_16: - case V_038004_COLOR_16_FLOAT: - case V_038004_COLOR_8_8: - case V_038004_COLOR_5_6_5: - case V_038004_COLOR_6_5_5: - case V_038004_COLOR_1_5_5_5: - case V_038004_COLOR_4_4_4_4: - case V_038004_COLOR_5_5_5_1: - *bpe = 2; - break; - case V_038004_FMT_8_8_8: - *bpe = 3; - break; - case V_038004_COLOR_32: - case V_038004_COLOR_32_FLOAT: - case V_038004_COLOR_16_16: - case V_038004_COLOR_16_16_FLOAT: - case V_038004_COLOR_8_24: - case V_038004_COLOR_8_24_FLOAT: - case V_038004_COLOR_24_8: - case V_038004_COLOR_24_8_FLOAT: - case V_038004_COLOR_10_11_11: - case V_038004_COLOR_10_11_11_FLOAT: - case V_038004_COLOR_11_11_10: - case V_038004_COLOR_11_11_10_FLOAT: - case V_038004_COLOR_2_10_10_10: - case V_038004_COLOR_8_8_8_8: - case V_038004_COLOR_10_10_10_2: - case V_038004_FMT_5_9_9_9_SHAREDEXP: - case V_038004_FMT_32_AS_8: - case V_038004_FMT_32_AS_8_8: - *bpe = 4; - break; - case V_038004_COLOR_X24_8_32_FLOAT: - case V_038004_COLOR_32_32: - case V_038004_COLOR_32_32_FLOAT: - case V_038004_COLOR_16_16_16_16: - case V_038004_COLOR_16_16_16_16_FLOAT: - *bpe = 8; - break; - case V_038004_FMT_16_16_16: - case V_038004_FMT_16_16_16_FLOAT: - *bpe = 6; - break; - case V_038004_FMT_32_32_32: - case V_038004_FMT_32_32_32_FLOAT: - *bpe = 12; - break; - case V_038004_COLOR_32_32_32_32: - case V_038004_COLOR_32_32_32_32_FLOAT: - *bpe = 16; - break; - case V_038004_FMT_GB_GR: - case V_038004_FMT_BG_RG: - case V_038004_COLOR_INVALID: - default: - *bpe = 16; - return -EINVAL; - } + unsigned res; + + if (format >= ARRAY_SIZE(color_formats_table)) + goto fail; + + res = color_formats_table[format].blocksize; + if (res == 0) + goto fail; + + *bpe = res; return 0; + +fail: + *bpe = 16; + return -EINVAL; } struct array_mode_checker { @@ -148,7 +240,7 @@ struct array_mode_checker { u32 nbanks; u32 npipes; u32 nsamples; - u32 bpe; + u32 blocksize; }; /* returns alignment in pixels for pitch/height/depth and bytes for base */ @@ -162,7 +254,7 @@ static inline int r600_get_array_mode_alignment(struct array_mode_checker *value u32 tile_height = 8; u32 macro_tile_width = values->nbanks; u32 macro_tile_height = values->npipes; - u32 tile_bytes = tile_width * tile_height * values->bpe * values->nsamples; + u32 tile_bytes = tile_width * tile_height * values->blocksize * values->nsamples; u32 macro_tile_bytes = macro_tile_width * macro_tile_height * tile_bytes; switch (values->array_mode) { @@ -174,7 +266,7 @@ static inline int r600_get_array_mode_alignment(struct array_mode_checker *value *base_align = 1; break; case ARRAY_LINEAR_ALIGNED: - *pitch_align = max((u32)64, (u32)(values->group_size / values->bpe)); + *pitch_align = max((u32)64, (u32)(values->group_size / values->blocksize)); *height_align = tile_height; *depth_align = 1; *base_align = values->group_size; @@ -182,7 +274,7 @@ static inline int r600_get_array_mode_alignment(struct array_mode_checker *value case ARRAY_1D_TILED_THIN1: *pitch_align = max((u32)tile_width, (u32)(values->group_size / - (tile_height * values->bpe * values->nsamples))); + (tile_height * values->blocksize * values->nsamples))); *height_align = tile_height; *depth_align = 1; *base_align = values->group_size; @@ -190,12 +282,12 @@ static inline int r600_get_array_mode_alignment(struct array_mode_checker *value case ARRAY_2D_TILED_THIN1: *pitch_align = max((u32)macro_tile_width, (u32)(((values->group_size / tile_height) / - (values->bpe * values->nsamples)) * + (values->blocksize * values->nsamples)) * values->nbanks)) * tile_width; *height_align = macro_tile_height * tile_height; *depth_align = 1; *base_align = max(macro_tile_bytes, - (*pitch_align) * values->bpe * (*height_align) * values->nsamples); + (*pitch_align) * values->blocksize * (*height_align) * values->nsamples); break; default: return -EINVAL; @@ -234,21 +326,22 @@ static void r600_cs_track_init(struct r600_cs_track *track) static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i) { struct r600_cs_track *track = p->track; - u32 bpe = 0, slice_tile_max, size, tmp; + u32 slice_tile_max, size, tmp; u32 height, height_align, pitch, pitch_align, depth_align; u64 base_offset, base_align; struct array_mode_checker array_check; volatile u32 *ib = p->ib->ptr; unsigned array_mode; - + u32 format; if (G_0280A0_TILE_MODE(track->cb_color_info[i])) { dev_warn(p->dev, "FMASK or CMASK buffer are not supported by this kernel\n"); return -EINVAL; } size = radeon_bo_size(track->cb_color_bo[i]) - track->cb_color_bo_offset[i]; - if (r600_bpe_from_format(&bpe, G_0280A0_FORMAT(track->cb_color_info[i]))) { + format = G_0280A0_FORMAT(track->cb_color_info[i]); + if (!fmt_is_valid_color(format)) { dev_warn(p->dev, "%s:%d cb invalid format %d for %d (0x%08X)\n", - __func__, __LINE__, G_0280A0_FORMAT(track->cb_color_info[i]), + __func__, __LINE__, format, i, track->cb_color_info[i]); return -EINVAL; } @@ -267,7 +360,7 @@ static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i) array_check.nbanks = track->nbanks; array_check.npipes = track->npipes; array_check.nsamples = track->nsamples; - array_check.bpe = bpe; + array_check.blocksize = fmt_get_blocksize(format); if (r600_get_array_mode_alignment(&array_check, &pitch_align, &height_align, &depth_align, &base_align)) { dev_warn(p->dev, "%s invalid tiling %d for %d (0x%08X)\n", __func__, @@ -295,22 +388,23 @@ static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i) } if (!IS_ALIGNED(pitch, pitch_align)) { - dev_warn(p->dev, "%s:%d cb pitch (%d) invalid\n", - __func__, __LINE__, pitch); + dev_warn(p->dev, "%s:%d cb pitch (%d, 0x%x, %d) invalid\n", + __func__, __LINE__, pitch, pitch_align, array_mode); return -EINVAL; } if (!IS_ALIGNED(height, height_align)) { - dev_warn(p->dev, "%s:%d cb height (%d) invalid\n", - __func__, __LINE__, height); + dev_warn(p->dev, "%s:%d cb height (%d, 0x%x, %d) invalid\n", + __func__, __LINE__, height, height_align, array_mode); return -EINVAL; } if (!IS_ALIGNED(base_offset, base_align)) { - dev_warn(p->dev, "%s offset[%d] 0x%llx not aligned\n", __func__, i, base_offset); + dev_warn(p->dev, "%s offset[%d] 0x%llx 0x%llx, %d not aligned\n", __func__, i, + base_offset, base_align, array_mode); return -EINVAL; } /* check offset */ - tmp = height * pitch * bpe; + tmp = fmt_get_nblocksy(format, height) * fmt_get_nblocksx(format, pitch) * fmt_get_blocksize(format); if ((tmp + track->cb_color_bo_offset[i]) > radeon_bo_size(track->cb_color_bo[i])) { if (array_mode == V_0280A0_ARRAY_LINEAR_GENERAL) { /* the initial DDX does bad things with the CB size occasionally */ @@ -320,7 +414,10 @@ static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i) * broken userspace. */ } else { - dev_warn(p->dev, "%s offset[%d] %d %d %lu too big\n", __func__, i, track->cb_color_bo_offset[i], tmp, radeon_bo_size(track->cb_color_bo[i])); + dev_warn(p->dev, "%s offset[%d] %d %d %d %lu too big\n", __func__, i, + array_mode, + track->cb_color_bo_offset[i], tmp, + radeon_bo_size(track->cb_color_bo[i])); return -EINVAL; } } @@ -432,7 +529,7 @@ static int r600_cs_track_check(struct radeon_cs_parser *p) array_check.nbanks = track->nbanks; array_check.npipes = track->npipes; array_check.nsamples = track->nsamples; - array_check.bpe = bpe; + array_check.blocksize = bpe; if (r600_get_array_mode_alignment(&array_check, &pitch_align, &height_align, &depth_align, &base_align)) { dev_warn(p->dev, "%s invalid tiling %d (0x%08X)\n", __func__, @@ -455,17 +552,18 @@ static int r600_cs_track_check(struct radeon_cs_parser *p) } if (!IS_ALIGNED(pitch, pitch_align)) { - dev_warn(p->dev, "%s:%d db pitch (%d) invalid\n", - __func__, __LINE__, pitch); + dev_warn(p->dev, "%s:%d db pitch (%d, 0x%x, %d) invalid\n", + __func__, __LINE__, pitch, pitch_align, array_mode); return -EINVAL; } if (!IS_ALIGNED(height, height_align)) { - dev_warn(p->dev, "%s:%d db height (%d) invalid\n", - __func__, __LINE__, height); + dev_warn(p->dev, "%s:%d db height (%d, 0x%x, %d) invalid\n", + __func__, __LINE__, height, height_align, array_mode); return -EINVAL; } if (!IS_ALIGNED(base_offset, base_align)) { - dev_warn(p->dev, "%s offset[%d] 0x%llx not aligned\n", __func__, i, base_offset); + dev_warn(p->dev, "%s offset[%d] 0x%llx, 0x%llx, %d not aligned\n", __func__, i, + base_offset, base_align, array_mode); return -EINVAL; } @@ -473,9 +571,10 @@ static int r600_cs_track_check(struct radeon_cs_parser *p) nviews = G_028004_SLICE_MAX(track->db_depth_view) + 1; tmp = ntiles * bpe * 64 * nviews; if ((tmp + track->db_offset) > radeon_bo_size(track->db_bo)) { - dev_warn(p->dev, "z/stencil buffer too small (0x%08X %d %d %d -> %u have %lu)\n", - track->db_depth_size, ntiles, nviews, bpe, tmp + track->db_offset, - radeon_bo_size(track->db_bo)); + dev_warn(p->dev, "z/stencil buffer (%d) too small (0x%08X %d %d %d -> %u have %lu)\n", + array_mode, + track->db_depth_size, ntiles, nviews, bpe, tmp + track->db_offset, + radeon_bo_size(track->db_bo)); return -EINVAL; } } @@ -681,33 +780,28 @@ static int r600_cs_packet_parse_vline(struct radeon_cs_parser *p) if (wait_reg_mem.type != PACKET_TYPE3 || wait_reg_mem.opcode != PACKET3_WAIT_REG_MEM) { DRM_ERROR("vline wait missing WAIT_REG_MEM segment\n"); - r = -EINVAL; - return r; + return -EINVAL; } wait_reg_mem_info = radeon_get_ib_value(p, wait_reg_mem.idx + 1); /* bit 4 is reg (0) or mem (1) */ if (wait_reg_mem_info & 0x10) { DRM_ERROR("vline WAIT_REG_MEM waiting on MEM rather than REG\n"); - r = -EINVAL; - return r; + return -EINVAL; } /* waiting for value to be equal */ if ((wait_reg_mem_info & 0x7) != 0x3) { DRM_ERROR("vline WAIT_REG_MEM function not equal\n"); - r = -EINVAL; - return r; + return -EINVAL; } if ((radeon_get_ib_value(p, wait_reg_mem.idx + 2) << 2) != AVIVO_D1MODE_VLINE_STATUS) { DRM_ERROR("vline WAIT_REG_MEM bad reg\n"); - r = -EINVAL; - return r; + return -EINVAL; } if (radeon_get_ib_value(p, wait_reg_mem.idx + 5) != AVIVO_D1MODE_VLINE_STAT) { DRM_ERROR("vline WAIT_REG_MEM bad bit mask\n"); - r = -EINVAL; - return r; + return -EINVAL; } /* jump over the NOP */ @@ -726,8 +820,7 @@ static int r600_cs_packet_parse_vline(struct radeon_cs_parser *p) obj = drm_mode_object_find(p->rdev->ddev, crtc_id, DRM_MODE_OBJECT_CRTC); if (!obj) { DRM_ERROR("cannot find crtc %d\n", crtc_id); - r = -EINVAL; - goto out; + return -EINVAL; } crtc = obj_to_crtc(obj); radeon_crtc = to_radeon_crtc(crtc); @@ -750,14 +843,13 @@ static int r600_cs_packet_parse_vline(struct radeon_cs_parser *p) break; default: DRM_ERROR("unknown crtc reloc\n"); - r = -EINVAL; - goto out; + return -EINVAL; } ib[h_idx] = header; ib[h_idx + 4] = AVIVO_D2MODE_VLINE_STATUS >> 2; } -out: - return r; + + return 0; } static int r600_packet0_check(struct radeon_cs_parser *p, @@ -1107,39 +1199,61 @@ static inline int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx return 0; } -static inline unsigned minify(unsigned size, unsigned levels) +static inline unsigned mip_minify(unsigned size, unsigned level) { - size = size >> levels; - if (size < 1) - size = 1; - return size; + unsigned val; + + val = max(1U, size >> level); + if (level > 0) + val = roundup_pow_of_two(val); + return val; } -static void r600_texture_size(unsigned nfaces, unsigned blevel, unsigned nlevels, - unsigned w0, unsigned h0, unsigned d0, unsigned bpe, - unsigned pitch_align, +static void r600_texture_size(unsigned nfaces, unsigned blevel, unsigned llevel, + unsigned w0, unsigned h0, unsigned d0, unsigned format, + unsigned block_align, unsigned height_align, unsigned base_align, unsigned *l0_size, unsigned *mipmap_size) { - unsigned offset, i, level, face; - unsigned width, height, depth, rowstride, size; - - w0 = minify(w0, 0); - h0 = minify(h0, 0); - d0 = minify(d0, 0); + unsigned offset, i, level; + unsigned width, height, depth, size; + unsigned blocksize; + unsigned nbx, nby; + unsigned nlevels = llevel - blevel + 1; + + *l0_size = -1; + blocksize = fmt_get_blocksize(format); + + w0 = mip_minify(w0, 0); + h0 = mip_minify(h0, 0); + d0 = mip_minify(d0, 0); for(i = 0, offset = 0, level = blevel; i < nlevels; i++, level++) { - width = minify(w0, i); - height = minify(h0, i); - depth = minify(d0, i); - for(face = 0; face < nfaces; face++) { - rowstride = ALIGN((width * bpe), pitch_align); - size = height * rowstride * depth; - offset += size; - offset = (offset + 0x1f) & ~0x1f; - } + width = mip_minify(w0, i); + nbx = fmt_get_nblocksx(format, width); + + nbx = round_up(nbx, block_align); + + height = mip_minify(h0, i); + nby = fmt_get_nblocksy(format, height); + nby = round_up(nby, height_align); + + depth = mip_minify(d0, i); + + size = nbx * nby * blocksize; + if (nfaces) + size *= nfaces; + else + size *= depth; + + if (i == 0) + *l0_size = size; + + if (i == 0 || i == 1) + offset = round_up(offset, base_align); + + offset += size; } - *l0_size = ALIGN((w0 * bpe), pitch_align) * h0 * d0; *mipmap_size = offset; - if (!nlevels) + if (llevel == 0) *mipmap_size = *l0_size; if (!blevel) *mipmap_size -= *l0_size; @@ -1163,11 +1277,13 @@ static inline int r600_check_texture_resource(struct radeon_cs_parser *p, u32 i u32 tiling_flags) { struct r600_cs_track *track = p->track; - u32 nfaces, nlevels, blevel, w0, h0, d0, bpe = 0; - u32 word0, word1, l0_size, mipmap_size; + u32 nfaces, llevel, blevel, w0, h0, d0; + u32 word0, word1, l0_size, mipmap_size, word2, word3; u32 height_align, pitch, pitch_align, depth_align; + u32 array, barray, larray; u64 base_align; struct array_mode_checker array_check; + u32 format; /* on legacy kernel we don't perform advanced check */ if (p->rdev == NULL) @@ -1193,19 +1309,25 @@ static inline int r600_check_texture_resource(struct radeon_cs_parser *p, u32 i case V_038000_SQ_TEX_DIM_3D: break; case V_038000_SQ_TEX_DIM_CUBEMAP: - nfaces = 6; + if (p->family >= CHIP_RV770) + nfaces = 8; + else + nfaces = 6; break; case V_038000_SQ_TEX_DIM_1D_ARRAY: case V_038000_SQ_TEX_DIM_2D_ARRAY: + array = 1; + break; case V_038000_SQ_TEX_DIM_2D_MSAA: case V_038000_SQ_TEX_DIM_2D_ARRAY_MSAA: default: dev_warn(p->dev, "this kernel doesn't support %d texture dim\n", G_038000_DIM(word0)); return -EINVAL; } - if (r600_bpe_from_format(&bpe, G_038004_DATA_FORMAT(word1))) { + format = G_038004_DATA_FORMAT(word1); + if (!fmt_is_valid_texture(format)) { dev_warn(p->dev, "%s:%d texture invalid format %d\n", - __func__, __LINE__, G_038004_DATA_FORMAT(word1)); + __func__, __LINE__, format); return -EINVAL; } @@ -1216,7 +1338,7 @@ static inline int r600_check_texture_resource(struct radeon_cs_parser *p, u32 i array_check.nbanks = track->nbanks; array_check.npipes = track->npipes; array_check.nsamples = 1; - array_check.bpe = bpe; + array_check.blocksize = fmt_get_blocksize(format); if (r600_get_array_mode_alignment(&array_check, &pitch_align, &height_align, &depth_align, &base_align)) { dev_warn(p->dev, "%s:%d tex array mode (%d) invalid\n", @@ -1227,40 +1349,49 @@ static inline int r600_check_texture_resource(struct radeon_cs_parser *p, u32 i /* XXX check height as well... */ if (!IS_ALIGNED(pitch, pitch_align)) { - dev_warn(p->dev, "%s:%d tex pitch (%d) invalid\n", - __func__, __LINE__, pitch); + dev_warn(p->dev, "%s:%d tex pitch (%d, 0x%x, %d) invalid\n", + __func__, __LINE__, pitch, pitch_align, G_038000_TILE_MODE(word0)); return -EINVAL; } if (!IS_ALIGNED(base_offset, base_align)) { - dev_warn(p->dev, "%s:%d tex base offset (0x%llx) invalid\n", - __func__, __LINE__, base_offset); + dev_warn(p->dev, "%s:%d tex base offset (0x%llx, 0x%llx, %d) invalid\n", + __func__, __LINE__, base_offset, base_align, G_038000_TILE_MODE(word0)); return -EINVAL; } if (!IS_ALIGNED(mip_offset, base_align)) { - dev_warn(p->dev, "%s:%d tex mip offset (0x%llx) invalid\n", - __func__, __LINE__, mip_offset); + dev_warn(p->dev, "%s:%d tex mip offset (0x%llx, 0x%llx, %d) invalid\n", + __func__, __LINE__, mip_offset, base_align, G_038000_TILE_MODE(word0)); return -EINVAL; } + word2 = radeon_get_ib_value(p, idx + 2) << 8; + word3 = radeon_get_ib_value(p, idx + 3) << 8; + word0 = radeon_get_ib_value(p, idx + 4); word1 = radeon_get_ib_value(p, idx + 5); blevel = G_038010_BASE_LEVEL(word0); - nlevels = G_038014_LAST_LEVEL(word1); - r600_texture_size(nfaces, blevel, nlevels, w0, h0, d0, bpe, - (pitch_align * bpe), + llevel = G_038014_LAST_LEVEL(word1); + if (array == 1) { + barray = G_038014_BASE_ARRAY(word1); + larray = G_038014_LAST_ARRAY(word1); + + nfaces = larray - barray + 1; + } + r600_texture_size(nfaces, blevel, llevel, w0, h0, d0, format, + pitch_align, height_align, base_align, &l0_size, &mipmap_size); /* using get ib will give us the offset into the texture bo */ - word0 = radeon_get_ib_value(p, idx + 2) << 8; - if ((l0_size + word0) > radeon_bo_size(texture)) { + if ((l0_size + word2) > radeon_bo_size(texture)) { dev_warn(p->dev, "texture bo too small (%d %d %d %d -> %d have %ld)\n", - w0, h0, bpe, word0, l0_size, radeon_bo_size(texture)); + w0, h0, format, word2, l0_size, radeon_bo_size(texture)); + dev_warn(p->dev, "alignments %d %d %d %lld\n", pitch, pitch_align, height_align, base_align); return -EINVAL; } /* using get ib will give us the offset into the mipmap bo */ - word0 = radeon_get_ib_value(p, idx + 3) << 8; - if ((mipmap_size + word0) > radeon_bo_size(mipmap)) { + word3 = radeon_get_ib_value(p, idx + 3) << 8; + if ((mipmap_size + word3) > radeon_bo_size(mipmap)) { /*dev_warn(p->dev, "mipmap bo too small (%d %d %d %d %d %d -> %d have %ld)\n", - w0, h0, bpe, blevel, nlevels, word0, mipmap_size, radeon_bo_size(texture));*/ + w0, h0, format, blevel, nlevels, word3, mipmap_size, radeon_bo_size(texture));*/ } return 0; } @@ -1283,6 +1414,38 @@ static int r600_packet3_check(struct radeon_cs_parser *p, idx_value = radeon_get_ib_value(p, idx); switch (pkt->opcode) { + case PACKET3_SET_PREDICATION: + { + int pred_op; + int tmp; + if (pkt->count != 1) { + DRM_ERROR("bad SET PREDICATION\n"); + return -EINVAL; + } + + tmp = radeon_get_ib_value(p, idx + 1); + pred_op = (tmp >> 16) & 0x7; + + /* for the clear predicate operation */ + if (pred_op == 0) + return 0; + + if (pred_op > 2) { + DRM_ERROR("bad SET PREDICATION operation %d\n", pred_op); + return -EINVAL; + } + + r = r600_cs_packet_next_reloc(p, &reloc); + if (r) { + DRM_ERROR("bad SET PREDICATION\n"); + return -EINVAL; + } + + ib[idx + 0] = idx_value + (u32)(reloc->lobj.gpu_offset & 0xffffffff); + ib[idx + 1] = tmp + (upper_32_bits(reloc->lobj.gpu_offset) & 0xff); + } + break; + case PACKET3_START_3D_CMDBUF: if (p->family >= CHIP_RV770 || pkt->count) { DRM_ERROR("bad START_3D\n"); diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c index e6a58ed..50db6d6 100644 --- a/drivers/gpu/drm/radeon/r600_hdmi.c +++ b/drivers/gpu/drm/radeon/r600_hdmi.c @@ -26,6 +26,7 @@ #include "drmP.h" #include "radeon_drm.h" #include "radeon.h" +#include "radeon_asic.h" #include "atom.h" /* diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h index a5d898b..b2b944bc 100644 --- a/drivers/gpu/drm/radeon/r600d.h +++ b/drivers/gpu/drm/radeon/r600d.h @@ -154,13 +154,14 @@ #define ROQ_IB2_START(x) ((x) << 8) #define CP_RB_BASE 0xC100 #define CP_RB_CNTL 0xC104 -#define RB_BUFSZ(x) ((x)<<0) -#define RB_BLKSZ(x) ((x)<<8) -#define RB_NO_UPDATE (1<<27) -#define RB_RPTR_WR_ENA (1<<31) +#define RB_BUFSZ(x) ((x) << 0) +#define RB_BLKSZ(x) ((x) << 8) +#define RB_NO_UPDATE (1 << 27) +#define RB_RPTR_WR_ENA (1 << 31) #define BUF_SWAP_32BIT (2 << 16) #define CP_RB_RPTR 0x8700 #define CP_RB_RPTR_ADDR 0xC10C +#define RB_RPTR_SWAP(x) ((x) << 0) #define CP_RB_RPTR_ADDR_HI 0xC110 #define CP_RB_RPTR_WR 0xC108 #define CP_RB_WPTR 0xC114 @@ -1303,6 +1304,11 @@ #define V_038004_FMT_16_16_16_FLOAT 0x0000002E #define V_038004_FMT_32_32_32 0x0000002F #define V_038004_FMT_32_32_32_FLOAT 0x00000030 +#define V_038004_FMT_BC1 0x00000031 +#define V_038004_FMT_BC2 0x00000032 +#define V_038004_FMT_BC3 0x00000033 +#define V_038004_FMT_BC4 0x00000034 +#define V_038004_FMT_BC5 0x00000035 #define R_038010_SQ_TEX_RESOURCE_WORD4_0 0x038010 #define S_038010_FORMAT_COMP_X(x) (((x) & 0x3) << 0) #define G_038010_FORMAT_COMP_X(x) (((x) >> 0) & 0x3) diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 56c48b6..cfe3af1 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -258,8 +258,9 @@ struct radeon_bo { int surface_reg; /* Constant after initialization */ struct radeon_device *rdev; - struct drm_gem_object *gobj; + struct drm_gem_object gem_base; }; +#define gem_to_radeon_bo(gobj) container_of((gobj), struct radeon_bo, gem_base) struct radeon_bo_list { struct ttm_validate_buffer tv; @@ -288,6 +289,15 @@ int radeon_gem_object_pin(struct drm_gem_object *obj, uint32_t pin_domain, uint64_t *gpu_addr); void radeon_gem_object_unpin(struct drm_gem_object *obj); +int radeon_mode_dumb_create(struct drm_file *file_priv, + struct drm_device *dev, + struct drm_mode_create_dumb *args); +int radeon_mode_dumb_mmap(struct drm_file *filp, + struct drm_device *dev, + uint32_t handle, uint64_t *offset_p); +int radeon_mode_dumb_destroy(struct drm_file *file_priv, + struct drm_device *dev, + uint32_t handle); /* * GART structures, functions & helpers @@ -319,6 +329,7 @@ struct radeon_gart { union radeon_gart_table table; struct page **pages; dma_addr_t *pages_addr; + bool *ttm_alloced; bool ready; }; @@ -331,7 +342,8 @@ void radeon_gart_fini(struct radeon_device *rdev); void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset, int pages); int radeon_gart_bind(struct radeon_device *rdev, unsigned offset, - int pages, struct page **pagelist); + int pages, struct page **pagelist, + dma_addr_t *dma_addr); /* @@ -345,7 +357,6 @@ struct radeon_mc { * about vram size near mc fb location */ u64 mc_vram_size; u64 visible_vram_size; - u64 active_vram_size; u64 gtt_size; u64 gtt_start; u64 gtt_end; @@ -652,6 +663,8 @@ struct radeon_wb { #define RADEON_WB_SCRATCH_OFFSET 0 #define RADEON_WB_CP_RPTR_OFFSET 1024 +#define RADEON_WB_CP1_RPTR_OFFSET 1280 +#define RADEON_WB_CP2_RPTR_OFFSET 1536 #define R600_WB_IH_WPTR_OFFSET 2048 #define R600_WB_EVENT_OFFSET 3072 @@ -1038,12 +1051,52 @@ struct evergreen_asic { struct r100_gpu_lockup lockup; }; +struct cayman_asic { + unsigned max_shader_engines; + unsigned max_pipes_per_simd; + unsigned max_tile_pipes; + unsigned max_simds_per_se; + unsigned max_backends_per_se; + unsigned max_texture_channel_caches; + unsigned max_gprs; + unsigned max_threads; + unsigned max_gs_threads; + unsigned max_stack_entries; + unsigned sx_num_of_sets; + unsigned sx_max_export_size; + unsigned sx_max_export_pos_size; + unsigned sx_max_export_smx_size; + unsigned max_hw_contexts; + unsigned sq_num_cf_insts; + unsigned sc_prim_fifo_size; + unsigned sc_hiz_tile_fifo_size; + unsigned sc_earlyz_tile_fifo_size; + + unsigned num_shader_engines; + unsigned num_shader_pipes_per_simd; + unsigned num_tile_pipes; + unsigned num_simds_per_se; + unsigned num_backends_per_se; + unsigned backend_disable_mask_per_asic; + unsigned backend_map; + unsigned num_texture_channel_caches; + unsigned mem_max_burst_length_bytes; + unsigned mem_row_size_in_kb; + unsigned shader_engine_tile_size; + unsigned num_gpus; + unsigned multi_gpu_tile_size; + + unsigned tile_config; + struct r100_gpu_lockup lockup; +}; + union radeon_asic_config { struct r300_asic r300; struct r100_asic r100; struct r600_asic r600; struct rv770_asic rv770; struct evergreen_asic evergreen; + struct cayman_asic cayman; }; /* @@ -1134,6 +1187,9 @@ struct radeon_device { struct radeon_mman mman; struct radeon_fence_driver fence_drv; struct radeon_cp cp; + /* cayman compute rings */ + struct radeon_cp cp1; + struct radeon_cp cp2; struct radeon_ib_pool ib_pool; struct radeon_irq irq; struct radeon_asic *asic; @@ -1186,19 +1242,6 @@ int radeon_device_init(struct radeon_device *rdev, void radeon_device_fini(struct radeon_device *rdev); int radeon_gpu_wait_for_idle(struct radeon_device *rdev); -/* r600 blit */ -int r600_blit_prepare_copy(struct radeon_device *rdev, int size_bytes); -void r600_blit_done_copy(struct radeon_device *rdev, struct radeon_fence *fence); -void r600_kms_blit_copy(struct radeon_device *rdev, - u64 src_gpu_addr, u64 dst_gpu_addr, - int size_bytes); -/* evergreen blit */ -int evergreen_blit_prepare_copy(struct radeon_device *rdev, int size_bytes); -void evergreen_blit_done_copy(struct radeon_device *rdev, struct radeon_fence *fence); -void evergreen_kms_blit_copy(struct radeon_device *rdev, - u64 src_gpu_addr, u64 dst_gpu_addr, - int size_bytes); - static inline uint32_t r100_mm_rreg(struct radeon_device *rdev, uint32_t reg) { if (reg < rdev->rmmio_size) @@ -1448,63 +1491,17 @@ extern void radeon_vram_location(struct radeon_device *rdev, struct radeon_mc *m extern void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc); extern int radeon_resume_kms(struct drm_device *dev); extern int radeon_suspend_kms(struct drm_device *dev, pm_message_t state); +extern void radeon_ttm_set_active_vram_size(struct radeon_device *rdev, u64 size); -/* r600, rv610, rv630, rv620, rv635, rv670, rs780, rs880 */ -extern bool r600_card_posted(struct radeon_device *rdev); -extern void r600_cp_stop(struct radeon_device *rdev); -extern int r600_cp_start(struct radeon_device *rdev); -extern void r600_ring_init(struct radeon_device *rdev, unsigned ring_size); -extern int r600_cp_resume(struct radeon_device *rdev); -extern void r600_cp_fini(struct radeon_device *rdev); -extern int r600_count_pipe_bits(uint32_t val); -extern int r600_mc_wait_for_idle(struct radeon_device *rdev); -extern int r600_pcie_gart_init(struct radeon_device *rdev); -extern void r600_pcie_gart_tlb_flush(struct radeon_device *rdev); -extern int r600_ib_test(struct radeon_device *rdev); -extern int r600_ring_test(struct radeon_device *rdev); -extern void r600_scratch_init(struct radeon_device *rdev); -extern int r600_blit_init(struct radeon_device *rdev); -extern void r600_blit_fini(struct radeon_device *rdev); -extern int r600_init_microcode(struct radeon_device *rdev); -extern int r600_asic_reset(struct radeon_device *rdev); -/* r600 irq */ -extern int r600_irq_init(struct radeon_device *rdev); -extern void r600_irq_fini(struct radeon_device *rdev); -extern void r600_ih_ring_init(struct radeon_device *rdev, unsigned ring_size); -extern int r600_irq_set(struct radeon_device *rdev); -extern void r600_irq_suspend(struct radeon_device *rdev); -extern void r600_disable_interrupts(struct radeon_device *rdev); -extern void r600_rlc_stop(struct radeon_device *rdev); -/* r600 audio */ -extern int r600_audio_init(struct radeon_device *rdev); -extern int r600_audio_tmds_index(struct drm_encoder *encoder); -extern void r600_audio_set_clock(struct drm_encoder *encoder, int clock); -extern int r600_audio_channels(struct radeon_device *rdev); -extern int r600_audio_bits_per_sample(struct radeon_device *rdev); -extern int r600_audio_rate(struct radeon_device *rdev); -extern uint8_t r600_audio_status_bits(struct radeon_device *rdev); -extern uint8_t r600_audio_category_code(struct radeon_device *rdev); -extern void r600_audio_schedule_polling(struct radeon_device *rdev); -extern void r600_audio_enable_polling(struct drm_encoder *encoder); -extern void r600_audio_disable_polling(struct drm_encoder *encoder); -extern void r600_audio_fini(struct radeon_device *rdev); -extern void r600_hdmi_init(struct drm_encoder *encoder); +/* + * r600 functions used by radeon_encoder.c + */ extern void r600_hdmi_enable(struct drm_encoder *encoder); extern void r600_hdmi_disable(struct drm_encoder *encoder); extern void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode); -extern int r600_hdmi_buffer_status_changed(struct drm_encoder *encoder); -extern void r600_hdmi_update_audio_settings(struct drm_encoder *encoder); - -extern void r700_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc); -extern void r700_cp_stop(struct radeon_device *rdev); -extern void r700_cp_fini(struct radeon_device *rdev); -extern void evergreen_disable_interrupt_state(struct radeon_device *rdev); -extern int evergreen_irq_set(struct radeon_device *rdev); -extern int evergreen_blit_init(struct radeon_device *rdev); -extern void evergreen_blit_fini(struct radeon_device *rdev); extern int ni_init_microcode(struct radeon_device *rdev); -extern int btc_mc_load_microcode(struct radeon_device *rdev); +extern int ni_mc_load_microcode(struct radeon_device *rdev); /* radeon_acpi.c */ #if defined(CONFIG_ACPI) @@ -1513,14 +1510,6 @@ extern int radeon_acpi_init(struct radeon_device *rdev); static inline int radeon_acpi_init(struct radeon_device *rdev) { return 0; } #endif -/* evergreen */ -struct evergreen_mc_save { - u32 vga_control[6]; - u32 vga_render_control; - u32 vga_hdp_control; - u32 crtc_control[6]; -}; - #include "radeon_object.h" #endif diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index e75d63b..eb888ee 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -834,6 +834,9 @@ static struct radeon_asic sumo_asic = { .pm_finish = &evergreen_pm_finish, .pm_init_profile = &rs780_pm_init_profile, .pm_get_dynpm_state = &r600_pm_get_dynpm_state, + .pre_page_flip = &evergreen_pre_page_flip, + .page_flip = &evergreen_page_flip, + .post_page_flip = &evergreen_post_page_flip, }; static struct radeon_asic btc_asic = { @@ -882,6 +885,52 @@ static struct radeon_asic btc_asic = { .post_page_flip = &evergreen_post_page_flip, }; +static struct radeon_asic cayman_asic = { + .init = &cayman_init, + .fini = &cayman_fini, + .suspend = &cayman_suspend, + .resume = &cayman_resume, + .cp_commit = &r600_cp_commit, + .gpu_is_lockup = &cayman_gpu_is_lockup, + .asic_reset = &cayman_asic_reset, + .vga_set_state = &r600_vga_set_state, + .gart_tlb_flush = &cayman_pcie_gart_tlb_flush, + .gart_set_page = &rs600_gart_set_page, + .ring_test = &r600_ring_test, + .ring_ib_execute = &evergreen_ring_ib_execute, + .irq_set = &evergreen_irq_set, + .irq_process = &evergreen_irq_process, + .get_vblank_counter = &evergreen_get_vblank_counter, + .fence_ring_emit = &r600_fence_ring_emit, + .cs_parse = &evergreen_cs_parse, + .copy_blit = NULL, + .copy_dma = NULL, + .copy = NULL, + .get_engine_clock = &radeon_atom_get_engine_clock, + .set_engine_clock = &radeon_atom_set_engine_clock, + .get_memory_clock = &radeon_atom_get_memory_clock, + .set_memory_clock = &radeon_atom_set_memory_clock, + .get_pcie_lanes = NULL, + .set_pcie_lanes = NULL, + .set_clock_gating = NULL, + .set_surface_reg = r600_set_surface_reg, + .clear_surface_reg = r600_clear_surface_reg, + .bandwidth_update = &evergreen_bandwidth_update, + .hpd_init = &evergreen_hpd_init, + .hpd_fini = &evergreen_hpd_fini, + .hpd_sense = &evergreen_hpd_sense, + .hpd_set_polarity = &evergreen_hpd_set_polarity, + .gui_idle = &r600_gui_idle, + .pm_misc = &evergreen_pm_misc, + .pm_prepare = &evergreen_pm_prepare, + .pm_finish = &evergreen_pm_finish, + .pm_init_profile = &r600_pm_init_profile, + .pm_get_dynpm_state = &r600_pm_get_dynpm_state, + .pre_page_flip = &evergreen_pre_page_flip, + .page_flip = &evergreen_page_flip, + .post_page_flip = &evergreen_post_page_flip, +}; + int radeon_asic_init(struct radeon_device *rdev) { radeon_register_accessor_init(rdev); @@ -974,6 +1023,9 @@ int radeon_asic_init(struct radeon_device *rdev) case CHIP_CAICOS: rdev->asic = &btc_asic; break; + case CHIP_CAYMAN: + rdev->asic = &cayman_asic; + break; default: /* FIXME: not supported yet */ return -EINVAL; diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h index c59bd98..3d7a0d7 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.h +++ b/drivers/gpu/drm/radeon/radeon_asic.h @@ -57,8 +57,6 @@ int r100_init(struct radeon_device *rdev); void r100_fini(struct radeon_device *rdev); int r100_suspend(struct radeon_device *rdev); int r100_resume(struct radeon_device *rdev); -uint32_t r100_mm_rreg(struct radeon_device *rdev, uint32_t reg); -void r100_mm_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); void r100_vga_set_state(struct radeon_device *rdev, bool state); bool r100_gpu_is_lockup(struct radeon_device *rdev); int r100_asic_reset(struct radeon_device *rdev); @@ -164,8 +162,6 @@ extern void r300_fence_ring_emit(struct radeon_device *rdev, 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 uint32_t rv370_pcie_rreg(struct radeon_device *rdev, uint32_t reg); -extern void rv370_pcie_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); 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); @@ -208,7 +204,6 @@ void rs400_gart_adjust_size(struct radeon_device *rdev); void rs400_gart_disable(struct radeon_device *rdev); void rs400_gart_fini(struct radeon_device *rdev); - /* * rs600. */ @@ -270,8 +265,6 @@ void rv515_fini(struct radeon_device *rdev); uint32_t rv515_mc_rreg(struct radeon_device *rdev, uint32_t reg); void rv515_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); void rv515_ring_start(struct radeon_device *rdev); -uint32_t rv515_pcie_rreg(struct radeon_device *rdev, uint32_t reg); -void rv515_pcie_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); void rv515_bandwidth_update(struct radeon_device *rdev); int rv515_resume(struct radeon_device *rdev); int rv515_suspend(struct radeon_device *rdev); @@ -307,14 +300,13 @@ void r600_pciep_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); int r600_cs_parse(struct radeon_cs_parser *p); void r600_fence_ring_emit(struct radeon_device *rdev, struct radeon_fence *fence); -int r600_irq_process(struct radeon_device *rdev); -int r600_irq_set(struct radeon_device *rdev); bool r600_gpu_is_lockup(struct radeon_device *rdev); int r600_asic_reset(struct radeon_device *rdev); int r600_set_surface_reg(struct radeon_device *rdev, int reg, uint32_t tiling_flags, uint32_t pitch, uint32_t offset, uint32_t obj_size); void r600_clear_surface_reg(struct radeon_device *rdev, int reg); +int r600_ib_test(struct radeon_device *rdev); void r600_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib); int r600_ring_test(struct radeon_device *rdev); int r600_copy_blit(struct radeon_device *rdev, @@ -333,6 +325,50 @@ extern void rs780_pm_init_profile(struct radeon_device *rdev); extern void r600_pm_get_dynpm_state(struct radeon_device *rdev); extern void r600_set_pcie_lanes(struct radeon_device *rdev, int lanes); extern int r600_get_pcie_lanes(struct radeon_device *rdev); +bool r600_card_posted(struct radeon_device *rdev); +void r600_cp_stop(struct radeon_device *rdev); +int r600_cp_start(struct radeon_device *rdev); +void r600_ring_init(struct radeon_device *rdev, unsigned ring_size); +int r600_cp_resume(struct radeon_device *rdev); +void r600_cp_fini(struct radeon_device *rdev); +int r600_count_pipe_bits(uint32_t val); +int r600_mc_wait_for_idle(struct radeon_device *rdev); +int r600_pcie_gart_init(struct radeon_device *rdev); +void r600_scratch_init(struct radeon_device *rdev); +int r600_blit_init(struct radeon_device *rdev); +void r600_blit_fini(struct radeon_device *rdev); +int r600_init_microcode(struct radeon_device *rdev); +/* r600 irq */ +int r600_irq_process(struct radeon_device *rdev); +int r600_irq_init(struct radeon_device *rdev); +void r600_irq_fini(struct radeon_device *rdev); +void r600_ih_ring_init(struct radeon_device *rdev, unsigned ring_size); +int r600_irq_set(struct radeon_device *rdev); +void r600_irq_suspend(struct radeon_device *rdev); +void r600_disable_interrupts(struct radeon_device *rdev); +void r600_rlc_stop(struct radeon_device *rdev); +/* r600 audio */ +int r600_audio_init(struct radeon_device *rdev); +int r600_audio_tmds_index(struct drm_encoder *encoder); +void r600_audio_set_clock(struct drm_encoder *encoder, int clock); +int r600_audio_channels(struct radeon_device *rdev); +int r600_audio_bits_per_sample(struct radeon_device *rdev); +int r600_audio_rate(struct radeon_device *rdev); +uint8_t r600_audio_status_bits(struct radeon_device *rdev); +uint8_t r600_audio_category_code(struct radeon_device *rdev); +void r600_audio_schedule_polling(struct radeon_device *rdev); +void r600_audio_enable_polling(struct drm_encoder *encoder); +void r600_audio_disable_polling(struct drm_encoder *encoder); +void r600_audio_fini(struct radeon_device *rdev); +void r600_hdmi_init(struct drm_encoder *encoder); +int r600_hdmi_buffer_status_changed(struct drm_encoder *encoder); +void r600_hdmi_update_audio_settings(struct drm_encoder *encoder); +/* r600 blit */ +int r600_blit_prepare_copy(struct radeon_device *rdev, int size_bytes); +void r600_blit_done_copy(struct radeon_device *rdev, struct radeon_fence *fence); +void r600_kms_blit_copy(struct radeon_device *rdev, + u64 src_gpu_addr, u64 dst_gpu_addr, + int size_bytes); /* * rv770,rv730,rv710,rv740 @@ -341,12 +377,21 @@ int rv770_init(struct radeon_device *rdev); void rv770_fini(struct radeon_device *rdev); int rv770_suspend(struct radeon_device *rdev); int rv770_resume(struct radeon_device *rdev); -extern void rv770_pm_misc(struct radeon_device *rdev); -extern u32 rv770_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base); +void rv770_pm_misc(struct radeon_device *rdev); +u32 rv770_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base); +void r700_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc); +void r700_cp_stop(struct radeon_device *rdev); +void r700_cp_fini(struct radeon_device *rdev); /* * evergreen */ +struct evergreen_mc_save { + u32 vga_control[6]; + u32 vga_render_control; + u32 vga_hdp_control; + u32 crtc_control[6]; +}; void evergreen_pcie_gart_tlb_flush(struct radeon_device *rdev); int evergreen_init(struct radeon_device *rdev); void evergreen_fini(struct radeon_device *rdev); @@ -374,5 +419,25 @@ extern void evergreen_pm_finish(struct radeon_device *rdev); extern void evergreen_pre_page_flip(struct radeon_device *rdev, int crtc); extern u32 evergreen_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base); extern void evergreen_post_page_flip(struct radeon_device *rdev, int crtc); +void evergreen_disable_interrupt_state(struct radeon_device *rdev); +int evergreen_blit_init(struct radeon_device *rdev); +void evergreen_blit_fini(struct radeon_device *rdev); +/* evergreen blit */ +int evergreen_blit_prepare_copy(struct radeon_device *rdev, int size_bytes); +void evergreen_blit_done_copy(struct radeon_device *rdev, struct radeon_fence *fence); +void evergreen_kms_blit_copy(struct radeon_device *rdev, + u64 src_gpu_addr, u64 dst_gpu_addr, + int size_bytes); + +/* + * cayman + */ +void cayman_pcie_gart_tlb_flush(struct radeon_device *rdev); +int cayman_init(struct radeon_device *rdev); +void cayman_fini(struct radeon_device *rdev); +int cayman_suspend(struct radeon_device *rdev); +int cayman_resume(struct radeon_device *rdev); +bool cayman_gpu_is_lockup(struct radeon_device *rdev); +int cayman_asic_reset(struct radeon_device *rdev); #endif diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index 5c1cc7a..02d5c41 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -88,7 +88,7 @@ static inline struct radeon_i2c_bus_rec radeon_lookup_i2c_gpio(struct radeon_dev /* some evergreen boards have bad data for this entry */ if (ASIC_IS_DCE4(rdev)) { if ((i == 7) && - (gpio->usClkMaskRegisterIndex == 0x1936) && + (le16_to_cpu(gpio->usClkMaskRegisterIndex) == 0x1936) && (gpio->sucI2cId.ucAccess == 0)) { gpio->sucI2cId.ucAccess = 0x97; gpio->ucDataMaskShift = 8; @@ -101,7 +101,7 @@ static inline struct radeon_i2c_bus_rec radeon_lookup_i2c_gpio(struct radeon_dev /* some DCE3 boards have bad data for this entry */ if (ASIC_IS_DCE3(rdev)) { if ((i == 4) && - (gpio->usClkMaskRegisterIndex == 0x1fda) && + (le16_to_cpu(gpio->usClkMaskRegisterIndex) == 0x1fda) && (gpio->sucI2cId.ucAccess == 0x94)) gpio->sucI2cId.ucAccess = 0x14; } @@ -172,7 +172,7 @@ void radeon_atombios_i2c_init(struct radeon_device *rdev) /* some evergreen boards have bad data for this entry */ if (ASIC_IS_DCE4(rdev)) { if ((i == 7) && - (gpio->usClkMaskRegisterIndex == 0x1936) && + (le16_to_cpu(gpio->usClkMaskRegisterIndex) == 0x1936) && (gpio->sucI2cId.ucAccess == 0)) { gpio->sucI2cId.ucAccess = 0x97; gpio->ucDataMaskShift = 8; @@ -185,7 +185,7 @@ void radeon_atombios_i2c_init(struct radeon_device *rdev) /* some DCE3 boards have bad data for this entry */ if (ASIC_IS_DCE3(rdev)) { if ((i == 4) && - (gpio->usClkMaskRegisterIndex == 0x1fda) && + (le16_to_cpu(gpio->usClkMaskRegisterIndex) == 0x1fda) && (gpio->sucI2cId.ucAccess == 0x94)) gpio->sucI2cId.ucAccess = 0x14; } @@ -252,7 +252,7 @@ static inline struct radeon_gpio_rec radeon_lookup_gpio(struct radeon_device *rd pin = &gpio_info->asGPIO_Pin[i]; if (id == pin->ucGPIO_ID) { gpio.id = pin->ucGPIO_ID; - gpio.reg = pin->usGpioPin_AIndex * 4; + gpio.reg = le16_to_cpu(pin->usGpioPin_AIndex) * 4; gpio.mask = (1 << pin->ucGpioPinBitShift); gpio.valid = true; break; @@ -1274,11 +1274,11 @@ bool radeon_atombios_sideport_present(struct radeon_device *rdev) data_offset); switch (crev) { case 1: - if (igp_info->info.ulBootUpMemoryClock) + if (le32_to_cpu(igp_info->info.ulBootUpMemoryClock)) return true; break; case 2: - if (igp_info->info_2.ulBootUpSidePortClock) + if (le32_to_cpu(igp_info->info_2.ulBootUpSidePortClock)) return true; break; default: @@ -1442,7 +1442,7 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev, for (i = 0; i < num_indices; i++) { if ((ss_info->info.asSpreadSpectrum[i].ucClockIndication == id) && - (clock <= ss_info->info.asSpreadSpectrum[i].ulTargetClockRange)) { + (clock <= le32_to_cpu(ss_info->info.asSpreadSpectrum[i].ulTargetClockRange))) { ss->percentage = le16_to_cpu(ss_info->info.asSpreadSpectrum[i].usSpreadSpectrumPercentage); ss->type = ss_info->info.asSpreadSpectrum[i].ucSpreadSpectrumMode; @@ -1456,7 +1456,7 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev, sizeof(ATOM_ASIC_SS_ASSIGNMENT_V2); for (i = 0; i < num_indices; i++) { if ((ss_info->info_2.asSpreadSpectrum[i].ucClockIndication == id) && - (clock <= ss_info->info_2.asSpreadSpectrum[i].ulTargetClockRange)) { + (clock <= le32_to_cpu(ss_info->info_2.asSpreadSpectrum[i].ulTargetClockRange))) { ss->percentage = le16_to_cpu(ss_info->info_2.asSpreadSpectrum[i].usSpreadSpectrumPercentage); ss->type = ss_info->info_2.asSpreadSpectrum[i].ucSpreadSpectrumMode; @@ -1470,7 +1470,7 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev, sizeof(ATOM_ASIC_SS_ASSIGNMENT_V3); for (i = 0; i < num_indices; i++) { if ((ss_info->info_3.asSpreadSpectrum[i].ucClockIndication == id) && - (clock <= ss_info->info_3.asSpreadSpectrum[i].ulTargetClockRange)) { + (clock <= le32_to_cpu(ss_info->info_3.asSpreadSpectrum[i].ulTargetClockRange))) { ss->percentage = le16_to_cpu(ss_info->info_3.asSpreadSpectrum[i].usSpreadSpectrumPercentage); ss->type = ss_info->info_3.asSpreadSpectrum[i].ucSpreadSpectrumMode; @@ -1553,8 +1553,8 @@ struct radeon_encoder_atom_dig *radeon_atombios_get_lvds_info(struct if (misc & ATOM_DOUBLE_CLOCK_MODE) lvds->native_mode.flags |= DRM_MODE_FLAG_DBLSCAN; - lvds->native_mode.width_mm = lvds_info->info.sLCDTiming.usImageHSize; - lvds->native_mode.height_mm = lvds_info->info.sLCDTiming.usImageVSize; + lvds->native_mode.width_mm = le16_to_cpu(lvds_info->info.sLCDTiming.usImageHSize); + lvds->native_mode.height_mm = le16_to_cpu(lvds_info->info.sLCDTiming.usImageVSize); /* set crtc values */ drm_mode_set_crtcinfo(&lvds->native_mode, CRTC_INTERLACE_HALVE_V); @@ -1569,13 +1569,13 @@ struct radeon_encoder_atom_dig *radeon_atombios_get_lvds_info(struct lvds->linkb = false; /* parse the lcd record table */ - if (lvds_info->info.usModePatchTableOffset) { + if (le16_to_cpu(lvds_info->info.usModePatchTableOffset)) { ATOM_FAKE_EDID_PATCH_RECORD *fake_edid_record; ATOM_PANEL_RESOLUTION_PATCH_RECORD *panel_res_record; bool bad_record = false; u8 *record = (u8 *)(mode_info->atom_context->bios + data_offset + - lvds_info->info.usModePatchTableOffset); + le16_to_cpu(lvds_info->info.usModePatchTableOffset)); while (*record != ATOM_RECORD_END_TYPE) { switch (*record) { case LCD_MODE_PATCH_RECORD_MODE_TYPE: @@ -2189,7 +2189,7 @@ static u16 radeon_atombios_get_default_vddc(struct radeon_device *rdev) firmware_info = (union firmware_info *)(mode_info->atom_context->bios + data_offset); - vddc = firmware_info->info_14.usBootUpVDDCVoltage; + vddc = le16_to_cpu(firmware_info->info_14.usBootUpVDDCVoltage); } return vddc; @@ -2284,7 +2284,7 @@ static bool radeon_atombios_parse_pplib_clock_info(struct radeon_device *rdev, rdev->pm.power_state[state_index].clock_info[mode_index].voltage.type = VOLTAGE_SW; rdev->pm.power_state[state_index].clock_info[mode_index].voltage.voltage = - clock_info->evergreen.usVDDC; + le16_to_cpu(clock_info->evergreen.usVDDC); } else { sclk = le16_to_cpu(clock_info->r600.usEngineClockLow); sclk |= clock_info->r600.ucEngineClockHigh << 16; @@ -2295,7 +2295,7 @@ static bool radeon_atombios_parse_pplib_clock_info(struct radeon_device *rdev, rdev->pm.power_state[state_index].clock_info[mode_index].voltage.type = VOLTAGE_SW; rdev->pm.power_state[state_index].clock_info[mode_index].voltage.voltage = - clock_info->r600.usVDDC; + le16_to_cpu(clock_info->r600.usVDDC); } if (rdev->flags & RADEON_IS_IGP) { @@ -2408,13 +2408,13 @@ static int radeon_atombios_parse_power_table_6(struct radeon_device *rdev) radeon_atombios_add_pplib_thermal_controller(rdev, &power_info->pplib.sThermalController); state_array = (struct StateArray *) (mode_info->atom_context->bios + data_offset + - power_info->pplib.usStateArrayOffset); + le16_to_cpu(power_info->pplib.usStateArrayOffset)); clock_info_array = (struct ClockInfoArray *) (mode_info->atom_context->bios + data_offset + - power_info->pplib.usClockInfoArrayOffset); + le16_to_cpu(power_info->pplib.usClockInfoArrayOffset)); non_clock_info_array = (struct NonClockInfoArray *) (mode_info->atom_context->bios + data_offset + - power_info->pplib.usNonClockInfoArrayOffset); + le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset)); rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) * state_array->ucNumEntries, GFP_KERNEL); if (!rdev->pm.power_state) @@ -2533,7 +2533,7 @@ uint32_t radeon_atom_get_engine_clock(struct radeon_device *rdev) int index = GetIndexIntoMasterTable(COMMAND, GetEngineClock); atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); - return args.ulReturnEngineClock; + return le32_to_cpu(args.ulReturnEngineClock); } uint32_t radeon_atom_get_memory_clock(struct radeon_device *rdev) @@ -2542,7 +2542,7 @@ uint32_t radeon_atom_get_memory_clock(struct radeon_device *rdev) int index = GetIndexIntoMasterTable(COMMAND, GetMemoryClock); atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); - return args.ulReturnMemoryClock; + return le32_to_cpu(args.ulReturnMemoryClock); } void radeon_atom_set_engine_clock(struct radeon_device *rdev, @@ -2551,7 +2551,7 @@ void radeon_atom_set_engine_clock(struct radeon_device *rdev, SET_ENGINE_CLOCK_PS_ALLOCATION args; int index = GetIndexIntoMasterTable(COMMAND, SetEngineClock); - args.ulTargetEngineClock = eng_clock; /* 10 khz */ + args.ulTargetEngineClock = cpu_to_le32(eng_clock); /* 10 khz */ atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); } @@ -2565,7 +2565,7 @@ void radeon_atom_set_memory_clock(struct radeon_device *rdev, if (rdev->flags & RADEON_IS_IGP) return; - args.ulTargetMemoryClock = mem_clock; /* 10 khz */ + args.ulTargetMemoryClock = cpu_to_le32(mem_clock); /* 10 khz */ atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args); } diff --git a/drivers/gpu/drm/radeon/radeon_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c index c558685..10191d9 100644 --- a/drivers/gpu/drm/radeon/radeon_benchmark.c +++ b/drivers/gpu/drm/radeon/radeon_benchmark.c @@ -41,7 +41,7 @@ void radeon_benchmark_move(struct radeon_device *rdev, unsigned bsize, size = bsize; n = 1024; - r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true, sdomain, &sobj); + r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, &sobj); if (r) { goto out_cleanup; } @@ -53,7 +53,7 @@ void radeon_benchmark_move(struct radeon_device *rdev, unsigned bsize, if (r) { goto out_cleanup; } - r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true, ddomain, &dobj); + r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, &dobj); if (r) { goto out_cleanup; } diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c index d27ef74..cf7c8d5 100644 --- a/drivers/gpu/drm/radeon/radeon_combios.c +++ b/drivers/gpu/drm/radeon/radeon_combios.c @@ -1504,6 +1504,11 @@ bool radeon_get_legacy_connector_info_from_table(struct drm_device *dev) (rdev->pdev->subsystem_device == 0x4a48)) { /* Mac X800 */ rdev->mode_info.connector_table = CT_MAC_X800; + } else if ((rdev->pdev->device == 0x4150) && + (rdev->pdev->subsystem_vendor == 0x1002) && + (rdev->pdev->subsystem_device == 0x4150)) { + /* Mac G5 9600 */ + rdev->mode_info.connector_table = CT_MAC_G5_9600; } else #endif /* CONFIG_PPC_PMAC */ #ifdef CONFIG_PPC64 @@ -2022,6 +2027,48 @@ bool radeon_get_legacy_connector_info_from_table(struct drm_device *dev) CONNECTOR_OBJECT_ID_DUAL_LINK_DVI_I, &hpd); break; + case CT_MAC_G5_9600: + DRM_INFO("Connector Table: %d (mac g5 9600)\n", + rdev->mode_info.connector_table); + /* DVI - tv dac, dvo */ + ddc_i2c = combios_setup_i2c_bus(rdev, DDC_DVI, 0, 0); + hpd.hpd = RADEON_HPD_1; /* ??? */ + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_DFP2_SUPPORT, + 0), + ATOM_DEVICE_DFP2_SUPPORT); + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_CRT2_SUPPORT, + 2), + ATOM_DEVICE_CRT2_SUPPORT); + radeon_add_legacy_connector(dev, 0, + ATOM_DEVICE_DFP2_SUPPORT | + ATOM_DEVICE_CRT2_SUPPORT, + DRM_MODE_CONNECTOR_DVII, &ddc_i2c, + CONNECTOR_OBJECT_ID_SINGLE_LINK_DVI_I, + &hpd); + /* ADC - primary dac, internal tmds */ + ddc_i2c = combios_setup_i2c_bus(rdev, DDC_VGA, 0, 0); + hpd.hpd = RADEON_HPD_2; /* ??? */ + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_DFP1_SUPPORT, + 0), + ATOM_DEVICE_DFP1_SUPPORT); + radeon_add_legacy_encoder(dev, + radeon_get_encoder_enum(dev, + ATOM_DEVICE_CRT1_SUPPORT, + 1), + ATOM_DEVICE_CRT1_SUPPORT); + radeon_add_legacy_connector(dev, 1, + ATOM_DEVICE_DFP1_SUPPORT | + ATOM_DEVICE_CRT1_SUPPORT, + DRM_MODE_CONNECTOR_DVII, &ddc_i2c, + CONNECTOR_OBJECT_ID_SINGLE_LINK_DVI_I, + &hpd); + break; default: DRM_INFO("Connector table: %d (invalid)\n", rdev->mode_info.connector_table); diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index 22b7e3d..3f3c9aa 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -972,7 +972,16 @@ 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 + else if (radeon_connector->connector_object_id == CONNECTOR_OBJECT_ID_HDMI_TYPE_A) { + if (ASIC_IS_DCE3(rdev)) { + /* HDMI 1.3+ supports max clock of 340 Mhz */ + if (mode->clock > 340000) + return MODE_CLOCK_HIGH; + else + return MODE_OK; + } else + return MODE_CLOCK_HIGH; + } else return MODE_CLOCK_HIGH; } return MODE_OK; diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c index eb6b9ee..3d599e3 100644 --- a/drivers/gpu/drm/radeon/radeon_cp.c +++ b/drivers/gpu/drm/radeon/radeon_cp.c @@ -2113,9 +2113,9 @@ int radeon_driver_load(struct drm_device *dev, unsigned long flags) break; } - if (drm_device_is_agp(dev)) + if (drm_pci_device_is_agp(dev)) dev_priv->flags |= RADEON_IS_AGP; - else if (drm_device_is_pcie(dev)) + else if (drm_pci_device_is_pcie(dev)) dev_priv->flags |= RADEON_IS_PCIE; else dev_priv->flags |= RADEON_IS_PCI; diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c index 35b5eb8..8c19169 100644 --- a/drivers/gpu/drm/radeon/radeon_cs.c +++ b/drivers/gpu/drm/radeon/radeon_cs.c @@ -75,7 +75,7 @@ int radeon_cs_parser_relocs(struct radeon_cs_parser *p) return -ENOENT; } p->relocs_ptr[i] = &p->relocs[i]; - p->relocs[i].robj = p->relocs[i].gobj->driver_private; + p->relocs[i].robj = gem_to_radeon_bo(p->relocs[i].gobj); p->relocs[i].lobj.bo = p->relocs[i].robj; p->relocs[i].lobj.wdomain = r->write_domain; p->relocs[i].lobj.rdomain = r->read_domains; diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 0d47893..f0209be 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -85,6 +85,7 @@ static const char radeon_family_name[][16] = { "BARTS", "TURKS", "CAICOS", + "CAYMAN", "LAST", }; @@ -184,7 +185,7 @@ int radeon_wb_init(struct radeon_device *rdev) int r; if (rdev->wb.wb_obj == NULL) { - r = radeon_bo_create(rdev, NULL, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true, + r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT, &rdev->wb.wb_obj); if (r) { dev_warn(rdev->dev, "(%d) create WB bo failed\n", r); @@ -860,7 +861,7 @@ int radeon_suspend_kms(struct drm_device *dev, pm_message_t state) if (rfb == NULL || rfb->obj == NULL) { continue; } - robj = rfb->obj->driver_private; + robj = gem_to_radeon_bo(rfb->obj); /* don't unpin kernel fb objects */ if (!radeon_fbdev_robj_is_fb(rdev, robj)) { r = radeon_bo_reserve(robj, false); @@ -936,8 +937,11 @@ int radeon_resume_kms(struct drm_device *dev) int radeon_gpu_reset(struct radeon_device *rdev) { int r; + int resched; radeon_save_bios_scratch_regs(rdev); + /* block TTM */ + resched = ttm_bo_lock_delayed_workqueue(&rdev->mman.bdev); radeon_suspend(rdev); r = radeon_asic_reset(rdev); @@ -946,6 +950,7 @@ int radeon_gpu_reset(struct radeon_device *rdev) radeon_resume(rdev); radeon_restore_bios_scratch_regs(rdev); drm_helper_resume_force_mode(rdev->ddev); + ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched); return 0; } /* bad news, how to tell it to userspace ? */ diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 2eff98cf..4be5879 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -371,7 +371,7 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc, new_radeon_fb = to_radeon_framebuffer(fb); /* schedule unpin of the old buffer */ obj = old_radeon_fb->obj; - rbo = obj->driver_private; + rbo = gem_to_radeon_bo(obj); work->old_rbo = rbo; INIT_WORK(&work->work, radeon_unpin_work_func); @@ -391,7 +391,7 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc, /* pin the new buffer */ obj = new_radeon_fb->obj; - rbo = obj->driver_private; + rbo = gem_to_radeon_bo(obj); DRM_DEBUG_DRIVER("flip-ioctl() cur_fbo = %p, cur_bbo = %p\n", work->old_rbo, rbo); @@ -793,6 +793,11 @@ static void avivo_get_fb_div(struct radeon_pll *pll, tmp *= target_clock; *fb_div = tmp / pll->reference_freq; *frac_fb_div = tmp % pll->reference_freq; + + if (*fb_div > pll->max_feedback_div) + *fb_div = pll->max_feedback_div; + else if (*fb_div < pll->min_feedback_div) + *fb_div = pll->min_feedback_div; } static u32 avivo_get_post_div(struct radeon_pll *pll, @@ -826,6 +831,11 @@ static u32 avivo_get_post_div(struct radeon_pll *pll, post_div--; } + if (post_div > pll->max_post_div) + post_div = pll->max_post_div; + else if (post_div < pll->min_post_div) + post_div = pll->min_post_div; + return post_div; } @@ -961,7 +971,7 @@ void radeon_compute_pll_legacy(struct radeon_pll *pll, max_fractional_feed_div = pll->max_frac_feedback_div; } - for (post_div = min_post_div; post_div <= max_post_div; ++post_div) { + for (post_div = max_post_div; post_div >= min_post_div; --post_div) { uint32_t ref_div; if ((pll->flags & RADEON_PLL_NO_ODD_POST_DIV) && (post_div & 1)) diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index 275b26a..63d2de8 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -49,9 +49,10 @@ * - 2.6.0 - add tiling config query (r6xx+), add initial HiZ support (r300->r500) * 2.7.0 - fixups for r600 2D tiling support. (no external ABI change), add eg dyn gpr regs * 2.8.0 - pageflip support, r500 US_FORMAT regs. r500 ARGB2101010 colorbuf, r300->r500 CMASK, clock crystal query + * 2.9.0 - r600 tiling (s3tc,rgtc) working, SET_PREDICATION packet 3 on r600 + eg, backend query */ #define KMS_DRIVER_MAJOR 2 -#define KMS_DRIVER_MINOR 8 +#define KMS_DRIVER_MINOR 9 #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); @@ -84,6 +85,16 @@ extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, extern struct drm_ioctl_desc radeon_ioctls_kms[]; extern int radeon_max_kms_ioctl; int radeon_mmap(struct file *filp, struct vm_area_struct *vma); +int radeon_mode_dumb_mmap(struct drm_file *filp, + struct drm_device *dev, + uint32_t handle, uint64_t *offset_p); +int radeon_mode_dumb_create(struct drm_file *file_priv, + struct drm_device *dev, + struct drm_mode_create_dumb *args); +int radeon_mode_dumb_destroy(struct drm_file *file_priv, + struct drm_device *dev, + uint32_t handle); + #if defined(CONFIG_DEBUG_FS) int radeon_debugfs_init(struct drm_minor *minor); void radeon_debugfs_cleanup(struct drm_minor *minor); @@ -228,11 +239,6 @@ static struct drm_driver driver_old = { .llseek = noop_llseek, }, - .pci_driver = { - .name = DRIVER_NAME, - .id_table = pciidlist, - }, - .name = DRIVER_NAME, .desc = DRIVER_DESC, .date = DRIVER_DATE, @@ -322,6 +328,9 @@ static struct drm_driver kms_driver = { .gem_init_object = radeon_gem_object_init, .gem_free_object = radeon_gem_object_free, .dma_ioctl = radeon_dma_ioctl_kms, + .dumb_create = radeon_mode_dumb_create, + .dumb_map_offset = radeon_mode_dumb_mmap, + .dumb_destroy = radeon_mode_dumb_destroy, .fops = { .owner = THIS_MODULE, .open = drm_open, @@ -336,15 +345,6 @@ static struct drm_driver kms_driver = { #endif }, - .pci_driver = { - .name = DRIVER_NAME, - .id_table = pciidlist, - .probe = radeon_pci_probe, - .remove = radeon_pci_remove, - .suspend = radeon_pci_suspend, - .resume = radeon_pci_resume, - }, - .name = DRIVER_NAME, .desc = DRIVER_DESC, .date = DRIVER_DATE, @@ -354,15 +354,32 @@ static struct drm_driver kms_driver = { }; static struct drm_driver *driver; +static struct pci_driver *pdriver; + +static struct pci_driver radeon_pci_driver = { + .name = DRIVER_NAME, + .id_table = pciidlist, +}; + +static struct pci_driver radeon_kms_pci_driver = { + .name = DRIVER_NAME, + .id_table = pciidlist, + .probe = radeon_pci_probe, + .remove = radeon_pci_remove, + .suspend = radeon_pci_suspend, + .resume = radeon_pci_resume, +}; static int __init radeon_init(void) { driver = &driver_old; + pdriver = &radeon_pci_driver; driver->num_ioctls = radeon_max_ioctl; #ifdef CONFIG_VGA_CONSOLE if (vgacon_text_force() && radeon_modeset == -1) { DRM_INFO("VGACON disable radeon kernel modesetting.\n"); driver = &driver_old; + pdriver = &radeon_pci_driver; driver->driver_features &= ~DRIVER_MODESET; radeon_modeset = 0; } @@ -380,18 +397,19 @@ static int __init radeon_init(void) if (radeon_modeset == 1) { DRM_INFO("radeon kernel modesetting enabled.\n"); driver = &kms_driver; + pdriver = &radeon_kms_pci_driver; driver->driver_features |= DRIVER_MODESET; driver->num_ioctls = radeon_max_kms_ioctl; radeon_register_atpx_handler(); } /* if the vga console setting is enabled still * let modprobe override it */ - return drm_init(driver); + return drm_pci_init(driver, pdriver); } static void __exit radeon_exit(void) { - drm_exit(driver); + drm_pci_exit(driver, pdriver); radeon_unregister_atpx_handler(); } diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h index 448eba8..5cba46b 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.h +++ b/drivers/gpu/drm/radeon/radeon_drv.h @@ -1524,6 +1524,7 @@ extern u32 radeon_get_scratch(drm_radeon_private_t *dev_priv, int index); #define R600_CP_RB_CNTL 0xc104 # define R600_RB_BUFSZ(x) ((x) << 0) # define R600_RB_BLKSZ(x) ((x) << 8) +# define R600_BUF_SWAP_32BIT (2 << 16) # define R600_RB_NO_UPDATE (1 << 27) # define R600_RB_RPTR_WR_ENA (1 << 31) #define R600_CP_RB_RPTR_WR 0xc108 diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index d4a5422..b427488 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -910,7 +910,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t args.v1.ucAction = action; if (action == ATOM_TRANSMITTER_ACTION_INIT) { - args.v1.usInitInfo = connector_object_id; + args.v1.usInitInfo = cpu_to_le16(connector_object_id); } else if (action == ATOM_TRANSMITTER_ACTION_SETUP_VSEMPH) { args.v1.asMode.ucLaneSel = lane_num; args.v1.asMode.ucLaneSet = lane_set; @@ -1140,7 +1140,7 @@ atombios_external_encoder_setup(struct drm_encoder *encoder, case 3: args.v3.sExtEncoder.ucAction = action; if (action == EXTERNAL_ENCODER_ACTION_V3_ENCODER_INIT) - args.v3.sExtEncoder.usConnectorId = connector_object_id; + args.v3.sExtEncoder.usConnectorId = cpu_to_le16(connector_object_id); else args.v3.sExtEncoder.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10); args.v3.sExtEncoder.ucEncoderMode = atombios_get_encoder_mode(encoder); @@ -1570,11 +1570,21 @@ atombios_apply_encoder_quirks(struct drm_encoder *encoder, } /* set scaler clears this on some chips */ - /* XXX check DCE4 */ - if (!(radeon_encoder->active_device & (ATOM_DEVICE_TV_SUPPORT))) { - if (ASIC_IS_AVIVO(rdev) && (mode->flags & DRM_MODE_FLAG_INTERLACE)) - WREG32(AVIVO_D1MODE_DATA_FORMAT + radeon_crtc->crtc_offset, - AVIVO_D1MODE_INTERLEAVE_EN); + if (ASIC_IS_AVIVO(rdev) && + (!(radeon_encoder->active_device & (ATOM_DEVICE_TV_SUPPORT)))) { + if (ASIC_IS_DCE4(rdev)) { + if (mode->flags & DRM_MODE_FLAG_INTERLACE) + WREG32(EVERGREEN_DATA_FORMAT + radeon_crtc->crtc_offset, + EVERGREEN_INTERLEAVE_EN); + else + WREG32(EVERGREEN_DATA_FORMAT + radeon_crtc->crtc_offset, 0); + } else { + if (mode->flags & DRM_MODE_FLAG_INTERLACE) + WREG32(AVIVO_D1MODE_DATA_FORMAT + radeon_crtc->crtc_offset, + AVIVO_D1MODE_INTERLEAVE_EN); + else + WREG32(AVIVO_D1MODE_DATA_FORMAT + radeon_crtc->crtc_offset, 0); + } } } diff --git a/drivers/gpu/drm/radeon/radeon_family.h b/drivers/gpu/drm/radeon/radeon_family.h index 1ca55eb..6f1d9e5 100644 --- a/drivers/gpu/drm/radeon/radeon_family.h +++ b/drivers/gpu/drm/radeon/radeon_family.h @@ -84,6 +84,7 @@ enum radeon_family { CHIP_BARTS, CHIP_TURKS, CHIP_CAICOS, + CHIP_CAYMAN, CHIP_LAST, }; diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index 66324b5..0b7b486 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -64,7 +64,7 @@ static struct fb_ops radeonfb_ops = { }; -static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled) +int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled) { int aligned = width; int align_large = (ASIC_IS_AVIVO(rdev)) || tiled; @@ -90,7 +90,7 @@ static int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bo static void radeonfb_destroy_pinned_object(struct drm_gem_object *gobj) { - struct radeon_bo *rbo = gobj->driver_private; + struct radeon_bo *rbo = gem_to_radeon_bo(gobj); int ret; ret = radeon_bo_reserve(rbo, false); @@ -113,11 +113,14 @@ static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev, u32 tiling_flags = 0; int ret; int aligned_size, size; + int height = mode_cmd->height; /* need to align pitch with crtc limits */ mode_cmd->pitch = radeon_align_pitch(rdev, mode_cmd->width, mode_cmd->bpp, fb_tiled) * ((mode_cmd->bpp + 1) / 8); - size = mode_cmd->pitch * mode_cmd->height; + if (rdev->family >= CHIP_R600) + height = ALIGN(mode_cmd->height, 8); + size = mode_cmd->pitch * height; aligned_size = ALIGN(size, PAGE_SIZE); ret = radeon_gem_object_create(rdev, aligned_size, 0, RADEON_GEM_DOMAIN_VRAM, @@ -128,7 +131,7 @@ static int radeonfb_create_pinned_object(struct radeon_fbdev *rfbdev, aligned_size); return -ENOMEM; } - rbo = gobj->driver_private; + rbo = gem_to_radeon_bo(gobj); if (fb_tiled) tiling_flags = RADEON_TILING_MACRO; @@ -202,7 +205,7 @@ static int radeonfb_create(struct radeon_fbdev *rfbdev, mode_cmd.depth = sizes->surface_depth; ret = radeonfb_create_pinned_object(rfbdev, &mode_cmd, &gobj); - rbo = gobj->driver_private; + rbo = gem_to_radeon_bo(gobj); /* okay we have an object now allocate the framebuffer */ info = framebuffer_alloc(0, device); @@ -403,14 +406,14 @@ int radeon_fbdev_total_size(struct radeon_device *rdev) struct radeon_bo *robj; int size = 0; - robj = rdev->mode_info.rfbdev->rfb.obj->driver_private; + robj = gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.obj); size += radeon_bo_size(robj); return size; } bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj) { - if (robj == rdev->mode_info.rfbdev->rfb.obj->driver_private) + if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.obj)) return true; return false; } diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 171b0b2..9e59868 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -60,8 +60,7 @@ int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence *fence) trace_radeon_fence_emit(rdev->ddev, fence->seq); fence->emited = true; - list_del(&fence->list); - list_add_tail(&fence->list, &rdev->fence_drv.emited); + list_move_tail(&fence->list, &rdev->fence_drv.emited); write_unlock_irqrestore(&rdev->fence_drv.lock, irq_flags); return 0; } @@ -121,8 +120,7 @@ static bool radeon_fence_poll_locked(struct radeon_device *rdev) i = n; do { n = i->prev; - list_del(i); - list_add_tail(i, &rdev->fence_drv.signaled); + list_move_tail(i, &rdev->fence_drv.signaled); fence = list_entry(i, struct radeon_fence, list); fence->signaled = true; i = n; diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c index 6501611..f0534ef 100644 --- a/drivers/gpu/drm/radeon/radeon_gart.c +++ b/drivers/gpu/drm/radeon/radeon_gart.c @@ -78,7 +78,7 @@ int radeon_gart_table_vram_alloc(struct radeon_device *rdev) int r; if (rdev->gart.table.vram.robj == NULL) { - r = radeon_bo_create(rdev, NULL, rdev->gart.table_size, + r = radeon_bo_create(rdev, rdev->gart.table_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, &rdev->gart.table.vram.robj); if (r) { @@ -149,8 +149,9 @@ void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset, p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); for (i = 0; i < pages; i++, p++) { if (rdev->gart.pages[p]) { - pci_unmap_page(rdev->pdev, rdev->gart.pages_addr[p], - PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); + if (!rdev->gart.ttm_alloced[p]) + pci_unmap_page(rdev->pdev, rdev->gart.pages_addr[p], + PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); rdev->gart.pages[p] = NULL; rdev->gart.pages_addr[p] = rdev->dummy_page.addr; page_base = rdev->gart.pages_addr[p]; @@ -165,7 +166,7 @@ void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset, } int radeon_gart_bind(struct radeon_device *rdev, unsigned offset, - int pages, struct page **pagelist) + int pages, struct page **pagelist, dma_addr_t *dma_addr) { unsigned t; unsigned p; @@ -180,15 +181,22 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset, p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); for (i = 0; i < pages; i++, p++) { - /* we need to support large memory configurations */ - /* assume that unbind have already been call on the range */ - rdev->gart.pages_addr[p] = pci_map_page(rdev->pdev, pagelist[i], + /* On TTM path, we only use the DMA API if TTM_PAGE_FLAG_DMA32 + * is requested. */ + if (dma_addr[i] != DMA_ERROR_CODE) { + rdev->gart.ttm_alloced[p] = true; + rdev->gart.pages_addr[p] = dma_addr[i]; + } else { + /* we need to support large memory configurations */ + /* assume that unbind have already been call on the range */ + rdev->gart.pages_addr[p] = pci_map_page(rdev->pdev, pagelist[i], 0, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); - if (pci_dma_mapping_error(rdev->pdev, rdev->gart.pages_addr[p])) { - /* FIXME: failed to map page (return -ENOMEM?) */ - radeon_gart_unbind(rdev, offset, pages); - return -ENOMEM; + if (pci_dma_mapping_error(rdev->pdev, rdev->gart.pages_addr[p])) { + /* FIXME: failed to map page (return -ENOMEM?) */ + radeon_gart_unbind(rdev, offset, pages); + return -ENOMEM; + } } rdev->gart.pages[p] = pagelist[i]; page_base = rdev->gart.pages_addr[p]; @@ -251,6 +259,12 @@ int radeon_gart_init(struct radeon_device *rdev) radeon_gart_fini(rdev); return -ENOMEM; } + rdev->gart.ttm_alloced = kzalloc(sizeof(bool) * + rdev->gart.num_cpu_pages, GFP_KERNEL); + if (rdev->gart.ttm_alloced == NULL) { + radeon_gart_fini(rdev); + return -ENOMEM; + } /* set GART entry to point to the dummy page by default */ for (i = 0; i < rdev->gart.num_cpu_pages; i++) { rdev->gart.pages_addr[i] = rdev->dummy_page.addr; @@ -267,6 +281,8 @@ void radeon_gart_fini(struct radeon_device *rdev) rdev->gart.ready = false; kfree(rdev->gart.pages); kfree(rdev->gart.pages_addr); + kfree(rdev->gart.ttm_alloced); rdev->gart.pages = NULL; rdev->gart.pages_addr = NULL; + rdev->gart.ttm_alloced = NULL; } diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index df95eb8..aa1ca2d 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -32,21 +32,18 @@ int radeon_gem_object_init(struct drm_gem_object *obj) { - /* we do nothings here */ + BUG(); + return 0; } void radeon_gem_object_free(struct drm_gem_object *gobj) { - struct radeon_bo *robj = gobj->driver_private; + struct radeon_bo *robj = gem_to_radeon_bo(gobj); - gobj->driver_private = NULL; if (robj) { radeon_bo_unref(&robj); } - - drm_gem_object_release(gobj); - kfree(gobj); } int radeon_gem_object_create(struct radeon_device *rdev, int size, @@ -54,36 +51,34 @@ int radeon_gem_object_create(struct radeon_device *rdev, int size, bool discardable, bool kernel, struct drm_gem_object **obj) { - struct drm_gem_object *gobj; struct radeon_bo *robj; int r; *obj = NULL; - gobj = drm_gem_object_alloc(rdev->ddev, size); - if (!gobj) { - return -ENOMEM; - } /* At least align on page size */ if (alignment < PAGE_SIZE) { alignment = PAGE_SIZE; } - r = radeon_bo_create(rdev, gobj, size, alignment, kernel, initial_domain, &robj); + r = radeon_bo_create(rdev, size, alignment, kernel, initial_domain, &robj); if (r) { if (r != -ERESTARTSYS) DRM_ERROR("Failed to allocate GEM object (%d, %d, %u, %d)\n", size, initial_domain, alignment, r); - drm_gem_object_unreference_unlocked(gobj); return r; } - gobj->driver_private = robj; - *obj = gobj; + *obj = &robj->gem_base; + + mutex_lock(&rdev->gem.mutex); + list_add_tail(&robj->list, &rdev->gem.objects); + mutex_unlock(&rdev->gem.mutex); + return 0; } int radeon_gem_object_pin(struct drm_gem_object *obj, uint32_t pin_domain, uint64_t *gpu_addr) { - struct radeon_bo *robj = obj->driver_private; + struct radeon_bo *robj = gem_to_radeon_bo(obj); int r; r = radeon_bo_reserve(robj, false); @@ -96,7 +91,7 @@ int radeon_gem_object_pin(struct drm_gem_object *obj, uint32_t pin_domain, void radeon_gem_object_unpin(struct drm_gem_object *obj) { - struct radeon_bo *robj = obj->driver_private; + struct radeon_bo *robj = gem_to_radeon_bo(obj); int r; r = radeon_bo_reserve(robj, false); @@ -114,7 +109,7 @@ int radeon_gem_set_domain(struct drm_gem_object *gobj, int r; /* FIXME: reeimplement */ - robj = gobj->driver_private; + robj = gem_to_radeon_bo(gobj); /* work out where to validate the buffer to */ domain = wdomain; if (!domain) { @@ -156,9 +151,12 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data, { struct radeon_device *rdev = dev->dev_private; struct drm_radeon_gem_info *args = data; + struct ttm_mem_type_manager *man; + + man = &rdev->mman.bdev.man[TTM_PL_VRAM]; args->vram_size = rdev->mc.real_vram_size; - args->vram_visible = rdev->mc.real_vram_size; + args->vram_visible = (u64)man->size << PAGE_SHIFT; if (rdev->stollen_vga_memory) args->vram_visible -= radeon_bo_size(rdev->stollen_vga_memory); args->vram_visible -= radeon_fbdev_total_size(rdev); @@ -228,7 +226,7 @@ int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data, if (gobj == NULL) { return -ENOENT; } - robj = gobj->driver_private; + robj = gem_to_radeon_bo(gobj); r = radeon_gem_set_domain(gobj, args->read_domains, args->write_domain); @@ -236,23 +234,31 @@ int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data, return r; } -int radeon_gem_mmap_ioctl(struct drm_device *dev, void *data, - struct drm_file *filp) +int radeon_mode_dumb_mmap(struct drm_file *filp, + struct drm_device *dev, + uint32_t handle, uint64_t *offset_p) { - struct drm_radeon_gem_mmap *args = data; struct drm_gem_object *gobj; struct radeon_bo *robj; - gobj = drm_gem_object_lookup(dev, filp, args->handle); + gobj = drm_gem_object_lookup(dev, filp, handle); if (gobj == NULL) { return -ENOENT; } - robj = gobj->driver_private; - args->addr_ptr = radeon_bo_mmap_offset(robj); + robj = gem_to_radeon_bo(gobj); + *offset_p = radeon_bo_mmap_offset(robj); drm_gem_object_unreference_unlocked(gobj); return 0; } +int radeon_gem_mmap_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp) +{ + struct drm_radeon_gem_mmap *args = data; + + return radeon_mode_dumb_mmap(filp, dev, args->handle, &args->addr_ptr); +} + int radeon_gem_busy_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) { @@ -266,7 +272,7 @@ int radeon_gem_busy_ioctl(struct drm_device *dev, void *data, if (gobj == NULL) { return -ENOENT; } - robj = gobj->driver_private; + robj = gem_to_radeon_bo(gobj); r = radeon_bo_wait(robj, &cur_placement, true); switch (cur_placement) { case TTM_PL_VRAM: @@ -296,7 +302,7 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data, if (gobj == NULL) { return -ENOENT; } - robj = gobj->driver_private; + robj = gem_to_radeon_bo(gobj); r = radeon_bo_wait(robj, NULL, false); /* callback hw specific functions if any */ if (robj->rdev->asic->ioctl_wait_idle) @@ -317,7 +323,7 @@ int radeon_gem_set_tiling_ioctl(struct drm_device *dev, void *data, gobj = drm_gem_object_lookup(dev, filp, args->handle); if (gobj == NULL) return -ENOENT; - robj = gobj->driver_private; + robj = gem_to_radeon_bo(gobj); r = radeon_bo_set_tiling_flags(robj, args->tiling_flags, args->pitch); drm_gem_object_unreference_unlocked(gobj); return r; @@ -335,7 +341,7 @@ int radeon_gem_get_tiling_ioctl(struct drm_device *dev, void *data, gobj = drm_gem_object_lookup(dev, filp, args->handle); if (gobj == NULL) return -ENOENT; - rbo = gobj->driver_private; + rbo = gem_to_radeon_bo(gobj); r = radeon_bo_reserve(rbo, false); if (unlikely(r != 0)) goto out; @@ -345,3 +351,40 @@ out: drm_gem_object_unreference_unlocked(gobj); return r; } + +int radeon_mode_dumb_create(struct drm_file *file_priv, + struct drm_device *dev, + struct drm_mode_create_dumb *args) +{ + struct radeon_device *rdev = dev->dev_private; + struct drm_gem_object *gobj; + uint32_t handle; + int r; + + args->pitch = radeon_align_pitch(rdev, args->width, args->bpp, 0) * ((args->bpp + 1) / 8); + args->size = args->pitch * args->height; + args->size = ALIGN(args->size, PAGE_SIZE); + + r = radeon_gem_object_create(rdev, args->size, 0, + RADEON_GEM_DOMAIN_VRAM, + false, ttm_bo_type_device, + &gobj); + if (r) + return -ENOMEM; + + r = drm_gem_handle_create(file_priv, gobj, &handle); + /* drop reference from allocate - handle holds it now */ + drm_gem_object_unreference_unlocked(gobj); + if (r) { + return r; + } + args->handle = handle; + return 0; +} + +int radeon_mode_dumb_destroy(struct drm_file *file_priv, + struct drm_device *dev, + uint32_t handle) +{ + return drm_gem_handle_delete(file_priv, handle); +} diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index 8387d32..bf7d4c0 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -58,9 +58,9 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags) dev->dev_private = (void *)rdev; /* update BUS flag */ - if (drm_device_is_agp(dev)) { + if (drm_pci_device_is_agp(dev)) { flags |= RADEON_IS_AGP; - } else if (drm_device_is_pcie(dev)) { + } else if (drm_pci_device_is_pcie(dev)) { flags |= RADEON_IS_PCIE; } else { flags |= RADEON_IS_PCI; @@ -169,7 +169,9 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) value = rdev->accel_working; break; case RADEON_INFO_TILING_CONFIG: - if (rdev->family >= CHIP_CEDAR) + if (rdev->family >= CHIP_CAYMAN) + value = rdev->config.cayman.tile_config; + else if (rdev->family >= CHIP_CEDAR) value = rdev->config.evergreen.tile_config; else if (rdev->family >= CHIP_RV770) value = rdev->config.rv770.tile_config; @@ -205,6 +207,20 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) /* return clock value in KHz */ value = rdev->clock.spll.reference_freq * 10; break; + case RADEON_INFO_NUM_BACKENDS: + if (rdev->family >= CHIP_CAYMAN) + value = rdev->config.cayman.max_backends_per_se * + rdev->config.cayman.max_shader_engines; + else if (rdev->family >= CHIP_CEDAR) + value = rdev->config.evergreen.max_backends; + else if (rdev->family >= CHIP_RV770) + value = rdev->config.rv770.max_backends; + else if (rdev->family >= CHIP_R600) + value = rdev->config.r600.max_backends; + else { + return -EINVAL; + } + break; default: DRM_DEBUG_KMS("Invalid request %d\n", info->request); return -EINVAL; diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c index cf0638c..66c9af1 100644 --- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c +++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c @@ -415,7 +415,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc, /* Pin framebuffer & get tilling informations */ obj = radeon_fb->obj; - rbo = obj->driver_private; + rbo = gem_to_radeon_bo(obj); r = radeon_bo_reserve(rbo, false); if (unlikely(r != 0)) return r; @@ -443,7 +443,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc, (target_fb->bits_per_pixel * 8)); crtc_pitch |= crtc_pitch << 16; - + crtc_offset_cntl |= RADEON_CRTC_GUI_TRIG_OFFSET_LEFT_EN; if (tiling_flags & RADEON_TILING_MACRO) { if (ASIC_IS_R300(rdev)) crtc_offset_cntl |= (R300_CRTC_X_Y_MODE_EN | @@ -502,6 +502,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc, gen_cntl_val = RREG32(gen_cntl_reg); gen_cntl_val &= ~(0xf << 8); gen_cntl_val |= (format << 8); + gen_cntl_val &= ~RADEON_CRTC_VSTAT_MODE_MASK; WREG32(gen_cntl_reg, gen_cntl_val); crtc_offset = (u32)base; @@ -520,7 +521,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc, if (!atomic && fb && fb != crtc->fb) { radeon_fb = to_radeon_framebuffer(fb); - rbo = radeon_fb->obj->driver_private; + rbo = gem_to_radeon_bo(radeon_fb->obj); r = radeon_bo_reserve(rbo, false); if (unlikely(r != 0)) return r; diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index 6794cdf..5067d18 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -209,6 +209,7 @@ enum radeon_connector_table { CT_EMAC, CT_RN50_POWER, CT_MAC_X800, + CT_MAC_G5_9600, }; enum radeon_dvo_chip { @@ -675,4 +676,5 @@ void radeon_fb_output_poll_changed(struct radeon_device *rdev); void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id); +int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled); #endif diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 7d6b8e8..976c3b1 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -55,6 +55,7 @@ static void radeon_ttm_bo_destroy(struct ttm_buffer_object *tbo) list_del_init(&bo->list); mutex_unlock(&bo->rdev->gem.mutex); radeon_bo_clear_surface_reg(bo); + drm_gem_object_release(&bo->gem_base); kfree(bo); } @@ -86,7 +87,7 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain) rbo->placement.num_busy_placement = c; } -int radeon_bo_create(struct radeon_device *rdev, struct drm_gem_object *gobj, +int radeon_bo_create(struct radeon_device *rdev, unsigned long size, int byte_align, bool kernel, u32 domain, struct radeon_bo **bo_ptr) { @@ -96,6 +97,8 @@ int radeon_bo_create(struct radeon_device *rdev, struct drm_gem_object *gobj, unsigned long max_size = 0; int r; + size = ALIGN(size, PAGE_SIZE); + if (unlikely(rdev->mman.bdev.dev_mapping == NULL)) { rdev->mman.bdev.dev_mapping = rdev->ddev->dev_mapping; } @@ -118,8 +121,13 @@ retry: bo = kzalloc(sizeof(struct radeon_bo), GFP_KERNEL); if (bo == NULL) return -ENOMEM; + r = drm_gem_object_init(rdev->ddev, &bo->gem_base, size); + if (unlikely(r)) { + kfree(bo); + return r; + } bo->rdev = rdev; - bo->gobj = gobj; + bo->gem_base.driver_private = NULL; bo->surface_reg = -1; INIT_LIST_HEAD(&bo->list); radeon_ttm_placement_from_domain(bo, domain); @@ -142,12 +150,9 @@ retry: return r; } *bo_ptr = bo; - if (gobj) { - mutex_lock(&bo->rdev->gem.mutex); - list_add_tail(&bo->list, &rdev->gem.objects); - mutex_unlock(&bo->rdev->gem.mutex); - } + trace_radeon_bo_create(bo); + return 0; } @@ -260,7 +265,6 @@ int radeon_bo_evict_vram(struct radeon_device *rdev) void radeon_bo_force_delete(struct radeon_device *rdev) { struct radeon_bo *bo, *n; - struct drm_gem_object *gobj; if (list_empty(&rdev->gem.objects)) { return; @@ -268,16 +272,14 @@ void radeon_bo_force_delete(struct radeon_device *rdev) dev_err(rdev->dev, "Userspace still has active objects !\n"); list_for_each_entry_safe(bo, n, &rdev->gem.objects, list) { mutex_lock(&rdev->ddev->struct_mutex); - gobj = bo->gobj; dev_err(rdev->dev, "%p %p %lu %lu force free\n", - gobj, bo, (unsigned long)gobj->size, - *((unsigned long *)&gobj->refcount)); + &bo->gem_base, bo, (unsigned long)bo->gem_base.size, + *((unsigned long *)&bo->gem_base.refcount)); mutex_lock(&bo->rdev->gem.mutex); list_del_init(&bo->list); mutex_unlock(&bo->rdev->gem.mutex); - radeon_bo_unref(&bo); - gobj->driver_private = NULL; - drm_gem_object_unreference(gobj); + /* this should unref the ttm bo */ + drm_gem_object_unreference(&bo->gem_base); mutex_unlock(&rdev->ddev->struct_mutex); } } diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h index 22d4c23..7f8e778 100644 --- a/drivers/gpu/drm/radeon/radeon_object.h +++ b/drivers/gpu/drm/radeon/radeon_object.h @@ -137,10 +137,9 @@ static inline int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type, } extern int radeon_bo_create(struct radeon_device *rdev, - struct drm_gem_object *gobj, unsigned long size, - int byte_align, - bool kernel, u32 domain, - struct radeon_bo **bo_ptr); + unsigned long size, int byte_align, + bool kernel, u32 domain, + struct radeon_bo **bo_ptr); extern int radeon_bo_kmap(struct radeon_bo *bo, void **ptr); extern void radeon_bo_kunmap(struct radeon_bo *bo); extern void radeon_bo_unref(struct radeon_bo **bo); diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index 239c3d7..bbc9cd8 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -175,7 +175,7 @@ int radeon_ib_pool_init(struct radeon_device *rdev) return 0; INIT_LIST_HEAD(&rdev->ib_pool.bogus_ib); /* Allocate 1M object buffer */ - r = radeon_bo_create(rdev, NULL, RADEON_IB_POOL_SIZE*64*1024, + r = radeon_bo_create(rdev, RADEON_IB_POOL_SIZE*64*1024, PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT, &rdev->ib_pool.robj); if (r) { @@ -332,7 +332,7 @@ int radeon_ring_init(struct radeon_device *rdev, unsigned ring_size) rdev->cp.ring_size = ring_size; /* Allocate ring buffer */ if (rdev->cp.ring_obj == NULL) { - r = radeon_bo_create(rdev, NULL, rdev->cp.ring_size, PAGE_SIZE, true, + r = radeon_bo_create(rdev, rdev->cp.ring_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT, &rdev->cp.ring_obj); if (r) { diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c index 5b44f65..dee4a0c 100644 --- a/drivers/gpu/drm/radeon/radeon_test.c +++ b/drivers/gpu/drm/radeon/radeon_test.c @@ -52,7 +52,7 @@ void radeon_test_moves(struct radeon_device *rdev) goto out_cleanup; } - r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, + r = radeon_bo_create(rdev, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, &vram_obj); if (r) { DRM_ERROR("Failed to create VRAM object\n"); @@ -71,7 +71,7 @@ void radeon_test_moves(struct radeon_device *rdev) void **gtt_start, **gtt_end; void **vram_start, **vram_end; - r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true, + r = radeon_bo_create(rdev, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT, gtt_obj + i); if (r) { DRM_ERROR("Failed to create GTT object %d\n", i); diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 1272e4b..60125dd 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -529,7 +529,7 @@ int radeon_ttm_init(struct radeon_device *rdev) DRM_ERROR("Failed initializing VRAM heap.\n"); return r; } - r = radeon_bo_create(rdev, NULL, 256 * 1024, PAGE_SIZE, true, + r = radeon_bo_create(rdev, 256 * 1024, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, &rdev->stollen_vga_memory); if (r) { @@ -589,6 +589,20 @@ void radeon_ttm_fini(struct radeon_device *rdev) DRM_INFO("radeon: ttm finalized\n"); } +/* this should only be called at bootup or when userspace + * isn't running */ +void radeon_ttm_set_active_vram_size(struct radeon_device *rdev, u64 size) +{ + struct ttm_mem_type_manager *man; + + if (!rdev->mman.initialized) + return; + + man = &rdev->mman.bdev.man[TTM_PL_VRAM]; + /* this just adjusts TTM size idea, which sets lpfn to the correct value */ + man->size = size >> PAGE_SHIFT; +} + static struct vm_operations_struct radeon_ttm_vm_ops; static const struct vm_operations_struct *ttm_vm_ops = NULL; @@ -647,6 +661,7 @@ struct radeon_ttm_backend { unsigned long num_pages; struct page **pages; struct page *dummy_read_page; + dma_addr_t *dma_addrs; bool populated; bool bound; unsigned offset; @@ -655,12 +670,14 @@ struct radeon_ttm_backend { static int radeon_ttm_backend_populate(struct ttm_backend *backend, unsigned long num_pages, struct page **pages, - struct page *dummy_read_page) + struct page *dummy_read_page, + dma_addr_t *dma_addrs) { struct radeon_ttm_backend *gtt; gtt = container_of(backend, struct radeon_ttm_backend, backend); gtt->pages = pages; + gtt->dma_addrs = dma_addrs; gtt->num_pages = num_pages; gtt->dummy_read_page = dummy_read_page; gtt->populated = true; @@ -673,6 +690,7 @@ static void radeon_ttm_backend_clear(struct ttm_backend *backend) gtt = container_of(backend, struct radeon_ttm_backend, backend); gtt->pages = NULL; + gtt->dma_addrs = NULL; gtt->num_pages = 0; gtt->dummy_read_page = NULL; gtt->populated = false; @@ -693,7 +711,7 @@ static int radeon_ttm_backend_bind(struct ttm_backend *backend, gtt->num_pages, bo_mem, backend); } r = radeon_gart_bind(gtt->rdev, gtt->offset, - gtt->num_pages, gtt->pages); + gtt->num_pages, gtt->pages, gtt->dma_addrs); if (r) { DRM_ERROR("failed to bind %lu pages at 0x%08X\n", gtt->num_pages, gtt->offset); @@ -787,9 +805,9 @@ static int radeon_ttm_debugfs_init(struct radeon_device *rdev) radeon_mem_types_list[i].show = &radeon_mm_dump_table; radeon_mem_types_list[i].driver_features = 0; if (i == 0) - radeon_mem_types_list[i].data = &rdev->mman.bdev.man[TTM_PL_VRAM].priv; + radeon_mem_types_list[i].data = rdev->mman.bdev.man[TTM_PL_VRAM].priv; else - radeon_mem_types_list[i].data = &rdev->mman.bdev.man[TTM_PL_TT].priv; + radeon_mem_types_list[i].data = rdev->mman.bdev.man[TTM_PL_TT].priv; } /* Add ttm page pool to debugfs */ diff --git a/drivers/gpu/drm/radeon/reg_srcs/cayman b/drivers/gpu/drm/radeon/reg_srcs/cayman new file mode 100644 index 0000000..6334f8a --- /dev/null +++ b/drivers/gpu/drm/radeon/reg_srcs/cayman @@ -0,0 +1,619 @@ +cayman 0x9400 +0x0000802C GRBM_GFX_INDEX +0x000088B0 VGT_VTX_VECT_EJECT_REG +0x000088C4 VGT_CACHE_INVALIDATION +0x000088D4 VGT_GS_VERTEX_REUSE +0x00008958 VGT_PRIMITIVE_TYPE +0x0000895C VGT_INDEX_TYPE +0x00008970 VGT_NUM_INDICES +0x00008974 VGT_NUM_INSTANCES +0x00008990 VGT_COMPUTE_DIM_X +0x00008994 VGT_COMPUTE_DIM_Y +0x00008998 VGT_COMPUTE_DIM_Z +0x0000899C VGT_COMPUTE_START_X +0x000089A0 VGT_COMPUTE_START_Y +0x000089A4 VGT_COMPUTE_START_Z +0x000089A8 VGT_COMPUTE_INDEX +0x000089AC VGT_COMPUTE_THREAD_GOURP_SIZE +0x000089B0 VGT_HS_OFFCHIP_PARAM +0x00008A14 PA_CL_ENHANCE +0x00008A60 PA_SC_LINE_STIPPLE_VALUE +0x00008B10 PA_SC_LINE_STIPPLE_STATE +0x00008BF0 PA_SC_ENHANCE +0x00008D8C SQ_DYN_GPR_CNTL_PS_FLUSH_REQ +0x00008D94 SQ_DYN_GPR_SIMD_LOCK_EN +0x00008C00 SQ_CONFIG +0x00008C04 SQ_GPR_RESOURCE_MGMT_1 +0x00008C10 SQ_GLOBAL_GPR_RESOURCE_MGMT_1 +0x00008C14 SQ_GLOBAL_GPR_RESOURCE_MGMT_2 +0x00008DF8 SQ_CONST_MEM_BASE +0x00008E20 SQ_STATIC_THREAD_MGMT_1 +0x00008E24 SQ_STATIC_THREAD_MGMT_2 +0x00008E28 SQ_STATIC_THREAD_MGMT_3 +0x00008E48 SQ_EX_ALLOC_TABLE_SLOTS +0x00009100 SPI_CONFIG_CNTL +0x0000913C SPI_CONFIG_CNTL_1 +0x00009830 DB_DEBUG +0x00009834 DB_DEBUG2 +0x00009838 DB_DEBUG3 +0x0000983C DB_DEBUG4 +0x00009854 DB_WATERMARKS +0x0000A400 TD_PS_BORDER_COLOR_INDEX +0x0000A404 TD_PS_BORDER_COLOR_RED +0x0000A408 TD_PS_BORDER_COLOR_GREEN +0x0000A40C TD_PS_BORDER_COLOR_BLUE +0x0000A410 TD_PS_BORDER_COLOR_ALPHA +0x0000A414 TD_VS_BORDER_COLOR_INDEX +0x0000A418 TD_VS_BORDER_COLOR_RED +0x0000A41C TD_VS_BORDER_COLOR_GREEN +0x0000A420 TD_VS_BORDER_COLOR_BLUE +0x0000A424 TD_VS_BORDER_COLOR_ALPHA +0x0000A428 TD_GS_BORDER_COLOR_INDEX +0x0000A42C TD_GS_BORDER_COLOR_RED +0x0000A430 TD_GS_BORDER_COLOR_GREEN +0x0000A434 TD_GS_BORDER_COLOR_BLUE +0x0000A438 TD_GS_BORDER_COLOR_ALPHA +0x0000A43C TD_HS_BORDER_COLOR_INDEX +0x0000A440 TD_HS_BORDER_COLOR_RED +0x0000A444 TD_HS_BORDER_COLOR_GREEN +0x0000A448 TD_HS_BORDER_COLOR_BLUE +0x0000A44C TD_HS_BORDER_COLOR_ALPHA +0x0000A450 TD_LS_BORDER_COLOR_INDEX +0x0000A454 TD_LS_BORDER_COLOR_RED +0x0000A458 TD_LS_BORDER_COLOR_GREEN +0x0000A45C TD_LS_BORDER_COLOR_BLUE +0x0000A460 TD_LS_BORDER_COLOR_ALPHA +0x0000A464 TD_CS_BORDER_COLOR_INDEX +0x0000A468 TD_CS_BORDER_COLOR_RED +0x0000A46C TD_CS_BORDER_COLOR_GREEN +0x0000A470 TD_CS_BORDER_COLOR_BLUE +0x0000A474 TD_CS_BORDER_COLOR_ALPHA +0x00028000 DB_RENDER_CONTROL +0x00028004 DB_COUNT_CONTROL +0x0002800C DB_RENDER_OVERRIDE +0x00028010 DB_RENDER_OVERRIDE2 +0x00028028 DB_STENCIL_CLEAR +0x0002802C DB_DEPTH_CLEAR +0x00028030 PA_SC_SCREEN_SCISSOR_TL +0x00028034 PA_SC_SCREEN_SCISSOR_BR +0x0002805C DB_DEPTH_SLICE +0x00028140 SQ_ALU_CONST_BUFFER_SIZE_PS_0 +0x00028144 SQ_ALU_CONST_BUFFER_SIZE_PS_1 +0x00028148 SQ_ALU_CONST_BUFFER_SIZE_PS_2 +0x0002814C SQ_ALU_CONST_BUFFER_SIZE_PS_3 +0x00028150 SQ_ALU_CONST_BUFFER_SIZE_PS_4 +0x00028154 SQ_ALU_CONST_BUFFER_SIZE_PS_5 +0x00028158 SQ_ALU_CONST_BUFFER_SIZE_PS_6 +0x0002815C SQ_ALU_CONST_BUFFER_SIZE_PS_7 +0x00028160 SQ_ALU_CONST_BUFFER_SIZE_PS_8 +0x00028164 SQ_ALU_CONST_BUFFER_SIZE_PS_9 +0x00028168 SQ_ALU_CONST_BUFFER_SIZE_PS_10 +0x0002816C SQ_ALU_CONST_BUFFER_SIZE_PS_11 +0x00028170 SQ_ALU_CONST_BUFFER_SIZE_PS_12 +0x00028174 SQ_ALU_CONST_BUFFER_SIZE_PS_13 +0x00028178 SQ_ALU_CONST_BUFFER_SIZE_PS_14 +0x0002817C SQ_ALU_CONST_BUFFER_SIZE_PS_15 +0x00028180 SQ_ALU_CONST_BUFFER_SIZE_VS_0 +0x00028184 SQ_ALU_CONST_BUFFER_SIZE_VS_1 +0x00028188 SQ_ALU_CONST_BUFFER_SIZE_VS_2 +0x0002818C SQ_ALU_CONST_BUFFER_SIZE_VS_3 +0x00028190 SQ_ALU_CONST_BUFFER_SIZE_VS_4 +0x00028194 SQ_ALU_CONST_BUFFER_SIZE_VS_5 +0x00028198 SQ_ALU_CONST_BUFFER_SIZE_VS_6 +0x0002819C SQ_ALU_CONST_BUFFER_SIZE_VS_7 +0x000281A0 SQ_ALU_CONST_BUFFER_SIZE_VS_8 +0x000281A4 SQ_ALU_CONST_BUFFER_SIZE_VS_9 +0x000281A8 SQ_ALU_CONST_BUFFER_SIZE_VS_10 +0x000281AC SQ_ALU_CONST_BUFFER_SIZE_VS_11 +0x000281B0 SQ_ALU_CONST_BUFFER_SIZE_VS_12 +0x000281B4 SQ_ALU_CONST_BUFFER_SIZE_VS_13 +0x000281B8 SQ_ALU_CONST_BUFFER_SIZE_VS_14 +0x000281BC SQ_ALU_CONST_BUFFER_SIZE_VS_15 +0x000281C0 SQ_ALU_CONST_BUFFER_SIZE_GS_0 +0x000281C4 SQ_ALU_CONST_BUFFER_SIZE_GS_1 +0x000281C8 SQ_ALU_CONST_BUFFER_SIZE_GS_2 +0x000281CC SQ_ALU_CONST_BUFFER_SIZE_GS_3 +0x000281D0 SQ_ALU_CONST_BUFFER_SIZE_GS_4 +0x000281D4 SQ_ALU_CONST_BUFFER_SIZE_GS_5 +0x000281D8 SQ_ALU_CONST_BUFFER_SIZE_GS_6 +0x000281DC SQ_ALU_CONST_BUFFER_SIZE_GS_7 +0x000281E0 SQ_ALU_CONST_BUFFER_SIZE_GS_8 +0x000281E4 SQ_ALU_CONST_BUFFER_SIZE_GS_9 +0x000281E8 SQ_ALU_CONST_BUFFER_SIZE_GS_10 +0x000281EC SQ_ALU_CONST_BUFFER_SIZE_GS_11 +0x000281F0 SQ_ALU_CONST_BUFFER_SIZE_GS_12 +0x000281F4 SQ_ALU_CONST_BUFFER_SIZE_GS_13 +0x000281F8 SQ_ALU_CONST_BUFFER_SIZE_GS_14 +0x000281FC SQ_ALU_CONST_BUFFER_SIZE_GS_15 +0x00028200 PA_SC_WINDOW_OFFSET +0x00028204 PA_SC_WINDOW_SCISSOR_TL +0x00028208 PA_SC_WINDOW_SCISSOR_BR +0x0002820C PA_SC_CLIPRECT_RULE +0x00028210 PA_SC_CLIPRECT_0_TL +0x00028214 PA_SC_CLIPRECT_0_BR +0x00028218 PA_SC_CLIPRECT_1_TL +0x0002821C PA_SC_CLIPRECT_1_BR +0x00028220 PA_SC_CLIPRECT_2_TL +0x00028224 PA_SC_CLIPRECT_2_BR +0x00028228 PA_SC_CLIPRECT_3_TL +0x0002822C PA_SC_CLIPRECT_3_BR +0x00028230 PA_SC_EDGERULE +0x00028234 PA_SU_HARDWARE_SCREEN_OFFSET +0x00028240 PA_SC_GENERIC_SCISSOR_TL +0x00028244 PA_SC_GENERIC_SCISSOR_BR +0x00028250 PA_SC_VPORT_SCISSOR_0_TL +0x00028254 PA_SC_VPORT_SCISSOR_0_BR +0x00028258 PA_SC_VPORT_SCISSOR_1_TL +0x0002825C PA_SC_VPORT_SCISSOR_1_BR +0x00028260 PA_SC_VPORT_SCISSOR_2_TL +0x00028264 PA_SC_VPORT_SCISSOR_2_BR +0x00028268 PA_SC_VPORT_SCISSOR_3_TL +0x0002826C PA_SC_VPORT_SCISSOR_3_BR +0x00028270 PA_SC_VPORT_SCISSOR_4_TL +0x00028274 PA_SC_VPORT_SCISSOR_4_BR +0x00028278 PA_SC_VPORT_SCISSOR_5_TL +0x0002827C PA_SC_VPORT_SCISSOR_5_BR +0x00028280 PA_SC_VPORT_SCISSOR_6_TL +0x00028284 PA_SC_VPORT_SCISSOR_6_BR +0x00028288 PA_SC_VPORT_SCISSOR_7_TL +0x0002828C PA_SC_VPORT_SCISSOR_7_BR +0x00028290 PA_SC_VPORT_SCISSOR_8_TL +0x00028294 PA_SC_VPORT_SCISSOR_8_BR +0x00028298 PA_SC_VPORT_SCISSOR_9_TL +0x0002829C PA_SC_VPORT_SCISSOR_9_BR +0x000282A0 PA_SC_VPORT_SCISSOR_10_TL +0x000282A4 PA_SC_VPORT_SCISSOR_10_BR +0x000282A8 PA_SC_VPORT_SCISSOR_11_TL +0x000282AC PA_SC_VPORT_SCISSOR_11_BR +0x000282B0 PA_SC_VPORT_SCISSOR_12_TL +0x000282B4 PA_SC_VPORT_SCISSOR_12_BR +0x000282B8 PA_SC_VPORT_SCISSOR_13_TL +0x000282BC PA_SC_VPORT_SCISSOR_13_BR +0x000282C0 PA_SC_VPORT_SCISSOR_14_TL +0x000282C4 PA_SC_VPORT_SCISSOR_14_BR +0x000282C8 PA_SC_VPORT_SCISSOR_15_TL +0x000282CC PA_SC_VPORT_SCISSOR_15_BR +0x000282D0 PA_SC_VPORT_ZMIN_0 +0x000282D4 PA_SC_VPORT_ZMAX_0 +0x000282D8 PA_SC_VPORT_ZMIN_1 +0x000282DC PA_SC_VPORT_ZMAX_1 +0x000282E0 PA_SC_VPORT_ZMIN_2 +0x000282E4 PA_SC_VPORT_ZMAX_2 +0x000282E8 PA_SC_VPORT_ZMIN_3 +0x000282EC PA_SC_VPORT_ZMAX_3 +0x000282F0 PA_SC_VPORT_ZMIN_4 +0x000282F4 PA_SC_VPORT_ZMAX_4 +0x000282F8 PA_SC_VPORT_ZMIN_5 +0x000282FC PA_SC_VPORT_ZMAX_5 +0x00028300 PA_SC_VPORT_ZMIN_6 +0x00028304 PA_SC_VPORT_ZMAX_6 +0x00028308 PA_SC_VPORT_ZMIN_7 +0x0002830C PA_SC_VPORT_ZMAX_7 +0x00028310 PA_SC_VPORT_ZMIN_8 +0x00028314 PA_SC_VPORT_ZMAX_8 +0x00028318 PA_SC_VPORT_ZMIN_9 +0x0002831C PA_SC_VPORT_ZMAX_9 +0x00028320 PA_SC_VPORT_ZMIN_10 +0x00028324 PA_SC_VPORT_ZMAX_10 +0x00028328 PA_SC_VPORT_ZMIN_11 +0x0002832C PA_SC_VPORT_ZMAX_11 +0x00028330 PA_SC_VPORT_ZMIN_12 +0x00028334 PA_SC_VPORT_ZMAX_12 +0x00028338 PA_SC_VPORT_ZMIN_13 +0x0002833C PA_SC_VPORT_ZMAX_13 +0x00028340 PA_SC_VPORT_ZMIN_14 +0x00028344 PA_SC_VPORT_ZMAX_14 +0x00028348 PA_SC_VPORT_ZMIN_15 +0x0002834C PA_SC_VPORT_ZMAX_15 +0x00028350 SX_MISC +0x00028354 SX_SURFACE_SYNC +0x00028380 SQ_VTX_SEMANTIC_0 +0x00028384 SQ_VTX_SEMANTIC_1 +0x00028388 SQ_VTX_SEMANTIC_2 +0x0002838C SQ_VTX_SEMANTIC_3 +0x00028390 SQ_VTX_SEMANTIC_4 +0x00028394 SQ_VTX_SEMANTIC_5 +0x00028398 SQ_VTX_SEMANTIC_6 +0x0002839C SQ_VTX_SEMANTIC_7 +0x000283A0 SQ_VTX_SEMANTIC_8 +0x000283A4 SQ_VTX_SEMANTIC_9 +0x000283A8 SQ_VTX_SEMANTIC_10 +0x000283AC SQ_VTX_SEMANTIC_11 +0x000283B0 SQ_VTX_SEMANTIC_12 +0x000283B4 SQ_VTX_SEMANTIC_13 +0x000283B8 SQ_VTX_SEMANTIC_14 +0x000283BC SQ_VTX_SEMANTIC_15 +0x000283C0 SQ_VTX_SEMANTIC_16 +0x000283C4 SQ_VTX_SEMANTIC_17 +0x000283C8 SQ_VTX_SEMANTIC_18 +0x000283CC SQ_VTX_SEMANTIC_19 +0x000283D0 SQ_VTX_SEMANTIC_20 +0x000283D4 SQ_VTX_SEMANTIC_21 +0x000283D8 SQ_VTX_SEMANTIC_22 +0x000283DC SQ_VTX_SEMANTIC_23 +0x000283E0 SQ_VTX_SEMANTIC_24 +0x000283E4 SQ_VTX_SEMANTIC_25 +0x000283E8 SQ_VTX_SEMANTIC_26 +0x000283EC SQ_VTX_SEMANTIC_27 +0x000283F0 SQ_VTX_SEMANTIC_28 +0x000283F4 SQ_VTX_SEMANTIC_29 +0x000283F8 SQ_VTX_SEMANTIC_30 +0x000283FC SQ_VTX_SEMANTIC_31 +0x00028400 VGT_MAX_VTX_INDX +0x00028404 VGT_MIN_VTX_INDX +0x00028408 VGT_INDX_OFFSET +0x0002840C VGT_MULTI_PRIM_IB_RESET_INDX +0x00028410 SX_ALPHA_TEST_CONTROL +0x00028414 CB_BLEND_RED +0x00028418 CB_BLEND_GREEN +0x0002841C CB_BLEND_BLUE +0x00028420 CB_BLEND_ALPHA +0x00028430 DB_STENCILREFMASK +0x00028434 DB_STENCILREFMASK_BF +0x00028438 SX_ALPHA_REF +0x0002843C PA_CL_VPORT_XSCALE_0 +0x00028440 PA_CL_VPORT_XOFFSET_0 +0x00028444 PA_CL_VPORT_YSCALE_0 +0x00028448 PA_CL_VPORT_YOFFSET_0 +0x0002844C PA_CL_VPORT_ZSCALE_0 +0x00028450 PA_CL_VPORT_ZOFFSET_0 +0x00028454 PA_CL_VPORT_XSCALE_1 +0x00028458 PA_CL_VPORT_XOFFSET_1 +0x0002845C PA_CL_VPORT_YSCALE_1 +0x00028460 PA_CL_VPORT_YOFFSET_1 +0x00028464 PA_CL_VPORT_ZSCALE_1 +0x00028468 PA_CL_VPORT_ZOFFSET_1 +0x0002846C PA_CL_VPORT_XSCALE_2 +0x00028470 PA_CL_VPORT_XOFFSET_2 +0x00028474 PA_CL_VPORT_YSCALE_2 +0x00028478 PA_CL_VPORT_YOFFSET_2 +0x0002847C PA_CL_VPORT_ZSCALE_2 +0x00028480 PA_CL_VPORT_ZOFFSET_2 +0x00028484 PA_CL_VPORT_XSCALE_3 +0x00028488 PA_CL_VPORT_XOFFSET_3 +0x0002848C PA_CL_VPORT_YSCALE_3 +0x00028490 PA_CL_VPORT_YOFFSET_3 +0x00028494 PA_CL_VPORT_ZSCALE_3 +0x00028498 PA_CL_VPORT_ZOFFSET_3 +0x0002849C PA_CL_VPORT_XSCALE_4 +0x000284A0 PA_CL_VPORT_XOFFSET_4 +0x000284A4 PA_CL_VPORT_YSCALE_4 +0x000284A8 PA_CL_VPORT_YOFFSET_4 +0x000284AC PA_CL_VPORT_ZSCALE_4 +0x000284B0 PA_CL_VPORT_ZOFFSET_4 +0x000284B4 PA_CL_VPORT_XSCALE_5 +0x000284B8 PA_CL_VPORT_XOFFSET_5 +0x000284BC PA_CL_VPORT_YSCALE_5 +0x000284C0 PA_CL_VPORT_YOFFSET_5 +0x000284C4 PA_CL_VPORT_ZSCALE_5 +0x000284C8 PA_CL_VPORT_ZOFFSET_5 +0x000284CC PA_CL_VPORT_XSCALE_6 +0x000284D0 PA_CL_VPORT_XOFFSET_6 +0x000284D4 PA_CL_VPORT_YSCALE_6 +0x000284D8 PA_CL_VPORT_YOFFSET_6 +0x000284DC PA_CL_VPORT_ZSCALE_6 +0x000284E0 PA_CL_VPORT_ZOFFSET_6 +0x000284E4 PA_CL_VPORT_XSCALE_7 +0x000284E8 PA_CL_VPORT_XOFFSET_7 +0x000284EC PA_CL_VPORT_YSCALE_7 +0x000284F0 PA_CL_VPORT_YOFFSET_7 +0x000284F4 PA_CL_VPORT_ZSCALE_7 +0x000284F8 PA_CL_VPORT_ZOFFSET_7 +0x000284FC PA_CL_VPORT_XSCALE_8 +0x00028500 PA_CL_VPORT_XOFFSET_8 +0x00028504 PA_CL_VPORT_YSCALE_8 +0x00028508 PA_CL_VPORT_YOFFSET_8 +0x0002850C PA_CL_VPORT_ZSCALE_8 +0x00028510 PA_CL_VPORT_ZOFFSET_8 +0x00028514 PA_CL_VPORT_XSCALE_9 +0x00028518 PA_CL_VPORT_XOFFSET_9 +0x0002851C PA_CL_VPORT_YSCALE_9 +0x00028520 PA_CL_VPORT_YOFFSET_9 +0x00028524 PA_CL_VPORT_ZSCALE_9 +0x00028528 PA_CL_VPORT_ZOFFSET_9 +0x0002852C PA_CL_VPORT_XSCALE_10 +0x00028530 PA_CL_VPORT_XOFFSET_10 +0x00028534 PA_CL_VPORT_YSCALE_10 +0x00028538 PA_CL_VPORT_YOFFSET_10 +0x0002853C PA_CL_VPORT_ZSCALE_10 +0x00028540 PA_CL_VPORT_ZOFFSET_10 +0x00028544 PA_CL_VPORT_XSCALE_11 +0x00028548 PA_CL_VPORT_XOFFSET_11 +0x0002854C PA_CL_VPORT_YSCALE_11 +0x00028550 PA_CL_VPORT_YOFFSET_11 +0x00028554 PA_CL_VPORT_ZSCALE_11 +0x00028558 PA_CL_VPORT_ZOFFSET_11 +0x0002855C PA_CL_VPORT_XSCALE_12 +0x00028560 PA_CL_VPORT_XOFFSET_12 +0x00028564 PA_CL_VPORT_YSCALE_12 +0x00028568 PA_CL_VPORT_YOFFSET_12 +0x0002856C PA_CL_VPORT_ZSCALE_12 +0x00028570 PA_CL_VPORT_ZOFFSET_12 +0x00028574 PA_CL_VPORT_XSCALE_13 +0x00028578 PA_CL_VPORT_XOFFSET_13 +0x0002857C PA_CL_VPORT_YSCALE_13 +0x00028580 PA_CL_VPORT_YOFFSET_13 +0x00028584 PA_CL_VPORT_ZSCALE_13 +0x00028588 PA_CL_VPORT_ZOFFSET_13 +0x0002858C PA_CL_VPORT_XSCALE_14 +0x00028590 PA_CL_VPORT_XOFFSET_14 +0x00028594 PA_CL_VPORT_YSCALE_14 +0x00028598 PA_CL_VPORT_YOFFSET_14 +0x0002859C PA_CL_VPORT_ZSCALE_14 +0x000285A0 PA_CL_VPORT_ZOFFSET_14 +0x000285A4 PA_CL_VPORT_XSCALE_15 +0x000285A8 PA_CL_VPORT_XOFFSET_15 +0x000285AC PA_CL_VPORT_YSCALE_15 +0x000285B0 PA_CL_VPORT_YOFFSET_15 +0x000285B4 PA_CL_VPORT_ZSCALE_15 +0x000285B8 PA_CL_VPORT_ZOFFSET_15 +0x000285BC PA_CL_UCP_0_X +0x000285C0 PA_CL_UCP_0_Y +0x000285C4 PA_CL_UCP_0_Z +0x000285C8 PA_CL_UCP_0_W +0x000285CC PA_CL_UCP_1_X +0x000285D0 PA_CL_UCP_1_Y +0x000285D4 PA_CL_UCP_1_Z +0x000285D8 PA_CL_UCP_1_W +0x000285DC PA_CL_UCP_2_X +0x000285E0 PA_CL_UCP_2_Y +0x000285E4 PA_CL_UCP_2_Z +0x000285E8 PA_CL_UCP_2_W +0x000285EC PA_CL_UCP_3_X +0x000285F0 PA_CL_UCP_3_Y +0x000285F4 PA_CL_UCP_3_Z +0x000285F8 PA_CL_UCP_3_W +0x000285FC PA_CL_UCP_4_X +0x00028600 PA_CL_UCP_4_Y +0x00028604 PA_CL_UCP_4_Z +0x00028608 PA_CL_UCP_4_W +0x0002860C PA_CL_UCP_5_X +0x00028610 PA_CL_UCP_5_Y +0x00028614 PA_CL_UCP_5_Z +0x00028618 PA_CL_UCP_5_W +0x0002861C SPI_VS_OUT_ID_0 +0x00028620 SPI_VS_OUT_ID_1 +0x00028624 SPI_VS_OUT_ID_2 +0x00028628 SPI_VS_OUT_ID_3 +0x0002862C SPI_VS_OUT_ID_4 +0x00028630 SPI_VS_OUT_ID_5 +0x00028634 SPI_VS_OUT_ID_6 +0x00028638 SPI_VS_OUT_ID_7 +0x0002863C SPI_VS_OUT_ID_8 +0x00028640 SPI_VS_OUT_ID_9 +0x00028644 SPI_PS_INPUT_CNTL_0 +0x00028648 SPI_PS_INPUT_CNTL_1 +0x0002864C SPI_PS_INPUT_CNTL_2 +0x00028650 SPI_PS_INPUT_CNTL_3 +0x00028654 SPI_PS_INPUT_CNTL_4 +0x00028658 SPI_PS_INPUT_CNTL_5 +0x0002865C SPI_PS_INPUT_CNTL_6 +0x00028660 SPI_PS_INPUT_CNTL_7 +0x00028664 SPI_PS_INPUT_CNTL_8 +0x00028668 SPI_PS_INPUT_CNTL_9 +0x0002866C SPI_PS_INPUT_CNTL_10 +0x00028670 SPI_PS_INPUT_CNTL_11 +0x00028674 SPI_PS_INPUT_CNTL_12 +0x00028678 SPI_PS_INPUT_CNTL_13 +0x0002867C SPI_PS_INPUT_CNTL_14 +0x00028680 SPI_PS_INPUT_CNTL_15 +0x00028684 SPI_PS_INPUT_CNTL_16 +0x00028688 SPI_PS_INPUT_CNTL_17 +0x0002868C SPI_PS_INPUT_CNTL_18 +0x00028690 SPI_PS_INPUT_CNTL_19 +0x00028694 SPI_PS_INPUT_CNTL_20 +0x00028698 SPI_PS_INPUT_CNTL_21 +0x0002869C SPI_PS_INPUT_CNTL_22 +0x000286A0 SPI_PS_INPUT_CNTL_23 +0x000286A4 SPI_PS_INPUT_CNTL_24 +0x000286A8 SPI_PS_INPUT_CNTL_25 +0x000286AC SPI_PS_INPUT_CNTL_26 +0x000286B0 SPI_PS_INPUT_CNTL_27 +0x000286B4 SPI_PS_INPUT_CNTL_28 +0x000286B8 SPI_PS_INPUT_CNTL_29 +0x000286BC SPI_PS_INPUT_CNTL_30 +0x000286C0 SPI_PS_INPUT_CNTL_31 +0x000286C4 SPI_VS_OUT_CONFIG +0x000286C8 SPI_THREAD_GROUPING +0x000286CC SPI_PS_IN_CONTROL_0 +0x000286D0 SPI_PS_IN_CONTROL_1 +0x000286D4 SPI_INTERP_CONTROL_0 +0x000286D8 SPI_INPUT_Z +0x000286DC SPI_FOG_CNTL +0x000286E0 SPI_BARYC_CNTL +0x000286E4 SPI_PS_IN_CONTROL_2 +0x000286E8 SPI_COMPUTE_INPUT_CNTL +0x000286EC SPI_COMPUTE_NUM_THREAD_X +0x000286F0 SPI_COMPUTE_NUM_THREAD_Y +0x000286F4 SPI_COMPUTE_NUM_THREAD_Z +0x000286F8 SPI_GPR_MGMT +0x000286FC SPI_LDS_MGMT +0x00028700 SPI_STACK_MGMT +0x00028704 SPI_WAVE_MGMT_1 +0x00028708 SPI_WAVE_MGMT_2 +0x00028724 GDS_ADDR_SIZE +0x00028780 CB_BLEND0_CONTROL +0x00028784 CB_BLEND1_CONTROL +0x00028788 CB_BLEND2_CONTROL +0x0002878C CB_BLEND3_CONTROL +0x00028790 CB_BLEND4_CONTROL +0x00028794 CB_BLEND5_CONTROL +0x00028798 CB_BLEND6_CONTROL +0x0002879C CB_BLEND7_CONTROL +0x000287CC CS_COPY_STATE +0x000287D0 GFX_COPY_STATE +0x000287D4 PA_CL_POINT_X_RAD +0x000287D8 PA_CL_POINT_Y_RAD +0x000287DC PA_CL_POINT_SIZE +0x000287E0 PA_CL_POINT_CULL_RAD +0x00028808 CB_COLOR_CONTROL +0x0002880C DB_SHADER_CONTROL +0x00028810 PA_CL_CLIP_CNTL +0x00028814 PA_SU_SC_MODE_CNTL +0x00028818 PA_CL_VTE_CNTL +0x0002881C PA_CL_VS_OUT_CNTL +0x00028820 PA_CL_NANINF_CNTL +0x00028824 PA_SU_LINE_STIPPLE_CNTL +0x00028828 PA_SU_LINE_STIPPLE_SCALE +0x0002882C PA_SU_PRIM_FILTER_CNTL +0x00028844 SQ_PGM_RESOURCES_PS +0x00028848 SQ_PGM_RESOURCES_2_PS +0x0002884C SQ_PGM_EXPORTS_PS +0x00028860 SQ_PGM_RESOURCES_VS +0x00028864 SQ_PGM_RESOURCES_2_VS +0x00028878 SQ_PGM_RESOURCES_GS +0x0002887C SQ_PGM_RESOURCES_2_GS +0x00028890 SQ_PGM_RESOURCES_ES +0x00028894 SQ_PGM_RESOURCES_2_ES +0x000288A8 SQ_PGM_RESOURCES_FS +0x000288BC SQ_PGM_RESOURCES_HS +0x000288C0 SQ_PGM_RESOURCES_2_HS +0x000288D4 SQ_PGM_RESOURCES_LS +0x000288D8 SQ_PGM_RESOURCES_2_LS +0x000288E8 SQ_LDS_ALLOC +0x000288EC SQ_LDS_ALLOC_PS +0x000288F0 SQ_VTX_SEMANTIC_CLEAR +0x00028A00 PA_SU_POINT_SIZE +0x00028A04 PA_SU_POINT_MINMAX +0x00028A08 PA_SU_LINE_CNTL +0x00028A0C PA_SC_LINE_STIPPLE +0x00028A10 VGT_OUTPUT_PATH_CNTL +0x00028A14 VGT_HOS_CNTL +0x00028A18 VGT_HOS_MAX_TESS_LEVEL +0x00028A1C VGT_HOS_MIN_TESS_LEVEL +0x00028A20 VGT_HOS_REUSE_DEPTH +0x00028A24 VGT_GROUP_PRIM_TYPE +0x00028A28 VGT_GROUP_FIRST_DECR +0x00028A2C VGT_GROUP_DECR +0x00028A30 VGT_GROUP_VECT_0_CNTL +0x00028A34 VGT_GROUP_VECT_1_CNTL +0x00028A38 VGT_GROUP_VECT_0_FMT_CNTL +0x00028A3C VGT_GROUP_VECT_1_FMT_CNTL +0x00028A40 VGT_GS_MODE +0x00028A48 PA_SC_MODE_CNTL_0 +0x00028A4C PA_SC_MODE_CNTL_1 +0x00028A50 VGT_ENHANCE +0x00028A54 VGT_GS_PER_ES +0x00028A58 VGT_ES_PER_GS +0x00028A5C VGT_GS_PER_VS +0x00028A6C VGT_GS_OUT_PRIM_TYPE +0x00028A70 IA_ENHANCE +0x00028A84 VGT_PRIMITIVEID_EN +0x00028A94 VGT_MULTI_PRIM_IB_RESET_EN +0x00028AA0 VGT_INSTANCE_STEP_RATE_0 +0x00028AA4 VGT_INSTANCE_STEP_RATE_1 +0x00028AA8 IA_MULTI_VGT_PARAM +0x00028AB4 VGT_REUSE_OFF +0x00028AB8 VGT_VTX_CNT_EN +0x00028ABC DB_HTILE_SURFACE +0x00028AC0 DB_SRESULTS_COMPARE_STATE0 +0x00028AC4 DB_SRESULTS_COMPARE_STATE1 +0x00028AC8 DB_PRELOAD_CONTROL +0x00028B38 VGT_GS_MAX_VERT_OUT +0x00028B54 VGT_SHADER_STAGES_EN +0x00028B58 VGT_LS_HS_CONFIG +0x00028B6C VGT_TF_PARAM +0x00028B70 DB_ALPHA_TO_MASK +0x00028B74 VGT_DISPATCH_INITIATOR +0x00028B78 PA_SU_POLY_OFFSET_DB_FMT_CNTL +0x00028B7C PA_SU_POLY_OFFSET_CLAMP +0x00028B80 PA_SU_POLY_OFFSET_FRONT_SCALE +0x00028B84 PA_SU_POLY_OFFSET_FRONT_OFFSET +0x00028B88 PA_SU_POLY_OFFSET_BACK_SCALE +0x00028B8C PA_SU_POLY_OFFSET_BACK_OFFSET +0x00028B74 VGT_GS_INSTANCE_CNT +0x00028BD4 PA_SC_CENTROID_PRIORITY_0 +0x00028BD8 PA_SC_CENTROID_PRIORITY_1 +0x00028BDC PA_SC_LINE_CNTL +0x00028BE4 PA_SU_VTX_CNTL +0x00028BE8 PA_CL_GB_VERT_CLIP_ADJ +0x00028BEC PA_CL_GB_VERT_DISC_ADJ +0x00028BF0 PA_CL_GB_HORZ_CLIP_ADJ +0x00028BF4 PA_CL_GB_HORZ_DISC_ADJ +0x00028BF8 PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y0_0 +0x00028BFC PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y0_1 +0x00028C00 PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y0_2 +0x00028C04 PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y0_3 +0x00028C08 PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y0_0 +0x00028C0C PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y0_1 +0x00028C10 PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y0_2 +0x00028C14 PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y0_3 +0x00028C18 PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y1_0 +0x00028C1C PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y1_1 +0x00028C20 PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y1_2 +0x00028C24 PA_SC_AA_SAMPLE_LOCS_PIXEL_X0_Y1_3 +0x00028C28 PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y1_0 +0x00028C2C PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y1_1 +0x00028C30 PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y1_2 +0x00028C34 PA_SC_AA_SAMPLE_LOCS_PIXEL_X1_Y1_3 +0x00028C38 PA_SC_AA_MASK_X0_Y0_X1_Y0 +0x00028C3C PA_SC_AA_MASK_X0_Y1_X1_Y1 +0x00028C8C CB_COLOR0_CLEAR_WORD0 +0x00028C90 CB_COLOR0_CLEAR_WORD1 +0x00028C94 CB_COLOR0_CLEAR_WORD2 +0x00028C98 CB_COLOR0_CLEAR_WORD3 +0x00028CC8 CB_COLOR1_CLEAR_WORD0 +0x00028CCC CB_COLOR1_CLEAR_WORD1 +0x00028CD0 CB_COLOR1_CLEAR_WORD2 +0x00028CD4 CB_COLOR1_CLEAR_WORD3 +0x00028D04 CB_COLOR2_CLEAR_WORD0 +0x00028D08 CB_COLOR2_CLEAR_WORD1 +0x00028D0C CB_COLOR2_CLEAR_WORD2 +0x00028D10 CB_COLOR2_CLEAR_WORD3 +0x00028D40 CB_COLOR3_CLEAR_WORD0 +0x00028D44 CB_COLOR3_CLEAR_WORD1 +0x00028D48 CB_COLOR3_CLEAR_WORD2 +0x00028D4C CB_COLOR3_CLEAR_WORD3 +0x00028D7C CB_COLOR4_CLEAR_WORD0 +0x00028D80 CB_COLOR4_CLEAR_WORD1 +0x00028D84 CB_COLOR4_CLEAR_WORD2 +0x00028D88 CB_COLOR4_CLEAR_WORD3 +0x00028DB8 CB_COLOR5_CLEAR_WORD0 +0x00028DBC CB_COLOR5_CLEAR_WORD1 +0x00028DC0 CB_COLOR5_CLEAR_WORD2 +0x00028DC4 CB_COLOR5_CLEAR_WORD3 +0x00028DF4 CB_COLOR6_CLEAR_WORD0 +0x00028DF8 CB_COLOR6_CLEAR_WORD1 +0x00028DFC CB_COLOR6_CLEAR_WORD2 +0x00028E00 CB_COLOR6_CLEAR_WORD3 +0x00028E30 CB_COLOR7_CLEAR_WORD0 +0x00028E34 CB_COLOR7_CLEAR_WORD1 +0x00028E38 CB_COLOR7_CLEAR_WORD2 +0x00028E3C CB_COLOR7_CLEAR_WORD3 +0x00028F80 SQ_ALU_CONST_BUFFER_SIZE_HS_0 +0x00028F84 SQ_ALU_CONST_BUFFER_SIZE_HS_1 +0x00028F88 SQ_ALU_CONST_BUFFER_SIZE_HS_2 +0x00028F8C SQ_ALU_CONST_BUFFER_SIZE_HS_3 +0x00028F90 SQ_ALU_CONST_BUFFER_SIZE_HS_4 +0x00028F94 SQ_ALU_CONST_BUFFER_SIZE_HS_5 +0x00028F98 SQ_ALU_CONST_BUFFER_SIZE_HS_6 +0x00028F9C SQ_ALU_CONST_BUFFER_SIZE_HS_7 +0x00028FA0 SQ_ALU_CONST_BUFFER_SIZE_HS_8 +0x00028FA4 SQ_ALU_CONST_BUFFER_SIZE_HS_9 +0x00028FA8 SQ_ALU_CONST_BUFFER_SIZE_HS_10 +0x00028FAC SQ_ALU_CONST_BUFFER_SIZE_HS_11 +0x00028FB0 SQ_ALU_CONST_BUFFER_SIZE_HS_12 +0x00028FB4 SQ_ALU_CONST_BUFFER_SIZE_HS_13 +0x00028FB8 SQ_ALU_CONST_BUFFER_SIZE_HS_14 +0x00028FBC SQ_ALU_CONST_BUFFER_SIZE_HS_15 +0x00028FC0 SQ_ALU_CONST_BUFFER_SIZE_LS_0 +0x00028FC4 SQ_ALU_CONST_BUFFER_SIZE_LS_1 +0x00028FC8 SQ_ALU_CONST_BUFFER_SIZE_LS_2 +0x00028FCC SQ_ALU_CONST_BUFFER_SIZE_LS_3 +0x00028FD0 SQ_ALU_CONST_BUFFER_SIZE_LS_4 +0x00028FD4 SQ_ALU_CONST_BUFFER_SIZE_LS_5 +0x00028FD8 SQ_ALU_CONST_BUFFER_SIZE_LS_6 +0x00028FDC SQ_ALU_CONST_BUFFER_SIZE_LS_7 +0x00028FE0 SQ_ALU_CONST_BUFFER_SIZE_LS_8 +0x00028FE4 SQ_ALU_CONST_BUFFER_SIZE_LS_9 +0x00028FE8 SQ_ALU_CONST_BUFFER_SIZE_LS_10 +0x00028FEC SQ_ALU_CONST_BUFFER_SIZE_LS_11 +0x00028FF0 SQ_ALU_CONST_BUFFER_SIZE_LS_12 +0x00028FF4 SQ_ALU_CONST_BUFFER_SIZE_LS_13 +0x00028FF8 SQ_ALU_CONST_BUFFER_SIZE_LS_14 +0x00028FFC SQ_ALU_CONST_BUFFER_SIZE_LS_15 +0x0003CFF0 SQ_VTX_BASE_VTX_LOC +0x0003CFF4 SQ_VTX_START_INST_LOC +0x0003FF00 SQ_TEX_SAMPLER_CLEAR +0x0003FF04 SQ_TEX_RESOURCE_CLEAR +0x0003FF08 SQ_LOOP_BOOL_CLEAR diff --git a/drivers/gpu/drm/radeon/reg_srcs/evergreen b/drivers/gpu/drm/radeon/reg_srcs/evergreen index 9177f91..7e16371 100644 --- a/drivers/gpu/drm/radeon/reg_srcs/evergreen +++ b/drivers/gpu/drm/radeon/reg_srcs/evergreen @@ -1,4 +1,5 @@ evergreen 0x9400 +0x0000802C GRBM_GFX_INDEX 0x00008040 WAIT_UNTIL 0x00008044 WAIT_UNTIL_POLL_CNTL 0x00008048 WAIT_UNTIL_POLL_MASK @@ -220,6 +221,7 @@ evergreen 0x9400 0x00028348 PA_SC_VPORT_ZMIN_15 0x0002834C PA_SC_VPORT_ZMAX_15 0x00028350 SX_MISC +0x00028354 SX_SURFACE_SYNC 0x00028380 SQ_VTX_SEMANTIC_0 0x00028384 SQ_VTX_SEMANTIC_1 0x00028388 SQ_VTX_SEMANTIC_2 diff --git a/drivers/gpu/drm/radeon/reg_srcs/r300 b/drivers/gpu/drm/radeon/reg_srcs/r300 index b506ec1..e8a1786 100644 --- a/drivers/gpu/drm/radeon/reg_srcs/r300 +++ b/drivers/gpu/drm/radeon/reg_srcs/r300 @@ -683,9 +683,7 @@ r300 0x4f60 0x4DF4 US_ALU_CONST_G_31 0x4DF8 US_ALU_CONST_B_31 0x4DFC US_ALU_CONST_A_31 -0x4E04 RB3D_BLENDCNTL_R3 0x4E08 RB3D_ABLENDCNTL_R3 -0x4E0C RB3D_COLOR_CHANNEL_MASK 0x4E10 RB3D_CONSTANT_COLOR 0x4E14 RB3D_COLOR_CLEAR_VALUE 0x4E18 RB3D_ROPCNTL_R3 @@ -706,13 +704,11 @@ r300 0x4f60 0x4E74 RB3D_CMASK_WRINDEX 0x4E78 RB3D_CMASK_DWORD 0x4E7C RB3D_CMASK_RDINDEX -0x4E80 RB3D_AARESOLVE_OFFSET -0x4E84 RB3D_AARESOLVE_PITCH -0x4E88 RB3D_AARESOLVE_CTL 0x4EA0 RB3D_DISCARD_SRC_PIXEL_LTE_THRESHOLD 0x4EA4 RB3D_DISCARD_SRC_PIXEL_GTE_THRESHOLD 0x4F04 ZB_ZSTENCILCNTL 0x4F08 ZB_STENCILREFMASK 0x4F14 ZB_ZTOP 0x4F18 ZB_ZCACHE_CTLSTAT +0x4F28 ZB_DEPTHCLEARVALUE 0x4F58 ZB_ZPASS_DATA diff --git a/drivers/gpu/drm/radeon/reg_srcs/r420 b/drivers/gpu/drm/radeon/reg_srcs/r420 index 8c1214c..722074e 100644 --- a/drivers/gpu/drm/radeon/reg_srcs/r420 +++ b/drivers/gpu/drm/radeon/reg_srcs/r420 @@ -130,7 +130,6 @@ r420 0x4f60 0x401C GB_SELECT 0x4020 GB_AA_CONFIG 0x4024 GB_FIFO_SIZE -0x4028 GB_Z_PEQ_CONFIG 0x4100 TX_INVALTAGS 0x4200 GA_POINT_S0 0x4204 GA_POINT_T0 @@ -750,9 +749,7 @@ r420 0x4f60 0x4DF4 US_ALU_CONST_G_31 0x4DF8 US_ALU_CONST_B_31 0x4DFC US_ALU_CONST_A_31 -0x4E04 RB3D_BLENDCNTL_R3 0x4E08 RB3D_ABLENDCNTL_R3 -0x4E0C RB3D_COLOR_CHANNEL_MASK 0x4E10 RB3D_CONSTANT_COLOR 0x4E14 RB3D_COLOR_CLEAR_VALUE 0x4E18 RB3D_ROPCNTL_R3 @@ -773,13 +770,11 @@ r420 0x4f60 0x4E74 RB3D_CMASK_WRINDEX 0x4E78 RB3D_CMASK_DWORD 0x4E7C RB3D_CMASK_RDINDEX -0x4E80 RB3D_AARESOLVE_OFFSET -0x4E84 RB3D_AARESOLVE_PITCH -0x4E88 RB3D_AARESOLVE_CTL 0x4EA0 RB3D_DISCARD_SRC_PIXEL_LTE_THRESHOLD 0x4EA4 RB3D_DISCARD_SRC_PIXEL_GTE_THRESHOLD 0x4F04 ZB_ZSTENCILCNTL 0x4F08 ZB_STENCILREFMASK 0x4F14 ZB_ZTOP 0x4F18 ZB_ZCACHE_CTLSTAT +0x4F28 ZB_DEPTHCLEARVALUE 0x4F58 ZB_ZPASS_DATA diff --git a/drivers/gpu/drm/radeon/reg_srcs/rs600 b/drivers/gpu/drm/radeon/reg_srcs/rs600 index 0828d80..d9f6286 100644 --- a/drivers/gpu/drm/radeon/reg_srcs/rs600 +++ b/drivers/gpu/drm/radeon/reg_srcs/rs600 @@ -749,9 +749,7 @@ rs600 0x6d40 0x4DF4 US_ALU_CONST_G_31 0x4DF8 US_ALU_CONST_B_31 0x4DFC US_ALU_CONST_A_31 -0x4E04 RB3D_BLENDCNTL_R3 0x4E08 RB3D_ABLENDCNTL_R3 -0x4E0C RB3D_COLOR_CHANNEL_MASK 0x4E10 RB3D_CONSTANT_COLOR 0x4E14 RB3D_COLOR_CLEAR_VALUE 0x4E18 RB3D_ROPCNTL_R3 @@ -772,13 +770,11 @@ rs600 0x6d40 0x4E74 RB3D_CMASK_WRINDEX 0x4E78 RB3D_CMASK_DWORD 0x4E7C RB3D_CMASK_RDINDEX -0x4E80 RB3D_AARESOLVE_OFFSET -0x4E84 RB3D_AARESOLVE_PITCH -0x4E88 RB3D_AARESOLVE_CTL 0x4EA0 RB3D_DISCARD_SRC_PIXEL_LTE_THRESHOLD 0x4EA4 RB3D_DISCARD_SRC_PIXEL_GTE_THRESHOLD 0x4F04 ZB_ZSTENCILCNTL 0x4F08 ZB_STENCILREFMASK 0x4F14 ZB_ZTOP 0x4F18 ZB_ZCACHE_CTLSTAT +0x4F28 ZB_DEPTHCLEARVALUE 0x4F58 ZB_ZPASS_DATA diff --git a/drivers/gpu/drm/radeon/reg_srcs/rv515 b/drivers/gpu/drm/radeon/reg_srcs/rv515 index ef422bb..911a8fb 100644 --- a/drivers/gpu/drm/radeon/reg_srcs/rv515 +++ b/drivers/gpu/drm/radeon/reg_srcs/rv515 @@ -164,7 +164,6 @@ rv515 0x6d40 0x401C GB_SELECT 0x4020 GB_AA_CONFIG 0x4024 GB_FIFO_SIZE -0x4028 GB_Z_PEQ_CONFIG 0x4100 TX_INVALTAGS 0x4114 SU_TEX_WRAP_PS3 0x4118 PS3_ENABLE @@ -461,9 +460,7 @@ rv515 0x6d40 0x4DF4 US_ALU_CONST_G_31 0x4DF8 US_ALU_CONST_B_31 0x4DFC US_ALU_CONST_A_31 -0x4E04 RB3D_BLENDCNTL_R3 0x4E08 RB3D_ABLENDCNTL_R3 -0x4E0C RB3D_COLOR_CHANNEL_MASK 0x4E10 RB3D_CONSTANT_COLOR 0x4E14 RB3D_COLOR_CLEAR_VALUE 0x4E18 RB3D_ROPCNTL_R3 @@ -484,9 +481,6 @@ rv515 0x6d40 0x4E74 RB3D_CMASK_WRINDEX 0x4E78 RB3D_CMASK_DWORD 0x4E7C RB3D_CMASK_RDINDEX -0x4E80 RB3D_AARESOLVE_OFFSET -0x4E84 RB3D_AARESOLVE_PITCH -0x4E88 RB3D_AARESOLVE_CTL 0x4EA0 RB3D_DISCARD_SRC_PIXEL_LTE_THRESHOLD 0x4EA4 RB3D_DISCARD_SRC_PIXEL_GTE_THRESHOLD 0x4EF8 RB3D_CONSTANT_COLOR_AR @@ -496,4 +490,5 @@ rv515 0x6d40 0x4F14 ZB_ZTOP 0x4F18 ZB_ZCACHE_CTLSTAT 0x4F58 ZB_ZPASS_DATA +0x4F28 ZB_DEPTHCLEARVALUE 0x4FD4 ZB_STENCILREFMASK_BF diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c index 26807bf..19763f5 100644 --- a/drivers/gpu/drm/radeon/rs600.c +++ b/drivers/gpu/drm/radeon/rs600.c @@ -751,7 +751,6 @@ void rs600_mc_init(struct radeon_device *rdev) rdev->mc.real_vram_size = RREG32(RADEON_CONFIG_MEMSIZE); rdev->mc.mc_vram_size = rdev->mc.real_vram_size; rdev->mc.visible_vram_size = rdev->mc.aper_size; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); base = RREG32_MC(R_000004_MC_FB_LOCATION); base = G_000004_MC_FB_START(base) << 16; diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c index 2e72b4f..a9049ed 100644 --- a/drivers/gpu/drm/radeon/rs690.c +++ b/drivers/gpu/drm/radeon/rs690.c @@ -77,9 +77,9 @@ void rs690_pm_info(struct radeon_device *rdev) switch (crev) { case 1: tmp.full = dfixed_const(100); - rdev->pm.igp_sideport_mclk.full = dfixed_const(info->info.ulBootUpMemoryClock); + rdev->pm.igp_sideport_mclk.full = dfixed_const(le32_to_cpu(info->info.ulBootUpMemoryClock)); rdev->pm.igp_sideport_mclk.full = dfixed_div(rdev->pm.igp_sideport_mclk, tmp); - if (info->info.usK8MemoryClock) + if (le16_to_cpu(info->info.usK8MemoryClock)) rdev->pm.igp_system_mclk.full = dfixed_const(le16_to_cpu(info->info.usK8MemoryClock)); else if (rdev->clock.default_mclk) { rdev->pm.igp_system_mclk.full = dfixed_const(rdev->clock.default_mclk); @@ -91,16 +91,16 @@ void rs690_pm_info(struct radeon_device *rdev) break; case 2: tmp.full = dfixed_const(100); - rdev->pm.igp_sideport_mclk.full = dfixed_const(info->info_v2.ulBootUpSidePortClock); + rdev->pm.igp_sideport_mclk.full = dfixed_const(le32_to_cpu(info->info_v2.ulBootUpSidePortClock)); rdev->pm.igp_sideport_mclk.full = dfixed_div(rdev->pm.igp_sideport_mclk, tmp); - if (info->info_v2.ulBootUpUMAClock) - rdev->pm.igp_system_mclk.full = dfixed_const(info->info_v2.ulBootUpUMAClock); + if (le32_to_cpu(info->info_v2.ulBootUpUMAClock)) + rdev->pm.igp_system_mclk.full = dfixed_const(le32_to_cpu(info->info_v2.ulBootUpUMAClock)); else if (rdev->clock.default_mclk) rdev->pm.igp_system_mclk.full = dfixed_const(rdev->clock.default_mclk); else rdev->pm.igp_system_mclk.full = dfixed_const(66700); rdev->pm.igp_system_mclk.full = dfixed_div(rdev->pm.igp_system_mclk, tmp); - rdev->pm.igp_ht_link_clk.full = dfixed_const(info->info_v2.ulHTLinkFreq); + rdev->pm.igp_ht_link_clk.full = dfixed_const(le32_to_cpu(info->info_v2.ulHTLinkFreq)); rdev->pm.igp_ht_link_clk.full = dfixed_div(rdev->pm.igp_ht_link_clk, tmp); rdev->pm.igp_ht_link_width.full = dfixed_const(le16_to_cpu(info->info_v2.usMinHTLinkWidth)); break; @@ -157,7 +157,6 @@ void rs690_mc_init(struct radeon_device *rdev) rdev->mc.aper_base = pci_resource_start(rdev->pdev, 0); rdev->mc.aper_size = pci_resource_len(rdev->pdev, 0); rdev->mc.visible_vram_size = rdev->mc.aper_size; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; base = RREG32_MC(R_000100_MCCFG_FB_LOCATION); base = G_000100_MC_FB_START(base) << 16; rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index 866cf2f..b974ac7 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -307,7 +307,7 @@ static void rv770_mc_program(struct radeon_device *rdev) */ void r700_cp_stop(struct radeon_device *rdev) { - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; + radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT)); WREG32(SCRATCH_UMSK, 0); } @@ -321,7 +321,11 @@ static int rv770_cp_load_microcode(struct radeon_device *rdev) return -EINVAL; r700_cp_stop(rdev); - WREG32(CP_RB_CNTL, RB_NO_UPDATE | (15 << 8) | (3 << 0)); + WREG32(CP_RB_CNTL, +#ifdef __BIG_ENDIAN + BUF_SWAP_32BIT | +#endif + RB_NO_UPDATE | RB_BLKSZ(15) | RB_BUFSZ(3)); /* Reset cp */ WREG32(GRBM_SOFT_RESET, SOFT_RESET_CP); @@ -999,7 +1003,7 @@ static int rv770_vram_scratch_init(struct radeon_device *rdev) u64 gpu_addr; if (rdev->vram_scratch.robj == NULL) { - r = radeon_bo_create(rdev, NULL, RADEON_GPU_PAGE_SIZE, + r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, &rdev->vram_scratch.robj); if (r) { @@ -1119,7 +1123,6 @@ int rv770_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; - rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r700_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); diff --git a/drivers/gpu/drm/radeon/rv770d.h b/drivers/gpu/drm/radeon/rv770d.h index abc8cf5..79fa588 100644 --- a/drivers/gpu/drm/radeon/rv770d.h +++ b/drivers/gpu/drm/radeon/rv770d.h @@ -76,10 +76,10 @@ #define ROQ_IB1_START(x) ((x) << 0) #define ROQ_IB2_START(x) ((x) << 8) #define CP_RB_CNTL 0xC104 -#define RB_BUFSZ(x) ((x)<<0) -#define RB_BLKSZ(x) ((x)<<8) -#define RB_NO_UPDATE (1<<27) -#define RB_RPTR_WR_ENA (1<<31) +#define RB_BUFSZ(x) ((x) << 0) +#define RB_BLKSZ(x) ((x) << 8) +#define RB_NO_UPDATE (1 << 27) +#define RB_RPTR_WR_ENA (1 << 31) #define BUF_SWAP_32BIT (2 << 16) #define CP_RB_RPTR 0x8700 #define CP_RB_RPTR_ADDR 0xC10C |
