diff options
| author | Linus Torvalds | 2014-10-14 09:39:08 +0200 |
|---|---|---|
| committer | Linus Torvalds | 2014-10-14 09:39:08 +0200 |
| commit | 2d65a9f48fcdf7866aab6457bc707ca233e0c791 (patch) | |
| tree | f93e5838d6ac2e59434367f4ff905f7d9c45fc2b /drivers/gpu/drm/radeon | |
| parent | Merge branch 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/k... (diff) | |
| parent | drm/mst: rework payload table allocation to conform better. (diff) | |
| download | kernel-qcow2-linux-2d65a9f48fcdf7866aab6457bc707ca233e0c791.tar.gz kernel-qcow2-linux-2d65a9f48fcdf7866aab6457bc707ca233e0c791.tar.xz kernel-qcow2-linux-2d65a9f48fcdf7866aab6457bc707ca233e0c791.zip | |
Merge branch 'drm-next' of git://people.freedesktop.org/~airlied/linux
Pull drm updates from Dave Airlie:
"This is the main git pull for the drm,
I pretty much froze major pulls at -rc5/6 time, and haven't had much
fallout, so will probably continue doing that.
Lots of changes all over, big internal header cleanup to make it clear
drm features are legacy things and what are things that modern KMS
drivers should be using. Also big move to use the new generic fences
in all the TTM drivers.
core:
atomic prep work,
vblank rework changes, allows immediate vblank disables
major header reworking and cleanups to better delinate legacy
interfaces from what KMS drivers should be using.
cursor planes locking fixes
ttm:
move to generic fences (affects all TTM drivers)
ppc64 caching fixes
radeon:
userptr support,
uvd for old asics,
reset rework for fence changes
better buffer placement changes,
dpm feature enablement
hdmi audio support fixes
intel:
Cherryview work,
180 degree rotation,
skylake prep work,
execlist command submission
full ppgtt prep work
cursor improvements
edid caching,
vdd handling improvements
nouveau:
fence reworking
kepler memory clock work
gt21x clock work
fan control improvements
hdmi infoframe fixes
DP audio
ast:
ppc64 fixes
caching fix
rcar:
rcar-du DT support
ipuv3:
prep work for capture support
msm:
LVDS support for mdp4, new panel, gpu refactoring
exynos:
exynos3250 SoC support, drop bad mmap interface,
mipi dsi changes, and component match support"
* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (640 commits)
drm/mst: rework payload table allocation to conform better.
drm/ast: Fix HW cursor image
drm/radeon/kv: add uvd/vce info to dpm debugfs output
drm/radeon/ci: add uvd/vce info to dpm debugfs output
drm/radeon: export reservation_object from dmabuf to ttm
drm/radeon: cope with foreign fences inside the reservation object
drm/radeon: cope with foreign fences inside display
drm/core: use helper to check driver features
drm/radeon/cik: write gfx ucode version to ucode addr reg
drm/radeon/si: print full CS when we hit a packet 0
drm/radeon: remove unecessary includes
drm/radeon/combios: declare legacy_connector_convert as static
drm/radeon/atombios: declare connector convert tables as static
drm/radeon: drop btc_get_max_clock_from_voltage_dependency_table
drm/radeon/dpm: drop clk/voltage dependency filters for BTC
drm/radeon/dpm: drop clk/voltage dependency filters for CI
drm/radeon/dpm: drop clk/voltage dependency filters for SI
drm/radeon/dpm: drop clk/voltage dependency filters for NI
drm/radeon: disable audio when we disable hdmi (v2)
drm/radeon: split audio enable between eg and r600 (v2)
...
Diffstat (limited to 'drivers/gpu/drm/radeon')
69 files changed, 2544 insertions, 1080 deletions
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index f77b7135ee4c..d01b87991422 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -60,7 +60,7 @@ radeon-y := radeon_drv.o # add UMS driver radeon-$(CONFIG_DRM_RADEON_UMS)+= radeon_cp.o radeon_state.o radeon_mem.o \ - radeon_irq.o r300_cmdbuf.o r600_cp.o r600_blit.o + radeon_irq.o r300_cmdbuf.o r600_cp.o r600_blit.o drm_buffer.o # add KMS driver radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ @@ -72,7 +72,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ radeon_cs.o radeon_bios.o radeon_benchmark.o r100.o r300.o r420.o \ rs400.o rs600.o rs690.o rv515.o r520.o r600.o rv770.o radeon_test.o \ r200.o radeon_legacy_tv.o r600_cs.o r600_blit_shaders.o \ - radeon_pm.o atombios_dp.o r600_audio.o r600_hdmi.o dce3_1_afmt.o \ + radeon_pm.o atombios_dp.o r600_hdmi.o dce3_1_afmt.o \ evergreen.o evergreen_cs.o evergreen_blit_shaders.o \ evergreen_hdmi.o radeon_trace_points.o ni.o cayman_blit_shaders.o \ atombios_encoders.o radeon_semaphore.o radeon_sa.o atombios_i2c.o si.o \ @@ -80,7 +80,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ r600_dpm.o rs780_dpm.o rv6xx_dpm.o rv770_dpm.o rv730_dpm.o rv740_dpm.o \ rv770_smc.o cypress_dpm.o btc_dpm.o sumo_dpm.o sumo_smc.o trinity_dpm.o \ trinity_smc.o ni_dpm.o si_smc.o si_dpm.o kv_smc.o kv_dpm.o ci_smc.o \ - ci_dpm.o dce6_afmt.o radeon_vm.o radeon_ucode.o radeon_ib.o + ci_dpm.o dce6_afmt.o radeon_vm.o radeon_ucode.o radeon_ib.o radeon_mn.o # add async DMA block radeon-y += \ diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index ac14b67621d3..95d5d4ab3335 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -232,8 +232,8 @@ void radeon_dp_aux_init(struct radeon_connector *radeon_connector) /***** general DP utility functions *****/ -#define DP_VOLTAGE_MAX DP_TRAIN_VOLTAGE_SWING_1200 -#define DP_PRE_EMPHASIS_MAX DP_TRAIN_PRE_EMPHASIS_9_5 +#define DP_VOLTAGE_MAX DP_TRAIN_VOLTAGE_SWING_LEVEL_3 +#define DP_PRE_EMPHASIS_MAX DP_TRAIN_PRE_EMPH_LEVEL_3 static void dp_get_adjust_train(u8 link_status[DP_LINK_STATUS_SIZE], int lane_count, diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index a7f2ddf09a9d..b8cd7975f797 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -291,29 +291,6 @@ static void radeon_atom_backlight_exit(struct radeon_encoder *encoder) bool radeon_atom_get_tv_timings(struct radeon_device *rdev, int index, struct drm_display_mode *mode); - -static inline bool radeon_encoder_is_digital(struct drm_encoder *encoder) -{ - struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); - switch (radeon_encoder->encoder_id) { - case ENCODER_OBJECT_ID_INTERNAL_LVDS: - case ENCODER_OBJECT_ID_INTERNAL_TMDS1: - case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: - case ENCODER_OBJECT_ID_INTERNAL_LVTM1: - case ENCODER_OBJECT_ID_INTERNAL_DVO1: - case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1: - case ENCODER_OBJECT_ID_INTERNAL_DDI: - case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: - case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: - case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: - case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: - case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3: - return true; - default: - return false; - } -} - static bool radeon_atom_mode_fixup(struct drm_encoder *encoder, const struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c index f81d7ca134db..300d971187c4 100644 --- a/drivers/gpu/drm/radeon/btc_dpm.c +++ b/drivers/gpu/drm/radeon/btc_dpm.c @@ -1170,23 +1170,6 @@ static const struct radeon_blacklist_clocks btc_blacklist_clocks[] = { 25000, 30000, RADEON_SCLK_UP } }; -void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table, - u32 *max_clock) -{ - u32 i, clock = 0; - - if ((table == NULL) || (table->count == 0)) { - *max_clock = clock; - return; - } - - for (i = 0; i < table->count; i++) { - if (clock < table->entries[i].clk) - clock = table->entries[i].clk; - } - *max_clock = clock; -} - void btc_apply_voltage_dependency_rules(struct radeon_clock_voltage_dependency_table *table, u32 clock, u16 max_voltage, u16 *voltage) { @@ -2099,7 +2082,6 @@ static void btc_apply_state_adjust_rules(struct radeon_device *rdev, bool disable_mclk_switching; u32 mclk, sclk; u16 vddc, vddci; - u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc; if ((rdev->pm.dpm.new_active_crtc_count > 1) || btc_dpm_vblank_too_short(rdev)) @@ -2141,39 +2123,6 @@ static void btc_apply_state_adjust_rules(struct radeon_device *rdev, ps->low.vddci = max_limits->vddci; } - /* limit clocks to max supported clocks based on voltage dependency tables */ - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk, - &max_sclk_vddc); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk, - &max_mclk_vddci); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk, - &max_mclk_vddc); - - if (max_sclk_vddc) { - if (ps->low.sclk > max_sclk_vddc) - ps->low.sclk = max_sclk_vddc; - if (ps->medium.sclk > max_sclk_vddc) - ps->medium.sclk = max_sclk_vddc; - if (ps->high.sclk > max_sclk_vddc) - ps->high.sclk = max_sclk_vddc; - } - if (max_mclk_vddci) { - if (ps->low.mclk > max_mclk_vddci) - ps->low.mclk = max_mclk_vddci; - if (ps->medium.mclk > max_mclk_vddci) - ps->medium.mclk = max_mclk_vddci; - if (ps->high.mclk > max_mclk_vddci) - ps->high.mclk = max_mclk_vddci; - } - if (max_mclk_vddc) { - if (ps->low.mclk > max_mclk_vddc) - ps->low.mclk = max_mclk_vddc; - if (ps->medium.mclk > max_mclk_vddc) - ps->medium.mclk = max_mclk_vddc; - if (ps->high.mclk > max_mclk_vddc) - ps->high.mclk = max_mclk_vddc; - } - /* XXX validate the min clocks required for display */ if (disable_mclk_switching) { diff --git a/drivers/gpu/drm/radeon/btc_dpm.h b/drivers/gpu/drm/radeon/btc_dpm.h index 3b6f12b7760b..1a15e0e41950 100644 --- a/drivers/gpu/drm/radeon/btc_dpm.h +++ b/drivers/gpu/drm/radeon/btc_dpm.h @@ -46,8 +46,6 @@ void btc_adjust_clock_combinations(struct radeon_device *rdev, struct rv7xx_pl *pl); void btc_apply_voltage_dependency_rules(struct radeon_clock_voltage_dependency_table *table, u32 clock, u16 max_voltage, u16 *voltage); -void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table, - u32 *max_clock); void btc_apply_voltage_delta_rules(struct radeon_device *rdev, u16 max_vddc, u16 max_vddci, u16 *vddc, u16 *vddci); diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c index d416bb2ff48d..f5c8c0445a94 100644 --- a/drivers/gpu/drm/radeon/ci_dpm.c +++ b/drivers/gpu/drm/radeon/ci_dpm.c @@ -162,8 +162,6 @@ static const struct ci_pt_config_reg didt_config_ci[] = }; extern u8 rv770_get_memory_module_index(struct radeon_device *rdev); -extern void btc_get_max_clock_from_voltage_dependency_table(struct radeon_clock_voltage_dependency_table *table, - u32 *max_clock); extern int ni_copy_and_switch_arb_sets(struct radeon_device *rdev, u32 arb_freq_src, u32 arb_freq_dest); extern u8 si_get_ddr3_mclk_frequency_ratio(u32 memory_clock); @@ -748,7 +746,6 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev, struct radeon_clock_and_voltage_limits *max_limits; bool disable_mclk_switching; u32 sclk, mclk; - u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc; int i; if (rps->vce_active) { @@ -784,29 +781,6 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev, } } - /* limit clocks to max supported clocks based on voltage dependency tables */ - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk, - &max_sclk_vddc); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk, - &max_mclk_vddci); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk, - &max_mclk_vddc); - - for (i = 0; i < ps->performance_level_count; i++) { - if (max_sclk_vddc) { - if (ps->performance_levels[i].sclk > max_sclk_vddc) - ps->performance_levels[i].sclk = max_sclk_vddc; - } - if (max_mclk_vddci) { - if (ps->performance_levels[i].mclk > max_mclk_vddci) - ps->performance_levels[i].mclk = max_mclk_vddci; - } - if (max_mclk_vddc) { - if (ps->performance_levels[i].mclk > max_mclk_vddc) - ps->performance_levels[i].mclk = max_mclk_vddc; - } - } - /* XXX validate the min clocks required for display */ if (disable_mclk_switching) { @@ -5293,9 +5267,13 @@ int ci_dpm_init(struct radeon_device *rdev) void ci_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev, struct seq_file *m) { + struct ci_power_info *pi = ci_get_pi(rdev); + struct radeon_ps *rps = &pi->current_rps; u32 sclk = ci_get_average_sclk_freq(rdev); u32 mclk = ci_get_average_mclk_freq(rdev); + seq_printf(m, "uvd %sabled\n", pi->uvd_enabled ? "en" : "dis"); + seq_printf(m, "vce %sabled\n", rps->vce_active ? "en" : "dis"); seq_printf(m, "power level avg sclk: %u mclk: %u\n", sclk, mclk); } diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c index 3d546c606b43..377afa504d2b 100644 --- a/drivers/gpu/drm/radeon/cik.c +++ b/drivers/gpu/drm/radeon/cik.c @@ -3959,18 +3959,19 @@ bool cik_semaphore_ring_emit(struct radeon_device *rdev, * @src_offset: src GPU address * @dst_offset: dst GPU address * @num_gpu_pages: number of GPU pages to xfer - * @fence: radeon fence object + * @resv: reservation object to sync to * * Copy GPU paging using the CP DMA engine (CIK+). * Used by the radeon ttm implementation to move pages if * registered as the asic copy callback. */ -int cik_copy_cpdma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *cik_copy_cpdma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_semaphore *sem = NULL; + struct radeon_fence *fence; int ring_index = rdev->asic->copy.blit_ring_index; struct radeon_ring *ring = &rdev->ring[ring_index]; u32 size_in_bytes, cur_size_in_bytes, control; @@ -3980,7 +3981,7 @@ int cik_copy_cpdma(struct radeon_device *rdev, r = radeon_semaphore_create(rdev, &sem); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT); @@ -3989,10 +3990,10 @@ int cik_copy_cpdma(struct radeon_device *rdev, if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } - radeon_semaphore_sync_to(sem, *fence); + radeon_semaphore_sync_resv(rdev, sem, resv, false); radeon_semaphore_sync_rings(rdev, sem, ring->idx); for (i = 0; i < num_loops; i++) { @@ -4014,17 +4015,17 @@ int cik_copy_cpdma(struct radeon_device *rdev, dst_offset += cur_size_in_bytes; } - r = radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, &fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - radeon_semaphore_free(rdev, &sem, *fence); + radeon_semaphore_free(rdev, &sem, fence); - return r; + return fence; } /* @@ -4234,7 +4235,7 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev) WREG32(CP_PFP_UCODE_ADDR, 0); for (i = 0; i < fw_size; i++) WREG32(CP_PFP_UCODE_DATA, le32_to_cpup(fw_data++)); - WREG32(CP_PFP_UCODE_ADDR, 0); + WREG32(CP_PFP_UCODE_ADDR, le32_to_cpu(pfp_hdr->header.ucode_version)); /* CE */ fw_data = (const __le32 *) @@ -4243,7 +4244,7 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev) WREG32(CP_CE_UCODE_ADDR, 0); for (i = 0; i < fw_size; i++) WREG32(CP_CE_UCODE_DATA, le32_to_cpup(fw_data++)); - WREG32(CP_CE_UCODE_ADDR, 0); + WREG32(CP_CE_UCODE_ADDR, le32_to_cpu(ce_hdr->header.ucode_version)); /* ME */ fw_data = (const __be32 *) @@ -4252,7 +4253,8 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev) WREG32(CP_ME_RAM_WADDR, 0); for (i = 0; i < fw_size; i++) WREG32(CP_ME_RAM_DATA, le32_to_cpup(fw_data++)); - WREG32(CP_ME_RAM_WADDR, 0); + WREG32(CP_ME_RAM_WADDR, le32_to_cpu(me_hdr->header.ucode_version)); + WREG32(CP_ME_RAM_RADDR, le32_to_cpu(me_hdr->header.ucode_version)); } else { const __be32 *fw_data; @@ -4278,10 +4280,6 @@ static int cik_cp_gfx_load_microcode(struct radeon_device *rdev) WREG32(CP_ME_RAM_WADDR, 0); } - WREG32(CP_PFP_UCODE_ADDR, 0); - WREG32(CP_CE_UCODE_ADDR, 0); - WREG32(CP_ME_RAM_WADDR, 0); - WREG32(CP_ME_RAM_RADDR, 0); return 0; } @@ -4563,7 +4561,7 @@ static int cik_cp_compute_load_microcode(struct radeon_device *rdev) WREG32(CP_MEC_ME1_UCODE_ADDR, 0); for (i = 0; i < fw_size; i++) WREG32(CP_MEC_ME1_UCODE_DATA, le32_to_cpup(fw_data++)); - WREG32(CP_MEC_ME1_UCODE_ADDR, 0); + WREG32(CP_MEC_ME1_UCODE_ADDR, le32_to_cpu(mec_hdr->header.ucode_version)); /* MEC2 */ if (rdev->family == CHIP_KAVERI) { @@ -4577,7 +4575,7 @@ static int cik_cp_compute_load_microcode(struct radeon_device *rdev) WREG32(CP_MEC_ME2_UCODE_ADDR, 0); for (i = 0; i < fw_size; i++) WREG32(CP_MEC_ME2_UCODE_DATA, le32_to_cpup(fw_data++)); - WREG32(CP_MEC_ME2_UCODE_ADDR, 0); + WREG32(CP_MEC_ME2_UCODE_ADDR, le32_to_cpu(mec2_hdr->header.ucode_version)); } } else { const __be32 *fw_data; @@ -4689,7 +4687,7 @@ static int cik_mec_init(struct radeon_device *rdev) r = radeon_bo_create(rdev, rdev->mec.num_mec *rdev->mec.num_pipe * MEC_HPD_SIZE * 2, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_GTT, 0, NULL, + RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL, &rdev->mec.hpd_eop_obj); if (r) { dev_warn(rdev->dev, "(%d) create HDP EOP bo failed\n", r); @@ -4860,7 +4858,7 @@ static int cik_cp_compute_resume(struct radeon_device *rdev) sizeof(struct bonaire_mqd), PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT, 0, NULL, - &rdev->ring[idx].mqd_obj); + NULL, &rdev->ring[idx].mqd_obj); if (r) { dev_warn(rdev->dev, "(%d) create MQD bo failed\n", r); return r; @@ -6226,7 +6224,7 @@ static int cik_rlc_resume(struct radeon_device *rdev) WREG32(RLC_GPM_UCODE_ADDR, 0); for (i = 0; i < size; i++) WREG32(RLC_GPM_UCODE_DATA, le32_to_cpup(fw_data++)); - WREG32(RLC_GPM_UCODE_ADDR, 0); + WREG32(RLC_GPM_UCODE_ADDR, le32_to_cpu(hdr->header.ucode_version)); } else { const __be32 *fw_data; @@ -8255,8 +8253,10 @@ restart_ih: } if (queue_hotplug) schedule_work(&rdev->hotplug_work); - if (queue_reset) - schedule_work(&rdev->reset_work); + if (queue_reset) { + rdev->needs_reset = true; + wake_up_all(&rdev->fence_queue); + } if (queue_thermal) schedule_work(&rdev->pm.dpm.thermal.work); rdev->ih.rptr = rptr; diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c index c4ffa54b1e3d..c77dad1a4576 100644 --- a/drivers/gpu/drm/radeon/cik_sdma.c +++ b/drivers/gpu/drm/radeon/cik_sdma.c @@ -530,18 +530,19 @@ void cik_sdma_fini(struct radeon_device *rdev) * @src_offset: src GPU address * @dst_offset: dst GPU address * @num_gpu_pages: number of GPU pages to xfer - * @fence: radeon fence object + * @resv: reservation object to sync to * * Copy GPU paging using the DMA engine (CIK). * Used by the radeon ttm implementation to move pages if * registered as the asic copy callback. */ -int cik_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *cik_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_semaphore *sem = NULL; + struct radeon_fence *fence; int ring_index = rdev->asic->copy.dma_ring_index; struct radeon_ring *ring = &rdev->ring[ring_index]; u32 size_in_bytes, cur_size_in_bytes; @@ -551,7 +552,7 @@ int cik_copy_dma(struct radeon_device *rdev, r = radeon_semaphore_create(rdev, &sem); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT); @@ -560,10 +561,10 @@ int cik_copy_dma(struct radeon_device *rdev, if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } - radeon_semaphore_sync_to(sem, *fence); + radeon_semaphore_sync_resv(rdev, sem, resv, false); radeon_semaphore_sync_rings(rdev, sem, ring->idx); for (i = 0; i < num_loops; i++) { @@ -582,17 +583,17 @@ int cik_copy_dma(struct radeon_device *rdev, dst_offset += cur_size_in_bytes; } - r = radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, &fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - radeon_semaphore_free(rdev, &sem, *fence); + radeon_semaphore_free(rdev, &sem, fence); - return r; + return fence; } /** diff --git a/drivers/gpu/drm/radeon/dce3_1_afmt.c b/drivers/gpu/drm/radeon/dce3_1_afmt.c index 51800e340a57..950af153f30e 100644 --- a/drivers/gpu/drm/radeon/dce3_1_afmt.c +++ b/drivers/gpu/drm/radeon/dce3_1_afmt.c @@ -165,7 +165,7 @@ void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *m /* disable audio prior to setting up hw */ dig->afmt->pin = r600_audio_get_pin(rdev); - r600_audio_enable(rdev, dig->afmt->pin, false); + r600_audio_enable(rdev, dig->afmt->pin, 0); r600_audio_set_dto(encoder, mode->clock); @@ -240,5 +240,5 @@ void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *m r600_hdmi_audio_workaround(encoder); /* enable audio after to setting up hw */ - r600_audio_enable(rdev, dig->afmt->pin, true); + r600_audio_enable(rdev, dig->afmt->pin, 0xf); } diff --git a/drivers/gpu/drm/radeon/dce6_afmt.c b/drivers/gpu/drm/radeon/dce6_afmt.c index ab29f953a767..c0bbf68dbc27 100644 --- a/drivers/gpu/drm/radeon/dce6_afmt.c +++ b/drivers/gpu/drm/radeon/dce6_afmt.c @@ -284,13 +284,13 @@ static int dce6_audio_chipset_supported(struct radeon_device *rdev) void dce6_audio_enable(struct radeon_device *rdev, struct r600_audio_pin *pin, - bool enable) + u8 enable_mask) { if (!pin) return; - WREG32_ENDPOINT(pin->offset, AZ_F0_CODEC_PIN_CONTROL_HOTPLUG_CONTROL, - enable ? AUDIO_ENABLED : 0); + WREG32_ENDPOINT(pin->offset, AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL, + enable_mask ? AUDIO_ENABLED : 0); } static const u32 pin_offsets[7] = diff --git a/drivers/gpu/drm/radeon/drm_buffer.c b/drivers/gpu/drm/radeon/drm_buffer.c new file mode 100644 index 000000000000..f4e0f3a3d7b1 --- /dev/null +++ b/drivers/gpu/drm/radeon/drm_buffer.c @@ -0,0 +1,177 @@ +/************************************************************************** + * + * Copyright 2010 Pauli Nieminen. + * All Rights Reserved. + * + * 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, sub license, 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 NON-INFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDERS, AUTHORS 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. + * + * + **************************************************************************/ +/* + * Multipart buffer for coping data which is larger than the page size. + * + * Authors: + * Pauli Nieminen <suokkos-at-gmail-dot-com> + */ + +#include <linux/export.h> +#include "drm_buffer.h" + +/** + * Allocate the drm buffer object. + * + * buf: Pointer to a pointer where the object is stored. + * size: The number of bytes to allocate. + */ +int drm_buffer_alloc(struct drm_buffer **buf, int size) +{ + int nr_pages = size / PAGE_SIZE + 1; + int idx; + + /* Allocating pointer table to end of structure makes drm_buffer + * variable sized */ + *buf = kzalloc(sizeof(struct drm_buffer) + nr_pages*sizeof(char *), + GFP_KERNEL); + + if (*buf == NULL) { + DRM_ERROR("Failed to allocate drm buffer object to hold" + " %d bytes in %d pages.\n", + size, nr_pages); + return -ENOMEM; + } + + (*buf)->size = size; + + for (idx = 0; idx < nr_pages; ++idx) { + + (*buf)->data[idx] = + kmalloc(min(PAGE_SIZE, size - idx * PAGE_SIZE), + GFP_KERNEL); + + + if ((*buf)->data[idx] == NULL) { + DRM_ERROR("Failed to allocate %dth page for drm" + " buffer with %d bytes and %d pages.\n", + idx + 1, size, nr_pages); + goto error_out; + } + + } + + return 0; + +error_out: + + for (; idx >= 0; --idx) + kfree((*buf)->data[idx]); + + kfree(*buf); + return -ENOMEM; +} + +/** + * Copy the user data to the begin of the buffer and reset the processing + * iterator. + * + * user_data: A pointer the data that is copied to the buffer. + * size: The Number of bytes to copy. + */ +int drm_buffer_copy_from_user(struct drm_buffer *buf, + void __user *user_data, int size) +{ + int nr_pages = size / PAGE_SIZE + 1; + int idx; + + if (size > buf->size) { + DRM_ERROR("Requesting to copy %d bytes to a drm buffer with" + " %d bytes space\n", + size, buf->size); + return -EFAULT; + } + + for (idx = 0; idx < nr_pages; ++idx) { + + if (copy_from_user(buf->data[idx], + user_data + idx * PAGE_SIZE, + min(PAGE_SIZE, size - idx * PAGE_SIZE))) { + DRM_ERROR("Failed to copy user data (%p) to drm buffer" + " (%p) %dth page.\n", + user_data, buf, idx); + return -EFAULT; + + } + } + buf->iterator = 0; + return 0; +} + +/** + * Free the drm buffer object + */ +void drm_buffer_free(struct drm_buffer *buf) +{ + + if (buf != NULL) { + + int nr_pages = buf->size / PAGE_SIZE + 1; + int idx; + for (idx = 0; idx < nr_pages; ++idx) + kfree(buf->data[idx]); + + kfree(buf); + } +} + +/** + * Read an object from buffer that may be split to multiple parts. If object + * is not split function just returns the pointer to object in buffer. But in + * case of split object data is copied to given stack object that is suplied + * by caller. + * + * The processing location of the buffer is also advanced to the next byte + * after the object. + * + * objsize: The size of the objet in bytes. + * stack_obj: A pointer to a memory location where object can be copied. + */ +void *drm_buffer_read_object(struct drm_buffer *buf, + int objsize, void *stack_obj) +{ + int idx = drm_buffer_index(buf); + int page = drm_buffer_page(buf); + void *obj = NULL; + + if (idx + objsize <= PAGE_SIZE) { + obj = &buf->data[page][idx]; + } else { + /* The object is split which forces copy to temporary object.*/ + int beginsz = PAGE_SIZE - idx; + memcpy(stack_obj, &buf->data[page][idx], beginsz); + + memcpy(stack_obj + beginsz, &buf->data[page + 1][0], + objsize - beginsz); + + obj = stack_obj; + } + + drm_buffer_advance(buf, objsize); + return obj; +} diff --git a/drivers/gpu/drm/radeon/drm_buffer.h b/drivers/gpu/drm/radeon/drm_buffer.h new file mode 100644 index 000000000000..c80d3a340b94 --- /dev/null +++ b/drivers/gpu/drm/radeon/drm_buffer.h @@ -0,0 +1,148 @@ +/************************************************************************** + * + * Copyright 2010 Pauli Nieminen. + * All Rights Reserved. + * + * 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, sub license, 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 NON-INFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDERS, AUTHORS 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. + * + * + **************************************************************************/ +/* + * Multipart buffer for coping data which is larger than the page size. + * + * Authors: + * Pauli Nieminen <suokkos-at-gmail-dot-com> + */ + +#ifndef _DRM_BUFFER_H_ +#define _DRM_BUFFER_H_ + +#include <drm/drmP.h> + +struct drm_buffer { + int iterator; + int size; + char *data[]; +}; + + +/** + * Return the index of page that buffer is currently pointing at. + */ +static inline int drm_buffer_page(struct drm_buffer *buf) +{ + return buf->iterator / PAGE_SIZE; +} +/** + * Return the index of the current byte in the page + */ +static inline int drm_buffer_index(struct drm_buffer *buf) +{ + return buf->iterator & (PAGE_SIZE - 1); +} +/** + * Return number of bytes that is left to process + */ +static inline int drm_buffer_unprocessed(struct drm_buffer *buf) +{ + return buf->size - buf->iterator; +} + +/** + * Advance the buffer iterator number of bytes that is given. + */ +static inline void drm_buffer_advance(struct drm_buffer *buf, int bytes) +{ + buf->iterator += bytes; +} + +/** + * Allocate the drm buffer object. + * + * buf: A pointer to a pointer where the object is stored. + * size: The number of bytes to allocate. + */ +extern int drm_buffer_alloc(struct drm_buffer **buf, int size); + +/** + * Copy the user data to the begin of the buffer and reset the processing + * iterator. + * + * user_data: A pointer the data that is copied to the buffer. + * size: The Number of bytes to copy. + */ +extern int drm_buffer_copy_from_user(struct drm_buffer *buf, + void __user *user_data, int size); + +/** + * Free the drm buffer object + */ +extern void drm_buffer_free(struct drm_buffer *buf); + +/** + * Read an object from buffer that may be split to multiple parts. If object + * is not split function just returns the pointer to object in buffer. But in + * case of split object data is copied to given stack object that is suplied + * by caller. + * + * The processing location of the buffer is also advanced to the next byte + * after the object. + * + * objsize: The size of the objet in bytes. + * stack_obj: A pointer to a memory location where object can be copied. + */ +extern void *drm_buffer_read_object(struct drm_buffer *buf, + int objsize, void *stack_obj); + +/** + * Returns the pointer to the dword which is offset number of elements from the + * current processing location. + * + * Caller must make sure that dword is not split in the buffer. This + * requirement is easily met if all the sizes of objects in buffer are + * multiples of dword and PAGE_SIZE is multiple dword. + * + * Call to this function doesn't change the processing location. + * + * offset: The index of the dword relative to the internat iterator. + */ +static inline void *drm_buffer_pointer_to_dword(struct drm_buffer *buffer, + int offset) +{ + int iter = buffer->iterator + offset * 4; + return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)]; +} +/** + * Returns the pointer to the dword which is offset number of elements from + * the current processing location. + * + * Call to this function doesn't change the processing location. + * + * offset: The index of the byte relative to the internat iterator. + */ +static inline void *drm_buffer_pointer_to_byte(struct drm_buffer *buffer, + int offset) +{ + int iter = buffer->iterator + offset; + return &buffer->data[iter / PAGE_SIZE][iter & (PAGE_SIZE - 1)]; +} + +#endif diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index e50807c29f69..a31f1ca40c6a 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -22,7 +22,6 @@ * Authors: Alex Deucher */ #include <linux/firmware.h> -#include <linux/platform_device.h> #include <linux/slab.h> #include <drm/drmP.h> #include "radeon.h" @@ -4023,7 +4022,7 @@ int sumo_rlc_init(struct radeon_device *rdev) if (rdev->rlc.save_restore_obj == NULL) { r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, 0, NULL, - &rdev->rlc.save_restore_obj); + NULL, &rdev->rlc.save_restore_obj); if (r) { dev_warn(rdev->dev, "(%d) create RLC sr bo failed\n", r); return r; @@ -4102,7 +4101,7 @@ int sumo_rlc_init(struct radeon_device *rdev) if (rdev->rlc.clear_state_obj == NULL) { r = radeon_bo_create(rdev, dws * 4, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, 0, NULL, - &rdev->rlc.clear_state_obj); + NULL, &rdev->rlc.clear_state_obj); if (r) { dev_warn(rdev->dev, "(%d) create RLC c bo failed\n", r); sumo_rlc_fini(rdev); @@ -4179,7 +4178,7 @@ int sumo_rlc_init(struct radeon_device *rdev) r = radeon_bo_create(rdev, rdev->rlc.cp_table_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, 0, NULL, - &rdev->rlc.cp_table_obj); + NULL, &rdev->rlc.cp_table_obj); if (r) { dev_warn(rdev->dev, "(%d) create RLC cp table bo failed\n", r); sumo_rlc_fini(rdev); diff --git a/drivers/gpu/drm/radeon/evergreen_dma.c b/drivers/gpu/drm/radeon/evergreen_dma.c index afaba388c36d..66bcfadeedd1 100644 --- a/drivers/gpu/drm/radeon/evergreen_dma.c +++ b/drivers/gpu/drm/radeon/evergreen_dma.c @@ -104,12 +104,14 @@ void evergreen_dma_ring_ib_execute(struct radeon_device *rdev, * Used by the radeon ttm implementation to move pages if * registered as the asic copy callback. */ -int evergreen_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *evergreen_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_semaphore *sem = NULL; + struct radeon_fence *fence; int ring_index = rdev->asic->copy.dma_ring_index; struct radeon_ring *ring = &rdev->ring[ring_index]; u32 size_in_dw, cur_size_in_dw; @@ -119,7 +121,7 @@ int evergreen_copy_dma(struct radeon_device *rdev, r = radeon_semaphore_create(rdev, &sem); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4; @@ -128,10 +130,10 @@ int evergreen_copy_dma(struct radeon_device *rdev, if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } - radeon_semaphore_sync_to(sem, *fence); + radeon_semaphore_sync_resv(rdev, sem, resv, false); radeon_semaphore_sync_rings(rdev, sem, ring->idx); for (i = 0; i < num_loops; i++) { @@ -148,17 +150,17 @@ int evergreen_copy_dma(struct radeon_device *rdev, dst_offset += cur_size_in_dw * 4; } - r = radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, &fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - radeon_semaphore_free(rdev, &sem, *fence); + radeon_semaphore_free(rdev, &sem, fence); - return r; + return fence; } /** diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c index 278c7a139d74..2514d659b1ba 100644 --- a/drivers/gpu/drm/radeon/evergreen_hdmi.c +++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c @@ -38,6 +38,37 @@ extern void dce6_afmt_select_pin(struct drm_encoder *encoder); extern void dce6_afmt_write_latency_fields(struct drm_encoder *encoder, struct drm_display_mode *mode); +/* enable the audio stream */ +static void dce4_audio_enable(struct radeon_device *rdev, + struct r600_audio_pin *pin, + u8 enable_mask) +{ + u32 tmp = RREG32(AZ_HOT_PLUG_CONTROL); + + if (!pin) + return; + + if (enable_mask) { + tmp |= AUDIO_ENABLED; + if (enable_mask & 1) + tmp |= PIN0_AUDIO_ENABLED; + if (enable_mask & 2) + tmp |= PIN1_AUDIO_ENABLED; + if (enable_mask & 4) + tmp |= PIN2_AUDIO_ENABLED; + if (enable_mask & 8) + tmp |= PIN3_AUDIO_ENABLED; + } else { + tmp &= ~(AUDIO_ENABLED | + PIN0_AUDIO_ENABLED | + PIN1_AUDIO_ENABLED | + PIN2_AUDIO_ENABLED | + PIN3_AUDIO_ENABLED); + } + + WREG32(AZ_HOT_PLUG_CONTROL, tmp); +} + /* * update the N and CTS parameters for a given pixel clock rate */ @@ -318,10 +349,10 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode /* disable audio prior to setting up hw */ if (ASIC_IS_DCE6(rdev)) { dig->afmt->pin = dce6_audio_get_pin(rdev); - dce6_audio_enable(rdev, dig->afmt->pin, false); + dce6_audio_enable(rdev, dig->afmt->pin, 0); } else { dig->afmt->pin = r600_audio_get_pin(rdev); - r600_audio_enable(rdev, dig->afmt->pin, false); + dce4_audio_enable(rdev, dig->afmt->pin, 0); } evergreen_audio_set_dto(encoder, mode->clock); @@ -463,13 +494,15 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode /* enable audio after to setting up hw */ if (ASIC_IS_DCE6(rdev)) - dce6_audio_enable(rdev, dig->afmt->pin, true); + dce6_audio_enable(rdev, dig->afmt->pin, 1); else - r600_audio_enable(rdev, dig->afmt->pin, true); + dce4_audio_enable(rdev, dig->afmt->pin, 0xf); } void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable) { + struct drm_device *dev = encoder->dev; + struct radeon_device *rdev = dev->dev_private; struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv; @@ -482,6 +515,14 @@ void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable) if (!enable && !dig->afmt->enabled) return; + if (!enable && dig->afmt->pin) { + if (ASIC_IS_DCE6(rdev)) + dce6_audio_enable(rdev, dig->afmt->pin, 0); + else + dce4_audio_enable(rdev, dig->afmt->pin, 0); + dig->afmt->pin = NULL; + } + dig->afmt->enabled = enable; DRM_DEBUG("%sabling HDMI interface @ 0x%04X for encoder 0x%x\n", diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c index 67cb472d188c..1dd976f447fa 100644 --- a/drivers/gpu/drm/radeon/kv_dpm.c +++ b/drivers/gpu/drm/radeon/kv_dpm.c @@ -2787,6 +2787,8 @@ void kv_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev, tmp = (RREG32_SMC(SMU_VOLTAGE_STATUS) & SMU_VOLTAGE_CURRENT_LEVEL_MASK) >> SMU_VOLTAGE_CURRENT_LEVEL_SHIFT; vddc = kv_convert_8bit_index_to_voltage(rdev, (u16)tmp); + seq_printf(m, "uvd %sabled\n", pi->uvd_power_gated ? "dis" : "en"); + seq_printf(m, "vce %sabled\n", pi->vce_power_gated ? "dis" : "en"); seq_printf(m, "power level %d sclk: %u vddc: %u\n", current_index, sclk, vddc); } diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c index 01fc4888e6fe..715b181c6243 100644 --- a/drivers/gpu/drm/radeon/ni_dpm.c +++ b/drivers/gpu/drm/radeon/ni_dpm.c @@ -789,7 +789,6 @@ static void ni_apply_state_adjust_rules(struct radeon_device *rdev, bool disable_mclk_switching; u32 mclk; u16 vddci; - u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc; int i; if ((rdev->pm.dpm.new_active_crtc_count > 1) || @@ -816,29 +815,6 @@ static void ni_apply_state_adjust_rules(struct radeon_device *rdev, } } - /* limit clocks to max supported clocks based on voltage dependency tables */ - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk, - &max_sclk_vddc); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk, - &max_mclk_vddci); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk, - &max_mclk_vddc); - - for (i = 0; i < ps->performance_level_count; i++) { - if (max_sclk_vddc) { - if (ps->performance_levels[i].sclk > max_sclk_vddc) - ps->performance_levels[i].sclk = max_sclk_vddc; - } - if (max_mclk_vddci) { - if (ps->performance_levels[i].mclk > max_mclk_vddci) - ps->performance_levels[i].mclk = max_mclk_vddci; - } - if (max_mclk_vddc) { - if (ps->performance_levels[i].mclk > max_mclk_vddc) - ps->performance_levels[i].mclk = max_mclk_vddc; - } - } - /* XXX validate the min clocks required for display */ /* adjust low state */ diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index b0098e792e62..10f8be0ee173 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -869,13 +869,14 @@ bool r100_semaphore_ring_emit(struct radeon_device *rdev, return false; } -int r100_copy_blit(struct radeon_device *rdev, - uint64_t src_offset, - uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *r100_copy_blit(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + struct radeon_fence *fence; uint32_t cur_pages; uint32_t stride_bytes = RADEON_GPU_PAGE_SIZE; uint32_t pitch; @@ -896,7 +897,7 @@ int r100_copy_blit(struct radeon_device *rdev, r = radeon_ring_lock(rdev, ring, ndw); if (r) { DRM_ERROR("radeon: moving bo (%d) asking for %u dw.\n", r, ndw); - return -EINVAL; + return ERR_PTR(-EINVAL); } while (num_gpu_pages > 0) { cur_pages = num_gpu_pages; @@ -936,11 +937,13 @@ int r100_copy_blit(struct radeon_device *rdev, RADEON_WAIT_2D_IDLECLEAN | RADEON_WAIT_HOST_IDLECLEAN | RADEON_WAIT_DMA_GUI_IDLE); - if (fence) { - r = radeon_fence_emit(rdev, fence, RADEON_RING_TYPE_GFX_INDEX); + r = radeon_fence_emit(rdev, &fence, RADEON_RING_TYPE_GFX_INDEX); + if (r) { + radeon_ring_unlock_undo(rdev, ring); + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - return r; + return fence; } static int r100_cp_wait_for_idle(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/r200.c b/drivers/gpu/drm/radeon/r200.c index 67780374a652..732d4938aab7 100644 --- a/drivers/gpu/drm/radeon/r200.c +++ b/drivers/gpu/drm/radeon/r200.c @@ -80,13 +80,14 @@ static int r200_get_vtx_size_0(uint32_t vtx_fmt_0) return vtx_size; } -int r200_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, - uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *r200_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; + struct radeon_fence *fence; uint32_t size; uint32_t cur_size; int i, num_loops; @@ -98,7 +99,7 @@ int r200_copy_dma(struct radeon_device *rdev, r = radeon_ring_lock(rdev, ring, num_loops * 4 + 64); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } /* Must wait for 2D idle & clean before DMA or hangs might happen */ radeon_ring_write(ring, PACKET0(RADEON_WAIT_UNTIL, 0)); @@ -118,11 +119,13 @@ int r200_copy_dma(struct radeon_device *rdev, } radeon_ring_write(ring, PACKET0(RADEON_WAIT_UNTIL, 0)); radeon_ring_write(ring, RADEON_WAIT_DMA_GUI_IDLE); - if (fence) { - r = radeon_fence_emit(rdev, fence, RADEON_RING_TYPE_GFX_INDEX); + r = radeon_fence_emit(rdev, &fence, RADEON_RING_TYPE_GFX_INDEX); + if (r) { + radeon_ring_unlock_undo(rdev, ring); + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - return r; + return fence; } diff --git a/drivers/gpu/drm/radeon/r300_cmdbuf.c b/drivers/gpu/drm/radeon/r300_cmdbuf.c index 84b1d5367a11..9418e388b045 100644 --- a/drivers/gpu/drm/radeon/r300_cmdbuf.c +++ b/drivers/gpu/drm/radeon/r300_cmdbuf.c @@ -34,10 +34,10 @@ */ #include <drm/drmP.h> -#include <drm/drm_buffer.h> #include <drm/radeon_drm.h> #include "radeon_drv.h" #include "r300_reg.h" +#include "drm_buffer.h" #include <asm/unaligned.h> diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index ea5c9af722ef..56b02927cd3d 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -122,6 +122,94 @@ u32 r600_get_xclk(struct radeon_device *rdev) int r600_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk) { + unsigned fb_div = 0, ref_div, vclk_div = 0, dclk_div = 0; + int r; + + /* bypass vclk and dclk with bclk */ + WREG32_P(CG_UPLL_FUNC_CNTL_2, + VCLK_SRC_SEL(1) | DCLK_SRC_SEL(1), + ~(VCLK_SRC_SEL_MASK | DCLK_SRC_SEL_MASK)); + + /* assert BYPASS_EN, deassert UPLL_RESET, UPLL_SLEEP and UPLL_CTLREQ */ + WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_BYPASS_EN_MASK, ~( + UPLL_RESET_MASK | UPLL_SLEEP_MASK | UPLL_CTLREQ_MASK)); + + if (rdev->family >= CHIP_RS780) + WREG32_P(GFX_MACRO_BYPASS_CNTL, UPLL_BYPASS_CNTL, + ~UPLL_BYPASS_CNTL); + + if (!vclk || !dclk) { + /* keep the Bypass mode, put PLL to sleep */ + WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_SLEEP_MASK, ~UPLL_SLEEP_MASK); + return 0; + } + + if (rdev->clock.spll.reference_freq == 10000) + ref_div = 34; + else + ref_div = 4; + + r = radeon_uvd_calc_upll_dividers(rdev, vclk, dclk, 50000, 160000, + ref_div + 1, 0xFFF, 2, 30, ~0, + &fb_div, &vclk_div, &dclk_div); + if (r) + return r; + + if (rdev->family >= CHIP_RV670 && rdev->family < CHIP_RS780) + fb_div >>= 1; + else + fb_div |= 1; + + r = radeon_uvd_send_upll_ctlreq(rdev, CG_UPLL_FUNC_CNTL); + if (r) + return r; + + /* assert PLL_RESET */ + WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_RESET_MASK, ~UPLL_RESET_MASK); + + /* For RS780 we have to choose ref clk */ + if (rdev->family >= CHIP_RS780) + WREG32_P(CG_UPLL_FUNC_CNTL, UPLL_REFCLK_SRC_SEL_MASK, + ~UPLL_REFCLK_SRC_SEL_MASK); + + /* set the required fb, ref and post divder values */ + WREG32_P(CG_UPLL_FUNC_CNTL, + UPLL_FB_DIV(fb_div) | + UPLL_REF_DIV(ref_div), + ~(UPLL_FB_DIV_MASK | UPLL_REF_DIV_MASK)); + WREG32_P(CG_UPLL_FUNC_CNTL_2, + UPLL_SW_HILEN(vclk_div >> 1) | + UPLL_SW_LOLEN((vclk_div >> 1) + (vclk_div & 1)) | + UPLL_SW_HILEN2(dclk_div >> 1) | + UPLL_SW_LOLEN2((dclk_div >> 1) + (dclk_div & 1)) | + UPLL_DIVEN_MASK | UPLL_DIVEN2_MASK, + ~UPLL_SW_MASK); + + /* give the PLL some time to settle */ + mdelay(15); + + /* deassert PLL_RESET */ + WREG32_P(CG_UPLL_FUNC_CNTL, 0, ~UPLL_RESET_MASK); + + mdelay(15); + + /* deassert BYPASS EN */ + WREG32_P(CG_UPLL_FUNC_CNTL, 0, ~UPLL_BYPASS_EN_MASK); + + if (rdev->family >= CHIP_RS780) + WREG32_P(GFX_MACRO_BYPASS_CNTL, 0, ~UPLL_BYPASS_CNTL); + + r = radeon_uvd_send_upll_ctlreq(rdev, CG_UPLL_FUNC_CNTL); + if (r) + return r; + + /* switch VCLK and DCLK selection */ + WREG32_P(CG_UPLL_FUNC_CNTL_2, + VCLK_SRC_SEL(2) | DCLK_SRC_SEL(2), + ~(VCLK_SRC_SEL_MASK | DCLK_SRC_SEL_MASK)); + + mdelay(100); + return 0; } @@ -992,6 +1080,8 @@ static int r600_pcie_gart_enable(struct radeon_device *rdev) WREG32(MC_VM_L1_TLB_MCB_WR_GFX_CNTL, tmp); WREG32(MC_VM_L1_TLB_MCB_RD_PDMA_CNTL, tmp); WREG32(MC_VM_L1_TLB_MCB_WR_PDMA_CNTL, tmp); + WREG32(MC_VM_L1_TLB_MCB_RD_UVD_CNTL, tmp); + WREG32(MC_VM_L1_TLB_MCB_WR_UVD_CNTL, tmp); WREG32(MC_VM_L1_TLB_MCB_RD_SEM_CNTL, tmp | ENABLE_SEMAPHORE_MODE); WREG32(MC_VM_L1_TLB_MCB_WR_SEM_CNTL, tmp | ENABLE_SEMAPHORE_MODE); WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, rdev->mc.gtt_start >> 12); @@ -1042,6 +1132,8 @@ static void r600_pcie_gart_disable(struct radeon_device *rdev) WREG32(MC_VM_L1_TLB_MCB_WR_SYS_CNTL, tmp); WREG32(MC_VM_L1_TLB_MCB_RD_HDP_CNTL, tmp); WREG32(MC_VM_L1_TLB_MCB_WR_HDP_CNTL, tmp); + WREG32(MC_VM_L1_TLB_MCB_RD_UVD_CNTL, tmp); + WREG32(MC_VM_L1_TLB_MCB_WR_UVD_CNTL, tmp); radeon_gart_table_vram_unpin(rdev); } @@ -1338,7 +1430,7 @@ int r600_vram_scratch_init(struct radeon_device *rdev) if (rdev->vram_scratch.robj == NULL) { r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, - 0, NULL, &rdev->vram_scratch.robj); + 0, NULL, NULL, &rdev->vram_scratch.robj); if (r) { return r; } @@ -2792,12 +2884,13 @@ bool r600_semaphore_ring_emit(struct radeon_device *rdev, * Used by the radeon ttm implementation to move pages if * registered as the asic copy callback. */ -int r600_copy_cpdma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *r600_copy_cpdma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_semaphore *sem = NULL; + struct radeon_fence *fence; int ring_index = rdev->asic->copy.blit_ring_index; struct radeon_ring *ring = &rdev->ring[ring_index]; u32 size_in_bytes, cur_size_in_bytes, tmp; @@ -2807,7 +2900,7 @@ int r600_copy_cpdma(struct radeon_device *rdev, r = radeon_semaphore_create(rdev, &sem); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT); @@ -2816,10 +2909,10 @@ int r600_copy_cpdma(struct radeon_device *rdev, if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } - radeon_semaphore_sync_to(sem, *fence); + radeon_semaphore_sync_resv(rdev, sem, resv, false); radeon_semaphore_sync_rings(rdev, sem, ring->idx); radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); @@ -2846,17 +2939,17 @@ int r600_copy_cpdma(struct radeon_device *rdev, radeon_ring_write(ring, (WAIT_UNTIL - PACKET3_SET_CONFIG_REG_OFFSET) >> 2); radeon_ring_write(ring, WAIT_CP_DMA_IDLE_bit); - r = radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, &fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - radeon_semaphore_free(rdev, &sem, *fence); + radeon_semaphore_free(rdev, &sem, fence); - return r; + return fence; } int r600_set_surface_reg(struct radeon_device *rdev, int reg, @@ -2907,6 +3000,18 @@ static int r600_startup(struct radeon_device *rdev) return r; } + if (rdev->has_uvd) { + r = uvd_v1_0_resume(rdev); + if (!r) { + r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_UVD_INDEX); + if (r) { + dev_err(rdev->dev, "failed initializing UVD fences (%d).\n", r); + } + } + if (r) + rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0; + } + /* Enable IRQ */ if (!rdev->irq.installed) { r = radeon_irq_kms_init(rdev); @@ -2935,6 +3040,18 @@ static int r600_startup(struct radeon_device *rdev) if (r) return r; + if (rdev->has_uvd) { + ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX]; + if (ring->ring_size) { + r = radeon_ring_init(rdev, ring, ring->ring_size, 0, + RADEON_CP_PACKET2); + if (!r) + r = uvd_v1_0_init(rdev); + if (r) + DRM_ERROR("radeon: failed initializing UVD (%d).\n", r); + } + } + r = radeon_ib_pool_init(rdev); if (r) { dev_err(rdev->dev, "IB initialization failed (%d).\n", r); @@ -2994,6 +3111,10 @@ int r600_suspend(struct radeon_device *rdev) radeon_pm_suspend(rdev); r600_audio_fini(rdev); r600_cp_stop(rdev); + if (rdev->has_uvd) { + uvd_v1_0_fini(rdev); + radeon_uvd_suspend(rdev); + } r600_irq_suspend(rdev); radeon_wb_disable(rdev); r600_pcie_gart_disable(rdev); @@ -3073,6 +3194,14 @@ int r600_init(struct radeon_device *rdev) rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL; r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024); + if (rdev->has_uvd) { + r = radeon_uvd_init(rdev); + if (!r) { + rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_obj = NULL; + r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX], 4096); + } + } + rdev->ih.ring_obj = NULL; r600_ih_ring_init(rdev, 64 * 1024); @@ -3102,6 +3231,10 @@ void r600_fini(struct radeon_device *rdev) r600_audio_fini(rdev); r600_cp_fini(rdev); r600_irq_fini(rdev); + if (rdev->has_uvd) { + uvd_v1_0_fini(rdev); + radeon_uvd_fini(rdev); + } radeon_wb_fini(rdev); radeon_ib_pool_fini(rdev); radeon_irq_kms_fini(rdev); @@ -3235,7 +3368,7 @@ int r600_ih_ring_alloc(struct radeon_device *rdev) r = radeon_bo_create(rdev, rdev->ih.ring_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT, 0, - NULL, &rdev->ih.ring_obj); + NULL, NULL, &rdev->ih.ring_obj); if (r) { DRM_ERROR("radeon: failed to create ih ring buffer (%d).\n", r); return r; diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c deleted file mode 100644 index bffac10c4296..000000000000 --- a/drivers/gpu/drm/radeon/r600_audio.c +++ /dev/null @@ -1,207 +0,0 @@ -/* - * Copyright 2008 Advanced Micro Devices, Inc. - * Copyright 2008 Red Hat Inc. - * Copyright 2009 Christian König. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice shall be included in - * all copies or substantial portions of the Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR - * OTHER DEALINGS IN THE SOFTWARE. - * - * Authors: Christian König - */ -#include <drm/drmP.h> -#include "radeon.h" -#include "radeon_reg.h" -#include "radeon_asic.h" -#include "atom.h" - -/* - * check if enc_priv stores radeon_encoder_atom_dig - */ -static bool radeon_dig_encoder(struct drm_encoder *encoder) -{ - struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); - switch (radeon_encoder->encoder_id) { - case ENCODER_OBJECT_ID_INTERNAL_LVDS: - case ENCODER_OBJECT_ID_INTERNAL_TMDS1: - case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: - case ENCODER_OBJECT_ID_INTERNAL_LVTM1: - case ENCODER_OBJECT_ID_INTERNAL_DVO1: - case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1: - case ENCODER_OBJECT_ID_INTERNAL_DDI: - case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: - case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: - case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: - case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: - return true; - } - return false; -} - -/* - * check if the chipset is supported - */ -static int r600_audio_chipset_supported(struct radeon_device *rdev) -{ - return ASIC_IS_DCE2(rdev) && !ASIC_IS_NODCE(rdev); -} - -struct r600_audio_pin r600_audio_status(struct radeon_device *rdev) -{ - struct r600_audio_pin status; - uint32_t value; - - value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL); - - /* number of channels */ - status.channels = (value & 0x7) + 1; - - /* bits per sample */ - switch ((value & 0xF0) >> 4) { - case 0x0: - status.bits_per_sample = 8; - break; - case 0x1: - status.bits_per_sample = 16; - break; - case 0x2: - status.bits_per_sample = 20; - break; - case 0x3: - status.bits_per_sample = 24; - break; - case 0x4: - status.bits_per_sample = 32; - break; - default: - dev_err(rdev->dev, "Unknown bits per sample 0x%x, using 16\n", - (int)value); - status.bits_per_sample = 16; - } - - /* current sampling rate in HZ */ - if (value & 0x4000) - status.rate = 44100; - else - status.rate = 48000; - status.rate *= ((value >> 11) & 0x7) + 1; - status.rate /= ((value >> 8) & 0x7) + 1; - - value = RREG32(R600_AUDIO_STATUS_BITS); - - /* iec 60958 status bits */ - status.status_bits = value & 0xff; - - /* iec 60958 category code */ - status.category_code = (value >> 8) & 0xff; - - return status; -} - -/* - * update all hdmi interfaces with current audio parameters - */ -void r600_audio_update_hdmi(struct work_struct *work) -{ - struct radeon_device *rdev = container_of(work, struct radeon_device, - audio_work); - struct drm_device *dev = rdev->ddev; - struct r600_audio_pin audio_status = r600_audio_status(rdev); - struct drm_encoder *encoder; - bool changed = false; - - if (rdev->audio.pin[0].channels != audio_status.channels || - rdev->audio.pin[0].rate != audio_status.rate || - rdev->audio.pin[0].bits_per_sample != audio_status.bits_per_sample || - rdev->audio.pin[0].status_bits != audio_status.status_bits || - rdev->audio.pin[0].category_code != audio_status.category_code) { - rdev->audio.pin[0] = audio_status; - changed = true; - } - - list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { - if (!radeon_dig_encoder(encoder)) - continue; - if (changed || r600_hdmi_buffer_status_changed(encoder)) - r600_hdmi_update_audio_settings(encoder); - } -} - -/* enable the audio stream */ -void r600_audio_enable(struct radeon_device *rdev, - struct r600_audio_pin *pin, - bool enable) -{ - u32 value = 0; - - if (!pin) - return; - - if (ASIC_IS_DCE4(rdev)) { - if (enable) { - value |= 0x81000000; /* Required to enable audio */ - value |= 0x0e1000f0; /* fglrx sets that too */ - } - WREG32(EVERGREEN_AUDIO_ENABLE, value); - } else { - WREG32_P(R600_AUDIO_ENABLE, - enable ? 0x81000000 : 0x0, ~0x81000000); - } -} - -/* - * initialize the audio vars - */ -int r600_audio_init(struct radeon_device *rdev) -{ - if (!radeon_audio || !r600_audio_chipset_supported(rdev)) - return 0; - - rdev->audio.enabled = true; - - rdev->audio.num_pins = 1; - rdev->audio.pin[0].channels = -1; - rdev->audio.pin[0].rate = -1; - rdev->audio.pin[0].bits_per_sample = -1; - rdev->audio.pin[0].status_bits = 0; - rdev->audio.pin[0].category_code = 0; - rdev->audio.pin[0].id = 0; - /* disable audio. it will be set up later */ - r600_audio_enable(rdev, &rdev->audio.pin[0], false); - - return 0; -} - -/* - * release the audio timer - * TODO: How to do this correctly on SMP systems? - */ -void r600_audio_fini(struct radeon_device *rdev) -{ - if (!rdev->audio.enabled) - return; - - r600_audio_enable(rdev, &rdev->audio.pin[0], false); - - rdev->audio.enabled = false; -} - -struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev) -{ - /* only one pin on 6xx-NI */ - return &rdev->audio.pin[0]; -} diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c index 8c9b7e26533c..09e3f39925fa 100644 --- a/drivers/gpu/drm/radeon/r600_cp.c +++ b/drivers/gpu/drm/radeon/r600_cp.c @@ -1949,15 +1949,15 @@ int r600_do_cleanup_cp(struct drm_device *dev) #if __OS_HAS_AGP if (dev_priv->flags & RADEON_IS_AGP) { if (dev_priv->cp_ring != NULL) { - drm_core_ioremapfree(dev_priv->cp_ring, dev); + drm_legacy_ioremapfree(dev_priv->cp_ring, dev); dev_priv->cp_ring = NULL; } if (dev_priv->ring_rptr != NULL) { - drm_core_ioremapfree(dev_priv->ring_rptr, dev); + drm_legacy_ioremapfree(dev_priv->ring_rptr, dev); dev_priv->ring_rptr = NULL; } if (dev->agp_buffer_map != NULL) { - drm_core_ioremapfree(dev->agp_buffer_map, dev); + drm_legacy_ioremapfree(dev->agp_buffer_map, dev); dev->agp_buffer_map = NULL; } } else @@ -1968,7 +1968,7 @@ int r600_do_cleanup_cp(struct drm_device *dev) r600_page_table_cleanup(dev, &dev_priv->gart_info); if (dev_priv->gart_info.gart_table_location == DRM_ATI_GART_FB) { - drm_core_ioremapfree(&dev_priv->gart_info.mapping, dev); + drm_legacy_ioremapfree(&dev_priv->gart_info.mapping, dev); dev_priv->gart_info.addr = NULL; } } @@ -2052,27 +2052,27 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, dev_priv->buffers_offset = init->buffers_offset; dev_priv->gart_textures_offset = init->gart_textures_offset; - master_priv->sarea = drm_getsarea(dev); + master_priv->sarea = drm_legacy_getsarea(dev); if (!master_priv->sarea) { DRM_ERROR("could not find sarea!\n"); r600_do_cleanup_cp(dev); return -EINVAL; } - dev_priv->cp_ring = drm_core_findmap(dev, init->ring_offset); + dev_priv->cp_ring = drm_legacy_findmap(dev, init->ring_offset); if (!dev_priv->cp_ring) { DRM_ERROR("could not find cp ring region!\n"); r600_do_cleanup_cp(dev); return -EINVAL; } - dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset); + dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset); if (!dev_priv->ring_rptr) { DRM_ERROR("could not find ring read pointer!\n"); r600_do_cleanup_cp(dev); return -EINVAL; } dev->agp_buffer_token = init->buffers_offset; - dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset); + dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset); if (!dev->agp_buffer_map) { DRM_ERROR("could not find dma buffer region!\n"); r600_do_cleanup_cp(dev); @@ -2081,7 +2081,7 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, if (init->gart_textures_offset) { dev_priv->gart_textures = - drm_core_findmap(dev, init->gart_textures_offset); + drm_legacy_findmap(dev, init->gart_textures_offset); if (!dev_priv->gart_textures) { DRM_ERROR("could not find GART texture region!\n"); r600_do_cleanup_cp(dev); @@ -2092,9 +2092,9 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, #if __OS_HAS_AGP /* XXX */ if (dev_priv->flags & RADEON_IS_AGP) { - drm_core_ioremap_wc(dev_priv->cp_ring, dev); - drm_core_ioremap_wc(dev_priv->ring_rptr, dev); - drm_core_ioremap_wc(dev->agp_buffer_map, dev); + drm_legacy_ioremap_wc(dev_priv->cp_ring, dev); + drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev); + drm_legacy_ioremap_wc(dev->agp_buffer_map, dev); if (!dev_priv->cp_ring->handle || !dev_priv->ring_rptr->handle || !dev->agp_buffer_map->handle) { @@ -2235,7 +2235,7 @@ int r600_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, dev_priv->gart_info.mapping.size = dev_priv->gart_info.table_size; - drm_core_ioremap_wc(&dev_priv->gart_info.mapping, dev); + drm_legacy_ioremap_wc(&dev_priv->gart_info.mapping, dev); if (!dev_priv->gart_info.mapping.handle) { DRM_ERROR("ioremap failed.\n"); r600_do_cleanup_cp(dev); diff --git a/drivers/gpu/drm/radeon/r600_dma.c b/drivers/gpu/drm/radeon/r600_dma.c index a908daa006d2..100189ec5fa8 100644 --- a/drivers/gpu/drm/radeon/r600_dma.c +++ b/drivers/gpu/drm/radeon/r600_dma.c @@ -427,18 +427,19 @@ void r600_dma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib) * @src_offset: src GPU address * @dst_offset: dst GPU address * @num_gpu_pages: number of GPU pages to xfer - * @fence: radeon fence object + * @resv: reservation object to sync to * * Copy GPU paging using the DMA engine (r6xx). * Used by the radeon ttm implementation to move pages if * registered as the asic copy callback. */ -int r600_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *r600_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_semaphore *sem = NULL; + struct radeon_fence *fence; int ring_index = rdev->asic->copy.dma_ring_index; struct radeon_ring *ring = &rdev->ring[ring_index]; u32 size_in_dw, cur_size_in_dw; @@ -448,7 +449,7 @@ int r600_copy_dma(struct radeon_device *rdev, r = radeon_semaphore_create(rdev, &sem); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4; @@ -457,10 +458,10 @@ int r600_copy_dma(struct radeon_device *rdev, if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } - radeon_semaphore_sync_to(sem, *fence); + radeon_semaphore_sync_resv(rdev, sem, resv, false); radeon_semaphore_sync_rings(rdev, sem, ring->idx); for (i = 0; i < num_loops; i++) { @@ -477,15 +478,15 @@ int r600_copy_dma(struct radeon_device *rdev, dst_offset += cur_size_in_dw * 4; } - r = radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, &fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - radeon_semaphore_free(rdev, &sem, *fence); + radeon_semaphore_free(rdev, &sem, fence); - return r; + return fence; } diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c index 26ef8ced6f89..b90dc0eb08e6 100644 --- a/drivers/gpu/drm/radeon/r600_hdmi.c +++ b/drivers/gpu/drm/radeon/r600_hdmi.c @@ -72,6 +72,169 @@ static const struct radeon_hdmi_acr r600_hdmi_predefined_acr[] = { /* + * check if the chipset is supported + */ +static int r600_audio_chipset_supported(struct radeon_device *rdev) +{ + return ASIC_IS_DCE2(rdev) && !ASIC_IS_NODCE(rdev); +} + +static struct r600_audio_pin r600_audio_status(struct radeon_device *rdev) +{ + struct r600_audio_pin status; + uint32_t value; + + value = RREG32(R600_AUDIO_RATE_BPS_CHANNEL); + + /* number of channels */ + status.channels = (value & 0x7) + 1; + + /* bits per sample */ + switch ((value & 0xF0) >> 4) { + case 0x0: + status.bits_per_sample = 8; + break; + case 0x1: + status.bits_per_sample = 16; + break; + case 0x2: + status.bits_per_sample = 20; + break; + case 0x3: + status.bits_per_sample = 24; + break; + case 0x4: + status.bits_per_sample = 32; + break; + default: + dev_err(rdev->dev, "Unknown bits per sample 0x%x, using 16\n", + (int)value); + status.bits_per_sample = 16; + } + + /* current sampling rate in HZ */ + if (value & 0x4000) + status.rate = 44100; + else + status.rate = 48000; + status.rate *= ((value >> 11) & 0x7) + 1; + status.rate /= ((value >> 8) & 0x7) + 1; + + value = RREG32(R600_AUDIO_STATUS_BITS); + + /* iec 60958 status bits */ + status.status_bits = value & 0xff; + + /* iec 60958 category code */ + status.category_code = (value >> 8) & 0xff; + + return status; +} + +/* + * update all hdmi interfaces with current audio parameters + */ +void r600_audio_update_hdmi(struct work_struct *work) +{ + struct radeon_device *rdev = container_of(work, struct radeon_device, + audio_work); + struct drm_device *dev = rdev->ddev; + struct r600_audio_pin audio_status = r600_audio_status(rdev); + struct drm_encoder *encoder; + bool changed = false; + + if (rdev->audio.pin[0].channels != audio_status.channels || + rdev->audio.pin[0].rate != audio_status.rate || + rdev->audio.pin[0].bits_per_sample != audio_status.bits_per_sample || + rdev->audio.pin[0].status_bits != audio_status.status_bits || + rdev->audio.pin[0].category_code != audio_status.category_code) { + rdev->audio.pin[0] = audio_status; + changed = true; + } + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (!radeon_encoder_is_digital(encoder)) + continue; + if (changed || r600_hdmi_buffer_status_changed(encoder)) + r600_hdmi_update_audio_settings(encoder); + } +} + +/* enable the audio stream */ +void r600_audio_enable(struct radeon_device *rdev, + struct r600_audio_pin *pin, + u8 enable_mask) +{ + u32 tmp = RREG32(AZ_HOT_PLUG_CONTROL); + + if (!pin) + return; + + if (enable_mask) { + tmp |= AUDIO_ENABLED; + if (enable_mask & 1) + tmp |= PIN0_AUDIO_ENABLED; + if (enable_mask & 2) + tmp |= PIN1_AUDIO_ENABLED; + if (enable_mask & 4) + tmp |= PIN2_AUDIO_ENABLED; + if (enable_mask & 8) + tmp |= PIN3_AUDIO_ENABLED; + } else { + tmp &= ~(AUDIO_ENABLED | + PIN0_AUDIO_ENABLED | + PIN1_AUDIO_ENABLED | + PIN2_AUDIO_ENABLED | + PIN3_AUDIO_ENABLED); + } + + WREG32(AZ_HOT_PLUG_CONTROL, tmp); +} + +/* + * initialize the audio vars + */ +int r600_audio_init(struct radeon_device *rdev) +{ + if (!radeon_audio || !r600_audio_chipset_supported(rdev)) + return 0; + + rdev->audio.enabled = true; + + rdev->audio.num_pins = 1; + rdev->audio.pin[0].channels = -1; + rdev->audio.pin[0].rate = -1; + rdev->audio.pin[0].bits_per_sample = -1; + rdev->audio.pin[0].status_bits = 0; + rdev->audio.pin[0].category_code = 0; + rdev->audio.pin[0].id = 0; + /* disable audio. it will be set up later */ + r600_audio_enable(rdev, &rdev->audio.pin[0], 0); + + return 0; +} + +/* + * release the audio timer + * TODO: How to do this correctly on SMP systems? + */ +void r600_audio_fini(struct radeon_device *rdev) +{ + if (!rdev->audio.enabled) + return; + + r600_audio_enable(rdev, &rdev->audio.pin[0], 0); + + rdev->audio.enabled = false; +} + +struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev) +{ + /* only one pin on 6xx-NI */ + return &rdev->audio.pin[0]; +} + +/* * calculate CTS and N values if they are not found in the table */ static void r600_hdmi_calc_cts(uint32_t clock, int *CTS, int *N, int freq) @@ -357,7 +520,7 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod /* disable audio prior to setting up hw */ dig->afmt->pin = r600_audio_get_pin(rdev); - r600_audio_enable(rdev, dig->afmt->pin, false); + r600_audio_enable(rdev, dig->afmt->pin, 0xf); r600_audio_set_dto(encoder, mode->clock); @@ -443,7 +606,7 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod WREG32(HDMI0_RAMP_CONTROL3 + offset, 0x00000001); /* enable audio after to setting up hw */ - r600_audio_enable(rdev, dig->afmt->pin, true); + r600_audio_enable(rdev, dig->afmt->pin, 0xf); } /** @@ -528,6 +691,11 @@ void r600_hdmi_enable(struct drm_encoder *encoder, bool enable) if (!enable && !dig->afmt->enabled) return; + if (!enable && dig->afmt->pin) { + r600_audio_enable(rdev, dig->afmt->pin, 0); + dig->afmt->pin = NULL; + } + /* Older chipsets require setting HDMI and routing manually */ if (!ASIC_IS_DCE3(rdev)) { if (enable) diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h index 31e1052ad3e3..1e8495cca41e 100644 --- a/drivers/gpu/drm/radeon/r600d.h +++ b/drivers/gpu/drm/radeon/r600d.h @@ -323,11 +323,12 @@ #define HDP_TILING_CONFIG 0x2F3C #define HDP_DEBUG1 0x2F34 +#define MC_CONFIG 0x2000 #define MC_VM_AGP_TOP 0x2184 #define MC_VM_AGP_BOT 0x2188 #define MC_VM_AGP_BASE 0x218C #define MC_VM_FB_LOCATION 0x2180 -#define MC_VM_L1_TLB_MCD_RD_A_CNTL 0x219C +#define MC_VM_L1_TLB_MCB_RD_UVD_CNTL 0x2124 #define ENABLE_L1_TLB (1 << 0) #define ENABLE_L1_FRAGMENT_PROCESSING (1 << 1) #define ENABLE_L1_STRICT_ORDERING (1 << 2) @@ -347,12 +348,14 @@ #define EFFECTIVE_L1_QUEUE_SIZE(x) (((x) & 7) << 15) #define EFFECTIVE_L1_QUEUE_SIZE_MASK 0x00038000 #define EFFECTIVE_L1_QUEUE_SIZE_SHIFT 15 +#define MC_VM_L1_TLB_MCD_RD_A_CNTL 0x219C #define MC_VM_L1_TLB_MCD_RD_B_CNTL 0x21A0 #define MC_VM_L1_TLB_MCB_RD_GFX_CNTL 0x21FC #define MC_VM_L1_TLB_MCB_RD_HDP_CNTL 0x2204 #define MC_VM_L1_TLB_MCB_RD_PDMA_CNTL 0x2208 #define MC_VM_L1_TLB_MCB_RD_SEM_CNTL 0x220C #define MC_VM_L1_TLB_MCB_RD_SYS_CNTL 0x2200 +#define MC_VM_L1_TLB_MCB_WR_UVD_CNTL 0x212c #define MC_VM_L1_TLB_MCD_WR_A_CNTL 0x21A4 #define MC_VM_L1_TLB_MCD_WR_B_CNTL 0x21A8 #define MC_VM_L1_TLB_MCB_WR_GFX_CNTL 0x2210 @@ -366,6 +369,8 @@ #define MC_VM_SYSTEM_APERTURE_HIGH_ADDR 0x2194 #define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR 0x2198 +#define RS_DQ_RD_RET_CONF 0x2348 + #define PA_CL_ENHANCE 0x8A14 #define CLIP_VTX_REORDER_ENA (1 << 0) #define NUM_CLIP_SEQ(x) ((x) << 1) @@ -922,6 +927,23 @@ # define TARGET_LINK_SPEED_MASK (0xf << 0) # define SELECTABLE_DEEMPHASIS (1 << 6) +/* Audio */ +#define AZ_HOT_PLUG_CONTROL 0x7300 +# define AZ_FORCE_CODEC_WAKE (1 << 0) +# define JACK_DETECTION_ENABLE (1 << 4) +# define UNSOLICITED_RESPONSE_ENABLE (1 << 8) +# define CODEC_HOT_PLUG_ENABLE (1 << 12) +# define AUDIO_ENABLED (1 << 31) +/* DCE3 adds */ +# define PIN0_JACK_DETECTION_ENABLE (1 << 4) +# define PIN1_JACK_DETECTION_ENABLE (1 << 5) +# define PIN2_JACK_DETECTION_ENABLE (1 << 6) +# define PIN3_JACK_DETECTION_ENABLE (1 << 7) +# define PIN0_AUDIO_ENABLED (1 << 24) +# define PIN1_AUDIO_ENABLED (1 << 25) +# define PIN2_AUDIO_ENABLED (1 << 26) +# define PIN3_AUDIO_ENABLED (1 << 27) + /* Audio clocks DCE 2.0/3.0 */ #define AUDIO_DTO 0x7340 # define AUDIO_DTO_PHASE(x) (((x) & 0xffff) << 0) @@ -1476,6 +1498,7 @@ #define UVD_CGC_GATE 0xf4a8 #define UVD_LMI_CTRL2 0xf4f4 #define UVD_MASTINT_EN 0xf500 +#define UVD_FW_START 0xf51C #define UVD_LMI_ADDR_EXT 0xf594 #define UVD_LMI_CTRL 0xf598 #define UVD_LMI_SWAP_CNTL 0xf5b4 @@ -1488,6 +1511,13 @@ #define UVD_MPC_SET_MUX 0xf5f4 #define UVD_MPC_SET_ALU 0xf5f8 +#define UVD_VCPU_CACHE_OFFSET0 0xf608 +#define UVD_VCPU_CACHE_SIZE0 0xf60c +#define UVD_VCPU_CACHE_OFFSET1 0xf610 +#define UVD_VCPU_CACHE_SIZE1 0xf614 +#define UVD_VCPU_CACHE_OFFSET2 0xf618 +#define UVD_VCPU_CACHE_SIZE2 0xf61c + #define UVD_VCPU_CNTL 0xf660 #define UVD_SOFT_RESET 0xf680 #define RBC_SOFT_RESET (1<<0) @@ -1517,9 +1547,35 @@ #define UVD_CONTEXT_ID 0xf6f4 +/* rs780 only */ +#define GFX_MACRO_BYPASS_CNTL 0x30c0 +#define SPLL_BYPASS_CNTL (1 << 0) +#define UPLL_BYPASS_CNTL (1 << 1) + +#define CG_UPLL_FUNC_CNTL 0x7e0 +# define UPLL_RESET_MASK 0x00000001 +# define UPLL_SLEEP_MASK 0x00000002 +# define UPLL_BYPASS_EN_MASK 0x00000004 # define UPLL_CTLREQ_MASK 0x00000008 +# define UPLL_FB_DIV(x) ((x) << 4) +# define UPLL_FB_DIV_MASK 0x0000FFF0 +# define UPLL_REF_DIV(x) ((x) << 16) +# define UPLL_REF_DIV_MASK 0x003F0000 +# define UPLL_REFCLK_SRC_SEL_MASK 0x20000000 # define UPLL_CTLACK_MASK 0x40000000 # define UPLL_CTLACK2_MASK 0x80000000 +#define CG_UPLL_FUNC_CNTL_2 0x7e4 +# define UPLL_SW_HILEN(x) ((x) << 0) +# define UPLL_SW_LOLEN(x) ((x) << 4) +# define UPLL_SW_HILEN2(x) ((x) << 8) +# define UPLL_SW_LOLEN2(x) ((x) << 12) +# define UPLL_DIVEN_MASK 0x00010000 +# define UPLL_DIVEN2_MASK 0x00020000 +# define UPLL_SW_MASK 0x0003FFFF +# define VCLK_SRC_SEL(x) ((x) << 20) +# define VCLK_SRC_SEL_MASK 0x01F00000 +# define DCLK_SRC_SEL(x) ((x) << 25) +# define DCLK_SRC_SEL_MASK 0x3E000000 /* * PM4 diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 3247bfd14410..f7c4b226a284 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -65,6 +65,8 @@ #include <linux/list.h> #include <linux/kref.h> #include <linux/interval_tree.h> +#include <linux/hashtable.h> +#include <linux/fence.h> #include <ttm/ttm_bo_api.h> #include <ttm/ttm_bo_driver.h> @@ -72,6 +74,8 @@ #include <ttm/ttm_module.h> #include <ttm/ttm_execbuf_util.h> +#include <drm/drm_gem.h> + #include "radeon_family.h" #include "radeon_mode.h" #include "radeon_reg.h" @@ -120,9 +124,6 @@ extern int radeon_backlight; #define RADEONFB_CONN_LIMIT 4 #define RADEON_BIOS_NUM_SCRATCH 8 -/* fence seq are set to this number when signaled */ -#define RADEON_FENCE_SIGNALED_SEQ 0LL - /* internal ring indices */ /* r1xx+ has gfx CP ring */ #define RADEON_RING_TYPE_GFX_INDEX 0 @@ -350,28 +351,32 @@ extern void evergreen_tiling_fields(unsigned tiling_flags, unsigned *bankw, * Fences. */ struct radeon_fence_driver { + struct radeon_device *rdev; uint32_t scratch_reg; uint64_t gpu_addr; volatile uint32_t *cpu_addr; /* sync_seq is protected by ring emission lock */ uint64_t sync_seq[RADEON_NUM_RINGS]; atomic64_t last_seq; - bool initialized; + bool initialized, delayed_irq; + struct delayed_work lockup_work; }; struct radeon_fence { + struct fence base; + struct radeon_device *rdev; - struct kref kref; - /* protected by radeon_fence.lock */ uint64_t seq; /* RB, DMA, etc. */ unsigned ring; + + wait_queue_t fence_wake; }; int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring); int radeon_fence_driver_init(struct radeon_device *rdev); void radeon_fence_driver_fini(struct radeon_device *rdev); -void radeon_fence_driver_force_completion(struct radeon_device *rdev); +void radeon_fence_driver_force_completion(struct radeon_device *rdev, int ring); int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence **fence, int ring); void radeon_fence_process(struct radeon_device *rdev, int ring); bool radeon_fence_signaled(struct radeon_fence *fence); @@ -469,7 +474,7 @@ struct radeon_bo { struct list_head list; /* Protected by tbo.reserved */ u32 initial_domain; - u32 placements[3]; + struct ttm_place placements[3]; struct ttm_placement placement; struct ttm_buffer_object tbo; struct ttm_bo_kmap_obj kmap; @@ -489,6 +494,9 @@ struct radeon_bo { struct ttm_bo_kmap_obj dma_buf_vmap; pid_t pid; + + struct radeon_mn *mn; + struct interval_tree_node mn_it; }; #define gem_to_radeon_bo(gobj) container_of((gobj), struct radeon_bo, gem_base) @@ -580,8 +588,12 @@ bool radeon_semaphore_emit_signal(struct radeon_device *rdev, int ring, struct radeon_semaphore *semaphore); bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ring, struct radeon_semaphore *semaphore); -void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore, - struct radeon_fence *fence); +void radeon_semaphore_sync_fence(struct radeon_semaphore *semaphore, + struct radeon_fence *fence); +int radeon_semaphore_sync_resv(struct radeon_device *rdev, + struct radeon_semaphore *semaphore, + struct reservation_object *resv, + bool shared); int radeon_semaphore_sync_rings(struct radeon_device *rdev, struct radeon_semaphore *semaphore, int waiting_ring); @@ -702,7 +714,7 @@ struct radeon_flip_work { uint64_t base; struct drm_pending_vblank_event *event; struct radeon_bo *old_rbo; - struct radeon_fence *fence; + struct fence *fence; }; struct r500_irq_stat_regs { @@ -780,6 +792,7 @@ struct radeon_irq { int radeon_irq_kms_init(struct radeon_device *rdev); void radeon_irq_kms_fini(struct radeon_device *rdev); void radeon_irq_kms_sw_irq_get(struct radeon_device *rdev, int ring); +bool radeon_irq_kms_sw_irq_get_delayed(struct radeon_device *rdev, int ring); void radeon_irq_kms_sw_irq_put(struct radeon_device *rdev, int ring); void radeon_irq_kms_pflip_irq_get(struct radeon_device *rdev, int crtc); void radeon_irq_kms_pflip_irq_put(struct radeon_device *rdev, int crtc); @@ -1642,7 +1655,8 @@ int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring, uint32_t handle, struct radeon_fence **fence); int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring, uint32_t handle, struct radeon_fence **fence); -void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo); +void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo, + uint32_t allowed_domains); void radeon_uvd_free_handles(struct radeon_device *rdev, struct drm_file *filp); int radeon_uvd_cs_parse(struct radeon_cs_parser *parser); @@ -1731,6 +1745,11 @@ void radeon_test_ring_sync(struct radeon_device *rdev, struct radeon_ring *cpB); void radeon_test_syncing(struct radeon_device *rdev); +/* + * MMU Notifier + */ +int radeon_mn_register(struct radeon_bo *bo, unsigned long addr); +void radeon_mn_unregister(struct radeon_bo *bo); /* * Debugfs @@ -1845,24 +1864,24 @@ struct radeon_asic { } display; /* copy functions for bo handling */ struct { - int (*blit)(struct radeon_device *rdev, - uint64_t src_offset, - uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); + struct radeon_fence *(*blit)(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); u32 blit_ring_index; - int (*dma)(struct radeon_device *rdev, - uint64_t src_offset, - uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); + struct radeon_fence *(*dma)(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); u32 dma_ring_index; /* method used for bo copy */ - int (*copy)(struct radeon_device *rdev, - uint64_t src_offset, - uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); + struct radeon_fence *(*copy)(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); /* ring used for bo copies */ u32 copy_ring_index; } copy; @@ -2144,6 +2163,8 @@ int radeon_gem_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp); int radeon_gem_create_ioctl(struct drm_device *dev, void *data, struct drm_file *filp); +int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp); int radeon_gem_pin_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); int radeon_gem_unpin_ioctl(struct drm_device *dev, void *data, @@ -2300,6 +2321,7 @@ struct radeon_device { struct radeon_mman mman; struct radeon_fence_driver fence_drv[RADEON_NUM_RINGS]; wait_queue_head_t fence_queue; + unsigned fence_context; struct mutex ring_lock; struct radeon_ring ring[RADEON_NUM_RINGS]; bool ib_pool_ready; @@ -2318,7 +2340,7 @@ struct radeon_device { bool need_dma32; bool accel_working; bool fastfb_working; /* IGP feature*/ - bool needs_reset; + bool needs_reset, in_reset; struct radeon_surface_reg surface_regs[RADEON_GEM_MAX_SURFACES]; const struct firmware *me_fw; /* all family ME firmware */ const struct firmware *pfp_fw; /* r6/700 PFP firmware */ @@ -2339,7 +2361,6 @@ struct radeon_device { struct radeon_mec mec; struct work_struct hotplug_work; struct work_struct audio_work; - struct work_struct reset_work; int num_crtc; /* number of crtcs */ struct mutex dc_hw_i2c_mutex; /* display controller hw i2c mutex */ bool has_uvd; @@ -2376,6 +2397,9 @@ struct radeon_device { /* tracking pinned memory */ u64 vram_pin_size; u64 gart_pin_size; + + struct mutex mn_lock; + DECLARE_HASHTABLE(mn_hash, 7); }; bool radeon_is_px(struct drm_device *dev); @@ -2431,7 +2455,17 @@ void cik_mm_wdoorbell(struct radeon_device *rdev, u32 index, u32 v); /* * Cast helper */ -#define to_radeon_fence(p) ((struct radeon_fence *)(p)) +extern const struct fence_ops radeon_fence_ops; + +static inline struct radeon_fence *to_radeon_fence(struct fence *f) +{ + struct radeon_fence *__f = container_of(f, struct radeon_fence, base); + + if (__f->base.ops == &radeon_fence_ops) + return __f; + + return NULL; +} /* * Registers read & write functions. @@ -2751,18 +2785,25 @@ void radeon_atombios_fini(struct radeon_device *rdev); /* * RING helpers. */ -#if DRM_DEBUG_CODE == 0 + +/** + * radeon_ring_write - write a value to the ring + * + * @ring: radeon_ring structure holding ring information + * @v: dword (dw) value to write + * + * Write a value to the requested ring buffer (all asics). + */ static inline void radeon_ring_write(struct radeon_ring *ring, uint32_t v) { + if (ring->count_dw <= 0) + DRM_ERROR("radeon: writing more dwords to the ring than expected!\n"); + ring->ring[ring->wptr++] = v; ring->wptr &= ring->ptr_mask; ring->count_dw--; ring->ring_free_dw--; } -#else -/* With debugging this is just too big to inline */ -void radeon_ring_write(struct radeon_ring *ring, uint32_t v); -#endif /* * ASICs macro. @@ -2801,9 +2842,9 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v); #define radeon_hdmi_setmode(rdev, e, m) (rdev)->asic->display.hdmi_setmode((e), (m)) #define radeon_fence_ring_emit(rdev, r, fence) (rdev)->asic->ring[(r)]->emit_fence((rdev), (fence)) #define radeon_semaphore_ring_emit(rdev, r, cp, semaphore, emit_wait) (rdev)->asic->ring[(r)]->emit_semaphore((rdev), (cp), (semaphore), (emit_wait)) -#define radeon_copy_blit(rdev, s, d, np, f) (rdev)->asic->copy.blit((rdev), (s), (d), (np), (f)) -#define radeon_copy_dma(rdev, s, d, np, f) (rdev)->asic->copy.dma((rdev), (s), (d), (np), (f)) -#define radeon_copy(rdev, s, d, np, f) (rdev)->asic->copy.copy((rdev), (s), (d), (np), (f)) +#define radeon_copy_blit(rdev, s, d, np, resv) (rdev)->asic->copy.blit((rdev), (s), (d), (np), (resv)) +#define radeon_copy_dma(rdev, s, d, np, resv) (rdev)->asic->copy.dma((rdev), (s), (d), (np), (resv)) +#define radeon_copy(rdev, s, d, np, resv) (rdev)->asic->copy.copy((rdev), (s), (d), (np), (resv)) #define radeon_copy_blit_ring_index(rdev) (rdev)->asic->copy.blit_ring_index #define radeon_copy_dma_ring_index(rdev) (rdev)->asic->copy.dma_ring_index #define radeon_copy_ring_index(rdev) (rdev)->asic->copy.copy_ring_index @@ -2877,6 +2918,10 @@ extern void radeon_legacy_set_clock_gating(struct radeon_device *rdev, int enabl extern void radeon_atom_set_clock_gating(struct radeon_device *rdev, int enable); extern void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain); extern bool radeon_ttm_bo_is_radeon_bo(struct ttm_buffer_object *bo); +extern int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr, + uint32_t flags); +extern bool radeon_ttm_tt_has_userptr(struct ttm_tt *ttm); +extern bool radeon_ttm_tt_is_readonly(struct ttm_tt *ttm); extern void radeon_vram_location(struct radeon_device *rdev, struct radeon_mc *mc, u64 base); extern void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc); extern int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon); @@ -2934,10 +2979,10 @@ struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev); struct r600_audio_pin *dce6_audio_get_pin(struct radeon_device *rdev); void r600_audio_enable(struct radeon_device *rdev, struct r600_audio_pin *pin, - bool enable); + u8 enable_mask); void dce6_audio_enable(struct radeon_device *rdev, struct r600_audio_pin *pin, - bool enable); + u8 enable_mask); /* * R600 vram scratch functions diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index 2dd5847f9b98..850de57069be 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -963,6 +963,19 @@ static struct radeon_asic r600_asic = { }, }; +static struct radeon_asic_ring rv6xx_uvd_ring = { + .ib_execute = &uvd_v1_0_ib_execute, + .emit_fence = &uvd_v1_0_fence_emit, + .emit_semaphore = &uvd_v1_0_semaphore_emit, + .cs_parse = &radeon_uvd_cs_parse, + .ring_test = &uvd_v1_0_ring_test, + .ib_test = &uvd_v1_0_ib_test, + .is_lockup = &radeon_ring_test_lockup, + .get_rptr = &uvd_v1_0_get_rptr, + .get_wptr = &uvd_v1_0_get_wptr, + .set_wptr = &uvd_v1_0_set_wptr, +}; + static struct radeon_asic rv6xx_asic = { .init = &r600_init, .fini = &r600_fini, @@ -982,6 +995,7 @@ static struct radeon_asic rv6xx_asic = { .ring = { [RADEON_RING_TYPE_GFX_INDEX] = &r600_gfx_ring, [R600_RING_TYPE_DMA_INDEX] = &r600_dma_ring, + [R600_RING_TYPE_UVD_INDEX] = &rv6xx_uvd_ring, }, .irq = { .set = &r600_irq_set, @@ -1072,6 +1086,7 @@ static struct radeon_asic rs780_asic = { .ring = { [RADEON_RING_TYPE_GFX_INDEX] = &r600_gfx_ring, [R600_RING_TYPE_DMA_INDEX] = &r600_dma_ring, + [R600_RING_TYPE_UVD_INDEX] = &rv6xx_uvd_ring, }, .irq = { .set = &r600_irq_set, @@ -2296,7 +2311,15 @@ int radeon_asic_init(struct radeon_device *rdev) case CHIP_RS780: case CHIP_RS880: rdev->asic = &rs780_asic; - rdev->has_uvd = true; + /* 760G/780V/880V don't have UVD */ + if ((rdev->pdev->device == 0x9616)|| + (rdev->pdev->device == 0x9611)|| + (rdev->pdev->device == 0x9613)|| + (rdev->pdev->device == 0x9711)|| + (rdev->pdev->device == 0x9713)) + rdev->has_uvd = false; + else + rdev->has_uvd = true; break; case CHIP_RV770: case CHIP_RV730: diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h index 7756bc1e1cd3..d8ace5b28a5b 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.h +++ b/drivers/gpu/drm/radeon/radeon_asic.h @@ -81,11 +81,11 @@ bool r100_semaphore_ring_emit(struct radeon_device *rdev, int r100_cs_parse(struct radeon_cs_parser *p); void r100_pll_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); uint32_t r100_pll_rreg(struct radeon_device *rdev, uint32_t reg); -int r100_copy_blit(struct radeon_device *rdev, - uint64_t src_offset, - uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); +struct radeon_fence *r100_copy_blit(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); int r100_set_surface_reg(struct radeon_device *rdev, int reg, uint32_t tiling_flags, uint32_t pitch, uint32_t offset, uint32_t obj_size); @@ -152,11 +152,11 @@ void r100_gfx_set_wptr(struct radeon_device *rdev, /* * r200,rv250,rs300,rv280 */ -extern int r200_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, - uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); +struct radeon_fence *r200_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, + uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); void r200_set_safe_registers(struct radeon_device *rdev); /* @@ -340,12 +340,14 @@ int r600_dma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring); void r600_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib); int r600_ring_test(struct radeon_device *rdev, struct radeon_ring *cp); int r600_dma_ring_test(struct radeon_device *rdev, struct radeon_ring *cp); -int r600_copy_cpdma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, struct radeon_fence **fence); -int r600_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, struct radeon_fence **fence); +struct radeon_fence *r600_copy_cpdma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); +struct radeon_fence *r600_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); void r600_hpd_init(struct radeon_device *rdev); void r600_hpd_fini(struct radeon_device *rdev); bool r600_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd); @@ -389,7 +391,6 @@ 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); -struct r600_audio_pin r600_audio_status(struct radeon_device *rdev); void r600_audio_fini(struct radeon_device *rdev); void r600_audio_set_dto(struct drm_encoder *encoder, u32 clock); void r600_hdmi_update_avi_infoframe(struct drm_encoder *encoder, void *buffer, @@ -461,10 +462,10 @@ bool rv770_page_flip_pending(struct radeon_device *rdev, int crtc); 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); -int rv770_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); +struct radeon_fence *rv770_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); u32 rv770_get_xclk(struct radeon_device *rdev); int rv770_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk); int rv770_get_temp(struct radeon_device *rdev); @@ -535,10 +536,10 @@ void evergreen_dma_fence_ring_emit(struct radeon_device *rdev, struct radeon_fence *fence); void evergreen_dma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib); -int evergreen_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); +struct radeon_fence *evergreen_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable); void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode); int evergreen_get_temp(struct radeon_device *rdev); @@ -700,10 +701,10 @@ int si_vm_init(struct radeon_device *rdev); void si_vm_fini(struct radeon_device *rdev); void si_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm); int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib); -int si_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); +struct radeon_fence *si_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); void si_dma_vm_copy_pages(struct radeon_device *rdev, struct radeon_ib *ib, @@ -759,14 +760,14 @@ bool cik_sdma_semaphore_ring_emit(struct radeon_device *rdev, struct radeon_semaphore *semaphore, bool emit_wait); void cik_sdma_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib); -int cik_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); -int cik_copy_cpdma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence); +struct radeon_fence *cik_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); +struct radeon_fence *cik_copy_cpdma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv); int cik_sdma_ring_test(struct radeon_device *rdev, struct radeon_ring *ring); int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring); bool cik_sdma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring); @@ -882,6 +883,7 @@ uint32_t uvd_v1_0_get_wptr(struct radeon_device *rdev, struct radeon_ring *ring); void uvd_v1_0_set_wptr(struct radeon_device *rdev, struct radeon_ring *ring); +int uvd_v1_0_resume(struct radeon_device *rdev); int uvd_v1_0_init(struct radeon_device *rdev); void uvd_v1_0_fini(struct radeon_device *rdev); @@ -889,6 +891,8 @@ int uvd_v1_0_start(struct radeon_device *rdev); void uvd_v1_0_stop(struct radeon_device *rdev); int uvd_v1_0_ring_test(struct radeon_device *rdev, struct radeon_ring *ring); +void uvd_v1_0_fence_emit(struct radeon_device *rdev, + struct radeon_fence *fence); int uvd_v1_0_ib_test(struct radeon_device *rdev, struct radeon_ring *ring); bool uvd_v1_0_semaphore_emit(struct radeon_device *rdev, struct radeon_ring *ring, diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index e74c7e387dde..df69b92ba164 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -458,7 +458,7 @@ static bool radeon_atom_apply_quirks(struct drm_device *dev, return true; } -const int supported_devices_connector_convert[] = { +static const int supported_devices_connector_convert[] = { DRM_MODE_CONNECTOR_Unknown, DRM_MODE_CONNECTOR_VGA, DRM_MODE_CONNECTOR_DVII, @@ -477,7 +477,7 @@ const int supported_devices_connector_convert[] = { DRM_MODE_CONNECTOR_DisplayPort }; -const uint16_t supported_devices_connector_object_id_convert[] = { +static const uint16_t supported_devices_connector_object_id_convert[] = { CONNECTOR_OBJECT_ID_NONE, CONNECTOR_OBJECT_ID_VGA, CONNECTOR_OBJECT_ID_DUAL_LINK_DVI_I, /* not all boards support DL */ @@ -494,7 +494,7 @@ const uint16_t supported_devices_connector_object_id_convert[] = { CONNECTOR_OBJECT_ID_SVIDEO }; -const int object_connector_convert[] = { +static const int object_connector_convert[] = { DRM_MODE_CONNECTOR_Unknown, DRM_MODE_CONNECTOR_DVII, DRM_MODE_CONNECTOR_DVII, diff --git a/drivers/gpu/drm/radeon/radeon_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c index 69f5695bdab9..9e7f23dd14bd 100644 --- a/drivers/gpu/drm/radeon/radeon_benchmark.c +++ b/drivers/gpu/drm/radeon/radeon_benchmark.c @@ -45,33 +45,29 @@ static int radeon_benchmark_do_move(struct radeon_device *rdev, unsigned size, for (i = 0; i < n; i++) { switch (flag) { case RADEON_BENCHMARK_COPY_DMA: - r = radeon_copy_dma(rdev, saddr, daddr, - size / RADEON_GPU_PAGE_SIZE, - &fence); + fence = radeon_copy_dma(rdev, saddr, daddr, + size / RADEON_GPU_PAGE_SIZE, + NULL); break; case RADEON_BENCHMARK_COPY_BLIT: - r = radeon_copy_blit(rdev, saddr, daddr, - size / RADEON_GPU_PAGE_SIZE, - &fence); + fence = radeon_copy_blit(rdev, saddr, daddr, + size / RADEON_GPU_PAGE_SIZE, + NULL); break; default: DRM_ERROR("Unknown copy method\n"); - r = -EINVAL; + return -EINVAL; } - if (r) - goto exit_do_move; + if (IS_ERR(fence)) + return PTR_ERR(fence); + r = radeon_fence_wait(fence, false); - if (r) - goto exit_do_move; radeon_fence_unref(&fence); + if (r) + return r; } end_jiffies = jiffies; - r = jiffies_to_msecs(end_jiffies - start_jiffies); - -exit_do_move: - if (fence) - radeon_fence_unref(&fence); - return r; + return jiffies_to_msecs(end_jiffies - start_jiffies); } @@ -97,7 +93,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size, int time; n = RADEON_BENCHMARK_ITERATIONS; - r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, 0, NULL, &sobj); + r = radeon_bo_create(rdev, size, PAGE_SIZE, true, sdomain, 0, NULL, NULL, &sobj); if (r) { goto out_cleanup; } @@ -109,7 +105,7 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size, if (r) { goto out_cleanup; } - r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, 0, NULL, &dobj); + r = radeon_bo_create(rdev, size, PAGE_SIZE, true, ddomain, 0, NULL, NULL, &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 6651177110f0..3e5f6b71f3ad 100644 --- a/drivers/gpu/drm/radeon/radeon_combios.c +++ b/drivers/gpu/drm/radeon/radeon_combios.c @@ -116,7 +116,7 @@ enum radeon_combios_connector { CONNECTOR_UNSUPPORTED_LEGACY }; -const int legacy_connector_convert[] = { +static const int legacy_connector_convert[] = { DRM_MODE_CONNECTOR_Unknown, DRM_MODE_CONNECTOR_DVID, DRM_MODE_CONNECTOR_VGA, diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c index bb0d5c3a8311..ea134a7d51a5 100644 --- a/drivers/gpu/drm/radeon/radeon_cp.c +++ b/drivers/gpu/drm/radeon/radeon_cp.c @@ -1298,27 +1298,27 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, dev_priv->buffers_offset = init->buffers_offset; dev_priv->gart_textures_offset = init->gart_textures_offset; - master_priv->sarea = drm_getsarea(dev); + master_priv->sarea = drm_legacy_getsarea(dev); if (!master_priv->sarea) { DRM_ERROR("could not find sarea!\n"); radeon_do_cleanup_cp(dev); return -EINVAL; } - dev_priv->cp_ring = drm_core_findmap(dev, init->ring_offset); + dev_priv->cp_ring = drm_legacy_findmap(dev, init->ring_offset); if (!dev_priv->cp_ring) { DRM_ERROR("could not find cp ring region!\n"); radeon_do_cleanup_cp(dev); return -EINVAL; } - dev_priv->ring_rptr = drm_core_findmap(dev, init->ring_rptr_offset); + dev_priv->ring_rptr = drm_legacy_findmap(dev, init->ring_rptr_offset); if (!dev_priv->ring_rptr) { DRM_ERROR("could not find ring read pointer!\n"); radeon_do_cleanup_cp(dev); return -EINVAL; } dev->agp_buffer_token = init->buffers_offset; - dev->agp_buffer_map = drm_core_findmap(dev, init->buffers_offset); + dev->agp_buffer_map = drm_legacy_findmap(dev, init->buffers_offset); if (!dev->agp_buffer_map) { DRM_ERROR("could not find dma buffer region!\n"); radeon_do_cleanup_cp(dev); @@ -1327,7 +1327,7 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, if (init->gart_textures_offset) { dev_priv->gart_textures = - drm_core_findmap(dev, init->gart_textures_offset); + drm_legacy_findmap(dev, init->gart_textures_offset); if (!dev_priv->gart_textures) { DRM_ERROR("could not find GART texture region!\n"); radeon_do_cleanup_cp(dev); @@ -1337,9 +1337,9 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, #if __OS_HAS_AGP if (dev_priv->flags & RADEON_IS_AGP) { - drm_core_ioremap_wc(dev_priv->cp_ring, dev); - drm_core_ioremap_wc(dev_priv->ring_rptr, dev); - drm_core_ioremap_wc(dev->agp_buffer_map, dev); + drm_legacy_ioremap_wc(dev_priv->cp_ring, dev); + drm_legacy_ioremap_wc(dev_priv->ring_rptr, dev); + drm_legacy_ioremap_wc(dev->agp_buffer_map, dev); if (!dev_priv->cp_ring->handle || !dev_priv->ring_rptr->handle || !dev->agp_buffer_map->handle) { @@ -1475,7 +1475,7 @@ static int radeon_do_init_cp(struct drm_device *dev, drm_radeon_init_t *init, dev_priv->gart_info.mapping.size = dev_priv->gart_info.table_size; - drm_core_ioremap_wc(&dev_priv->gart_info.mapping, dev); + drm_legacy_ioremap_wc(&dev_priv->gart_info.mapping, dev); dev_priv->gart_info.addr = dev_priv->gart_info.mapping.handle; @@ -1569,15 +1569,15 @@ static int radeon_do_cleanup_cp(struct drm_device * dev) #if __OS_HAS_AGP if (dev_priv->flags & RADEON_IS_AGP) { if (dev_priv->cp_ring != NULL) { - drm_core_ioremapfree(dev_priv->cp_ring, dev); + drm_legacy_ioremapfree(dev_priv->cp_ring, dev); dev_priv->cp_ring = NULL; } if (dev_priv->ring_rptr != NULL) { - drm_core_ioremapfree(dev_priv->ring_rptr, dev); + drm_legacy_ioremapfree(dev_priv->ring_rptr, dev); dev_priv->ring_rptr = NULL; } if (dev->agp_buffer_map != NULL) { - drm_core_ioremapfree(dev->agp_buffer_map, dev); + drm_legacy_ioremapfree(dev->agp_buffer_map, dev); dev->agp_buffer_map = NULL; } } else @@ -1597,7 +1597,7 @@ static int radeon_do_cleanup_cp(struct drm_device * dev) if (dev_priv->gart_info.gart_table_location == DRM_ATI_GART_FB) { - drm_core_ioremapfree(&dev_priv->gart_info.mapping, dev); + drm_legacy_ioremapfree(&dev_priv->gart_info.mapping, dev); dev_priv->gart_info.addr = NULL; } } @@ -2106,9 +2106,9 @@ int radeon_driver_load(struct drm_device *dev, unsigned long flags) else dev_priv->flags |= RADEON_IS_PCI; - ret = drm_addmap(dev, pci_resource_start(dev->pdev, 2), - pci_resource_len(dev->pdev, 2), _DRM_REGISTERS, - _DRM_READ_ONLY | _DRM_DRIVER, &dev_priv->mmio); + ret = drm_legacy_addmap(dev, pci_resource_start(dev->pdev, 2), + pci_resource_len(dev->pdev, 2), _DRM_REGISTERS, + _DRM_READ_ONLY | _DRM_DRIVER, &dev_priv->mmio); if (ret != 0) return ret; @@ -2135,8 +2135,8 @@ int radeon_master_create(struct drm_device *dev, struct drm_master *master) /* prebuild the SAREA */ sareapage = max_t(unsigned long, SAREA_MAX, PAGE_SIZE); - ret = drm_addmap(dev, 0, sareapage, _DRM_SHM, _DRM_CONTAINS_LOCK, - &master_priv->sarea); + ret = drm_legacy_addmap(dev, 0, sareapage, _DRM_SHM, _DRM_CONTAINS_LOCK, + &master_priv->sarea); if (ret) { DRM_ERROR("SAREA setup failed\n"); kfree(master_priv); @@ -2162,7 +2162,7 @@ void radeon_master_destroy(struct drm_device *dev, struct drm_master *master) master_priv->sarea_priv = NULL; if (master_priv->sarea) - drm_rmmap_locked(dev, master_priv->sarea); + drm_legacy_rmmap_locked(dev, master_priv->sarea); kfree(master_priv); @@ -2181,9 +2181,9 @@ int radeon_driver_firstopen(struct drm_device *dev) dev_priv->gart_info.table_size = RADEON_PCIGART_TABLE_SIZE; dev_priv->fb_aper_offset = pci_resource_start(dev->pdev, 0); - ret = drm_addmap(dev, dev_priv->fb_aper_offset, - pci_resource_len(dev->pdev, 0), _DRM_FRAME_BUFFER, - _DRM_WRITE_COMBINING, &map); + ret = drm_legacy_addmap(dev, dev_priv->fb_aper_offset, + pci_resource_len(dev->pdev, 0), + _DRM_FRAME_BUFFER, _DRM_WRITE_COMBINING, &map); if (ret != 0) return ret; @@ -2196,7 +2196,7 @@ int radeon_driver_unload(struct drm_device *dev) DRM_DEBUG("\n"); - drm_rmmap(dev, dev_priv->mmio); + drm_legacy_rmmap(dev, dev_priv->mmio); kfree(dev_priv); diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c index 83f382e8e40e..1c893447d7cd 100644 --- a/drivers/gpu/drm/radeon/radeon_cs.c +++ b/drivers/gpu/drm/radeon/radeon_cs.c @@ -78,7 +78,8 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) struct radeon_cs_chunk *chunk; struct radeon_cs_buckets buckets; unsigned i, j; - bool duplicate; + bool duplicate, need_mmap_lock = false; + int r; if (p->chunk_relocs_idx == -1) { return 0; @@ -136,10 +137,13 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) + !!r->write_domain; /* the first reloc of an UVD job is the msg and that must be in - VRAM, also but everything into VRAM on AGP cards to avoid - image corruptions */ + VRAM, also but everything into VRAM on AGP cards and older + IGP chips to avoid image corruptions */ if (p->ring == R600_RING_TYPE_UVD_INDEX && - (i == 0 || drm_pci_device_is_agp(p->rdev->ddev))) { + (i == 0 || drm_pci_device_is_agp(p->rdev->ddev) || + p->rdev->family == CHIP_RS780 || + p->rdev->family == CHIP_RS880)) { + /* TODO: is this still needed for NI+ ? */ p->relocs[i].prefered_domains = RADEON_GEM_DOMAIN_VRAM; @@ -165,7 +169,21 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) p->relocs[i].allowed_domains = domain; } + if (radeon_ttm_tt_has_userptr(p->relocs[i].robj->tbo.ttm)) { + uint32_t domain = p->relocs[i].prefered_domains; + if (!(domain & RADEON_GEM_DOMAIN_GTT)) { + DRM_ERROR("Only RADEON_GEM_DOMAIN_GTT is " + "allowed for userptr BOs\n"); + return -EINVAL; + } + need_mmap_lock = true; + domain = RADEON_GEM_DOMAIN_GTT; + p->relocs[i].prefered_domains = domain; + p->relocs[i].allowed_domains = domain; + } + p->relocs[i].tv.bo = &p->relocs[i].robj->tbo; + p->relocs[i].tv.shared = !r->write_domain; p->relocs[i].handle = r->handle; radeon_cs_buckets_add(&buckets, &p->relocs[i].tv.head, @@ -177,8 +195,15 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) if (p->cs_flags & RADEON_CS_USE_VM) p->vm_bos = radeon_vm_get_bos(p->rdev, p->ib.vm, &p->validated); + if (need_mmap_lock) + down_read(¤t->mm->mmap_sem); - return radeon_bo_list_validate(p->rdev, &p->ticket, &p->validated, p->ring); + r = radeon_bo_list_validate(p->rdev, &p->ticket, &p->validated, p->ring); + + if (need_mmap_lock) + up_read(¤t->mm->mmap_sem); + + return r; } static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority) @@ -224,17 +249,24 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority return 0; } -static void radeon_cs_sync_rings(struct radeon_cs_parser *p) +static int radeon_cs_sync_rings(struct radeon_cs_parser *p) { - int i; + int i, r = 0; for (i = 0; i < p->nrelocs; i++) { + struct reservation_object *resv; + if (!p->relocs[i].robj) continue; - radeon_semaphore_sync_to(p->ib.semaphore, - p->relocs[i].robj->tbo.sync_obj); + resv = p->relocs[i].robj->tbo.resv; + r = radeon_semaphore_sync_resv(p->rdev, p->ib.semaphore, resv, + p->relocs[i].tv.shared); + + if (r) + break; } + return r; } /* XXX: note that this is called from the legacy UMS CS ioctl as well */ @@ -403,7 +435,7 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo ttm_eu_fence_buffer_objects(&parser->ticket, &parser->validated, - parser->ib.fence); + &parser->ib.fence->base); } else if (backoff) { ttm_eu_backoff_reservation(&parser->ticket, &parser->validated); @@ -444,13 +476,19 @@ static int radeon_cs_ib_chunk(struct radeon_device *rdev, return r; } + r = radeon_cs_sync_rings(parser); + if (r) { + if (r != -ERESTARTSYS) + DRM_ERROR("Failed to sync rings: %i\n", r); + return r; + } + if (parser->ring == R600_RING_TYPE_UVD_INDEX) radeon_uvd_note_usage(rdev); else if ((parser->ring == TN_RING_TYPE_VCE1_INDEX) || (parser->ring == TN_RING_TYPE_VCE2_INDEX)) radeon_vce_note_usage(rdev); - radeon_cs_sync_rings(parser); r = radeon_ib_schedule(rdev, &parser->ib, NULL, true); if (r) { DRM_ERROR("Failed to schedule IB !\n"); @@ -537,8 +575,14 @@ static int radeon_cs_ib_vm_chunk(struct radeon_device *rdev, if (r) { goto out; } - radeon_cs_sync_rings(parser); - radeon_semaphore_sync_to(parser->ib.semaphore, vm->fence); + + r = radeon_cs_sync_rings(parser); + if (r) { + if (r != -ERESTARTSYS) + DRM_ERROR("Failed to sync rings: %i\n", r); + goto out; + } + radeon_semaphore_sync_fence(parser->ib.semaphore, vm->fence); if ((rdev->family >= CHIP_TAHITI) && (parser->chunk_const_ib_idx != -1)) { @@ -629,6 +673,13 @@ int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) up_read(&rdev->exclusive_lock); return -EBUSY; } + if (rdev->in_reset) { + up_read(&rdev->exclusive_lock); + r = radeon_gpu_reset(rdev); + if (!r) + r = -EAGAIN; + return r; + } /* initialize parser */ memset(&parser, 0, sizeof(struct radeon_cs_parser)); parser.filp = filp; diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 12c8329644c4..f41cc1538e48 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -434,7 +434,7 @@ int radeon_wb_init(struct radeon_device *rdev) if (rdev->wb.wb_obj == NULL) { r = radeon_bo_create(rdev, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_GTT, 0, NULL, + RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL, &rdev->wb.wb_obj); if (r) { dev_warn(rdev->dev, "(%d) create WB bo failed\n", r); @@ -1257,6 +1257,7 @@ int radeon_device_init(struct radeon_device *rdev, for (i = 0; i < RADEON_NUM_RINGS; i++) { rdev->ring[i].idx = i; } + rdev->fence_context = fence_context_alloc(RADEON_NUM_RINGS); DRM_INFO("initializing kernel modesetting (%s 0x%04X:0x%04X 0x%04X:0x%04X).\n", radeon_family_name[rdev->family], pdev->vendor, pdev->device, @@ -1274,6 +1275,8 @@ int radeon_device_init(struct radeon_device *rdev, init_rwsem(&rdev->pm.mclk_lock); init_rwsem(&rdev->exclusive_lock); init_waitqueue_head(&rdev->irq.vblank_queue); + mutex_init(&rdev->mn_lock); + hash_init(rdev->mn_hash); r = radeon_gem_init(rdev); if (r) return r; @@ -1399,10 +1402,6 @@ int radeon_device_init(struct radeon_device *rdev, if (r) goto failed; - r = radeon_ib_ring_tests(rdev); - if (r) - DRM_ERROR("ib ring test failed (%d).\n", r); - r = radeon_gem_debugfs_init(rdev); if (r) { DRM_ERROR("registering gem debugfs failed (%d).\n", r); @@ -1420,6 +1419,10 @@ int radeon_device_init(struct radeon_device *rdev, goto failed; } + r = radeon_ib_ring_tests(rdev); + if (r) + DRM_ERROR("ib ring test failed (%d).\n", r); + if ((radeon_testing & 1)) { if (rdev->accel_working) radeon_test_moves(rdev); @@ -1497,7 +1500,6 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon) struct drm_crtc *crtc; struct drm_connector *connector; int i, r; - bool force_completion = false; if (dev == NULL || dev->dev_private == NULL) { return -ENODEV; @@ -1541,12 +1543,9 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon) r = radeon_fence_wait_empty(rdev, i); if (r) { /* delay GPU reset to resume */ - force_completion = true; + radeon_fence_driver_force_completion(rdev, i); } } - if (force_completion) { - radeon_fence_driver_force_completion(rdev); - } radeon_save_bios_scratch_regs(rdev); @@ -1686,8 +1685,6 @@ int radeon_gpu_reset(struct radeon_device *rdev) return 0; } - rdev->needs_reset = false; - radeon_save_bios_scratch_regs(rdev); /* block TTM */ resched = ttm_bo_lock_delayed_workqueue(&rdev->mman.bdev); @@ -1704,7 +1701,6 @@ int radeon_gpu_reset(struct radeon_device *rdev) } } -retry: r = radeon_asic_reset(rdev); if (!r) { dev_info(rdev->dev, "GPU reset succeeded, trying to resume\n"); @@ -1713,26 +1709,12 @@ retry: radeon_restore_bios_scratch_regs(rdev); - if (!r) { - for (i = 0; i < RADEON_NUM_RINGS; ++i) { + for (i = 0; i < RADEON_NUM_RINGS; ++i) { + if (!r && ring_data[i]) { radeon_ring_restore(rdev, &rdev->ring[i], ring_sizes[i], ring_data[i]); - ring_sizes[i] = 0; - ring_data[i] = NULL; - } - - r = radeon_ib_ring_tests(rdev); - if (r) { - dev_err(rdev->dev, "ib ring test failed (%d).\n", r); - if (saved) { - saved = false; - radeon_suspend(rdev); - goto retry; - } - } - } else { - radeon_fence_driver_force_completion(rdev); - for (i = 0; i < RADEON_NUM_RINGS; ++i) { + } else { + radeon_fence_driver_force_completion(rdev, i); kfree(ring_data[i]); } } @@ -1764,19 +1746,32 @@ retry: /* reset hpd state */ radeon_hpd_init(rdev); + ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched); + + rdev->in_reset = true; + rdev->needs_reset = false; + + downgrade_write(&rdev->exclusive_lock); + drm_helper_resume_force_mode(rdev->ddev); /* set the power state here in case we are a PX system or headless */ if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) radeon_pm_compute_clocks(rdev); - ttm_bo_unlock_delayed_workqueue(&rdev->mman.bdev, resched); - if (r) { + if (!r) { + r = radeon_ib_ring_tests(rdev); + if (r && saved) + r = -EAGAIN; + } else { /* bad news, how to tell it to userspace ? */ dev_info(rdev->dev, "GPU reset failed\n"); } - up_write(&rdev->exclusive_lock); + rdev->needs_reset = r == -EAGAIN; + rdev->in_reset = false; + + up_read(&rdev->exclusive_lock); return r; } diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 3fdf87318069..00ead8c2758a 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -402,12 +402,21 @@ static void radeon_flip_work_func(struct work_struct *__work) down_read(&rdev->exclusive_lock); if (work->fence) { - r = radeon_fence_wait(work->fence, false); - if (r == -EDEADLK) { - up_read(&rdev->exclusive_lock); - r = radeon_gpu_reset(rdev); - down_read(&rdev->exclusive_lock); - } + struct radeon_fence *fence; + + fence = to_radeon_fence(work->fence); + if (fence && fence->rdev == rdev) { + r = radeon_fence_wait(fence, false); + if (r == -EDEADLK) { + up_read(&rdev->exclusive_lock); + do { + r = radeon_gpu_reset(rdev); + } while (r == -EAGAIN); + down_read(&rdev->exclusive_lock); + } + } else + r = fence_wait(work->fence, false); + if (r) DRM_ERROR("failed to wait on page flip fence (%d)!\n", r); @@ -416,7 +425,8 @@ static void radeon_flip_work_func(struct work_struct *__work) * confused about which BO the CRTC is scanning out */ - radeon_fence_unref(&work->fence); + fence_put(work->fence); + work->fence = NULL; } /* We borrow the event spin lock for protecting flip_status */ @@ -474,11 +484,6 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc, obj = new_radeon_fb->obj; new_rbo = gem_to_radeon_bo(obj); - spin_lock(&new_rbo->tbo.bdev->fence_lock); - if (new_rbo->tbo.sync_obj) - work->fence = radeon_fence_ref(new_rbo->tbo.sync_obj); - spin_unlock(&new_rbo->tbo.bdev->fence_lock); - /* pin the new buffer */ DRM_DEBUG_DRIVER("flip-ioctl() cur_rbo = %p, new_rbo = %p\n", work->old_rbo, new_rbo); @@ -497,6 +502,7 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc, DRM_ERROR("failed to pin new rbo buffer before flip\n"); goto cleanup; } + work->fence = fence_get(reservation_object_get_excl(new_rbo->tbo.resv)); radeon_bo_get_tiling_flags(new_rbo, &tiling_flags, NULL); radeon_bo_unreserve(new_rbo); @@ -578,9 +584,8 @@ pflip_cleanup: cleanup: drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base); - radeon_fence_unref(&work->fence); + fence_put(work->fence); kfree(work); - return r; } @@ -1917,7 +1922,7 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, unsigned int fl /* In vblank? */ if (in_vbl) - ret |= DRM_SCANOUTPOS_INVBL; + ret |= DRM_SCANOUTPOS_IN_VBLANK; /* Is vpos outside nominal vblank area, but less than * 1/100 of a frame height away from start of vblank? diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index f9d17b29b343..dcffa30ee2db 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -38,6 +38,8 @@ #include <linux/module.h> #include <linux/pm_runtime.h> #include <linux/vga_switcheroo.h> +#include <drm/drm_gem.h> + #include "drm_crtc_helper.h" /* * KMS wrapper. @@ -114,6 +116,9 @@ int radeon_gem_object_open(struct drm_gem_object *obj, struct drm_file *file_priv); void radeon_gem_object_close(struct drm_gem_object *obj, struct drm_file *file_priv); +struct dma_buf *radeon_gem_prime_export(struct drm_device *dev, + struct drm_gem_object *gobj, + int flags); extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, unsigned int flags, int *vpos, int *hpos, ktime_t *stime, @@ -130,7 +135,7 @@ int radeon_mode_dumb_create(struct drm_file *file_priv, struct drm_mode_create_dumb *args); struct sg_table *radeon_gem_prime_get_sg_table(struct drm_gem_object *obj); struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev, - size_t size, + struct dma_buf_attachment *, struct sg_table *sg); int radeon_gem_prime_pin(struct drm_gem_object *obj); void radeon_gem_prime_unpin(struct drm_gem_object *obj); @@ -309,7 +314,7 @@ static const struct file_operations radeon_driver_old_fops = { .open = drm_open, .release = drm_release, .unlocked_ioctl = drm_ioctl, - .mmap = drm_mmap, + .mmap = drm_legacy_mmap, .poll = drm_poll, .read = drm_read, #ifdef CONFIG_COMPAT @@ -329,6 +334,7 @@ static struct drm_driver driver_old = { .preclose = radeon_driver_preclose, .postclose = radeon_driver_postclose, .lastclose = radeon_driver_lastclose, + .set_busid = drm_pci_set_busid, .unload = radeon_driver_unload, .suspend = radeon_suspend, .resume = radeon_resume, @@ -553,6 +559,7 @@ static struct drm_driver kms_driver = { .preclose = radeon_driver_preclose_kms, .postclose = radeon_driver_postclose_kms, .lastclose = radeon_driver_lastclose_kms, + .set_busid = drm_pci_set_busid, .unload = radeon_driver_unload_kms, .get_vblank_counter = radeon_get_vblank_counter_kms, .enable_vblank = radeon_enable_vblank_kms, @@ -578,7 +585,7 @@ static struct drm_driver kms_driver = { .prime_handle_to_fd = drm_gem_prime_handle_to_fd, .prime_fd_to_handle = drm_gem_prime_fd_to_handle, - .gem_prime_export = drm_gem_prime_export, + .gem_prime_export = radeon_gem_prime_export, .gem_prime_import = drm_gem_prime_import, .gem_prime_pin = radeon_gem_prime_pin, .gem_prime_unpin = radeon_gem_prime_unpin, diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h index dafd812e4571..46bd3938282c 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.h +++ b/drivers/gpu/drm/radeon/radeon_drv.h @@ -33,7 +33,9 @@ #include <linux/firmware.h> #include <linux/platform_device.h> +#include <drm/drm_legacy.h> +#include <drm/ati_pcigart.h> #include "radeon_family.h" /* General customization: diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index 15edf23b465c..9a19e52cc655 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -410,3 +410,24 @@ bool radeon_dig_monitor_is_duallink(struct drm_encoder *encoder, } } +bool radeon_encoder_is_digital(struct drm_encoder *encoder) +{ + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + switch (radeon_encoder->encoder_id) { + case ENCODER_OBJECT_ID_INTERNAL_LVDS: + case ENCODER_OBJECT_ID_INTERNAL_TMDS1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_TMDS1: + case ENCODER_OBJECT_ID_INTERNAL_LVTM1: + case ENCODER_OBJECT_ID_INTERNAL_DVO1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1: + case ENCODER_OBJECT_ID_INTERNAL_DDI: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3: + return true; + default: + return false; + } +} diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index 94b0f2aa3d7c..0ea1db83d573 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -189,7 +189,8 @@ out_unref: static int radeonfb_create(struct drm_fb_helper *helper, struct drm_fb_helper_surface_size *sizes) { - struct radeon_fbdev *rfbdev = (struct radeon_fbdev *)helper; + struct radeon_fbdev *rfbdev = + container_of(helper, struct radeon_fbdev, helper); struct radeon_device *rdev = rfbdev->rdev; struct fb_info *info; struct drm_framebuffer *fb = NULL; diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 913787085dfa..995167025282 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -98,6 +98,25 @@ static u32 radeon_fence_read(struct radeon_device *rdev, int ring) } /** + * radeon_fence_schedule_check - schedule lockup check + * + * @rdev: radeon_device pointer + * @ring: ring index we should work with + * + * Queues a delayed work item to check for lockups. + */ +static void radeon_fence_schedule_check(struct radeon_device *rdev, int ring) +{ + /* + * Do not reset the timer here with mod_delayed_work, + * this can livelock in an interaction with TTM delayed destroy. + */ + queue_delayed_work(system_power_efficient_wq, + &rdev->fence_drv[ring].lockup_work, + RADEON_FENCE_JIFFIES_TIMEOUT); +} + +/** * radeon_fence_emit - emit a fence on the requested ring * * @rdev: radeon_device pointer @@ -111,30 +130,70 @@ int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence **fence, int ring) { + u64 seq = ++rdev->fence_drv[ring].sync_seq[ring]; + /* we are protected by the ring emission mutex */ *fence = kmalloc(sizeof(struct radeon_fence), GFP_KERNEL); if ((*fence) == NULL) { return -ENOMEM; } - kref_init(&((*fence)->kref)); (*fence)->rdev = rdev; - (*fence)->seq = ++rdev->fence_drv[ring].sync_seq[ring]; + (*fence)->seq = seq; (*fence)->ring = ring; + fence_init(&(*fence)->base, &radeon_fence_ops, + &rdev->fence_queue.lock, rdev->fence_context + ring, seq); radeon_fence_ring_emit(rdev, ring, *fence); trace_radeon_fence_emit(rdev->ddev, ring, (*fence)->seq); + radeon_fence_schedule_check(rdev, ring); return 0; } /** - * radeon_fence_process - process a fence + * radeon_fence_check_signaled - callback from fence_queue + * + * this function is called with fence_queue lock held, which is also used + * for the fence locking itself, so unlocked variants are used for + * fence_signal, and remove_wait_queue. + */ +static int radeon_fence_check_signaled(wait_queue_t *wait, unsigned mode, int flags, void *key) +{ + struct radeon_fence *fence; + u64 seq; + + fence = container_of(wait, struct radeon_fence, fence_wake); + + /* + * We cannot use radeon_fence_process here because we're already + * in the waitqueue, in a call from wake_up_all. + */ + seq = atomic64_read(&fence->rdev->fence_drv[fence->ring].last_seq); + if (seq >= fence->seq) { + int ret = fence_signal_locked(&fence->base); + + if (!ret) + FENCE_TRACE(&fence->base, "signaled from irq context\n"); + else + FENCE_TRACE(&fence->base, "was already signaled\n"); + + radeon_irq_kms_sw_irq_put(fence->rdev, fence->ring); + __remove_wait_queue(&fence->rdev->fence_queue, &fence->fence_wake); + fence_put(&fence->base); + } else + FENCE_TRACE(&fence->base, "pending\n"); + return 0; +} + +/** + * radeon_fence_activity - check for fence activity * * @rdev: radeon_device pointer * @ring: ring index the fence is associated with * - * Checks the current fence value and wakes the fence queue - * if the sequence number has increased (all asics). + * Checks the current fence value and calculates the last + * signalled fence value. Returns true if activity occured + * on the ring, and the fence_queue should be waken up. */ -void radeon_fence_process(struct radeon_device *rdev, int ring) +static bool radeon_fence_activity(struct radeon_device *rdev, int ring) { uint64_t seq, last_seq, last_emitted; unsigned count_loop = 0; @@ -190,23 +249,77 @@ void radeon_fence_process(struct radeon_device *rdev, int ring) } } while (atomic64_xchg(&rdev->fence_drv[ring].last_seq, seq) > seq); - if (wake) - wake_up_all(&rdev->fence_queue); + if (seq < last_emitted) + radeon_fence_schedule_check(rdev, ring); + + return wake; } /** - * radeon_fence_destroy - destroy a fence + * radeon_fence_check_lockup - check for hardware lockup * - * @kref: fence kref + * @work: delayed work item * - * Frees the fence object (all asics). + * Checks for fence activity and if there is none probe + * the hardware if a lockup occured. */ -static void radeon_fence_destroy(struct kref *kref) +static void radeon_fence_check_lockup(struct work_struct *work) { - struct radeon_fence *fence; + struct radeon_fence_driver *fence_drv; + struct radeon_device *rdev; + int ring; + + fence_drv = container_of(work, struct radeon_fence_driver, + lockup_work.work); + rdev = fence_drv->rdev; + ring = fence_drv - &rdev->fence_drv[0]; + + if (!down_read_trylock(&rdev->exclusive_lock)) { + /* just reschedule the check if a reset is going on */ + radeon_fence_schedule_check(rdev, ring); + return; + } + + if (fence_drv->delayed_irq && rdev->ddev->irq_enabled) { + unsigned long irqflags; + + fence_drv->delayed_irq = false; + spin_lock_irqsave(&rdev->irq.lock, irqflags); + radeon_irq_set(rdev); + spin_unlock_irqrestore(&rdev->irq.lock, irqflags); + } + + if (radeon_fence_activity(rdev, ring)) + wake_up_all(&rdev->fence_queue); - fence = container_of(kref, struct radeon_fence, kref); - kfree(fence); + else if (radeon_ring_is_lockup(rdev, ring, &rdev->ring[ring])) { + + /* good news we believe it's a lockup */ + dev_warn(rdev->dev, "GPU lockup (current fence id " + "0x%016llx last fence id 0x%016llx on ring %d)\n", + (uint64_t)atomic64_read(&fence_drv->last_seq), + fence_drv->sync_seq[ring], ring); + + /* remember that we need an reset */ + rdev->needs_reset = true; + wake_up_all(&rdev->fence_queue); + } + up_read(&rdev->exclusive_lock); +} + +/** + * radeon_fence_process - process a fence + * + * @rdev: radeon_device pointer + * @ring: ring index the fence is associated with + * + * Checks the current fence value and wakes the fence queue + * if the sequence number has increased (all asics). + */ +void radeon_fence_process(struct radeon_device *rdev, int ring) +{ + if (radeon_fence_activity(rdev, ring)) + wake_up_all(&rdev->fence_queue); } /** @@ -237,6 +350,75 @@ static bool radeon_fence_seq_signaled(struct radeon_device *rdev, return false; } +static bool radeon_fence_is_signaled(struct fence *f) +{ + struct radeon_fence *fence = to_radeon_fence(f); + struct radeon_device *rdev = fence->rdev; + unsigned ring = fence->ring; + u64 seq = fence->seq; + + if (atomic64_read(&rdev->fence_drv[ring].last_seq) >= seq) { + return true; + } + + if (down_read_trylock(&rdev->exclusive_lock)) { + radeon_fence_process(rdev, ring); + up_read(&rdev->exclusive_lock); + + if (atomic64_read(&rdev->fence_drv[ring].last_seq) >= seq) { + return true; + } + } + return false; +} + +/** + * radeon_fence_enable_signaling - enable signalling on fence + * @fence: fence + * + * This function is called with fence_queue lock held, and adds a callback + * to fence_queue that checks if this fence is signaled, and if so it + * signals the fence and removes itself. + */ +static bool radeon_fence_enable_signaling(struct fence *f) +{ + struct radeon_fence *fence = to_radeon_fence(f); + struct radeon_device *rdev = fence->rdev; + + if (atomic64_read(&rdev->fence_drv[fence->ring].last_seq) >= fence->seq) + return false; + + if (down_read_trylock(&rdev->exclusive_lock)) { + radeon_irq_kms_sw_irq_get(rdev, fence->ring); + + if (radeon_fence_activity(rdev, fence->ring)) + wake_up_all_locked(&rdev->fence_queue); + + /* did fence get signaled after we enabled the sw irq? */ + if (atomic64_read(&rdev->fence_drv[fence->ring].last_seq) >= fence->seq) { + radeon_irq_kms_sw_irq_put(rdev, fence->ring); + up_read(&rdev->exclusive_lock); + return false; + } + + up_read(&rdev->exclusive_lock); + } else { + /* we're probably in a lockup, lets not fiddle too much */ + if (radeon_irq_kms_sw_irq_get_delayed(rdev, fence->ring)) + rdev->fence_drv[fence->ring].delayed_irq = true; + radeon_fence_schedule_check(rdev, fence->ring); + } + + fence->fence_wake.flags = 0; + fence->fence_wake.private = NULL; + fence->fence_wake.func = radeon_fence_check_signaled; + __add_wait_queue(&rdev->fence_queue, &fence->fence_wake); + fence_get(f); + + FENCE_TRACE(&fence->base, "armed on ring %i!\n", fence->ring); + return true; +} + /** * radeon_fence_signaled - check if a fence has signaled * @@ -247,14 +429,15 @@ static bool radeon_fence_seq_signaled(struct radeon_device *rdev, */ bool radeon_fence_signaled(struct radeon_fence *fence) { - if (!fence) { + if (!fence) return true; - } - if (fence->seq == RADEON_FENCE_SIGNALED_SEQ) { - return true; - } + if (radeon_fence_seq_signaled(fence->rdev, fence->seq, fence->ring)) { - fence->seq = RADEON_FENCE_SIGNALED_SEQ; + int ret; + + ret = fence_signal(&fence->base); + if (!ret) + FENCE_TRACE(&fence->base, "signaled from radeon_fence_signaled\n"); return true; } return false; @@ -283,110 +466,70 @@ static bool radeon_fence_any_seq_signaled(struct radeon_device *rdev, u64 *seq) } /** - * radeon_fence_wait_seq - wait for a specific sequence numbers + * radeon_fence_wait_seq_timeout - wait for a specific sequence numbers * * @rdev: radeon device pointer * @target_seq: sequence number(s) we want to wait for * @intr: use interruptable sleep + * @timeout: maximum time to wait, or MAX_SCHEDULE_TIMEOUT for infinite wait * * Wait for the requested sequence number(s) to be written by any ring * (all asics). Sequnce number array is indexed by ring id. * @intr selects whether to use interruptable (true) or non-interruptable * (false) sleep when waiting for the sequence number. Helper function * for radeon_fence_wait_*(). - * Returns 0 if the sequence number has passed, error for all other cases. + * Returns remaining time if the sequence number has passed, 0 when + * the wait timeout, or an error for all other cases. * -EDEADLK is returned when a GPU lockup has been detected. */ -static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq, - bool intr) +static long radeon_fence_wait_seq_timeout(struct radeon_device *rdev, + u64 *target_seq, bool intr, + long timeout) { - uint64_t last_seq[RADEON_NUM_RINGS]; - bool signaled; - int i, r; - - while (!radeon_fence_any_seq_signaled(rdev, target_seq)) { + long r; + int i; - /* Save current sequence values, used to check for GPU lockups */ - for (i = 0; i < RADEON_NUM_RINGS; ++i) { - if (!target_seq[i]) - continue; + if (radeon_fence_any_seq_signaled(rdev, target_seq)) + return timeout; - last_seq[i] = atomic64_read(&rdev->fence_drv[i].last_seq); - trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]); - radeon_irq_kms_sw_irq_get(rdev, i); - } + /* enable IRQs and tracing */ + for (i = 0; i < RADEON_NUM_RINGS; ++i) { + if (!target_seq[i]) + continue; - if (intr) { - r = wait_event_interruptible_timeout(rdev->fence_queue, ( - (signaled = radeon_fence_any_seq_signaled(rdev, target_seq)) - || rdev->needs_reset), RADEON_FENCE_JIFFIES_TIMEOUT); - } else { - r = wait_event_timeout(rdev->fence_queue, ( - (signaled = radeon_fence_any_seq_signaled(rdev, target_seq)) - || rdev->needs_reset), RADEON_FENCE_JIFFIES_TIMEOUT); - } + trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]); + radeon_irq_kms_sw_irq_get(rdev, i); + } - for (i = 0; i < RADEON_NUM_RINGS; ++i) { - if (!target_seq[i]) - continue; + if (intr) { + r = wait_event_interruptible_timeout(rdev->fence_queue, ( + radeon_fence_any_seq_signaled(rdev, target_seq) + || rdev->needs_reset), timeout); + } else { + r = wait_event_timeout(rdev->fence_queue, ( + radeon_fence_any_seq_signaled(rdev, target_seq) + || rdev->needs_reset), timeout); + } - radeon_irq_kms_sw_irq_put(rdev, i); - trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]); - } + if (rdev->needs_reset) + r = -EDEADLK; - if (unlikely(r < 0)) - return r; + for (i = 0; i < RADEON_NUM_RINGS; ++i) { + if (!target_seq[i]) + continue; - if (unlikely(!signaled)) { - if (rdev->needs_reset) - return -EDEADLK; - - /* we were interrupted for some reason and fence - * isn't signaled yet, resume waiting */ - if (r) - continue; - - for (i = 0; i < RADEON_NUM_RINGS; ++i) { - if (!target_seq[i]) - continue; - - if (last_seq[i] != atomic64_read(&rdev->fence_drv[i].last_seq)) - break; - } - - if (i != RADEON_NUM_RINGS) - continue; - - for (i = 0; i < RADEON_NUM_RINGS; ++i) { - if (!target_seq[i]) - continue; - - if (radeon_ring_is_lockup(rdev, i, &rdev->ring[i])) - break; - } - - if (i < RADEON_NUM_RINGS) { - /* good news we believe it's a lockup */ - dev_warn(rdev->dev, "GPU lockup (waiting for " - "0x%016llx last fence id 0x%016llx on" - " ring %d)\n", - target_seq[i], last_seq[i], i); - - /* remember that we need an reset */ - rdev->needs_reset = true; - wake_up_all(&rdev->fence_queue); - return -EDEADLK; - } - } + radeon_irq_kms_sw_irq_put(rdev, i); + trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]); } - return 0; + + return r; } /** * radeon_fence_wait - wait for a fence to signal * * @fence: radeon fence object - * @intr: use interruptable sleep + * @intr: use interruptible sleep * * Wait for the requested fence to signal (all asics). * @intr selects whether to use interruptable (true) or non-interruptable @@ -396,22 +539,26 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq, int radeon_fence_wait(struct radeon_fence *fence, bool intr) { uint64_t seq[RADEON_NUM_RINGS] = {}; - int r; + long r; - if (fence == NULL) { - WARN(1, "Querying an invalid fence : %p !\n", fence); - return -EINVAL; - } + /* + * This function should not be called on !radeon fences. + * If this is the case, it would mean this function can + * also be called on radeon fences belonging to another card. + * exclusive_lock is not held in that case. + */ + if (WARN_ON_ONCE(!to_radeon_fence(&fence->base))) + return fence_wait(&fence->base, intr); seq[fence->ring] = fence->seq; - if (seq[fence->ring] == RADEON_FENCE_SIGNALED_SEQ) - return 0; - - r = radeon_fence_wait_seq(fence->rdev, seq, intr); - if (r) + r = radeon_fence_wait_seq_timeout(fence->rdev, seq, intr, MAX_SCHEDULE_TIMEOUT); + if (r < 0) { return r; + } - fence->seq = RADEON_FENCE_SIGNALED_SEQ; + r = fence_signal(&fence->base); + if (!r) + FENCE_TRACE(&fence->base, "signaled from fence_wait\n"); return 0; } @@ -434,7 +581,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev, { uint64_t seq[RADEON_NUM_RINGS]; unsigned i, num_rings = 0; - int r; + long r; for (i = 0; i < RADEON_NUM_RINGS; ++i) { seq[i] = 0; @@ -445,18 +592,14 @@ int radeon_fence_wait_any(struct radeon_device *rdev, seq[i] = fences[i]->seq; ++num_rings; - - /* test if something was allready signaled */ - if (seq[i] == RADEON_FENCE_SIGNALED_SEQ) - return 0; } /* nothing to wait for ? */ if (num_rings == 0) return -ENOENT; - r = radeon_fence_wait_seq(rdev, seq, intr); - if (r) { + r = radeon_fence_wait_seq_timeout(rdev, seq, intr, MAX_SCHEDULE_TIMEOUT); + if (r < 0) { return r; } return 0; @@ -475,6 +618,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev, int radeon_fence_wait_next(struct radeon_device *rdev, int ring) { uint64_t seq[RADEON_NUM_RINGS] = {}; + long r; seq[ring] = atomic64_read(&rdev->fence_drv[ring].last_seq) + 1ULL; if (seq[ring] >= rdev->fence_drv[ring].sync_seq[ring]) { @@ -482,7 +626,10 @@ int radeon_fence_wait_next(struct radeon_device *rdev, int ring) already the last emited fence */ return -ENOENT; } - return radeon_fence_wait_seq(rdev, seq, false); + r = radeon_fence_wait_seq_timeout(rdev, seq, false, MAX_SCHEDULE_TIMEOUT); + if (r < 0) + return r; + return 0; } /** @@ -498,18 +645,18 @@ int radeon_fence_wait_next(struct radeon_device *rdev, int ring) int radeon_fence_wait_empty(struct radeon_device *rdev, int ring) { uint64_t seq[RADEON_NUM_RINGS] = {}; - int r; + long r; seq[ring] = rdev->fence_drv[ring].sync_seq[ring]; if (!seq[ring]) return 0; - r = radeon_fence_wait_seq(rdev, seq, false); - if (r) { + r = radeon_fence_wait_seq_timeout(rdev, seq, false, MAX_SCHEDULE_TIMEOUT); + if (r < 0) { if (r == -EDEADLK) return -EDEADLK; - dev_err(rdev->dev, "error waiting for ring[%d] to become idle (%d)\n", + dev_err(rdev->dev, "error waiting for ring[%d] to become idle (%ld)\n", ring, r); } return 0; @@ -525,7 +672,7 @@ int radeon_fence_wait_empty(struct radeon_device *rdev, int ring) */ struct radeon_fence *radeon_fence_ref(struct radeon_fence *fence) { - kref_get(&fence->kref); + fence_get(&fence->base); return fence; } @@ -542,7 +689,7 @@ void radeon_fence_unref(struct radeon_fence **fence) *fence = NULL; if (tmp) { - kref_put(&tmp->kref, radeon_fence_destroy); + fence_put(&tmp->base); } } @@ -711,6 +858,9 @@ static void radeon_fence_driver_init_ring(struct radeon_device *rdev, int ring) rdev->fence_drv[ring].sync_seq[i] = 0; atomic64_set(&rdev->fence_drv[ring].last_seq, 0); rdev->fence_drv[ring].initialized = false; + INIT_DELAYED_WORK(&rdev->fence_drv[ring].lockup_work, + radeon_fence_check_lockup); + rdev->fence_drv[ring].rdev = rdev; } /** @@ -758,8 +908,9 @@ void radeon_fence_driver_fini(struct radeon_device *rdev) r = radeon_fence_wait_empty(rdev, ring); if (r) { /* no need to trigger GPU reset as we are unloading */ - radeon_fence_driver_force_completion(rdev); + radeon_fence_driver_force_completion(rdev, ring); } + cancel_delayed_work_sync(&rdev->fence_drv[ring].lockup_work); wake_up_all(&rdev->fence_queue); radeon_scratch_free(rdev, rdev->fence_drv[ring].scratch_reg); rdev->fence_drv[ring].initialized = false; @@ -771,18 +922,16 @@ void radeon_fence_driver_fini(struct radeon_device *rdev) * radeon_fence_driver_force_completion - force all fence waiter to complete * * @rdev: radeon device pointer + * @ring: the ring to complete * * In case of GPU reset failure make sure no process keep waiting on fence * that will never complete. */ -void radeon_fence_driver_force_completion(struct radeon_device *rdev) +void radeon_fence_driver_force_completion(struct radeon_device *rdev, int ring) { - int ring; - - for (ring = 0; ring < RADEON_NUM_RINGS; ring++) { - if (!rdev->fence_drv[ring].initialized) - continue; + if (rdev->fence_drv[ring].initialized) { radeon_fence_write(rdev, rdev->fence_drv[ring].sync_seq[ring], ring); + cancel_delayed_work_sync(&rdev->fence_drv[ring].lockup_work); } } @@ -833,6 +982,7 @@ static int radeon_debugfs_gpu_reset(struct seq_file *m, void *data) down_read(&rdev->exclusive_lock); seq_printf(m, "%d\n", rdev->needs_reset); rdev->needs_reset = true; + wake_up_all(&rdev->fence_queue); up_read(&rdev->exclusive_lock); return 0; @@ -852,3 +1002,72 @@ int radeon_debugfs_fence_init(struct radeon_device *rdev) return 0; #endif } + +static const char *radeon_fence_get_driver_name(struct fence *fence) +{ + return "radeon"; +} + +static const char *radeon_fence_get_timeline_name(struct fence *f) +{ + struct radeon_fence *fence = to_radeon_fence(f); + switch (fence->ring) { + case RADEON_RING_TYPE_GFX_INDEX: return "radeon.gfx"; + case CAYMAN_RING_TYPE_CP1_INDEX: return "radeon.cp1"; + case CAYMAN_RING_TYPE_CP2_INDEX: return "radeon.cp2"; + case R600_RING_TYPE_DMA_INDEX: return "radeon.dma"; + case CAYMAN_RING_TYPE_DMA1_INDEX: return "radeon.dma1"; + case R600_RING_TYPE_UVD_INDEX: return "radeon.uvd"; + case TN_RING_TYPE_VCE1_INDEX: return "radeon.vce1"; + case TN_RING_TYPE_VCE2_INDEX: return "radeon.vce2"; + default: WARN_ON_ONCE(1); return "radeon.unk"; + } +} + +static inline bool radeon_test_signaled(struct radeon_fence *fence) +{ + return test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->base.flags); +} + +static signed long radeon_fence_default_wait(struct fence *f, bool intr, + signed long t) +{ + struct radeon_fence *fence = to_radeon_fence(f); + struct radeon_device *rdev = fence->rdev; + bool signaled; + + fence_enable_sw_signaling(&fence->base); + + /* + * This function has to return -EDEADLK, but cannot hold + * exclusive_lock during the wait because some callers + * may already hold it. This means checking needs_reset without + * lock, and not fiddling with any gpu internals. + * + * The callback installed with fence_enable_sw_signaling will + * run before our wait_event_*timeout call, so we will see + * both the signaled fence and the changes to needs_reset. + */ + + if (intr) + t = wait_event_interruptible_timeout(rdev->fence_queue, + ((signaled = radeon_test_signaled(fence)) || + rdev->needs_reset), t); + else + t = wait_event_timeout(rdev->fence_queue, + ((signaled = radeon_test_signaled(fence)) || + rdev->needs_reset), t); + + if (t > 0 && !signaled) + return -EDEADLK; + return t; +} + +const struct fence_ops radeon_fence_ops = { + .get_driver_name = radeon_fence_get_driver_name, + .get_timeline_name = radeon_fence_get_timeline_name, + .enable_signaling = radeon_fence_enable_signaling, + .signaled = radeon_fence_is_signaled, + .wait = radeon_fence_default_wait, + .release = NULL, +}; diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c index a053a0779aac..84146d5901aa 100644 --- a/drivers/gpu/drm/radeon/radeon_gart.c +++ b/drivers/gpu/drm/radeon/radeon_gart.c @@ -128,7 +128,7 @@ int radeon_gart_table_vram_alloc(struct radeon_device *rdev) if (rdev->gart.robj == NULL) { r = radeon_bo_create(rdev, rdev->gart.table_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, - 0, NULL, &rdev->gart.robj); + 0, NULL, NULL, &rdev->gart.robj); if (r) { return r; } diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index bfd7e1b0ff3f..c194497aa586 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -67,7 +67,7 @@ int radeon_gem_object_create(struct radeon_device *rdev, unsigned long size, retry: r = radeon_bo_create(rdev, size, alignment, kernel, initial_domain, - flags, NULL, &robj); + flags, NULL, NULL, &robj); if (r) { if (r != -ERESTARTSYS) { if (initial_domain == RADEON_GEM_DOMAIN_VRAM) { @@ -94,7 +94,7 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj, { struct radeon_bo *robj; uint32_t domain; - int r; + long r; /* FIXME: reeimplement */ robj = gem_to_radeon_bo(gobj); @@ -110,9 +110,12 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj, } if (domain == RADEON_GEM_DOMAIN_CPU) { /* Asking for cpu access wait for object idle */ - r = radeon_bo_wait(robj, NULL, false); - if (r) { - printk(KERN_ERR "Failed to wait for object !\n"); + r = reservation_object_wait_timeout_rcu(robj->tbo.resv, true, true, 30 * HZ); + if (!r) + r = -EBUSY; + + if (r < 0 && r != -EINTR) { + printk(KERN_ERR "Failed to wait for object: %li\n", r); return r; } } @@ -272,6 +275,94 @@ int radeon_gem_create_ioctl(struct drm_device *dev, void *data, return 0; } +int radeon_gem_userptr_ioctl(struct drm_device *dev, void *data, + struct drm_file *filp) +{ + struct radeon_device *rdev = dev->dev_private; + struct drm_radeon_gem_userptr *args = data; + struct drm_gem_object *gobj; + struct radeon_bo *bo; + uint32_t handle; + int r; + + if (offset_in_page(args->addr | args->size)) + return -EINVAL; + + /* reject unknown flag values */ + if (args->flags & ~(RADEON_GEM_USERPTR_READONLY | + RADEON_GEM_USERPTR_ANONONLY | RADEON_GEM_USERPTR_VALIDATE | + RADEON_GEM_USERPTR_REGISTER)) + return -EINVAL; + + if (args->flags & RADEON_GEM_USERPTR_READONLY) { + /* readonly pages not tested on older hardware */ + if (rdev->family < CHIP_R600) + return -EINVAL; + + } else if (!(args->flags & RADEON_GEM_USERPTR_ANONONLY) || + !(args->flags & RADEON_GEM_USERPTR_REGISTER)) { + + /* if we want to write to it we must require anonymous + memory and install a MMU notifier */ + return -EACCES; + } + + down_read(&rdev->exclusive_lock); + + /* create a gem object to contain this object in */ + r = radeon_gem_object_create(rdev, args->size, 0, + RADEON_GEM_DOMAIN_CPU, 0, + false, &gobj); + if (r) + goto handle_lockup; + + bo = gem_to_radeon_bo(gobj); + r = radeon_ttm_tt_set_userptr(bo->tbo.ttm, args->addr, args->flags); + if (r) + goto release_object; + + if (args->flags & RADEON_GEM_USERPTR_REGISTER) { + r = radeon_mn_register(bo, args->addr); + if (r) + goto release_object; + } + + if (args->flags & RADEON_GEM_USERPTR_VALIDATE) { + down_read(¤t->mm->mmap_sem); + r = radeon_bo_reserve(bo, true); + if (r) { + up_read(¤t->mm->mmap_sem); + goto release_object; + } + + radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_GTT); + r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false); + radeon_bo_unreserve(bo); + up_read(¤t->mm->mmap_sem); + if (r) + goto release_object; + } + + r = drm_gem_handle_create(filp, gobj, &handle); + /* drop reference from allocate - handle holds it now */ + drm_gem_object_unreference_unlocked(gobj); + if (r) + goto handle_lockup; + + args->handle = handle; + up_read(&rdev->exclusive_lock); + return 0; + +release_object: + drm_gem_object_unreference_unlocked(gobj); + +handle_lockup: + up_read(&rdev->exclusive_lock); + r = radeon_gem_handle_lockup(rdev, r); + + return r; +} + int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) { @@ -315,6 +406,10 @@ int radeon_mode_dumb_mmap(struct drm_file *filp, return -ENOENT; } robj = gem_to_radeon_bo(gobj); + if (radeon_ttm_tt_has_userptr(robj->tbo.ttm)) { + drm_gem_object_unreference_unlocked(gobj); + return -EPERM; + } *offset_p = radeon_bo_mmap_offset(robj); drm_gem_object_unreference_unlocked(gobj); return 0; @@ -357,15 +452,22 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data, struct drm_radeon_gem_wait_idle *args = data; struct drm_gem_object *gobj; struct radeon_bo *robj; - int r; + int r = 0; uint32_t cur_placement = 0; + long ret; gobj = drm_gem_object_lookup(dev, filp, args->handle); if (gobj == NULL) { return -ENOENT; } robj = gem_to_radeon_bo(gobj); - r = radeon_bo_wait(robj, &cur_placement, false); + + ret = reservation_object_wait_timeout_rcu(robj->tbo.resv, true, true, 30 * HZ); + if (ret == 0) + r = -EBUSY; + else if (ret < 0) + r = ret; + /* Flush HDP cache via MMIO if necessary */ if (rdev->asic->mmio_hdp_flush && radeon_mem_type_to_domain(cur_placement) == RADEON_GEM_DOMAIN_VRAM) @@ -532,6 +634,11 @@ int radeon_gem_op_ioctl(struct drm_device *dev, void *data, return -ENOENT; } robj = gem_to_radeon_bo(gobj); + + r = -EPERM; + if (radeon_ttm_tt_has_userptr(robj->tbo.ttm)) + goto out; + r = radeon_bo_reserve(robj, false); if (unlikely(r)) goto out; diff --git a/drivers/gpu/drm/radeon/radeon_ib.c b/drivers/gpu/drm/radeon/radeon_ib.c index 5bf2c0a05827..3f39fcca4d07 100644 --- a/drivers/gpu/drm/radeon/radeon_ib.c +++ b/drivers/gpu/drm/radeon/radeon_ib.c @@ -145,7 +145,7 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib, if (ib->vm) { struct radeon_fence *vm_id_fence; vm_id_fence = radeon_vm_grab_id(rdev, ib->vm, ib->ring); - radeon_semaphore_sync_to(ib->semaphore, vm_id_fence); + radeon_semaphore_sync_fence(ib->semaphore, vm_id_fence); } /* sync with other rings */ @@ -269,6 +269,7 @@ int radeon_ib_ring_tests(struct radeon_device *rdev) r = radeon_ib_test(rdev, i, ring); if (r) { + radeon_fence_driver_force_completion(rdev, i); ring->ready = false; rdev->needs_reset = false; diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c index 16807afab362..7784911d78ef 100644 --- a/drivers/gpu/drm/radeon/radeon_irq_kms.c +++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c @@ -88,23 +88,6 @@ static void radeon_hotplug_work_func(struct work_struct *work) } /** - * radeon_irq_reset_work_func - execute gpu reset - * - * @work: work struct - * - * Execute scheduled gpu reset (cayman+). - * This function is called when the irq handler - * thinks we need a gpu reset. - */ -static void radeon_irq_reset_work_func(struct work_struct *work) -{ - struct radeon_device *rdev = container_of(work, struct radeon_device, - reset_work); - - radeon_gpu_reset(rdev); -} - -/** * radeon_driver_irq_preinstall_kms - drm irq preinstall callback * * @dev: drm dev pointer @@ -284,7 +267,6 @@ int radeon_irq_kms_init(struct radeon_device *rdev) INIT_WORK(&rdev->hotplug_work, radeon_hotplug_work_func); INIT_WORK(&rdev->audio_work, r600_audio_update_hdmi); - INIT_WORK(&rdev->reset_work, radeon_irq_reset_work_func); rdev->irq.installed = true; r = drm_irq_install(rdev->ddev, rdev->ddev->pdev->irq); @@ -342,6 +324,21 @@ void radeon_irq_kms_sw_irq_get(struct radeon_device *rdev, int ring) } /** + * radeon_irq_kms_sw_irq_get_delayed - enable software interrupt + * + * @rdev: radeon device pointer + * @ring: ring whose interrupt you want to enable + * + * Enables the software interrupt for a specific ring (all asics). + * The software interrupt is generally used to signal a fence on + * a particular ring. + */ +bool radeon_irq_kms_sw_irq_get_delayed(struct radeon_device *rdev, int ring) +{ + return atomic_inc_return(&rdev->irq.ring_int[ring]) == 1; +} + +/** * radeon_irq_kms_sw_irq_put - disable software interrupt * * @rdev: radeon device pointer diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index eb7164d07985..8309b11e674d 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -885,5 +885,6 @@ const struct drm_ioctl_desc radeon_ioctls_kms[] = { DRM_IOCTL_DEF_DRV(RADEON_GEM_BUSY, radeon_gem_busy_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(RADEON_GEM_VA, radeon_gem_va_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(RADEON_GEM_OP, radeon_gem_op_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(RADEON_GEM_USERPTR, radeon_gem_userptr_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW), }; int radeon_max_kms_ioctl = ARRAY_SIZE(radeon_ioctls_kms); diff --git a/drivers/gpu/drm/radeon/radeon_mn.c b/drivers/gpu/drm/radeon/radeon_mn.c new file mode 100644 index 000000000000..a69bd441dd2d --- /dev/null +++ b/drivers/gpu/drm/radeon/radeon_mn.c @@ -0,0 +1,274 @@ +/* + * Copyright 2014 Advanced Micro Devices, Inc. + * All Rights Reserved. + * + * 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, sub license, 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 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 NON-INFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDERS, AUTHORS 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. + * + * The above copyright notice and this permission notice (including the + * next paragraph) shall be included in all copies or substantial portions + * of the Software. + * + */ +/* + * Authors: + * Christian König <christian.koenig@amd.com> + */ + +#include <linux/firmware.h> +#include <linux/module.h> +#include <linux/mmu_notifier.h> +#include <drm/drmP.h> +#include <drm/drm.h> + +#include "radeon.h" + +struct radeon_mn { + /* constant after initialisation */ + struct radeon_device *rdev; + struct mm_struct *mm; + struct mmu_notifier mn; + + /* only used on destruction */ + struct work_struct work; + + /* protected by rdev->mn_lock */ + struct hlist_node node; + + /* objects protected by lock */ + struct mutex lock; + struct rb_root objects; +}; + +/** + * radeon_mn_destroy - destroy the rmn + * + * @work: previously sheduled work item + * + * Lazy destroys the notifier from a work item + */ +static void radeon_mn_destroy(struct work_struct *work) +{ + struct radeon_mn *rmn = container_of(work, struct radeon_mn, work); + struct radeon_device *rdev = rmn->rdev; + struct radeon_bo *bo, *next; + + mutex_lock(&rdev->mn_lock); + mutex_lock(&rmn->lock); + hash_del(&rmn->node); + rbtree_postorder_for_each_entry_safe(bo, next, &rmn->objects, mn_it.rb) { + interval_tree_remove(&bo->mn_it, &rmn->objects); + bo->mn = NULL; + } + mutex_unlock(&rmn->lock); + mutex_unlock(&rdev->mn_lock); + mmu_notifier_unregister(&rmn->mn, rmn->mm); + kfree(rmn); +} + +/** + * radeon_mn_release - callback to notify about mm destruction + * + * @mn: our notifier + * @mn: the mm this callback is about + * + * Shedule a work item to lazy destroy our notifier. + */ +static void radeon_mn_release(struct mmu_notifier *mn, + struct mm_struct *mm) +{ + struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn); + INIT_WORK(&rmn->work, radeon_mn_destroy); + schedule_work(&rmn->work); +} + +/** + * radeon_mn_invalidate_range_start - callback to notify about mm change + * + * @mn: our notifier + * @mn: the mm this callback is about + * @start: start of updated range + * @end: end of updated range + * + * We block for all BOs between start and end to be idle and + * unmap them by move them into system domain again. + */ +static void radeon_mn_invalidate_range_start(struct mmu_notifier *mn, + struct mm_struct *mm, + unsigned long start, + unsigned long end) +{ + struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn); + struct interval_tree_node *it; + + /* notification is exclusive, but interval is inclusive */ + end -= 1; + + mutex_lock(&rmn->lock); + + it = interval_tree_iter_first(&rmn->objects, start, end); + while (it) { + struct radeon_bo *bo; + struct fence *fence; + int r; + + bo = container_of(it, struct radeon_bo, mn_it); + it = interval_tree_iter_next(it, start, end); + + r = radeon_bo_reserve(bo, true); + if (r) { + DRM_ERROR("(%d) failed to reserve user bo\n", r); + continue; + } + + fence = reservation_object_get_excl(bo->tbo.resv); + if (fence) { + r = radeon_fence_wait((struct radeon_fence *)fence, false); + if (r) + DRM_ERROR("(%d) failed to wait for user bo\n", r); + } + + radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_CPU); + r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false); + if (r) + DRM_ERROR("(%d) failed to validate user bo\n", r); + + radeon_bo_unreserve(bo); + } + + mutex_unlock(&rmn->lock); +} + +static const struct mmu_notifier_ops radeon_mn_ops = { + .release = radeon_mn_release, + .invalidate_range_start = radeon_mn_invalidate_range_start, +}; + +/** + * radeon_mn_get - create notifier context + * + * @rdev: radeon device pointer + * + * Creates a notifier context for current->mm. + */ +static struct radeon_mn *radeon_mn_get(struct radeon_device *rdev) +{ + struct mm_struct *mm = current->mm; + struct radeon_mn *rmn; + int r; + + down_write(&mm->mmap_sem); + mutex_lock(&rdev->mn_lock); + + hash_for_each_possible(rdev->mn_hash, rmn, node, (unsigned long)mm) + if (rmn->mm == mm) + goto release_locks; + + rmn = kzalloc(sizeof(*rmn), GFP_KERNEL); + if (!rmn) { + rmn = ERR_PTR(-ENOMEM); + goto release_locks; + } + + rmn->rdev = rdev; + rmn->mm = mm; + rmn->mn.ops = &radeon_mn_ops; + mutex_init(&rmn->lock); + rmn->objects = RB_ROOT; + + r = __mmu_notifier_register(&rmn->mn, mm); + if (r) + goto free_rmn; + + hash_add(rdev->mn_hash, &rmn->node, (unsigned long)mm); + +release_locks: + mutex_unlock(&rdev->mn_lock); + up_write(&mm->mmap_sem); + + return rmn; + +free_rmn: + mutex_unlock(&rdev->mn_lock); + up_write(&mm->mmap_sem); + kfree(rmn); + + return ERR_PTR(r); +} + +/** + * radeon_mn_register - register a BO for notifier updates + * + * @bo: radeon buffer object + * @addr: userptr addr we should monitor + * + * Registers an MMU notifier for the given BO at the specified address. + * Returns 0 on success, -ERRNO if anything goes wrong. + */ +int radeon_mn_register(struct radeon_bo *bo, unsigned long addr) +{ + unsigned long end = addr + radeon_bo_size(bo) - 1; + struct radeon_device *rdev = bo->rdev; + struct radeon_mn *rmn; + struct interval_tree_node *it; + + rmn = radeon_mn_get(rdev); + if (IS_ERR(rmn)) + return PTR_ERR(rmn); + + mutex_lock(&rmn->lock); + + it = interval_tree_iter_first(&rmn->objects, addr, end); + if (it) { + mutex_unlock(&rmn->lock); + return -EEXIST; + } + + bo->mn = rmn; + bo->mn_it.start = addr; + bo->mn_it.last = end; + interval_tree_insert(&bo->mn_it, &rmn->objects); + + mutex_unlock(&rmn->lock); + + return 0; +} + +/** + * radeon_mn_unregister - unregister a BO for notifier updates + * + * @bo: radeon buffer object + * + * Remove any registration of MMU notifier updates from the buffer object. + */ +void radeon_mn_unregister(struct radeon_bo *bo) +{ + struct radeon_device *rdev = bo->rdev; + struct radeon_mn *rmn; + + mutex_lock(&rdev->mn_lock); + rmn = bo->mn; + if (rmn == NULL) { + mutex_unlock(&rdev->mn_lock); + return; + } + + mutex_lock(&rmn->lock); + interval_tree_remove(&bo->mn_it, &rmn->objects); + bo->mn = NULL; + mutex_unlock(&rmn->lock); + mutex_unlock(&rdev->mn_lock); +} diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index e27608c29c11..04db2fdd8692 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -777,6 +777,7 @@ extern void atombios_digital_setup(struct drm_encoder *encoder, int action); extern int atombios_get_encoder_mode(struct drm_encoder *encoder); extern bool atombios_set_edp_panel_power(struct drm_connector *connector, int action); extern void radeon_encoder_set_active_device(struct drm_encoder *encoder); +extern bool radeon_encoder_is_digital(struct drm_encoder *encoder); extern void radeon_crtc_load_lut(struct drm_crtc *crtc); extern int atombios_crtc_set_base(struct drm_crtc *crtc, int x, int y, diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 480c87d8edc5..99a960a4f302 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -75,6 +75,7 @@ static void radeon_ttm_bo_destroy(struct ttm_buffer_object *tbo) bo = container_of(tbo, struct radeon_bo, tbo); radeon_update_memory_usage(bo, bo->tbo.mem.mem_type, -1); + radeon_mn_unregister(bo); mutex_lock(&bo->rdev->gem.mutex); list_del_init(&bo->list); @@ -96,55 +97,80 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain) { u32 c = 0, i; - rbo->placement.fpfn = 0; - rbo->placement.lpfn = 0; rbo->placement.placement = rbo->placements; rbo->placement.busy_placement = rbo->placements; if (domain & RADEON_GEM_DOMAIN_VRAM) - rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | - TTM_PL_FLAG_VRAM; + rbo->placements[c++].flags = TTM_PL_FLAG_WC | + TTM_PL_FLAG_UNCACHED | + TTM_PL_FLAG_VRAM; + if (domain & RADEON_GEM_DOMAIN_GTT) { if (rbo->flags & RADEON_GEM_GTT_UC) { - rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_TT; + rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED | + TTM_PL_FLAG_TT; + } else if ((rbo->flags & RADEON_GEM_GTT_WC) || (rbo->rdev->flags & RADEON_IS_AGP)) { - rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | + rbo->placements[c++].flags = TTM_PL_FLAG_WC | + TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_TT; } else { - rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_TT; + rbo->placements[c++].flags = TTM_PL_FLAG_CACHED | + TTM_PL_FLAG_TT; } } + if (domain & RADEON_GEM_DOMAIN_CPU) { if (rbo->flags & RADEON_GEM_GTT_UC) { - rbo->placements[c++] = TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_SYSTEM; + rbo->placements[c++].flags = TTM_PL_FLAG_UNCACHED | + TTM_PL_FLAG_SYSTEM; + } else if ((rbo->flags & RADEON_GEM_GTT_WC) || rbo->rdev->flags & RADEON_IS_AGP) { - rbo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | + rbo->placements[c++].flags = TTM_PL_FLAG_WC | + TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_SYSTEM; } else { - rbo->placements[c++] = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM; + rbo->placements[c++].flags = TTM_PL_FLAG_CACHED | + TTM_PL_FLAG_SYSTEM; } } if (!c) - rbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM; + rbo->placements[c++].flags = TTM_PL_MASK_CACHING | + TTM_PL_FLAG_SYSTEM; + rbo->placement.num_placement = c; rbo->placement.num_busy_placement = c; + for (i = 0; i < c; ++i) { + rbo->placements[i].fpfn = 0; + if ((rbo->flags & RADEON_GEM_CPU_ACCESS) && + (rbo->placements[i].flags & TTM_PL_FLAG_VRAM)) + rbo->placements[i].lpfn = + rbo->rdev->mc.visible_vram_size >> PAGE_SHIFT; + else + rbo->placements[i].lpfn = 0; + } + /* * Use two-ended allocation depending on the buffer size to * improve fragmentation quality. * 512kb was measured as the most optimal number. */ - if (rbo->tbo.mem.size > 512 * 1024) { + if (!((rbo->flags & RADEON_GEM_CPU_ACCESS) && + (rbo->placements[i].flags & TTM_PL_FLAG_VRAM)) && + rbo->tbo.mem.size > 512 * 1024) { for (i = 0; i < c; i++) { - rbo->placements[i] |= TTM_PL_FLAG_TOPDOWN; + rbo->placements[i].flags |= TTM_PL_FLAG_TOPDOWN; } } } int radeon_bo_create(struct radeon_device *rdev, - unsigned long size, int byte_align, bool kernel, u32 domain, - u32 flags, struct sg_table *sg, struct radeon_bo **bo_ptr) + unsigned long size, int byte_align, bool kernel, + u32 domain, u32 flags, struct sg_table *sg, + struct reservation_object *resv, + struct radeon_bo **bo_ptr) { struct radeon_bo *bo; enum ttm_bo_type type; @@ -192,7 +218,7 @@ int radeon_bo_create(struct radeon_device *rdev, down_read(&rdev->pm.mclk_lock); r = ttm_bo_init(&rdev->mman.bdev, &bo->tbo, size, type, &bo->placement, page_align, !kernel, NULL, - acc_size, sg, &radeon_ttm_bo_destroy); + acc_size, sg, resv, &radeon_ttm_bo_destroy); up_read(&rdev->pm.mclk_lock); if (unlikely(r != 0)) { return r; @@ -264,6 +290,9 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset, { int r, i; + if (radeon_ttm_tt_has_userptr(bo->tbo.ttm)) + return -EPERM; + if (bo->pin_count) { bo->pin_count++; if (gpu_addr) @@ -283,21 +312,19 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset, return 0; } radeon_ttm_placement_from_domain(bo, domain); - if (domain == RADEON_GEM_DOMAIN_VRAM) { + for (i = 0; i < bo->placement.num_placement; i++) { /* force to pin into visible video ram */ - bo->placement.lpfn = bo->rdev->mc.visible_vram_size >> PAGE_SHIFT; - } - if (max_offset) { - u64 lpfn = max_offset >> PAGE_SHIFT; - - if (!bo->placement.lpfn) - bo->placement.lpfn = bo->rdev->mc.gtt_size >> PAGE_SHIFT; + if ((bo->placements[i].flags & TTM_PL_FLAG_VRAM) && + !(bo->flags & RADEON_GEM_NO_CPU_ACCESS) && + (!max_offset || max_offset > bo->rdev->mc.visible_vram_size)) + bo->placements[i].lpfn = + bo->rdev->mc.visible_vram_size >> PAGE_SHIFT; + else + bo->placements[i].lpfn = max_offset >> PAGE_SHIFT; - if (lpfn < bo->placement.lpfn) - bo->placement.lpfn = lpfn; + bo->placements[i].flags |= TTM_PL_FLAG_NO_EVICT; } - for (i = 0; i < bo->placement.num_placement; i++) - bo->placements[i] |= TTM_PL_FLAG_NO_EVICT; + r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false); if (likely(r == 0)) { bo->pin_count = 1; @@ -329,8 +356,10 @@ int radeon_bo_unpin(struct radeon_bo *bo) bo->pin_count--; if (bo->pin_count) return 0; - for (i = 0; i < bo->placement.num_placement; i++) - bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT; + for (i = 0; i < bo->placement.num_placement; i++) { + bo->placements[i].lpfn = 0; + bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT; + } r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false); if (likely(r == 0)) { if (bo->tbo.mem.mem_type == TTM_PL_VRAM) @@ -459,7 +488,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev, u64 bytes_moved = 0, initial_bytes_moved; u64 bytes_moved_threshold = radeon_bo_get_threshold_for_moves(rdev); - r = ttm_eu_reserve_buffers(ticket, head); + r = ttm_eu_reserve_buffers(ticket, head, true); if (unlikely(r != 0)) { return r; } @@ -468,6 +497,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev, bo = lobj->robj; if (!bo->pin_count) { u32 domain = lobj->prefered_domains; + u32 allowed = lobj->allowed_domains; u32 current_domain = radeon_mem_type_to_domain(bo->tbo.mem.mem_type); @@ -479,7 +509,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev, * into account. We don't want to disallow buffer moves * completely. */ - if ((lobj->allowed_domains & current_domain) != 0 && + if ((allowed & current_domain) != 0 && (domain & current_domain) == 0 && /* will be moved */ bytes_moved > bytes_moved_threshold) { /* don't move it */ @@ -489,7 +519,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev, retry: radeon_ttm_placement_from_domain(bo, domain); if (ring == R600_RING_TYPE_UVD_INDEX) - radeon_uvd_force_into_uvd_segment(bo); + radeon_uvd_force_into_uvd_segment(bo, allowed); initial_bytes_moved = atomic64_read(&rdev->num_bytes_moved); r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false); @@ -731,7 +761,7 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo) /* hurrah the memory is not visible ! */ radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM); - rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT; + rbo->placements[0].lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT; r = ttm_bo_validate(bo, &rbo->placement, false, false); if (unlikely(r == -ENOMEM)) { radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT); @@ -755,12 +785,10 @@ int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type, bool no_wait) r = ttm_bo_reserve(&bo->tbo, true, no_wait, false, NULL); if (unlikely(r != 0)) return r; - spin_lock(&bo->tbo.bdev->fence_lock); if (mem_type) *mem_type = bo->tbo.mem.mem_type; - if (bo->tbo.sync_obj) - r = ttm_bo_wait(&bo->tbo, true, true, no_wait); - spin_unlock(&bo->tbo.bdev->fence_lock); + + r = ttm_bo_wait(&bo->tbo, true, true, no_wait); ttm_bo_unreserve(&bo->tbo); return r; } diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h index 98a47fdf3625..1b8ec7917154 100644 --- a/drivers/gpu/drm/radeon/radeon_object.h +++ b/drivers/gpu/drm/radeon/radeon_object.h @@ -126,6 +126,7 @@ extern int radeon_bo_create(struct radeon_device *rdev, unsigned long size, int byte_align, bool kernel, u32 domain, u32 flags, struct sg_table *sg, + struct reservation_object *resv, 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); diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index 164898b0010c..32522cc940a1 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -1556,7 +1556,7 @@ static bool radeon_pm_in_vbl(struct radeon_device *rdev) if (rdev->pm.active_crtcs & (1 << crtc)) { vbl_status = radeon_get_crtc_scanoutpos(rdev->ddev, crtc, 0, &vpos, &hpos, NULL, NULL); if ((vbl_status & DRM_SCANOUTPOS_VALID) && - !(vbl_status & DRM_SCANOUTPOS_INVBL)) + !(vbl_status & DRM_SCANOUTPOS_IN_VBLANK)) in_vbl = false; } } diff --git a/drivers/gpu/drm/radeon/radeon_prime.c b/drivers/gpu/drm/radeon/radeon_prime.c index 0b16f2cbcf17..f3609c97496b 100644 --- a/drivers/gpu/drm/radeon/radeon_prime.c +++ b/drivers/gpu/drm/radeon/radeon_prime.c @@ -27,6 +27,7 @@ #include "radeon.h" #include <drm/radeon_drm.h> +#include <linux/dma-buf.h> struct sg_table *radeon_gem_prime_get_sg_table(struct drm_gem_object *obj) { @@ -57,15 +58,18 @@ void radeon_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr) } struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev, - size_t size, + struct dma_buf_attachment *attach, struct sg_table *sg) { + struct reservation_object *resv = attach->dmabuf->resv; struct radeon_device *rdev = dev->dev_private; struct radeon_bo *bo; int ret; - ret = radeon_bo_create(rdev, size, PAGE_SIZE, false, - RADEON_GEM_DOMAIN_GTT, 0, sg, &bo); + ww_mutex_lock(&resv->lock, NULL); + ret = radeon_bo_create(rdev, attach->dmabuf->size, PAGE_SIZE, false, + RADEON_GEM_DOMAIN_GTT, 0, sg, resv, &bo); + ww_mutex_unlock(&resv->lock); if (ret) return ERR_PTR(ret); @@ -111,3 +115,13 @@ struct reservation_object *radeon_gem_prime_res_obj(struct drm_gem_object *obj) return bo->tbo.resv; } + +struct dma_buf *radeon_gem_prime_export(struct drm_device *dev, + struct drm_gem_object *gobj, + int flags) +{ + struct radeon_bo *bo = gem_to_radeon_bo(gobj); + if (radeon_ttm_tt_has_userptr(bo->tbo.ttm)) + return ERR_PTR(-EPERM); + return drm_gem_prime_export(dev, gobj, flags); +} diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index d65607902537..3d17af34afa7 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -45,27 +45,6 @@ static int radeon_debugfs_ring_init(struct radeon_device *rdev, struct radeon_ring *ring); /** - * radeon_ring_write - write a value to the ring - * - * @ring: radeon_ring structure holding ring information - * @v: dword (dw) value to write - * - * Write a value to the requested ring buffer (all asics). - */ -void radeon_ring_write(struct radeon_ring *ring, uint32_t v) -{ -#if DRM_DEBUG_CODE - if (ring->count_dw <= 0) { - DRM_ERROR("radeon: writing more dwords to the ring than expected!\n"); - } -#endif - ring->ring[ring->wptr++] = v; - ring->wptr &= ring->ptr_mask; - ring->count_dw--; - ring->ring_free_dw--; -} - -/** * radeon_ring_supports_scratch_reg - check if the ring supports * writing to scratch registers * @@ -404,7 +383,7 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig /* Allocate ring buffer */ if (ring->ring_obj == NULL) { r = radeon_bo_create(rdev, ring->ring_size, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_GTT, 0, + RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL, &ring->ring_obj); if (r) { dev_err(rdev->dev, "(%d) ring create failed\n", r); diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c index b84f97c8718c..c507896aca45 100644 --- a/drivers/gpu/drm/radeon/radeon_sa.c +++ b/drivers/gpu/drm/radeon/radeon_sa.c @@ -65,7 +65,7 @@ int radeon_sa_bo_manager_init(struct radeon_device *rdev, } r = radeon_bo_create(rdev, size, align, true, - domain, flags, NULL, &sa_manager->bo); + domain, flags, NULL, NULL, &sa_manager->bo); if (r) { dev_err(rdev->dev, "(%d) failed to allocate bo for manager\n", r); return r; diff --git a/drivers/gpu/drm/radeon/radeon_semaphore.c b/drivers/gpu/drm/radeon/radeon_semaphore.c index abd6753a570a..6deb08f045b7 100644 --- a/drivers/gpu/drm/radeon/radeon_semaphore.c +++ b/drivers/gpu/drm/radeon/radeon_semaphore.c @@ -96,15 +96,15 @@ bool radeon_semaphore_emit_wait(struct radeon_device *rdev, int ridx, } /** - * radeon_semaphore_sync_to - use the semaphore to sync to a fence + * radeon_semaphore_sync_fence - use the semaphore to sync to a fence * * @semaphore: semaphore object to add fence to * @fence: fence to sync to * * Sync to the fence using this semaphore object */ -void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore, - struct radeon_fence *fence) +void radeon_semaphore_sync_fence(struct radeon_semaphore *semaphore, + struct radeon_fence *fence) { struct radeon_fence *other; @@ -116,6 +116,53 @@ void radeon_semaphore_sync_to(struct radeon_semaphore *semaphore, } /** + * radeon_semaphore_sync_to - use the semaphore to sync to a reservation object + * + * @sema: semaphore object to add fence from reservation object to + * @resv: reservation object with embedded fence + * @shared: true if we should onyl sync to the exclusive fence + * + * Sync to the fence using this semaphore object + */ +int radeon_semaphore_sync_resv(struct radeon_device *rdev, + struct radeon_semaphore *sema, + struct reservation_object *resv, + bool shared) +{ + struct reservation_object_list *flist; + struct fence *f; + struct radeon_fence *fence; + unsigned i; + int r = 0; + + /* always sync to the exclusive fence */ + f = reservation_object_get_excl(resv); + fence = f ? to_radeon_fence(f) : NULL; + if (fence && fence->rdev == rdev) + radeon_semaphore_sync_fence(sema, fence); + else if (f) + r = fence_wait(f, true); + + flist = reservation_object_get_list(resv); + if (shared || !flist || r) + return r; + + for (i = 0; i < flist->shared_count; ++i) { + f = rcu_dereference_protected(flist->shared[i], + reservation_object_held(resv)); + fence = to_radeon_fence(f); + if (fence && fence->rdev == rdev) + radeon_semaphore_sync_fence(sema, fence); + else + r = fence_wait(f, true); + + if (r) + break; + } + return r; +} + +/** * radeon_semaphore_sync_rings - sync ring to all registered fences * * @rdev: radeon_device pointer diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c index 23bb64fd775f..535403e0c8a2 100644 --- a/drivers/gpu/drm/radeon/radeon_state.c +++ b/drivers/gpu/drm/radeon/radeon_state.c @@ -30,9 +30,9 @@ */ #include <drm/drmP.h> -#include <drm/drm_buffer.h> #include <drm/radeon_drm.h> #include "radeon_drv.h" +#include "drm_buffer.h" /* ================================================================ * Helper functions for client state checking and fixup diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c index 17bc3dced9f1..07b506b41008 100644 --- a/drivers/gpu/drm/radeon/radeon_test.c +++ b/drivers/gpu/drm/radeon/radeon_test.c @@ -67,7 +67,7 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag) } r = radeon_bo_create(rdev, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM, - 0, NULL, &vram_obj); + 0, NULL, NULL, &vram_obj); if (r) { DRM_ERROR("Failed to create VRAM object\n"); goto out_cleanup; @@ -87,7 +87,8 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag) struct radeon_fence *fence = NULL; r = radeon_bo_create(rdev, size, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_GTT, 0, NULL, gtt_obj + i); + RADEON_GEM_DOMAIN_GTT, 0, NULL, NULL, + gtt_obj + i); if (r) { DRM_ERROR("Failed to create GTT object %d\n", i); goto out_lclean; @@ -116,11 +117,16 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag) radeon_bo_kunmap(gtt_obj[i]); if (ring == R600_RING_TYPE_DMA_INDEX) - r = radeon_copy_dma(rdev, gtt_addr, vram_addr, size / RADEON_GPU_PAGE_SIZE, &fence); + fence = radeon_copy_dma(rdev, gtt_addr, vram_addr, + size / RADEON_GPU_PAGE_SIZE, + NULL); else - r = radeon_copy_blit(rdev, gtt_addr, vram_addr, size / RADEON_GPU_PAGE_SIZE, &fence); - if (r) { + fence = radeon_copy_blit(rdev, gtt_addr, vram_addr, + size / RADEON_GPU_PAGE_SIZE, + NULL); + if (IS_ERR(fence)) { DRM_ERROR("Failed GTT->VRAM copy %d\n", i); + r = PTR_ERR(fence); goto out_lclean_unpin; } @@ -162,11 +168,16 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag) radeon_bo_kunmap(vram_obj); if (ring == R600_RING_TYPE_DMA_INDEX) - r = radeon_copy_dma(rdev, vram_addr, gtt_addr, size / RADEON_GPU_PAGE_SIZE, &fence); + fence = radeon_copy_dma(rdev, vram_addr, gtt_addr, + size / RADEON_GPU_PAGE_SIZE, + NULL); else - r = radeon_copy_blit(rdev, vram_addr, gtt_addr, size / RADEON_GPU_PAGE_SIZE, &fence); - if (r) { + fence = radeon_copy_blit(rdev, vram_addr, gtt_addr, + size / RADEON_GPU_PAGE_SIZE, + NULL); + if (IS_ERR(fence)) { DRM_ERROR("Failed VRAM->GTT copy %d\n", i); + r = PTR_ERR(fence); goto out_lclean_unpin; } @@ -222,7 +233,7 @@ out_lclean: radeon_bo_unreserve(gtt_obj[i]); radeon_bo_unref(>t_obj[i]); } - if (fence) + if (fence && !IS_ERR(fence)) radeon_fence_unref(&fence); break; } diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 72afe82a95c9..8624979afb65 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -39,6 +39,8 @@ #include <linux/seq_file.h> #include <linux/slab.h> #include <linux/swiotlb.h> +#include <linux/swap.h> +#include <linux/pagemap.h> #include <linux/debugfs.h> #include "radeon_reg.h" #include "radeon.h" @@ -176,12 +178,15 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, static void radeon_evict_flags(struct ttm_buffer_object *bo, struct ttm_placement *placement) { + static struct ttm_place placements = { + .fpfn = 0, + .lpfn = 0, + .flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM + }; + struct radeon_bo *rbo; - static u32 placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM; if (!radeon_ttm_bo_is_radeon_bo(bo)) { - placement->fpfn = 0; - placement->lpfn = 0; placement->placement = &placements; placement->busy_placement = &placements; placement->num_placement = 1; @@ -228,6 +233,7 @@ static int radeon_move_blit(struct ttm_buffer_object *bo, struct radeon_device *rdev; uint64_t old_start, new_start; struct radeon_fence *fence; + unsigned num_pages; int r, ridx; rdev = radeon_get_rdev(bo->bdev); @@ -264,13 +270,12 @@ static int radeon_move_blit(struct ttm_buffer_object *bo, BUILD_BUG_ON((PAGE_SIZE % RADEON_GPU_PAGE_SIZE) != 0); - /* sync other rings */ - fence = bo->sync_obj; - r = radeon_copy(rdev, old_start, new_start, - new_mem->num_pages * (PAGE_SIZE / RADEON_GPU_PAGE_SIZE), /* GPU pages */ - &fence); - /* FIXME: handle copy error */ - r = ttm_bo_move_accel_cleanup(bo, (void *)fence, + num_pages = new_mem->num_pages * (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); + fence = radeon_copy(rdev, old_start, new_start, num_pages, bo->resv); + if (IS_ERR(fence)) + return PTR_ERR(fence); + + r = ttm_bo_move_accel_cleanup(bo, &fence->base, evict, no_wait_gpu, new_mem); radeon_fence_unref(&fence); return r; @@ -284,20 +289,20 @@ static int radeon_move_vram_ram(struct ttm_buffer_object *bo, struct radeon_device *rdev; struct ttm_mem_reg *old_mem = &bo->mem; struct ttm_mem_reg tmp_mem; - u32 placements; + struct ttm_place placements; struct ttm_placement placement; int r; rdev = radeon_get_rdev(bo->bdev); tmp_mem = *new_mem; tmp_mem.mm_node = NULL; - placement.fpfn = 0; - placement.lpfn = 0; placement.num_placement = 1; placement.placement = &placements; placement.num_busy_placement = 1; placement.busy_placement = &placements; - placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; + placements.fpfn = 0; + placements.lpfn = 0; + placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait_gpu); if (unlikely(r)) { @@ -332,19 +337,19 @@ static int radeon_move_ram_vram(struct ttm_buffer_object *bo, struct ttm_mem_reg *old_mem = &bo->mem; struct ttm_mem_reg tmp_mem; struct ttm_placement placement; - u32 placements; + struct ttm_place placements; int r; rdev = radeon_get_rdev(bo->bdev); tmp_mem = *new_mem; tmp_mem.mm_node = NULL; - placement.fpfn = 0; - placement.lpfn = 0; placement.num_placement = 1; placement.placement = &placements; placement.num_busy_placement = 1; placement.busy_placement = &placements; - placements = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; + placements.fpfn = 0; + placements.lpfn = 0; + placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait_gpu); if (unlikely(r)) { @@ -483,39 +488,108 @@ static void radeon_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_re { } -static int radeon_sync_obj_wait(void *sync_obj, bool lazy, bool interruptible) -{ - return radeon_fence_wait((struct radeon_fence *)sync_obj, interruptible); -} +/* + * TTM backend functions. + */ +struct radeon_ttm_tt { + struct ttm_dma_tt ttm; + struct radeon_device *rdev; + u64 offset; -static int radeon_sync_obj_flush(void *sync_obj) + uint64_t userptr; + struct mm_struct *usermm; + uint32_t userflags; +}; + +/* prepare the sg table with the user pages */ +static int radeon_ttm_tt_pin_userptr(struct ttm_tt *ttm) { + struct radeon_device *rdev = radeon_get_rdev(ttm->bdev); + struct radeon_ttm_tt *gtt = (void *)ttm; + unsigned pinned = 0, nents; + int r; + + int write = !(gtt->userflags & RADEON_GEM_USERPTR_READONLY); + enum dma_data_direction direction = write ? + DMA_BIDIRECTIONAL : DMA_TO_DEVICE; + + if (current->mm != gtt->usermm) + return -EPERM; + + if (gtt->userflags & RADEON_GEM_USERPTR_ANONONLY) { + /* check that we only pin down anonymous memory + to prevent problems with writeback */ + unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE; + struct vm_area_struct *vma; + vma = find_vma(gtt->usermm, gtt->userptr); + if (!vma || vma->vm_file || vma->vm_end < end) + return -EPERM; + } + + do { + unsigned num_pages = ttm->num_pages - pinned; + uint64_t userptr = gtt->userptr + pinned * PAGE_SIZE; + struct page **pages = ttm->pages + pinned; + + r = get_user_pages(current, current->mm, userptr, num_pages, + write, 0, pages, NULL); + if (r < 0) + goto release_pages; + + pinned += r; + + } while (pinned < ttm->num_pages); + + r = sg_alloc_table_from_pages(ttm->sg, ttm->pages, ttm->num_pages, 0, + ttm->num_pages << PAGE_SHIFT, + GFP_KERNEL); + if (r) + goto release_sg; + + r = -ENOMEM; + nents = dma_map_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction); + if (nents != ttm->sg->nents) + goto release_sg; + + drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages, + gtt->ttm.dma_address, ttm->num_pages); + return 0; -} -static void radeon_sync_obj_unref(void **sync_obj) -{ - radeon_fence_unref((struct radeon_fence **)sync_obj); -} +release_sg: + kfree(ttm->sg); -static void *radeon_sync_obj_ref(void *sync_obj) -{ - return radeon_fence_ref((struct radeon_fence *)sync_obj); +release_pages: + release_pages(ttm->pages, pinned, 0); + return r; } -static bool radeon_sync_obj_signaled(void *sync_obj) +static void radeon_ttm_tt_unpin_userptr(struct ttm_tt *ttm) { - return radeon_fence_signaled((struct radeon_fence *)sync_obj); -} + struct radeon_device *rdev = radeon_get_rdev(ttm->bdev); + struct radeon_ttm_tt *gtt = (void *)ttm; + struct scatterlist *sg; + int i; -/* - * TTM backend functions. - */ -struct radeon_ttm_tt { - struct ttm_dma_tt ttm; - struct radeon_device *rdev; - u64 offset; -}; + int write = !(gtt->userflags & RADEON_GEM_USERPTR_READONLY); + enum dma_data_direction direction = write ? + DMA_BIDIRECTIONAL : DMA_TO_DEVICE; + + /* free the sg table and pages again */ + dma_unmap_sg(rdev->dev, ttm->sg->sgl, ttm->sg->nents, direction); + + for_each_sg(ttm->sg->sgl, sg, ttm->sg->nents, i) { + struct page *page = sg_page(sg); + + if (!(gtt->userflags & RADEON_GEM_USERPTR_READONLY)) + set_page_dirty(page); + + mark_page_accessed(page); + page_cache_release(page); + } + + sg_free_table(ttm->sg); +} static int radeon_ttm_backend_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem) @@ -525,6 +599,11 @@ static int radeon_ttm_backend_bind(struct ttm_tt *ttm, RADEON_GART_PAGE_WRITE; int r; + if (gtt->userptr) { + radeon_ttm_tt_pin_userptr(ttm); + flags &= ~RADEON_GART_PAGE_WRITE; + } + gtt->offset = (unsigned long)(bo_mem->start << PAGE_SHIFT); if (!ttm->num_pages) { WARN(1, "nothing to bind %lu pages for mreg %p back %p!\n", @@ -547,6 +626,10 @@ static int radeon_ttm_backend_unbind(struct ttm_tt *ttm) struct radeon_ttm_tt *gtt = (void *)ttm; radeon_gart_unbind(gtt->rdev, gtt->offset, ttm->num_pages); + + if (gtt->userptr) + radeon_ttm_tt_unpin_userptr(ttm); + return 0; } @@ -592,10 +675,17 @@ static struct ttm_tt *radeon_ttm_tt_create(struct ttm_bo_device *bdev, return >t->ttm.ttm; } +static struct radeon_ttm_tt *radeon_ttm_tt_to_gtt(struct ttm_tt *ttm) +{ + if (!ttm || ttm->func != &radeon_backend_func) + return NULL; + return (struct radeon_ttm_tt *)ttm; +} + static int radeon_ttm_tt_populate(struct ttm_tt *ttm) { + struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm); struct radeon_device *rdev; - struct radeon_ttm_tt *gtt = (void *)ttm; unsigned i; int r; bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG); @@ -603,6 +693,16 @@ static int radeon_ttm_tt_populate(struct ttm_tt *ttm) if (ttm->state != tt_unpopulated) return 0; + if (gtt && gtt->userptr) { + ttm->sg = kcalloc(1, sizeof(struct sg_table), GFP_KERNEL); + if (!ttm->sg) + return -ENOMEM; + + ttm->page_flags |= TTM_PAGE_FLAG_SG; + ttm->state = tt_unbound; + return 0; + } + if (slave && ttm->sg) { drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages, gtt->ttm.dma_address, ttm->num_pages); @@ -648,10 +748,16 @@ static int radeon_ttm_tt_populate(struct ttm_tt *ttm) static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm) { struct radeon_device *rdev; - struct radeon_ttm_tt *gtt = (void *)ttm; + struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm); unsigned i; bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG); + if (gtt && gtt->userptr) { + kfree(ttm->sg); + ttm->page_flags &= ~TTM_PAGE_FLAG_SG; + return; + } + if (slave) return; @@ -680,6 +786,40 @@ static void radeon_ttm_tt_unpopulate(struct ttm_tt *ttm) ttm_pool_unpopulate(ttm); } +int radeon_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr, + uint32_t flags) +{ + struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm); + + if (gtt == NULL) + return -EINVAL; + + gtt->userptr = addr; + gtt->usermm = current->mm; + gtt->userflags = flags; + return 0; +} + +bool radeon_ttm_tt_has_userptr(struct ttm_tt *ttm) +{ + struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm); + + if (gtt == NULL) + return false; + + return !!gtt->userptr; +} + +bool radeon_ttm_tt_is_readonly(struct ttm_tt *ttm) +{ + struct radeon_ttm_tt *gtt = radeon_ttm_tt_to_gtt(ttm); + + if (gtt == NULL) + return false; + + return !!(gtt->userflags & RADEON_GEM_USERPTR_READONLY); +} + static struct ttm_bo_driver radeon_bo_driver = { .ttm_tt_create = &radeon_ttm_tt_create, .ttm_tt_populate = &radeon_ttm_tt_populate, @@ -689,11 +829,6 @@ static struct ttm_bo_driver radeon_bo_driver = { .evict_flags = &radeon_evict_flags, .move = &radeon_bo_move, .verify_access = &radeon_verify_access, - .sync_obj_signaled = &radeon_sync_obj_signaled, - .sync_obj_wait = &radeon_sync_obj_wait, - .sync_obj_flush = &radeon_sync_obj_flush, - .sync_obj_unref = &radeon_sync_obj_unref, - .sync_obj_ref = &radeon_sync_obj_ref, .move_notify = &radeon_bo_move_notify, .fault_reserve_notify = &radeon_bo_fault_reserve_notify, .io_mem_reserve = &radeon_ttm_io_mem_reserve, @@ -730,7 +865,7 @@ int radeon_ttm_init(struct radeon_device *rdev) radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size); r = radeon_bo_create(rdev, 256 * 1024, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_VRAM, 0, + RADEON_GEM_DOMAIN_VRAM, 0, NULL, NULL, &rdev->stollen_vga_memory); if (r) { return r; @@ -828,7 +963,7 @@ int radeon_mmap(struct file *filp, struct vm_area_struct *vma) int r; if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET)) { - return drm_mmap(filp, vma); + return -EINVAL; } file_priv = filp->private_data; diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c index 341848a14376..11b662469253 100644 --- a/drivers/gpu/drm/radeon/radeon_uvd.c +++ b/drivers/gpu/drm/radeon/radeon_uvd.c @@ -40,12 +40,18 @@ #define UVD_IDLE_TIMEOUT_MS 1000 /* Firmware Names */ +#define FIRMWARE_R600 "radeon/R600_uvd.bin" +#define FIRMWARE_RS780 "radeon/RS780_uvd.bin" +#define FIRMWARE_RV770 "radeon/RV770_uvd.bin" #define FIRMWARE_RV710 "radeon/RV710_uvd.bin" #define FIRMWARE_CYPRESS "radeon/CYPRESS_uvd.bin" #define FIRMWARE_SUMO "radeon/SUMO_uvd.bin" #define FIRMWARE_TAHITI "radeon/TAHITI_uvd.bin" #define FIRMWARE_BONAIRE "radeon/BONAIRE_uvd.bin" +MODULE_FIRMWARE(FIRMWARE_R600); +MODULE_FIRMWARE(FIRMWARE_RS780); +MODULE_FIRMWARE(FIRMWARE_RV770); MODULE_FIRMWARE(FIRMWARE_RV710); MODULE_FIRMWARE(FIRMWARE_CYPRESS); MODULE_FIRMWARE(FIRMWARE_SUMO); @@ -63,6 +69,23 @@ int radeon_uvd_init(struct radeon_device *rdev) INIT_DELAYED_WORK(&rdev->uvd.idle_work, radeon_uvd_idle_work_handler); switch (rdev->family) { + case CHIP_RV610: + case CHIP_RV630: + case CHIP_RV670: + case CHIP_RV620: + case CHIP_RV635: + fw_name = FIRMWARE_R600; + break; + + case CHIP_RS780: + case CHIP_RS880: + fw_name = FIRMWARE_RS780; + break; + + case CHIP_RV770: + fw_name = FIRMWARE_RV770; + break; + case CHIP_RV710: case CHIP_RV730: case CHIP_RV740: @@ -115,9 +138,11 @@ int radeon_uvd_init(struct radeon_device *rdev) } bo_size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size + 8) + - RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE; + RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE + + RADEON_GPU_PAGE_SIZE; r = radeon_bo_create(rdev, bo_size, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->uvd.vcpu_bo); + RADEON_GEM_DOMAIN_VRAM, 0, NULL, + NULL, &rdev->uvd.vcpu_bo); if (r) { dev_err(rdev->dev, "(%d) failed to allocate UVD bo\n", r); return r; @@ -231,10 +256,30 @@ int radeon_uvd_resume(struct radeon_device *rdev) return 0; } -void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo) +void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo, + uint32_t allowed_domains) { - rbo->placement.fpfn = 0 >> PAGE_SHIFT; - rbo->placement.lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT; + int i; + + for (i = 0; i < rbo->placement.num_placement; ++i) { + rbo->placements[i].fpfn = 0 >> PAGE_SHIFT; + rbo->placements[i].lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT; + } + + /* If it must be in VRAM it must be in the first segment as well */ + if (allowed_domains == RADEON_GEM_DOMAIN_VRAM) + return; + + /* abort if we already have more than one placement */ + if (rbo->placement.num_placement > 1) + return; + + /* add another 256MB segment */ + rbo->placements[1] = rbo->placements[0]; + rbo->placements[1].fpfn += (256 * 1024 * 1024) >> PAGE_SHIFT; + rbo->placements[1].lpfn += (256 * 1024 * 1024) >> PAGE_SHIFT; + rbo->placement.num_placement++; + rbo->placement.num_busy_placement++; } void radeon_uvd_free_handles(struct radeon_device *rdev, struct drm_file *filp) @@ -356,6 +401,7 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo, { int32_t *msg, msg_type, handle; unsigned img_size = 0; + struct fence *f; void *ptr; int i, r; @@ -365,8 +411,9 @@ static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo, return -EINVAL; } - if (bo->tbo.sync_obj) { - r = radeon_fence_wait(bo->tbo.sync_obj, false); + f = reservation_object_get_excl(bo->tbo.resv); + if (f) { + r = radeon_fence_wait((struct radeon_fence *)f, false); if (r) { DRM_ERROR("Failed waiting for UVD message (%d)!\n", r); return r; @@ -604,38 +651,16 @@ int radeon_uvd_cs_parse(struct radeon_cs_parser *p) } static int radeon_uvd_send_msg(struct radeon_device *rdev, - int ring, struct radeon_bo *bo, + int ring, uint64_t addr, struct radeon_fence **fence) { - struct ttm_validate_buffer tv; - struct ww_acquire_ctx ticket; - struct list_head head; struct radeon_ib ib; - uint64_t addr; int i, r; - memset(&tv, 0, sizeof(tv)); - tv.bo = &bo->tbo; - - INIT_LIST_HEAD(&head); - list_add(&tv.head, &head); - - r = ttm_eu_reserve_buffers(&ticket, &head); - if (r) - return r; - - radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_VRAM); - radeon_uvd_force_into_uvd_segment(bo); - - r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false); - if (r) - goto err; - r = radeon_ib_get(rdev, ring, &ib, NULL, 64); if (r) - goto err; + return r; - addr = radeon_bo_gpu_offset(bo); ib.ptr[0] = PACKET0(UVD_GPCOM_VCPU_DATA0, 0); ib.ptr[1] = addr; ib.ptr[2] = PACKET0(UVD_GPCOM_VCPU_DATA1, 0); @@ -647,19 +672,11 @@ static int radeon_uvd_send_msg(struct radeon_device *rdev, ib.length_dw = 16; r = radeon_ib_schedule(rdev, &ib, NULL, false); - if (r) - goto err; - ttm_eu_fence_buffer_objects(&ticket, &head, ib.fence); if (fence) *fence = radeon_fence_ref(ib.fence); radeon_ib_free(rdev, &ib); - radeon_bo_unref(&bo); - return 0; - -err: - ttm_eu_backoff_reservation(&ticket, &head); return r; } @@ -669,27 +686,18 @@ err: int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring, uint32_t handle, struct radeon_fence **fence) { - struct radeon_bo *bo; - uint32_t *msg; - int r, i; + /* we use the last page of the vcpu bo for the UVD message */ + uint64_t offs = radeon_bo_size(rdev->uvd.vcpu_bo) - + RADEON_GPU_PAGE_SIZE; - r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo); - if (r) - return r; + uint32_t *msg = rdev->uvd.cpu_addr + offs; + uint64_t addr = rdev->uvd.gpu_addr + offs; - r = radeon_bo_reserve(bo, false); - if (r) { - radeon_bo_unref(&bo); - return r; - } + int r, i; - r = radeon_bo_kmap(bo, (void **)&msg); - if (r) { - radeon_bo_unreserve(bo); - radeon_bo_unref(&bo); + r = radeon_bo_reserve(rdev->uvd.vcpu_bo, true); + if (r) return r; - } /* stitch together an UVD create msg */ msg[0] = cpu_to_le32(0x00000de4); @@ -706,36 +714,26 @@ int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring, for (i = 11; i < 1024; ++i) msg[i] = cpu_to_le32(0x0); - radeon_bo_kunmap(bo); - radeon_bo_unreserve(bo); - - return radeon_uvd_send_msg(rdev, ring, bo, fence); + r = radeon_uvd_send_msg(rdev, ring, addr, fence); + radeon_bo_unreserve(rdev->uvd.vcpu_bo); + return r; } int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring, uint32_t handle, struct radeon_fence **fence) { - struct radeon_bo *bo; - uint32_t *msg; - int r, i; + /* we use the last page of the vcpu bo for the UVD message */ + uint64_t offs = radeon_bo_size(rdev->uvd.vcpu_bo) - + RADEON_GPU_PAGE_SIZE; - r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_VRAM, 0, NULL, &bo); - if (r) - return r; + uint32_t *msg = rdev->uvd.cpu_addr + offs; + uint64_t addr = rdev->uvd.gpu_addr + offs; - r = radeon_bo_reserve(bo, false); - if (r) { - radeon_bo_unref(&bo); - return r; - } + int r, i; - r = radeon_bo_kmap(bo, (void **)&msg); - if (r) { - radeon_bo_unreserve(bo); - radeon_bo_unref(&bo); + r = radeon_bo_reserve(rdev->uvd.vcpu_bo, true); + if (r) return r; - } /* stitch together an UVD destroy msg */ msg[0] = cpu_to_le32(0x00000de4); @@ -745,10 +743,9 @@ int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring, for (i = 4; i < 1024; ++i) msg[i] = cpu_to_le32(0x0); - radeon_bo_kunmap(bo); - radeon_bo_unreserve(bo); - - return radeon_uvd_send_msg(rdev, ring, bo, fence); + r = radeon_uvd_send_msg(rdev, ring, addr, fence); + radeon_bo_unreserve(rdev->uvd.vcpu_bo); + return r; } /** diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c index c7190aadbd89..9e85757d5599 100644 --- a/drivers/gpu/drm/radeon/radeon_vce.c +++ b/drivers/gpu/drm/radeon/radeon_vce.c @@ -126,7 +126,8 @@ int radeon_vce_init(struct radeon_device *rdev) size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size) + RADEON_VCE_STACK_SIZE + RADEON_VCE_HEAP_SIZE; r = radeon_bo_create(rdev, size, PAGE_SIZE, true, - RADEON_GEM_DOMAIN_VRAM, 0, NULL, &rdev->vce.vcpu_bo); + RADEON_GEM_DOMAIN_VRAM, 0, NULL, NULL, + &rdev->vce.vcpu_bo); if (r) { dev_err(rdev->dev, "(%d) failed to allocate VCE bo\n", r); return r; diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c index 088ffdc2f577..4532cc76a0a6 100644 --- a/drivers/gpu/drm/radeon/radeon_vm.c +++ b/drivers/gpu/drm/radeon/radeon_vm.c @@ -143,6 +143,7 @@ struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev, list[0].prefered_domains = RADEON_GEM_DOMAIN_VRAM; list[0].allowed_domains = RADEON_GEM_DOMAIN_VRAM; list[0].tv.bo = &vm->page_directory->tbo; + list[0].tv.shared = false; list[0].tiling_flags = 0; list[0].handle = 0; list_add(&list[0].tv.head, head); @@ -156,6 +157,7 @@ struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev, list[idx].prefered_domains = RADEON_GEM_DOMAIN_VRAM; list[idx].allowed_domains = RADEON_GEM_DOMAIN_VRAM; list[idx].tv.bo = &list[idx].robj->tbo; + list[idx].tv.shared = false; list[idx].tiling_flags = 0; list[idx].handle = 0; list_add(&list[idx++].tv.head, head); @@ -395,11 +397,12 @@ static int radeon_vm_clear_bo(struct radeon_device *rdev, memset(&tv, 0, sizeof(tv)); tv.bo = &bo->tbo; + tv.shared = false; INIT_LIST_HEAD(&head); list_add(&tv.head, &head); - r = ttm_eu_reserve_buffers(&ticket, &head); + r = ttm_eu_reserve_buffers(&ticket, &head, true); if (r) return r; @@ -424,7 +427,7 @@ static int radeon_vm_clear_bo(struct radeon_device *rdev, if (r) goto error; - ttm_eu_fence_buffer_objects(&ticket, &head, ib.fence); + ttm_eu_fence_buffer_objects(&ticket, &head, &ib.fence->base); radeon_ib_free(rdev, &ib); return 0; @@ -545,7 +548,8 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev, r = radeon_bo_create(rdev, RADEON_VM_PTE_COUNT * 8, RADEON_GPU_PAGE_SIZE, true, - RADEON_GEM_DOMAIN_VRAM, 0, NULL, &pt); + RADEON_GEM_DOMAIN_VRAM, 0, + NULL, NULL, &pt); if (r) return r; @@ -694,8 +698,9 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev, if (ib.length_dw != 0) { radeon_asic_vm_pad_ib(rdev, &ib); - radeon_semaphore_sync_to(ib.semaphore, pd->tbo.sync_obj); - radeon_semaphore_sync_to(ib.semaphore, vm->last_id_use); + + radeon_semaphore_sync_resv(rdev, ib.semaphore, pd->tbo.resv, false); + radeon_semaphore_sync_fence(ib.semaphore, vm->last_id_use); WARN_ON(ib.length_dw > ndw); r = radeon_ib_schedule(rdev, &ib, NULL, false); if (r) { @@ -821,7 +826,7 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev, unsigned nptes; uint64_t pte; - radeon_semaphore_sync_to(ib->semaphore, pt->tbo.sync_obj); + radeon_semaphore_sync_resv(rdev, ib->semaphore, pt->tbo.resv, false); if ((addr & ~mask) == (end & ~mask)) nptes = end - addr; @@ -892,6 +897,9 @@ int radeon_vm_bo_update(struct radeon_device *rdev, bo_va->flags &= ~RADEON_VM_PAGE_VALID; bo_va->flags &= ~RADEON_VM_PAGE_SYSTEM; bo_va->flags &= ~RADEON_VM_PAGE_SNOOPED; + if (bo_va->bo && radeon_ttm_tt_is_readonly(bo_va->bo->tbo.ttm)) + bo_va->flags &= ~RADEON_VM_PAGE_WRITEABLE; + if (mem) { addr = mem->start << PAGE_SHIFT; if (mem->mem_type != TTM_PL_SYSTEM) { @@ -960,7 +968,7 @@ int radeon_vm_bo_update(struct radeon_device *rdev, radeon_asic_vm_pad_ib(rdev, &ib); WARN_ON(ib.length_dw > ndw); - radeon_semaphore_sync_to(ib.semaphore, vm->fence); + radeon_semaphore_sync_fence(ib.semaphore, vm->fence); r = radeon_ib_schedule(rdev, &ib, NULL, false); if (r) { radeon_ib_free(rdev, &ib); @@ -1120,7 +1128,7 @@ int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm) r = radeon_bo_create(rdev, pd_size, align, true, RADEON_GEM_DOMAIN_VRAM, 0, NULL, - &vm->page_directory); + NULL, &vm->page_directory); if (r) return r; diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index d9f5ce715c9b..372016e266d0 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -26,7 +26,6 @@ * Jerome Glisse */ #include <linux/firmware.h> -#include <linux/platform_device.h> #include <linux/slab.h> #include <drm/drmP.h> #include "radeon.h" diff --git a/drivers/gpu/drm/radeon/rv770_dma.c b/drivers/gpu/drm/radeon/rv770_dma.c index 74426ac2bb5c..7f34bad2e724 100644 --- a/drivers/gpu/drm/radeon/rv770_dma.c +++ b/drivers/gpu/drm/radeon/rv770_dma.c @@ -33,18 +33,19 @@ * @src_offset: src GPU address * @dst_offset: dst GPU address * @num_gpu_pages: number of GPU pages to xfer - * @fence: radeon fence object + * @resv: reservation object to sync to * * Copy GPU paging using the DMA engine (r7xx). * Used by the radeon ttm implementation to move pages if * registered as the asic copy callback. */ -int rv770_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *rv770_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_semaphore *sem = NULL; + struct radeon_fence *fence; int ring_index = rdev->asic->copy.dma_ring_index; struct radeon_ring *ring = &rdev->ring[ring_index]; u32 size_in_dw, cur_size_in_dw; @@ -54,7 +55,7 @@ int rv770_copy_dma(struct radeon_device *rdev, r = radeon_semaphore_create(rdev, &sem); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } size_in_dw = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT) / 4; @@ -63,10 +64,10 @@ int rv770_copy_dma(struct radeon_device *rdev, if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } - radeon_semaphore_sync_to(sem, *fence); + radeon_semaphore_sync_resv(rdev, sem, resv, false); radeon_semaphore_sync_rings(rdev, sem, ring->idx); for (i = 0; i < num_loops; i++) { @@ -83,15 +84,15 @@ int rv770_copy_dma(struct radeon_device *rdev, dst_offset += cur_size_in_dw * 4; } - r = radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, &fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - radeon_semaphore_free(rdev, &sem, *fence); + radeon_semaphore_free(rdev, &sem, fence); - return r; + return fence; } diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c index 3a0b973e8a96..eeea5b6a1775 100644 --- a/drivers/gpu/drm/radeon/si.c +++ b/drivers/gpu/drm/radeon/si.c @@ -4684,7 +4684,7 @@ static int si_vm_packet3_compute_check(struct radeon_device *rdev, int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib) { int ret = 0; - u32 idx = 0; + u32 idx = 0, i; struct radeon_cs_packet pkt; do { @@ -4695,6 +4695,12 @@ int si_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib) switch (pkt.type) { case RADEON_PACKET_TYPE0: dev_err(rdev->dev, "Packet0 not allowed!\n"); + for (i = 0; i < ib->length_dw; i++) { + if (i == idx) + printk("\t0x%08x <---\n", ib->ptr[i]); + else + printk("\t0x%08x\n", ib->ptr[i]); + } ret = -EINVAL; break; case RADEON_PACKET_TYPE2: diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c index 7c22baaf94db..b58f12b762d7 100644 --- a/drivers/gpu/drm/radeon/si_dma.c +++ b/drivers/gpu/drm/radeon/si_dma.c @@ -218,18 +218,19 @@ void si_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm) * @src_offset: src GPU address * @dst_offset: dst GPU address * @num_gpu_pages: number of GPU pages to xfer - * @fence: radeon fence object + * @resv: reservation object to sync to * * Copy GPU paging using the DMA engine (SI). * Used by the radeon ttm implementation to move pages if * registered as the asic copy callback. */ -int si_copy_dma(struct radeon_device *rdev, - uint64_t src_offset, uint64_t dst_offset, - unsigned num_gpu_pages, - struct radeon_fence **fence) +struct radeon_fence *si_copy_dma(struct radeon_device *rdev, + uint64_t src_offset, uint64_t dst_offset, + unsigned num_gpu_pages, + struct reservation_object *resv) { struct radeon_semaphore *sem = NULL; + struct radeon_fence *fence; int ring_index = rdev->asic->copy.dma_ring_index; struct radeon_ring *ring = &rdev->ring[ring_index]; u32 size_in_bytes, cur_size_in_bytes; @@ -239,7 +240,7 @@ int si_copy_dma(struct radeon_device *rdev, r = radeon_semaphore_create(rdev, &sem); if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); - return r; + return ERR_PTR(r); } size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT); @@ -248,10 +249,10 @@ int si_copy_dma(struct radeon_device *rdev, if (r) { DRM_ERROR("radeon: moving bo (%d).\n", r); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } - radeon_semaphore_sync_to(sem, *fence); + radeon_semaphore_sync_resv(rdev, sem, resv, false); radeon_semaphore_sync_rings(rdev, sem, ring->idx); for (i = 0; i < num_loops; i++) { @@ -268,16 +269,16 @@ int si_copy_dma(struct radeon_device *rdev, dst_offset += cur_size_in_bytes; } - r = radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, &fence, ring->idx); if (r) { radeon_ring_unlock_undo(rdev, ring); radeon_semaphore_free(rdev, &sem, NULL); - return r; + return ERR_PTR(r); } radeon_ring_unlock_commit(rdev, ring, false); - radeon_semaphore_free(rdev, &sem, *fence); + radeon_semaphore_free(rdev, &sem, fence); - return r; + return fence; } diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c index 70e61ffeace2..9e4d5d7d348f 100644 --- a/drivers/gpu/drm/radeon/si_dpm.c +++ b/drivers/gpu/drm/radeon/si_dpm.c @@ -2916,7 +2916,6 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev, bool disable_sclk_switching = false; u32 mclk, sclk; u16 vddc, vddci; - u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc; int i; if ((rdev->pm.dpm.new_active_crtc_count > 1) || @@ -2950,29 +2949,6 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev, } } - /* limit clocks to max supported clocks based on voltage dependency tables */ - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk, - &max_sclk_vddc); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddci_dependency_on_mclk, - &max_mclk_vddci); - btc_get_max_clock_from_voltage_dependency_table(&rdev->pm.dpm.dyn_state.vddc_dependency_on_mclk, - &max_mclk_vddc); - - for (i = 0; i < ps->performance_level_count; i++) { - if (max_sclk_vddc) { - if (ps->performance_levels[i].sclk > max_sclk_vddc) - ps->performance_levels[i].sclk = max_sclk_vddc; - } - if (max_mclk_vddci) { - if (ps->performance_levels[i].mclk > max_mclk_vddci) - ps->performance_levels[i].mclk = max_mclk_vddci; - } - if (max_mclk_vddc) { - if (ps->performance_levels[i].mclk > max_mclk_vddc) - ps->performance_levels[i].mclk = max_mclk_vddc; - } - } - /* XXX validate the min clocks required for display */ if (disable_mclk_switching) { diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h index fd414d34d885..6635da9ec986 100644 --- a/drivers/gpu/drm/radeon/sid.h +++ b/drivers/gpu/drm/radeon/sid.h @@ -736,7 +736,7 @@ # define DESCRIPTION16(x) (((x) & 0xff) << 0) # define DESCRIPTION17(x) (((x) & 0xff) << 8) -#define AZ_F0_CODEC_PIN_CONTROL_HOTPLUG_CONTROL 0x54 +#define AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL 0x54 # define AUDIO_ENABLED (1 << 31) #define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_CONFIGURATION_DEFAULT 0x56 diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c index cda391347286..e72b3cb59358 100644 --- a/drivers/gpu/drm/radeon/uvd_v1_0.c +++ b/drivers/gpu/drm/radeon/uvd_v1_0.c @@ -22,6 +22,7 @@ * Authors: Christian König <christian.koenig@amd.com> */ +#include <linux/firmware.h> #include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" @@ -70,6 +71,82 @@ void uvd_v1_0_set_wptr(struct radeon_device *rdev, } /** + * uvd_v1_0_fence_emit - emit an fence & trap command + * + * @rdev: radeon_device pointer + * @fence: fence to emit + * + * Write a fence and a trap command to the ring. + */ +void uvd_v1_0_fence_emit(struct radeon_device *rdev, + struct radeon_fence *fence) +{ + struct radeon_ring *ring = &rdev->ring[fence->ring]; + uint64_t addr = rdev->fence_drv[fence->ring].gpu_addr; + + radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0)); + radeon_ring_write(ring, addr & 0xffffffff); + radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0)); + radeon_ring_write(ring, fence->seq); + radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0)); + radeon_ring_write(ring, 0); + + radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0)); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0)); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0)); + radeon_ring_write(ring, 2); + return; +} + +/** + * uvd_v1_0_resume - memory controller programming + * + * @rdev: radeon_device pointer + * + * Let the UVD memory controller know it's offsets + */ +int uvd_v1_0_resume(struct radeon_device *rdev) +{ + uint64_t addr; + uint32_t size; + int r; + + r = radeon_uvd_resume(rdev); + if (r) + return r; + + /* programm the VCPU memory controller bits 0-27 */ + addr = (rdev->uvd.gpu_addr >> 3) + 16; + size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size) >> 3; + WREG32(UVD_VCPU_CACHE_OFFSET0, addr); + WREG32(UVD_VCPU_CACHE_SIZE0, size); + + addr += size; + size = RADEON_UVD_STACK_SIZE >> 3; + WREG32(UVD_VCPU_CACHE_OFFSET1, addr); + WREG32(UVD_VCPU_CACHE_SIZE1, size); + + addr += size; + size = RADEON_UVD_HEAP_SIZE >> 3; + WREG32(UVD_VCPU_CACHE_OFFSET2, addr); + WREG32(UVD_VCPU_CACHE_SIZE2, size); + + /* bits 28-31 */ + addr = (rdev->uvd.gpu_addr >> 28) & 0xF; + WREG32(UVD_LMI_ADDR_EXT, (addr << 12) | (addr << 0)); + + /* bits 32-39 */ + addr = (rdev->uvd.gpu_addr >> 32) & 0xFF; + WREG32(UVD_LMI_EXT40_ADDR, addr | (0x9 << 16) | (0x1 << 31)); + + WREG32(UVD_FW_START, *((uint32_t*)rdev->uvd.cpu_addr)); + + return 0; +} + +/** * uvd_v1_0_init - start and test UVD block * * @rdev: radeon_device pointer @@ -130,8 +207,32 @@ done: /* lower clocks again */ radeon_set_uvd_clocks(rdev, 0, 0); - if (!r) + if (!r) { + switch (rdev->family) { + case CHIP_RV610: + case CHIP_RV630: + case CHIP_RV620: + /* 64byte granularity workaround */ + WREG32(MC_CONFIG, 0); + WREG32(MC_CONFIG, 1 << 4); + WREG32(RS_DQ_RD_RET_CONF, 0x3f); + WREG32(MC_CONFIG, 0x1f); + + /* fall through */ + case CHIP_RV670: + case CHIP_RV635: + + /* write clean workaround */ + WREG32_P(UVD_VCPU_CNTL, 0x10, ~0x10); + break; + + default: + /* TODO: Do we need more? */ + break; + } + DRM_INFO("UVD initialized successfully.\n"); + } return r; } @@ -218,12 +319,12 @@ int uvd_v1_0_start(struct radeon_device *rdev) /* enable UMC */ WREG32_P(UVD_LMI_CTRL2, 0, ~(1 << 8)); + WREG32_P(UVD_RB_ARB_CTRL, 0, ~(1 << 3)); + /* boot up the VCPU */ WREG32(UVD_SOFT_RESET, 0); mdelay(10); - WREG32_P(UVD_RB_ARB_CTRL, 0, ~(1 << 3)); - for (i = 0; i < 10; ++i) { uint32_t status; for (j = 0; j < 100; ++j) { diff --git a/drivers/gpu/drm/radeon/uvd_v2_2.c b/drivers/gpu/drm/radeon/uvd_v2_2.c index 8bfdadd56598..89193519f8a1 100644 --- a/drivers/gpu/drm/radeon/uvd_v2_2.c +++ b/drivers/gpu/drm/radeon/uvd_v2_2.c @@ -72,6 +72,10 @@ int uvd_v2_2_resume(struct radeon_device *rdev) uint32_t chip_id, size; int r; + /* RV770 uses V1.0 MC */ + if (rdev->family == CHIP_RV770) + return uvd_v1_0_resume(rdev); + r = radeon_uvd_resume(rdev); if (r) return r; |