diff options
Diffstat (limited to 'drivers/gpu/drm/radeon')
29 files changed, 391 insertions, 97 deletions
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index d59ec491dbb9..ed644a4f6f57 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -1851,10 +1851,9 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc) return pll; } /* otherwise, pick one of the plls */ - if ((rdev->family == CHIP_KAVERI) || - (rdev->family == CHIP_KABINI) || + if ((rdev->family == CHIP_KABINI) || (rdev->family == CHIP_MULLINS)) { - /* KB/KV/ML has PPLL1 and PPLL2 */ + /* KB/ML has PPLL1 and PPLL2 */ pll_in_use = radeon_get_pll_use_mask(crtc); if (!(pll_in_use & (1 << ATOM_PPLL2))) return ATOM_PPLL2; @@ -1863,7 +1862,7 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc) DRM_ERROR("unable to allocate a PPLL\n"); return ATOM_PPLL_INVALID; } else { - /* CI has PPLL0, PPLL1, and PPLL2 */ + /* CI/KV has PPLL0, PPLL1, and PPLL2 */ pll_in_use = radeon_get_pll_use_mask(crtc); if (!(pll_in_use & (1 << ATOM_PPLL2))) return ATOM_PPLL2; @@ -2155,6 +2154,7 @@ static void atombios_crtc_disable(struct drm_crtc *crtc) case ATOM_PPLL0: /* disable the ppll */ if ((rdev->family == CHIP_ARUBA) || + (rdev->family == CHIP_KAVERI) || (rdev->family == CHIP_BONAIRE) || (rdev->family == CHIP_HAWAII)) atombios_crtc_program_pll(crtc, radeon_crtc->crtc_id, radeon_crtc->pll_id, diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index 11ba9d21b89b..db42a670f995 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -492,6 +492,10 @@ int radeon_dp_mode_valid_helper(struct drm_connector *connector, struct radeon_connector_atom_dig *dig_connector; int dp_clock; + if ((mode->clock > 340000) && + (!radeon_connector_is_dp12_capable(connector))) + return MODE_CLOCK_HIGH; + if (!radeon_connector->con_priv) return MODE_CLOCK_HIGH; dig_connector = radeon_connector->con_priv; diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c index 6dcde3798b45..64fdae558d36 100644 --- a/drivers/gpu/drm/radeon/cik.c +++ b/drivers/gpu/drm/radeon/cik.c @@ -6033,6 +6033,17 @@ void cik_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring, radeon_ring_write(ring, 0); radeon_ring_write(ring, 1 << vm_id); + /* wait for the invalidate to complete */ + radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); + radeon_ring_write(ring, (WAIT_REG_MEM_OPERATION(0) | /* wait */ + WAIT_REG_MEM_FUNCTION(0) | /* always */ + WAIT_REG_MEM_ENGINE(0))); /* me */ + radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); /* ref */ + radeon_ring_write(ring, 0); /* mask */ + radeon_ring_write(ring, 0x20); /* poll interval */ + /* compute doesn't have PFP */ if (usepfp) { /* sync PFP to ME, otherwise we might get invalid PFP reads */ diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c index dde5c7e29eb2..42cd0cffe210 100644 --- a/drivers/gpu/drm/radeon/cik_sdma.c +++ b/drivers/gpu/drm/radeon/cik_sdma.c @@ -816,7 +816,6 @@ void cik_sdma_vm_write_pages(struct radeon_device *rdev, for (; ndw > 0; ndw -= 2, --count, pe += 8) { if (flags & R600_PTE_SYSTEM) { value = radeon_vm_map_gart(rdev, addr); - value &= 0xFFFFFFFFFFFFF000ULL; } else if (flags & R600_PTE_VALID) { value = addr; } else { @@ -903,6 +902,9 @@ void cik_sdma_vm_pad_ib(struct radeon_ib *ib) void cik_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring, unsigned vm_id, uint64_t pd_addr) { + u32 extra_bits = (SDMA_POLL_REG_MEM_EXTRA_OP(0) | + SDMA_POLL_REG_MEM_EXTRA_FUNC(0)); /* always */ + radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000)); if (vm_id < 8) { radeon_ring_write(ring, (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2)) >> 2); @@ -943,5 +945,12 @@ void cik_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring, radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000)); radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2); radeon_ring_write(ring, 1 << vm_id); + + radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_POLL_REG_MEM, 0, extra_bits)); + radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); /* reference */ + radeon_ring_write(ring, 0); /* mask */ + radeon_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */ } diff --git a/drivers/gpu/drm/radeon/cikd.h b/drivers/gpu/drm/radeon/cikd.h index ba85986febea..03003f8a6de6 100644 --- a/drivers/gpu/drm/radeon/cikd.h +++ b/drivers/gpu/drm/radeon/cikd.h @@ -2156,4 +2156,6 @@ #define ATC_VM_APERTURE1_HIGH_ADDR 0x330Cu #define ATC_VM_APERTURE1_LOW_ADDR 0x3304u +#define IH_VMID_0_LUT 0x3D40u + #endif diff --git a/drivers/gpu/drm/radeon/dce3_1_afmt.c b/drivers/gpu/drm/radeon/dce3_1_afmt.c index 2fe8cfc966d9..bafdf92a5732 100644 --- a/drivers/gpu/drm/radeon/dce3_1_afmt.c +++ b/drivers/gpu/drm/radeon/dce3_1_afmt.c @@ -103,7 +103,7 @@ static void dce3_2_afmt_write_sad_regs(struct drm_encoder *encoder) } sad_count = drm_edid_to_sad(radeon_connector->edid, &sads); - if (sad_count < 0) { + if (sad_count <= 0) { DRM_ERROR("Couldn't read SADs: %d\n", sad_count); return; } diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c index 9b42001295ba..e3e9c10cfba9 100644 --- a/drivers/gpu/drm/radeon/kv_dpm.c +++ b/drivers/gpu/drm/radeon/kv_dpm.c @@ -2745,13 +2745,11 @@ int kv_dpm_init(struct radeon_device *rdev) pi->enable_auto_thermal_throttling = true; pi->disable_nb_ps3_in_battery = false; if (radeon_bapm == -1) { - /* There are stability issues reported on with - * bapm enabled on an asrock system. - */ - if (rdev->pdev->subsystem_vendor == 0x1849) - pi->bapm_enable = false; - else + /* only enable bapm on KB, ML by default */ + if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) pi->bapm_enable = true; + else + pi->bapm_enable = false; } else if (radeon_bapm == 0) { pi->bapm_enable = false; } else { diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c index 360de9f1f491..aea48c89b241 100644 --- a/drivers/gpu/drm/radeon/ni.c +++ b/drivers/gpu/drm/radeon/ni.c @@ -2516,6 +2516,16 @@ void cayman_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring, radeon_ring_write(ring, PACKET0(VM_INVALIDATE_REQUEST, 0)); radeon_ring_write(ring, 1 << vm_id); + /* wait for the invalidate to complete */ + radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); + radeon_ring_write(ring, (WAIT_REG_MEM_FUNCTION(0) | /* always */ + WAIT_REG_MEM_ENGINE(0))); /* me */ + radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); /* ref */ + radeon_ring_write(ring, 0); /* mask */ + radeon_ring_write(ring, 0x20); /* poll interval */ + /* sync PFP to ME, otherwise we might get invalid PFP reads */ radeon_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0)); radeon_ring_write(ring, 0x0); diff --git a/drivers/gpu/drm/radeon/ni_dma.c b/drivers/gpu/drm/radeon/ni_dma.c index 50f88611ff60..ce787a9f12c0 100644 --- a/drivers/gpu/drm/radeon/ni_dma.c +++ b/drivers/gpu/drm/radeon/ni_dma.c @@ -372,7 +372,6 @@ void cayman_dma_vm_write_pages(struct radeon_device *rdev, for (; ndw > 0; ndw -= 2, --count, pe += 8) { if (flags & R600_PTE_SYSTEM) { value = radeon_vm_map_gart(rdev, addr); - value &= 0xFFFFFFFFFFFFF000ULL; } else if (flags & R600_PTE_VALID) { value = addr; } else { @@ -463,5 +462,11 @@ void cayman_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring, radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0)); radeon_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST >> 2)); radeon_ring_write(ring, 1 << vm_id); + + /* wait for invalidate to complete */ + radeon_ring_write(ring, DMA_SRBM_READ_PACKET); + radeon_ring_write(ring, (0xff << 20) | (VM_INVALIDATE_REQUEST >> 2)); + radeon_ring_write(ring, 0); /* mask */ + radeon_ring_write(ring, 0); /* value */ } diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h index 2e12e4d69253..ad7125486894 100644 --- a/drivers/gpu/drm/radeon/nid.h +++ b/drivers/gpu/drm/radeon/nid.h @@ -1133,6 +1133,23 @@ #define PACKET3_MEM_SEMAPHORE 0x39 #define PACKET3_MPEG_INDEX 0x3A #define PACKET3_WAIT_REG_MEM 0x3C +#define WAIT_REG_MEM_FUNCTION(x) ((x) << 0) + /* 0 - always + * 1 - < + * 2 - <= + * 3 - == + * 4 - != + * 5 - >= + * 6 - > + */ +#define WAIT_REG_MEM_MEM_SPACE(x) ((x) << 4) + /* 0 - reg + * 1 - mem + */ +#define WAIT_REG_MEM_ENGINE(x) ((x) << 8) + /* 0 - me + * 1 - pfp + */ #define PACKET3_MEM_WRITE 0x3D #define PACKET3_PFP_SYNC_ME 0x42 #define PACKET3_SURFACE_SYNC 0x43 @@ -1272,6 +1289,13 @@ (1 << 21) | \ (((n) & 0xFFFFF) << 0)) +#define DMA_SRBM_POLL_PACKET ((9 << 28) | \ + (1 << 27) | \ + (1 << 26)) + +#define DMA_SRBM_READ_PACKET ((9 << 28) | \ + (1 << 27)) + /* async DMA Packet types */ #define DMA_PACKET_WRITE 0x2 #define DMA_PACKET_COPY 0x3 diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index 74f06d540591..279801ca5110 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -644,6 +644,7 @@ int r100_pci_gart_init(struct radeon_device *rdev) return r; rdev->gart.table_size = rdev->gart.num_gpu_pages * 4; rdev->asic->gart.tlb_flush = &r100_pci_gart_tlb_flush; + rdev->asic->gart.get_page_entry = &r100_pci_gart_get_page_entry; rdev->asic->gart.set_page = &r100_pci_gart_set_page; return radeon_gart_table_ram_alloc(rdev); } @@ -681,11 +682,16 @@ void r100_pci_gart_disable(struct radeon_device *rdev) WREG32(RADEON_AIC_HI_ADDR, 0); } +uint64_t r100_pci_gart_get_page_entry(uint64_t addr, uint32_t flags) +{ + return addr; +} + void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags) + uint64_t entry) { u32 *gtt = rdev->gart.ptr; - gtt[i] = cpu_to_le32(lower_32_bits(addr)); + gtt[i] = cpu_to_le32(lower_32_bits(entry)); } void r100_pci_gart_fini(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c index 064ad5569cca..08d68f3e13e9 100644 --- a/drivers/gpu/drm/radeon/r300.c +++ b/drivers/gpu/drm/radeon/r300.c @@ -73,11 +73,8 @@ void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev) #define R300_PTE_WRITEABLE (1 << 2) #define R300_PTE_READABLE (1 << 3) -void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags) +uint64_t rv370_pcie_gart_get_page_entry(uint64_t addr, uint32_t flags) { - void __iomem *ptr = rdev->gart.ptr; - addr = (lower_32_bits(addr) >> 8) | ((upper_32_bits(addr) & 0xff) << 24); if (flags & RADEON_GART_PAGE_READ) @@ -86,10 +83,18 @@ void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i, addr |= R300_PTE_WRITEABLE; if (!(flags & RADEON_GART_PAGE_SNOOP)) addr |= R300_PTE_UNSNOOPED; + return addr; +} + +void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t entry) +{ + void __iomem *ptr = rdev->gart.ptr; + /* on x86 we want this to be CPU endian, on powerpc * on powerpc without HW swappers, it'll get swapped on way * into VRAM - so no need for cpu_to_le32 on VRAM tables */ - writel(addr, ((void __iomem *)ptr) + (i * 4)); + writel(entry, ((void __iomem *)ptr) + (i * 4)); } int rv370_pcie_gart_init(struct radeon_device *rdev) @@ -109,6 +114,7 @@ int rv370_pcie_gart_init(struct radeon_device *rdev) DRM_ERROR("Failed to register debugfs file for PCIE gart !\n"); rdev->gart.table_size = rdev->gart.num_gpu_pages * 4; rdev->asic->gart.tlb_flush = &rv370_pcie_gart_tlb_flush; + rdev->asic->gart.get_page_entry = &rv370_pcie_gart_get_page_entry; rdev->asic->gart.set_page = &rv370_pcie_gart_set_page; return radeon_gart_table_vram_alloc(rdev); } diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 54529b837afa..3f2a8d3febca 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -242,6 +242,7 @@ bool radeon_get_bios(struct radeon_device *rdev); * Dummy page */ struct radeon_dummy_page { + uint64_t entry; struct page *page; dma_addr_t addr; }; @@ -645,7 +646,7 @@ struct radeon_gart { unsigned num_cpu_pages; unsigned table_size; struct page **pages; - dma_addr_t *pages_addr; + uint64_t *pages_entry; bool ready; }; @@ -1847,8 +1848,9 @@ struct radeon_asic { /* gart */ struct { void (*tlb_flush)(struct radeon_device *rdev); + uint64_t (*get_page_entry)(uint64_t addr, uint32_t flags); void (*set_page)(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags); + uint64_t entry); } gart; struct { int (*init)(struct radeon_device *rdev); @@ -2852,7 +2854,8 @@ static inline void radeon_ring_write(struct radeon_ring *ring, uint32_t v) #define radeon_vga_set_state(rdev, state) (rdev)->asic->vga_set_state((rdev), (state)) #define radeon_asic_reset(rdev) (rdev)->asic->asic_reset((rdev)) #define radeon_gart_tlb_flush(rdev) (rdev)->asic->gart.tlb_flush((rdev)) -#define radeon_gart_set_page(rdev, i, p, f) (rdev)->asic->gart.set_page((rdev), (i), (p), (f)) +#define radeon_gart_get_page_entry(a, f) (rdev)->asic->gart.get_page_entry((a), (f)) +#define radeon_gart_set_page(rdev, i, e) (rdev)->asic->gart.set_page((rdev), (i), (e)) #define radeon_asic_vm_init(rdev) (rdev)->asic->vm.init((rdev)) #define radeon_asic_vm_fini(rdev) (rdev)->asic->vm.fini((rdev)) #define radeon_asic_vm_copy_pages(rdev, ib, pe, src, count) ((rdev)->asic->vm.copy_pages((rdev), (ib), (pe), (src), (count))) diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index 850de57069be..ed0e10eee2dc 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -159,11 +159,13 @@ void radeon_agp_disable(struct radeon_device *rdev) DRM_INFO("Forcing AGP to PCIE mode\n"); rdev->flags |= RADEON_IS_PCIE; rdev->asic->gart.tlb_flush = &rv370_pcie_gart_tlb_flush; + rdev->asic->gart.get_page_entry = &rv370_pcie_gart_get_page_entry; rdev->asic->gart.set_page = &rv370_pcie_gart_set_page; } else { DRM_INFO("Forcing AGP to PCI mode\n"); rdev->flags |= RADEON_IS_PCI; rdev->asic->gart.tlb_flush = &r100_pci_gart_tlb_flush; + rdev->asic->gart.get_page_entry = &r100_pci_gart_get_page_entry; rdev->asic->gart.set_page = &r100_pci_gart_set_page; } rdev->mc.gtt_size = radeon_gart_size * 1024 * 1024; @@ -199,6 +201,7 @@ static struct radeon_asic r100_asic = { .mc_wait_for_idle = &r100_mc_wait_for_idle, .gart = { .tlb_flush = &r100_pci_gart_tlb_flush, + .get_page_entry = &r100_pci_gart_get_page_entry, .set_page = &r100_pci_gart_set_page, }, .ring = { @@ -265,6 +268,7 @@ static struct radeon_asic r200_asic = { .mc_wait_for_idle = &r100_mc_wait_for_idle, .gart = { .tlb_flush = &r100_pci_gart_tlb_flush, + .get_page_entry = &r100_pci_gart_get_page_entry, .set_page = &r100_pci_gart_set_page, }, .ring = { @@ -333,6 +337,20 @@ static struct radeon_asic_ring r300_gfx_ring = { .set_wptr = &r100_gfx_set_wptr, }; +static struct radeon_asic_ring rv515_gfx_ring = { + .ib_execute = &r100_ring_ib_execute, + .emit_fence = &r300_fence_ring_emit, + .emit_semaphore = &r100_semaphore_ring_emit, + .cs_parse = &r300_cs_parse, + .ring_start = &rv515_ring_start, + .ring_test = &r100_ring_test, + .ib_test = &r100_ib_test, + .is_lockup = &r100_gpu_is_lockup, + .get_rptr = &r100_gfx_get_rptr, + .get_wptr = &r100_gfx_get_wptr, + .set_wptr = &r100_gfx_set_wptr, +}; + static struct radeon_asic r300_asic = { .init = &r300_init, .fini = &r300_fini, @@ -345,6 +363,7 @@ static struct radeon_asic r300_asic = { .mc_wait_for_idle = &r300_mc_wait_for_idle, .gart = { .tlb_flush = &r100_pci_gart_tlb_flush, + .get_page_entry = &r100_pci_gart_get_page_entry, .set_page = &r100_pci_gart_set_page, }, .ring = { @@ -411,6 +430,7 @@ static struct radeon_asic r300_asic_pcie = { .mc_wait_for_idle = &r300_mc_wait_for_idle, .gart = { .tlb_flush = &rv370_pcie_gart_tlb_flush, + .get_page_entry = &rv370_pcie_gart_get_page_entry, .set_page = &rv370_pcie_gart_set_page, }, .ring = { @@ -477,6 +497,7 @@ static struct radeon_asic r420_asic = { .mc_wait_for_idle = &r300_mc_wait_for_idle, .gart = { .tlb_flush = &rv370_pcie_gart_tlb_flush, + .get_page_entry = &rv370_pcie_gart_get_page_entry, .set_page = &rv370_pcie_gart_set_page, }, .ring = { @@ -543,6 +564,7 @@ static struct radeon_asic rs400_asic = { .mc_wait_for_idle = &rs400_mc_wait_for_idle, .gart = { .tlb_flush = &rs400_gart_tlb_flush, + .get_page_entry = &rs400_gart_get_page_entry, .set_page = &rs400_gart_set_page, }, .ring = { @@ -609,6 +631,7 @@ static struct radeon_asic rs600_asic = { .mc_wait_for_idle = &rs600_mc_wait_for_idle, .gart = { .tlb_flush = &rs600_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -677,6 +700,7 @@ static struct radeon_asic rs690_asic = { .mc_wait_for_idle = &rs690_mc_wait_for_idle, .gart = { .tlb_flush = &rs400_gart_tlb_flush, + .get_page_entry = &rs400_gart_get_page_entry, .set_page = &rs400_gart_set_page, }, .ring = { @@ -745,10 +769,11 @@ static struct radeon_asic rv515_asic = { .mc_wait_for_idle = &rv515_mc_wait_for_idle, .gart = { .tlb_flush = &rv370_pcie_gart_tlb_flush, + .get_page_entry = &rv370_pcie_gart_get_page_entry, .set_page = &rv370_pcie_gart_set_page, }, .ring = { - [RADEON_RING_TYPE_GFX_INDEX] = &r300_gfx_ring + [RADEON_RING_TYPE_GFX_INDEX] = &rv515_gfx_ring }, .irq = { .set = &rs600_irq_set, @@ -811,10 +836,11 @@ static struct radeon_asic r520_asic = { .mc_wait_for_idle = &r520_mc_wait_for_idle, .gart = { .tlb_flush = &rv370_pcie_gart_tlb_flush, + .get_page_entry = &rv370_pcie_gart_get_page_entry, .set_page = &rv370_pcie_gart_set_page, }, .ring = { - [RADEON_RING_TYPE_GFX_INDEX] = &r300_gfx_ring + [RADEON_RING_TYPE_GFX_INDEX] = &rv515_gfx_ring }, .irq = { .set = &rs600_irq_set, @@ -905,6 +931,7 @@ static struct radeon_asic r600_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &r600_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -990,6 +1017,7 @@ static struct radeon_asic rv6xx_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &r600_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -1081,6 +1109,7 @@ static struct radeon_asic rs780_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &r600_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -1185,6 +1214,7 @@ static struct radeon_asic rv770_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &r600_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -1303,6 +1333,7 @@ static struct radeon_asic evergreen_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &evergreen_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -1395,6 +1426,7 @@ static struct radeon_asic sumo_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &evergreen_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -1486,6 +1518,7 @@ static struct radeon_asic btc_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &evergreen_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .ring = { @@ -1621,6 +1654,7 @@ static struct radeon_asic cayman_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &cayman_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .vm = { @@ -1724,6 +1758,7 @@ static struct radeon_asic trinity_asic = { .get_gpu_clock_counter = &r600_get_gpu_clock_counter, .gart = { .tlb_flush = &cayman_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .vm = { @@ -1857,6 +1892,7 @@ static struct radeon_asic si_asic = { .get_gpu_clock_counter = &si_get_gpu_clock_counter, .gart = { .tlb_flush = &si_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .vm = { @@ -2018,6 +2054,7 @@ static struct radeon_asic ci_asic = { .get_gpu_clock_counter = &cik_get_gpu_clock_counter, .gart = { .tlb_flush = &cik_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .vm = { @@ -2125,6 +2162,7 @@ static struct radeon_asic kv_asic = { .get_gpu_clock_counter = &cik_get_gpu_clock_counter, .gart = { .tlb_flush = &cik_pcie_gart_tlb_flush, + .get_page_entry = &rs600_gart_get_page_entry, .set_page = &rs600_gart_set_page, }, .vm = { diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h index 2a45d548d5ec..8d787d115653 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.h +++ b/drivers/gpu/drm/radeon/radeon_asic.h @@ -67,8 +67,9 @@ bool r100_gpu_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp); int r100_asic_reset(struct radeon_device *rdev); u32 r100_get_vblank_counter(struct radeon_device *rdev, int crtc); void r100_pci_gart_tlb_flush(struct radeon_device *rdev); +uint64_t r100_pci_gart_get_page_entry(uint64_t addr, uint32_t flags); void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags); + uint64_t entry); void r100_ring_start(struct radeon_device *rdev, struct radeon_ring *ring); int r100_irq_set(struct radeon_device *rdev); int r100_irq_process(struct radeon_device *rdev); @@ -172,8 +173,9 @@ extern void r300_fence_ring_emit(struct radeon_device *rdev, struct radeon_fence *fence); extern int r300_cs_parse(struct radeon_cs_parser *p); extern void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev); +extern uint64_t rv370_pcie_gart_get_page_entry(uint64_t addr, uint32_t flags); extern void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags); + uint64_t entry); extern void rv370_set_pcie_lanes(struct radeon_device *rdev, int lanes); extern int rv370_get_pcie_lanes(struct radeon_device *rdev); extern void r300_set_reg_safe(struct radeon_device *rdev); @@ -208,8 +210,9 @@ extern void rs400_fini(struct radeon_device *rdev); extern int rs400_suspend(struct radeon_device *rdev); extern int rs400_resume(struct radeon_device *rdev); void rs400_gart_tlb_flush(struct radeon_device *rdev); +uint64_t rs400_gart_get_page_entry(uint64_t addr, uint32_t flags); void rs400_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags); + uint64_t entry); uint32_t rs400_mc_rreg(struct radeon_device *rdev, uint32_t reg); void rs400_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); int rs400_gart_init(struct radeon_device *rdev); @@ -232,8 +235,9 @@ int rs600_irq_process(struct radeon_device *rdev); void rs600_irq_disable(struct radeon_device *rdev); u32 rs600_get_vblank_counter(struct radeon_device *rdev, int crtc); void rs600_gart_tlb_flush(struct radeon_device *rdev); +uint64_t rs600_gart_get_page_entry(uint64_t addr, uint32_t flags); void rs600_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags); + uint64_t entry); uint32_t rs600_mc_rreg(struct radeon_device *rdev, uint32_t reg); void rs600_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v); void rs600_bandwidth_update(struct radeon_device *rdev); diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 0ec65168f331..bd7519fdd3f4 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -774,6 +774,8 @@ int radeon_dummy_page_init(struct radeon_device *rdev) rdev->dummy_page.page = NULL; return -ENOMEM; } + rdev->dummy_page.entry = radeon_gart_get_page_entry(rdev->dummy_page.addr, + RADEON_GART_PAGE_DUMMY); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c index 84146d5901aa..5450fa95a47e 100644 --- a/drivers/gpu/drm/radeon/radeon_gart.c +++ b/drivers/gpu/drm/radeon/radeon_gart.c @@ -165,6 +165,19 @@ int radeon_gart_table_vram_pin(struct radeon_device *rdev) radeon_bo_unpin(rdev->gart.robj); radeon_bo_unreserve(rdev->gart.robj); rdev->gart.table_addr = gpu_addr; + + if (!r) { + int i; + + /* We might have dropped some GART table updates while it wasn't + * mapped, restore all entries + */ + for (i = 0; i < rdev->gart.num_gpu_pages; i++) + radeon_gart_set_page(rdev, i, rdev->gart.pages_entry[i]); + mb(); + radeon_gart_tlb_flush(rdev); + } + return r; } @@ -228,7 +241,6 @@ void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset, unsigned t; unsigned p; int i, j; - u64 page_base; if (!rdev->gart.ready) { WARN(1, "trying to unbind memory from uninitialized GART !\n"); @@ -239,14 +251,12 @@ void radeon_gart_unbind(struct radeon_device *rdev, unsigned offset, for (i = 0; i < pages; i++, p++) { if (rdev->gart.pages[p]) { rdev->gart.pages[p] = NULL; - rdev->gart.pages_addr[p] = rdev->dummy_page.addr; - page_base = rdev->gart.pages_addr[p]; for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) { + rdev->gart.pages_entry[t] = rdev->dummy_page.entry; if (rdev->gart.ptr) { - radeon_gart_set_page(rdev, t, page_base, - RADEON_GART_PAGE_DUMMY); + radeon_gart_set_page(rdev, t, + rdev->dummy_page.entry); } - page_base += RADEON_GPU_PAGE_SIZE; } } } @@ -274,7 +284,7 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset, { unsigned t; unsigned p; - uint64_t page_base; + uint64_t page_base, page_entry; int i, j; if (!rdev->gart.ready) { @@ -285,14 +295,15 @@ int radeon_gart_bind(struct radeon_device *rdev, unsigned offset, p = t / (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); for (i = 0; i < pages; i++, p++) { - rdev->gart.pages_addr[p] = dma_addr[i]; rdev->gart.pages[p] = pagelist[i]; - if (rdev->gart.ptr) { - page_base = rdev->gart.pages_addr[p]; - for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) { - radeon_gart_set_page(rdev, t, page_base, flags); - page_base += RADEON_GPU_PAGE_SIZE; + page_base = dma_addr[i]; + for (j = 0; j < (PAGE_SIZE / RADEON_GPU_PAGE_SIZE); j++, t++) { + page_entry = radeon_gart_get_page_entry(page_base, flags); + rdev->gart.pages_entry[t] = page_entry; + if (rdev->gart.ptr) { + radeon_gart_set_page(rdev, t, page_entry); } + page_base += RADEON_GPU_PAGE_SIZE; } } mb(); @@ -334,16 +345,15 @@ int radeon_gart_init(struct radeon_device *rdev) radeon_gart_fini(rdev); return -ENOMEM; } - rdev->gart.pages_addr = vzalloc(sizeof(dma_addr_t) * - rdev->gart.num_cpu_pages); - if (rdev->gart.pages_addr == NULL) { + rdev->gart.pages_entry = vmalloc(sizeof(uint64_t) * + rdev->gart.num_gpu_pages); + if (rdev->gart.pages_entry == NULL) { radeon_gart_fini(rdev); return -ENOMEM; } /* set GART entry to point to the dummy page by default */ - for (i = 0; i < rdev->gart.num_cpu_pages; i++) { - rdev->gart.pages_addr[i] = rdev->dummy_page.addr; - } + for (i = 0; i < rdev->gart.num_gpu_pages; i++) + rdev->gart.pages_entry[i] = rdev->dummy_page.entry; return 0; } @@ -356,15 +366,15 @@ int radeon_gart_init(struct radeon_device *rdev) */ void radeon_gart_fini(struct radeon_device *rdev) { - if (rdev->gart.pages && rdev->gart.pages_addr && rdev->gart.ready) { + if (rdev->gart.ready) { /* unbind pages */ radeon_gart_unbind(rdev, 0, rdev->gart.num_cpu_pages); } rdev->gart.ready = false; vfree(rdev->gart.pages); - vfree(rdev->gart.pages_addr); + vfree(rdev->gart.pages_entry); rdev->gart.pages = NULL; - rdev->gart.pages_addr = NULL; + rdev->gart.pages_entry = NULL; radeon_dummy_page_fini(rdev); } diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index fe48f229043e..d0b4f7d1140d 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -394,10 +394,9 @@ int radeon_gem_set_domain_ioctl(struct drm_device *dev, void *data, return r; } -static int radeon_mode_mmap(struct drm_file *filp, - struct drm_device *dev, - uint32_t handle, bool dumb, - uint64_t *offset_p) +int radeon_mode_dumb_mmap(struct drm_file *filp, + struct drm_device *dev, + uint32_t handle, uint64_t *offset_p) { struct drm_gem_object *gobj; struct radeon_bo *robj; @@ -406,14 +405,6 @@ static int radeon_mode_mmap(struct drm_file *filp, if (gobj == NULL) { return -ENOENT; } - - /* - * We don't allow dumb mmaps on objects created using another - * interface. - */ - WARN_ONCE(dumb && !(gobj->dumb || gobj->import_attach), - "Illegal dumb map of GPU buffer.\n"); - robj = gem_to_radeon_bo(gobj); if (radeon_ttm_tt_has_userptr(robj->tbo.ttm)) { drm_gem_object_unreference_unlocked(gobj); @@ -424,20 +415,12 @@ static int radeon_mode_mmap(struct drm_file *filp, return 0; } -int radeon_mode_dumb_mmap(struct drm_file *filp, - struct drm_device *dev, - uint32_t handle, uint64_t *offset_p) -{ - return radeon_mode_mmap(filp, dev, handle, true, offset_p); -} - int radeon_gem_mmap_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) { struct drm_radeon_gem_mmap *args = data; - return radeon_mode_mmap(filp, dev, args->handle, false, - &args->addr_ptr); + return radeon_mode_dumb_mmap(filp, dev, args->handle, &args->addr_ptr); } int radeon_gem_busy_ioctl(struct drm_device *dev, void *data, @@ -593,7 +576,7 @@ error_unreserve: error_free: drm_free_large(vm_bos); - if (r) + if (r && r != -ERESTARTSYS) DRM_ERROR("Couldn't update BO_VA (%d)\n", r); } @@ -763,7 +746,6 @@ int radeon_mode_dumb_create(struct drm_file *file_priv, return -ENOMEM; r = drm_gem_handle_create(file_priv, gobj, &handle); - gobj->dumb = true; /* drop reference from allocate - handle holds it now */ drm_gem_object_unreference_unlocked(gobj); if (r) { diff --git a/drivers/gpu/drm/radeon/radeon_kfd.c b/drivers/gpu/drm/radeon/radeon_kfd.c index 065d02068ec3..bef9a0953284 100644 --- a/drivers/gpu/drm/radeon/radeon_kfd.c +++ b/drivers/gpu/drm/radeon/radeon_kfd.c @@ -28,6 +28,8 @@ #include "cikd.h" #include "cik_reg.h" #include "radeon_kfd.h" +#include "radeon_ucode.h" +#include <linux/firmware.h> #define CIK_PIPE_PER_MEC (4) @@ -49,6 +51,7 @@ static uint64_t get_vmem_size(struct kgd_dev *kgd); static uint64_t get_gpu_clock_counter(struct kgd_dev *kgd); static uint32_t get_max_engine_clock_in_mhz(struct kgd_dev *kgd); +static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type); /* * Register access functions @@ -69,7 +72,7 @@ static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id, static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id, uint32_t queue_id, uint32_t __user *wptr); -static bool kgd_hqd_is_occupies(struct kgd_dev *kgd, uint64_t queue_address, +static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address, uint32_t pipe_id, uint32_t queue_id); static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, @@ -89,14 +92,16 @@ static const struct kfd2kgd_calls kfd2kgd = { .init_memory = kgd_init_memory, .init_pipeline = kgd_init_pipeline, .hqd_load = kgd_hqd_load, - .hqd_is_occupies = kgd_hqd_is_occupies, + .hqd_is_occupied = kgd_hqd_is_occupied, .hqd_destroy = kgd_hqd_destroy, + .get_fw_version = get_fw_version }; static const struct kgd2kfd_calls *kgd2kfd; bool radeon_kfd_init(void) { +#if defined(CONFIG_HSA_AMD_MODULE) bool (*kgd2kfd_init_p)(unsigned, const struct kfd2kgd_calls*, const struct kgd2kfd_calls**); @@ -113,6 +118,17 @@ bool radeon_kfd_init(void) } return true; +#elif defined(CONFIG_HSA_AMD) + if (!kgd2kfd_init(KFD_INTERFACE_VERSION, &kfd2kgd, &kgd2kfd)) { + kgd2kfd = NULL; + + return false; + } + + return true; +#else + return false; +#endif } void radeon_kfd_fini(void) @@ -374,6 +390,10 @@ static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid, cpu_relax(); write_register(kgd, ATC_VMID_PASID_MAPPING_UPDATE_STATUS, 1U << vmid); + /* Mapping vmid to pasid also for IH block */ + write_register(kgd, IH_VMID_0_LUT + vmid * sizeof(uint32_t), + pasid_mapping); + return 0; } @@ -416,7 +436,7 @@ static int kgd_init_memory(struct kgd_dev *kgd) static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id, uint32_t hpd_size, uint64_t hpd_gpu_addr) { - uint32_t mec = (++pipe_id / CIK_PIPE_PER_MEC) + 1; + uint32_t mec = (pipe_id / CIK_PIPE_PER_MEC) + 1; uint32_t pipe = (pipe_id % CIK_PIPE_PER_MEC); lock_srbm(kgd, mec, pipe, 0, 0); @@ -513,7 +533,7 @@ static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id, return 0; } -static bool kgd_hqd_is_occupies(struct kgd_dev *kgd, uint64_t queue_address, +static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address, uint32_t pipe_id, uint32_t queue_id) { uint32_t act; @@ -552,6 +572,7 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, if (timeout == 0) { pr_err("kfd: cp queue preemption time out (%dms)\n", temp); + release_queue(kgd); return -ETIME; } msleep(20); @@ -561,3 +582,52 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, release_queue(kgd); return 0; } + +static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type) +{ + struct radeon_device *rdev = (struct radeon_device *) kgd; + const union radeon_firmware_header *hdr; + + BUG_ON(kgd == NULL || rdev->mec_fw == NULL); + + switch (type) { + case KGD_ENGINE_PFP: + hdr = (const union radeon_firmware_header *) rdev->pfp_fw->data; + break; + + case KGD_ENGINE_ME: + hdr = (const union radeon_firmware_header *) rdev->me_fw->data; + break; + + case KGD_ENGINE_CE: + hdr = (const union radeon_firmware_header *) rdev->ce_fw->data; + break; + + case KGD_ENGINE_MEC1: + hdr = (const union radeon_firmware_header *) rdev->mec_fw->data; + break; + + case KGD_ENGINE_MEC2: + hdr = (const union radeon_firmware_header *) + rdev->mec2_fw->data; + break; + + case KGD_ENGINE_RLC: + hdr = (const union radeon_firmware_header *) rdev->rlc_fw->data; + break; + + case KGD_ENGINE_SDMA: + hdr = (const union radeon_firmware_header *) + rdev->sdma_fw->data; + break; + + default: + return 0; + } + + if (hdr == NULL) + return 0; + + /* Only 12 bit in use*/ + return hdr->common.ucode_version; +} diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 7d68223eb469..86fc56434b28 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -529,9 +529,6 @@ int radeon_bo_list_validate(struct radeon_device *rdev, u32 current_domain = radeon_mem_type_to_domain(bo->tbo.mem.mem_type); - WARN_ONCE(bo->gem_base.dumb, - "GPU use of dumb buffer is illegal.\n"); - /* Check if this buffer will be moved and don't move it * if we have moved too many buffers for this IB already. * diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index 32522cc940a1..f7da8fe96a66 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -1287,8 +1287,39 @@ dpm_failed: return ret; } +struct radeon_dpm_quirk { + u32 chip_vendor; + u32 chip_device; + u32 subsys_vendor; + u32 subsys_device; +}; + +/* cards with dpm stability problems */ +static struct radeon_dpm_quirk radeon_dpm_quirk_list[] = { + /* TURKS - https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1386534 */ + { PCI_VENDOR_ID_ATI, 0x6759, 0x1682, 0x3195 }, + /* TURKS - https://bugzilla.kernel.org/show_bug.cgi?id=83731 */ + { PCI_VENDOR_ID_ATI, 0x6840, 0x1179, 0xfb81 }, + { 0, 0, 0, 0 }, +}; + int radeon_pm_init(struct radeon_device *rdev) { + struct radeon_dpm_quirk *p = radeon_dpm_quirk_list; + bool disable_dpm = false; + + /* Apply dpm quirks */ + while (p && p->chip_device != 0) { + if (rdev->pdev->vendor == p->chip_vendor && + rdev->pdev->device == p->chip_device && + rdev->pdev->subsystem_vendor == p->subsys_vendor && + rdev->pdev->subsystem_device == p->subsys_device) { + disable_dpm = true; + break; + } + ++p; + } + /* enable dpm on rv6xx+ */ switch (rdev->family) { case CHIP_RV610: @@ -1344,6 +1375,8 @@ int radeon_pm_init(struct radeon_device *rdev) (!(rdev->flags & RADEON_IS_IGP)) && (!rdev->smc_fw)) rdev->pm.pm_method = PM_METHOD_PROFILE; + else if (disable_dpm && (radeon_dpm == -1)) + rdev->pm.pm_method = PM_METHOD_PROFILE; else if (radeon_dpm == 0) rdev->pm.pm_method = PM_METHOD_PROFILE; else diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c index 535403e0c8a2..15aee723db77 100644 --- a/drivers/gpu/drm/radeon/radeon_state.c +++ b/drivers/gpu/drm/radeon/radeon_state.c @@ -1703,7 +1703,7 @@ static int radeon_cp_dispatch_texture(struct drm_device * dev, u32 format; u32 *buffer; const u8 __user *data; - int size, dwords, tex_width, blit_width, spitch; + unsigned int size, dwords, tex_width, blit_width, spitch; u32 height; int i; u32 texpitch, microtile; diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c index cde48c42b30a..06d2246d07f1 100644 --- a/drivers/gpu/drm/radeon/radeon_vm.c +++ b/drivers/gpu/drm/radeon/radeon_vm.c @@ -587,10 +587,8 @@ uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr) uint64_t result; /* page table offset */ - result = rdev->gart.pages_addr[addr >> PAGE_SHIFT]; - - /* in case cpu page size != gpu page size*/ - result |= addr & (~PAGE_MASK); + result = rdev->gart.pages_entry[addr >> RADEON_GPU_PAGE_SHIFT]; + result &= ~RADEON_GPU_PAGE_MASK; return result; } diff --git a/drivers/gpu/drm/radeon/rs400.c b/drivers/gpu/drm/radeon/rs400.c index c5799f16aa4b..34e3235f41d2 100644 --- a/drivers/gpu/drm/radeon/rs400.c +++ b/drivers/gpu/drm/radeon/rs400.c @@ -212,11 +212,9 @@ void rs400_gart_fini(struct radeon_device *rdev) #define RS400_PTE_WRITEABLE (1 << 2) #define RS400_PTE_READABLE (1 << 3) -void rs400_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags) +uint64_t rs400_gart_get_page_entry(uint64_t addr, uint32_t flags) { uint32_t entry; - u32 *gtt = rdev->gart.ptr; entry = (lower_32_bits(addr) & PAGE_MASK) | ((upper_32_bits(addr) & 0xff) << 4); @@ -226,8 +224,14 @@ void rs400_gart_set_page(struct radeon_device *rdev, unsigned i, entry |= RS400_PTE_WRITEABLE; if (!(flags & RADEON_GART_PAGE_SNOOP)) entry |= RS400_PTE_UNSNOOPED; - entry = cpu_to_le32(entry); - gtt[i] = entry; + return entry; +} + +void rs400_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t entry) +{ + u32 *gtt = rdev->gart.ptr; + gtt[i] = cpu_to_le32(lower_32_bits(entry)); } int rs400_mc_wait_for_idle(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c index 9acb1c3c005b..74bce91aecc1 100644 --- a/drivers/gpu/drm/radeon/rs600.c +++ b/drivers/gpu/drm/radeon/rs600.c @@ -625,11 +625,8 @@ static void rs600_gart_fini(struct radeon_device *rdev) radeon_gart_table_vram_free(rdev); } -void rs600_gart_set_page(struct radeon_device *rdev, unsigned i, - uint64_t addr, uint32_t flags) +uint64_t rs600_gart_get_page_entry(uint64_t addr, uint32_t flags) { - void __iomem *ptr = (void *)rdev->gart.ptr; - addr = addr & 0xFFFFFFFFFFFFF000ULL; addr |= R600_PTE_SYSTEM; if (flags & RADEON_GART_PAGE_VALID) @@ -640,7 +637,14 @@ void rs600_gart_set_page(struct radeon_device *rdev, unsigned i, addr |= R600_PTE_WRITEABLE; if (flags & RADEON_GART_PAGE_SNOOP) addr |= R600_PTE_SNOOPED; - writeq(addr, ptr + (i * 8)); + return addr; +} + +void rs600_gart_set_page(struct radeon_device *rdev, unsigned i, + uint64_t entry) +{ + void __iomem *ptr = (void *)rdev->gart.ptr; + writeq(entry, ptr + (i * 8)); } int rs600_irq_set(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c index 60df444bd075..5d89b874a1a2 100644 --- a/drivers/gpu/drm/radeon/si.c +++ b/drivers/gpu/drm/radeon/si.c @@ -5057,6 +5057,16 @@ void si_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring, radeon_ring_write(ring, 0); radeon_ring_write(ring, 1 << vm_id); + /* wait for the invalidate to complete */ + radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); + radeon_ring_write(ring, (WAIT_REG_MEM_FUNCTION(0) | /* always */ + WAIT_REG_MEM_ENGINE(0))); /* me */ + radeon_ring_write(ring, VM_INVALIDATE_REQUEST >> 2); + radeon_ring_write(ring, 0); + radeon_ring_write(ring, 0); /* ref */ + radeon_ring_write(ring, 0); /* mask */ + radeon_ring_write(ring, 0x20); /* poll interval */ + /* sync PFP to ME, otherwise we might get invalid PFP reads */ radeon_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0)); radeon_ring_write(ring, 0x0); diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c index f5cc777e1c5f..83207929fc62 100644 --- a/drivers/gpu/drm/radeon/si_dma.c +++ b/drivers/gpu/drm/radeon/si_dma.c @@ -123,7 +123,6 @@ void si_dma_vm_write_pages(struct radeon_device *rdev, for (; ndw > 0; ndw -= 2, --count, pe += 8) { if (flags & R600_PTE_SYSTEM) { value = radeon_vm_map_gart(rdev, addr); - value &= 0xFFFFFFFFFFFFF000ULL; } else if (flags & R600_PTE_VALID) { value = addr; } else { @@ -206,6 +205,14 @@ void si_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring, radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); radeon_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST >> 2)); radeon_ring_write(ring, 1 << vm_id); + + /* wait for invalidate to complete */ + radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0)); + radeon_ring_write(ring, VM_INVALIDATE_REQUEST); + radeon_ring_write(ring, 0xff << 16); /* retry */ + radeon_ring_write(ring, 1 << vm_id); /* mask */ + radeon_ring_write(ring, 0); /* value */ + radeon_ring_write(ring, (0 << 28) | 0x20); /* func(always) | poll interval */ } /** diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c index 32e354b8b0ab..eff8a6444956 100644 --- a/drivers/gpu/drm/radeon/si_dpm.c +++ b/drivers/gpu/drm/radeon/si_dpm.c @@ -2908,6 +2908,22 @@ static int si_init_smc_spll_table(struct radeon_device *rdev) return ret; } +struct si_dpm_quirk { + u32 chip_vendor; + u32 chip_device; + u32 subsys_vendor; + u32 subsys_device; + u32 max_sclk; + u32 max_mclk; +}; + +/* cards with dpm stability problems */ +static struct si_dpm_quirk si_dpm_quirk_list[] = { + /* PITCAIRN - https://bugs.freedesktop.org/show_bug.cgi?id=76490 */ + { PCI_VENDOR_ID_ATI, 0x6810, 0x1462, 0x3036, 0, 120000 }, + { 0, 0, 0, 0 }, +}; + static void si_apply_state_adjust_rules(struct radeon_device *rdev, struct radeon_ps *rps) { @@ -2918,7 +2934,22 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev, u32 mclk, sclk; u16 vddc, vddci; u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc; + u32 max_sclk = 0, max_mclk = 0; int i; + struct si_dpm_quirk *p = si_dpm_quirk_list; + + /* Apply dpm quirks */ + while (p && p->chip_device != 0) { + if (rdev->pdev->vendor == p->chip_vendor && + rdev->pdev->device == p->chip_device && + rdev->pdev->subsystem_vendor == p->subsys_vendor && + rdev->pdev->subsystem_device == p->subsys_device) { + max_sclk = p->max_sclk; + max_mclk = p->max_mclk; + break; + } + ++p; + } if ((rdev->pm.dpm.new_active_crtc_count > 1) || ni_dpm_vblank_too_short(rdev)) @@ -2972,6 +3003,14 @@ static void si_apply_state_adjust_rules(struct radeon_device *rdev, if (ps->performance_levels[i].mclk > max_mclk_vddc) ps->performance_levels[i].mclk = max_mclk_vddc; } + if (max_mclk) { + if (ps->performance_levels[i].mclk > max_mclk) + ps->performance_levels[i].mclk = max_mclk; + } + if (max_sclk) { + if (ps->performance_levels[i].sclk > max_sclk) + ps->performance_levels[i].sclk = max_sclk; + } } /* XXX validate the min clocks required for display */ diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h index 4069be89e585..84999242c747 100644 --- a/drivers/gpu/drm/radeon/sid.h +++ b/drivers/gpu/drm/radeon/sid.h @@ -1632,6 +1632,23 @@ #define PACKET3_MPEG_INDEX 0x3A #define PACKET3_COPY_DW 0x3B #define PACKET3_WAIT_REG_MEM 0x3C +#define WAIT_REG_MEM_FUNCTION(x) ((x) << 0) + /* 0 - always + * 1 - < + * 2 - <= + * 3 - == + * 4 - != + * 5 - >= + * 6 - > + */ +#define WAIT_REG_MEM_MEM_SPACE(x) ((x) << 4) + /* 0 - reg + * 1 - mem + */ +#define WAIT_REG_MEM_ENGINE(x) ((x) << 8) + /* 0 - me + * 1 - pfp + */ #define PACKET3_MEM_WRITE 0x3D #define PACKET3_COPY_DATA 0x40 #define PACKET3_CP_DMA 0x41 @@ -1835,6 +1852,7 @@ #define DMA_PACKET_TRAP 0x7 #define DMA_PACKET_SRBM_WRITE 0x9 #define DMA_PACKET_CONSTANT_FILL 0xd +#define DMA_PACKET_POLL_REG_MEM 0xe #define DMA_PACKET_NOP 0xf #define VCE_STATUS 0x20004 |
