summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/radeon
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/radeon')
-rw-r--r--drivers/gpu/drm/radeon/cik.c11
-rw-r--r--drivers/gpu/drm/radeon/cik_sdma.c11
-rw-r--r--drivers/gpu/drm/radeon/ni.c10
-rw-r--r--drivers/gpu/drm/radeon/ni_dma.c7
-rw-r--r--drivers/gpu/drm/radeon/nid.h24
-rw-r--r--drivers/gpu/drm/radeon/r100.c10
-rw-r--r--drivers/gpu/drm/radeon/r300.c16
-rw-r--r--drivers/gpu/drm/radeon/radeon.h9
-rw-r--r--drivers/gpu/drm/radeon/radeon_asic.c42
-rw-r--r--drivers/gpu/drm/radeon/radeon_asic.h12
-rw-r--r--drivers/gpu/drm/radeon/radeon_benchmark.c13
-rw-r--r--drivers/gpu/drm/radeon/radeon_device.c2
-rw-r--r--drivers/gpu/drm/radeon/radeon_display.c3
-rw-r--r--drivers/gpu/drm/radeon/radeon_gart.c54
-rw-r--r--drivers/gpu/drm/radeon/radeon_gem.c8
-rw-r--r--drivers/gpu/drm/radeon/radeon_kfd.c2
-rw-r--r--drivers/gpu/drm/radeon/radeon_kms.c16
-rw-r--r--drivers/gpu/drm/radeon/radeon_pm.c33
-rw-r--r--drivers/gpu/drm/radeon/radeon_test.c8
-rw-r--r--drivers/gpu/drm/radeon/radeon_vm.c12
-rw-r--r--drivers/gpu/drm/radeon/rs400.c14
-rw-r--r--drivers/gpu/drm/radeon/rs600.c14
-rw-r--r--drivers/gpu/drm/radeon/si.c10
-rw-r--r--drivers/gpu/drm/radeon/si_dma.c9
-rw-r--r--drivers/gpu/drm/radeon/si_dpm.c39
-rw-r--r--drivers/gpu/drm/radeon/sid.h18
26 files changed, 329 insertions, 78 deletions
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/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_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c
index 9e7f23dd14bd..87d5fb21cb61 100644
--- a/drivers/gpu/drm/radeon/radeon_benchmark.c
+++ b/drivers/gpu/drm/radeon/radeon_benchmark.c
@@ -34,7 +34,8 @@
static int radeon_benchmark_do_move(struct radeon_device *rdev, unsigned size,
uint64_t saddr, uint64_t daddr,
- int flag, int n)
+ int flag, int n,
+ struct reservation_object *resv)
{
unsigned long start_jiffies;
unsigned long end_jiffies;
@@ -47,12 +48,12 @@ static int radeon_benchmark_do_move(struct radeon_device *rdev, unsigned size,
case RADEON_BENCHMARK_COPY_DMA:
fence = radeon_copy_dma(rdev, saddr, daddr,
size / RADEON_GPU_PAGE_SIZE,
- NULL);
+ resv);
break;
case RADEON_BENCHMARK_COPY_BLIT:
fence = radeon_copy_blit(rdev, saddr, daddr,
size / RADEON_GPU_PAGE_SIZE,
- NULL);
+ resv);
break;
default:
DRM_ERROR("Unknown copy method\n");
@@ -120,7 +121,8 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
if (rdev->asic->copy.dma) {
time = radeon_benchmark_do_move(rdev, size, saddr, daddr,
- RADEON_BENCHMARK_COPY_DMA, n);
+ RADEON_BENCHMARK_COPY_DMA, n,
+ dobj->tbo.resv);
if (time < 0)
goto out_cleanup;
if (time > 0)
@@ -130,7 +132,8 @@ static void radeon_benchmark_move(struct radeon_device *rdev, unsigned size,
if (rdev->asic->copy.blit) {
time = radeon_benchmark_do_move(rdev, size, saddr, daddr,
- RADEON_BENCHMARK_COPY_BLIT, n);
+ RADEON_BENCHMARK_COPY_BLIT, n,
+ dobj->tbo.resv);
if (time < 0)
goto out_cleanup;
if (time > 0)
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_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 102116902a07..913fafa597ad 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -960,6 +960,9 @@ void radeon_compute_pll_avivo(struct radeon_pll *pll,
if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV &&
pll->flags & RADEON_PLL_USE_REF_DIV)
ref_div_max = pll->reference_div;
+ else if (pll->flags & RADEON_PLL_PREFER_MINM_OVER_MAXP)
+ /* fix for problems on RS880 */
+ ref_div_max = min(pll->max_ref_div, 7u);
else
ref_div_max = pll->max_ref_div;
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 a46f73737994..ac3c1310b953 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -146,7 +146,8 @@ int radeon_gem_object_open(struct drm_gem_object *obj, struct drm_file *file_pri
struct radeon_bo_va *bo_va;
int r;
- if (rdev->family < CHIP_CAYMAN) {
+ if ((rdev->family < CHIP_CAYMAN) ||
+ (!rdev->accel_working)) {
return 0;
}
@@ -176,7 +177,8 @@ void radeon_gem_object_close(struct drm_gem_object *obj,
struct radeon_bo_va *bo_va;
int r;
- if (rdev->family < CHIP_CAYMAN) {
+ if ((rdev->family < CHIP_CAYMAN) ||
+ (!rdev->accel_working)) {
return;
}
@@ -576,7 +578,7 @@ 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);
}
diff --git a/drivers/gpu/drm/radeon/radeon_kfd.c b/drivers/gpu/drm/radeon/radeon_kfd.c
index 8bf87f1203cc..bef9a0953284 100644
--- a/drivers/gpu/drm/radeon/radeon_kfd.c
+++ b/drivers/gpu/drm/radeon/radeon_kfd.c
@@ -436,7 +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);
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 3cf9c1fa6475..686411e4e4f6 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -605,14 +605,14 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
return -ENOMEM;
}
- vm = &fpriv->vm;
- r = radeon_vm_init(rdev, vm);
- if (r) {
- kfree(fpriv);
- return r;
- }
-
if (rdev->accel_working) {
+ vm = &fpriv->vm;
+ r = radeon_vm_init(rdev, vm);
+ if (r) {
+ kfree(fpriv);
+ return r;
+ }
+
r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
if (r) {
radeon_vm_fini(rdev, vm);
@@ -668,9 +668,9 @@ void radeon_driver_postclose_kms(struct drm_device *dev,
radeon_vm_bo_rmv(rdev, vm->ib_bo_va);
radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
}
+ radeon_vm_fini(rdev, vm);
}
- radeon_vm_fini(rdev, vm);
kfree(fpriv);
file_priv->driver_priv = NULL;
}
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_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index 07b506b41008..791818165c76 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -119,11 +119,11 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
if (ring == R600_RING_TYPE_DMA_INDEX)
fence = radeon_copy_dma(rdev, gtt_addr, vram_addr,
size / RADEON_GPU_PAGE_SIZE,
- NULL);
+ vram_obj->tbo.resv);
else
fence = radeon_copy_blit(rdev, gtt_addr, vram_addr,
size / RADEON_GPU_PAGE_SIZE,
- NULL);
+ vram_obj->tbo.resv);
if (IS_ERR(fence)) {
DRM_ERROR("Failed GTT->VRAM copy %d\n", i);
r = PTR_ERR(fence);
@@ -170,11 +170,11 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
if (ring == R600_RING_TYPE_DMA_INDEX)
fence = radeon_copy_dma(rdev, vram_addr, gtt_addr,
size / RADEON_GPU_PAGE_SIZE,
- NULL);
+ vram_obj->tbo.resv);
else
fence = radeon_copy_blit(rdev, vram_addr, gtt_addr,
size / RADEON_GPU_PAGE_SIZE,
- NULL);
+ vram_obj->tbo.resv);
if (IS_ERR(fence)) {
DRM_ERROR("Failed VRAM->GTT copy %d\n", i);
r = PTR_ERR(fence);
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
index cde48c42b30a..2a5a4a9e772d 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;
}
@@ -745,9 +743,11 @@ static void radeon_vm_frag_ptes(struct radeon_device *rdev,
*/
/* NI is optimized for 256KB fragments, SI and newer for 64KB */
- uint64_t frag_flags = rdev->family == CHIP_CAYMAN ?
+ uint64_t frag_flags = ((rdev->family == CHIP_CAYMAN) ||
+ (rdev->family == CHIP_ARUBA)) ?
R600_PTE_FRAG_256KB : R600_PTE_FRAG_64KB;
- uint64_t frag_align = rdev->family == CHIP_CAYMAN ? 0x200 : 0x80;
+ uint64_t frag_align = ((rdev->family == CHIP_CAYMAN) ||
+ (rdev->family == CHIP_ARUBA)) ? 0x200 : 0x80;
uint64_t frag_start = ALIGN(pe_start, frag_align);
uint64_t frag_end = pe_end & ~(frag_align - 1);
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