Merge remote branch 'korg/drm-radeon-next' into drm-linus

* korg/drm-radeon-next:
  drm/radeon/kms: fix legacy get_engine/memory clock
  drm/radeon/kms/atom: atom parser fixes
  drm/radeon/kms: clean up atombios pll code
  drm/radeon/kms: clean up pll struct
  drm/radeon/kms/atom: fix crtc lock ordering
  drm/radeon: r6xx/r7xx possible security issue, system ram access
  drm/radeon/kms: r600/r700 don't test ib if ib initialization fails
  drm/radeon/kms: Forbid creation of framebuffer with no valid GEM object
  drm/radeon/kms: r600 handle irq vector ring overflow
  drm/radeon/kms: r600/r700 don't process IRQ if not initialized
  drm/radeon/kms: r600/r700 disable irq at suspend
  drm/radeon/kms/r4xx: cleanup atom path
  drm/radeon/kms: fix atombios_crtc_set_base
  drm/radeon/kms/atom: upstream parser updates
  drm/radeon/kms/atom: fix some parser bugs
  drm/radeon/kms: fix hardcoded mmio size in register functions
  drm/radeon/kms/r100: fix bug in CS parser
  drm/radeon/kms/r200: fix bug in CS parser
  drm/radeon/kms/r200: fix bug in CS parser

drivers/gpu/drm/radeon/atom.c

@@ -246,6 +246,9 @@ static uint32_t atom_get_src_int(atom_exec_context *ctx, uint8_t attr,
 		case ATOM_WS_ATTRIBUTES:
 			val = gctx->io_attr;
 			break;
+		case ATOM_WS_REGPTR:
+			val = gctx->reg_block;
+			break;
 		default:
 			val = ctx->ws[idx];
 		}
@@ -385,6 +388,32 @@ static uint32_t atom_get_src(atom_exec_context *ctx, uint8_t attr, int *ptr)
 	return atom_get_src_int(ctx, attr, ptr, NULL, 1);
 }
 
+static uint32_t atom_get_src_direct(atom_exec_context *ctx, uint8_t align, int *ptr)
+{
+	uint32_t val = 0xCDCDCDCD;
+
+	switch (align) {
+	case ATOM_SRC_DWORD:
+		val = U32(*ptr);
+		(*ptr) += 4;
+		break;
+	case ATOM_SRC_WORD0:
+	case ATOM_SRC_WORD8:
+	case ATOM_SRC_WORD16:
+		val = U16(*ptr);
+		(*ptr) += 2;
+		break;
+	case ATOM_SRC_BYTE0:
+	case ATOM_SRC_BYTE8:
+	case ATOM_SRC_BYTE16:
+	case ATOM_SRC_BYTE24:
+		val = U8(*ptr);
+		(*ptr)++;
+		break;
+	}
+	return val;
+}
+
 static uint32_t atom_get_dst(atom_exec_context *ctx, int arg, uint8_t attr,
 			     int *ptr, uint32_t *saved, int print)
 {
@@ -482,6 +511,9 @@ static void atom_put_dst(atom_exec_context *ctx, int arg, uint8_t attr,
 		case ATOM_WS_ATTRIBUTES:
 			gctx->io_attr = val;
 			break;
+		case ATOM_WS_REGPTR:
+			gctx->reg_block = val;
+			break;
 		default:
 			ctx->ws[idx] = val;
 		}
@@ -677,7 +709,7 @@ static void atom_op_mask(atom_exec_context *ctx, int *ptr, int arg)
 	SDEBUG("   dst: ");
 	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
 	SDEBUG("   src1: ");
-	src1 = atom_get_src(ctx, attr, ptr);
+	src1 = atom_get_src_direct(ctx, ((attr >> 3) & 7), ptr);
 	SDEBUG("   src2: ");
 	src2 = atom_get_src(ctx, attr, ptr);
 	dst &= src1;
@@ -809,6 +841,38 @@ static void atom_op_setregblock(atom_exec_context *ctx, int *ptr, int arg)
 	SDEBUG("   base: 0x%04X\n", ctx->ctx->reg_block);
 }
 
+static void atom_op_shift_left(atom_exec_context *ctx, int *ptr, int arg)
+{
+	uint8_t attr = U8((*ptr)++), shift;
+	uint32_t saved, dst;
+	int dptr = *ptr;
+	attr &= 0x38;
+	attr |= atom_def_dst[attr >> 3] << 6;
+	SDEBUG("   dst: ");
+	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
+	shift = atom_get_src_direct(ctx, ATOM_SRC_BYTE0, ptr);
+	SDEBUG("   shift: %d\n", shift);
+	dst <<= shift;
+	SDEBUG("   dst: ");
+	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
+}
+
+static void atom_op_shift_right(atom_exec_context *ctx, int *ptr, int arg)
+{
+	uint8_t attr = U8((*ptr)++), shift;
+	uint32_t saved, dst;
+	int dptr = *ptr;
+	attr &= 0x38;
+	attr |= atom_def_dst[attr >> 3] << 6;
+	SDEBUG("   dst: ");
+	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
+	shift = atom_get_src_direct(ctx, ATOM_SRC_BYTE0, ptr);
+	SDEBUG("   shift: %d\n", shift);
+	dst >>= shift;
+	SDEBUG("   dst: ");
+	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
+}
+
 static void atom_op_shl(atom_exec_context *ctx, int *ptr, int arg)
 {
 	uint8_t attr = U8((*ptr)++), shift;
@@ -818,7 +882,7 @@ static void atom_op_shl(atom_exec_context *ctx, int *ptr, int arg)
 	attr |= atom_def_dst[attr >> 3] << 6;
 	SDEBUG("   dst: ");
 	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
-	shift = U8((*ptr)++);
+	shift = atom_get_src(ctx, attr, ptr);
 	SDEBUG("   shift: %d\n", shift);
 	dst <<= shift;
 	SDEBUG("   dst: ");
@@ -834,7 +898,7 @@ static void atom_op_shr(atom_exec_context *ctx, int *ptr, int arg)
 	attr |= atom_def_dst[attr >> 3] << 6;
 	SDEBUG("   dst: ");
 	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
-	shift = U8((*ptr)++);
+	shift = atom_get_src(ctx, attr, ptr);
 	SDEBUG("   shift: %d\n", shift);
 	dst >>= shift;
 	SDEBUG("   dst: ");
@@ -937,18 +1001,18 @@ static struct {
 	atom_op_or, ATOM_ARG_FB}, {
 	atom_op_or, ATOM_ARG_PLL}, {
 	atom_op_or, ATOM_ARG_MC}, {
-	atom_op_shl, ATOM_ARG_REG}, {
-	atom_op_shl, ATOM_ARG_PS}, {
-	atom_op_shl, ATOM_ARG_WS}, {
-	atom_op_shl, ATOM_ARG_FB}, {
-	atom_op_shl, ATOM_ARG_PLL}, {
-	atom_op_shl, ATOM_ARG_MC}, {
-	atom_op_shr, ATOM_ARG_REG}, {
-	atom_op_shr, ATOM_ARG_PS}, {
-	atom_op_shr, ATOM_ARG_WS}, {
-	atom_op_shr, ATOM_ARG_FB}, {
-	atom_op_shr, ATOM_ARG_PLL}, {
-	atom_op_shr, ATOM_ARG_MC}, {
+	atom_op_shift_left, ATOM_ARG_REG}, {
+	atom_op_shift_left, ATOM_ARG_PS}, {
+	atom_op_shift_left, ATOM_ARG_WS}, {
+	atom_op_shift_left, ATOM_ARG_FB}, {
+	atom_op_shift_left, ATOM_ARG_PLL}, {
+	atom_op_shift_left, ATOM_ARG_MC}, {
+	atom_op_shift_right, ATOM_ARG_REG}, {
+	atom_op_shift_right, ATOM_ARG_PS}, {
+	atom_op_shift_right, ATOM_ARG_WS}, {
+	atom_op_shift_right, ATOM_ARG_FB}, {
+	atom_op_shift_right, ATOM_ARG_PLL}, {
+	atom_op_shift_right, ATOM_ARG_MC}, {
 	atom_op_mul, ATOM_ARG_REG}, {
 	atom_op_mul, ATOM_ARG_PS}, {
 	atom_op_mul, ATOM_ARG_WS}, {
@@ -1058,8 +1122,6 @@ static void atom_execute_table_locked(struct atom_context *ctx, int index, uint3
 
 	SDEBUG(">> execute %04X (len %d, WS %d, PS %d)\n", base, len, ws, ps);
 
-	/* reset reg block */
-	ctx->reg_block = 0;
 	ectx.ctx = ctx;
 	ectx.ps_shift = ps / 4;
 	ectx.start = base;
@@ -1096,6 +1158,12 @@ static void atom_execute_table_locked(struct atom_context *ctx, int index, uint3
 void atom_execute_table(struct atom_context *ctx, int index, uint32_t * params)
 {
 	mutex_lock(&ctx->mutex);
+	/* reset reg block */
+	ctx->reg_block = 0;
+	/* reset fb window */
+	ctx->fb_base = 0;
+	/* reset io mode */
+	ctx->io_mode = ATOM_IO_MM;
 	atom_execute_table_locked(ctx, index, params);
 	mutex_unlock(&ctx->mutex);
 }

drivers/gpu/drm/radeon/atom.h

@@ -91,6 +91,7 @@
 #define ATOM_WS_AND_MASK	0x45
 #define ATOM_WS_FB_WINDOW	0x46
 #define ATOM_WS_ATTRIBUTES	0x47
+#define ATOM_WS_REGPTR  	0x48
 
 #define ATOM_IIO_NOP		0
 #define ATOM_IIO_START		1

drivers/gpu/drm/radeon/r100.c

@@ -1504,6 +1504,7 @@ static int r100_packet3_check(struct radeon_cs_parser *p,
 			DRM_ERROR("PRIM_WALK must be 3 for IMMD draw\n");
 			return -EINVAL;
 		}
+		track->vtx_size = r100_get_vtx_size(radeon_get_ib_value(p, idx + 0));
 		track->vap_vf_cntl = radeon_get_ib_value(p, idx + 1);
 		track->immd_dwords = pkt->count - 1;
 		r = r100_cs_track_check(p->rdev, track);

drivers/gpu/drm/radeon/r200.c

@@ -371,13 +371,16 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 		case 5:
 		case 6:
 		case 7:
+			/* 1D/2D */
 			track->textures[i].tex_coord_type = 0;
 			break;
 		case 1:
-			track->textures[i].tex_coord_type = 1;
+			/* CUBE */
+			track->textures[i].tex_coord_type = 2;
 			break;
 		case 2:
-			track->textures[i].tex_coord_type = 2;
+			/* 3D */
+			track->textures[i].tex_coord_type = 1;
 			break;
 		}
 		break;

drivers/gpu/drm/radeon/r600.c

@@ -1954,6 +1954,7 @@ int r600_suspend(struct radeon_device *rdev)
 	/* FIXME: we should wait for ring to be empty */
 	r600_cp_stop(rdev);
 	rdev->cp.ready = false;
+	r600_irq_suspend(rdev);
 	r600_wb_disable(rdev);
 	r600_pcie_gart_disable(rdev);
 	/* unpin shaders bo */
@@ -2063,15 +2064,16 @@ int r600_init(struct radeon_device *rdev)
 	if (rdev->accel_working) {
 		r = radeon_ib_pool_init(rdev);
 		if (r) {
-			DRM_ERROR("radeon: failed initializing IB pool (%d).\n", r);
+			dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
 			rdev->accel_working = false;
-		}
+		} else {
 			r = r600_ib_test(rdev);
 			if (r) {
-			DRM_ERROR("radeon: failed testing IB (%d).\n", r);
+				dev_err(rdev->dev, "IB test failed (%d).\n", r);
 				rdev->accel_working = false;
 			}
 		}
+	}
 
 	r = r600_audio_init(rdev);
 	if (r)
@@ -2200,14 +2202,14 @@ void r600_ih_ring_init(struct radeon_device *rdev, unsigned ring_size)
 	rb_bufsz = drm_order(ring_size / 4);
 	ring_size = (1 << rb_bufsz) * 4;
 	rdev->ih.ring_size = ring_size;
-	rdev->ih.align_mask = 4 - 1;
+	rdev->ih.ptr_mask = rdev->ih.ring_size - 1;
+	rdev->ih.rptr = 0;
 }
 
-static int r600_ih_ring_alloc(struct radeon_device *rdev, unsigned ring_size)
+static int r600_ih_ring_alloc(struct radeon_device *rdev)
 {
 	int r;
 
-	rdev->ih.ring_size = ring_size;
 	/* Allocate ring buffer */
 	if (rdev->ih.ring_obj == NULL) {
 		r = radeon_bo_create(rdev, NULL, rdev->ih.ring_size,
@@ -2237,9 +2239,6 @@ static int r600_ih_ring_alloc(struct radeon_device *rdev, unsigned ring_size)
 			return r;
 		}
 	}
-	rdev->ih.ptr_mask = (rdev->cp.ring_size / 4) - 1;
-	rdev->ih.rptr = 0;
-
 	return 0;
 }
 
@@ -2389,7 +2388,7 @@ int r600_irq_init(struct radeon_device *rdev)
 	u32 interrupt_cntl, ih_cntl, ih_rb_cntl;
 
 	/* allocate ring */
-	ret = r600_ih_ring_alloc(rdev, rdev->ih.ring_size);
+	ret = r600_ih_ring_alloc(rdev);
 	if (ret)
 		return ret;
 
@@ -2452,10 +2451,15 @@ int r600_irq_init(struct radeon_device *rdev)
 	return ret;
 }
 
-void r600_irq_fini(struct radeon_device *rdev)
+void r600_irq_suspend(struct radeon_device *rdev)
 {
 	r600_disable_interrupts(rdev);
 	r600_rlc_stop(rdev);
+}
+
+void r600_irq_fini(struct radeon_device *rdev)
+{
+	r600_irq_suspend(rdev);
 	r600_ih_ring_fini(rdev);
 }
 
@@ -2470,8 +2474,12 @@ int r600_irq_set(struct radeon_device *rdev)
 		return -EINVAL;
 	}
 	/* don't enable anything if the ih is disabled */
-	if (!rdev->ih.enabled)
+	if (!rdev->ih.enabled) {
+		r600_disable_interrupts(rdev);
+		/* force the active interrupt state to all disabled */
+		r600_disable_interrupt_state(rdev);
 		return 0;
+	}
 
 	if (ASIC_IS_DCE3(rdev)) {
 		hpd1 = RREG32(DC_HPD1_INT_CONTROL) & ~DC_HPDx_INT_EN;
@@ -2641,16 +2649,18 @@ static inline u32 r600_get_ih_wptr(struct radeon_device *rdev)
 	wptr = RREG32(IH_RB_WPTR);
 
 	if (wptr & RB_OVERFLOW) {
-		WARN_ON(1);
-		/* XXX deal with overflow */
-		DRM_ERROR("IH RB overflow\n");
+		/* When a ring buffer overflow happen start parsing interrupt
+		 * from the last not overwritten vector (wptr + 16). Hopefully
+		 * this should allow us to catchup.
+		 */
+		dev_warn(rdev->dev, "IH ring buffer overflow (0x%08X, %d, %d)\n",
+			wptr, rdev->ih.rptr, (wptr + 16) + rdev->ih.ptr_mask);
+		rdev->ih.rptr = (wptr + 16) & rdev->ih.ptr_mask;
 		tmp = RREG32(IH_RB_CNTL);
 		tmp |= IH_WPTR_OVERFLOW_CLEAR;
 		WREG32(IH_RB_CNTL, tmp);
 	}
-	wptr = wptr & WPTR_OFFSET_MASK;
-
-	return wptr;
+	return (wptr & rdev->ih.ptr_mask);
 }
 
 /*        r600 IV Ring
@@ -2686,12 +2696,13 @@ int r600_irq_process(struct radeon_device *rdev)
 	u32 wptr = r600_get_ih_wptr(rdev);
 	u32 rptr = rdev->ih.rptr;
 	u32 src_id, src_data;
-	u32 last_entry = rdev->ih.ring_size - 16;
 	u32 ring_index, disp_int, disp_int_cont, disp_int_cont2;
 	unsigned long flags;
 	bool queue_hotplug = false;
 
 	DRM_DEBUG("r600_irq_process start: rptr %d, wptr %d\n", rptr, wptr);
+	if (!rdev->ih.enabled)
+		return IRQ_NONE;
 
 	spin_lock_irqsave(&rdev->ih.lock, flags);
 
@@ -2820,10 +2831,8 @@ restart_ih:
 		}
 
 		/* wptr/rptr are in bytes! */
-		if (rptr == last_entry)
-			rptr = 0;
-		else
 		rptr += 16;
+		rptr &= rdev->ih.ptr_mask;
 	}
 	/* make sure wptr hasn't changed while processing */
 	wptr = r600_get_ih_wptr(rdev);

drivers/gpu/drm/radeon/r600_cs.c

@@ -36,6 +36,10 @@ static int r600_cs_packet_next_reloc_nomm(struct radeon_cs_parser *p,
 typedef int (*next_reloc_t)(struct radeon_cs_parser*, struct radeon_cs_reloc**);
 static next_reloc_t r600_cs_packet_next_reloc = &r600_cs_packet_next_reloc_mm;
 
+struct r600_cs_track {
+	u32	cb_color0_base_last;
+};
+
 /**
  * r600_cs_packet_parse() - parse cp packet and point ib index to next packet
  * @parser:	parser structure holding parsing context.
@@ -176,6 +180,28 @@ static int r600_cs_packet_next_reloc_nomm(struct radeon_cs_parser *p,
 	return 0;
 }
 
+/**
+ * r600_cs_packet_next_is_pkt3_nop() - test if next packet is packet3 nop for reloc
+ * @parser:		parser structure holding parsing context.
+ *
+ * Check next packet is relocation packet3, do bo validation and compute
+ * GPU offset using the provided start.
+ **/
+static inline int r600_cs_packet_next_is_pkt3_nop(struct radeon_cs_parser *p)
+{
+	struct radeon_cs_packet p3reloc;
+	int r;
+
+	r = r600_cs_packet_parse(p, &p3reloc, p->idx);
+	if (r) {
+		return 0;
+	}
+	if (p3reloc.type != PACKET_TYPE3 || p3reloc.opcode != PACKET3_NOP) {
+		return 0;
+	}
+	return 1;
+}
+
 /**
  * r600_cs_packet_next_vline() - parse userspace VLINE packet
  * @parser:		parser structure holding parsing context.
@@ -337,6 +363,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 				struct radeon_cs_packet *pkt)
 {
 	struct radeon_cs_reloc *reloc;
+	struct r600_cs_track *track;
 	volatile u32 *ib;
 	unsigned idx;
 	unsigned i;
@@ -344,6 +371,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 	int r;
 	u32 idx_value;
 
+	track = (struct r600_cs_track *)p->track;
 	ib = p->ib->ptr;
 	idx = pkt->idx + 1;
 	idx_value = radeon_get_ib_value(p, idx);
@@ -503,9 +531,60 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 		for (i = 0; i < pkt->count; i++) {
 			reg = start_reg + (4 * i);
 			switch (reg) {
+			/* This register were added late, there is userspace
+			 * which does provide relocation for those but set
+			 * 0 offset. In order to avoid breaking old userspace
+			 * we detect this and set address to point to last
+			 * CB_COLOR0_BASE, note that if userspace doesn't set
+			 * CB_COLOR0_BASE before this register we will report
+			 * error. Old userspace always set CB_COLOR0_BASE
+			 * before any of this.
+			 */
+			case R_0280E0_CB_COLOR0_FRAG:
+			case R_0280E4_CB_COLOR1_FRAG:
+			case R_0280E8_CB_COLOR2_FRAG:
+			case R_0280EC_CB_COLOR3_FRAG:
+			case R_0280F0_CB_COLOR4_FRAG:
+			case R_0280F4_CB_COLOR5_FRAG:
+			case R_0280F8_CB_COLOR6_FRAG:
+			case R_0280FC_CB_COLOR7_FRAG:
+			case R_0280C0_CB_COLOR0_TILE:
+			case R_0280C4_CB_COLOR1_TILE:
+			case R_0280C8_CB_COLOR2_TILE:
+			case R_0280CC_CB_COLOR3_TILE:
+			case R_0280D0_CB_COLOR4_TILE:
+			case R_0280D4_CB_COLOR5_TILE:
+			case R_0280D8_CB_COLOR6_TILE:
+			case R_0280DC_CB_COLOR7_TILE:
+				if (!r600_cs_packet_next_is_pkt3_nop(p)) {
+					if (!track->cb_color0_base_last) {
+						dev_err(p->dev, "Broken old userspace ? no cb_color0_base supplied before trying to write 0x%08X\n", reg);
+						return -EINVAL;
+					}
+					ib[idx+1+i] = track->cb_color0_base_last;
+					printk_once(KERN_WARNING "You have old & broken userspace "
+						"please consider updating mesa & xf86-video-ati\n");
+				} else {
+					r = r600_cs_packet_next_reloc(p, &reloc);
+					if (r) {
+						dev_err(p->dev, "bad SET_CONTEXT_REG 0x%04X\n", reg);
+						return -EINVAL;
+					}
+					ib[idx+1+i] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+				}
+				break;
 			case DB_DEPTH_BASE:
 			case DB_HTILE_DATA_BASE:
 			case CB_COLOR0_BASE:
+				r = r600_cs_packet_next_reloc(p, &reloc);
+				if (r) {
+					DRM_ERROR("bad SET_CONTEXT_REG "
+							"0x%04X\n", reg);
+					return -EINVAL;
+				}
+				ib[idx+1+i] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+				track->cb_color0_base_last = ib[idx+1+i];
+				break;
 			case CB_COLOR1_BASE:
 			case CB_COLOR2_BASE:
 			case CB_COLOR3_BASE:
@@ -678,8 +757,11 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 int r600_cs_parse(struct radeon_cs_parser *p)
 {
 	struct radeon_cs_packet pkt;
+	struct r600_cs_track *track;
 	int r;
 
+	track = kzalloc(sizeof(*track), GFP_KERNEL);
+	p->track = track;
 	do {
 		r = r600_cs_packet_parse(p, &pkt, p->idx);
 		if (r) {
@@ -757,6 +839,7 @@ int r600_cs_legacy(struct drm_device *dev, void *data, struct drm_file *filp,
 	/* initialize parser */
 	memset(&parser, 0, sizeof(struct radeon_cs_parser));
 	parser.filp = filp;
+	parser.dev = &dev->pdev->dev;
 	parser.rdev = NULL;
 	parser.family = family;
 	parser.ib = &fake_ib;

drivers/gpu/drm/radeon/r600d.h

@@ -882,4 +882,29 @@
 #define		S_000E60_SOFT_RESET_VMC(x)		(((x) & 1) << 17)
 
 #define R_005480_HDP_MEM_COHERENCY_FLUSH_CNTL		0x5480
+
+#define R_0280E0_CB_COLOR0_FRAG                      0x0280E0
+#define   S_0280E0_BASE_256B(x)                        (((x) & 0xFFFFFFFF) << 0)
+#define   G_0280E0_BASE_256B(x)                        (((x) >> 0) & 0xFFFFFFFF)
+#define   C_0280E0_BASE_256B                           0x00000000
+#define R_0280E4_CB_COLOR1_FRAG                      0x0280E4
+#define R_0280E8_CB_COLOR2_FRAG                      0x0280E8
+#define R_0280EC_CB_COLOR3_FRAG                      0x0280EC
+#define R_0280F0_CB_COLOR4_FRAG                      0x0280F0
+#define R_0280F4_CB_COLOR5_FRAG                      0x0280F4
+#define R_0280F8_CB_COLOR6_FRAG                      0x0280F8
+#define R_0280FC_CB_COLOR7_FRAG                      0x0280FC
+#define R_0280C0_CB_COLOR0_TILE                      0x0280C0
+#define   S_0280C0_BASE_256B(x)                        (((x) & 0xFFFFFFFF) << 0)
+#define   G_0280C0_BASE_256B(x)                        (((x) >> 0) & 0xFFFFFFFF)
+#define   C_0280C0_BASE_256B                           0x00000000
+#define R_0280C4_CB_COLOR1_TILE                      0x0280C4
+#define R_0280C8_CB_COLOR2_TILE                      0x0280C8
+#define R_0280CC_CB_COLOR3_TILE                      0x0280CC
+#define R_0280D0_CB_COLOR4_TILE                      0x0280D0
+#define R_0280D4_CB_COLOR5_TILE                      0x0280D4
+#define R_0280D8_CB_COLOR6_TILE                      0x0280D8
+#define R_0280DC_CB_COLOR7_TILE                      0x0280DC
+
+
 #endif

drivers/gpu/drm/radeon/radeon.h

@@ -410,7 +410,6 @@ struct r600_ih {
 	unsigned		wptr_old;
 	unsigned		ring_size;
 	uint64_t		gpu_addr;
-	uint32_t		align_mask;
 	uint32_t		ptr_mask;
 	spinlock_t              lock;
 	bool                    enabled;
@@ -465,6 +464,7 @@ struct radeon_cs_chunk {
 };
 
 struct radeon_cs_parser {
+	struct device		*dev;
 	struct radeon_device	*rdev;
 	struct drm_file		*filp;
 	/* chunks */
@@ -847,7 +847,7 @@ void r600_kms_blit_copy(struct radeon_device *rdev,
 
 static inline uint32_t r100_mm_rreg(struct radeon_device *rdev, uint32_t reg)
 {
-	if (reg < 0x10000)
+	if (reg < rdev->rmmio_size)
 		return readl(((void __iomem *)rdev->rmmio) + reg);
 	else {
 		writel(reg, ((void __iomem *)rdev->rmmio) + RADEON_MM_INDEX);
@@ -857,7 +857,7 @@ static inline uint32_t r100_mm_rreg(struct radeon_device *rdev, uint32_t reg)
 
 static inline void r100_mm_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v)
 {
-	if (reg < 0x10000)
+	if (reg < rdev->rmmio_size)
 		writel(v, ((void __iomem *)rdev->rmmio) + reg);
 	else {
 		writel(reg, ((void __iomem *)rdev->rmmio) + RADEON_MM_INDEX);
@@ -1162,7 +1162,8 @@ extern int r600_irq_init(struct radeon_device *rdev);
 extern void r600_irq_fini(struct radeon_device *rdev);
 extern void r600_ih_ring_init(struct radeon_device *rdev, unsigned ring_size);
 extern int r600_irq_set(struct radeon_device *rdev);
-
+extern void r600_irq_suspend(struct radeon_device *rdev);
+/* r600 audio */
 extern int r600_audio_init(struct radeon_device *rdev);
 extern int r600_audio_tmds_index(struct drm_encoder *encoder);
 extern void r600_audio_set_clock(struct drm_encoder *encoder, int clock);

drivers/gpu/drm/radeon/radeon_clocks.c

@@ -56,7 +56,7 @@ uint32_t radeon_legacy_get_engine_clock(struct radeon_device *rdev)
 	else if (post_div == 3)
 		sclk >>= 2;
 	else if (post_div == 4)
-		sclk >>= 4;
+		sclk >>= 3;
 
 	return sclk;
 }
@@ -86,7 +86,7 @@ uint32_t radeon_legacy_get_memory_clock(struct radeon_device *rdev)
 	else if (post_div == 3)
 		mclk >>= 2;
 	else if (post_div == 4)
-		mclk >>= 4;
+		mclk >>= 3;
 
 	return mclk;
 }

drivers/gpu/drm/radeon/radeon_cs.c

@@ -231,6 +231,7 @@ int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 	memset(&parser, 0, sizeof(struct radeon_cs_parser));
 	parser.filp = filp;
 	parser.rdev = rdev;
+	parser.dev = rdev->dev;
 	r = radeon_cs_parser_init(&parser, data);
 	if (r) {
 		DRM_ERROR("Failed to initialize parser !\n");

drivers/gpu/drm/radeon/radeon_display.c

@@ -411,11 +411,12 @@ void radeon_compute_pll(struct radeon_pll *pll,
 			uint32_t *fb_div_p,
 			uint32_t *frac_fb_div_p,
 			uint32_t *ref_div_p,
-			uint32_t *post_div_p,
-			int flags)
+			uint32_t *post_div_p)
 {
 	uint32_t min_ref_div = pll->min_ref_div;
 	uint32_t max_ref_div = pll->max_ref_div;
+	uint32_t min_post_div = pll->min_post_div;
+	uint32_t max_post_div = pll->max_post_div;
 	uint32_t min_fractional_feed_div = 0;
 	uint32_t max_fractional_feed_div = 0;
 	uint32_t best_vco = pll->best_vco;
@@ -431,7 +432,7 @@ void radeon_compute_pll(struct radeon_pll *pll,
 	DRM_DEBUG("PLL freq %llu %u %u\n", freq, pll->min_ref_div, pll->max_ref_div);
 	freq = freq * 1000;
 
-	if (flags & RADEON_PLL_USE_REF_DIV)
+	if (pll->flags & RADEON_PLL_USE_REF_DIV)
 		min_ref_div = max_ref_div = pll->reference_div;
 	else {
 		while (min_ref_div < max_ref_div-1) {
@@ -446,19 +447,22 @@ void radeon_compute_pll(struct radeon_pll *pll,
 		}
 	}
 
-	if (flags & RADEON_PLL_USE_FRAC_FB_DIV) {
+	if (pll->flags & RADEON_PLL_USE_POST_DIV)
+		min_post_div = max_post_div = pll->post_div;
+
+	if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV) {
 		min_fractional_feed_div = pll->min_frac_feedback_div;
 		max_fractional_feed_div = pll->max_frac_feedback_div;
 	}
 
-	for (post_div = pll->min_post_div; post_div <= pll->max_post_div; ++post_div) {
+	for (post_div = min_post_div; post_div <= max_post_div; ++post_div) {
 		uint32_t ref_div;
 
-		if ((flags & RADEON_PLL_NO_ODD_POST_DIV) && (post_div & 1))
+		if ((pll->flags & RADEON_PLL_NO_ODD_POST_DIV) && (post_div & 1))
 			continue;
 
 		/* legacy radeons only have a few post_divs */
-		if (flags & RADEON_PLL_LEGACY) {
+		if (pll->flags & RADEON_PLL_LEGACY) {
 			if ((post_div == 5) ||
 			    (post_div == 7) ||
 			    (post_div == 9) ||
@@ -505,7 +509,7 @@ void radeon_compute_pll(struct radeon_pll *pll,
 					tmp += (uint64_t)pll->reference_freq * 1000 * frac_feedback_div;
 					current_freq = radeon_div(tmp, ref_div * post_div);
 
-					if (flags & RADEON_PLL_PREFER_CLOSEST_LOWER) {
+					if (pll->flags & RADEON_PLL_PREFER_CLOSEST_LOWER) {
 						error = freq - current_freq;
 						error = error < 0 ? 0xffffffff : error;
 					} else
@@ -532,12 +536,12 @@ void radeon_compute_pll(struct radeon_pll *pll,
 							best_freq = current_freq;
 							best_error = error;
 							best_vco_diff = vco_diff;
-						} else if (((flags & RADEON_PLL_PREFER_LOW_REF_DIV) && (ref_div < best_ref_div)) ||
-							   ((flags & RADEON_PLL_PREFER_HIGH_REF_DIV) && (ref_div > best_ref_div)) ||
-							   ((flags & RADEON_PLL_PREFER_LOW_FB_DIV) && (feedback_div < best_feedback_div)) ||
-							   ((flags & RADEON_PLL_PREFER_HIGH_FB_DIV) && (feedback_div > best_feedback_div)) ||
-							   ((flags & RADEON_PLL_PREFER_LOW_POST_DIV) && (post_div < best_post_div)) ||
-							   ((flags & RADEON_PLL_PREFER_HIGH_POST_DIV) && (post_div > best_post_div))) {
+						} else if (((pll->flags & RADEON_PLL_PREFER_LOW_REF_DIV) && (ref_div < best_ref_div)) ||
+							   ((pll->flags & RADEON_PLL_PREFER_HIGH_REF_DIV) && (ref_div > best_ref_div)) ||
+							   ((pll->flags & RADEON_PLL_PREFER_LOW_FB_DIV) && (feedback_div < best_feedback_div)) ||
+							   ((pll->flags & RADEON_PLL_PREFER_HIGH_FB_DIV) && (feedback_div > best_feedback_div)) ||
+							   ((pll->flags & RADEON_PLL_PREFER_LOW_POST_DIV) && (post_div < best_post_div)) ||
+							   ((pll->flags & RADEON_PLL_PREFER_HIGH_POST_DIV) && (post_div > best_post_div))) {
 							best_post_div = post_div;
 							best_ref_div = ref_div;
 							best_feedback_div = feedback_div;
@@ -573,8 +577,7 @@ void radeon_compute_pll_avivo(struct radeon_pll *pll,
 			      uint32_t *fb_div_p,
 			      uint32_t *frac_fb_div_p,
 			      uint32_t *ref_div_p,
-			      uint32_t *post_div_p,
-			      int flags)
+			      uint32_t *post_div_p)
 {
 	fixed20_12 m, n, frac_n, p, f_vco, f_pclk, best_freq;
 	fixed20_12 pll_out_max, pll_out_min;
@@ -715,7 +718,11 @@ radeon_user_framebuffer_create(struct drm_device *dev,
 	struct drm_gem_object *obj;
 
 	obj = drm_gem_object_lookup(dev, file_priv, mode_cmd->handle);
-
+	if (obj ==  NULL) {
+		dev_err(&dev->pdev->dev, "No GEM object associated to handle 0x%08X, "
+			"can't create framebuffer\n", mode_cmd->handle);
+		return NULL;
+	}
 	return radeon_framebuffer_create(dev, mode_cmd, obj);
 }
 

drivers/gpu/drm/radeon/radeon_legacy_crtc.c

@@ -339,69 +339,6 @@ void radeon_crtc_dpms(struct drm_crtc *crtc, int mode)
 	}
 }
 
-/* properly set crtc bpp when using atombios */
-void radeon_legacy_atom_set_surface(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	int format;
-	uint32_t crtc_gen_cntl;
-	uint32_t disp_merge_cntl;
-	uint32_t crtc_pitch;
-
-	switch (crtc->fb->bits_per_pixel) {
-	case 8:
-		format = 2;
-		break;
-	case 15:      /*  555 */
-		format = 3;
-		break;
-	case 16:      /*  565 */
-		format = 4;
-		break;
-	case 24:      /*  RGB */
-		format = 5;
-		break;
-	case 32:      /* xRGB */
-		format = 6;
-		break;
-	default:
-		return;
-	}
-
-	crtc_pitch  = ((((crtc->fb->pitch / (crtc->fb->bits_per_pixel / 8)) * crtc->fb->bits_per_pixel) +
-			((crtc->fb->bits_per_pixel * 8) - 1)) /
-		       (crtc->fb->bits_per_pixel * 8));
-	crtc_pitch |= crtc_pitch << 16;
-
-	WREG32(RADEON_CRTC_PITCH + radeon_crtc->crtc_offset, crtc_pitch);
-
-	switch (radeon_crtc->crtc_id) {
-	case 0:
-		disp_merge_cntl = RREG32(RADEON_DISP_MERGE_CNTL);
-		disp_merge_cntl &= ~RADEON_DISP_RGB_OFFSET_EN;
-		WREG32(RADEON_DISP_MERGE_CNTL, disp_merge_cntl);
-
-		crtc_gen_cntl = RREG32(RADEON_CRTC_GEN_CNTL) & 0xfffff0ff;
-		crtc_gen_cntl |= (format << 8);
-		crtc_gen_cntl |= RADEON_CRTC_EXT_DISP_EN;
-		WREG32(RADEON_CRTC_GEN_CNTL, crtc_gen_cntl);
-		break;
-	case 1:
-		disp_merge_cntl = RREG32(RADEON_DISP2_MERGE_CNTL);
-		disp_merge_cntl &= ~RADEON_DISP2_RGB_OFFSET_EN;
-		WREG32(RADEON_DISP2_MERGE_CNTL, disp_merge_cntl);
-
-		crtc_gen_cntl = RREG32(RADEON_CRTC2_GEN_CNTL) & 0xfffff0ff;
-		crtc_gen_cntl |= (format << 8);
-		WREG32(RADEON_CRTC2_GEN_CNTL, crtc_gen_cntl);
-		WREG32(RADEON_FP_H2_SYNC_STRT_WID,   RREG32(RADEON_CRTC2_H_SYNC_STRT_WID));
-		WREG32(RADEON_FP_V2_SYNC_STRT_WID,   RREG32(RADEON_CRTC2_V_SYNC_STRT_WID));
-		break;
-	}
-}
-
 int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 			 struct drm_framebuffer *old_fb)
 {
@@ -755,7 +692,6 @@ static void radeon_set_pll(struct drm_crtc *crtc, struct drm_display_mode *mode)
 	uint32_t post_divider = 0;
 	uint32_t freq = 0;
 	uint8_t pll_gain;
-	int pll_flags = RADEON_PLL_LEGACY;
 	bool use_bios_divs = false;
 	/* PLL registers */
 	uint32_t pll_ref_div = 0;
@@ -789,10 +725,12 @@ static void radeon_set_pll(struct drm_crtc *crtc, struct drm_display_mode *mode)
 	else
 		pll = &rdev->clock.p1pll;
 
+	pll->flags = RADEON_PLL_LEGACY;
+
 	if (mode->clock > 200000) /* range limits??? */
-		pll_flags |= RADEON_PLL_PREFER_HIGH_FB_DIV;
+		pll->flags |= RADEON_PLL_PREFER_HIGH_FB_DIV;
 	else
-		pll_flags |= RADEON_PLL_PREFER_LOW_REF_DIV;
+		pll->flags |= RADEON_PLL_PREFER_LOW_REF_DIV;
 
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 		if (encoder->crtc == crtc) {
@@ -804,7 +742,7 @@ static void radeon_set_pll(struct drm_crtc *crtc, struct drm_display_mode *mode)
 			}
 
 			if (encoder->encoder_type != DRM_MODE_ENCODER_DAC)
-				pll_flags |= RADEON_PLL_NO_ODD_POST_DIV;
+				pll->flags |= RADEON_PLL_NO_ODD_POST_DIV;
 			if (encoder->encoder_type == DRM_MODE_ENCODER_LVDS) {
 				if (!rdev->is_atom_bios) {
 					struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
@@ -819,7 +757,7 @@ static void radeon_set_pll(struct drm_crtc *crtc, struct drm_display_mode *mode)
 						}
 					}
 				}
-				pll_flags |= RADEON_PLL_USE_REF_DIV;
+				pll->flags |= RADEON_PLL_USE_REF_DIV;
 			}
 		}
 	}
@@ -829,8 +767,7 @@ static void radeon_set_pll(struct drm_crtc *crtc, struct drm_display_mode *mode)
 	if (!use_bios_divs) {
 		radeon_compute_pll(pll, mode->clock,
 				   &freq, &feedback_div, &frac_fb_div,
-				   &reference_div, &post_divider,
-				   pll_flags);
+				   &reference_div, &post_divider);
 
 		for (post_div = &post_divs[0]; post_div->divider; ++post_div) {
 			if (post_div->divider == post_divider)

drivers/gpu/drm/radeon/radeon_mode.h

@@ -125,16 +125,24 @@ struct radeon_tmds_pll {
 #define RADEON_PLL_PREFER_HIGH_POST_DIV (1 << 9)
 #define RADEON_PLL_USE_FRAC_FB_DIV      (1 << 10)
 #define RADEON_PLL_PREFER_CLOSEST_LOWER (1 << 11)
+#define RADEON_PLL_USE_POST_DIV         (1 << 12)
 
 struct radeon_pll {
-	uint16_t reference_freq;
-	uint16_t reference_div;
+	/* reference frequency */
+	uint32_t reference_freq;
+
+	/* fixed dividers */
+	uint32_t reference_div;
+	uint32_t post_div;
+
+	/* pll in/out limits */
 	uint32_t pll_in_min;
 	uint32_t pll_in_max;
 	uint32_t pll_out_min;
 	uint32_t pll_out_max;
-	uint16_t xclk;
+	uint32_t best_vco;
 
+	/* divider limits */
 	uint32_t min_ref_div;
 	uint32_t max_ref_div;
 	uint32_t min_post_div;
@@ -143,7 +151,12 @@ struct radeon_pll {
 	uint32_t max_feedback_div;
 	uint32_t min_frac_feedback_div;
 	uint32_t max_frac_feedback_div;
-	uint32_t best_vco;
+
+	/* flags for the current clock */
+	uint32_t flags;
+
+	/* pll id */
+	uint32_t id;
 };
 
 struct radeon_i2c_chan {
@@ -417,8 +430,7 @@ extern void radeon_compute_pll(struct radeon_pll *pll,
 			       uint32_t *fb_div_p,
 			       uint32_t *frac_fb_div_p,
 			       uint32_t *ref_div_p,
-			       uint32_t *post_div_p,
-			       int flags);
+			       uint32_t *post_div_p);
 
 extern void radeon_compute_pll_avivo(struct radeon_pll *pll,
 				     uint64_t freq,
@@ -426,8 +438,7 @@ extern void radeon_compute_pll_avivo(struct radeon_pll *pll,
 				     uint32_t *fb_div_p,
 				     uint32_t *frac_fb_div_p,
 				     uint32_t *ref_div_p,
-				     uint32_t *post_div_p,
-				     int flags);
+				     uint32_t *post_div_p);
 
 extern void radeon_setup_encoder_clones(struct drm_device *dev);
 
@@ -453,7 +464,6 @@ extern void atombios_crtc_dpms(struct drm_crtc *crtc, int mode);
 
 extern int radeon_crtc_set_base(struct drm_crtc *crtc, int x, int y,
 				 struct drm_framebuffer *old_fb);
-extern void radeon_legacy_atom_set_surface(struct drm_crtc *crtc);
 
 extern int radeon_crtc_cursor_set(struct drm_crtc *crtc,
 				  struct drm_file *file_priv,

drivers/gpu/drm/radeon/reg_srcs/r200

@@ -91,6 +91,8 @@ r200 0x3294
 0x22b8 SE_TCL_TEX_CYL_WRAP_CTL
 0x22c0 SE_TCL_UCP_VERT_BLEND_CNTL
 0x22c4 SE_TCL_POINT_SPRITE_CNTL
+0x22d0 SE_PVS_CNTL
+0x22d4 SE_PVS_CONST_CNTL
 0x2648 RE_POINTSIZE
 0x26c0 RE_TOP_LEFT
 0x26c4 RE_MISC

drivers/gpu/drm/radeon/rv770.c

@@ -968,6 +968,7 @@ int rv770_suspend(struct radeon_device *rdev)
 	/* FIXME: we should wait for ring to be empty */
 	r700_cp_stop(rdev);
 	rdev->cp.ready = false;
+	r600_irq_suspend(rdev);
 	r600_wb_disable(rdev);
 	rv770_pcie_gart_disable(rdev);
 	/* unpin shaders bo */
@@ -1074,15 +1075,16 @@ int rv770_init(struct radeon_device *rdev)
 	if (rdev->accel_working) {
 		r = radeon_ib_pool_init(rdev);
 		if (r) {
-			DRM_ERROR("radeon: failed initializing IB pool (%d).\n", r);
+			dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
 			rdev->accel_working = false;
-		}
+		} else {
 			r = r600_ib_test(rdev);
 			if (r) {
-			DRM_ERROR("radeon: failed testing IB (%d).\n", r);
+				dev_err(rdev->dev, "IB test failed (%d).\n", r);
 				rdev->accel_working = false;
 			}
 		}
+	}
 	return 0;
 }