diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 895776c421d4de..37fbef271db338 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -2237,6 +2237,13 @@ void radeon_agp_disable(struct radeon_device *rdev);
 int radeon_asic_init(struct radeon_device *rdev);
 
 
+/*
+ * memcpy_io and memset_io functions that work on a raspberry pi 4
+ */
+void memcpy_fromio_pcie(void *to, const volatile void __iomem *from, size_t count);
+void memcpy_toio_pcie(volatile void __iomem *to, const void *from, size_t count);
+void memset_io_pcie(volatile void __iomem *dst, int c, size_t count);
+
 /*
  * IOCTL.
  */
diff --git a/drivers/gpu/drm/radeon/radeon_bios.c b/drivers/gpu/drm/radeon/radeon_bios.c
index 33121655d50bbe..85cdec1f59ae73 100644
--- a/drivers/gpu/drm/radeon/radeon_bios.c
+++ b/drivers/gpu/drm/radeon/radeon_bios.c
@@ -72,7 +72,7 @@ static bool igp_read_bios_from_vram(struct radeon_device *rdev)
 		iounmap(bios);
 		return false;
 	}
-	memcpy_fromio(rdev->bios, bios, size);
+	memcpy_fromio_pcie(rdev->bios, bios, size);
 	iounmap(bios);
 	return true;
 }
@@ -101,7 +101,7 @@ static bool radeon_read_bios(struct radeon_device *rdev)
 		pci_unmap_rom(rdev->pdev, bios);
 		return false;
 	}
-	memcpy_fromio(rdev->bios, bios, size);
+	memcpy_fromio_pcie(rdev->bios, bios, size);
 	pci_unmap_rom(rdev->pdev, bios);
 	return true;
 }
@@ -125,7 +125,7 @@ static bool radeon_read_platform_bios(struct radeon_device *rdev)
 	if (!bios)
 		goto free_bios;
 
-	memcpy_fromio(rdev->bios, bios, romlen);
+	memcpy_fromio_pcie(rdev->bios, bios, romlen);
 	iounmap(bios);
 
 	if (rdev->bios[0] != 0x55 || rdev->bios[1] != 0xaa)
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 4f0fbf66743160..680415b98ad6e1 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -112,6 +112,98 @@ static const char radeon_family_name[][16] = {
 	"LAST",
 };
 
+/**
+ * DOC: memcpy_fromio_pcie
+ *
+ * like memcpy_fromio, but it only uses 8-bit and 32-bit wide accesses to work on a raspberry pi 4
+ */
+
+void memcpy_fromio_pcie(void *to, const volatile void __iomem *from, size_t count)
+{
+	while (count && !IS_ALIGNED((unsigned long)from, 8)) {
+		*(u8 *)to = __raw_readb(from);
+		from++;
+		to++;
+		count--;
+	}
+
+	while (count >= 4) {
+		*(u32 *)to = __raw_readl(from);
+		from += 4;
+		to += 4;
+		count -= 4;
+	}
+
+	while (count) {
+		*(u8 *)to = __raw_readb(from);
+		from++;
+		to++;
+		count--;
+	}
+}
+
+/**
+ * DOC: memcpy_toio_pcie
+ *
+ * like memcpy_toio, but it only uses 8-bit and 32-bit wide accesses to work on a raspberry pi 4
+ */
+
+void memcpy_toio_pcie(volatile void __iomem *to, const void *from, size_t count)
+{
+	while (count && !IS_ALIGNED((unsigned long)to, 8)) {
+		__raw_writeb(*(u8 *)from, to);
+		from++;
+		to++;
+		count--;
+	}
+
+	while (count >= 4) {
+		__raw_writel(*(u64 *)from, to);
+		from += 4;
+		to += 4;
+		count -= 4;
+	}
+
+	while (count) {
+		__raw_writeb(*(u8 *)from, to);
+		from++;
+		to++;
+		count--;
+	}
+}
+
+/**
+ * DOC: memset_io_pcie
+ *
+ * like memset_io, but it only uses 8-bit and 32-bit wide accesses to work on a raspberry pi 4
+ */
+
+void memset_io_pcie(volatile void __iomem *dst, int c, size_t count)
+{
+	u32 qc = (u8)c;
+
+	qc |= qc << 8;
+	qc |= qc << 16;
+
+	while (count && !IS_ALIGNED((unsigned long)dst, 8)) {
+		__raw_writeb(c, dst);
+		dst++;
+		count--;
+	}
+
+	while (count >= 4) {
+		__raw_writel(qc, dst);
+		dst += 4;
+		count -= 4;
+	}
+
+	while (count) {
+		__raw_writeb(c, dst);
+		dst++;
+		count--;
+	}
+}
+
 #if defined(CONFIG_VGA_SWITCHEROO)
 bool radeon_has_atpx_dgpu_power_cntl(void);
 bool radeon_is_atpx_hybrid(void);
@@ -490,7 +582,7 @@ int radeon_wb_init(struct radeon_device *rdev)
 	}
 
 	/* clear wb memory */
-	memset((char *)rdev->wb.wb, 0, RADEON_GPU_PAGE_SIZE);
+	memset_io_pcie((char *)rdev->wb.wb, 0, RADEON_GPU_PAGE_SIZE);
 	/* disable event_write fences */
 	rdev->wb.use_event = false;
 	/* disabled via module param */
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index ca382fbf7a86a3..7165c67ec45d1f 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -262,7 +262,7 @@ static int radeonfb_create(struct drm_fb_helper *helper,
 	/* setup helper */
 	rfbdev->helper.fb = fb;
 
-	memset_io(rbo->kptr, 0x0, radeon_bo_size(rbo));
+	memset_io_pcie(rbo->kptr, 0x0, radeon_bo_size(rbo));
 
 	info->fbops = &radeonfb_ops;
 
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index a36a4f2c76b097..7ec6a9140b12be 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -711,7 +711,7 @@ int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 	 * otherwise we will endup with broken userspace and we won't be able
 	 * to enable this feature without adding new interface
 	 */
-	invalid_flags = RADEON_VM_PAGE_VALID | RADEON_VM_PAGE_SYSTEM;
+	invalid_flags = RADEON_VM_PAGE_VALID | RADEON_VM_PAGE_SYSTEM | RADEON_VM_PAGE_SNOOPED;
 	if ((args->flags & invalid_flags)) {
 		dev_err(dev->dev, "invalid flags 0x%08X vs 0x%08X\n",
 			args->flags, invalid_flags);
diff --git a/drivers/gpu/drm/radeon/radeon_ib.c b/drivers/gpu/drm/radeon/radeon_ib.c
index 62b116727b4f50..342cbe4479a5ec 100644
--- a/drivers/gpu/drm/radeon/radeon_ib.c
+++ b/drivers/gpu/drm/radeon/radeon_ib.c
@@ -212,7 +212,7 @@ int radeon_ib_pool_init(struct radeon_device *rdev)
 		r = radeon_sa_bo_manager_init(rdev, &rdev->ring_tmp_bo,
 					      RADEON_IB_POOL_SIZE*64*1024,
 					      RADEON_GPU_PAGE_SIZE,
-					      RADEON_GEM_DOMAIN_GTT, 0);
+					      RADEON_GEM_DOMAIN_GTT, RADEON_GEM_GTT_UC);
 	}
 	if (r) {
 		return r;
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 56ede9d63b12c5..b6859869167556 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -109,29 +109,29 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 			rbo->placements[c].fpfn =
 				rbo->rdev->mc.visible_vram_size >> PAGE_SHIFT;
 			rbo->placements[c].mem_type = TTM_PL_VRAM;
-			rbo->placements[c++].flags = 0;
+			rbo->placements[c++].flags = RADEON_GEM_GTT_UC;
 		}
 
 		rbo->placements[c].fpfn = 0;
 		rbo->placements[c].mem_type = TTM_PL_VRAM;
-		rbo->placements[c++].flags = 0;
+		rbo->placements[c++].flags = RADEON_GEM_GTT_UC;
 	}
 
 	if (domain & RADEON_GEM_DOMAIN_GTT) {
 		rbo->placements[c].fpfn = 0;
 		rbo->placements[c].mem_type = TTM_PL_TT;
-		rbo->placements[c++].flags = 0;
+		rbo->placements[c++].flags = RADEON_GEM_GTT_UC;
 	}
 
 	if (domain & RADEON_GEM_DOMAIN_CPU) {
 		rbo->placements[c].fpfn = 0;
 		rbo->placements[c].mem_type = TTM_PL_SYSTEM;
-		rbo->placements[c++].flags = 0;
+		rbo->placements[c++].flags = RADEON_GEM_GTT_UC;
 	}
 	if (!c) {
 		rbo->placements[c].fpfn = 0;
 		rbo->placements[c].mem_type = TTM_PL_SYSTEM;
-		rbo->placements[c++].flags = 0;
+		rbo->placements[c++].flags = RADEON_GEM_GTT_UC;
 	}
 
 	rbo->placement.num_placement = c;
@@ -184,6 +184,7 @@ int radeon_bo_create(struct radeon_device *rdev,
 
 	bo->flags = flags;
 	/* PCI GART is always snooped */
+	/* Don't be so sure. TODO */
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC);
 
@@ -220,6 +221,10 @@ int radeon_bo_create(struct radeon_device *rdev,
 		bo->flags &= ~RADEON_GEM_GTT_WC;
 #endif
 
+	//Write combining may cause issues on the raspberry pi
+	bo->flags &= ~(RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC);
+	bo->flags |= RADEON_GEM_GTT_UC;
+
 	radeon_ttm_placement_from_domain(bo, domain);
 	/* Kernel allocation are uninterruptible */
 	down_read(&rdev->pm.mclk_lock);
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index 7e207276df374f..4eab7460a9ce62 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -390,7 +390,7 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig
 	/* Allocate ring buffer */
 	if (ring->ring_obj == NULL) {
 		r = radeon_bo_create(rdev, ring->ring_size, PAGE_SIZE, true,
-				     RADEON_GEM_DOMAIN_GTT, 0, NULL,
+				     RADEON_GEM_DOMAIN_GTT, RADEON_GEM_GTT_UC, NULL,
 				     NULL, &ring->ring_obj);
 		if (r) {
 			dev_err(rdev->dev, "(%d) ring create failed\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index a06d4cc2fb1c43..8d2266ea17b22c 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -295,7 +295,7 @@ static int radeon_ttm_io_mem_reserve(struct ttm_device *bdev, struct ttm_resourc
 			return -EINVAL;
 		mem->bus.offset += rdev->mc.aper_base;
 		mem->bus.is_iomem = true;
-		mem->bus.caching = ttm_write_combined;
+		mem->bus.caching = ttm_uncached;
 #ifdef __alpha__
 		/*
 		 * Alpha: use bus.addr to hold the ioremap() return,
@@ -456,7 +456,8 @@ static int radeon_ttm_backend_bind(struct ttm_device *bdev,
 		     ttm->num_pages, bo_mem, ttm);
 	}
 	if (ttm->caching == ttm_cached)
-		flags |= RADEON_GART_PAGE_SNOOP;
+		printk("TTM Page would've been snooped\n");
+		// flags |= RADEON_GART_PAGE_SNOOP;
 	r = radeon_gart_bind(rdev, gtt->offset, ttm->num_pages,
 			     ttm->pages, gtt->ttm.dma_address, flags);
 	if (r) {
@@ -515,13 +516,14 @@ static struct ttm_tt *radeon_ttm_tt_create(struct ttm_buffer_object *bo,
 		return NULL;
 	}
 
+	// TODO: Highly suspect.
 	if (rbo->flags & RADEON_GEM_GTT_UC)
 		caching = ttm_uncached;
 	else if (rbo->flags & RADEON_GEM_GTT_WC)
 		caching = ttm_write_combined;
 	else
 		caching = ttm_cached;
-
+	caching = ttm_uncached;
 	if (ttm_sg_tt_init(&gtt->ttm, bo, page_flags, caching)) {
 		kfree(gtt);
 		return NULL;
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index 2ea86919d95364..e3804be3680d8c 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -288,7 +288,7 @@ int radeon_uvd_resume(struct radeon_device *rdev)
 	if (rdev->uvd.vcpu_bo == NULL)
 		return -EINVAL;
 
-	memcpy_toio((void __iomem *)rdev->uvd.cpu_addr, rdev->uvd_fw->data, rdev->uvd_fw->size);
+	memcpy_toio_pcie((void __iomem *)rdev->uvd.cpu_addr, rdev->uvd_fw->data, rdev->uvd_fw->size);
 
 	size = radeon_bo_size(rdev->uvd.vcpu_bo);
 	size -= rdev->uvd_fw->size;
@@ -296,7 +296,7 @@ int radeon_uvd_resume(struct radeon_device *rdev)
 	ptr = rdev->uvd.cpu_addr;
 	ptr += rdev->uvd_fw->size;
 
-	memset_io((void __iomem *)ptr, 0, size);
+	memset_io_pcie((void __iomem *)ptr, 0, size);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
index bb53016f3138a2..a5065df377e7d6 100644
--- a/drivers/gpu/drm/radeon/radeon_vm.c
+++ b/drivers/gpu/drm/radeon/radeon_vm.c
@@ -952,7 +952,8 @@ int radeon_vm_bo_update(struct radeon_device *rdev,
 		if (mem->mem_type == TTM_PL_TT) {
 			bo_va->flags |= RADEON_VM_PAGE_SYSTEM;
 			if (!(bo_va->bo->flags & (RADEON_GEM_GTT_WC | RADEON_GEM_GTT_UC)))
-				bo_va->flags |= RADEON_VM_PAGE_SNOOPED;
+				printk("VM Page would've been set to snooped\n");
+	      //bo_va->flags |= RADEON_VM_PAGE_SNOOPED;
 
 		} else {
 			addr += rdev->vm_manager.vram_base_offset;
diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c
index b87dd551e93977..0b62d1f9dc307f 100644
--- a/drivers/gpu/drm/radeon/rs600.c
+++ b/drivers/gpu/drm/radeon/rs600.c
@@ -651,8 +651,8 @@ uint64_t rs600_gart_get_page_entry(uint64_t addr, uint32_t flags)
 		addr |= R600_PTE_READABLE;
 	if (flags & RADEON_GART_PAGE_WRITE)
 		addr |= R600_PTE_WRITEABLE;
-	if (flags & RADEON_GART_PAGE_SNOOP)
-		addr |= R600_PTE_SNOOPED;
+	// if (flags & RADEON_GART_PAGE_SNOOP) // no snooping around
+	// 	addr |= R600_PTE_SNOOPED;
 	return addr;
 }
 
@@ -660,7 +660,10 @@ 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));
+	uint32_t high = entry >> 32;
+	writel(entry,ptr+(i*8));
+	writel(high,ptr + (i*8) + 4);
+	//writeq(entry, ptr + (i * 8));
 }
 
 int rs600_irq_set(struct radeon_device *rdev)
diff --git a/drivers/video/fbdev/core/cfbcopyarea.c b/drivers/video/fbdev/core/cfbcopyarea.c
index 6d4bfeecee3508..3e2369ff697fa1 100644
--- a/drivers/video/fbdev/core/cfbcopyarea.c
+++ b/drivers/video/fbdev/core/cfbcopyarea.c
@@ -34,10 +34,22 @@
 #  define FB_WRITEL fb_writel
 #  define FB_READL  fb_readl
 #else
-#  define FB_WRITEL fb_writeq
-#  define FB_READL  fb_readq
+#  define FB_WRITEL fb_writel_writeq
+#  define FB_READL  fb_readl_readq
 #endif
 
+static void fb_writel_writeq(u64 val, volatile void __iomem *addr){
+  fb_writel(val,addr);
+  fb_writel(val >> 32, addr + 4);
+}
+
+static u64 fb_readl_readq(volatile void __iomem *addr){
+  u64 val;
+  val = fb_readl(addr);
+  val |= fb_readl(addr + 4) << 32;
+  return val;
+}
+
     /*
      *  Generic bitwise copy algorithm
      */
diff --git a/drivers/video/fbdev/core/cfbfillrect.c b/drivers/video/fbdev/core/cfbfillrect.c
index ba9f58b2a5e86e..8dda2665ec7f2e 100644
--- a/drivers/video/fbdev/core/cfbfillrect.c
+++ b/drivers/video/fbdev/core/cfbfillrect.c
@@ -23,10 +23,22 @@
 #  define FB_WRITEL fb_writel
 #  define FB_READL  fb_readl
 #else
-#  define FB_WRITEL fb_writeq
-#  define FB_READL  fb_readq
+#  define FB_WRITEL fb_writel_writeq
+#  define FB_READL  fb_readl_readq
 #endif
 
+static void fb_writel_writeq(u64 val, volatile void __iomem *addr){
+  fb_writel(val,addr);
+  fb_writel(val >> 32, addr + 4);
+}
+
+static u64 fb_readl_readq(volatile void __iomem *addr){
+  u64 val;
+  val = fb_readl(addr);
+  val |= fb_readl(addr + 4) << 32;
+  return val;
+}
+
     /*
      *  Aligned pattern fill using 32/64-bit memory accesses
      */