drm/radeon: fix endian handling in rlc buffer setup

The buffers needs to be in little endian format.

Noticed-by: Sylvain BERTRAND <sylware@legeek.net>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index f48fe14..2ddb01d 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -5625,7 +5625,7 @@
 		}
 
 		for (i = 0; i < CP_ME_TABLE_SIZE; i ++) {
-			dst_ptr[bo_offset + i] = be32_to_cpu(fw_data[table_offset + i]);
+			dst_ptr[bo_offset + i] = cpu_to_le32(be32_to_cpu(fw_data[table_offset + i]));
 		}
 		bo_offset += CP_ME_TABLE_SIZE;
 	}
@@ -5847,52 +5847,53 @@
 	if (buffer == NULL)
 		return;
 
-	buffer[count++] = PACKET3(PACKET3_PREAMBLE_CNTL, 0);
-	buffer[count++] = PACKET3_PREAMBLE_BEGIN_CLEAR_STATE;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+	buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_BEGIN_CLEAR_STATE);
 
-	buffer[count++] = PACKET3(PACKET3_CONTEXT_CONTROL, 1);
-	buffer[count++] = 0x80000000;
-	buffer[count++] = 0x80000000;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CONTEXT_CONTROL, 1));
+	buffer[count++] = cpu_to_le32(0x80000000);
+	buffer[count++] = cpu_to_le32(0x80000000);
 
 	for (sect = rdev->rlc.cs_data; sect->section != NULL; ++sect) {
 		for (ext = sect->section; ext->extent != NULL; ++ext) {
 			if (sect->id == SECT_CONTEXT) {
-				buffer[count++] = PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count);
-				buffer[count++] = ext->reg_index - 0xa000;
+				buffer[count++] =
+					cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count));
+				buffer[count++] = cpu_to_le32(ext->reg_index - 0xa000);
 				for (i = 0; i < ext->reg_count; i++)
-					buffer[count++] = ext->extent[i];
+					buffer[count++] = cpu_to_le32(ext->extent[i]);
 			} else {
 				return;
 			}
 		}
 	}
 
-	buffer[count++] = PACKET3(PACKET3_SET_CONTEXT_REG, 2);
-	buffer[count++] = PA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, 2));
+	buffer[count++] = cpu_to_le32(PA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START);
 	switch (rdev->family) {
 	case CHIP_BONAIRE:
-		buffer[count++] = 0x16000012;
-		buffer[count++] = 0x00000000;
+		buffer[count++] = cpu_to_le32(0x16000012);
+		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	case CHIP_KAVERI:
-		buffer[count++] = 0x00000000; /* XXX */
-		buffer[count++] = 0x00000000;
+		buffer[count++] = cpu_to_le32(0x00000000); /* XXX */
+		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	case CHIP_KABINI:
-		buffer[count++] = 0x00000000; /* XXX */
-		buffer[count++] = 0x00000000;
+		buffer[count++] = cpu_to_le32(0x00000000); /* XXX */
+		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	default:
-		buffer[count++] = 0x00000000;
-		buffer[count++] = 0x00000000;
+		buffer[count++] = cpu_to_le32(0x00000000);
+		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	}
 
-	buffer[count++] = PACKET3(PACKET3_PREAMBLE_CNTL, 0);
-	buffer[count++] = PACKET3_PREAMBLE_END_CLEAR_STATE;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+	buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_END_CLEAR_STATE);
 
-	buffer[count++] = PACKET3(PACKET3_CLEAR_STATE, 0);
-	buffer[count++] = 0;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CLEAR_STATE, 0));
+	buffer[count++] = cpu_to_le32(0);
 }
 
 static void cik_init_pg(struct radeon_device *rdev)
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index 52fa3be..7a84d0c 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -4019,7 +4019,7 @@
 		if (rdev->family >= CHIP_TAHITI) {
 			/* SI */
 			for (i = 0; i < rdev->rlc.reg_list_size; i++)
-				dst_ptr[i] = src_ptr[i];
+				dst_ptr[i] = cpu_to_le32(src_ptr[i]);
 		} else {
 			/* ON/LN/TN */
 			/* format:
@@ -4033,10 +4033,10 @@
 				if (i < dws)
 					data |= (src_ptr[i] >> 2) << 16;
 				j = (((i - 1) * 3) / 2);
-				dst_ptr[j] = data;
+				dst_ptr[j] = cpu_to_le32(data);
 			}
 			j = ((i * 3) / 2);
-			dst_ptr[j] = RLC_SAVE_RESTORE_LIST_END_MARKER;
+			dst_ptr[j] = cpu_to_le32(RLC_SAVE_RESTORE_LIST_END_MARKER);
 		}
 		radeon_bo_kunmap(rdev->rlc.save_restore_obj);
 		radeon_bo_unreserve(rdev->rlc.save_restore_obj);
@@ -4098,40 +4098,40 @@
 			cik_get_csb_buffer(rdev, dst_ptr);
 		} else if (rdev->family >= CHIP_TAHITI) {
 			reg_list_mc_addr = rdev->rlc.clear_state_gpu_addr + 256;
-			dst_ptr[0] = upper_32_bits(reg_list_mc_addr);
-			dst_ptr[1] = lower_32_bits(reg_list_mc_addr);
-			dst_ptr[2] = rdev->rlc.clear_state_size;
+			dst_ptr[0] = cpu_to_le32(upper_32_bits(reg_list_mc_addr));
+			dst_ptr[1] = cpu_to_le32(lower_32_bits(reg_list_mc_addr));
+			dst_ptr[2] = cpu_to_le32(rdev->rlc.clear_state_size);
 			si_get_csb_buffer(rdev, &dst_ptr[(256/4)]);
 		} else {
 			reg_list_hdr_blk_index = 0;
 			reg_list_mc_addr = rdev->rlc.clear_state_gpu_addr + (reg_list_blk_index * 4);
 			data = upper_32_bits(reg_list_mc_addr);
-			dst_ptr[reg_list_hdr_blk_index] = data;
+			dst_ptr[reg_list_hdr_blk_index] = cpu_to_le32(data);
 			reg_list_hdr_blk_index++;
 			for (i = 0; cs_data[i].section != NULL; i++) {
 				for (j = 0; cs_data[i].section[j].extent != NULL; j++) {
 					reg_num = cs_data[i].section[j].reg_count;
 					data = reg_list_mc_addr & 0xffffffff;
-					dst_ptr[reg_list_hdr_blk_index] = data;
+					dst_ptr[reg_list_hdr_blk_index] = cpu_to_le32(data);
 					reg_list_hdr_blk_index++;
 
 					data = (cs_data[i].section[j].reg_index * 4) & 0xffffffff;
-					dst_ptr[reg_list_hdr_blk_index] = data;
+					dst_ptr[reg_list_hdr_blk_index] = cpu_to_le32(data);
 					reg_list_hdr_blk_index++;
 
 					data = 0x08000000 | (reg_num * 4);
-					dst_ptr[reg_list_hdr_blk_index] = data;
+					dst_ptr[reg_list_hdr_blk_index] = cpu_to_le32(data);
 					reg_list_hdr_blk_index++;
 
 					for (k = 0; k < reg_num; k++) {
 						data = cs_data[i].section[j].extent[k];
-						dst_ptr[reg_list_blk_index + k] = data;
+						dst_ptr[reg_list_blk_index + k] = cpu_to_le32(data);
 					}
 					reg_list_mc_addr += reg_num * 4;
 					reg_list_blk_index += reg_num;
 				}
 			}
-			dst_ptr[reg_list_hdr_blk_index] = RLC_CLEAR_STATE_END_MARKER;
+			dst_ptr[reg_list_hdr_blk_index] = cpu_to_le32(RLC_CLEAR_STATE_END_MARKER);
 		}
 		radeon_bo_kunmap(rdev->rlc.clear_state_obj);
 		radeon_bo_unreserve(rdev->rlc.clear_state_obj);
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index c354c10..b2071af 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -5361,52 +5361,53 @@
 	if (buffer == NULL)
 		return;
 
-	buffer[count++] = PACKET3(PACKET3_PREAMBLE_CNTL, 0);
-	buffer[count++] = PACKET3_PREAMBLE_BEGIN_CLEAR_STATE;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+	buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_BEGIN_CLEAR_STATE);
 
-	buffer[count++] = PACKET3(PACKET3_CONTEXT_CONTROL, 1);
-	buffer[count++] = 0x80000000;
-	buffer[count++] = 0x80000000;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CONTEXT_CONTROL, 1));
+	buffer[count++] = cpu_to_le32(0x80000000);
+	buffer[count++] = cpu_to_le32(0x80000000);
 
 	for (sect = rdev->rlc.cs_data; sect->section != NULL; ++sect) {
 		for (ext = sect->section; ext->extent != NULL; ++ext) {
 			if (sect->id == SECT_CONTEXT) {
-				buffer[count++] = PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count);
-				buffer[count++] = ext->reg_index - 0xa000;
+				buffer[count++] =
+					cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count));
+				buffer[count++] = cpu_to_le32(ext->reg_index - 0xa000);
 				for (i = 0; i < ext->reg_count; i++)
-					buffer[count++] = ext->extent[i];
+					buffer[count++] = cpu_to_le32(ext->extent[i]);
 			} else {
 				return;
 			}
 		}
 	}
 
-	buffer[count++] = PACKET3(PACKET3_SET_CONTEXT_REG, 1);
-	buffer[count++] = PA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, 1));
+	buffer[count++] = cpu_to_le32(PA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START);
 	switch (rdev->family) {
 	case CHIP_TAHITI:
 	case CHIP_PITCAIRN:
-		buffer[count++] = 0x2a00126a;
+		buffer[count++] = cpu_to_le32(0x2a00126a);
 		break;
 	case CHIP_VERDE:
-		buffer[count++] = 0x0000124a;
+		buffer[count++] = cpu_to_le32(0x0000124a);
 		break;
 	case CHIP_OLAND:
-		buffer[count++] = 0x00000082;
+		buffer[count++] = cpu_to_le32(0x00000082);
 		break;
 	case CHIP_HAINAN:
-		buffer[count++] = 0x00000000;
+		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	default:
-		buffer[count++] = 0x00000000;
+		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	}
 
-	buffer[count++] = PACKET3(PACKET3_PREAMBLE_CNTL, 0);
-	buffer[count++] = PACKET3_PREAMBLE_END_CLEAR_STATE;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+	buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_END_CLEAR_STATE);
 
-	buffer[count++] = PACKET3(PACKET3_CLEAR_STATE, 0);
-	buffer[count++] = 0;
+	buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CLEAR_STATE, 0));
+	buffer[count++] = cpu_to_le32(0);
 }
 
 static void si_init_pg(struct radeon_device *rdev)