Merge tag 'drm-vc4-next-2016-07-15' of https://github.com/anholt/linux into drm-next

This pull request brings in vc4 shader validation for branching,
allowing GLSL shaders with non-unrolled loops.

* tag 'drm-vc4-next-2016-07-15' of https://github.com/anholt/linux:
  drm/vc4: Fix a "the the" typo in a comment.
  drm/vc4: Fix definition of QPU_R_MS_REV_FLAGS
  drm/vc4: Add a getparam to signal support for branches.
  drm/vc4: Add support for branching in shader validation.
  drm/vc4: Add a bitmap of branch targets during shader validation.
  drm/vc4: Move validation's current/max ip into the validation struct.
  drm/vc4: Add a getparam ioctl for getting the V3D identity regs.

drivers/gpu/drm/vc4/vc4_drv.c

@@ -14,6 +14,7 @@
 #include <linux/module.h>
 #include <linux/of_platform.h>
 #include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
 #include "drm_fb_cma_helper.h"
 
 #include "uapi/drm/vc4_drm.h"
@@ -43,6 +44,49 @@ void __iomem *vc4_ioremap_regs(struct platform_device *dev, int index)
 	return map;
 }
 
+static int vc4_get_param_ioctl(struct drm_device *dev, void *data,
+			       struct drm_file *file_priv)
+{
+	struct vc4_dev *vc4 = to_vc4_dev(dev);
+	struct drm_vc4_get_param *args = data;
+	int ret;
+
+	if (args->pad != 0)
+		return -EINVAL;
+
+	switch (args->param) {
+	case DRM_VC4_PARAM_V3D_IDENT0:
+		ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
+		if (ret)
+			return ret;
+		args->value = V3D_READ(V3D_IDENT0);
+		pm_runtime_put(&vc4->v3d->pdev->dev);
+		break;
+	case DRM_VC4_PARAM_V3D_IDENT1:
+		ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
+		if (ret)
+			return ret;
+		args->value = V3D_READ(V3D_IDENT1);
+		pm_runtime_put(&vc4->v3d->pdev->dev);
+		break;
+	case DRM_VC4_PARAM_V3D_IDENT2:
+		ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
+		if (ret)
+			return ret;
+		args->value = V3D_READ(V3D_IDENT2);
+		pm_runtime_put(&vc4->v3d->pdev->dev);
+		break;
+	case DRM_VC4_PARAM_SUPPORTS_BRANCHES:
+		args->value = true;
+		break;
+	default:
+		DRM_DEBUG("Unknown parameter %d\n", args->param);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 static void vc4_lastclose(struct drm_device *dev)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
@@ -74,6 +118,7 @@ static const struct drm_ioctl_desc vc4_drm_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(VC4_CREATE_SHADER_BO, vc4_create_shader_bo_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(VC4_GET_HANG_STATE, vc4_get_hang_state_ioctl,
 			  DRM_ROOT_ONLY),
+	DRM_IOCTL_DEF_DRV(VC4_GET_PARAM, vc4_get_param_ioctl, DRM_RENDER_ALLOW),
 };
 
 static struct drm_driver vc4_drm_driver = {

drivers/gpu/drm/vc4/vc4_drv.h

@@ -355,6 +355,9 @@ struct vc4_validated_shader_info {
 	uint32_t uniforms_src_size;
 	uint32_t num_texture_samples;
 	struct vc4_texture_sample_info *texture_samples;
+
+	uint32_t num_uniform_addr_offsets;
+	uint32_t *uniform_addr_offsets;
 };
 
 /**

drivers/gpu/drm/vc4/vc4_qpu_defines.h

@@ -70,7 +70,7 @@ enum qpu_raddr {
 	QPU_R_ELEM_QPU = 38,
 	QPU_R_NOP,
 	QPU_R_XY_PIXEL_COORD = 41,
-	QPU_R_MS_REV_FLAGS = 41,
+	QPU_R_MS_REV_FLAGS = 42,
 	QPU_R_VPM = 48,
 	QPU_R_VPM_LD_BUSY,
 	QPU_R_VPM_LD_WAIT,
@@ -230,6 +230,15 @@ enum qpu_unpack_r4 {
 #define QPU_COND_MUL_SHIFT              46
 #define QPU_COND_MUL_MASK               QPU_MASK(48, 46)
 
+#define QPU_BRANCH_COND_SHIFT           52
+#define QPU_BRANCH_COND_MASK            QPU_MASK(55, 52)
+
+#define QPU_BRANCH_REL                  ((uint64_t)1 << 51)
+#define QPU_BRANCH_REG                  ((uint64_t)1 << 50)
+
+#define QPU_BRANCH_RADDR_A_SHIFT        45
+#define QPU_BRANCH_RADDR_A_MASK         QPU_MASK(49, 45)
+
 #define QPU_SF                          ((uint64_t)1 << 45)
 
 #define QPU_WADDR_ADD_SHIFT             38
@@ -261,4 +270,10 @@ enum qpu_unpack_r4 {
 #define QPU_OP_ADD_SHIFT                24
 #define QPU_OP_ADD_MASK                 QPU_MASK(28, 24)
 
+#define QPU_LOAD_IMM_SHIFT              0
+#define QPU_LOAD_IMM_MASK               QPU_MASK(31, 0)
+
+#define QPU_BRANCH_TARGET_SHIFT         0
+#define QPU_BRANCH_TARGET_MASK          QPU_MASK(31, 0)
+
 #endif /* VC4_QPU_DEFINES_H */

drivers/gpu/drm/vc4/vc4_validate.c

@@ -802,7 +802,7 @@ validate_gl_shader_rec(struct drm_device *dev,
 		uint32_t src_offset = *(uint32_t *)(pkt_u + o);
 		uint32_t *texture_handles_u;
 		void *uniform_data_u;
-		uint32_t tex;
+		uint32_t tex, uni;
 
 		*(uint32_t *)(pkt_v + o) = bo[i]->paddr + src_offset;
 
@@ -840,6 +840,17 @@ validate_gl_shader_rec(struct drm_device *dev,
 			}
 		}
 
+		/* Fill in the uniform slots that need this shader's
+		 * start-of-uniforms address (used for resetting the uniform
+		 * stream in the presence of control flow).
+		 */
+		for (uni = 0;
+		     uni < validated_shader->num_uniform_addr_offsets;
+		     uni++) {
+			uint32_t o = validated_shader->uniform_addr_offsets[uni];
+			((uint32_t *)exec->uniforms_v)[o] = exec->uniforms_p;
+		}
+
 		*(uint32_t *)(pkt_v + o + 4) = exec->uniforms_p;
 
 		exec->uniforms_u += validated_shader->uniforms_src_size;

include/uapi/drm/vc4_drm.h

@@ -37,6 +37,7 @@ extern "C" {
 #define DRM_VC4_MMAP_BO                           0x04
 #define DRM_VC4_CREATE_SHADER_BO                  0x05
 #define DRM_VC4_GET_HANG_STATE                    0x06
+#define DRM_VC4_GET_PARAM                         0x07
 
 #define DRM_IOCTL_VC4_SUBMIT_CL           DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_SUBMIT_CL, struct drm_vc4_submit_cl)
 #define DRM_IOCTL_VC4_WAIT_SEQNO          DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_WAIT_SEQNO, struct drm_vc4_wait_seqno)
@@ -45,6 +46,7 @@ extern "C" {
 #define DRM_IOCTL_VC4_MMAP_BO             DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_MMAP_BO, struct drm_vc4_mmap_bo)
 #define DRM_IOCTL_VC4_CREATE_SHADER_BO    DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_CREATE_SHADER_BO, struct drm_vc4_create_shader_bo)
 #define DRM_IOCTL_VC4_GET_HANG_STATE      DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_GET_HANG_STATE, struct drm_vc4_get_hang_state)
+#define DRM_IOCTL_VC4_GET_PARAM           DRM_IOWR(DRM_COMMAND_BASE + DRM_VC4_GET_PARAM, struct drm_vc4_get_param)
 
 struct drm_vc4_submit_rcl_surface {
 	__u32 hindex; /* Handle index, or ~0 if not present. */
@@ -280,6 +282,17 @@ struct drm_vc4_get_hang_state {
 	__u32 pad[16];
 };
 
+#define DRM_VC4_PARAM_V3D_IDENT0		0
+#define DRM_VC4_PARAM_V3D_IDENT1		1
+#define DRM_VC4_PARAM_V3D_IDENT2		2
+#define DRM_VC4_PARAM_SUPPORTS_BRANCHES		3
+
+struct drm_vc4_get_param {
+	__u32 param;
+	__u32 pad;
+	__u64 value;
+};
+
 #if defined(__cplusplus)
 }
 #endif