drm/vc4: Add support for branching in shader validation.

We're already checking that branch instructions are between the start
of the shader and the proper PROG_END sequence.  The other thing we
need to make branching safe is to verify that the shader doesn't read
past the end of the uniforms stream.

To do that, we require that at any basic block reading uniforms have
the following instructions:

load_imm temp, <next offset within uniform stream>
add unif_addr, temp, unif

The instructions are generated by userspace, and the kernel verifies
that the load_imm is of the expected offset, and that the add adds it
to a uniform.  We track which uniform in the stream that is, and at
draw call time fix up the uniform stream to have the address of the
start of the shader's uniforms at that location.

Signed-off-by: Eric Anholt <[email protected]>

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

@@ -270,6 +270,9 @@ 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)
 

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;