Skip to contentSkip to conversation

Patch "drm/vc4: Fix termination of the initial scan for branch targets." has been added to the 4.9-stable tree

PATCH [email protected] 1 participant 1 message Mar 22, 2017

Subsystems (4)
lore
Share & tools
lore

[email protected]

author maintainer Applied
#0 lore raw

This is a note to let you know that I've just added the patch titled

drm/vc4: Fix termination of the initial scan for branch targets.

to the 4.9-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
drm-vc4-fix-termination-of-the-initial-scan-for-branch-targets.patch
and it can be found in the queue-4.9 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree, please let <[email protected]> know about it.

Quoted 1 line

From 457e67a728696c4f8e6423c64e93def50530db9a Mon Sep 17 00:00:00 2001

From: Eric Anholt <[email protected]>
Date: Thu, 20 Oct 2016 16:48:12 -0700
Subject: drm/vc4: Fix termination of the initial scan for branch targets.

From: Eric Anholt <[email protected]>

commit 457e67a728696c4f8e6423c64e93def50530db9a upstream.

The loop is scanning until the original max_ip (size of the BO), but we want to not examine any code after the PROG_END's delay slots. There was a block trying to do that, except that we had some early continue statements if the signal wasn't a PROG_END or a BRANCH.

The failure mode would be that a valid shader is rejected because some undefined memory after the PROG_END slots is parsed as a branch and the rest of its setup is illegal. I haven't seen this in the wild, but valgrind was complaining when about this up in the userland simulator mode.

---
 drivers/gpu/drm/vc4/vc4_validate_shaders.c |   19 ++++++++-----------
 1 file changed, 8 insertions(+), 11 deletions(-)
drivers/gpu/drm/vc4/vc4_validate_shaders.c +8 -11
@@ -608,9 +608,7 @@ static bool
 vc4_validate_branches(struct vc4_shader_validation_state *validation_state)
 {
 	uint32_t max_branch_target = 0;
-	bool found_shader_end = false;
 	int ip;
-	int shader_end_ip = 0;
 	int last_branch = -2;
 
 	for (ip = 0; ip < validation_state->max_ip; ip++) {
@@ -621,8 +619,13 @@ vc4_validate_branches(struct vc4_shader_
 		uint32_t branch_target_ip;
 
 		if (sig == QPU_SIG_PROG_END) {
-			shader_end_ip = ip;
-			found_shader_end = true;
+			/* There are two delay slots after program end is
+			 * signaled that are still executed, then we're
+			 * finished.  validation_state->max_ip is the
+			 * instruction after the last valid instruction in the
+			 * program.
+			 */
+			validation_state->max_ip = ip + 3;
 			continue;
 		}
 
@@ -676,15 +679,9 @@ vc4_validate_branches(struct vc4_shader_
 		}
 		set_bit(after_delay_ip, validation_state->branch_targets);
 		max_branch_target = max(max_branch_target, after_delay_ip);
-
-		/* There are two delay slots after program end is signaled
-		 * that are still executed, then we're finished.
-		 */
-		if (found_shader_end && ip == shader_end_ip + 2)
-			break;
 	}
 
-	if (max_branch_target > shader_end_ip) {
+	if (max_branch_target > validation_state->max_ip - 3) {
 		DRM_ERROR("Branch landed after QPU_SIG_PROG_END");
 		return false;
 	}

Patches currently in stable-queue which might be from [email protected] are

queue-4.9/drm-vc4-fix-termination-of-the-initial-scan-for-branch-targets.patch queue-4.9/drm-vc4-use-runtime-autosuspend-to-avoid-thrashing-v3d-power-state.patch