From: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Date: Tue, 15 Oct 2019 12:52:02 +0000 (+0200)
Subject: radv: dump trace files earlier if a GPU hang is detected
X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=cf5e55558efab33792c3312ffdc78b4c19bb4ff2;p=mesa.git

radv: dump trace files earlier if a GPU hang is detected

To make sure a trace file is generated in case the driver crashes
during the hang report generation (which happens sometimes).

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
---

diff --git a/src/amd/vulkan/radv_debug.c b/src/amd/vulkan/radv_debug.c
index 9ab7c8638e9..ce480794560 100644
--- a/src/amd/vulkan/radv_debug.c
+++ b/src/amd/vulkan/radv_debug.c
@@ -604,6 +604,8 @@ radv_check_gpu_hangs(struct radv_queue *queue, struct radeon_cmdbuf *cs)
 	graphics_pipeline = radv_get_saved_graphics_pipeline(device);
 	compute_pipeline = radv_get_saved_compute_pipeline(device);
 
+	radv_dump_trace(queue->device, cs);
+
 	fprintf(stderr, "GPU hang report:\n\n");
 	radv_dump_device_name(device, stderr);
 
@@ -632,7 +634,6 @@ radv_check_gpu_hangs(struct radv_queue *queue, struct radeon_cmdbuf *cs)
 		break;
 	}
 
-	radv_dump_trace(queue->device, cs);
 	abort();
 }