From: jsg <jsg@openbsd.org>
Date: Tue, 13 Jun 2023 03:21:05 +0000 (+0000)
Subject: drm/amdgpu: Fix vram recover doesn't work after whole GPU reset (v2)
X-Git-Url: http://artulab.com/gitweb/?a=commitdiff_plain;h=5f70b624d3c0a44f5703418108f6eff5b9908f89;p=openbsd

drm/amdgpu: Fix vram recover doesn't work after whole GPU reset (v2)

From Lin.Cao
7a8248317b322d3cb56c64332062bae91460978a in linux-6.1.y/6.1.29
6c032c37ac3ef3b7df30937c785ecc4da428edc0 in mainline linux
---

diff --git a/sys/dev/pci/drm/amd/amdgpu/amdgpu_device.c b/sys/dev/pci/drm/amd/amdgpu/amdgpu_device.c
index a15539e21ff..9693660acf5 100644
--- a/sys/dev/pci/drm/amd/amdgpu/amdgpu_device.c
+++ b/sys/dev/pci/drm/amd/amdgpu/amdgpu_device.c
@@ -4596,7 +4596,11 @@ static int amdgpu_device_recover_vram(struct amdgpu_device *adev)
 	dev_info(adev->dev, "recover vram bo from shadow start\n");
 	mutex_lock(&adev->shadow_list_lock);
 	list_for_each_entry(vmbo, &adev->shadow_list, shadow_list) {
-		shadow = &vmbo->bo;
+		/* If vm is compute context or adev is APU, shadow will be NULL */
+		if (!vmbo->shadow)
+			continue;
+		shadow = vmbo->shadow;
+
 		/* No need to recover an evicted BO */
 		if (shadow->tbo.resource->mem_type != TTM_PL_TT ||
 		    shadow->tbo.resource->start == AMDGPU_BO_INVALID_OFFSET ||