Re: [PATCH] drm/amdgpu: Fix amdgpu_vm_alloc_pts failed

["Zhu, Rex" <Rex.Zhu@xxxxxxx>]

Hi Christian,

You can reproduce this issue by allocate CSA buffer in baremetal and map them to reserved VM address.

Please see the attached patch.

Best Regards

Rex

---

From: Christian König <ckoenig.leichtzumerken@xxxxxxxxx>
Sent: Tuesday, October 23, 2018 5:01 PM
To: Zhu, Rex; Zhang, Jerry; amd-gfx@xxxxxxxxxxxxxxxxxxxxx; Deucher, Alexander; Koenig, Christian
Subject: Re: [PATCH] drm/amdgpu: Fix amdgpu_vm_alloc_pts failed

 

Hi guys,

yeah the root PD doesn't necessarily have a power of two entries.

But what exactly was the problem with the original code? Why does 0xffffffff doesn't work?

The only possible explanation I can see is that somebody tried to use an address which is above max_pfn, or how did that trigger?

Thanks,
Christian.

Am 23.10.18 um 07:42 schrieb Zhu, Rex:

No, if the vm size is small, there may only on root pd entry.

we need to make sure the mask >= 0;

Maybe this change revert Christian's commit:

commit 72af632549b97ead9251bb155f08fefd1fb6f5c3
Author: Christian König <christian.koenig@xxxxxxx>
Date:   Sat Sep 15 10:02:13 2018 +0200

    drm/amdgpu: add amdgpu_vm_entries_mask v2
    
    We can't get the mask for the root directory from the number of entries.
    
    So add a new function to avoid that problem.

Best Regards

Rex

---

From: amd-gfx <amd-gfx-bounces@xxxxxxxxxxxxxxxxxxxxx> on behalf of Zhang, Jerry(Junwei) <Jerry.Zhang@xxxxxxx>
Sent: Tuesday, October 23, 2018 1:12 PM
To: Zhu, Rex; amd-gfx@xxxxxxxxxxxxxxxxxxxxx; Deucher, Alexander; Koenig, Christian
Subject: Re: [PATCH] drm/amdgpu: Fix amdgpu_vm_alloc_pts failed

 

On 10/23/2018 11:29 AM, Rex Zhu wrote:
> when the VA address located in the last PD entries,
> the alloc_pts will faile.
>
> Use the right PD mask instand of hardcode, suggested
> by jerry.zhang.
>
> Signed-off-by: Rex Zhu <Rex.Zhu@xxxxxxx>

Thanks to verify that.
Feel free to add
Reviewed-by: Junwei Zhang <Jerry.Zhang@xxxxxxx>

Also like to get to know some background about these two functions from
Christian.
Perhaps we may make it more simple, e.g. merging them together.

Regards,
Jerry

> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 5 ++++-
>   1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> index 054633b..3939013 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> @@ -202,8 +202,11 @@ static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev,
>   static uint32_t amdgpu_vm_entries_mask(struct amdgpu_device *adev,
>                                       unsigned int level)
>   {
> +     unsigned shift = amdgpu_vm_level_shift(adev,
> +                                            adev->vm_manager.root_level);
> +
>        if (level <= adev->vm_manager.root_level)
> -             return 0xffffffff;
> +             return (round_up(adev->vm_manager.max_pfn, 1 << shift) >> shift) - 1;
>        else if (level != AMDGPU_VM_PTB)
>                return 0x1ff;
>        else

_______________________________________________
amd-gfx mailing list
amd-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

amd-gfx Info Page - freedesktop.org

lists.freedesktop.org

To see the collection of prior postings to the list, visit the amd-gfx Archives.. Using amd-gfx: To post a message to all the list members, send email to amd-gfx@xxxxxxxxxxxxxxxxxxxxx. You can subscribe to the list, or change your existing subscription, in the sections below.

_______________________________________________
amd-gfx mailing list
amd-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

From a68c029e2bc5f60fd26d1b0705fee2be80b6f604 Mon Sep 17 00:00:00 2001
From: Rex Zhu <Rex.Zhu@xxxxxxx>
Date: Tue, 23 Oct 2018 17:07:12 +0800
Subject: [PATCH] debug: create and map csa buffer

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ++--
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c    | 8 ++++----
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 450b0b7..e0d46b9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1658,7 +1658,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 			adev->ip_blocks[i].status.hw = true;
 
 			/* right after GMC hw init, we create CSA */
-			if (amdgpu_sriov_vf(adev)) {
+//			if (amdgpu_sriov_vf(adev)) {
 				r = amdgpu_allocate_static_csa(adev, &adev->virt.csa_obj,
 								AMDGPU_GEM_DOMAIN_VRAM,
 								AMDGPU_CSA_SIZE);
@@ -1666,7 +1666,7 @@ static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 					DRM_ERROR("allocate CSA failed %d\n", r);
 					return r;
 				}
-			}
+//			}
 		}
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index f2ef9a1..b332262 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -977,13 +977,13 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 		goto error_vm;
 	}
 
-	if (amdgpu_sriov_vf(adev)) {
+//	if (amdgpu_sriov_vf(adev)) {
 		uint64_t csa_addr = amdgpu_csa_vaddr(adev, 1) & AMDGPU_GMC_HOLE_MASK;
 		r = amdgpu_map_static_csa(adev, &fpriv->vm, adev->virt.csa_obj,
 						&fpriv->csa_va, csa_addr, AMDGPU_CSA_SIZE);
 		if (r)
 			goto error_vm;
-	}
+//	}
 
 	mutex_init(&fpriv->bo_list_lock);
 	idr_init(&fpriv->bo_list_handles);
@@ -1043,13 +1043,13 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 
 	amdgpu_vm_bo_rmv(adev, fpriv->prt_va);
 
-	if (amdgpu_sriov_vf(adev)) {
+//	if (amdgpu_sriov_vf(adev)) {
 		/* TODO: how to handle reserve failure */
 		BUG_ON(amdgpu_bo_reserve(adev->virt.csa_obj, true));
 		amdgpu_vm_bo_rmv(adev, fpriv->csa_va);
 		fpriv->csa_va = NULL;
 		amdgpu_bo_unreserve(adev->virt.csa_obj);
-	}
+//	}
 
 	pasid = fpriv->vm.pasid;
 	pd = amdgpu_bo_ref(fpriv->vm.root.base.bo);
-- 
1.9.1

_______________________________________________
amd-gfx mailing list
amd-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/amd-gfx