Skip to contentSkip to conversation

drm/amdgpu: force using sdma to update vm page table when mmio is blocked

PATCH In review Yang Wang 3 participants 7 messages Jan 19, 2022 – Jan 25, 2022

Subsystems (2)
lore
Share & tools
lore

Yang Wang

author
#0 lore raw

PATCH 1/1 drm/amdgpu: force using sdma to update vm page table when mmio is blocked

when mmio protection feature is enabled in hypervisor, it will cause guest OS can't R/W HDP regiters, and using cpu to update page table is not working well.

force using sdma to update page table when mmio is blocked.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +7 -0
index b23cb463b106..0f86f0b2e183 100644
@@ -2959,6 +2959,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 	vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
 				    AMDGPU_VM_USE_CPU_FOR_GFX);
 
+	if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+		vm->use_cpu_for_update = false;
+
 	DRM_DEBUG_DRIVER("VM update mode is %s\n",
 			 vm->use_cpu_for_update ? "CPU" : "SDMA");
 	WARN_ONCE((vm->use_cpu_for_update &&
@@ -3094,6 +3097,10 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 	/* Update VM state */
 	vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
 				    AMDGPU_VM_USE_CPU_FOR_COMPUTE);
+
+	if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+		vm->use_cpu_for_update = false;
+
 	DRM_DEBUG_DRIVER("VM update mode is %s\n",
 			 vm->use_cpu_for_update ? "CPU" : "SDMA");
 	WARN_ONCE((vm->use_cpu_for_update &&
Signature · 1 line
2.25.1
#1 lore raw

[AMD Official Use Only]

ping...

add @Liu, Monk<mailto:[email protected]> @Koenig, Christian<mailto:[email protected]> @Deucher, Alexander<mailto:[email protected]>

Best Regards,
Kevin
________________________________
From: Wang, Yang(Kevin) <[email protected]>
Sent: Wednesday, January 19, 2022 11:16 AM
To: [email protected] <[email protected]>
Cc: Min, Frank <[email protected]>; Chen, Horace <[email protected]>; Wang, Yang(Kevin) <[email protected]>
Subject: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked

when mmio protection feature is enabled in hypervisor, it will cause guest OS can't R/W HDP regiters, and using cpu to update page table is not working well.

force using sdma to update page table when mmio is blocked.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +7 -0
index b23cb463b106..0f86f0b2e183 100644
@@ -2959,6 +2959,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
                                     AMDGPU_VM_USE_CPU_FOR_GFX);
 
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n",
                          vm->use_cpu_for_update ? "CPU" : "SDMA");
         WARN_ONCE((vm->use_cpu_for_update &&
@@ -3094,6 +3097,10 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
         /* Update VM state */
         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
                                     AMDGPU_VM_USE_CPU_FOR_COMPUTE);
+
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n",
                          vm->use_cpu_for_update ? "CPU" : "SDMA");
         WARN_ONCE((vm->use_cpu_for_update &&

--
2.25.1

maintainer
#2 lore raw

Well NAK.

Even when we can't R/W HDP registers we need a way to invalidate the HDP or quite a bunch of functions won't work correctly.

Blocking CPU base page table updates only works around the symptoms, but doesn't really solve anything.

Regards,
Christian.

Am 20.01.22 um 09:46 schrieb Wang, Yang(Kevin):

Quoted 59 lines

[AMD Official Use Only]

ping...

add @Liu, Monk <mailto:[email protected]> @Koenig, Christian <mailto:[email protected]> @Deucher, Alexander <mailto:[email protected]>

Best Regards,
Kevin
------------------------------------------------------------------------
*From:* Wang, Yang(Kevin) <[email protected]>
*Sent:* Wednesday, January 19, 2022 11:16 AM
*To:* [email protected] <[email protected]>
*Cc:* Min, Frank <[email protected]>; Chen, Horace <[email protected]>; Wang, Yang(Kevin) <[email protected]> *Subject:* [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked
when mmio protection feature is enabled in hypervisor,
it will cause guest OS can't R/W HDP regiters,
and using cpu to update page table is not working well.

force using sdma to update page table when mmio is blocked.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 +++++++

 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
index b23cb463b106..0f86f0b2e183 100644
@@ -2959,6 +2959,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, 
 struct amdgpu_vm *vm)

         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_GFX);

+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update &&

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
@@ -3094,6 +3097,10 @@ int amdgpu_vm_make_compute(struct amdgpu_device 
 *adev, struct amdgpu_vm *vm)
          /* Update VM state */

         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_COMPUTE);
+
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update &&

Signature · 1 line
2.25.1
#3 lore raw

[AMD Official Use Only]

Hi Chris,

yes, I agree with your point.
and another way is that we can use KIQ to write HDP register to resolve HDP can't R/W issue. but it will cause some performance drop if use KIQ to programing register.

what is your ideas?

Best Regards,
Kevin
________________________________
From: Koenig, Christian <[email protected]>
Sent: Thursday, January 20, 2022 4:58 PM
To: Wang, Yang(Kevin) <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>; Liu, Monk <[email protected]>
Cc: Min, Frank <[email protected]>; Chen, Horace <[email protected]>
Subject: Re: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked

Well NAK.

Even when we can't R/W HDP registers we need a way to invalidate the HDP or quite a bunch of functions won't work correctly.

Blocking CPU base page table updates only works around the symptoms, but doesn't really solve anything.

Regards,
Christian.

Am 20.01.22 um 09:46 schrieb Wang, Yang(Kevin):

[AMD Official Use Only]

ping...

add @Liu, Monk<mailto:[email protected]> @Koenig, Christian<mailto:[email protected]> @Deucher, Alexander<mailto:[email protected]>

Best Regards,
Kevin
________________________________
From: Wang, Yang(Kevin) <[email protected]><mailto:[email protected]>
Sent: Wednesday, January 19, 2022 11:16 AM
To: [email protected]<mailto:[email protected]> <[email protected]><mailto:[email protected]>
Cc: Min, Frank <[email protected]><mailto:[email protected]>; Chen, Horace <[email protected]><mailto:[email protected]>; Wang, Yang(Kevin) <[email protected]><mailto:[email protected]>
Subject: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked

when mmio protection feature is enabled in hypervisor, it will cause guest OS can't R/W HDP regiters, and using cpu to update page table is not working well.

force using sdma to update page table when mmio is blocked.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +7 -0
index b23cb463b106..0f86f0b2e183 100644
@@ -2959,6 +2959,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
                                     AMDGPU_VM_USE_CPU_FOR_GFX);
 
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n",
                          vm->use_cpu_for_update ? "CPU" : "SDMA");
         WARN_ONCE((vm->use_cpu_for_update &&
@@ -3094,6 +3097,10 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
         /* Update VM state */
         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
                                     AMDGPU_VM_USE_CPU_FOR_COMPUTE);
+
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n",
                          vm->use_cpu_for_update ? "CPU" : "SDMA");
         WARN_ONCE((vm->use_cpu_for_update &&

--
2.25.1

maintainer
#4 lore raw

Hi Kevin,

well at least the HDP flush function has to work correctly or otherwise the driver won't work correctly.

If the registers are not accessible any more we need to find a proper workaround for this.

One possibility would be to use the KIQ, another is a dummy write/read to make sure the HDP is flushed (check the hardware docs).

The third option would be to question if blocking the HDP registers is really a good idea.

The solution is up to you, but a workaround like proposed below doesn't really help in any way.

Regards,
Christian.

Am 20.01.22 um 10:07 schrieb Wang, Yang(Kevin):

Quoted 90 lines

[AMD Official Use Only]

Hi Chris,

yes, I agree with your point.
and another way is that we can use KIQ to write HDP register to resolve HDP can't R/W issue. but it will cause some performance drop if use KIQ to programing register.

what is your ideas?

Best Regards,
Kevin
------------------------------------------------------------------------
*From:* Koenig, Christian <[email protected]>
*Sent:* Thursday, January 20, 2022 4:58 PM
*To:* Wang, Yang(Kevin) <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>; Liu, Monk <[email protected]> *Cc:* Min, Frank <[email protected]>; Chen, Horace <[email protected]>
*Subject:* Re: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked
Well NAK.

Even when we can't R/W HDP registers we need a way to invalidate the HDP or quite a bunch of functions won't work correctly.

Blocking CPU base page table updates only works around the symptoms, but doesn't really solve anything.

Regards,
Christian.

Am 20.01.22 um 09:46 schrieb Wang, Yang(Kevin):

Quoted 59 lines

[AMD Official Use Only]

ping...

add @Liu, Monk <mailto:[email protected]> @Koenig, Christian <mailto:[email protected]> @Deucher, Alexander <mailto:[email protected]>

Best Regards,
Kevin
------------------------------------------------------------------------
*From:* Wang, Yang(Kevin) <[email protected]> <mailto:[email protected]>
*Sent:* Wednesday, January 19, 2022 11:16 AM
*To:* [email protected] <mailto:[email protected]> <[email protected]> <mailto:[email protected]>
*Cc:* Min, Frank <[email protected]> <mailto:[email protected]>; Chen, Horace <[email protected]> <mailto:[email protected]>; Wang, Yang(Kevin) <[email protected]> <mailto:[email protected]> *Subject:* [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked
when mmio protection feature is enabled in hypervisor,
it will cause guest OS can't R/W HDP regiters,
and using cpu to update page table is not working well.

force using sdma to update page table when mmio is blocked.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 +++++++

 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
index b23cb463b106..0f86f0b2e183 100644
@@ -2959,6 +2959,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, 
 struct amdgpu_vm *vm)

         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_GFX);

+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update &&

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
@@ -3094,6 +3097,10 @@ int amdgpu_vm_make_compute(struct 
 amdgpu_device *adev, struct amdgpu_vm *vm)
          /* Update VM state */

         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_COMPUTE);
+
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update &&

Signature · 1 line
2.25.1
#5 lore raw

[AMD Official Use Only]

Thanks @Koenig, Christian<mailto:[email protected]>.

Best Regards,
Kevin
________________________________
From: Koenig, Christian <[email protected]>
Sent: Thursday, January 20, 2022 5:11 PM
To: Wang, Yang(Kevin) <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>; Liu, Monk <[email protected]>
Cc: Min, Frank <[email protected]>; Chen, Horace <[email protected]>
Subject: Re: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked

Hi Kevin,

well at least the HDP flush function has to work correctly or otherwise the driver won't work correctly.

If the registers are not accessible any more we need to find a proper workaround for this.

One possibility would be to use the KIQ, another is a dummy write/read to make sure the HDP is flushed (check the hardware docs).

The third option would be to question if blocking the HDP registers is really a good idea.

The solution is up to you, but a workaround like proposed below doesn't really help in any way.

Regards,
Christian.

Am 20.01.22 um 10:07 schrieb Wang, Yang(Kevin):

[AMD Official Use Only]

Hi Chris,

yes, I agree with your point.
and another way is that we can use KIQ to write HDP register to resolve HDP can't R/W issue. but it will cause some performance drop if use KIQ to programing register.

what is your ideas?

Best Regards,
Kevin
________________________________
From: Koenig, Christian <[email protected]><mailto:[email protected]>
Sent: Thursday, January 20, 2022 4:58 PM
To: Wang, Yang(Kevin) <[email protected]><mailto:[email protected]>; [email protected]<mailto:[email protected]> <[email protected]><mailto:[email protected]>; Deucher, Alexander <[email protected]><mailto:[email protected]>; Liu, Monk <[email protected]><mailto:[email protected]>
Cc: Min, Frank <[email protected]><mailto:[email protected]>; Chen, Horace <[email protected]><mailto:[email protected]>
Subject: Re: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked

Well NAK.

Even when we can't R/W HDP registers we need a way to invalidate the HDP or quite a bunch of functions won't work correctly.

Blocking CPU base page table updates only works around the symptoms, but doesn't really solve anything.

Regards,
Christian.

Am 20.01.22 um 09:46 schrieb Wang, Yang(Kevin):

[AMD Official Use Only]

ping...

add @Liu, Monk<mailto:[email protected]> @Koenig, Christian<mailto:[email protected]> @Deucher, Alexander<mailto:[email protected]>

Best Regards,
Kevin
________________________________
From: Wang, Yang(Kevin) <[email protected]><mailto:[email protected]>
Sent: Wednesday, January 19, 2022 11:16 AM
To: [email protected]<mailto:[email protected]> <[email protected]><mailto:[email protected]>
Cc: Min, Frank <[email protected]><mailto:[email protected]>; Chen, Horace <[email protected]><mailto:[email protected]>; Wang, Yang(Kevin) <[email protected]><mailto:[email protected]>
Subject: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked

when mmio protection feature is enabled in hypervisor, it will cause guest OS can't R/W HDP regiters, and using cpu to update page table is not working well.

force using sdma to update page table when mmio is blocked.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +7 -0
index b23cb463b106..0f86f0b2e183 100644
@@ -2959,6 +2959,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
                                     AMDGPU_VM_USE_CPU_FOR_GFX);
 
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n",
                          vm->use_cpu_for_update ? "CPU" : "SDMA");
         WARN_ONCE((vm->use_cpu_for_update &&
@@ -3094,6 +3097,10 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
         /* Update VM state */
         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode &
                                     AMDGPU_VM_USE_CPU_FOR_COMPUTE);
+
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n",
                          vm->use_cpu_for_update ? "CPU" : "SDMA");
         WARN_ONCE((vm->use_cpu_for_update &&

--
2.25.1

Tested-by Bokun Zhang Reviewed-by Lijo Lazar
#6 lore raw

I made a related fix for HDP flushing under SRIOV a while ago, and that was confirmed to be working at the time. What changed since then? Is this on different HW or with a different BIOS? Or are you missing this patch? For reference:

commit 6295c4064d8d18c4480636076dbd5afb48885c02

Author: Felix Kuehling <[email protected]>
Date:   Thu Nov 4 16:15:43 2021 -0400

drm/amdgpu: Fix MMIO HDP flush on SRIOV

Disable HDP register remapping on SRIOV and set rmmio_remap.reg_offset to the fixed address of the VF register for hdp_v*_flush_hdp.

Regards,
  Felix

Am 2022-01-20 um 04:39 schrieb Wang, Yang(Kevin):

Quoted 124 lines

[AMD Official Use Only]

[AMD Official Use Only]

Thanks @Koenig, Christian <mailto:[email protected]>.

Best Regards,
Kevin
------------------------------------------------------------------------
*From:* Koenig, Christian <[email protected]>
*Sent:* Thursday, January 20, 2022 5:11 PM
*To:* Wang, Yang(Kevin) <[email protected]>; [email protected] <[email protected]>; Deucher, Alexander <[email protected]>; Liu, Monk <[email protected]> *Cc:* Min, Frank <[email protected]>; Chen, Horace <[email protected]>
*Subject:* Re: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked
Hi Kevin,

well at least the HDP flush function has to work correctly or otherwise the driver won't work correctly.

If the registers are not accessible any more we need to find a proper workaround for this.

One possibility would be to use the KIQ, another is a dummy write/read to make sure the HDP is flushed (check the hardware docs).

The third option would be to question if blocking the HDP registers is really a good idea.

The solution is up to you, but a workaround like proposed below doesn't really help in any way.

Regards,
Christian.

Am 20.01.22 um 10:07 schrieb Wang, Yang(Kevin):

Quoted 90 lines

[AMD Official Use Only]

Hi Chris,

yes, I agree with your point.
and another way is that we can use KIQ to write HDP register to resolve HDP can't R/W issue. but it will cause some performance drop if use KIQ to programing register.

what is your ideas?

Best Regards,
Kevin
------------------------------------------------------------------------
*From:* Koenig, Christian <[email protected]> <mailto:[email protected]>
*Sent:* Thursday, January 20, 2022 4:58 PM
*To:* Wang, Yang(Kevin) <[email protected]> <mailto:[email protected]>; [email protected] <mailto:[email protected]> <[email protected]> <mailto:[email protected]>; Deucher, Alexander <[email protected]> <mailto:[email protected]>; Liu, Monk <[email protected]> <mailto:[email protected]> *Cc:* Min, Frank <[email protected]> <mailto:[email protected]>; Chen, Horace <[email protected]> <mailto:[email protected]>
*Subject:* Re: [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked
Well NAK.

Even when we can't R/W HDP registers we need a way to invalidate the HDP or quite a bunch of functions won't work correctly.

Blocking CPU base page table updates only works around the symptoms, but doesn't really solve anything.

Regards,
Christian.

Am 20.01.22 um 09:46 schrieb Wang, Yang(Kevin):

Quoted 59 lines

[AMD Official Use Only]

ping...

add @Liu, Monk <mailto:[email protected]> @Koenig, Christian <mailto:[email protected]> @Deucher, Alexander <mailto:[email protected]>

Best Regards,
Kevin
------------------------------------------------------------------------
*From:* Wang, Yang(Kevin) <[email protected]> <mailto:[email protected]>
*Sent:* Wednesday, January 19, 2022 11:16 AM
*To:* [email protected] <mailto:[email protected]> <[email protected]> <mailto:[email protected]>
*Cc:* Min, Frank <[email protected]> <mailto:[email protected]>; Chen, Horace <[email protected]> <mailto:[email protected]>; Wang, Yang(Kevin) <[email protected]> <mailto:[email protected]> *Subject:* [PATCH] drm/amdgpu: force using sdma to update vm page table when mmio is blocked
when mmio protection feature is enabled in hypervisor,
it will cause guest OS can't R/W HDP regiters,
and using cpu to update page table is not working well.

force using sdma to update page table when mmio is blocked.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 7 +++++++

 1 file changed, 7 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
index b23cb463b106..0f86f0b2e183 100644
@@ -2959,6 +2959,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, 
 struct amdgpu_vm *vm)

         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_GFX);

+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update &&

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +0 -0
@@ -3094,6 +3097,10 @@ int amdgpu_vm_make_compute(struct 
 amdgpu_device *adev, struct amdgpu_vm *vm)
          /* Update VM state */

         vm->use_cpu_for_update = !!(adev->vm_manager.vm_update_mode & AMDGPU_VM_USE_CPU_FOR_COMPUTE);
+
+       if (vm->use_cpu_for_update && amdgpu_sriov_vf(adev) && amdgpu_virt_mmio_blocked(adev))
+               vm->use_cpu_for_update = false;
+
         DRM_DEBUG_DRIVER("VM update mode is %s\n", vm->use_cpu_for_update ? "CPU" : "SDMA"); WARN_ONCE((vm->use_cpu_for_update &&

Signature · 1 line
2.25.1