Skip to contentSkip to conversation

drm/amd/display: Initialize num_pkrs on VANGOGH.

PATCH Applied Bas Nieuwenhuizen 4 participants 6 messages Oct 20, 2020 – Oct 21, 2020

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 drm/amd/display: Initialize num_pkrs on VANGOGH.

As far a I can tell uses a variant of DCN3xx which uses num_pkrs.

If we do not initialize the variable we will set the register field to ilog2(0) = -1, though the mask will reduce that to 7. Pretty sure 7 is not the value we want here.

---

Found while rebasing my modifiers series. Not tested on HW.

 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +2 -1
index 6855aad7f312..2713caac4f2a 100644
@@ -4074,7 +4074,8 @@ fill_plane_buffer_attributes(struct amdgpu_device *adev,
 #ifdef CONFIG_DRM_AMD_DC_DCN3_0
 		if (adev->asic_type == CHIP_SIENNA_CICHLID ||
 		    adev->asic_type == CHIP_NAVY_FLOUNDER ||
-		    adev->asic_type == CHIP_DIMGREY_CAVEFISH)
+		    adev->asic_type == CHIP_DIMGREY_CAVEFISH ||
+		    adev->asic_type == CHIP_VANGOGH)
 			tiling_info->gfx9.num_pkrs = adev->gfx.config.gb_addr_config_fields.num_pkrs;
 #endif
 		ret = fill_plane_dcc_attributes(adev, afb, format, rotation,
Signature · 6 lines
2.28.0

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Reviewed-by on v1 1/1
#1 lore raw

FYI, this fixes tiling on VanGogh.

Marek

Quoted 35 lines · Bas Nieuwenhuizen

As far a I can tell uses a variant of DCN3xx which uses num_pkrs.

If we do not initialize the variable we will set the register field to ilog2(0) = -1, though the mask will reduce that to 7. Pretty sure 7 is not the value we want here.

---

Found while rebasing my modifiers series. Not tested on HW.

 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +2 -1
index 6855aad7f312..2713caac4f2a 100644
@@ -4074,7 +4074,8 @@ fill_plane_buffer_attributes(struct amdgpu_device
 *adev,
 #ifdef CONFIG_DRM_AMD_DC_DCN3_0
                if (adev->asic_type == CHIP_SIENNA_CICHLID ||
                    adev->asic_type == CHIP_NAVY_FLOUNDER ||
-                   adev->asic_type == CHIP_DIMGREY_CAVEFISH)
+                   adev->asic_type == CHIP_DIMGREY_CAVEFISH ||
+                   adev->asic_type == CHIP_VANGOGH)
                        tiling_info->gfx9.num_pkrs =
 adev->gfx.config.gb_addr_config_fields.num_pkrs;

#endif
ret = fill_plane_dcc_attributes(adev, afb, format,
rotation,
--
2.28.0

Quoted 5 lines

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

Applied
#2 lore raw
Quoted 19 lines · Marek Olšák

FYI, this fixes tiling on VanGogh.

Marek

Quoted 10 lines · Bas Nieuwenhuizen

As far a I can tell uses a variant of DCN3xx which uses num_pkrs.

If we do not initialize the variable we will set the register field to ilog2(0) = -1, though the mask will reduce that to 7. Pretty sure 7 is not the value we want here.

---

Found while rebasing my modifiers series. Not tested on HW.

Yes! Thank you so much, Bas!

I just verified the tiling issue got resovled with your patch! :-)

Applied.

Reviewed-and-Tested-by Huang Rui <[email protected]>
Quoted 25 lines
Quoted 25 lines
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +2 -1
index 6855aad7f312..2713caac4f2a 100644
@@ -4074,7 +4074,8 @@ fill_plane_buffer_attributes(struct amdgpu_device
 *adev,
 #ifdef CONFIG_DRM_AMD_DC_DCN3_0
                if (adev->asic_type == CHIP_SIENNA_CICHLID ||
                    adev->asic_type == CHIP_NAVY_FLOUNDER ||
-                   adev->asic_type == CHIP_DIMGREY_CAVEFISH)
+                   adev->asic_type == CHIP_DIMGREY_CAVEFISH ||
+                   adev->asic_type == CHIP_VANGOGH)
                        tiling_info->gfx9.num_pkrs =
 adev->gfx.config.gb_addr_config_fields.num_pkrs;

#endif
ret = fill_plane_dcc_attributes(adev, afb, format,
rotation,
--
2.28.0

Quoted 10 lines
Quoted 5 lines

_______________________________________________
amd-gfx mailing list
[email protected]
https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.freedesktop.org%2Fmailman%2Flistinfo%2Famd-gfx&amp;data=04%7C01%7Cray.huang%40amd.com%7C9afdb4e0a2e949655c9b08d87552a1bf%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637388345070265382%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=SbojOTLaOXt85OSbYjyRPozLAsqBwXpvoYVBu7c0Ido%3D&amp;reserved=0

_______________________________________________
amd-gfx mailing list
[email protected]
https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.freedesktop.org%2Fmailman%2Flistinfo%2Famd-gfx&amp;data=04%7C01%7Cray.huang%40amd.com%7C9afdb4e0a2e949655c9b08d87552a1bf%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C637388345070305360%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=uPui9S6bxwp%2B4J4uav4FekR1zOZ9VeHYcdrOSA9ZFKg%3D&amp;reserved=0

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

maintainer
#3 lore raw

PATCH 1/1 drm/amd/display: Interpret log_2(0) as 0

In reply to Bas Nieuwenhuizen

Even though log(0) is technically undefined our code assumes that log_2(0) is 0. This mirrors the current behavior of our log_2 implementation on non-Linux platforms.

---

What's num_pkrs value is upstream Mesa providing for CHIP_VANGOGH?

I saw that problem at bringup with an internal Mesa and had this fix for it.

Harry

 drivers/gpu/drm/amd/display/dc/basics/conversion.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/basics/conversion.h +5 -1
index ade785c4fdc7..da9883ec7b1c 100644
@@ -40,7 +40,11 @@ void convert_float_matrix(
 
 static inline unsigned int log_2(unsigned int num)
 {
-	return ilog2(num);
+	/*
+	 * Technically log(0) is undefined, but our code is structured
+	 * in a way that assumes log(0) = 0
+	 */
+	return num ? ilog2(num) : num;
 }
 
 #endif
Signature · 6 lines
2.28.0

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
#4 lore raw

Re: [PATCH] drm/amd/display: Interpret log_2(0) as 0

Quoted 9 lines · Harry Wentland

Even though log(0) is technically undefined our code assumes that log_2(0) is 0. This mirrors the current behavior of our log_2 implementation on non-Linux platforms.

---

What's num_pkrs value is upstream Mesa providing for CHIP_VANGOGH?

So mesa isn't quite providing a num_pkrs value to the HW as the render HW doesn't require you to set one, it assumes something.

In practice mesa still uses num_pkrs, which is provided by a readback of the GB_ADDR_CFG(CONFIG?) register info provided by the kernel. This can be used for e.g. determining texture coordinates (like we do for the displayable DCC transform) but otherwise isn't really needed, since the tile size for texture allocation is not dependent on num_pkrs.

so outside of displayable DCC I think you'd be able to get far in mesa with a wrong num_pkrs.

Quoted 27 lines

I saw that problem at bringup with an internal Mesa and had this fix for it.

Harry

 drivers/gpu/drm/amd/display/dc/basics/conversion.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/basics/conversion.h +5 -2
index ade785c4fdc7..da9883ec7b1c 100644
@@ -40,7 +40,11 @@ void convert_float_matrix(
 
 static inline unsigned int log_2(unsigned int num)
 {
-       return ilog2(num);
+       /*
+        * Technically log(0) is undefined, but our code is structured
+        * in a way that assumes log(0) = 0
+        */
+       return num ? ilog2(num) : num;
 }
 
 #endif
--
Quoted 2 lines

2.28.0

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

#5 lore raw

Re: [PATCH] drm/amd/display: Interpret log_2(0) as 0

num_pkrs is a hardware config parameter like the number of shader engines. It's constant for each chip.

Marek

Quoted 51 lines · Bas Nieuwenhuizen
Quoted 9 lines · Harry Wentland

Even though log(0) is technically undefined our code assumes that log_2(0) is 0. This mirrors the current behavior of our log_2 implementation on non-Linux platforms.

---

What's num_pkrs value is upstream Mesa providing for CHIP_VANGOGH?

So mesa isn't quite providing a num_pkrs value to the HW as the render HW doesn't require you to set one, it assumes something.

In practice mesa still uses num_pkrs, which is provided by a readback of the GB_ADDR_CFG(CONFIG?) register info provided by the kernel. This can be used for e.g. determining texture coordinates (like we do for the displayable DCC transform) but otherwise isn't really needed, since the tile size for texture allocation is not dependent on num_pkrs.

so outside of displayable DCC I think you'd be able to get far in mesa with a wrong num_pkrs.

Review

Quoted 10 lines

I saw that problem at bringup with an internal Mesa and had this fix for it.

Harry

 drivers/gpu/drm/amd/display/dc/basics/conversion.h | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
(quoted diff) +0 -0

b/drivers/gpu/drm/amd/display/dc/basics/conversion.h

Quoted 15 lines

index ade785c4fdc7..da9883ec7b1c 100644

drivers/gpu/drm/amd/display/dc/basics/conversion.h +5 -1
@@ -40,7 +40,11 @@ void convert_float_matrix(
 
 static inline unsigned int log_2(unsigned int num)
 {
-       return ilog2(num);
+       /*
+        * Technically log(0) is undefined, but our code is structured
+        * in a way that assumes log(0) = 0
+        */
+       return num ? ilog2(num) : num;
 }
 
Quoted 9 lines
Quoted 4 lines
drivers/gpu/drm/amd/display/dc/basics/conversion.h +0 -1
 #endif
--

2.28.0

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx