Skip to contentSkip to conversation

iommu/arm-smmu: Add Adreno SMMU specific implementation

PATCH Superseded Jordan Crouse 6 participants 73 messages Aug 10, 2020 – Sep 1, 2020

Subsystems (9)

v12 of 19 Superseded Latest is v19 →

lore
Share & tools
lore
#0 lore raw

v12 cover iommu/arm-smmu: Add Adreno SMMU specific implementation

This series adds an Adreno SMMU implementation to arm-smmu to allow GPU hardware pagetable switching.

The Adreno GPU has built in capabilities to switch the TTBR0 pagetable during runtime to allow each individual instance or application to have its own pagetable. In order to take advantage of the HW capabilities there are certain requirements needed of the SMMU hardware.

This series adds support for an Adreno specific arm-smmu implementation. The new implementation 1) ensures that the GPU domain is always assigned context bank 0,
2) enables split pagetable support (TTBR1) so that the instance specific pagetable can be swapped while the global memory remains in place and 3) shares the current pagetable configuration with the GPU driver to allow it to create its own io-pgtable instances.

The series then adds the drm/msm code to enable these features. For targets that support it allocate new pagetables using the io-pgtable configuration shared by the arm-smmu driver and swap them in during runtime.

This version of the series merges the previous patchset(s) [1] and [2] with the following improvements:

v12:
- Nitpick cleanups in gpu/drm/msm/msm_iommu.c (Rob Clark)
- Reorg in gpu/drm/msm/msm_gpu.c (Rob Clark)
- Use the default asid for the context bank so that iommu_tlb_flush_all works
- Flush the UCHE after a page switch
- Add the SCTLR.HUPCF patch at the end of the series
v11:
- Add implementation specific get_attr/set_attr functions (per Rob Clark)
- Fix context bank allocation (per Bjorn Andersson)
v10:
- arm-smmu: add implementation hook to allocate context banks
- arm-smmu: Match the GPU domain by stream ID instead of compatible string
- arm-smmu: Make DOMAIN_ATTR_PGTABLE_CFG bi-directional. The leaf driver queries the configuration to create a pagetable and then sends the newly created configuration back to the smmu-driver to enable TTBR0
- drm/msm: Add context reference counting for submissions
- drm/msm: Use dummy functions to skip TLB operations on per-instance pagetables

[1] https://lists.linuxfoundation.org/pipermail/iommu/2020-June/045653.html
[2] https://lists.linuxfoundation.org/pipermail/iommu/2020-June/045659.html

Jordan Crouse (13):
iommu/arm-smmu: Pass io-pgtable config to implementation specific
function
iommu/arm-smmu: Add support for split pagetables iommu/arm-smmu: Prepare for the adreno-smmu implementation
iommu: Add a domain attribute to get/set a pagetable configuration iommu/arm-smmu-qcom: Add implementation for the adreno GPU SMMU
dt-bindings: arm-smmu: Add compatible string for Adreno GPU SMMU drm/msm: Add a context pointer to the submitqueue drm/msm: Set the global virtual address range from the IOMMU domain drm/msm: Add support to create a local pagetable drm/msm: Add support for private address space instances drm/msm/a6xx: Add support for per-instance pagetables
arm: dts: qcom: sm845: Set the compatible string for the GPU SMMU iommu/arm-smmu: Add a init_context_bank implementation hook

 .../devicetree/bindings/iommu/arm,smmu.yaml   |   4 +
 arch/arm64/boot/dts/qcom/sdm845.dtsi          |   2 +-
 drivers/gpu/drm/msm/adreno/a5xx_gpu.c         |  12 +-
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c         |  75 ++++++-
 drivers/gpu/drm/msm/adreno/a6xx_gpu.h         |   1 +
 drivers/gpu/drm/msm/adreno/adreno_gpu.c       |  18 +-
 drivers/gpu/drm/msm/adreno/adreno_gpu.h       |   3 +-
 drivers/gpu/drm/msm/msm_drv.c                 |  16 +-
 drivers/gpu/drm/msm/msm_drv.h                 |  13 ++
 drivers/gpu/drm/msm/msm_gem.h                 |   1 +
 drivers/gpu/drm/msm/msm_gem_submit.c          |   8 +-
 drivers/gpu/drm/msm/msm_gem_vma.c             |   9 +
 drivers/gpu/drm/msm/msm_gpu.c                 |  31 ++-
 drivers/gpu/drm/msm/msm_gpu.h                 |  12 +-
 drivers/gpu/drm/msm/msm_gpummu.c              |   2 +-
 drivers/gpu/drm/msm/msm_iommu.c               | 197 +++++++++++++++++-
 drivers/gpu/drm/msm/msm_mmu.h                 |  16 +-
 drivers/gpu/drm/msm/msm_ringbuffer.h          |   1 +
 drivers/gpu/drm/msm/msm_submitqueue.c         |   8 +-
 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c    |   6 +-
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c    | 172 ++++++++++++++-
 drivers/iommu/arm/arm-smmu/arm-smmu.c         | 134 ++++++------
 drivers/iommu/arm/arm-smmu/arm-smmu.h         |  87 +++++++-
 include/linux/iommu.h                         |   1 +
 24 files changed, 708 insertions(+), 121 deletions(-)
Signature · 1 line
2.25.1
#1 lore raw

PATCH v12 1/13 iommu/arm-smmu: Pass io-pgtable config to implementation specific function

Construct the io-pgtable config before calling the implementation specific init_context function and pass it so the implementation specific function can get a chance to change it before the io-pgtable is created.

---

 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |  3 ++-
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 11 ++++++-----
 drivers/iommu/arm/arm-smmu/arm-smmu.h      |  3 ++-
 3 files changed, 10 insertions(+), 7 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +2 -1
index f4ff124a1967..a9861dcd0884 100644
@@ -68,7 +68,8 @@ static int cavium_cfg_probe(struct arm_smmu_device *smmu)
 	return 0;
 }
 
-static int cavium_init_context(struct arm_smmu_domain *smmu_domain)
+static int cavium_init_context(struct arm_smmu_domain *smmu_domain,
+		struct io_pgtable_cfg *pgtbl_cfg)
 {
 	struct cavium_smmu *cs = container_of(smmu_domain->smmu,
 					      struct cavium_smmu, smmu);
drivers/iommu/arm/arm-smmu/arm-smmu.c +6 -5
index 09c42af9f31e..37d8d49299b4 100644
@@ -795,11 +795,6 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 		cfg->asid = cfg->cbndx;
 
 	smmu_domain->smmu = smmu;
-	if (smmu->impl && smmu->impl->init_context) {
-		ret = smmu->impl->init_context(smmu_domain);
-		if (ret)
-			goto out_unlock;
-	}
 
 	pgtbl_cfg = (struct io_pgtable_cfg) {
 		.pgsize_bitmap	= smmu->pgsize_bitmap,
@@ -810,6 +805,12 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 		.iommu_dev	= smmu->dev,
 	};
 
+	if (smmu->impl && smmu->impl->init_context) {
+		ret = smmu->impl->init_context(smmu_domain, &pgtbl_cfg);
+		if (ret)
+			goto out_clear_smmu;
+	}
+
 	if (smmu_domain->non_strict)
 		pgtbl_cfg.quirks |= IO_PGTABLE_QUIRK_NON_STRICT;
 
drivers/iommu/arm/arm-smmu/arm-smmu.h +2 -1
index d890a4a968e8..83294516ac08 100644
@@ -386,7 +386,8 @@ struct arm_smmu_impl {
 			    u64 val);
 	int (*cfg_probe)(struct arm_smmu_device *smmu);
 	int (*reset)(struct arm_smmu_device *smmu);
-	int (*init_context)(struct arm_smmu_domain *smmu_domain);
+	int (*init_context)(struct arm_smmu_domain *smmu_domain,
+			struct io_pgtable_cfg *cfg);
 	void (*tlb_sync)(struct arm_smmu_device *smmu, int page, int sync,
 			 int status);
 	int (*def_domain_type)(struct device *dev);
Signature · 1 line
2.25.1
#2 lore raw

PATCH v12 2/13 iommu/arm-smmu: Add support for split pagetables

In reply to Jordan Crouse

Enable TTBR1 for a context bank if IO_PGTABLE_QUIRK_ARM_TTBR1 is selected by the io-pgtable configuration.

---

 drivers/iommu/arm/arm-smmu/arm-smmu.c | 21 ++++++++++++++++-----
 drivers/iommu/arm/arm-smmu/arm-smmu.h | 25 +++++++++++++++++++------
 2 files changed, 35 insertions(+), 11 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu.c +16 -5
index 37d8d49299b4..976d43a7f2ff 100644
@@ -552,11 +552,15 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain,
 			cb->ttbr[0] = pgtbl_cfg->arm_v7s_cfg.ttbr;
 			cb->ttbr[1] = 0;
 		} else {
-			cb->ttbr[0] = pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
-			cb->ttbr[0] |= FIELD_PREP(ARM_SMMU_TTBRn_ASID,
-						  cfg->asid);
+			cb->ttbr[0] = FIELD_PREP(ARM_SMMU_TTBRn_ASID,
+				cfg->asid);
 			cb->ttbr[1] = FIELD_PREP(ARM_SMMU_TTBRn_ASID,
-						 cfg->asid);
+				cfg->asid);
+
+			if (pgtbl_cfg->quirks & IO_PGTABLE_QUIRK_ARM_TTBR1)
+				cb->ttbr[1] |= pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
+			else
+				cb->ttbr[0] |= pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
 		}
 	} else {
 		cb->ttbr[0] = pgtbl_cfg->arm_lpae_s2_cfg.vttbr;
@@ -822,7 +826,14 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 
 	/* Update the domain's page sizes to reflect the page table format */
 	domain->pgsize_bitmap = pgtbl_cfg.pgsize_bitmap;
-	domain->geometry.aperture_end = (1UL << ias) - 1;
+
+	if (pgtbl_cfg.quirks & IO_PGTABLE_QUIRK_ARM_TTBR1) {
+		domain->geometry.aperture_start = ~0UL << ias;
+		domain->geometry.aperture_end = ~0UL;
+	} else {
+		domain->geometry.aperture_end = (1UL << ias) - 1;
+	}
+
 	domain->geometry.force_aperture = true;
 
 	/* Initialise the context bank with our page table cfg */
drivers/iommu/arm/arm-smmu/arm-smmu.h +19 -6
index 83294516ac08..f3e456893f28 100644
@@ -169,10 +169,12 @@ enum arm_smmu_cbar_type {
 #define ARM_SMMU_CB_TCR			0x30
 #define ARM_SMMU_TCR_EAE		BIT(31)
 #define ARM_SMMU_TCR_EPD1		BIT(23)
+#define ARM_SMMU_TCR_A1			BIT(22)
 #define ARM_SMMU_TCR_TG0		GENMASK(15, 14)
 #define ARM_SMMU_TCR_SH0		GENMASK(13, 12)
 #define ARM_SMMU_TCR_ORGN0		GENMASK(11, 10)
 #define ARM_SMMU_TCR_IRGN0		GENMASK(9, 8)
+#define ARM_SMMU_TCR_EPD0		BIT(7)
 #define ARM_SMMU_TCR_T0SZ		GENMASK(5, 0)
 
 #define ARM_SMMU_VTCR_RES1		BIT(31)
@@ -350,12 +352,23 @@ struct arm_smmu_domain {
 
 static inline u32 arm_smmu_lpae_tcr(struct io_pgtable_cfg *cfg)
 {
-	return ARM_SMMU_TCR_EPD1 |
-	       FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
-	       FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
-	       FIELD_PREP(ARM_SMMU_TCR_ORGN0, cfg->arm_lpae_s1_cfg.tcr.orgn) |
-	       FIELD_PREP(ARM_SMMU_TCR_IRGN0, cfg->arm_lpae_s1_cfg.tcr.irgn) |
-	       FIELD_PREP(ARM_SMMU_TCR_T0SZ, cfg->arm_lpae_s1_cfg.tcr.tsz);
+	u32 tcr = FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
+		FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
+		FIELD_PREP(ARM_SMMU_TCR_ORGN0, cfg->arm_lpae_s1_cfg.tcr.orgn) |
+		FIELD_PREP(ARM_SMMU_TCR_IRGN0, cfg->arm_lpae_s1_cfg.tcr.irgn) |
+		FIELD_PREP(ARM_SMMU_TCR_T0SZ, cfg->arm_lpae_s1_cfg.tcr.tsz);
+
+       /*
+	* When TTBR1 is selected shift the TCR fields by 16 bits and disable
+	* translation in TTBR0
+	*/
+	if (cfg->quirks & IO_PGTABLE_QUIRK_ARM_TTBR1) {
+		tcr = (tcr << 16) & ~ARM_SMMU_TCR_A1;
+		tcr |= ARM_SMMU_TCR_EPD0;
+	} else
+		tcr |= ARM_SMMU_TCR_EPD1;
+
+	return tcr;
 }
 
 static inline u32 arm_smmu_lpae_tcr2(struct io_pgtable_cfg *cfg)
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#3 lore raw

PATCH v12 3/13 iommu/arm-smmu: Prepare for the adreno-smmu implementation

In reply to Jordan Crouse

Do a bit of prep work to add the upcoming adreno-smmu implementation.

Add an hook to allow the implementation to choose which context banks to allocate. Then, add domain_attr_get / domain_attr_set hooks to allow for implementation specific domain attributes.

Move some of the common structs to arm-smmu.h in anticipation of them being used by the implementations and update some of the existing hooks to pass more information that the implementation will need.

These modifications will be used by the upcoming Adreno SMMU implementation to identify the GPU device and properly configure it for pagetable switching.

---

 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |  2 +-
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 83 ++++++++--------------
 drivers/iommu/arm/arm-smmu/arm-smmu.h      | 56 ++++++++++++++-
 3 files changed, 87 insertions(+), 54 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +1 -1
index a9861dcd0884..88f17cc33023 100644
@@ -69,7 +69,7 @@ static int cavium_cfg_probe(struct arm_smmu_device *smmu)
 }
 
 static int cavium_init_context(struct arm_smmu_domain *smmu_domain,
-		struct io_pgtable_cfg *pgtbl_cfg)
+		struct io_pgtable_cfg *pgtbl_cfg, struct device *dev)
 {
 	struct cavium_smmu *cs = container_of(smmu_domain->smmu,
 					      struct cavium_smmu, smmu);
drivers/iommu/arm/arm-smmu/arm-smmu.c +31 -52

149 lines not on this page · raw message

drivers/iommu/arm/arm-smmu/arm-smmu.h +55 -1

94 lines not on this page · raw message

Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#4 lore raw

PATCH v12 4/13 iommu: Add a domain attribute to get/set a pagetable configuration

In reply to Jordan Crouse

Add domain attribute DOMAIN_ATTR_PGTABLE_CFG. This will be used by arm-smmu to share the current pagetable configuration with the leaf driver and to allow the leaf driver to set up a new pagetable configuration under certain circumstances.

---

 include/linux/iommu.h | 1 +
 1 file changed, 1 insertion(+)
include/linux/iommu.h +1 -0
index fee209efb756..995ab8c47ef2 100644
@@ -118,6 +118,7 @@ enum iommu_attr {
 	DOMAIN_ATTR_FSL_PAMUV1,
 	DOMAIN_ATTR_NESTING,	/* two stages of translation */
 	DOMAIN_ATTR_DMA_USE_FLUSH_QUEUE,
+	DOMAIN_ATTR_PGTABLE_CFG,
 	DOMAIN_ATTR_MAX,
 };
 
Signature · 1 line
2.25.1
maintainer
#5 lore raw

Re: [PATCH v12 04/13] iommu: Add a domain attribute to get/set a pagetable configuration

Quoted 16 lines · Jordan Crouse

Add domain attribute DOMAIN_ATTR_PGTABLE_CFG. This will be used by arm-smmu to share the current pagetable configuration with the leaf driver and to allow the leaf driver to set up a new pagetable configuration under certain circumstances.

---

 include/linux/iommu.h | 1 +
 1 file changed, 1 insertion(+)
include/linux/iommu.h +0 -0
index fee209efb756..995ab8c47ef2 100644

include/linux/iommu.h enum iommu_attr { · line 123

Quoted 6 lines
include/linux/iommu.h +1 -0
@@ -118,6 +118,7 @@ enum iommu_attr {
 	DOMAIN_ATTR_FSL_PAMUV1,
 	DOMAIN_ATTR_NESTING,	/* two stages of translation */
 	DOMAIN_ATTR_DMA_USE_FLUSH_QUEUE,
+	DOMAIN_ATTR_PGTABLE_CFG,
 	DOMAIN_ATTR_MAX,
 };

Nobody other than the adreno gpu uses this, so can we avoid exposing it in the IOMMU API, please? Given that you have a reference to the adreno GPU device in the SMMU implementation code thanks to .alloc_context_bank(), can you squirrel some function pointers away in the driver data (i.e. with dev_set_drvdata()) instead?

Will

#6 lore raw

Re: [Freedreno] [PATCH v12 04/13] iommu: Add a domain attribute to get/set a pagetable configuration

Quoted 22 lines · Will Deacon
Quoted 20 lines · Jordan Crouse

Add domain attribute DOMAIN_ATTR_PGTABLE_CFG. This will be used by arm-smmu to share the current pagetable configuration with the leaf driver and to allow the leaf driver to set up a new pagetable configuration under certain circumstances.

---

 include/linux/iommu.h | 1 +
 1 file changed, 1 insertion(+)
include/linux/iommu.h +1 -0
index fee209efb756..995ab8c47ef2 100644
@@ -118,6 +118,7 @@ enum iommu_attr {
      DOMAIN_ATTR_FSL_PAMUV1,
      DOMAIN_ATTR_NESTING,    /* two stages of translation */
      DOMAIN_ATTR_DMA_USE_FLUSH_QUEUE,
+     DOMAIN_ATTR_PGTABLE_CFG,
Quoted 9 lines

include/linux/iommu.h enum iommu_attr { · line 123

Quoted 2 lines
include/linux/iommu.h +0 -0
      DOMAIN_ATTR_MAX,
 };

Nobody other than the adreno gpu uses this, so can we avoid exposing it in the IOMMU API, please? Given that you have a reference to the adreno GPU device in the SMMU implementation code thanks to .alloc_context_bank(), can you squirrel some function pointers away in the driver data (i.e. with dev_set_drvdata()) instead?

Hmm, we are already using drvdata on the gpu side, and it looks like arm-smmu is also using it. Could we get away with stashing an extra 'void *' in iommu_domain itself?

Or alternatively, if we had a is_arm_smmu_domain(domain), then we could just directly call some exported private fxns with the domain ptr (which could then verify that the domain is actually an arm_smmu_domain, and then from there that the smmu is indeed qcom,adreno-smmu, to keep things sane)

BR,
-R

maintainer
#7 lore raw

Re: [Freedreno] [PATCH v12 04/13] iommu: Add a domain attribute to get/set a pagetable configuration

Quoted 33 lines · Rob Clark
Quoted 22 lines · Will Deacon
Quoted 20 lines · Jordan Crouse

Add domain attribute DOMAIN_ATTR_PGTABLE_CFG. This will be used by arm-smmu to share the current pagetable configuration with the leaf driver and to allow the leaf driver to set up a new pagetable configuration under certain circumstances.

---

 include/linux/iommu.h | 1 +
 1 file changed, 1 insertion(+)
include/linux/iommu.h +1 -0
index fee209efb756..995ab8c47ef2 100644
@@ -118,6 +118,7 @@ enum iommu_attr {
      DOMAIN_ATTR_FSL_PAMUV1,
      DOMAIN_ATTR_NESTING,    /* two stages of translation */
      DOMAIN_ATTR_DMA_USE_FLUSH_QUEUE,
+     DOMAIN_ATTR_PGTABLE_CFG,
Quoted 9 lines

include/linux/iommu.h enum iommu_attr { · line 123

Quoted 2 lines
include/linux/iommu.h +0 -0
      DOMAIN_ATTR_MAX,
 };

Nobody other than the adreno gpu uses this, so can we avoid exposing it in the IOMMU API, please? Given that you have a reference to the adreno GPU device in the SMMU implementation code thanks to .alloc_context_bank(), can you squirrel some function pointers away in the driver data (i.e. with dev_set_drvdata()) instead?

Quoted 3 lines

Hmm, we are already using drvdata on the gpu side, and it looks like arm-smmu is also using it. Could we get away with stashing an extra 'void *' in iommu_domain itself?

What I meant was, expose the type of whatever you put in there on the GPU side so that the SMMU impl can install its function pointers into a field of that structure. As far as I'm concerned, the SMMU impl code and the GPU driver are the same entity and we should keep their communication private, rather than expose it up the stack. After all, the GPU writes to the SMMU registers!

If you really don't want to expose all of your gubbins, I suppose you could have a structure just for the SMMU view and container_of() out of that on the GPU side.

Will
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#8 lore raw

Re: [Freedreno] [PATCH v12 04/13] iommu: Add a domain attribute to get/set a pagetable configuration

Quoted 46 lines · Will Deacon
Quoted 33 lines · Rob Clark
Quoted 22 lines · Will Deacon
Quoted 20 lines · Jordan Crouse

Add domain attribute DOMAIN_ATTR_PGTABLE_CFG. This will be used by arm-smmu to share the current pagetable configuration with the leaf driver and to allow the leaf driver to set up a new pagetable configuration under certain circumstances.

---

 include/linux/iommu.h | 1 +
 1 file changed, 1 insertion(+)
include/linux/iommu.h +1 -0
index fee209efb756..995ab8c47ef2 100644
@@ -118,6 +118,7 @@ enum iommu_attr {
      DOMAIN_ATTR_FSL_PAMUV1,
      DOMAIN_ATTR_NESTING,    /* two stages of translation */
      DOMAIN_ATTR_DMA_USE_FLUSH_QUEUE,
+     DOMAIN_ATTR_PGTABLE_CFG,
Quoted 9 lines

include/linux/iommu.h enum iommu_attr { · line 123

Quoted 2 lines
include/linux/iommu.h +0 -0
      DOMAIN_ATTR_MAX,
 };

Nobody other than the adreno gpu uses this, so can we avoid exposing it in the IOMMU API, please? Given that you have a reference to the adreno GPU device in the SMMU implementation code thanks to .alloc_context_bank(), can you squirrel some function pointers away in the driver data (i.e. with dev_set_drvdata()) instead?

Quoted 3 lines

Hmm, we are already using drvdata on the gpu side, and it looks like arm-smmu is also using it. Could we get away with stashing an extra 'void *' in iommu_domain itself?

What I meant was, expose the type of whatever you put in there on the GPU side so that the SMMU impl can install its function pointers into a field of that structure. As far as I'm concerned, the SMMU impl code and the GPU driver are the same entity and we should keep their communication private, rather than expose it up the stack. After all, the GPU writes to the SMMU registers!

Quoted 3 lines

If you really don't want to expose all of your gubbins, I suppose you could have a structure just for the SMMU view and container_of() out of that on the GPU side.

yeah, msm_gpu has a lot of internal state.. but I suppose we could define a 'struct adreno_smmu_priv' and embed that in msm_gpu, and throw in a get_gpu_drvdata() type wrapper for get_drvdata() to make this not totally horrible in the various cases places that use get_drvdata() currently.

BR,
-R
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#9 lore raw

PATCH v12 5/13 iommu/arm-smmu-qcom: Add implementation for the adreno GPU SMMU

In reply to Jordan Crouse

Add a special implementation for the SMMU attached to most Adreno GPU target triggered from the qcom,adreno-smmu compatible string.

The new Adreno SMMU implementation will enable split pagetables (TTBR1) for the domain attached to the GPU device (SID 0) and hard code it context bank 0 so the GPU hardware can implement per-instance pagetables.

---

 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |   3 +
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 156 ++++++++++++++++++++-
 2 files changed, 157 insertions(+), 2 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +3 -0
index 88f17cc33023..d199b4bff15d 100644
@@ -223,6 +223,9 @@ struct arm_smmu_device *arm_smmu_impl_init(struct arm_smmu_device *smmu)
 	    of_device_is_compatible(np, "qcom,sm8250-smmu-500"))
 		return qcom_smmu_impl_init(smmu);
 
+	if (of_device_is_compatible(smmu->dev->of_node, "qcom,adreno-smmu"))
+		return qcom_adreno_smmu_impl_init(smmu);
+
 	if (of_device_is_compatible(np, "marvell,ap806-smmu-500"))
 		smmu->impl = &mrvl_mmu500_impl;
 
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +154 -2

175 lines not on this page · raw message

Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
maintainer
#10 lore raw

Re: [PATCH v12 05/13] iommu/arm-smmu-qcom: Add implementation for the adreno GPU SMMU

Quoted 45 lines · Jordan Crouse

Add a special implementation for the SMMU attached to most Adreno GPU target triggered from the qcom,adreno-smmu compatible string.

The new Adreno SMMU implementation will enable split pagetables (TTBR1) for the domain attached to the GPU device (SID 0) and hard code it context bank 0 so the GPU hardware can implement per-instance pagetables.

---

 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |   3 +
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 156 ++++++++++++++++++++-
 2 files changed, 157 insertions(+), 2 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +3 -0
index 88f17cc33023..d199b4bff15d 100644
@@ -223,6 +223,9 @@ struct arm_smmu_device *arm_smmu_impl_init(struct arm_smmu_device *smmu)
 	    of_device_is_compatible(np, "qcom,sm8250-smmu-500"))
 		return qcom_smmu_impl_init(smmu);
 
+	if (of_device_is_compatible(smmu->dev->of_node, "qcom,adreno-smmu"))
+		return qcom_adreno_smmu_impl_init(smmu);
+
 	if (of_device_is_compatible(np, "marvell,ap806-smmu-500"))
 		smmu->impl = &mrvl_mmu500_impl;
 
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +8 -0
index be4318044f96..3be10145bf57 100644
@@ -12,6 +12,138 @@ struct qcom_smmu {
 	struct arm_smmu_device smmu;
 };
 
+#define QCOM_ADRENO_SMMU_GPU_SID 0
+
+static bool qcom_adreno_smmu_is_gpu_device(struct device *dev)
+{
+	struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+	struct arm_smmu_master_cfg *cfg = dev_iommu_priv_get(dev);
+	int idx, i;
+

drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c struct qcom_smmu { · line 32

Quoted 10 lines
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +10 -0
+	/*
+	 * The GPU will always use SID 0 so that is a handy way to uniquely
+	 * identify it and configure it for per-instance pagetables
+	 */
+	for_each_cfg_sme(cfg, fwspec, i, idx) {
+		u16 sid = FIELD_GET(ARM_SMMU_SMR_ID, fwspec->ids[i]);
+
+		if (sid == QCOM_ADRENO_SMMU_GPU_SID)
+			return true;
+	}

Is for_each_cfg_sme() really what you want here? You're not using idx for anything, so I guess it should really be a loop over the sids (e.g. a bog standard for loop from 0 to fw->num_ids - 1)?

Will

#11 lore raw

PATCH v12 6/13 dt-bindings: arm-smmu: Add compatible string for Adreno GPU SMMU

In reply to Jordan Crouse

Every Qcom Adreno GPU has an embedded SMMU for its own use. These devices depend on unique features such as split pagetables, different stall/halt requirements and other settings. Identify them with a compatible string so that they can be identified in the arm-smmu implementation specific code.

---

 Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 4 ++++
 1 file changed, 4 insertions(+)
Documentation/devicetree/bindings/iommu/arm,smmu.yaml +4 -0
index 503160a7b9a0..70996348c1d8 100644
@@ -49,6 +49,10 @@ properties:
           - enum:
               - nvidia,tegra194-smmu
           - const: nvidia,smmu-500
+      - description: Qcom Adreno GPUs implementing "arm,smmu-v2"
+        items:
+          - const: qcom,adreno-smmu
+          - const: qcom,smmu-v2
       - items:
           - const: arm,mmu-500
           - const: arm,smmu-v2
Signature · 1 line
2.25.1
#12 lore raw

PATCH v12 7/13 drm/msm: Add a context pointer to the submitqueue

In reply to Jordan Crouse

Each submitqueue is attached to a context. Add a pointer to the context to the submitqueue at create time and refcount it so that it stays around through the life of the queue.

GPU submissions can access the active context via the submitqueue instead of requiring it to be passed around from function to function.

---

 drivers/gpu/drm/msm/adreno/a5xx_gpu.c   | 12 +++++-------
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c   |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.c |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.h |  3 +--
 drivers/gpu/drm/msm/msm_drv.c           |  3 ++-
 drivers/gpu/drm/msm/msm_drv.h           |  8 ++++++++
 drivers/gpu/drm/msm/msm_gem.h           |  1 +
 drivers/gpu/drm/msm/msm_gem_submit.c    |  8 ++++----
 drivers/gpu/drm/msm/msm_gpu.c           |  9 ++++-----
 drivers/gpu/drm/msm/msm_gpu.h           |  7 +++----
 drivers/gpu/drm/msm/msm_submitqueue.c   |  8 +++++++-
 11 files changed, 39 insertions(+), 30 deletions(-)
drivers/gpu/drm/msm/adreno/a5xx_gpu.c +5 -7
index 9e63a190642c..eff2439ea57b 100644
@@ -43,8 +43,7 @@ static void a5xx_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring)
 		gpu_write(gpu, REG_A5XX_CP_RB_WPTR, wptr);
 }
 
-static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct msm_drm_private *priv = gpu->dev->dev_private;
 	struct msm_ringbuffer *ring = submit->ring;
@@ -57,7 +56,7 @@ static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
@@ -103,8 +102,7 @@ static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
 	msm_gpu_retire(gpu);
 }
 
-static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
 	struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu);
@@ -114,7 +112,7 @@ static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 
 	if (IS_ENABLED(CONFIG_DRM_MSM_GPU_SUDO) && submit->in_rb) {
 		priv->lastctx = NULL;
-		a5xx_submit_in_rb(gpu, submit, ctx);
+		a5xx_submit_in_rb(gpu, submit);
 		return;
 	}
 
@@ -148,7 +146,7 @@ static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +2 -3
index c5a3e4d4c007..5eabb0109577 100644
@@ -81,8 +81,7 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
 	OUT_RING(ring, upper_32_bits(iova));
 }
 
-static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
 	struct msm_drm_private *priv = gpu->dev->dev_private;
@@ -115,7 +114,7 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/adreno_gpu.c +2 -3
index e23641a5ec84..b38a8126541a 100644
@@ -457,8 +457,7 @@ void adreno_recover(struct msm_gpu *gpu)
 	}
 }
 
-void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx)
+void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
 	struct msm_drm_private *priv = gpu->dev->dev_private;
@@ -472,7 +471,7 @@ void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
 			/* ignore if there has not been a ctx switch: */
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/adreno_gpu.h +1 -2
index 99bb468f5f24..0ae8b373c428 100644
@@ -267,8 +267,7 @@ struct drm_gem_object *adreno_fw_create_bo(struct msm_gpu *gpu,
 		const struct firmware *fw, u64 *iova);
 int adreno_hw_init(struct msm_gpu *gpu);
 void adreno_recover(struct msm_gpu *gpu);
-void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx);
+void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 void adreno_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 bool adreno_idle(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 #if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP)
drivers/gpu/drm/msm/msm_drv.c +2 -1
index 7d641c7e3514..108b663c3ef2 100644
@@ -594,6 +594,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	if (!ctx)
 		return -ENOMEM;
 
+	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
 	ctx->aspace = priv->gpu ? priv->gpu->aspace : NULL;
@@ -615,7 +616,7 @@ static int msm_open(struct drm_device *dev, struct drm_file *file)
 static void context_close(struct msm_file_private *ctx)
 {
 	msm_submitqueue_close(ctx);
-	kfree(ctx);
+	kref_put(&ctx->ref, msm_file_private_destroy);
 }
 
 static void msm_postclose(struct drm_device *dev, struct drm_file *file)
drivers/gpu/drm/msm/msm_drv.h +8 -0
index af259b0573ea..f69c6d62584d 100644
@@ -57,6 +57,7 @@ struct msm_file_private {
 	struct list_head submitqueues;
 	int queueid;
 	struct msm_gem_address_space *aspace;
+	struct kref ref;
 };
 
 enum msm_mdp_plane_property {
@@ -428,6 +429,13 @@ void msm_submitqueue_close(struct msm_file_private *ctx);
 
 void msm_submitqueue_destroy(struct kref *kref);
 
+static inline void msm_file_private_destroy(struct kref *kref)
+{
+	struct msm_file_private *ctx = container_of(kref,
+		struct msm_file_private, ref);
+
+	kfree(ctx);
+}
 
 #define DBG(fmt, ...) DRM_DEBUG_DRIVER(fmt"\n", ##__VA_ARGS__)
 #define VERB(fmt, ...) if (0) DRM_DEBUG_DRIVER(fmt"\n", ##__VA_ARGS__)
drivers/gpu/drm/msm/msm_gem.h +1 -0
index 972490b14ba5..9c573c4269cb 100644
@@ -142,6 +142,7 @@ struct msm_gem_submit {
 	bool valid;         /* true if no cmdstream patching needed */
 	bool in_rb;         /* "sudo" mode, copy cmds into RB */
 	struct msm_ringbuffer *ring;
+	struct msm_file_private *ctx;
 	unsigned int nr_cmds;
 	unsigned int nr_bos;
 	u32 ident;	   /* A "identifier" for the submit for logging */
drivers/gpu/drm/msm/msm_gem_submit.c +4 -4
index 8cb9aa15ff90..aa5c60a7132d 100644
@@ -27,7 +27,7 @@
 #define BO_PINNED   0x2000
 
 static struct msm_gem_submit *submit_create(struct drm_device *dev,
-		struct msm_gpu *gpu, struct msm_gem_address_space *aspace,
+		struct msm_gpu *gpu,
 		struct msm_gpu_submitqueue *queue, uint32_t nr_bos,
 		uint32_t nr_cmds)
 {
@@ -43,7 +43,7 @@ static struct msm_gem_submit *submit_create(struct drm_device *dev,
 		return NULL;
 
 	submit->dev = dev;
-	submit->aspace = aspace;
+	submit->aspace = queue->ctx->aspace;
 	submit->gpu = gpu;
 	submit->fence = NULL;
 	submit->cmd = (void *)&submit->bos[nr_bos];
@@ -677,7 +677,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		}
 	}
 
-	submit = submit_create(dev, gpu, ctx->aspace, queue, args->nr_bos,
+	submit = submit_create(dev, gpu, queue, args->nr_bos,
 		args->nr_cmds);
 	if (!submit) {
 		ret = -ENOMEM;
@@ -785,7 +785,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		}
 	}
 
-	msm_gpu_submit(gpu, submit, ctx);
+	msm_gpu_submit(gpu, submit);
 
 	args->fence = submit->fence->seqno;
 
drivers/gpu/drm/msm/msm_gpu.c +4 -5
index d5645472b25d..a1f3da6550e5 100644
@@ -520,7 +520,7 @@ static void recover_worker(struct work_struct *work)
 			struct msm_ringbuffer *ring = gpu->rb[i];
 
 			list_for_each_entry(submit, &ring->submits, node)
-				gpu->funcs->submit(gpu, submit, NULL);
+				gpu->funcs->submit(gpu, submit);
 		}
 	}
 
@@ -747,8 +747,7 @@ void msm_gpu_retire(struct msm_gpu *gpu)
 }
 
 /* add bo's to gpu's ring, and kick gpu: */
-void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx)
+void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct drm_device *dev = gpu->dev;
 	struct msm_drm_private *priv = dev->dev_private;
@@ -788,8 +787,8 @@ void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 			msm_gem_move_to_active(&msm_obj->base, gpu, false, submit->fence);
 	}
 
-	gpu->funcs->submit(gpu, submit, ctx);
-	priv->lastctx = ctx;
+	gpu->funcs->submit(gpu, submit);
+	priv->lastctx = submit->queue->ctx;
 
 	hangcheck_timer_reset(gpu);
 }
drivers/gpu/drm/msm/msm_gpu.h +3 -4
index 0db117a7339b..d496d488222c 100644
@@ -44,8 +44,7 @@ struct msm_gpu_funcs {
 	int (*hw_init)(struct msm_gpu *gpu);
 	int (*pm_suspend)(struct msm_gpu *gpu);
 	int (*pm_resume)(struct msm_gpu *gpu);
-	void (*submit)(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-			struct msm_file_private *ctx);
+	void (*submit)(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 	void (*flush)(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 	irqreturn_t (*irq)(struct msm_gpu *irq);
 	struct msm_ringbuffer *(*active_ring)(struct msm_gpu *gpu);
@@ -181,6 +180,7 @@ struct msm_gpu_submitqueue {
 	u32 flags;
 	u32 prio;
 	int faults;
+	struct msm_file_private *ctx;
 	struct list_head node;
 	struct kref ref;
 };
@@ -280,8 +280,7 @@ int msm_gpu_perfcntr_sample(struct msm_gpu *gpu, uint32_t *activetime,
 		uint32_t *totaltime, uint32_t ncntrs, uint32_t *cntrs);
 
 void msm_gpu_retire(struct msm_gpu *gpu);
-void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx);
+void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 
 int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
drivers/gpu/drm/msm/msm_submitqueue.c +7 -1
index a1d94be7883a..10f557225a3e 100644
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
 	 * No lock needed in close and there won't
 	 * be any more user ioctls coming our way
 	 */
-	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+		kref_put(&ctx->ref, msm_file_private_destroy);
 		msm_submitqueue_put(entry);
+	}
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
@@ -81,6 +83,9 @@ int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
 
 	write_lock(&ctx->queuelock);
 
+	kref_get(&ctx->ref);
+
+	queue->ctx = ctx;
 	queue->id = ctx->queueid++;
 
 	if (id)
@@ -177,6 +182,7 @@ int msm_submitqueue_remove(struct msm_file_private *ctx, u32 id)
 			list_del(&entry->node);
 			write_unlock(&ctx->queuelock);
 
+			kref_put(&ctx->ref, msm_file_private_destroy);
 			msm_submitqueue_put(entry);
 			return 0;
 		}
Signature · 6 lines
2.25.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#13 lore raw

Re: [Freedreno] [PATCH v12 07/13] drm/msm: Add a context pointer to the submitqueue

Quoted 25 lines · Jordan Crouse

Each submitqueue is attached to a context. Add a pointer to the context to the submitqueue at create time and refcount it so that it stays around through the life of the queue.

GPU submissions can access the active context via the submitqueue instead of requiring it to be passed around from function to function.

---

 drivers/gpu/drm/msm/adreno/a5xx_gpu.c   | 12 +++++-------
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c   |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.c |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.h |  3 +--
 drivers/gpu/drm/msm/msm_drv.c           |  3 ++-
 drivers/gpu/drm/msm/msm_drv.h           |  8 ++++++++
 drivers/gpu/drm/msm/msm_gem.h           |  1 +
 drivers/gpu/drm/msm/msm_gem_submit.c    |  8 ++++----
 drivers/gpu/drm/msm/msm_gpu.c           |  9 ++++-----
 drivers/gpu/drm/msm/msm_gpu.h           |  7 +++----
 drivers/gpu/drm/msm/msm_submitqueue.c   |  8 +++++++-
 11 files changed, 39 insertions(+), 30 deletions(-)

[snip]

drivers/gpu/drm/msm/msm_gem.h struct msm_gem_submit { · line 145

Quoted 9 lines
drivers/gpu/drm/msm/msm_gem.h +1 -0
index 972490b14ba5..9c573c4269cb 100644
@@ -142,6 +142,7 @@ struct msm_gem_submit {
        bool valid;         /* true if no cmdstream patching needed */
        bool in_rb;         /* "sudo" mode, copy cmds into RB */
        struct msm_ringbuffer *ring;
+       struct msm_file_private *ctx;

So, it looks like this is (currently) unused, and everything is instead using submit->queue->ctx

That said, changing this so the submit also holds a ref to the ctx seems to fix the intermittent splat I can trigger by repeatedly hanging the gpu. Which (from the pile of additional tracepoints I've added on top of this series) seems to be related to re-playing submits after the userspace process has crashed and/or closed the device.

It seems like the reference the submit holds to the queue should keep the ctx (and therefore address space) alive, but I need to dig through that a bit more.

BR,
-R

drivers/gpu/drm/msm/msm_gem.h struct msm_gem_submit { · line 148

Quoted 3 lines
drivers/gpu/drm/msm/msm_gem.h +0 -0
        unsigned int nr_cmds;
        unsigned int nr_bos;
        u32 ident;         /* A "identifier" for the submit for logging */

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

#14 lore raw

Re: [Freedreno] [PATCH v12 07/13] drm/msm: Add a context pointer to the submitqueue

In reply to Jordan Crouse
Quoted 25 lines · Jordan Crouse

Each submitqueue is attached to a context. Add a pointer to the context to the submitqueue at create time and refcount it so that it stays around through the life of the queue.

GPU submissions can access the active context via the submitqueue instead of requiring it to be passed around from function to function.

---

 drivers/gpu/drm/msm/adreno/a5xx_gpu.c   | 12 +++++-------
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c   |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.c |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.h |  3 +--
 drivers/gpu/drm/msm/msm_drv.c           |  3 ++-
 drivers/gpu/drm/msm/msm_drv.h           |  8 ++++++++
 drivers/gpu/drm/msm/msm_gem.h           |  1 +
 drivers/gpu/drm/msm/msm_gem_submit.c    |  8 ++++----
 drivers/gpu/drm/msm/msm_gpu.c           |  9 ++++-----
 drivers/gpu/drm/msm/msm_gpu.h           |  7 +++----
 drivers/gpu/drm/msm/msm_submitqueue.c   |  8 +++++++-
 11 files changed, 39 insertions(+), 30 deletions(-)

[snip]

drivers/gpu/drm/msm/msm_submitqueue.c void msm_submitqueue_close(struct msm_file_private *ctx) · line 55

Quoted 13 lines
drivers/gpu/drm/msm/msm_submitqueue.c +3 -1
index a1d94be7883a..10f557225a3e 100644
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
         * No lock needed in close and there won't
         * be any more user ioctls coming our way
         */
-       list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+       list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+               kref_put(&ctx->ref, msm_file_private_destroy);
                msm_submitqueue_put(entry);
+       }

oh, this is the problem I mentioned in the last email.. we are dropping the queue's reference to the ctx, when the device file is closed, not on the last unref of the queue. So the queue stays live until all associated submits are retired, but the ctx ref (and therefore the aspace) get destroyed earlier

BR,
-R

drivers/gpu/drm/msm/msm_submitqueue.c void msm_submitqueue_close(struct msm_file_private *ctx) · line 58

Quoted 3 lines
drivers/gpu/drm/msm/msm_submitqueue.c +0 -0
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,

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

#15 lore raw

PATCH v12 8/13 drm/msm: Set the global virtual address range from the IOMMU domain

In reply to Jordan Crouse

Use the aperture settings from the IOMMU domain to set up the virtual address range for the GPU. This allows us to transparently deal with IOMMU side features (like split pagetables).

---

 drivers/gpu/drm/msm/adreno/adreno_gpu.c | 13 +++++++++++--
 drivers/gpu/drm/msm/msm_iommu.c         |  7 +++++++
 2 files changed, 18 insertions(+), 2 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_gpu.c +11 -2
index b38a8126541a..f9e3badf2fca 100644
@@ -192,9 +192,18 @@ adreno_iommu_create_address_space(struct msm_gpu *gpu,
 	struct iommu_domain *iommu = iommu_domain_alloc(&platform_bus_type);
 	struct msm_mmu *mmu = msm_iommu_new(&pdev->dev, iommu);
 	struct msm_gem_address_space *aspace;
+	u64 start, size;
 
-	aspace = msm_gem_address_space_create(mmu, "gpu", SZ_16M,
-		0xffffffff - SZ_16M);
+	/*
+	 * Use the aperture start or SZ_16M, whichever is greater. This will
+	 * ensure that we align with the allocated pagetable range while still
+	 * allowing room in the lower 32 bits for GMEM and whatnot
+	 */
+	start = max_t(u64, SZ_16M, iommu->geometry.aperture_start);
+	size = iommu->geometry.aperture_end - start + 1;
+
+	aspace = msm_gem_address_space_create(mmu, "gpu",
+		start & GENMASK(48, 0), size);
 
 	if (IS_ERR(aspace) && !IS_ERR(mmu))
 		mmu->funcs->destroy(mmu);
drivers/gpu/drm/msm/msm_iommu.c +7 -0
index 3a381a9674c9..1b6635504069 100644
@@ -36,6 +36,10 @@ static int msm_iommu_map(struct msm_mmu *mmu, uint64_t iova,
 	struct msm_iommu *iommu = to_msm_iommu(mmu);
 	size_t ret;
 
+	/* The arm-smmu driver expects the addresses to be sign extended */
+	if (iova & BIT_ULL(48))
+		iova |= GENMASK_ULL(63, 49);
+
 	ret = iommu_map_sg(iommu->domain, iova, sgt->sgl, sgt->nents, prot);
 	WARN_ON(!ret);
 
@@ -46,6 +50,9 @@ static int msm_iommu_unmap(struct msm_mmu *mmu, uint64_t iova, size_t len)
 {
 	struct msm_iommu *iommu = to_msm_iommu(mmu);
 
+	if (iova & BIT_ULL(48))
+		iova |= GENMASK_ULL(63, 49);
+
 	iommu_unmap(iommu->domain, iova, len);
 
 	return 0;
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#16 lore raw

PATCH v12 9/13 drm/msm: Add support to create a local pagetable

In reply to Jordan Crouse

Add support to create a io-pgtable for use by targets that support per-instance pagetables. In order to support per-instance pagetables the GPU SMMU device needs to have the qcom,adreno-smmu compatible string and split pagetables enabled.

---

 drivers/gpu/drm/msm/msm_gpummu.c |   2 +-
 drivers/gpu/drm/msm/msm_iommu.c  | 190 ++++++++++++++++++++++++++++++-
 drivers/gpu/drm/msm/msm_mmu.h    |  16 ++-
 3 files changed, 205 insertions(+), 3 deletions(-)
drivers/gpu/drm/msm/msm_gpummu.c +1 -1
index 310a31b05faa..aab121f4beb7 100644
@@ -102,7 +102,7 @@ struct msm_mmu *msm_gpummu_new(struct device *dev, struct msm_gpu *gpu)
 	}
 
 	gpummu->gpu = gpu;
-	msm_mmu_init(&gpummu->base, dev, &funcs);
+	msm_mmu_init(&gpummu->base, dev, &funcs, MSM_MMU_GPUMMU);
 
 	return &gpummu->base;
 }
drivers/gpu/drm/msm/msm_iommu.c +189 -1

213 lines not on this page · raw message

drivers/gpu/drm/msm/msm_mmu.h +15 -1
index 3a534ee59bf6..61ade89d9e48 100644
@@ -17,18 +17,26 @@ struct msm_mmu_funcs {
 	void (*destroy)(struct msm_mmu *mmu);
 };
 
+enum msm_mmu_type {
+	MSM_MMU_GPUMMU,
+	MSM_MMU_IOMMU,
+	MSM_MMU_IOMMU_PAGETABLE,
+};
+
 struct msm_mmu {
 	const struct msm_mmu_funcs *funcs;
 	struct device *dev;
 	int (*handler)(void *arg, unsigned long iova, int flags);
 	void *arg;
+	enum msm_mmu_type type;
 };
 
 static inline void msm_mmu_init(struct msm_mmu *mmu, struct device *dev,
-		const struct msm_mmu_funcs *funcs)
+		const struct msm_mmu_funcs *funcs, enum msm_mmu_type type)
 {
 	mmu->dev = dev;
 	mmu->funcs = funcs;
+	mmu->type = type;
 }
 
 struct msm_mmu *msm_iommu_new(struct device *dev, struct iommu_domain *domain);
@@ -41,7 +49,13 @@ static inline void msm_mmu_set_fault_handler(struct msm_mmu *mmu, void *arg,
 	mmu->handler = handler;
 }
 
+struct msm_mmu *msm_iommu_pagetable_create(struct msm_mmu *parent);
+
 void msm_gpummu_params(struct msm_mmu *mmu, dma_addr_t *pt_base,
 		dma_addr_t *tran_error);
 
+
+int msm_iommu_pagetable_params(struct msm_mmu *mmu, phys_addr_t *ttbr,
+		int *asid);
+
 #endif /* __MSM_MMU_H__ */
Signature · 1 line
2.25.1
#17 lore raw

Re: [PATCH v12 09/13] drm/msm: Add support to create a local pagetable

Quoted 6 lines · Jordan Crouse

Add support to create a io-pgtable for use by targets that support per-instance pagetables. In order to support per-instance pagetables the GPU SMMU device needs to have the qcom,adreno-smmu compatible string and split pagetables enabled.

This patch adds a hard dependency on IOMMU_IO_PGTABLE (for alloc_io_pgtable_ops and free_io_pgtable_ops).

Guenter

Quoted 285 lines

---

 drivers/gpu/drm/msm/msm_gpummu.c |   2 +-
 drivers/gpu/drm/msm/msm_iommu.c  | 190 ++++++++++++++++++++++++++++++-
 drivers/gpu/drm/msm/msm_mmu.h    |  16 ++-
 3 files changed, 205 insertions(+), 3 deletions(-)
drivers/gpu/drm/msm/msm_gpummu.c +1 -1
index 310a31b05faa..aab121f4beb7 100644
@@ -102,7 +102,7 @@ struct msm_mmu *msm_gpummu_new(struct device *dev, struct msm_gpu *gpu)
 	}
 
 	gpummu->gpu = gpu;
-	msm_mmu_init(&gpummu->base, dev, &funcs);
+	msm_mmu_init(&gpummu->base, dev, &funcs, MSM_MMU_GPUMMU);
 
 	return &gpummu->base;
 }
drivers/gpu/drm/msm/msm_iommu.c +189 -1
index 1b6635504069..bc04dda8a198 100644
@@ -4,15 +4,201 @@
  * Author: Rob Clark <[email protected]>
  */
 
+#include <linux/io-pgtable.h>
 #include "msm_drv.h"
 #include "msm_mmu.h"
 
 struct msm_iommu {
 	struct msm_mmu base;
 	struct iommu_domain *domain;
+	atomic_t pagetables;
 };
+
 #define to_msm_iommu(x) container_of(x, struct msm_iommu, base)
 
+struct msm_iommu_pagetable {
+	struct msm_mmu base;
+	struct msm_mmu *parent;
+	struct io_pgtable_ops *pgtbl_ops;
+	phys_addr_t ttbr;
+	u32 asid;
+};
+static struct msm_iommu_pagetable *to_pagetable(struct msm_mmu *mmu)
+{
+	return container_of(mmu, struct msm_iommu_pagetable, base);
+}
+
+static int msm_iommu_pagetable_unmap(struct msm_mmu *mmu, u64 iova,
+		size_t size)
+{
+	struct msm_iommu_pagetable *pagetable = to_pagetable(mmu);
+	struct io_pgtable_ops *ops = pagetable->pgtbl_ops;
+	size_t unmapped = 0;
+
+	/* Unmap the block one page at a time */
+	while (size) {
+		unmapped += ops->unmap(ops, iova, 4096, NULL);
+		iova += 4096;
+		size -= 4096;
+	}
+
+	iommu_flush_tlb_all(to_msm_iommu(pagetable->parent)->domain);
+
+	return (unmapped == size) ? 0 : -EINVAL;
+}
+
+static int msm_iommu_pagetable_map(struct msm_mmu *mmu, u64 iova,
+		struct sg_table *sgt, size_t len, int prot)
+{
+	struct msm_iommu_pagetable *pagetable = to_pagetable(mmu);
+	struct io_pgtable_ops *ops = pagetable->pgtbl_ops;
+	struct scatterlist *sg;
+	size_t mapped = 0;
+	u64 addr = iova;
+	unsigned int i;
+
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		size_t size = sg->length;
+		phys_addr_t phys = sg_phys(sg);
+
+		/* Map the block one page at a time */
+		while (size) {
+			if (ops->map(ops, addr, phys, 4096, prot, GFP_KERNEL)) {
+				msm_iommu_pagetable_unmap(mmu, iova, mapped);
+				return -EINVAL;
+			}
+
+			phys += 4096;
+			addr += 4096;
+			size -= 4096;
+			mapped += 4096;
+		}
+	}
+
+	return 0;
+}
+
+static void msm_iommu_pagetable_destroy(struct msm_mmu *mmu)
+{
+	struct msm_iommu_pagetable *pagetable = to_pagetable(mmu);
+	struct msm_iommu *iommu = to_msm_iommu(pagetable->parent);
+
+	/*
+	 * If this is the last attached pagetable for the parent,
+	 * disable TTBR0 in the arm-smmu driver
+	 */
+	if (atomic_dec_return(&iommu->pagetables) == 0)
+		iommu_domain_set_attr(iommu->domain,
+			DOMAIN_ATTR_PGTABLE_CFG, NULL);
+
+	free_io_pgtable_ops(pagetable->pgtbl_ops);
+	kfree(pagetable);
+}
+
+int msm_iommu_pagetable_params(struct msm_mmu *mmu,
+		phys_addr_t *ttbr, int *asid)
+{
+	struct msm_iommu_pagetable *pagetable;
+
+	if (mmu->type != MSM_MMU_IOMMU_PAGETABLE)
+		return -EINVAL;
+
+	pagetable = to_pagetable(mmu);
+
+	if (ttbr)
+		*ttbr = pagetable->ttbr;
+
+	if (asid)
+		*asid = pagetable->asid;
+
+	return 0;
+}
+
+static const struct msm_mmu_funcs pagetable_funcs = {
+		.map = msm_iommu_pagetable_map,
+		.unmap = msm_iommu_pagetable_unmap,
+		.destroy = msm_iommu_pagetable_destroy,
+};
+
+static void msm_iommu_tlb_flush_all(void *cookie)
+{
+}
+
+static void msm_iommu_tlb_flush_walk(unsigned long iova, size_t size,
+		size_t granule, void *cookie)
+{
+}
+
+static void msm_iommu_tlb_add_page(struct iommu_iotlb_gather *gather,
+		unsigned long iova, size_t granule, void *cookie)
+{
+}
+
+static const struct iommu_flush_ops null_tlb_ops = {
+	.tlb_flush_all = msm_iommu_tlb_flush_all,
+	.tlb_flush_walk = msm_iommu_tlb_flush_walk,
+	.tlb_flush_leaf = msm_iommu_tlb_flush_walk,
+	.tlb_add_page = msm_iommu_tlb_add_page,
+};
+
+struct msm_mmu *msm_iommu_pagetable_create(struct msm_mmu *parent)
+{
+	struct msm_iommu *iommu = to_msm_iommu(parent);
+	static int next_asid = 16;
+	struct msm_iommu_pagetable *pagetable;
+	struct io_pgtable_cfg cfg;
+	int ret;
+
+	/* Get the pagetable configuration from the domain */
+	ret = iommu_domain_get_attr(iommu->domain,
+		DOMAIN_ATTR_PGTABLE_CFG, &cfg);
+	if (ret)
+		return ERR_PTR(ret);
+
+	pagetable = kzalloc(sizeof(*pagetable), GFP_KERNEL);
+	if (!pagetable)
+		return ERR_PTR(-ENOMEM);
+
+	msm_mmu_init(&pagetable->base, parent->dev, &pagetable_funcs,
+		MSM_MMU_IOMMU_PAGETABLE);
+
+	/* The incoming cfg will have the TTBR1 quirk enabled */
+	cfg.quirks &= ~IO_PGTABLE_QUIRK_ARM_TTBR1;
+	cfg.tlb = &null_tlb_ops;
+
+	pagetable->pgtbl_ops = alloc_io_pgtable_ops(ARM_64_LPAE_S1,
+		&cfg, iommu->domain);
+
+	if (!pagetable->pgtbl_ops) {
+		kfree(pagetable);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	/*
+	 * If this is the first pagetable that we've allocated, send it back to
+	 * the arm-smmu driver as a trigger to set up TTBR0
+	 */
+	if (atomic_inc_return(&iommu->pagetables) == 1) {
+		ret = iommu_domain_set_attr(iommu->domain,
+			DOMAIN_ATTR_PGTABLE_CFG, &cfg);
+		if (ret) {
+			free_io_pgtable_ops(pagetable->pgtbl_ops);
+			kfree(pagetable);
+			return ERR_PTR(ret);
+		}
+	}
+
+	/* Needed later for TLB flush */
+	pagetable->parent = parent;
+	pagetable->ttbr = cfg.arm_lpae_s1_cfg.ttbr;
+
+	pagetable->asid = next_asid;
+	next_asid = (next_asid + 1) % 255;
+	next_asid = min(16, next_asid);
+
+	return &pagetable->base;
+}
+
 static int msm_fault_handler(struct iommu_domain *domain, struct device *dev,
 		unsigned long iova, int flags, void *arg)
 {
@@ -85,9 +271,11 @@ struct msm_mmu *msm_iommu_new(struct device *dev, struct iommu_domain *domain)
 		return ERR_PTR(-ENOMEM);
 
 	iommu->domain = domain;
-	msm_mmu_init(&iommu->base, dev, &funcs);
+	msm_mmu_init(&iommu->base, dev, &funcs, MSM_MMU_IOMMU);
 	iommu_set_fault_handler(domain, msm_fault_handler, iommu);
 
+	atomic_set(&iommu->pagetables, 0);
+
 	ret = iommu_attach_device(iommu->domain, dev);
 	if (ret) {
 		kfree(iommu);
drivers/gpu/drm/msm/msm_mmu.h +15 -1
index 3a534ee59bf6..61ade89d9e48 100644
@@ -17,18 +17,26 @@ struct msm_mmu_funcs {
 	void (*destroy)(struct msm_mmu *mmu);
 };
 
+enum msm_mmu_type {
+	MSM_MMU_GPUMMU,
+	MSM_MMU_IOMMU,
+	MSM_MMU_IOMMU_PAGETABLE,
+};
+
 struct msm_mmu {
 	const struct msm_mmu_funcs *funcs;
 	struct device *dev;
 	int (*handler)(void *arg, unsigned long iova, int flags);
 	void *arg;
+	enum msm_mmu_type type;
 };
 
 static inline void msm_mmu_init(struct msm_mmu *mmu, struct device *dev,
-		const struct msm_mmu_funcs *funcs)
+		const struct msm_mmu_funcs *funcs, enum msm_mmu_type type)
 {
 	mmu->dev = dev;
 	mmu->funcs = funcs;
+	mmu->type = type;
 }
 
 struct msm_mmu *msm_iommu_new(struct device *dev, struct iommu_domain *domain);
@@ -41,7 +49,13 @@ static inline void msm_mmu_set_fault_handler(struct msm_mmu *mmu, void *arg,
 	mmu->handler = handler;
 }
 
+struct msm_mmu *msm_iommu_pagetable_create(struct msm_mmu *parent);
+
 void msm_gpummu_params(struct msm_mmu *mmu, dma_addr_t *pt_base,
 		dma_addr_t *tran_error);
 
+
+int msm_iommu_pagetable_params(struct msm_mmu *mmu, phys_addr_t *ttbr,
+		int *asid);
+
 #endif /* __MSM_MMU_H__ */
#18 lore raw

PATCH v12 10/13 drm/msm: Add support for private address space instances

In reply to Jordan Crouse

Add support for allocating private address space instances. Targets that support per-context pagetables should implement their own function to allocate private address spaces.

The default will return a pointer to the global address space.

---

 drivers/gpu/drm/msm/msm_drv.c     | 13 +++++++------
 drivers/gpu/drm/msm/msm_drv.h     |  5 +++++
 drivers/gpu/drm/msm/msm_gem_vma.c |  9 +++++++++
 drivers/gpu/drm/msm/msm_gpu.c     | 22 ++++++++++++++++++++++
 drivers/gpu/drm/msm/msm_gpu.h     |  5 +++++
 5 files changed, 48 insertions(+), 6 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +7 -6
index 108b663c3ef2..f072306f1260 100644
@@ -597,7 +597,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
-	ctx->aspace = priv->gpu ? priv->gpu->aspace : NULL;
+	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu);
 	file->driver_priv = ctx;
 
 	return 0;
@@ -780,18 +780,19 @@ static int msm_ioctl_gem_cpu_fini(struct drm_device *dev, void *data,
 }
 
 static int msm_ioctl_gem_info_iova(struct drm_device *dev,
-		struct drm_gem_object *obj, uint64_t *iova)
+		struct drm_file *file, struct drm_gem_object *obj,
+		uint64_t *iova)
 {
-	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_file_private *ctx = file->driver_priv;
 
-	if (!priv->gpu)
+	if (!ctx->aspace)
 		return -EINVAL;
 
 	/*
 	 * Don't pin the memory here - just get an address so that userspace can
 	 * be productive
 	 */
-	return msm_gem_get_iova(obj, priv->gpu->aspace, iova);
+	return msm_gem_get_iova(obj, ctx->aspace, iova);
 }
 
 static int msm_ioctl_gem_info(struct drm_device *dev, void *data,
@@ -830,7 +831,7 @@ static int msm_ioctl_gem_info(struct drm_device *dev, void *data,
 		args->value = msm_gem_mmap_offset(obj);
 		break;
 	case MSM_INFO_GET_IOVA:
-		ret = msm_ioctl_gem_info_iova(dev, obj, &args->value);
+		ret = msm_ioctl_gem_info_iova(dev, file, obj, &args->value);
 		break;
 	case MSM_INFO_SET_NAME:
 		/* length check should leave room for terminating null: */
drivers/gpu/drm/msm/msm_drv.h +5 -0
index f69c6d62584d..51a5c9083e13 100644
@@ -249,6 +249,10 @@ int msm_gem_map_vma(struct msm_gem_address_space *aspace,
 void msm_gem_close_vma(struct msm_gem_address_space *aspace,
 		struct msm_gem_vma *vma);
 
+
+struct msm_gem_address_space *
+msm_gem_address_space_get(struct msm_gem_address_space *aspace);
+
 void msm_gem_address_space_put(struct msm_gem_address_space *aspace);
 
 struct msm_gem_address_space *
@@ -434,6 +438,7 @@ static inline void msm_file_private_destroy(struct kref *kref)
 	struct msm_file_private *ctx = container_of(kref,
 		struct msm_file_private, ref);
 
+	msm_gem_address_space_put(ctx->aspace);
 	kfree(ctx);
 }
 
drivers/gpu/drm/msm/msm_gem_vma.c +9 -0
index 5f6a11211b64..29cc1305cf37 100644
@@ -27,6 +27,15 @@ void msm_gem_address_space_put(struct msm_gem_address_space *aspace)
 		kref_put(&aspace->kref, msm_gem_address_space_destroy);
 }
 
+struct msm_gem_address_space *
+msm_gem_address_space_get(struct msm_gem_address_space *aspace)
+{
+	if (!IS_ERR_OR_NULL(aspace))
+		kref_get(&aspace->kref);
+
+	return aspace;
+}
+
 /* Actually unmap memory for the vma */
 void msm_gem_purge_vma(struct msm_gem_address_space *aspace,
 		struct msm_gem_vma *vma)
drivers/gpu/drm/msm/msm_gpu.c +22 -0
index a1f3da6550e5..b070355369d8 100644
@@ -823,6 +823,28 @@ static int get_clocks(struct platform_device *pdev, struct msm_gpu *gpu)
 	return 0;
 }
 
+/* Return a new address space for a msm_drm_private instance */
+struct msm_gem_address_space *
+msm_gpu_create_private_address_space(struct msm_gpu *gpu)
+{
+	struct msm_gem_address_space *aspace = NULL;
+
+	if (!gpu)
+		return NULL;
+
+	/*
+	 * If the target doesn't support private address spaces then return
+	 * the global one
+	 */
+	if (gpu->funcs->create_private_address_space)
+		aspace = gpu->funcs->create_private_address_space(gpu);
+
+	if (IS_ERR_OR_NULL(aspace))
+		aspace = msm_gem_address_space_get(gpu->aspace);
+
+	return aspace;
+}
+
 int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, struct msm_gpu_config *config)
drivers/gpu/drm/msm/msm_gpu.h +5 -0
index d496d488222c..d298657b4730 100644
@@ -64,6 +64,8 @@ struct msm_gpu_funcs {
 	void (*gpu_set_freq)(struct msm_gpu *gpu, struct dev_pm_opp *opp);
 	struct msm_gem_address_space *(*create_address_space)
 		(struct msm_gpu *gpu, struct platform_device *pdev);
+	struct msm_gem_address_space *(*create_private_address_space)
+		(struct msm_gpu *gpu);
 };
 
 struct msm_gpu {
@@ -286,6 +288,9 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, struct msm_gpu_config *config);
 
+struct msm_gem_address_space *
+msm_gpu_create_private_address_space(struct msm_gpu *gpu);
+
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
 struct msm_gpu *adreno_load_gpu(struct drm_device *dev);
Signature · 1 line
2.25.1
#19 lore raw

PATCH v12 11/13 drm/msm/a6xx: Add support for per-instance pagetables

In reply to Jordan Crouse

Add support for using per-instance pagetables if all the dependencies are available.

---

 drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 70 +++++++++++++++++++++++++++
 drivers/gpu/drm/msm/adreno/a6xx_gpu.h |  1 +
 drivers/gpu/drm/msm/msm_ringbuffer.h  |  1 +
 3 files changed, 72 insertions(+)
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +70 -0
index 5eabb0109577..9653ac9b3cb8 100644
@@ -81,6 +81,56 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
 	OUT_RING(ring, upper_32_bits(iova));
 }
 
+static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
+		struct msm_ringbuffer *ring, struct msm_file_private *ctx)
+{
+	phys_addr_t ttbr;
+	u32 asid;
+	u64 memptr = rbmemptr(ring, ttbr0);
+
+	if (ctx == a6xx_gpu->cur_ctx)
+		return;
+
+	if (msm_iommu_pagetable_params(ctx->aspace->mmu, &ttbr, &asid))
+		return;
+
+	/* Execute the table update */
+	OUT_PKT7(ring, CP_SMMU_TABLE_UPDATE, 4);
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_0_TTBR0_LO(lower_32_bits(ttbr)));
+
+	/*
+	 * For now ignore the asid since the smmu driver uses a TLBIASID to
+	 * flush the TLB when we use iommu_flush_tlb_all() and the smmu driver
+	 * isn't aware that the asid changed.  Instead, keep the default asid
+	 * (0, same as the context bank) to make sure the TLB is properly
+	 * flushed.
+	 */
+	OUT_RING(ring,
+		CP_SMMU_TABLE_UPDATE_1_TTBR0_HI(upper_32_bits(ttbr)) |
+		CP_SMMU_TABLE_UPDATE_1_ASID(0));
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_2_CONTEXTIDR(0));
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_3_CONTEXTBANK(0));
+
+	/*
+	 * Write the new TTBR0 to the memstore. This is good for debugging.
+	 */
+	OUT_PKT7(ring, CP_MEM_WRITE, 4);
+	OUT_RING(ring, CP_MEM_WRITE_0_ADDR_LO(lower_32_bits(memptr)));
+	OUT_RING(ring, CP_MEM_WRITE_1_ADDR_HI(upper_32_bits(memptr)));
+	OUT_RING(ring, lower_32_bits(ttbr));
+	OUT_RING(ring, (0 << 16) | upper_32_bits(ttbr));
+
+	/*
+	 * And finally, trigger a uche flush to be sure there isn't anything
+	 * lingering in that part of the GPU
+	 */
+
+	OUT_PKT7(ring, CP_EVENT_WRITE, 1);
+	OUT_RING(ring, 0x31);
+
+	a6xx_gpu->cur_ctx = ctx;
+}
+
 static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
@@ -90,6 +140,8 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 	struct msm_ringbuffer *ring = submit->ring;
 	unsigned int i;
 
+	a6xx_set_pagetable(a6xx_gpu, ring, submit->queue->ctx);
+
 	get_stats_counter(ring, REG_A6XX_RBBM_PERFCTR_CP_0_LO,
 		rbmemptr_stats(ring, index, cpcycles_start));
 
@@ -696,6 +748,8 @@ static int a6xx_hw_init(struct msm_gpu *gpu)
 	/* Always come up on rb 0 */
 	a6xx_gpu->cur_ring = gpu->rb[0];
 
+	a6xx_gpu->cur_ctx = NULL;
+
 	/* Enable the SQE_to start the CP engine */
 	gpu_write(gpu, REG_A6XX_CP_SQE_CNTL, 1);
 
@@ -1008,6 +1062,21 @@ static unsigned long a6xx_gpu_busy(struct msm_gpu *gpu)
 	return (unsigned long)busy_time;
 }
 
+static struct msm_gem_address_space *
+a6xx_create_private_address_space(struct msm_gpu *gpu)
+{
+	struct msm_gem_address_space *aspace = NULL;
+	struct msm_mmu *mmu;
+
+	mmu = msm_iommu_pagetable_create(gpu->aspace->mmu);
+
+	if (!IS_ERR(mmu))
+		aspace = msm_gem_address_space_create(mmu,
+			"gpu", 0x100000000ULL, 0x1ffffffffULL);
+
+	return aspace;
+}
+
 static const struct adreno_gpu_funcs funcs = {
 	.base = {
 		.get_param = adreno_get_param,
@@ -1031,6 +1100,7 @@ static const struct adreno_gpu_funcs funcs = {
 		.gpu_state_put = a6xx_gpu_state_put,
 #endif
 		.create_address_space = adreno_iommu_create_address_space,
+		.create_private_address_space = a6xx_create_private_address_space,
 	},
 	.get_timestamp = a6xx_get_timestamp,
 };
drivers/gpu/drm/msm/adreno/a6xx_gpu.h +1 -0
index 03ba60d5b07f..da22d7549d9b 100644
@@ -19,6 +19,7 @@ struct a6xx_gpu {
 	uint64_t sqe_iova;
 
 	struct msm_ringbuffer *cur_ring;
+	struct msm_file_private *cur_ctx;
 
 	struct a6xx_gmu gmu;
 };
drivers/gpu/drm/msm/msm_ringbuffer.h +1 -0
index 7764373d0ed2..0987d6bf848c 100644
@@ -31,6 +31,7 @@ struct msm_rbmemptrs {
 	volatile uint32_t fence;
 
 	volatile struct msm_gpu_submit_stats stats[MSM_GPU_SUBMIT_STATS_COUNT];
+	volatile u64 ttbr0;
 };
 
 struct msm_ringbuffer {
Signature · 1 line
2.25.1
#20 lore raw

PATCH v12 12/13 arm: dts: qcom: sm845: Set the compatible string for the GPU SMMU

In reply to Jordan Crouse

Set the qcom,adreno-smmu compatible string for the GPU SMMU to enable split pagetables and per-instance pagetables for drm/msm.

---

 arch/arm64/boot/dts/qcom/sdm845.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/boot/dts/qcom/sdm845.dtsi +1 -1
index 2884577dcb77..6a9adaa401a9 100644
@@ -4058,7 +4058,7 @@ opp-257000000 {
 		};
 
 		adreno_smmu: iommu@5040000 {
-			compatible = "qcom,sdm845-smmu-v2", "qcom,smmu-v2";
+			compatible = "qcom,adreno-smmu", "qcom,smmu-v2";
 			reg = <0 0x5040000 0 0x10000>;
 			#iommu-cells = <1>;
 			#global-interrupts = <2>;
Signature · 6 lines
2.25.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#21 lore raw

RFC v12 13/13 iommu/arm-smmu: Add a init_context_bank implementation hook

In reply to Jordan Crouse

Add a new implementation hook to allow the implementation specific code to tweek the context bank configuration just before it gets written. The first user will be the Adreno GPU implementation to turn on SCTLR.HUPCF to ensure that a page fault doesn't terminating pending transactions. Doing so could hang the GPU if one of the terminated transactions is a CP read.

---

 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 16 ++++++++++++++++
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 21 +++++++++++++--------
 drivers/iommu/arm/arm-smmu/arm-smmu.h      |  5 +++++
 3 files changed, 34 insertions(+), 8 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +16 -0
index 3be10145bf57..baa026ddca1c 100644
@@ -34,6 +34,19 @@ static bool qcom_adreno_smmu_is_gpu_device(struct device *dev)
 	return false;
 }
 
+#define QCOM_ADRENO_SMMU_GPU 1
+
+static void qcom_adreno_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain,
+		struct arm_smmu_cb *cb)
+{
+	/*
+	 * On the GPU device we want to process subsequent transactions after a
+	 * fault to keep the GPU from hanging
+	 */
+	if (cb->cfg->priv == QCOM_ADRENO_SMMU_GPU)
+		cb->sctlr |= ARM_SMMU_SCTLR_HUPCF;
+}
+
 /*
  * Local implementation to configure TTBR0 with the specified pagetable config.
  * The GPU driver will call this to enable TTBR0 when per-instance pagetables
@@ -120,6 +133,7 @@ static int qcom_adreno_smmu_alloc_context_bank(struct arm_smmu_domain *smmu_doma
 		count = 1;
 	} else {
 		start = 1;
+		smmu_domain->cfg.priv = QCOM_ADRENO_SMMU_GPU;
 	}
 
 	return __arm_smmu_alloc_bitmap(smmu->context_map, start, count);
@@ -141,6 +155,7 @@ static int qcom_adreno_smmu_init_context(struct arm_smmu_domain *smmu_domain,
 	    (smmu_domain->cfg.fmt == ARM_SMMU_CTX_FMT_AARCH64))
 		pgtbl_cfg->quirks |= IO_PGTABLE_QUIRK_ARM_TTBR1;
 
+
 	return 0;
 }
 
@@ -204,6 +219,7 @@ static const struct arm_smmu_impl qcom_adreno_smmu_impl = {
 	.alloc_context_bank = qcom_adreno_smmu_alloc_context_bank,
 	.domain_set_attr = qcom_adreno_smmu_domain_set_attr,
 	.domain_get_attr = qcom_adreno_smmu_domain_get_attr,
+	.init_context_bank = qcom_adreno_smmu_init_context_bank,
 };
 
 static struct arm_smmu_device *qcom_smmu_create(struct arm_smmu_device *smmu,
drivers/iommu/arm/arm-smmu/arm-smmu.c +13 -8
index e0a3e0da885b..6225649bbfef 100644
@@ -532,6 +532,18 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain,
 			cb->mair[1] = pgtbl_cfg->arm_lpae_s1_cfg.mair >> 32;
 		}
 	}
+
+	cb->sctlr = ARM_SMMU_SCTLR_CFIE | ARM_SMMU_SCTLR_CFRE | ARM_SMMU_SCTLR_AFE |
+		ARM_SMMU_SCTLR_TRE | ARM_SMMU_SCTLR_M;
+
+	if (stage1)
+		cb->sctlr |= ARM_SMMU_SCTLR_S1_ASIDPNE;
+	if (IS_ENABLED(CONFIG_CPU_BIG_ENDIAN))
+		cb->sctlr |= ARM_SMMU_SCTLR_E;
+
+	/* Give the implementation a chance to adjust the configuration */
+	if (smmu_domain->smmu->impl && smmu_domain->smmu->impl->init_context_bank)
+		smmu_domain->smmu->impl->init_context_bank(smmu_domain, cb);
 }
 
 void arm_smmu_write_context_bank(struct arm_smmu_device *smmu, int idx)
@@ -610,14 +622,7 @@ void arm_smmu_write_context_bank(struct arm_smmu_device *smmu, int idx)
 	}
 
 	/* SCTLR */
-	reg = ARM_SMMU_SCTLR_CFIE | ARM_SMMU_SCTLR_CFRE | ARM_SMMU_SCTLR_AFE |
-	      ARM_SMMU_SCTLR_TRE | ARM_SMMU_SCTLR_M;
-	if (stage1)
-		reg |= ARM_SMMU_SCTLR_S1_ASIDPNE;
-	if (IS_ENABLED(CONFIG_CPU_BIG_ENDIAN))
-		reg |= ARM_SMMU_SCTLR_E;
-
-	arm_smmu_cb_write(smmu, idx, ARM_SMMU_CB_SCTLR, reg);
+	arm_smmu_cb_write(smmu, idx, ARM_SMMU_CB_SCTLR, cb->sctlr);
 }
 
 static int arm_smmu_init_domain_context(struct iommu_domain *domain,
drivers/iommu/arm/arm-smmu/arm-smmu.h +5 -0
index 870f0fd060a5..e84b8da8b93b 100644
@@ -143,6 +143,7 @@ enum arm_smmu_cbar_type {
 
 #define ARM_SMMU_CB_SCTLR		0x0
 #define ARM_SMMU_SCTLR_S1_ASIDPNE	BIT(12)
+#define ARM_SMMU_SCTLR_HUPCF		BIT(8)
 #define ARM_SMMU_SCTLR_CFCFG		BIT(7)
 #define ARM_SMMU_SCTLR_CFIE		BIT(6)
 #define ARM_SMMU_SCTLR_CFRE		BIT(5)
@@ -343,6 +344,7 @@ struct arm_smmu_cfg {
 	};
 	enum arm_smmu_cbar_type		cbar;
 	enum arm_smmu_context_fmt	fmt;
+	unsigned long			priv;
 };
 #define ARM_SMMU_INVALID_IRPTNDX	0xff
 
@@ -350,6 +352,7 @@ struct arm_smmu_cb {
 	u64				ttbr[2];
 	u32				tcr[2];
 	u32				mair[2];
+	u32				sctlr;
 	struct arm_smmu_cfg		*cfg;
 };
 
@@ -439,6 +442,8 @@ struct arm_smmu_impl {
 			enum iommu_attr attr, void *data);
 	int (*domain_set_attr)(struct arm_smmu_domain *smmu_domain,
 			enum iommu_attr attr, void *data);
+	void (*init_context_bank)(struct arm_smmu_domain *smmu_domain,
+			struct arm_smmu_cb *cb);
 };
 
 #define INVALID_SMENDX			-1
Signature · 1 line
2.25.1
maintainer
#22 lore raw

Re: [RFC v12 13/13] iommu/arm-smmu: Add a init_context_bank implementation hook

Quoted 7 lines · Jordan Crouse

Add a new implementation hook to allow the implementation specific code to tweek the context bank configuration just before it gets written. The first user will be the Adreno GPU implementation to turn on SCTLR.HUPCF to ensure that a page fault doesn't terminating pending transactions. Doing so could hang the GPU if one of the terminated transactions is a CP read.

Quoted 7 lines

---

 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 16 ++++++++++++++++
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 21 +++++++++++++--------
 drivers/iommu/arm/arm-smmu/arm-smmu.h      |  5 +++++
 3 files changed, 34 insertions(+), 8 deletions(-)

We already have ->init_context(), so I'd prefer to use that instead of adding another callback. Could we stick a couple of fields in smmu_domain->cfg (e.g. sctlr_set, sctlr_clr) and handle those a bit like we do for the asid/vmid on cavium implementations?

Will

maintainer
#23 lore raw
In reply to Jordan Crouse
Quoted 7 lines · Jordan Crouse

This series adds an Adreno SMMU implementation to arm-smmu to allow GPU hardware pagetable switching.

The Adreno GPU has built in capabilities to switch the TTBR0 pagetable during runtime to allow each individual instance or application to have its own pagetable. In order to take advantage of the HW capabilities there are certain requirements needed of the SMMU hardware.

"capabilities" is a polite way of putting it ;)

Anyway, modulo two design comments, I think this is about as nice as we're going to get this. Thanks for persevering, and sorry that you have to deal with such dreadful hardware.

Hopefully the next version will be the one, although I'd like Robin to take a quick look as well if he gets a chance.

Will
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#24 lore raw

cover iommu/arm-smmu + drm/msm: per-process GPU pgtables

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

NOTE: Since Jordan was out today, and I wanted to keep things moving on
this, I took the liberty of respinning his series (originally titled "iommu/arm-smmu: Add Adreno SMMU specific implementation") with updates based on Will's review comments, and some fixes and extra bits that I found in stress testing the series. Original commit msg and updated version history below.

In general I like the private interface between adreno-smmu and the GPU driver. It should make for a more straightforward way to extend things to optimize TLB invalidation in the future, for example, rather than shoe-horning everything thru domain attrs. And it lets us describe the get_ttbr1_cfg/set_ttrb0_cfg dance more clearly. Although it is going to make landing this via iommu vs drm tree a bit more difficult. Maybe there are some arm-smmu parts of this series that could be pulled out to make it not conflicty to land the private interface and adreno-smmu bits via the drm tree? (But I'm jumping a bit ahead here. Just wanted to point out that issue.)

The complete series can be found at:
https://gitlab.freedesktop.org/drm/msm/-/commits/msm-next-pgtables

This series adds an Adreno SMMU implementation to arm-smmu to allow GPU hardware pagetable switching.

The Adreno GPU has built in capabilities to switch the TTBR0 pagetable during runtime to allow each individual instance or application to have its own pagetable. In order to take advantage of the HW capabilities there are certain requirements needed of the SMMU hardware.

This series adds support for an Adreno specific arm-smmu implementation. The new implementation 1) ensures that the GPU domain is always assigned context bank 0,
2) enables split pagetable support (TTBR1) so that the instance specific pagetable can be swapped while the global memory remains in place and 3) shares the current pagetable configuration with the GPU driver to allow it to create its own io-pgtable instances.

The series then adds the drm/msm code to enable these features. For targets that support it allocate new pagetables using the io-pgtable configuration shared by the arm-smmu driver and swap them in during runtime.

This version of the series merges the previous patchset(s) [1] and [2] with the following improvements:

v13: (Respin by Rob)
- Switch to a private interface between adreno-smmu and GPU driver, dropping the custom domain attr (Will Deacon)
- Rework the SCTLR.HUPCF patch to add new fields in smmu_domain->cfg rather than adding new impl hook (Will Deacon)
- Drop for_each_cfg_sme() in favor of plain for() loop (Will Deacon)
- Fix context refcnt'ing issue which was causing problems with GPU crash recover stress testing.
- Spiff up $debugfs/gem to show process information associated with VMAs
v12:
- Nitpick cleanups in gpu/drm/msm/msm_iommu.c (Rob Clark)
- Reorg in gpu/drm/msm/msm_gpu.c (Rob Clark)
- Use the default asid for the context bank so that iommu_tlb_flush_all works
- Flush the UCHE after a page switch
- Add the SCTLR.HUPCF patch at the end of the series
v11:
- Add implementation specific get_attr/set_attr functions (per Rob Clark)
- Fix context bank allocation (per Bjorn Andersson)
v10:
- arm-smmu: add implementation hook to allocate context banks
- arm-smmu: Match the GPU domain by stream ID instead of compatible string
- arm-smmu: Make DOMAIN_ATTR_PGTABLE_CFG bi-directional. The leaf driver queries the configuration to create a pagetable and then sends the newly created configuration back to the smmu-driver to enable TTBR0
- drm/msm: Add context reference counting for submissions
- drm/msm: Use dummy functions to skip TLB operations on per-instance pagetables

[1] https://lists.linuxfoundation.org/pipermail/iommu/2020-June/045653.html
[2] https://lists.linuxfoundation.org/pipermail/iommu/2020-June/045659.html

Jordan Crouse (12):
iommu/arm-smmu: Pass io-pgtable config to implementation specific
function
iommu/arm-smmu: Add support for split pagetables iommu/arm-smmu: Prepare for the adreno-smmu implementation iommu/arm-smmu-qcom: Add implementation for the adreno GPU SMMU
dt-bindings: arm-smmu: Add compatible string for Adreno GPU SMMU drm/msm: Add a context pointer to the submitqueue drm/msm: Drop context arg to gpu->submit()
drm/msm: Set the global virtual address range from the IOMMU domain drm/msm: Add support to create a local pagetable drm/msm: Add support for private address space instances drm/msm/a6xx: Add support for per-instance pagetables
arm: dts: qcom: sm845: Set the compatible string for the GPU SMMU

Rob Clark (7):
drm/msm: remove dangling submitqueue references
iommu: add private interface for adreno-smmu drm/msm/gpu: add dev_to_gpu() helper
drm/msm: set adreno_smmu as gpu's drvdata iommu/arm-smmu: constify some helpers
iommu/arm-smmu: add a way for implementations to influence SCTLR drm/msm: show process names in gem_describe

 .../devicetree/bindings/iommu/arm,smmu.yaml   |   4 +
 arch/arm64/boot/dts/qcom/sdm845.dtsi          |   2 +-
 drivers/gpu/drm/msm/adreno/a5xx_gpu.c         |  12 +-
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c         |  75 ++++++-
 drivers/gpu/drm/msm/adreno/a6xx_gpu.h         |   1 +
 drivers/gpu/drm/msm/adreno/adreno_device.c    |  12 +-
 drivers/gpu/drm/msm/adreno/adreno_gpu.c       |  18 +-
 drivers/gpu/drm/msm/adreno/adreno_gpu.h       |   3 +-
 drivers/gpu/drm/msm/msm_drv.c                 |  16 +-
 drivers/gpu/drm/msm/msm_drv.h                 |  25 +++
 drivers/gpu/drm/msm/msm_gem.c                 |  25 ++-
 drivers/gpu/drm/msm/msm_gem.h                 |   6 +
 drivers/gpu/drm/msm/msm_gem_submit.c          |   8 +-
 drivers/gpu/drm/msm/msm_gem_vma.c             |  10 +
 drivers/gpu/drm/msm/msm_gpu.c                 |  41 +++-
 drivers/gpu/drm/msm/msm_gpu.h                 |  21 +-
 drivers/gpu/drm/msm/msm_gpummu.c              |   2 +-
 drivers/gpu/drm/msm/msm_iommu.c               | 206 +++++++++++++++++-
 drivers/gpu/drm/msm/msm_mmu.h                 |  16 +-
 drivers/gpu/drm/msm/msm_ringbuffer.h          |   1 +
 drivers/gpu/drm/msm/msm_submitqueue.c         |   7 +-
 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c    |   6 +-
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c    | 155 ++++++++++++-
 drivers/iommu/arm/arm-smmu/arm-smmu.c         | 102 ++++-----
 drivers/iommu/arm/arm-smmu/arm-smmu.h         |  87 +++++++-
 include/linux/adreno-smmu-priv.h              |  36 +++
 26 files changed, 765 insertions(+), 132 deletions(-)
 create mode 100644 include/linux/adreno-smmu-priv.h
Signature · 1 line
2.26.2
#25 lore raw

Re: [PATCH 00/19] iommu/arm-smmu + drm/msm: per-process GPU pgtables

Quoted 46 lines · Rob Clark

From: Rob Clark <[email protected]>

NOTE: Since Jordan was out today, and I wanted to keep things moving on
this, I took the liberty of respinning his series (originally titled "iommu/arm-smmu: Add Adreno SMMU specific implementation") with updates based on Will's review comments, and some fixes and extra bits that I found in stress testing the series. Original commit msg and updated version history below.

In general I like the private interface between adreno-smmu and the GPU driver. It should make for a more straightforward way to extend things to optimize TLB invalidation in the future, for example, rather than shoe-horning everything thru domain attrs. And it lets us describe the get_ttbr1_cfg/set_ttrb0_cfg dance more clearly. Although it is going to make landing this via iommu vs drm tree a bit more difficult. Maybe there are some arm-smmu parts of this series that could be pulled out to make it not conflicty to land the private interface and adreno-smmu bits via the drm tree? (But I'm jumping a bit ahead here. Just wanted to point out that issue.)

The complete series can be found at:
https://gitlab.freedesktop.org/drm/msm/-/commits/msm-next-pgtables

This series adds an Adreno SMMU implementation to arm-smmu to allow GPU hardware pagetable switching.

The Adreno GPU has built in capabilities to switch the TTBR0 pagetable during runtime to allow each individual instance or application to have its own pagetable. In order to take advantage of the HW capabilities there are certain requirements needed of the SMMU hardware.

This series adds support for an Adreno specific arm-smmu implementation. The new implementation 1) ensures that the GPU domain is always assigned context bank 0,
2) enables split pagetable support (TTBR1) so that the instance specific pagetable can be swapped while the global memory remains in place and 3) shares the current pagetable configuration with the GPU driver to allow it to create its own io-pgtable instances.

The series then adds the drm/msm code to enable these features. For targets that support it allocate new pagetables using the io-pgtable configuration shared by the arm-smmu driver and swap them in during runtime.

This version of the series merges the previous patchset(s) [1] and [2] with the following improvements:

Quoted 10 lines

v13: (Respin by Rob)
- Switch to a private interface between adreno-smmu and GPU driver, dropping the custom domain attr (Will Deacon)
- Rework the SCTLR.HUPCF patch to add new fields in smmu_domain->cfg rather than adding new impl hook (Will Deacon)
- Drop for_each_cfg_sme() in favor of plain for() loop (Will Deacon)
- Fix context refcnt'ing issue which was causing problems with GPU crash recover stress testing.
- Spiff up $debugfs/gem to show process information associated with VMAs

I'll add the tags to Rob's code but in general I ack all these changes. I also like the private interface - it gives us the most flexibility without either changing the IOMMU API or giving up entirely and making an internal SMMU implementation inside drm/msm.

<snip>

Jordan

Signature · 6 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#26 lore raw

PATCH 1/19 drm/msm: remove dangling submitqueue references

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

Currently it doesn't matter, since we free the ctx immediately. But when we start refcnt'ing the ctx, we don't want old dangling list entries to hang around.

---
 drivers/gpu/drm/msm/msm_submitqueue.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/msm/msm_submitqueue.c +3 -1
index a1d94be7883a..90c9d84e6155 100644
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
 	 * No lock needed in close and there won't
 	 * be any more user ioctls coming our way
 	 */
-	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+		list_del(&entry->node);
 		msm_submitqueue_put(entry);
+	}
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
Signature · 1 line
2.26.2
author Reviewed-by on v1 1/19
#27 lore raw

Re: [PATCH 01/19] drm/msm: remove dangling submitqueue references

Quoted 5 lines · Rob Clark

From: Rob Clark <[email protected]>

Currently it doesn't matter, since we free the ctx immediately. But when we start refcnt'ing the ctx, we don't want old dangling list entries to hang around.

Quoted 28 lines
---
 drivers/gpu/drm/msm/msm_submitqueue.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/msm/msm_submitqueue.c +3 -1
index a1d94be7883a..90c9d84e6155 100644
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
 	 * No lock needed in close and there won't
 	 * be any more user ioctls coming our way
 	 */
-	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+		list_del(&entry->node);
 		msm_submitqueue_put(entry);
+	}
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
Signature · 6 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#28 lore raw

Re: [PATCH 01/19] drm/msm: remove dangling submitqueue references

In reply to Rob Clark
Quoted 16 lines · Rob Clark

From: Rob Clark <[email protected]>

Currently it doesn't matter, since we free the ctx immediately. But when we start refcnt'ing the ctx, we don't want old dangling list entries to hang around.

---
 drivers/gpu/drm/msm/msm_submitqueue.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/msm/msm_submitqueue.c +0 -0
index a1d94be7883a..90c9d84e6155 100644

drivers/gpu/drm/msm/msm_submitqueue.c void msm_submitqueue_close(struct msm_file_private *ctx) · line 53

Quoted 6 lines
drivers/gpu/drm/msm/msm_submitqueue.c +2 -1
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
 	 * No lock needed in close and there won't
 	 * be any more user ioctls coming our way
 	 */
-	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+		list_del(&entry->node);

If you refcount ctx, what does that do for the entries in the submit queue?

"entry" here is kref'ed, but you're popping it off the list regardless of the put ends up freeing the object or not - which afaict would mean leaking the object.

On the other hand, with the current implementation an object with higher refcount with adjacent objects of single refcount would end up with dangling pointers after the put. So in itself this change seems like a net gain, but I'm wondering about the plan described in the commit message.

Regards,
Bjorn

Quoted 8 lines
drivers/gpu/drm/msm/msm_submitqueue.c +1 -0
 		msm_submitqueue_put(entry);
+	}
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
Signature · 1 line
2.26.2
#29 lore raw

Re: [PATCH 01/19] drm/msm: remove dangling submitqueue references

Quoted 29 lines · Bjorn Andersson
Quoted 16 lines · Rob Clark

From: Rob Clark <[email protected]>

Currently it doesn't matter, since we free the ctx immediately. But when we start refcnt'ing the ctx, we don't want old dangling list entries to hang around.

---
 drivers/gpu/drm/msm/msm_submitqueue.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/msm/msm_submitqueue.c +0 -0
index a1d94be7883a..90c9d84e6155 100644

drivers/gpu/drm/msm/msm_submitqueue.c void msm_submitqueue_close(struct msm_file_private *ctx) · line 53

Quoted 6 lines
drivers/gpu/drm/msm/msm_submitqueue.c +2 -1
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
       * No lock needed in close and there won't
       * be any more user ioctls coming our way
       */
-     list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+     list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+             list_del(&entry->node);

If you refcount ctx, what does that do for the entries in the submit queue?

Quoted 4 lines

"entry" here is kref'ed, but you're popping it off the list regardless of the put ends up freeing the object or not - which afaict would mean leaking the object.

What ends up happening is the submit has reference to submit-queue, which has reference to the ctx.. the submitqueue could be alive still pending in-flight submits (in a later patch), but dead from the PoV of userspace interface.

We aren't relying (or at least aren't in the end, and I *think* I didn't miss anything in the middle) relying on ctx->submitqueues list to clean anything up in the end, just track what is still a valid submitqueue from userspace PoV

BR,
-R

Quoted 18 lines

On the other hand, with the current implementation an object with higher refcount with adjacent objects of single refcount would end up with dangling pointers after the put. So in itself this change seems like a net gain, but I'm wondering about the plan described in the commit message.

Regards,
Bjorn

Quoted 8 lines
drivers/gpu/drm/msm/msm_submitqueue.c +1 -1
              msm_submitqueue_put(entry);
+     }
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
--

2.26.2

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

#30 lore raw

Re: [PATCH 01/19] drm/msm: remove dangling submitqueue references

Quoted 41 lines · Rob Clark
Quoted 29 lines · Bjorn Andersson
Quoted 16 lines · Rob Clark

From: Rob Clark <[email protected]>

Currently it doesn't matter, since we free the ctx immediately. But when we start refcnt'ing the ctx, we don't want old dangling list entries to hang around.

---
 drivers/gpu/drm/msm/msm_submitqueue.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/msm/msm_submitqueue.c +0 -0
index a1d94be7883a..90c9d84e6155 100644

drivers/gpu/drm/msm/msm_submitqueue.c void msm_submitqueue_close(struct msm_file_private *ctx) · line 53

Quoted 6 lines
drivers/gpu/drm/msm/msm_submitqueue.c +2 -1
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
       * No lock needed in close and there won't
       * be any more user ioctls coming our way
       */
-     list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+     list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+             list_del(&entry->node);

If you refcount ctx, what does that do for the entries in the submit queue?

Quoted 4 lines

"entry" here is kref'ed, but you're popping it off the list regardless of the put ends up freeing the object or not - which afaict would mean leaking the object.

What ends up happening is the submit has reference to submit-queue, which has reference to the ctx.. the submitqueue could be alive still pending in-flight submits (in a later patch), but dead from the PoV of userspace interface.

Quoted 5 lines

We aren't relying (or at least aren't in the end, and I *think* I didn't miss anything in the middle) relying on ctx->submitqueues list to clean anything up in the end, just track what is still a valid submitqueue from userspace PoV

Looks reasonable, thanks for the explanation.

Quoted 21 lines

BR,
-R

Quoted 18 lines

On the other hand, with the current implementation an object with higher refcount with adjacent objects of single refcount would end up with dangling pointers after the put. So in itself this change seems like a net gain, but I'm wondering about the plan described in the commit message.

Regards,
Bjorn

Quoted 8 lines
drivers/gpu/drm/msm/msm_submitqueue.c +1 -1
              msm_submitqueue_put(entry);
+     }
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
--

2.26.2

Reviewed-by on v1 1/19
#31 lore raw

Re: [PATCH 01/19] drm/msm: remove dangling submitqueue references

In reply to Rob Clark
Quoted 7 lines · Rob Clark

From: Rob Clark <[email protected]>

Currently it doesn't matter, since we free the ctx immediately. But when we start refcnt'ing the ctx, we don't want old dangling list entries to hang around.

Quoted 23 lines
---
 drivers/gpu/drm/msm/msm_submitqueue.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/msm/msm_submitqueue.c +3 -1
index a1d94be7883a..90c9d84e6155 100644
@@ -49,8 +49,10 @@ void msm_submitqueue_close(struct msm_file_private *ctx)
 	 * No lock needed in close and there won't
 	 * be any more user ioctls coming our way
 	 */
-	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node)
+	list_for_each_entry_safe(entry, tmp, &ctx->submitqueues, node) {
+		list_del(&entry->node);
 		msm_submitqueue_put(entry);
+	}
 }
 
 int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
Signature · 1 line
2.26.2
#32 lore raw

PATCH 2/19 iommu/arm-smmu: Pass io-pgtable config to implementation specific function

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Construct the io-pgtable config before calling the implementation specific init_context function and pass it so the implementation specific function can get a chance to change it before the io-pgtable is created.

---
 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |  3 ++-
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 11 ++++++-----
 drivers/iommu/arm/arm-smmu/arm-smmu.h      |  3 ++-
 3 files changed, 10 insertions(+), 7 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +2 -1
index f4ff124a1967..a9861dcd0884 100644
@@ -68,7 +68,8 @@ static int cavium_cfg_probe(struct arm_smmu_device *smmu)
 	return 0;
 }
 
-static int cavium_init_context(struct arm_smmu_domain *smmu_domain)
+static int cavium_init_context(struct arm_smmu_domain *smmu_domain,
+		struct io_pgtable_cfg *pgtbl_cfg)
 {
 	struct cavium_smmu *cs = container_of(smmu_domain->smmu,
 					      struct cavium_smmu, smmu);
drivers/iommu/arm/arm-smmu/arm-smmu.c +6 -5
index 09c42af9f31e..37d8d49299b4 100644
@@ -795,11 +795,6 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 		cfg->asid = cfg->cbndx;
 
 	smmu_domain->smmu = smmu;
-	if (smmu->impl && smmu->impl->init_context) {
-		ret = smmu->impl->init_context(smmu_domain);
-		if (ret)
-			goto out_unlock;
-	}
 
 	pgtbl_cfg = (struct io_pgtable_cfg) {
 		.pgsize_bitmap	= smmu->pgsize_bitmap,
@@ -810,6 +805,12 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 		.iommu_dev	= smmu->dev,
 	};
 
+	if (smmu->impl && smmu->impl->init_context) {
+		ret = smmu->impl->init_context(smmu_domain, &pgtbl_cfg);
+		if (ret)
+			goto out_clear_smmu;
+	}
+
 	if (smmu_domain->non_strict)
 		pgtbl_cfg.quirks |= IO_PGTABLE_QUIRK_NON_STRICT;
 
drivers/iommu/arm/arm-smmu/arm-smmu.h +2 -1
index d890a4a968e8..83294516ac08 100644
@@ -386,7 +386,8 @@ struct arm_smmu_impl {
 			    u64 val);
 	int (*cfg_probe)(struct arm_smmu_device *smmu);
 	int (*reset)(struct arm_smmu_device *smmu);
-	int (*init_context)(struct arm_smmu_domain *smmu_domain);
+	int (*init_context)(struct arm_smmu_domain *smmu_domain,
+			struct io_pgtable_cfg *cfg);
 	void (*tlb_sync)(struct arm_smmu_device *smmu, int page, int sync,
 			 int status);
 	int (*def_domain_type)(struct device *dev);
Signature · 1 line
2.26.2
Reviewed-by on v1 2/19
#33 lore raw

Re: [PATCH 02/19] iommu/arm-smmu: Pass io-pgtable config to implementation specific function

Quoted 8 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Construct the io-pgtable config before calling the implementation specific init_context function and pass it so the implementation specific function can get a chance to change it before the io-pgtable is created.

Regards,
Bjorn

Quoted 66 lines
---
 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |  3 ++-
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 11 ++++++-----
 drivers/iommu/arm/arm-smmu/arm-smmu.h      |  3 ++-
 3 files changed, 10 insertions(+), 7 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +2 -1
index f4ff124a1967..a9861dcd0884 100644
@@ -68,7 +68,8 @@ static int cavium_cfg_probe(struct arm_smmu_device *smmu)
 	return 0;
 }
 
-static int cavium_init_context(struct arm_smmu_domain *smmu_domain)
+static int cavium_init_context(struct arm_smmu_domain *smmu_domain,
+		struct io_pgtable_cfg *pgtbl_cfg)
 {
 	struct cavium_smmu *cs = container_of(smmu_domain->smmu,
 					      struct cavium_smmu, smmu);
drivers/iommu/arm/arm-smmu/arm-smmu.c +6 -5
index 09c42af9f31e..37d8d49299b4 100644
@@ -795,11 +795,6 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 		cfg->asid = cfg->cbndx;
 
 	smmu_domain->smmu = smmu;
-	if (smmu->impl && smmu->impl->init_context) {
-		ret = smmu->impl->init_context(smmu_domain);
-		if (ret)
-			goto out_unlock;
-	}
 
 	pgtbl_cfg = (struct io_pgtable_cfg) {
 		.pgsize_bitmap	= smmu->pgsize_bitmap,
@@ -810,6 +805,12 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 		.iommu_dev	= smmu->dev,
 	};
 
+	if (smmu->impl && smmu->impl->init_context) {
+		ret = smmu->impl->init_context(smmu_domain, &pgtbl_cfg);
+		if (ret)
+			goto out_clear_smmu;
+	}
+
 	if (smmu_domain->non_strict)
 		pgtbl_cfg.quirks |= IO_PGTABLE_QUIRK_NON_STRICT;
 
drivers/iommu/arm/arm-smmu/arm-smmu.h +2 -1
index d890a4a968e8..83294516ac08 100644
@@ -386,7 +386,8 @@ struct arm_smmu_impl {
 			    u64 val);
 	int (*cfg_probe)(struct arm_smmu_device *smmu);
 	int (*reset)(struct arm_smmu_device *smmu);
-	int (*init_context)(struct arm_smmu_domain *smmu_domain);
+	int (*init_context)(struct arm_smmu_domain *smmu_domain,
+			struct io_pgtable_cfg *cfg);
 	void (*tlb_sync)(struct arm_smmu_device *smmu, int page, int sync,
 			 int status);
 	int (*def_domain_type)(struct device *dev);
Signature · 1 line
2.26.2
#34 lore raw

PATCH 3/19 iommu/arm-smmu: Add support for split pagetables

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Enable TTBR1 for a context bank if IO_PGTABLE_QUIRK_ARM_TTBR1 is selected by the io-pgtable configuration.

---
 drivers/iommu/arm/arm-smmu/arm-smmu.c | 21 ++++++++++++++++-----
 drivers/iommu/arm/arm-smmu/arm-smmu.h | 25 +++++++++++++++++++------
 2 files changed, 35 insertions(+), 11 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu.c +16 -5
index 37d8d49299b4..976d43a7f2ff 100644
@@ -552,11 +552,15 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain,
 			cb->ttbr[0] = pgtbl_cfg->arm_v7s_cfg.ttbr;
 			cb->ttbr[1] = 0;
 		} else {
-			cb->ttbr[0] = pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
-			cb->ttbr[0] |= FIELD_PREP(ARM_SMMU_TTBRn_ASID,
-						  cfg->asid);
+			cb->ttbr[0] = FIELD_PREP(ARM_SMMU_TTBRn_ASID,
+				cfg->asid);
 			cb->ttbr[1] = FIELD_PREP(ARM_SMMU_TTBRn_ASID,
-						 cfg->asid);
+				cfg->asid);
+
+			if (pgtbl_cfg->quirks & IO_PGTABLE_QUIRK_ARM_TTBR1)
+				cb->ttbr[1] |= pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
+			else
+				cb->ttbr[0] |= pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
 		}
 	} else {
 		cb->ttbr[0] = pgtbl_cfg->arm_lpae_s2_cfg.vttbr;
@@ -822,7 +826,14 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 
 	/* Update the domain's page sizes to reflect the page table format */
 	domain->pgsize_bitmap = pgtbl_cfg.pgsize_bitmap;
-	domain->geometry.aperture_end = (1UL << ias) - 1;
+
+	if (pgtbl_cfg.quirks & IO_PGTABLE_QUIRK_ARM_TTBR1) {
+		domain->geometry.aperture_start = ~0UL << ias;
+		domain->geometry.aperture_end = ~0UL;
+	} else {
+		domain->geometry.aperture_end = (1UL << ias) - 1;
+	}
+
 	domain->geometry.force_aperture = true;
 
 	/* Initialise the context bank with our page table cfg */
drivers/iommu/arm/arm-smmu/arm-smmu.h +19 -6
index 83294516ac08..f3e456893f28 100644
@@ -169,10 +169,12 @@ enum arm_smmu_cbar_type {
 #define ARM_SMMU_CB_TCR			0x30
 #define ARM_SMMU_TCR_EAE		BIT(31)
 #define ARM_SMMU_TCR_EPD1		BIT(23)
+#define ARM_SMMU_TCR_A1			BIT(22)
 #define ARM_SMMU_TCR_TG0		GENMASK(15, 14)
 #define ARM_SMMU_TCR_SH0		GENMASK(13, 12)
 #define ARM_SMMU_TCR_ORGN0		GENMASK(11, 10)
 #define ARM_SMMU_TCR_IRGN0		GENMASK(9, 8)
+#define ARM_SMMU_TCR_EPD0		BIT(7)
 #define ARM_SMMU_TCR_T0SZ		GENMASK(5, 0)
 
 #define ARM_SMMU_VTCR_RES1		BIT(31)
@@ -350,12 +352,23 @@ struct arm_smmu_domain {
 
 static inline u32 arm_smmu_lpae_tcr(struct io_pgtable_cfg *cfg)
 {
-	return ARM_SMMU_TCR_EPD1 |
-	       FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
-	       FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
-	       FIELD_PREP(ARM_SMMU_TCR_ORGN0, cfg->arm_lpae_s1_cfg.tcr.orgn) |
-	       FIELD_PREP(ARM_SMMU_TCR_IRGN0, cfg->arm_lpae_s1_cfg.tcr.irgn) |
-	       FIELD_PREP(ARM_SMMU_TCR_T0SZ, cfg->arm_lpae_s1_cfg.tcr.tsz);
+	u32 tcr = FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
+		FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
+		FIELD_PREP(ARM_SMMU_TCR_ORGN0, cfg->arm_lpae_s1_cfg.tcr.orgn) |
+		FIELD_PREP(ARM_SMMU_TCR_IRGN0, cfg->arm_lpae_s1_cfg.tcr.irgn) |
+		FIELD_PREP(ARM_SMMU_TCR_T0SZ, cfg->arm_lpae_s1_cfg.tcr.tsz);
+
+       /*
+	* When TTBR1 is selected shift the TCR fields by 16 bits and disable
+	* translation in TTBR0
+	*/
+	if (cfg->quirks & IO_PGTABLE_QUIRK_ARM_TTBR1) {
+		tcr = (tcr << 16) & ~ARM_SMMU_TCR_A1;
+		tcr |= ARM_SMMU_TCR_EPD0;
+	} else
+		tcr |= ARM_SMMU_TCR_EPD1;
+
+	return tcr;
 }
 
 static inline u32 arm_smmu_lpae_tcr2(struct io_pgtable_cfg *cfg)
Signature · 1 line
2.26.2
Reviewed-by on v1 3/19
#35 lore raw

Re: [PATCH 03/19] iommu/arm-smmu: Add support for split pagetables

Quoted 18 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Enable TTBR1 for a context bank if IO_PGTABLE_QUIRK_ARM_TTBR1 is selected by the io-pgtable configuration.

---
 drivers/iommu/arm/arm-smmu/arm-smmu.c | 21 ++++++++++++++++-----
 drivers/iommu/arm/arm-smmu/arm-smmu.h | 25 +++++++++++++++++++------
 2 files changed, 35 insertions(+), 11 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu.c +0 -0
index 37d8d49299b4..976d43a7f2ff 100644
@@ -552,11 +552,15 @@ static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain,
 			cb->ttbr[0] = pgtbl_cfg->arm_v7s_cfg.ttbr;

drivers/iommu/arm/arm-smmu/arm-smmu.c static void arm_smmu_init_context_bank(struct arm_smmu_domain *smmu_domain, · line 558

Quoted 10 lines
drivers/iommu/arm/arm-smmu/arm-smmu.c +3 -4
 			cb->ttbr[1] = 0;
 		} else {
-			cb->ttbr[0] = pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
-			cb->ttbr[0] |= FIELD_PREP(ARM_SMMU_TTBRn_ASID,
-						  cfg->asid);
+			cb->ttbr[0] = FIELD_PREP(ARM_SMMU_TTBRn_ASID,
+				cfg->asid);
 			cb->ttbr[1] = FIELD_PREP(ARM_SMMU_TTBRn_ASID,
-						 cfg->asid);
+				cfg->asid);

The old indentation seems more appropriate.

Apart from that this looks sensible.

Regards,
Bjorn

Quoted 74 lines
drivers/iommu/arm/arm-smmu/arm-smmu.c +13 -1
+
+			if (pgtbl_cfg->quirks & IO_PGTABLE_QUIRK_ARM_TTBR1)
+				cb->ttbr[1] |= pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
+			else
+				cb->ttbr[0] |= pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
 		}
 	} else {
 		cb->ttbr[0] = pgtbl_cfg->arm_lpae_s2_cfg.vttbr;
@@ -822,7 +826,14 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
 
 	/* Update the domain's page sizes to reflect the page table format */
 	domain->pgsize_bitmap = pgtbl_cfg.pgsize_bitmap;
-	domain->geometry.aperture_end = (1UL << ias) - 1;
+
+	if (pgtbl_cfg.quirks & IO_PGTABLE_QUIRK_ARM_TTBR1) {
+		domain->geometry.aperture_start = ~0UL << ias;
+		domain->geometry.aperture_end = ~0UL;
+	} else {
+		domain->geometry.aperture_end = (1UL << ias) - 1;
+	}
+
 	domain->geometry.force_aperture = true;
 
 	/* Initialise the context bank with our page table cfg */
drivers/iommu/arm/arm-smmu/arm-smmu.h +19 -6
index 83294516ac08..f3e456893f28 100644
@@ -169,10 +169,12 @@ enum arm_smmu_cbar_type {
 #define ARM_SMMU_CB_TCR			0x30
 #define ARM_SMMU_TCR_EAE		BIT(31)
 #define ARM_SMMU_TCR_EPD1		BIT(23)
+#define ARM_SMMU_TCR_A1			BIT(22)
 #define ARM_SMMU_TCR_TG0		GENMASK(15, 14)
 #define ARM_SMMU_TCR_SH0		GENMASK(13, 12)
 #define ARM_SMMU_TCR_ORGN0		GENMASK(11, 10)
 #define ARM_SMMU_TCR_IRGN0		GENMASK(9, 8)
+#define ARM_SMMU_TCR_EPD0		BIT(7)
 #define ARM_SMMU_TCR_T0SZ		GENMASK(5, 0)
 
 #define ARM_SMMU_VTCR_RES1		BIT(31)
@@ -350,12 +352,23 @@ struct arm_smmu_domain {
 
 static inline u32 arm_smmu_lpae_tcr(struct io_pgtable_cfg *cfg)
 {
-	return ARM_SMMU_TCR_EPD1 |
-	       FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
-	       FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
-	       FIELD_PREP(ARM_SMMU_TCR_ORGN0, cfg->arm_lpae_s1_cfg.tcr.orgn) |
-	       FIELD_PREP(ARM_SMMU_TCR_IRGN0, cfg->arm_lpae_s1_cfg.tcr.irgn) |
-	       FIELD_PREP(ARM_SMMU_TCR_T0SZ, cfg->arm_lpae_s1_cfg.tcr.tsz);
+	u32 tcr = FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
+		FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
+		FIELD_PREP(ARM_SMMU_TCR_ORGN0, cfg->arm_lpae_s1_cfg.tcr.orgn) |
+		FIELD_PREP(ARM_SMMU_TCR_IRGN0, cfg->arm_lpae_s1_cfg.tcr.irgn) |
+		FIELD_PREP(ARM_SMMU_TCR_T0SZ, cfg->arm_lpae_s1_cfg.tcr.tsz);
+
+       /*
+	* When TTBR1 is selected shift the TCR fields by 16 bits and disable
+	* translation in TTBR0
+	*/
+	if (cfg->quirks & IO_PGTABLE_QUIRK_ARM_TTBR1) {
+		tcr = (tcr << 16) & ~ARM_SMMU_TCR_A1;
+		tcr |= ARM_SMMU_TCR_EPD0;
+	} else
+		tcr |= ARM_SMMU_TCR_EPD1;
+
+	return tcr;
 }
 
 static inline u32 arm_smmu_lpae_tcr2(struct io_pgtable_cfg *cfg)
Signature · 1 line
2.26.2
#36 lore raw

PATCH 4/19 iommu/arm-smmu: Prepare for the adreno-smmu implementation

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Do a bit of prep work to add the upcoming adreno-smmu implementation.

Add an hook to allow the implementation to choose which context banks to allocate.

Move some of the common structs to arm-smmu.h in anticipation of them being used by the implementations and update some of the existing hooks to pass more information that the implementation will need.

These modifications will be used by the upcoming Adreno SMMU implementation to identify the GPU device and properly configure it for pagetable switching.

---
 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |  2 +-
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 69 ++++++----------------
 drivers/iommu/arm/arm-smmu/arm-smmu.h      | 51 +++++++++++++++-
 3 files changed, 68 insertions(+), 54 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +1 -1
index a9861dcd0884..88f17cc33023 100644
@@ -69,7 +69,7 @@ static int cavium_cfg_probe(struct arm_smmu_device *smmu)
 }
 
 static int cavium_init_context(struct arm_smmu_domain *smmu_domain,
-		struct io_pgtable_cfg *pgtbl_cfg)
+		struct io_pgtable_cfg *pgtbl_cfg, struct device *dev)
 {
 	struct cavium_smmu *cs = container_of(smmu_domain->smmu,
 					      struct cavium_smmu, smmu);
drivers/iommu/arm/arm-smmu/arm-smmu.c +17 -52

123 lines not on this page · raw message

drivers/iommu/arm/arm-smmu/arm-smmu.h +50 -1

88 lines not on this page · raw message

Signature · 1 line
2.26.2
#37 lore raw

PATCH 5/19 iommu: add private interface for adreno-smmu

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

This interface will be used for drm/msm to coordinate with the qcom_adreno_smmu_impl to enable/disable TTBR0 translation.

Once TTBR0 translation is enabled, the GPU's CP (Command Processor) will directly switch TTBR0 pgtables (and do the necessary TLB inv) synchronized to the GPU's operation. But help from the SMMU driver is needed to initially bootstrap TTBR0 translation, which cannot be done from the GPU.

Since this is a very special case, a private interface is used to avoid adding highly driver specific things to the public iommu interface.

---
 include/linux/adreno-smmu-priv.h | 36 ++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)
 create mode 100644 include/linux/adreno-smmu-priv.h
include/linux/adreno-smmu-priv.h new file +36 -0
new file mode 100644
index 000000000000..a889f28afb42
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2020 Google, Inc
+ */
+
+#ifndef __ADRENO_SMMU_PRIV_H
+#define __ADRENO_SMMU_PRIV_H
+
+#include <linux/io-pgtable.h>
+
+/**
+ * struct adreno_smmu_priv - private interface between adreno-smmu and GPU
+ *
+ * @cookie:        An opque token provided by adreno-smmu and passed
+ *                 back into the callbacks
+ * @get_ttbr1_cfg: Get the TTBR1 config for the GPUs context-bank
+ * @set_ttbr0_cfg: Set the TTBR0 config for the GPUs context bank.  A
+ *                 NULL config disables TTBR0 translation, otherwise
+ *                 TTBR0 translation is enabled with the specified cfg
+ *
+ * The GPU driver (drm/msm) and adreno-smmu work together for controlling
+ * the GPU's SMMU instance.  This is by necessity, as the GPU is directly
+ * updating the SMMU for context switches, while on the other hand we do
+ * not want to duplicate all of the initial setup logic from arm-smmu.
+ *
+ * This private interface is used for the two drivers to coordinate.  The
+ * cookie and callback functions are populated when the GPU driver attaches
+ * it's domain.
+ */
+struct adreno_smmu_priv {
+    const void *cookie;
+    const struct io_pgtable_cfg *(*get_ttbr1_cfg)(const void *cookie);
+    int (*set_ttbr0_cfg)(const void *cookie, const struct io_pgtable_cfg *cfg);
+};
+
+#endif /* __ADRENO_SMMU_PRIV_H */
\ No newline at end of file
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author Reviewed-by on v1 5/19
#38 lore raw

Re: [Freedreno] [PATCH 05/19] iommu: add private interface for adreno-smmu

Quoted 11 lines · Rob Clark

From: Rob Clark <[email protected]>

This interface will be used for drm/msm to coordinate with the qcom_adreno_smmu_impl to enable/disable TTBR0 translation.

Once TTBR0 translation is enabled, the GPU's CP (Command Processor) will directly switch TTBR0 pgtables (and do the necessary TLB inv) synchronized to the GPU's operation. But help from the SMMU driver is needed to initially bootstrap TTBR0 translation, which cannot be done from the GPU.

Quoted 4 lines

Since this is a very special case, a private interface is used to avoid adding highly driver specific things to the public iommu interface.

Quoted 56 lines
---
 include/linux/adreno-smmu-priv.h | 36 ++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)
 create mode 100644 include/linux/adreno-smmu-priv.h
include/linux/adreno-smmu-priv.h new file +36 -0
new file mode 100644
index 000000000000..a889f28afb42
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2020 Google, Inc
+ */
+
+#ifndef __ADRENO_SMMU_PRIV_H
+#define __ADRENO_SMMU_PRIV_H
+
+#include <linux/io-pgtable.h>
+
+/**
+ * struct adreno_smmu_priv - private interface between adreno-smmu and GPU
+ *
+ * @cookie:        An opque token provided by adreno-smmu and passed
+ *                 back into the callbacks
+ * @get_ttbr1_cfg: Get the TTBR1 config for the GPUs context-bank
+ * @set_ttbr0_cfg: Set the TTBR0 config for the GPUs context bank.  A
+ *                 NULL config disables TTBR0 translation, otherwise
+ *                 TTBR0 translation is enabled with the specified cfg
+ *
+ * The GPU driver (drm/msm) and adreno-smmu work together for controlling
+ * the GPU's SMMU instance.  This is by necessity, as the GPU is directly
+ * updating the SMMU for context switches, while on the other hand we do
+ * not want to duplicate all of the initial setup logic from arm-smmu.
+ *
+ * This private interface is used for the two drivers to coordinate.  The
+ * cookie and callback functions are populated when the GPU driver attaches
+ * it's domain.
+ */
+struct adreno_smmu_priv {
+    const void *cookie;
+    const struct io_pgtable_cfg *(*get_ttbr1_cfg)(const void *cookie);
+    int (*set_ttbr0_cfg)(const void *cookie, const struct io_pgtable_cfg *cfg);
+};
+
+#endif /* __ADRENO_SMMU_PRIV_H */
\ No newline at end of file
Signature · 6 lines
2.26.2

_______________________________________________
Freedreno mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/freedreno
Signature · 6 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Reviewed-by on v1 5/19
#39 lore raw

Re: [PATCH 05/19] iommu: add private interface for adreno-smmu

In reply to Rob Clark
Quoted 11 lines · Rob Clark

From: Rob Clark <[email protected]>

This interface will be used for drm/msm to coordinate with the qcom_adreno_smmu_impl to enable/disable TTBR0 translation.

Once TTBR0 translation is enabled, the GPU's CP (Command Processor) will directly switch TTBR0 pgtables (and do the necessary TLB inv) synchronized to the GPU's operation. But help from the SMMU driver is needed to initially bootstrap TTBR0 translation, which cannot be done from the GPU.

Quoted 5 lines

Since this is a very special case, a private interface is used to avoid adding highly driver specific things to the public iommu interface.

Quoted 51 lines
---
 include/linux/adreno-smmu-priv.h | 36 ++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)
 create mode 100644 include/linux/adreno-smmu-priv.h
include/linux/adreno-smmu-priv.h new file +36 -0
new file mode 100644
index 000000000000..a889f28afb42
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2020 Google, Inc
+ */
+
+#ifndef __ADRENO_SMMU_PRIV_H
+#define __ADRENO_SMMU_PRIV_H
+
+#include <linux/io-pgtable.h>
+
+/**
+ * struct adreno_smmu_priv - private interface between adreno-smmu and GPU
+ *
+ * @cookie:        An opque token provided by adreno-smmu and passed
+ *                 back into the callbacks
+ * @get_ttbr1_cfg: Get the TTBR1 config for the GPUs context-bank
+ * @set_ttbr0_cfg: Set the TTBR0 config for the GPUs context bank.  A
+ *                 NULL config disables TTBR0 translation, otherwise
+ *                 TTBR0 translation is enabled with the specified cfg
+ *
+ * The GPU driver (drm/msm) and adreno-smmu work together for controlling
+ * the GPU's SMMU instance.  This is by necessity, as the GPU is directly
+ * updating the SMMU for context switches, while on the other hand we do
+ * not want to duplicate all of the initial setup logic from arm-smmu.
+ *
+ * This private interface is used for the two drivers to coordinate.  The
+ * cookie and callback functions are populated when the GPU driver attaches
+ * it's domain.
+ */
+struct adreno_smmu_priv {
+    const void *cookie;
+    const struct io_pgtable_cfg *(*get_ttbr1_cfg)(const void *cookie);
+    int (*set_ttbr0_cfg)(const void *cookie, const struct io_pgtable_cfg *cfg);
+};
+
+#endif /* __ADRENO_SMMU_PRIV_H */
\ No newline at end of file
Signature · 1 line
2.26.2
#40 lore raw

PATCH 6/19 drm/msm/gpu: add dev_to_gpu() helper

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

In a later patch, the drvdata will not directly be 'struct msm_gpu *', so add a helper to reduce the churn.

---
 drivers/gpu/drm/msm/adreno/adreno_device.c | 10 ++++------
 drivers/gpu/drm/msm/msm_gpu.c              |  6 +++---
 drivers/gpu/drm/msm/msm_gpu.h              |  5 +++++
 3 files changed, 12 insertions(+), 9 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_device.c +4 -6
index 9eeb46bf2a5d..26664e1b30c0 100644
@@ -282,7 +282,7 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
 	int ret;
 
 	if (pdev)
-		gpu = platform_get_drvdata(pdev);
+		gpu = dev_to_gpu(&pdev->dev);
 
 	if (!gpu) {
 		dev_err_once(dev->dev, "no GPU device was found\n");
@@ -425,7 +425,7 @@ static int adreno_bind(struct device *dev, struct device *master, void *data)
 static void adreno_unbind(struct device *dev, struct device *master,
 		void *data)
 {
-	struct msm_gpu *gpu = dev_get_drvdata(dev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	pm_runtime_force_suspend(dev);
 	gpu->funcs->destroy(gpu);
@@ -490,16 +490,14 @@ static const struct of_device_id dt_match[] = {
 #ifdef CONFIG_PM
 static int adreno_resume(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct msm_gpu *gpu = platform_get_drvdata(pdev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	return gpu->funcs->pm_resume(gpu);
 }
 
 static int adreno_suspend(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct msm_gpu *gpu = platform_get_drvdata(pdev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	return gpu->funcs->pm_suspend(gpu);
 }
drivers/gpu/drm/msm/msm_gpu.c +3 -3
index d5645472b25d..6aa9e04e52e7 100644
@@ -24,7 +24,7 @@
 static int msm_devfreq_target(struct device *dev, unsigned long *freq,
 		u32 flags)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 	struct dev_pm_opp *opp;
 
 	opp = devfreq_recommended_opp(dev, freq, flags);
@@ -45,7 +45,7 @@ static int msm_devfreq_target(struct device *dev, unsigned long *freq,
 static int msm_devfreq_get_dev_status(struct device *dev,
 		struct devfreq_dev_status *status)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 	ktime_t time;
 
 	if (gpu->funcs->gpu_get_freq)
@@ -64,7 +64,7 @@ static int msm_devfreq_get_dev_status(struct device *dev,
 
 static int msm_devfreq_get_cur_freq(struct device *dev, unsigned long *freq)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	if (gpu->funcs->gpu_get_freq)
 		*freq = gpu->funcs->gpu_get_freq(gpu);
drivers/gpu/drm/msm/msm_gpu.h +5 -0
index 0db117a7339b..8bda7beaed4b 100644
@@ -141,6 +141,11 @@ struct msm_gpu {
 	struct msm_gpu_state *crashstate;
 };
 
+static inline struct msm_gpu *dev_to_gpu(struct device *dev)
+{
+	return dev_get_drvdata(dev);
+}
+
 /* It turns out that all targets use the same ringbuffer size */
 #define MSM_GPU_RINGBUFFER_SZ SZ_32K
 #define MSM_GPU_RINGBUFFER_BLKSIZE 32
Signature · 1 line
2.26.2
author Reviewed-by on v1 6/19
#41 lore raw

Re: [Freedreno] [PATCH 06/19] drm/msm/gpu: add dev_to_gpu() helper

Quoted 5 lines · Rob Clark

From: Rob Clark <[email protected]>

In a later patch, the drvdata will not directly be 'struct msm_gpu *', so add a helper to reduce the churn.

Quoted 102 lines
---
 drivers/gpu/drm/msm/adreno/adreno_device.c | 10 ++++------
 drivers/gpu/drm/msm/msm_gpu.c              |  6 +++---
 drivers/gpu/drm/msm/msm_gpu.h              |  5 +++++
 3 files changed, 12 insertions(+), 9 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_device.c +4 -6
index 9eeb46bf2a5d..26664e1b30c0 100644
@@ -282,7 +282,7 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
 	int ret;
 
 	if (pdev)
-		gpu = platform_get_drvdata(pdev);
+		gpu = dev_to_gpu(&pdev->dev);
 
 	if (!gpu) {
 		dev_err_once(dev->dev, "no GPU device was found\n");
@@ -425,7 +425,7 @@ static int adreno_bind(struct device *dev, struct device *master, void *data)
 static void adreno_unbind(struct device *dev, struct device *master,
 		void *data)
 {
-	struct msm_gpu *gpu = dev_get_drvdata(dev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	pm_runtime_force_suspend(dev);
 	gpu->funcs->destroy(gpu);
@@ -490,16 +490,14 @@ static const struct of_device_id dt_match[] = {
 #ifdef CONFIG_PM
 static int adreno_resume(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct msm_gpu *gpu = platform_get_drvdata(pdev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	return gpu->funcs->pm_resume(gpu);
 }
 
 static int adreno_suspend(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct msm_gpu *gpu = platform_get_drvdata(pdev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	return gpu->funcs->pm_suspend(gpu);
 }
drivers/gpu/drm/msm/msm_gpu.c +3 -3
index d5645472b25d..6aa9e04e52e7 100644
@@ -24,7 +24,7 @@
 static int msm_devfreq_target(struct device *dev, unsigned long *freq,
 		u32 flags)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 	struct dev_pm_opp *opp;
 
 	opp = devfreq_recommended_opp(dev, freq, flags);
@@ -45,7 +45,7 @@ static int msm_devfreq_target(struct device *dev, unsigned long *freq,
 static int msm_devfreq_get_dev_status(struct device *dev,
 		struct devfreq_dev_status *status)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 	ktime_t time;
 
 	if (gpu->funcs->gpu_get_freq)
@@ -64,7 +64,7 @@ static int msm_devfreq_get_dev_status(struct device *dev,
 
 static int msm_devfreq_get_cur_freq(struct device *dev, unsigned long *freq)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	if (gpu->funcs->gpu_get_freq)
 		*freq = gpu->funcs->gpu_get_freq(gpu);
drivers/gpu/drm/msm/msm_gpu.h +5 -0
index 0db117a7339b..8bda7beaed4b 100644
@@ -141,6 +141,11 @@ struct msm_gpu {
 	struct msm_gpu_state *crashstate;
 };
 
+static inline struct msm_gpu *dev_to_gpu(struct device *dev)
+{
+	return dev_get_drvdata(dev);
+}
+
 /* It turns out that all targets use the same ringbuffer size */
 #define MSM_GPU_RINGBUFFER_SZ SZ_32K
 #define MSM_GPU_RINGBUFFER_BLKSIZE 32
Signature · 6 lines
2.26.2

_______________________________________________
Freedreno mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/freedreno
Signature · 6 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Reviewed-by on v1 6/19
#42 lore raw

Re: [PATCH 06/19] drm/msm/gpu: add dev_to_gpu() helper

In reply to Rob Clark
Quoted 89 lines · Rob Clark

From: Rob Clark <[email protected]>

In a later patch, the drvdata will not directly be 'struct msm_gpu *', so add a helper to reduce the churn.

---
 drivers/gpu/drm/msm/adreno/adreno_device.c | 10 ++++------
 drivers/gpu/drm/msm/msm_gpu.c              |  6 +++---
 drivers/gpu/drm/msm/msm_gpu.h              |  5 +++++
 3 files changed, 12 insertions(+), 9 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_device.c +4 -6
index 9eeb46bf2a5d..26664e1b30c0 100644
@@ -282,7 +282,7 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
 	int ret;
 
 	if (pdev)
-		gpu = platform_get_drvdata(pdev);
+		gpu = dev_to_gpu(&pdev->dev);
 
 	if (!gpu) {
 		dev_err_once(dev->dev, "no GPU device was found\n");
@@ -425,7 +425,7 @@ static int adreno_bind(struct device *dev, struct device *master, void *data)
 static void adreno_unbind(struct device *dev, struct device *master,
 		void *data)
 {
-	struct msm_gpu *gpu = dev_get_drvdata(dev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	pm_runtime_force_suspend(dev);
 	gpu->funcs->destroy(gpu);
@@ -490,16 +490,14 @@ static const struct of_device_id dt_match[] = {
 #ifdef CONFIG_PM
 static int adreno_resume(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct msm_gpu *gpu = platform_get_drvdata(pdev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	return gpu->funcs->pm_resume(gpu);
 }
 
 static int adreno_suspend(struct device *dev)
 {
-	struct platform_device *pdev = to_platform_device(dev);
-	struct msm_gpu *gpu = platform_get_drvdata(pdev);
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	return gpu->funcs->pm_suspend(gpu);
 }
drivers/gpu/drm/msm/msm_gpu.c +3 -3
index d5645472b25d..6aa9e04e52e7 100644
@@ -24,7 +24,7 @@
 static int msm_devfreq_target(struct device *dev, unsigned long *freq,
 		u32 flags)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 	struct dev_pm_opp *opp;
 
 	opp = devfreq_recommended_opp(dev, freq, flags);
@@ -45,7 +45,7 @@ static int msm_devfreq_target(struct device *dev, unsigned long *freq,
 static int msm_devfreq_get_dev_status(struct device *dev,
 		struct devfreq_dev_status *status)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 	ktime_t time;
 
 	if (gpu->funcs->gpu_get_freq)
@@ -64,7 +64,7 @@ static int msm_devfreq_get_dev_status(struct device *dev,
 
 static int msm_devfreq_get_cur_freq(struct device *dev, unsigned long *freq)
 {
-	struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+	struct msm_gpu *gpu = dev_to_gpu(dev);
 
 	if (gpu->funcs->gpu_get_freq)
 		*freq = gpu->funcs->gpu_get_freq(gpu);
drivers/gpu/drm/msm/msm_gpu.h +0 -0
index 0db117a7339b..8bda7beaed4b 100644

drivers/gpu/drm/msm/msm_gpu.h struct msm_gpu { · line 144

Quoted 4 lines
drivers/gpu/drm/msm/msm_gpu.h +1 -0
@@ -141,6 +141,11 @@ struct msm_gpu {
 	struct msm_gpu_state *crashstate;
 };
 
+static inline struct msm_gpu *dev_to_gpu(struct device *dev)

That's a fairly generic name for a driver-global helper :)

Regards,
Bjorn

Quoted 10 lines
drivers/gpu/drm/msm/msm_gpu.h +4 -0
+{
+	return dev_get_drvdata(dev);
+}
+
 /* It turns out that all targets use the same ringbuffer size */
 #define MSM_GPU_RINGBUFFER_SZ SZ_32K
 #define MSM_GPU_RINGBUFFER_BLKSIZE 32
Signature · 1 line
2.26.2
#43 lore raw

Re: [PATCH 06/19] drm/msm/gpu: add dev_to_gpu() helper

Quoted 95 lines · Bjorn Andersson
Quoted 92 lines · Rob Clark

From: Rob Clark <[email protected]>

In a later patch, the drvdata will not directly be 'struct msm_gpu *', so add a helper to reduce the churn.

---
 drivers/gpu/drm/msm/adreno/adreno_device.c | 10 ++++------
 drivers/gpu/drm/msm/msm_gpu.c              |  6 +++---
 drivers/gpu/drm/msm/msm_gpu.h              |  5 +++++
 3 files changed, 12 insertions(+), 9 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_device.c +4 -6
index 9eeb46bf2a5d..26664e1b30c0 100644
@@ -282,7 +282,7 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
      int ret;
 
      if (pdev)
-             gpu = platform_get_drvdata(pdev);
+             gpu = dev_to_gpu(&pdev->dev);
 
      if (!gpu) {
              dev_err_once(dev->dev, "no GPU device was found\n");
@@ -425,7 +425,7 @@ static int adreno_bind(struct device *dev, struct device *master, void *data)
 static void adreno_unbind(struct device *dev, struct device *master,
              void *data)
 {
-     struct msm_gpu *gpu = dev_get_drvdata(dev);
+     struct msm_gpu *gpu = dev_to_gpu(dev);
 
      pm_runtime_force_suspend(dev);
      gpu->funcs->destroy(gpu);
@@ -490,16 +490,14 @@ static const struct of_device_id dt_match[] = {
 #ifdef CONFIG_PM
 static int adreno_resume(struct device *dev)
 {
-     struct platform_device *pdev = to_platform_device(dev);
-     struct msm_gpu *gpu = platform_get_drvdata(pdev);
+     struct msm_gpu *gpu = dev_to_gpu(dev);
 
      return gpu->funcs->pm_resume(gpu);
 }
 
 static int adreno_suspend(struct device *dev)
 {
-     struct platform_device *pdev = to_platform_device(dev);
-     struct msm_gpu *gpu = platform_get_drvdata(pdev);
+     struct msm_gpu *gpu = dev_to_gpu(dev);
 
      return gpu->funcs->pm_suspend(gpu);
 }
drivers/gpu/drm/msm/msm_gpu.c +3 -3
index d5645472b25d..6aa9e04e52e7 100644
@@ -24,7 +24,7 @@
 static int msm_devfreq_target(struct device *dev, unsigned long *freq,
              u32 flags)
 {
-     struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+     struct msm_gpu *gpu = dev_to_gpu(dev);
      struct dev_pm_opp *opp;
 
      opp = devfreq_recommended_opp(dev, freq, flags);
@@ -45,7 +45,7 @@ static int msm_devfreq_target(struct device *dev, unsigned long *freq,
 static int msm_devfreq_get_dev_status(struct device *dev,
              struct devfreq_dev_status *status)
 {
-     struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+     struct msm_gpu *gpu = dev_to_gpu(dev);
      ktime_t time;
 
      if (gpu->funcs->gpu_get_freq)
@@ -64,7 +64,7 @@ static int msm_devfreq_get_dev_status(struct device *dev,
 
 static int msm_devfreq_get_cur_freq(struct device *dev, unsigned long *freq)
 {
-     struct msm_gpu *gpu = platform_get_drvdata(to_platform_device(dev));
+     struct msm_gpu *gpu = dev_to_gpu(dev);
 
      if (gpu->funcs->gpu_get_freq)
              *freq = gpu->funcs->gpu_get_freq(gpu);
drivers/gpu/drm/msm/msm_gpu.h +0 -0
index 0db117a7339b..8bda7beaed4b 100644
@@ -141,6 +141,11 @@ struct msm_gpu {
      struct msm_gpu_state *crashstate;
 };
 
Quoted 3 lines

drivers/gpu/drm/msm/msm_gpu.h struct msm_gpu { · line 144

Quoted 1 line
drivers/gpu/drm/msm/msm_gpu.h +1 -0
+static inline struct msm_gpu *dev_to_gpu(struct device *dev)

That's a fairly generic name for a driver-global helper :)

tbf, it is only global to the gpu part of the driver..

thanks for the review

BR,
-R

Quoted 5 lines

Regards,
Bjorn

Quoted 10 lines
Quoted 10 lines
drivers/gpu/drm/msm/msm_gpu.h +4 -1
+{
+     return dev_get_drvdata(dev);
+}
+
 /* It turns out that all targets use the same ringbuffer size */
 #define MSM_GPU_RINGBUFFER_SZ SZ_32K
 #define MSM_GPU_RINGBUFFER_BLKSIZE 32
--

2.26.2

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

#44 lore raw

PATCH 7/19 drm/msm: set adreno_smmu as gpu's drvdata

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

This will be populated by adreno-smmu, to provide a way for coordinating enabling/disabling TTBR0 translation.

---
 drivers/gpu/drm/msm/adreno/adreno_device.c | 2 --
 drivers/gpu/drm/msm/msm_gpu.c              | 2 +-
 drivers/gpu/drm/msm/msm_gpu.h              | 6 +++++-
 3 files changed, 6 insertions(+), 4 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_device.c +0 -2
index 26664e1b30c0..58e03b20e1c7 100644
@@ -417,8 +417,6 @@ static int adreno_bind(struct device *dev, struct device *master, void *data)
 		return PTR_ERR(gpu);
 	}
 
-	dev_set_drvdata(dev, gpu);
-
 	return 0;
 }
 
drivers/gpu/drm/msm/msm_gpu.c +1 -1
index 6aa9e04e52e7..806eb0957280 100644
@@ -892,7 +892,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		gpu->gpu_cx = NULL;
 
 	gpu->pdev = pdev;
-	platform_set_drvdata(pdev, gpu);
+	platform_set_drvdata(pdev, &gpu->adreno_smmu);
 
 	msm_devfreq_init(gpu);
 
drivers/gpu/drm/msm/msm_gpu.h +5 -1
index 8bda7beaed4b..f91b141add75 100644
@@ -7,6 +7,7 @@
 #ifndef __MSM_GPU_H__
 #define __MSM_GPU_H__
 
+#include <linux/adreno-smmu-priv.h>
 #include <linux/clk.h>
 #include <linux/interconnect.h>
 #include <linux/pm_opp.h>
@@ -73,6 +74,8 @@ struct msm_gpu {
 	struct platform_device *pdev;
 	const struct msm_gpu_funcs *funcs;
 
+	struct adreno_smmu_priv adreno_smmu;
+
 	/* performance counters (hw & sw): */
 	spinlock_t perf_lock;
 	bool perfcntr_active;
@@ -143,7 +146,8 @@ struct msm_gpu {
 
 static inline struct msm_gpu *dev_to_gpu(struct device *dev)
 {
-	return dev_get_drvdata(dev);
+	struct adreno_smmu_priv *adreno_smmu = dev_get_drvdata(dev);
+	return container_of(adreno_smmu, struct msm_gpu, adreno_smmu);
 }
 
 /* It turns out that all targets use the same ringbuffer size */
Signature · 1 line
2.26.2
author Reviewed-by on v1 7/19
#45 lore raw

Re: [PATCH 07/19] drm/msm: set adreno_smmu as gpu's drvdata

Quoted 5 lines · Rob Clark

From: Rob Clark <[email protected]>

This will be populated by adreno-smmu, to provide a way for coordinating enabling/disabling TTBR0 translation.

Quoted 67 lines
---
 drivers/gpu/drm/msm/adreno/adreno_device.c | 2 --
 drivers/gpu/drm/msm/msm_gpu.c              | 2 +-
 drivers/gpu/drm/msm/msm_gpu.h              | 6 +++++-
 3 files changed, 6 insertions(+), 4 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_device.c +0 -2
index 26664e1b30c0..58e03b20e1c7 100644
@@ -417,8 +417,6 @@ static int adreno_bind(struct device *dev, struct device *master, void *data)
 		return PTR_ERR(gpu);
 	}
 
-	dev_set_drvdata(dev, gpu);
-
 	return 0;
 }
 
drivers/gpu/drm/msm/msm_gpu.c +1 -1
index 6aa9e04e52e7..806eb0957280 100644
@@ -892,7 +892,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		gpu->gpu_cx = NULL;
 
 	gpu->pdev = pdev;
-	platform_set_drvdata(pdev, gpu);
+	platform_set_drvdata(pdev, &gpu->adreno_smmu);
 
 	msm_devfreq_init(gpu);
 
drivers/gpu/drm/msm/msm_gpu.h +5 -1
index 8bda7beaed4b..f91b141add75 100644
@@ -7,6 +7,7 @@
 #ifndef __MSM_GPU_H__
 #define __MSM_GPU_H__
 
+#include <linux/adreno-smmu-priv.h>
 #include <linux/clk.h>
 #include <linux/interconnect.h>
 #include <linux/pm_opp.h>
@@ -73,6 +74,8 @@ struct msm_gpu {
 	struct platform_device *pdev;
 	const struct msm_gpu_funcs *funcs;
 
+	struct adreno_smmu_priv adreno_smmu;
+
 	/* performance counters (hw & sw): */
 	spinlock_t perf_lock;
 	bool perfcntr_active;
@@ -143,7 +146,8 @@ struct msm_gpu {
 
 static inline struct msm_gpu *dev_to_gpu(struct device *dev)
 {
-	return dev_get_drvdata(dev);
+	struct adreno_smmu_priv *adreno_smmu = dev_get_drvdata(dev);
+	return container_of(adreno_smmu, struct msm_gpu, adreno_smmu);
 }
 
 /* It turns out that all targets use the same ringbuffer size */
Signature · 1 line
2.26.2
Signature · 6 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Reviewed-by on v1 7/19
#46 lore raw

Re: [PATCH 07/19] drm/msm: set adreno_smmu as gpu's drvdata

In reply to Rob Clark
Quoted 5 lines · Rob Clark

From: Rob Clark <[email protected]>

This will be populated by adreno-smmu, to provide a way for coordinating enabling/disabling TTBR0 translation.

Quoted 67 lines
---
 drivers/gpu/drm/msm/adreno/adreno_device.c | 2 --
 drivers/gpu/drm/msm/msm_gpu.c              | 2 +-
 drivers/gpu/drm/msm/msm_gpu.h              | 6 +++++-
 3 files changed, 6 insertions(+), 4 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_device.c +0 -2
index 26664e1b30c0..58e03b20e1c7 100644
@@ -417,8 +417,6 @@ static int adreno_bind(struct device *dev, struct device *master, void *data)
 		return PTR_ERR(gpu);
 	}
 
-	dev_set_drvdata(dev, gpu);
-
 	return 0;
 }
 
drivers/gpu/drm/msm/msm_gpu.c +1 -1
index 6aa9e04e52e7..806eb0957280 100644
@@ -892,7 +892,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		gpu->gpu_cx = NULL;
 
 	gpu->pdev = pdev;
-	platform_set_drvdata(pdev, gpu);
+	platform_set_drvdata(pdev, &gpu->adreno_smmu);
 
 	msm_devfreq_init(gpu);
 
drivers/gpu/drm/msm/msm_gpu.h +5 -1
index 8bda7beaed4b..f91b141add75 100644
@@ -7,6 +7,7 @@
 #ifndef __MSM_GPU_H__
 #define __MSM_GPU_H__
 
+#include <linux/adreno-smmu-priv.h>
 #include <linux/clk.h>
 #include <linux/interconnect.h>
 #include <linux/pm_opp.h>
@@ -73,6 +74,8 @@ struct msm_gpu {
 	struct platform_device *pdev;
 	const struct msm_gpu_funcs *funcs;
 
+	struct adreno_smmu_priv adreno_smmu;
+
 	/* performance counters (hw & sw): */
 	spinlock_t perf_lock;
 	bool perfcntr_active;
@@ -143,7 +146,8 @@ struct msm_gpu {
 
 static inline struct msm_gpu *dev_to_gpu(struct device *dev)
 {
-	return dev_get_drvdata(dev);
+	struct adreno_smmu_priv *adreno_smmu = dev_get_drvdata(dev);
+	return container_of(adreno_smmu, struct msm_gpu, adreno_smmu);
 }
 
 /* It turns out that all targets use the same ringbuffer size */
Signature · 1 line
2.26.2
#47 lore raw

PATCH 8/19 iommu/arm-smmu: constify some helpers

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

Sprinkle a few consts where helpers don't need write access.

---
 drivers/iommu/arm/arm-smmu/arm-smmu.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu.h +3 -3
index 59ff3fc5c6c8..27c83333fc50 100644
@@ -377,7 +377,7 @@ struct arm_smmu_master_cfg {
 	s16				smendx[];
 };
 
-static inline u32 arm_smmu_lpae_tcr(struct io_pgtable_cfg *cfg)
+static inline u32 arm_smmu_lpae_tcr(const struct io_pgtable_cfg *cfg)
 {
 	u32 tcr = FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
 		FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
@@ -398,13 +398,13 @@ static inline u32 arm_smmu_lpae_tcr(struct io_pgtable_cfg *cfg)
 	return tcr;
 }
 
-static inline u32 arm_smmu_lpae_tcr2(struct io_pgtable_cfg *cfg)
+static inline u32 arm_smmu_lpae_tcr2(const struct io_pgtable_cfg *cfg)
 {
 	return FIELD_PREP(ARM_SMMU_TCR2_PASIZE, cfg->arm_lpae_s1_cfg.tcr.ips) |
 	       FIELD_PREP(ARM_SMMU_TCR2_SEP, ARM_SMMU_TCR2_SEP_UPSTREAM);
 }
 
-static inline u32 arm_smmu_lpae_vtcr(struct io_pgtable_cfg *cfg)
+static inline u32 arm_smmu_lpae_vtcr(const struct io_pgtable_cfg *cfg)
 {
 	return ARM_SMMU_VTCR_RES1 |
 	       FIELD_PREP(ARM_SMMU_VTCR_PS, cfg->arm_lpae_s2_cfg.vtcr.ps) |
Signature · 1 line
2.26.2
Reviewed-by on v1 8/19
#48 lore raw

Re: [PATCH 08/19] iommu/arm-smmu: constify some helpers

Quoted 4 lines · Rob Clark

From: Rob Clark <[email protected]>

Sprinkle a few consts where helpers don't need write access.

Quoted 37 lines
---
 drivers/iommu/arm/arm-smmu/arm-smmu.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu.h +3 -3
index 59ff3fc5c6c8..27c83333fc50 100644
@@ -377,7 +377,7 @@ struct arm_smmu_master_cfg {
 	s16				smendx[];
 };
 
-static inline u32 arm_smmu_lpae_tcr(struct io_pgtable_cfg *cfg)
+static inline u32 arm_smmu_lpae_tcr(const struct io_pgtable_cfg *cfg)
 {
 	u32 tcr = FIELD_PREP(ARM_SMMU_TCR_TG0, cfg->arm_lpae_s1_cfg.tcr.tg) |
 		FIELD_PREP(ARM_SMMU_TCR_SH0, cfg->arm_lpae_s1_cfg.tcr.sh) |
@@ -398,13 +398,13 @@ static inline u32 arm_smmu_lpae_tcr(struct io_pgtable_cfg *cfg)
 	return tcr;
 }
 
-static inline u32 arm_smmu_lpae_tcr2(struct io_pgtable_cfg *cfg)
+static inline u32 arm_smmu_lpae_tcr2(const struct io_pgtable_cfg *cfg)
 {
 	return FIELD_PREP(ARM_SMMU_TCR2_PASIZE, cfg->arm_lpae_s1_cfg.tcr.ips) |
 	       FIELD_PREP(ARM_SMMU_TCR2_SEP, ARM_SMMU_TCR2_SEP_UPSTREAM);
 }
 
-static inline u32 arm_smmu_lpae_vtcr(struct io_pgtable_cfg *cfg)
+static inline u32 arm_smmu_lpae_vtcr(const struct io_pgtable_cfg *cfg)
 {
 	return ARM_SMMU_VTCR_RES1 |
 	       FIELD_PREP(ARM_SMMU_VTCR_PS, cfg->arm_lpae_s2_cfg.vtcr.ps) |
Signature · 1 line
2.26.2
#49 lore raw

PATCH 9/19 iommu/arm-smmu-qcom: Add implementation for the adreno GPU SMMU

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Add a special implementation for the SMMU attached to most Adreno GPU target triggered from the qcom,adreno-smmu compatible string.

The new Adreno SMMU implementation will enable split pagetables (TTBR1) for the domain attached to the GPU device (SID 0) and hard code it context bank 0 so the GPU hardware can implement per-instance pagetables.

---
 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |   3 +
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 149 ++++++++++++++++++++-
 drivers/iommu/arm/arm-smmu/arm-smmu.h      |   1 +
 3 files changed, 151 insertions(+), 2 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +3 -0
index 88f17cc33023..d199b4bff15d 100644
@@ -223,6 +223,9 @@ struct arm_smmu_device *arm_smmu_impl_init(struct arm_smmu_device *smmu)
 	    of_device_is_compatible(np, "qcom,sm8250-smmu-500"))
 		return qcom_smmu_impl_init(smmu);
 
+	if (of_device_is_compatible(smmu->dev->of_node, "qcom,adreno-smmu"))
+		return qcom_adreno_smmu_impl_init(smmu);
+
 	if (of_device_is_compatible(np, "marvell,ap806-smmu-500"))
 		smmu->impl = &mrvl_mmu500_impl;
 
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +147 -2

174 lines not on this page · raw message

drivers/iommu/arm/arm-smmu/arm-smmu.h +1 -0
index 27c83333fc50..cd75a33967bb 100644
@@ -519,6 +519,7 @@ static inline void arm_smmu_writeq(struct arm_smmu_device *smmu, int page,
 struct arm_smmu_device *arm_smmu_impl_init(struct arm_smmu_device *smmu);
 struct arm_smmu_device *nvidia_smmu_impl_init(struct arm_smmu_device *smmu);
 struct arm_smmu_device *qcom_smmu_impl_init(struct arm_smmu_device *smmu);
+struct arm_smmu_device *qcom_adreno_smmu_impl_init(struct arm_smmu_device *smmu);
 
 void arm_smmu_write_context_bank(struct arm_smmu_device *smmu, int idx);
 int arm_mmu500_reset(struct arm_smmu_device *smmu);
Signature · 1 line
2.26.2
Reviewed-by on v1 9/19
#50 lore raw

Re: [PATCH 09/19] iommu/arm-smmu-qcom: Add implementation for the adreno GPU SMMU

Quoted 10 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Add a special implementation for the SMMU attached to most Adreno GPU target triggered from the qcom,adreno-smmu compatible string.

The new Adreno SMMU implementation will enable split pagetables (TTBR1) for the domain attached to the GPU device (SID 0) and hard code it context bank 0 so the GPU hardware can implement per-instance pagetables.

Quoted 220 lines
---
 drivers/iommu/arm/arm-smmu/arm-smmu-impl.c |   3 +
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 149 ++++++++++++++++++++-
 drivers/iommu/arm/arm-smmu/arm-smmu.h      |   1 +
 3 files changed, 151 insertions(+), 2 deletions(-)
drivers/iommu/arm/arm-smmu/arm-smmu-impl.c +3 -0
index 88f17cc33023..d199b4bff15d 100644
@@ -223,6 +223,9 @@ struct arm_smmu_device *arm_smmu_impl_init(struct arm_smmu_device *smmu)
 	    of_device_is_compatible(np, "qcom,sm8250-smmu-500"))
 		return qcom_smmu_impl_init(smmu);
 
+	if (of_device_is_compatible(smmu->dev->of_node, "qcom,adreno-smmu"))
+		return qcom_adreno_smmu_impl_init(smmu);
+
 	if (of_device_is_compatible(np, "marvell,ap806-smmu-500"))
 		smmu->impl = &mrvl_mmu500_impl;
 
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +147 -2
index be4318044f96..5640d9960610 100644
@@ -3,6 +3,7 @@
  * Copyright (c) 2019, The Linux Foundation. All rights reserved.
  */
 
+#include <linux/adreno-smmu-priv.h>
 #include <linux/of_device.h>
 #include <linux/qcom_scm.h>
 
@@ -12,6 +13,132 @@ struct qcom_smmu {
 	struct arm_smmu_device smmu;
 };
 
+#define QCOM_ADRENO_SMMU_GPU_SID 0
+
+static bool qcom_adreno_smmu_is_gpu_device(struct device *dev)
+{
+	struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+	int i;
+
+	/*
+	 * The GPU will always use SID 0 so that is a handy way to uniquely
+	 * identify it and configure it for per-instance pagetables
+	 */
+	for (i = 0; i < fwspec->num_ids; i++) {
+		u16 sid = FIELD_GET(ARM_SMMU_SMR_ID, fwspec->ids[i]);
+
+		if (sid == QCOM_ADRENO_SMMU_GPU_SID)
+			return true;
+	}
+
+	return false;
+}
+
+static const struct io_pgtable_cfg *qcom_adreno_smmu_get_ttbr1_cfg(
+		const void *cookie)
+{
+	struct arm_smmu_domain *smmu_domain = (void *)cookie;
+	struct io_pgtable *pgtable =
+		io_pgtable_ops_to_pgtable(smmu_domain->pgtbl_ops);
+	return &pgtable->cfg;
+}
+
+/*
+ * Local implementation to configure TTBR0 with the specified pagetable config.
+ * The GPU driver will call this to enable TTBR0 when per-instance pagetables
+ * are active
+ */
+
+static int qcom_adreno_smmu_set_ttbr0_cfg(const void *cookie,
+		const struct io_pgtable_cfg *pgtbl_cfg)
+{
+	struct arm_smmu_domain *smmu_domain = (void *)cookie;
+	struct io_pgtable *pgtable = io_pgtable_ops_to_pgtable(smmu_domain->pgtbl_ops);
+	struct arm_smmu_cfg *cfg = &smmu_domain->cfg;
+	struct arm_smmu_cb *cb = &smmu_domain->smmu->cbs[cfg->cbndx];
+
+	/* The domain must have split pagetables already enabled */
+	if (cb->tcr[0] & ARM_SMMU_TCR_EPD1)
+		return -EINVAL;
+
+	/* If the pagetable config is NULL, disable TTBR0 */
+	if (!pgtbl_cfg) {
+		/* Do nothing if it is already disabled */
+		if ((cb->tcr[0] & ARM_SMMU_TCR_EPD0))
+			return -EINVAL;
+
+		/* Set TCR to the original configuration */
+		cb->tcr[0] = arm_smmu_lpae_tcr(&pgtable->cfg);
+		cb->ttbr[0] = FIELD_PREP(ARM_SMMU_TTBRn_ASID, cb->cfg->asid);
+	} else {
+		u32 tcr = cb->tcr[0];
+
+		/* Don't call this again if TTBR0 is already enabled */
+		if (!(cb->tcr[0] & ARM_SMMU_TCR_EPD0))
+			return -EINVAL;
+
+		tcr |= arm_smmu_lpae_tcr(pgtbl_cfg);
+		tcr &= ~(ARM_SMMU_TCR_EPD0 | ARM_SMMU_TCR_EPD1);
+
+		cb->tcr[0] = tcr;
+		cb->ttbr[0] = pgtbl_cfg->arm_lpae_s1_cfg.ttbr;
+		cb->ttbr[0] |= FIELD_PREP(ARM_SMMU_TTBRn_ASID, cb->cfg->asid);
+	}
+
+	arm_smmu_write_context_bank(smmu_domain->smmu, cb->cfg->cbndx);
+
+	return 0;
+}
+
+static int qcom_adreno_smmu_alloc_context_bank(struct arm_smmu_domain *smmu_domain,
+		struct device *dev, int start, int count)
+{
+	struct arm_smmu_device *smmu = smmu_domain->smmu;
+
+	/*
+	 * Assign context bank 0 to the GPU device so the GPU hardware can
+	 * switch pagetables
+	 */
+	if (qcom_adreno_smmu_is_gpu_device(dev)) {
+		start = 0;
+		count = 1;
+	} else {
+		start = 1;
+	}
+
+	return __arm_smmu_alloc_bitmap(smmu->context_map, start, count);
+}
+
+static int qcom_adreno_smmu_init_context(struct arm_smmu_domain *smmu_domain,
+		struct io_pgtable_cfg *pgtbl_cfg, struct device *dev)
+{
+	struct adreno_smmu_priv *priv;
+
+	/* Only enable split pagetables for the GPU device (SID 0) */
+	if (!qcom_adreno_smmu_is_gpu_device(dev))
+		return 0;
+
+	/*
+	 * All targets that use the qcom,adreno-smmu compatible string *should*
+	 * be AARCH64 stage 1 but double check because the arm-smmu code assumes
+	 * that is the case when the TTBR1 quirk is enabled
+	 */
+	if ((smmu_domain->stage == ARM_SMMU_DOMAIN_S1) &&
+	    (smmu_domain->cfg.fmt == ARM_SMMU_CTX_FMT_AARCH64))
+		pgtbl_cfg->quirks |= IO_PGTABLE_QUIRK_ARM_TTBR1;
+
+	/*
+	 * Initialize private interface with GPU:
+	 */
+
+	priv = dev_get_drvdata(dev);
+	priv->cookie = smmu_domain;
+	priv->get_ttbr1_cfg = qcom_adreno_smmu_get_ttbr1_cfg;
+	priv->set_ttbr0_cfg = qcom_adreno_smmu_set_ttbr0_cfg;
+
+	return 0;
+}
+
 static const struct of_device_id qcom_smmu_client_of_match[] __maybe_unused = {
 	{ .compatible = "qcom,adreno" },
 	{ .compatible = "qcom,mdp4" },
@@ -65,7 +192,15 @@ static const struct arm_smmu_impl qcom_smmu_impl = {
 	.reset = qcom_smmu500_reset,
 };
 
-struct arm_smmu_device *qcom_smmu_impl_init(struct arm_smmu_device *smmu)
+static const struct arm_smmu_impl qcom_adreno_smmu_impl = {
+	.init_context = qcom_adreno_smmu_init_context,
+	.def_domain_type = qcom_smmu_def_domain_type,
+	.reset = qcom_smmu500_reset,
+	.alloc_context_bank = qcom_adreno_smmu_alloc_context_bank,
+};
+
+static struct arm_smmu_device *qcom_smmu_create(struct arm_smmu_device *smmu,
+		const struct arm_smmu_impl *impl)
 {
 	struct qcom_smmu *qsmmu;
 
@@ -75,8 +210,18 @@ struct arm_smmu_device *qcom_smmu_impl_init(struct arm_smmu_device *smmu)
 
 	qsmmu->smmu = *smmu;
 
-	qsmmu->smmu.impl = &qcom_smmu_impl;
+	qsmmu->smmu.impl = impl;
 	devm_kfree(smmu->dev, smmu);
 
 	return &qsmmu->smmu;
 }
+
+struct arm_smmu_device *qcom_smmu_impl_init(struct arm_smmu_device *smmu)
+{
+	return qcom_smmu_create(smmu, &qcom_smmu_impl);
+}
+
+struct arm_smmu_device *qcom_adreno_smmu_impl_init(struct arm_smmu_device *smmu)
+{
+	return qcom_smmu_create(smmu, &qcom_adreno_smmu_impl);
+}
drivers/iommu/arm/arm-smmu/arm-smmu.h +1 -0
index 27c83333fc50..cd75a33967bb 100644
@@ -519,6 +519,7 @@ static inline void arm_smmu_writeq(struct arm_smmu_device *smmu, int page,
 struct arm_smmu_device *arm_smmu_impl_init(struct arm_smmu_device *smmu);
 struct arm_smmu_device *nvidia_smmu_impl_init(struct arm_smmu_device *smmu);
 struct arm_smmu_device *qcom_smmu_impl_init(struct arm_smmu_device *smmu);
+struct arm_smmu_device *qcom_adreno_smmu_impl_init(struct arm_smmu_device *smmu);
 
 void arm_smmu_write_context_bank(struct arm_smmu_device *smmu, int idx);
 int arm_mmu500_reset(struct arm_smmu_device *smmu);
Signature · 1 line
2.26.2
#51 lore raw

PATCH 10/19 dt-bindings: arm-smmu: Add compatible string for Adreno GPU SMMU

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Every Qcom Adreno GPU has an embedded SMMU for its own use. These devices depend on unique features such as split pagetables, different stall/halt requirements and other settings. Identify them with a compatible string so that they can be identified in the arm-smmu implementation specific code.

---
 Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 4 ++++
 1 file changed, 4 insertions(+)
Documentation/devicetree/bindings/iommu/arm,smmu.yaml +4 -0
index 503160a7b9a0..5ec5d0d691f6 100644
@@ -40,6 +40,10 @@ properties:
               - qcom,sm8150-smmu-500
               - qcom,sm8250-smmu-500
           - const: arm,mmu-500
+      - description: Qcom Adreno GPUs implementing "arm,smmu-v2"
+        items:
+          - const: qcom,adreno-smmu
+          - const: qcom,smmu-v2
       - description: Marvell SoCs implementing "arm,mmu-500"
         items:
           - const: marvell,ap806-smmu-500
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Reviewed-by on v1 10/19
#52 lore raw

Re: [PATCH 10/19] dt-bindings: arm-smmu: Add compatible string for Adreno GPU SMMU

Quoted 8 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Every Qcom Adreno GPU has an embedded SMMU for its own use. These devices depend on unique features such as split pagetables, different stall/halt requirements and other settings. Identify them with a compatible string so that they can be identified in the arm-smmu implementation specific code.

Quoted 25 lines
---
 Documentation/devicetree/bindings/iommu/arm,smmu.yaml | 4 ++++
 1 file changed, 4 insertions(+)
Documentation/devicetree/bindings/iommu/arm,smmu.yaml +4 -0
index 503160a7b9a0..5ec5d0d691f6 100644
@@ -40,6 +40,10 @@ properties:
               - qcom,sm8150-smmu-500
               - qcom,sm8250-smmu-500
           - const: arm,mmu-500
+      - description: Qcom Adreno GPUs implementing "arm,smmu-v2"
+        items:
+          - const: qcom,adreno-smmu
+          - const: qcom,smmu-v2
       - description: Marvell SoCs implementing "arm,mmu-500"
         items:
           - const: marvell,ap806-smmu-500
Signature · 1 line
2.26.2
#53 lore raw

PATCH 11/19 drm/msm: Add a context pointer to the submitqueue

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Each submitqueue is attached to a context. Add a pointer to the context to the submitqueue at create time and refcount it so that it stays around through the life of the queue.

---
 drivers/gpu/drm/msm/msm_drv.c         |  3 ++-
 drivers/gpu/drm/msm/msm_drv.h         | 20 ++++++++++++++++++++
 drivers/gpu/drm/msm/msm_gem.h         |  1 +
 drivers/gpu/drm/msm/msm_gem_submit.c  |  6 +++---
 drivers/gpu/drm/msm/msm_gpu.h         |  1 +
 drivers/gpu/drm/msm/msm_submitqueue.c |  3 +++
 6 files changed, 30 insertions(+), 4 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +2 -1
index 7d641c7e3514..01845a3b8d52 100644
@@ -594,6 +594,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	if (!ctx)
 		return -ENOMEM;
 
+	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
 	ctx->aspace = priv->gpu ? priv->gpu->aspace : NULL;
@@ -615,7 +616,7 @@ static int msm_open(struct drm_device *dev, struct drm_file *file)
 static void context_close(struct msm_file_private *ctx)
 {
 	msm_submitqueue_close(ctx);
-	kfree(ctx);
+	msm_file_private_put(ctx);
 }
 
 static void msm_postclose(struct drm_device *dev, struct drm_file *file)
drivers/gpu/drm/msm/msm_drv.h +20 -0
index af259b0573ea..4561bfb5e745 100644
@@ -57,6 +57,7 @@ struct msm_file_private {
 	struct list_head submitqueues;
 	int queueid;
 	struct msm_gem_address_space *aspace;
+	struct kref ref;
 };
 
 enum msm_mdp_plane_property {
@@ -428,6 +429,25 @@ void msm_submitqueue_close(struct msm_file_private *ctx);
 
 void msm_submitqueue_destroy(struct kref *kref);
 
+static inline void __msm_file_private_destroy(struct kref *kref)
+{
+	struct msm_file_private *ctx = container_of(kref,
+		struct msm_file_private, ref);
+
+	kfree(ctx);
+}
+
+static inline void msm_file_private_put(struct msm_file_private *ctx)
+{
+	kref_put(&ctx->ref, __msm_file_private_destroy);
+}
+
+static inline struct msm_file_private *msm_file_private_get(
+	struct msm_file_private *ctx)
+{
+	kref_get(&ctx->ref);
+	return ctx;
+}
 
 #define DBG(fmt, ...) DRM_DEBUG_DRIVER(fmt"\n", ##__VA_ARGS__)
 #define VERB(fmt, ...) if (0) DRM_DEBUG_DRIVER(fmt"\n", ##__VA_ARGS__)
drivers/gpu/drm/msm/msm_gem.h +1 -0
index 972490b14ba5..9c573c4269cb 100644
@@ -142,6 +142,7 @@ struct msm_gem_submit {
 	bool valid;         /* true if no cmdstream patching needed */
 	bool in_rb;         /* "sudo" mode, copy cmds into RB */
 	struct msm_ringbuffer *ring;
+	struct msm_file_private *ctx;
 	unsigned int nr_cmds;
 	unsigned int nr_bos;
 	u32 ident;	   /* A "identifier" for the submit for logging */
drivers/gpu/drm/msm/msm_gem_submit.c +3 -3
index 8cb9aa15ff90..1464b04d25d3 100644
@@ -27,7 +27,7 @@
 #define BO_PINNED   0x2000
 
 static struct msm_gem_submit *submit_create(struct drm_device *dev,
-		struct msm_gpu *gpu, struct msm_gem_address_space *aspace,
+		struct msm_gpu *gpu,
 		struct msm_gpu_submitqueue *queue, uint32_t nr_bos,
 		uint32_t nr_cmds)
 {
@@ -43,7 +43,7 @@ static struct msm_gem_submit *submit_create(struct drm_device *dev,
 		return NULL;
 
 	submit->dev = dev;
-	submit->aspace = aspace;
+	submit->aspace = queue->ctx->aspace;
 	submit->gpu = gpu;
 	submit->fence = NULL;
 	submit->cmd = (void *)&submit->bos[nr_bos];
@@ -677,7 +677,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		}
 	}
 
-	submit = submit_create(dev, gpu, ctx->aspace, queue, args->nr_bos,
+	submit = submit_create(dev, gpu, queue, args->nr_bos,
 		args->nr_cmds);
 	if (!submit) {
 		ret = -ENOMEM;
drivers/gpu/drm/msm/msm_gpu.h +1 -0
index f91b141add75..97c527e98391 100644
@@ -190,6 +190,7 @@ struct msm_gpu_submitqueue {
 	u32 flags;
 	u32 prio;
 	int faults;
+	struct msm_file_private *ctx;
 	struct list_head node;
 	struct kref ref;
 };
drivers/gpu/drm/msm/msm_submitqueue.c +3 -0
index 90c9d84e6155..c3d206105d28 100644
@@ -12,6 +12,8 @@ void msm_submitqueue_destroy(struct kref *kref)
 	struct msm_gpu_submitqueue *queue = container_of(kref,
 		struct msm_gpu_submitqueue, ref);
 
+	msm_file_private_put(queue->ctx);
+
 	kfree(queue);
 }
 
@@ -83,6 +85,7 @@ int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
 
 	write_lock(&ctx->queuelock);
 
+	queue->ctx = msm_file_private_get(ctx);
 	queue->id = ctx->queueid++;
 
 	if (id)
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Reviewed-by on v1 11/19
#54 lore raw

Re: [PATCH 11/19] drm/msm: Add a context pointer to the submitqueue

Quoted 6 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Each submitqueue is attached to a context. Add a pointer to the context to the submitqueue at create time and refcount it so that it stays around through the life of the queue.

Quoted 150 lines
---
 drivers/gpu/drm/msm/msm_drv.c         |  3 ++-
 drivers/gpu/drm/msm/msm_drv.h         | 20 ++++++++++++++++++++
 drivers/gpu/drm/msm/msm_gem.h         |  1 +
 drivers/gpu/drm/msm/msm_gem_submit.c  |  6 +++---
 drivers/gpu/drm/msm/msm_gpu.h         |  1 +
 drivers/gpu/drm/msm/msm_submitqueue.c |  3 +++
 6 files changed, 30 insertions(+), 4 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +2 -1
index 7d641c7e3514..01845a3b8d52 100644
@@ -594,6 +594,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	if (!ctx)
 		return -ENOMEM;
 
+	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
 	ctx->aspace = priv->gpu ? priv->gpu->aspace : NULL;
@@ -615,7 +616,7 @@ static int msm_open(struct drm_device *dev, struct drm_file *file)
 static void context_close(struct msm_file_private *ctx)
 {
 	msm_submitqueue_close(ctx);
-	kfree(ctx);
+	msm_file_private_put(ctx);
 }
 
 static void msm_postclose(struct drm_device *dev, struct drm_file *file)
drivers/gpu/drm/msm/msm_drv.h +20 -0
index af259b0573ea..4561bfb5e745 100644
@@ -57,6 +57,7 @@ struct msm_file_private {
 	struct list_head submitqueues;
 	int queueid;
 	struct msm_gem_address_space *aspace;
+	struct kref ref;
 };
 
 enum msm_mdp_plane_property {
@@ -428,6 +429,25 @@ void msm_submitqueue_close(struct msm_file_private *ctx);
 
 void msm_submitqueue_destroy(struct kref *kref);
 
+static inline void __msm_file_private_destroy(struct kref *kref)
+{
+	struct msm_file_private *ctx = container_of(kref,
+		struct msm_file_private, ref);
+
+	kfree(ctx);
+}
+
+static inline void msm_file_private_put(struct msm_file_private *ctx)
+{
+	kref_put(&ctx->ref, __msm_file_private_destroy);
+}
+
+static inline struct msm_file_private *msm_file_private_get(
+	struct msm_file_private *ctx)
+{
+	kref_get(&ctx->ref);
+	return ctx;
+}
 
 #define DBG(fmt, ...) DRM_DEBUG_DRIVER(fmt"\n", ##__VA_ARGS__)
 #define VERB(fmt, ...) if (0) DRM_DEBUG_DRIVER(fmt"\n", ##__VA_ARGS__)
drivers/gpu/drm/msm/msm_gem.h +1 -0
index 972490b14ba5..9c573c4269cb 100644
@@ -142,6 +142,7 @@ struct msm_gem_submit {
 	bool valid;         /* true if no cmdstream patching needed */
 	bool in_rb;         /* "sudo" mode, copy cmds into RB */
 	struct msm_ringbuffer *ring;
+	struct msm_file_private *ctx;
 	unsigned int nr_cmds;
 	unsigned int nr_bos;
 	u32 ident;	   /* A "identifier" for the submit for logging */
drivers/gpu/drm/msm/msm_gem_submit.c +3 -3
index 8cb9aa15ff90..1464b04d25d3 100644
@@ -27,7 +27,7 @@
 #define BO_PINNED   0x2000
 
 static struct msm_gem_submit *submit_create(struct drm_device *dev,
-		struct msm_gpu *gpu, struct msm_gem_address_space *aspace,
+		struct msm_gpu *gpu,
 		struct msm_gpu_submitqueue *queue, uint32_t nr_bos,
 		uint32_t nr_cmds)
 {
@@ -43,7 +43,7 @@ static struct msm_gem_submit *submit_create(struct drm_device *dev,
 		return NULL;
 
 	submit->dev = dev;
-	submit->aspace = aspace;
+	submit->aspace = queue->ctx->aspace;
 	submit->gpu = gpu;
 	submit->fence = NULL;
 	submit->cmd = (void *)&submit->bos[nr_bos];
@@ -677,7 +677,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		}
 	}
 
-	submit = submit_create(dev, gpu, ctx->aspace, queue, args->nr_bos,
+	submit = submit_create(dev, gpu, queue, args->nr_bos,
 		args->nr_cmds);
 	if (!submit) {
 		ret = -ENOMEM;
drivers/gpu/drm/msm/msm_gpu.h +1 -0
index f91b141add75..97c527e98391 100644
@@ -190,6 +190,7 @@ struct msm_gpu_submitqueue {
 	u32 flags;
 	u32 prio;
 	int faults;
+	struct msm_file_private *ctx;
 	struct list_head node;
 	struct kref ref;
 };
drivers/gpu/drm/msm/msm_submitqueue.c +3 -0
index 90c9d84e6155..c3d206105d28 100644
@@ -12,6 +12,8 @@ void msm_submitqueue_destroy(struct kref *kref)
 	struct msm_gpu_submitqueue *queue = container_of(kref,
 		struct msm_gpu_submitqueue, ref);
 
+	msm_file_private_put(queue->ctx);
+
 	kfree(queue);
 }
 
@@ -83,6 +85,7 @@ int msm_submitqueue_create(struct drm_device *drm, struct msm_file_private *ctx,
 
 	write_lock(&ctx->queuelock);
 
+	queue->ctx = msm_file_private_get(ctx);
 	queue->id = ctx->queueid++;
 
 	if (id)
Signature · 1 line
2.26.2
#55 lore raw

PATCH 12/19 drm/msm: Drop context arg to gpu->submit()

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Now that we can get the ctx from the submitqueue, the extra arg is redundant.

[split out of previous patch to reduce churny noise]
---
 drivers/gpu/drm/msm/adreno/a5xx_gpu.c   | 12 +++++-------
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c   |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.c |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.h |  3 +--
 drivers/gpu/drm/msm/msm_gem_submit.c    |  2 +-
 drivers/gpu/drm/msm/msm_gpu.c           |  9 ++++-----
 drivers/gpu/drm/msm/msm_gpu.h           |  6 ++----
 7 files changed, 17 insertions(+), 25 deletions(-)
drivers/gpu/drm/msm/adreno/a5xx_gpu.c +5 -7
index 9e63a190642c..eff2439ea57b 100644
@@ -43,8 +43,7 @@ static void a5xx_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring)
 		gpu_write(gpu, REG_A5XX_CP_RB_WPTR, wptr);
 }
 
-static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct msm_drm_private *priv = gpu->dev->dev_private;
 	struct msm_ringbuffer *ring = submit->ring;
@@ -57,7 +56,7 @@ static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
@@ -103,8 +102,7 @@ static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
 	msm_gpu_retire(gpu);
 }
 
-static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
 	struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu);
@@ -114,7 +112,7 @@ static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 
 	if (IS_ENABLED(CONFIG_DRM_MSM_GPU_SUDO) && submit->in_rb) {
 		priv->lastctx = NULL;
-		a5xx_submit_in_rb(gpu, submit, ctx);
+		a5xx_submit_in_rb(gpu, submit);
 		return;
 	}
 
@@ -148,7 +146,7 @@ static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +2 -3
index c5a3e4d4c007..5eabb0109577 100644
@@ -81,8 +81,7 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
 	OUT_RING(ring, upper_32_bits(iova));
 }
 
-static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
 	struct msm_drm_private *priv = gpu->dev->dev_private;
@@ -115,7 +114,7 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/adreno_gpu.c +2 -3
index d2dbb6968cba..533a34b4cce2 100644
@@ -457,8 +457,7 @@ void adreno_recover(struct msm_gpu *gpu)
 	}
 }
 
-void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx)
+void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
 	struct msm_drm_private *priv = gpu->dev->dev_private;
@@ -472,7 +471,7 @@ void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
 			/* ignore if there has not been a ctx switch: */
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/adreno_gpu.h +1 -2
index e55abae365b5..848632758450 100644
@@ -269,8 +269,7 @@ struct drm_gem_object *adreno_fw_create_bo(struct msm_gpu *gpu,
 		const struct firmware *fw, u64 *iova);
 int adreno_hw_init(struct msm_gpu *gpu);
 void adreno_recover(struct msm_gpu *gpu);
-void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx);
+void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 void adreno_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 bool adreno_idle(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 #if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP)
drivers/gpu/drm/msm/msm_gem_submit.c +1 -1
index 1464b04d25d3..aa5c60a7132d 100644
@@ -785,7 +785,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		}
 	}
 
-	msm_gpu_submit(gpu, submit, ctx);
+	msm_gpu_submit(gpu, submit);
 
 	args->fence = submit->fence->seqno;
 
drivers/gpu/drm/msm/msm_gpu.c +4 -5
index 806eb0957280..e1a3cbe25a0c 100644
@@ -520,7 +520,7 @@ static void recover_worker(struct work_struct *work)
 			struct msm_ringbuffer *ring = gpu->rb[i];
 
 			list_for_each_entry(submit, &ring->submits, node)
-				gpu->funcs->submit(gpu, submit, NULL);
+				gpu->funcs->submit(gpu, submit);
 		}
 	}
 
@@ -747,8 +747,7 @@ void msm_gpu_retire(struct msm_gpu *gpu)
 }
 
 /* add bo's to gpu's ring, and kick gpu: */
-void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx)
+void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct drm_device *dev = gpu->dev;
 	struct msm_drm_private *priv = dev->dev_private;
@@ -788,8 +787,8 @@ void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 			msm_gem_move_to_active(&msm_obj->base, gpu, false, submit->fence);
 	}
 
-	gpu->funcs->submit(gpu, submit, ctx);
-	priv->lastctx = ctx;
+	gpu->funcs->submit(gpu, submit);
+	priv->lastctx = submit->queue->ctx;
 
 	hangcheck_timer_reset(gpu);
 }
drivers/gpu/drm/msm/msm_gpu.h +2 -4
index 97c527e98391..1f96ac0d9049 100644
@@ -45,8 +45,7 @@ struct msm_gpu_funcs {
 	int (*hw_init)(struct msm_gpu *gpu);
 	int (*pm_suspend)(struct msm_gpu *gpu);
 	int (*pm_resume)(struct msm_gpu *gpu);
-	void (*submit)(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-			struct msm_file_private *ctx);
+	void (*submit)(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 	void (*flush)(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 	irqreturn_t (*irq)(struct msm_gpu *irq);
 	struct msm_ringbuffer *(*active_ring)(struct msm_gpu *gpu);
@@ -290,8 +289,7 @@ int msm_gpu_perfcntr_sample(struct msm_gpu *gpu, uint32_t *activetime,
 		uint32_t *totaltime, uint32_t ncntrs, uint32_t *cntrs);
 
 void msm_gpu_retire(struct msm_gpu *gpu);
-void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx);
+void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 
 int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
Signature · 1 line
2.26.2
Reviewed-by on v1 12/19
#56 lore raw

Re: [PATCH 12/19] drm/msm: Drop context arg to gpu->submit()

Quoted 5 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Now that we can get the ctx from the submitqueue, the extra arg is redundant.

Quoted 198 lines
[split out of previous patch to reduce churny noise]
---
 drivers/gpu/drm/msm/adreno/a5xx_gpu.c   | 12 +++++-------
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c   |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.c |  5 ++---
 drivers/gpu/drm/msm/adreno/adreno_gpu.h |  3 +--
 drivers/gpu/drm/msm/msm_gem_submit.c    |  2 +-
 drivers/gpu/drm/msm/msm_gpu.c           |  9 ++++-----
 drivers/gpu/drm/msm/msm_gpu.h           |  6 ++----
 7 files changed, 17 insertions(+), 25 deletions(-)
drivers/gpu/drm/msm/adreno/a5xx_gpu.c +5 -7
index 9e63a190642c..eff2439ea57b 100644
@@ -43,8 +43,7 @@ static void a5xx_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring)
 		gpu_write(gpu, REG_A5XX_CP_RB_WPTR, wptr);
 }
 
-static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct msm_drm_private *priv = gpu->dev->dev_private;
 	struct msm_ringbuffer *ring = submit->ring;
@@ -57,7 +56,7 @@ static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
@@ -103,8 +102,7 @@ static void a5xx_submit_in_rb(struct msm_gpu *gpu, struct msm_gem_submit *submit
 	msm_gpu_retire(gpu);
 }
 
-static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
 	struct a5xx_gpu *a5xx_gpu = to_a5xx_gpu(adreno_gpu);
@@ -114,7 +112,7 @@ static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 
 	if (IS_ENABLED(CONFIG_DRM_MSM_GPU_SUDO) && submit->in_rb) {
 		priv->lastctx = NULL;
-		a5xx_submit_in_rb(gpu, submit, ctx);
+		a5xx_submit_in_rb(gpu, submit);
 		return;
 	}
 
@@ -148,7 +146,7 @@ static void a5xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +2 -3
index c5a3e4d4c007..5eabb0109577 100644
@@ -81,8 +81,7 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
 	OUT_RING(ring, upper_32_bits(iova));
 }
 
-static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-	struct msm_file_private *ctx)
+static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
 	struct msm_drm_private *priv = gpu->dev->dev_private;
@@ -115,7 +114,7 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/adreno_gpu.c +2 -3
index d2dbb6968cba..533a34b4cce2 100644
@@ -457,8 +457,7 @@ void adreno_recover(struct msm_gpu *gpu)
 	}
 }
 
-void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx)
+void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
 	struct msm_drm_private *priv = gpu->dev->dev_private;
@@ -472,7 +471,7 @@ void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 			break;
 		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
 			/* ignore if there has not been a ctx switch: */
-			if (priv->lastctx == ctx)
+			if (priv->lastctx == submit->queue->ctx)
 				break;
 			/* fall-thru */
 		case MSM_SUBMIT_CMD_BUF:
drivers/gpu/drm/msm/adreno/adreno_gpu.h +1 -2
index e55abae365b5..848632758450 100644
@@ -269,8 +269,7 @@ struct drm_gem_object *adreno_fw_create_bo(struct msm_gpu *gpu,
 		const struct firmware *fw, u64 *iova);
 int adreno_hw_init(struct msm_gpu *gpu);
 void adreno_recover(struct msm_gpu *gpu);
-void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx);
+void adreno_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 void adreno_flush(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 bool adreno_idle(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 #if defined(CONFIG_DEBUG_FS) || defined(CONFIG_DEV_COREDUMP)
drivers/gpu/drm/msm/msm_gem_submit.c +1 -1
index 1464b04d25d3..aa5c60a7132d 100644
@@ -785,7 +785,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		}
 	}
 
-	msm_gpu_submit(gpu, submit, ctx);
+	msm_gpu_submit(gpu, submit);
 
 	args->fence = submit->fence->seqno;
 
drivers/gpu/drm/msm/msm_gpu.c +4 -5
index 806eb0957280..e1a3cbe25a0c 100644
@@ -520,7 +520,7 @@ static void recover_worker(struct work_struct *work)
 			struct msm_ringbuffer *ring = gpu->rb[i];
 
 			list_for_each_entry(submit, &ring->submits, node)
-				gpu->funcs->submit(gpu, submit, NULL);
+				gpu->funcs->submit(gpu, submit);
 		}
 	}
 
@@ -747,8 +747,7 @@ void msm_gpu_retire(struct msm_gpu *gpu)
 }
 
 /* add bo's to gpu's ring, and kick gpu: */
-void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx)
+void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	struct drm_device *dev = gpu->dev;
 	struct msm_drm_private *priv = dev->dev_private;
@@ -788,8 +787,8 @@ void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 			msm_gem_move_to_active(&msm_obj->base, gpu, false, submit->fence);
 	}
 
-	gpu->funcs->submit(gpu, submit, ctx);
-	priv->lastctx = ctx;
+	gpu->funcs->submit(gpu, submit);
+	priv->lastctx = submit->queue->ctx;
 
 	hangcheck_timer_reset(gpu);
 }
drivers/gpu/drm/msm/msm_gpu.h +2 -4
index 97c527e98391..1f96ac0d9049 100644
@@ -45,8 +45,7 @@ struct msm_gpu_funcs {
 	int (*hw_init)(struct msm_gpu *gpu);
 	int (*pm_suspend)(struct msm_gpu *gpu);
 	int (*pm_resume)(struct msm_gpu *gpu);
-	void (*submit)(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-			struct msm_file_private *ctx);
+	void (*submit)(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 	void (*flush)(struct msm_gpu *gpu, struct msm_ringbuffer *ring);
 	irqreturn_t (*irq)(struct msm_gpu *irq);
 	struct msm_ringbuffer *(*active_ring)(struct msm_gpu *gpu);
@@ -290,8 +289,7 @@ int msm_gpu_perfcntr_sample(struct msm_gpu *gpu, uint32_t *activetime,
 		uint32_t *totaltime, uint32_t ncntrs, uint32_t *cntrs);
 
 void msm_gpu_retire(struct msm_gpu *gpu);
-void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
-		struct msm_file_private *ctx);
+void msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit);
 
 int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
Signature · 1 line
2.26.2
#57 lore raw

PATCH 13/19 drm/msm: Set the global virtual address range from the IOMMU domain

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Use the aperture settings from the IOMMU domain to set up the virtual address range for the GPU. This allows us to transparently deal with IOMMU side features (like split pagetables).

---
 drivers/gpu/drm/msm/adreno/adreno_gpu.c | 13 +++++++++++--
 drivers/gpu/drm/msm/msm_iommu.c         |  7 +++++++
 2 files changed, 18 insertions(+), 2 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_gpu.c +11 -2
index 533a34b4cce2..34e6242c1767 100644
@@ -192,9 +192,18 @@ adreno_iommu_create_address_space(struct msm_gpu *gpu,
 	struct iommu_domain *iommu = iommu_domain_alloc(&platform_bus_type);
 	struct msm_mmu *mmu = msm_iommu_new(&pdev->dev, iommu);
 	struct msm_gem_address_space *aspace;
+	u64 start, size;
 
-	aspace = msm_gem_address_space_create(mmu, "gpu", SZ_16M,
-		0xffffffff - SZ_16M);
+	/*
+	 * Use the aperture start or SZ_16M, whichever is greater. This will
+	 * ensure that we align with the allocated pagetable range while still
+	 * allowing room in the lower 32 bits for GMEM and whatnot
+	 */
+	start = max_t(u64, SZ_16M, iommu->geometry.aperture_start);
+	size = iommu->geometry.aperture_end - start + 1;
+
+	aspace = msm_gem_address_space_create(mmu, "gpu",
+		start & GENMASK(48, 0), size);
 
 	if (IS_ERR(aspace) && !IS_ERR(mmu))
 		mmu->funcs->destroy(mmu);
drivers/gpu/drm/msm/msm_iommu.c +7 -0
index 3a381a9674c9..1b6635504069 100644
@@ -36,6 +36,10 @@ static int msm_iommu_map(struct msm_mmu *mmu, uint64_t iova,
 	struct msm_iommu *iommu = to_msm_iommu(mmu);
 	size_t ret;
 
+	/* The arm-smmu driver expects the addresses to be sign extended */
+	if (iova & BIT_ULL(48))
+		iova |= GENMASK_ULL(63, 49);
+
 	ret = iommu_map_sg(iommu->domain, iova, sgt->sgl, sgt->nents, prot);
 	WARN_ON(!ret);
 
@@ -46,6 +50,9 @@ static int msm_iommu_unmap(struct msm_mmu *mmu, uint64_t iova, size_t len)
 {
 	struct msm_iommu *iommu = to_msm_iommu(mmu);
 
+	if (iova & BIT_ULL(48))
+		iova |= GENMASK_ULL(63, 49);
+
 	iommu_unmap(iommu->domain, iova, len);
 
 	return 0;
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Reviewed-by on v1 13/19
#58 lore raw

Re: [PATCH 13/19] drm/msm: Set the global virtual address range from the IOMMU domain

Quoted 6 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Use the aperture settings from the IOMMU domain to set up the virtual address range for the GPU. This allows us to transparently deal with IOMMU side features (like split pagetables).

Quoted 60 lines
---
 drivers/gpu/drm/msm/adreno/adreno_gpu.c | 13 +++++++++++--
 drivers/gpu/drm/msm/msm_iommu.c         |  7 +++++++
 2 files changed, 18 insertions(+), 2 deletions(-)
drivers/gpu/drm/msm/adreno/adreno_gpu.c +11 -2
index 533a34b4cce2..34e6242c1767 100644
@@ -192,9 +192,18 @@ adreno_iommu_create_address_space(struct msm_gpu *gpu,
 	struct iommu_domain *iommu = iommu_domain_alloc(&platform_bus_type);
 	struct msm_mmu *mmu = msm_iommu_new(&pdev->dev, iommu);
 	struct msm_gem_address_space *aspace;
+	u64 start, size;
 
-	aspace = msm_gem_address_space_create(mmu, "gpu", SZ_16M,
-		0xffffffff - SZ_16M);
+	/*
+	 * Use the aperture start or SZ_16M, whichever is greater. This will
+	 * ensure that we align with the allocated pagetable range while still
+	 * allowing room in the lower 32 bits for GMEM and whatnot
+	 */
+	start = max_t(u64, SZ_16M, iommu->geometry.aperture_start);
+	size = iommu->geometry.aperture_end - start + 1;
+
+	aspace = msm_gem_address_space_create(mmu, "gpu",
+		start & GENMASK(48, 0), size);
 
 	if (IS_ERR(aspace) && !IS_ERR(mmu))
 		mmu->funcs->destroy(mmu);
drivers/gpu/drm/msm/msm_iommu.c +7 -0
index 3a381a9674c9..1b6635504069 100644
@@ -36,6 +36,10 @@ static int msm_iommu_map(struct msm_mmu *mmu, uint64_t iova,
 	struct msm_iommu *iommu = to_msm_iommu(mmu);
 	size_t ret;
 
+	/* The arm-smmu driver expects the addresses to be sign extended */
+	if (iova & BIT_ULL(48))
+		iova |= GENMASK_ULL(63, 49);
+
 	ret = iommu_map_sg(iommu->domain, iova, sgt->sgl, sgt->nents, prot);
 	WARN_ON(!ret);
 
@@ -46,6 +50,9 @@ static int msm_iommu_unmap(struct msm_mmu *mmu, uint64_t iova, size_t len)
 {
 	struct msm_iommu *iommu = to_msm_iommu(mmu);
 
+	if (iova & BIT_ULL(48))
+		iova |= GENMASK_ULL(63, 49);
+
 	iommu_unmap(iommu->domain, iova, len);
 
 	return 0;
Signature · 1 line
2.26.2
#59 lore raw

PATCH 14/19 drm/msm: Add support to create a local pagetable

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Add support to create a io-pgtable for use by targets that support per-instance pagetables. In order to support per-instance pagetables the GPU SMMU device needs to have the qcom,adreno-smmu compatible string and split pagetables enabled.

---
 drivers/gpu/drm/msm/msm_gpummu.c |   2 +-
 drivers/gpu/drm/msm/msm_iommu.c  | 199 ++++++++++++++++++++++++++++++-
 drivers/gpu/drm/msm/msm_mmu.h    |  16 ++-
 3 files changed, 214 insertions(+), 3 deletions(-)
drivers/gpu/drm/msm/msm_gpummu.c +1 -1
index 310a31b05faa..aab121f4beb7 100644
@@ -102,7 +102,7 @@ struct msm_mmu *msm_gpummu_new(struct device *dev, struct msm_gpu *gpu)
 	}
 
 	gpummu->gpu = gpu;
-	msm_mmu_init(&gpummu->base, dev, &funcs);
+	msm_mmu_init(&gpummu->base, dev, &funcs, MSM_MMU_GPUMMU);
 
 	return &gpummu->base;
 }
drivers/gpu/drm/msm/msm_iommu.c +198 -1

222 lines not on this page · raw message

drivers/gpu/drm/msm/msm_mmu.h +15 -1
index 3a534ee59bf6..61ade89d9e48 100644
@@ -17,18 +17,26 @@ struct msm_mmu_funcs {
 	void (*destroy)(struct msm_mmu *mmu);
 };
 
+enum msm_mmu_type {
+	MSM_MMU_GPUMMU,
+	MSM_MMU_IOMMU,
+	MSM_MMU_IOMMU_PAGETABLE,
+};
+
 struct msm_mmu {
 	const struct msm_mmu_funcs *funcs;
 	struct device *dev;
 	int (*handler)(void *arg, unsigned long iova, int flags);
 	void *arg;
+	enum msm_mmu_type type;
 };
 
 static inline void msm_mmu_init(struct msm_mmu *mmu, struct device *dev,
-		const struct msm_mmu_funcs *funcs)
+		const struct msm_mmu_funcs *funcs, enum msm_mmu_type type)
 {
 	mmu->dev = dev;
 	mmu->funcs = funcs;
+	mmu->type = type;
 }
 
 struct msm_mmu *msm_iommu_new(struct device *dev, struct iommu_domain *domain);
@@ -41,7 +49,13 @@ static inline void msm_mmu_set_fault_handler(struct msm_mmu *mmu, void *arg,
 	mmu->handler = handler;
 }
 
+struct msm_mmu *msm_iommu_pagetable_create(struct msm_mmu *parent);
+
 void msm_gpummu_params(struct msm_mmu *mmu, dma_addr_t *pt_base,
 		dma_addr_t *tran_error);
 
+
+int msm_iommu_pagetable_params(struct msm_mmu *mmu, phys_addr_t *ttbr,
+		int *asid);
+
 #endif /* __MSM_MMU_H__ */
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
Reviewed-by on v1 14/19
#60 lore raw

Re: [PATCH 14/19] drm/msm: Add support to create a local pagetable

Quoted 7 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Add support to create a io-pgtable for use by targets that support per-instance pagetables. In order to support per-instance pagetables the GPU SMMU device needs to have the qcom,adreno-smmu compatible string and split pagetables enabled.

Quoted 298 lines
---
 drivers/gpu/drm/msm/msm_gpummu.c |   2 +-
 drivers/gpu/drm/msm/msm_iommu.c  | 199 ++++++++++++++++++++++++++++++-
 drivers/gpu/drm/msm/msm_mmu.h    |  16 ++-
 3 files changed, 214 insertions(+), 3 deletions(-)
drivers/gpu/drm/msm/msm_gpummu.c +1 -1
index 310a31b05faa..aab121f4beb7 100644
@@ -102,7 +102,7 @@ struct msm_mmu *msm_gpummu_new(struct device *dev, struct msm_gpu *gpu)
 	}
 
 	gpummu->gpu = gpu;
-	msm_mmu_init(&gpummu->base, dev, &funcs);
+	msm_mmu_init(&gpummu->base, dev, &funcs, MSM_MMU_GPUMMU);
 
 	return &gpummu->base;
 }
drivers/gpu/drm/msm/msm_iommu.c +198 -1
index 1b6635504069..697cc0a059d6 100644
@@ -4,15 +4,210 @@
  * Author: Rob Clark <[email protected]>
  */
 
+#include <linux/adreno-smmu-priv.h>
+#include <linux/io-pgtable.h>
 #include "msm_drv.h"
 #include "msm_mmu.h"
 
 struct msm_iommu {
 	struct msm_mmu base;
 	struct iommu_domain *domain;
+	atomic_t pagetables;
 };
+
 #define to_msm_iommu(x) container_of(x, struct msm_iommu, base)
 
+struct msm_iommu_pagetable {
+	struct msm_mmu base;
+	struct msm_mmu *parent;
+	struct io_pgtable_ops *pgtbl_ops;
+	phys_addr_t ttbr;
+	u32 asid;
+};
+static struct msm_iommu_pagetable *to_pagetable(struct msm_mmu *mmu)
+{
+	return container_of(mmu, struct msm_iommu_pagetable, base);
+}
+
+static int msm_iommu_pagetable_unmap(struct msm_mmu *mmu, u64 iova,
+		size_t size)
+{
+	struct msm_iommu_pagetable *pagetable = to_pagetable(mmu);
+	struct io_pgtable_ops *ops = pagetable->pgtbl_ops;
+	size_t unmapped = 0;
+
+	/* Unmap the block one page at a time */
+	while (size) {
+		unmapped += ops->unmap(ops, iova, 4096, NULL);
+		iova += 4096;
+		size -= 4096;
+	}
+
+	iommu_flush_tlb_all(to_msm_iommu(pagetable->parent)->domain);
+
+	return (unmapped == size) ? 0 : -EINVAL;
+}
+
+static int msm_iommu_pagetable_map(struct msm_mmu *mmu, u64 iova,
+		struct sg_table *sgt, size_t len, int prot)
+{
+	struct msm_iommu_pagetable *pagetable = to_pagetable(mmu);
+	struct io_pgtable_ops *ops = pagetable->pgtbl_ops;
+	struct scatterlist *sg;
+	size_t mapped = 0;
+	u64 addr = iova;
+	unsigned int i;
+
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		size_t size = sg->length;
+		phys_addr_t phys = sg_phys(sg);
+
+		/* Map the block one page at a time */
+		while (size) {
+			if (ops->map(ops, addr, phys, 4096, prot, GFP_KERNEL)) {
+				msm_iommu_pagetable_unmap(mmu, iova, mapped);
+				return -EINVAL;
+			}
+
+			phys += 4096;
+			addr += 4096;
+			size -= 4096;
+			mapped += 4096;
+		}
+	}
+
+	return 0;
+}
+
+static void msm_iommu_pagetable_destroy(struct msm_mmu *mmu)
+{
+	struct msm_iommu_pagetable *pagetable = to_pagetable(mmu);
+	struct msm_iommu *iommu = to_msm_iommu(pagetable->parent);
+	struct adreno_smmu_priv *adreno_smmu =
+		dev_get_drvdata(pagetable->parent->dev);
+
+	/*
+	 * If this is the last attached pagetable for the parent,
+	 * disable TTBR0 in the arm-smmu driver
+	 */
+	if (atomic_dec_return(&iommu->pagetables) == 0)
+		adreno_smmu->set_ttbr0_cfg(adreno_smmu->cookie, NULL);
+
+	free_io_pgtable_ops(pagetable->pgtbl_ops);
+	kfree(pagetable);
+}
+
+int msm_iommu_pagetable_params(struct msm_mmu *mmu,
+		phys_addr_t *ttbr, int *asid)
+{
+	struct msm_iommu_pagetable *pagetable;
+
+	if (mmu->type != MSM_MMU_IOMMU_PAGETABLE)
+		return -EINVAL;
+
+	pagetable = to_pagetable(mmu);
+
+	if (ttbr)
+		*ttbr = pagetable->ttbr;
+
+	if (asid)
+		*asid = pagetable->asid;
+
+	return 0;
+}
+
+static const struct msm_mmu_funcs pagetable_funcs = {
+		.map = msm_iommu_pagetable_map,
+		.unmap = msm_iommu_pagetable_unmap,
+		.destroy = msm_iommu_pagetable_destroy,
+};
+
+static void msm_iommu_tlb_flush_all(void *cookie)
+{
+}
+
+static void msm_iommu_tlb_flush_walk(unsigned long iova, size_t size,
+		size_t granule, void *cookie)
+{
+}
+
+static void msm_iommu_tlb_add_page(struct iommu_iotlb_gather *gather,
+		unsigned long iova, size_t granule, void *cookie)
+{
+}
+
+static const struct iommu_flush_ops null_tlb_ops = {
+	.tlb_flush_all = msm_iommu_tlb_flush_all,
+	.tlb_flush_walk = msm_iommu_tlb_flush_walk,
+	.tlb_flush_leaf = msm_iommu_tlb_flush_walk,
+	.tlb_add_page = msm_iommu_tlb_add_page,
+};
+
+struct msm_mmu *msm_iommu_pagetable_create(struct msm_mmu *parent)
+{
+	struct adreno_smmu_priv *adreno_smmu = dev_get_drvdata(parent->dev);
+	struct msm_iommu *iommu = to_msm_iommu(parent);
+	struct msm_iommu_pagetable *pagetable;
+	const struct io_pgtable_cfg *ttbr1_cfg = NULL;
+	struct io_pgtable_cfg ttbr0_cfg;
+	int ret;
+
+	/* Get the pagetable configuration from the domain */
+	if (adreno_smmu->cookie)
+		ttbr1_cfg = adreno_smmu->get_ttbr1_cfg(adreno_smmu->cookie);
+	if (!ttbr1_cfg)
+		return ERR_PTR(-ENODEV);
+
+	pagetable = kzalloc(sizeof(*pagetable), GFP_KERNEL);
+	if (!pagetable)
+		return ERR_PTR(-ENOMEM);
+
+	msm_mmu_init(&pagetable->base, parent->dev, &pagetable_funcs,
+		MSM_MMU_IOMMU_PAGETABLE);
+
+	/* Clone the TTBR1 cfg as starting point for TTBR0 cfg: */
+	ttbr0_cfg = *ttbr1_cfg;
+
+	/* The incoming cfg will have the TTBR1 quirk enabled */
+	ttbr0_cfg.quirks &= ~IO_PGTABLE_QUIRK_ARM_TTBR1;
+	ttbr0_cfg.tlb = &null_tlb_ops;
+
+	pagetable->pgtbl_ops = alloc_io_pgtable_ops(ARM_64_LPAE_S1,
+		&ttbr0_cfg, iommu->domain);
+
+	if (!pagetable->pgtbl_ops) {
+		kfree(pagetable);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	/*
+	 * If this is the first pagetable that we've allocated, send it back to
+	 * the arm-smmu driver as a trigger to set up TTBR0
+	 */
+	if (atomic_inc_return(&iommu->pagetables) == 1) {
+		ret = adreno_smmu->set_ttbr0_cfg(adreno_smmu->cookie, &ttbr0_cfg);
+		if (ret) {
+			free_io_pgtable_ops(pagetable->pgtbl_ops);
+			kfree(pagetable);
+			return ERR_PTR(ret);
+		}
+	}
+
+	/* Needed later for TLB flush */
+	pagetable->parent = parent;
+	pagetable->ttbr = ttbr0_cfg.arm_lpae_s1_cfg.ttbr;
+
+	/*
+	 * TODO we would like each set of page tables to have a unique ASID
+	 * to optimize TLB invalidation.  But iommu_flush_tlb_all() will
+	 * end up flushing the ASID used for TTBR1 pagetables, which is not
+	 * what we want.  So for now just use the same ASID as TTBR1.
+	 */
+	pagetable->asid = 0;
+
+	return &pagetable->base;
+}
+
 static int msm_fault_handler(struct iommu_domain *domain, struct device *dev,
 		unsigned long iova, int flags, void *arg)
 {
@@ -85,9 +280,11 @@ struct msm_mmu *msm_iommu_new(struct device *dev, struct iommu_domain *domain)
 		return ERR_PTR(-ENOMEM);
 
 	iommu->domain = domain;
-	msm_mmu_init(&iommu->base, dev, &funcs);
+	msm_mmu_init(&iommu->base, dev, &funcs, MSM_MMU_IOMMU);
 	iommu_set_fault_handler(domain, msm_fault_handler, iommu);
 
+	atomic_set(&iommu->pagetables, 0);
+
 	ret = iommu_attach_device(iommu->domain, dev);
 	if (ret) {
 		kfree(iommu);
drivers/gpu/drm/msm/msm_mmu.h +15 -1
index 3a534ee59bf6..61ade89d9e48 100644
@@ -17,18 +17,26 @@ struct msm_mmu_funcs {
 	void (*destroy)(struct msm_mmu *mmu);
 };
 
+enum msm_mmu_type {
+	MSM_MMU_GPUMMU,
+	MSM_MMU_IOMMU,
+	MSM_MMU_IOMMU_PAGETABLE,
+};
+
 struct msm_mmu {
 	const struct msm_mmu_funcs *funcs;
 	struct device *dev;
 	int (*handler)(void *arg, unsigned long iova, int flags);
 	void *arg;
+	enum msm_mmu_type type;
 };
 
 static inline void msm_mmu_init(struct msm_mmu *mmu, struct device *dev,
-		const struct msm_mmu_funcs *funcs)
+		const struct msm_mmu_funcs *funcs, enum msm_mmu_type type)
 {
 	mmu->dev = dev;
 	mmu->funcs = funcs;
+	mmu->type = type;
 }
 
 struct msm_mmu *msm_iommu_new(struct device *dev, struct iommu_domain *domain);
@@ -41,7 +49,13 @@ static inline void msm_mmu_set_fault_handler(struct msm_mmu *mmu, void *arg,
 	mmu->handler = handler;
 }
 
+struct msm_mmu *msm_iommu_pagetable_create(struct msm_mmu *parent);
+
 void msm_gpummu_params(struct msm_mmu *mmu, dma_addr_t *pt_base,
 		dma_addr_t *tran_error);
 
+
+int msm_iommu_pagetable_params(struct msm_mmu *mmu, phys_addr_t *ttbr,
+		int *asid);
+
 #endif /* __MSM_MMU_H__ */
Signature · 1 line
2.26.2
#61 lore raw

PATCH 15/19 drm/msm: Add support for private address space instances

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Add support for allocating private address space instances. Targets that support per-context pagetables should implement their own function to allocate private address spaces.

The default will return a pointer to the global address space.

---
 drivers/gpu/drm/msm/msm_drv.c     | 13 +++++++------
 drivers/gpu/drm/msm/msm_drv.h     |  5 +++++
 drivers/gpu/drm/msm/msm_gem_vma.c |  9 +++++++++
 drivers/gpu/drm/msm/msm_gpu.c     | 22 ++++++++++++++++++++++
 drivers/gpu/drm/msm/msm_gpu.h     |  5 +++++
 5 files changed, 48 insertions(+), 6 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +7 -6
index 01845a3b8d52..8e70d220bba8 100644
@@ -597,7 +597,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
-	ctx->aspace = priv->gpu ? priv->gpu->aspace : NULL;
+	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu);
 	file->driver_priv = ctx;
 
 	return 0;
@@ -780,18 +780,19 @@ static int msm_ioctl_gem_cpu_fini(struct drm_device *dev, void *data,
 }
 
 static int msm_ioctl_gem_info_iova(struct drm_device *dev,
-		struct drm_gem_object *obj, uint64_t *iova)
+		struct drm_file *file, struct drm_gem_object *obj,
+		uint64_t *iova)
 {
-	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_file_private *ctx = file->driver_priv;
 
-	if (!priv->gpu)
+	if (!ctx->aspace)
 		return -EINVAL;
 
 	/*
 	 * Don't pin the memory here - just get an address so that userspace can
 	 * be productive
 	 */
-	return msm_gem_get_iova(obj, priv->gpu->aspace, iova);
+	return msm_gem_get_iova(obj, ctx->aspace, iova);
 }
 
 static int msm_ioctl_gem_info(struct drm_device *dev, void *data,
@@ -830,7 +831,7 @@ static int msm_ioctl_gem_info(struct drm_device *dev, void *data,
 		args->value = msm_gem_mmap_offset(obj);
 		break;
 	case MSM_INFO_GET_IOVA:
-		ret = msm_ioctl_gem_info_iova(dev, obj, &args->value);
+		ret = msm_ioctl_gem_info_iova(dev, file, obj, &args->value);
 		break;
 	case MSM_INFO_SET_NAME:
 		/* length check should leave room for terminating null: */
drivers/gpu/drm/msm/msm_drv.h +5 -0
index 4561bfb5e745..2ca9c3c03845 100644
@@ -249,6 +249,10 @@ int msm_gem_map_vma(struct msm_gem_address_space *aspace,
 void msm_gem_close_vma(struct msm_gem_address_space *aspace,
 		struct msm_gem_vma *vma);
 
+
+struct msm_gem_address_space *
+msm_gem_address_space_get(struct msm_gem_address_space *aspace);
+
 void msm_gem_address_space_put(struct msm_gem_address_space *aspace);
 
 struct msm_gem_address_space *
@@ -434,6 +438,7 @@ static inline void __msm_file_private_destroy(struct kref *kref)
 	struct msm_file_private *ctx = container_of(kref,
 		struct msm_file_private, ref);
 
+	msm_gem_address_space_put(ctx->aspace);
 	kfree(ctx);
 }
 
drivers/gpu/drm/msm/msm_gem_vma.c +9 -0
index 5f6a11211b64..29cc1305cf37 100644
@@ -27,6 +27,15 @@ void msm_gem_address_space_put(struct msm_gem_address_space *aspace)
 		kref_put(&aspace->kref, msm_gem_address_space_destroy);
 }
 
+struct msm_gem_address_space *
+msm_gem_address_space_get(struct msm_gem_address_space *aspace)
+{
+	if (!IS_ERR_OR_NULL(aspace))
+		kref_get(&aspace->kref);
+
+	return aspace;
+}
+
 /* Actually unmap memory for the vma */
 void msm_gem_purge_vma(struct msm_gem_address_space *aspace,
 		struct msm_gem_vma *vma)
drivers/gpu/drm/msm/msm_gpu.c +22 -0
index e1a3cbe25a0c..951850804d77 100644
@@ -823,6 +823,28 @@ static int get_clocks(struct platform_device *pdev, struct msm_gpu *gpu)
 	return 0;
 }
 
+/* Return a new address space for a msm_drm_private instance */
+struct msm_gem_address_space *
+msm_gpu_create_private_address_space(struct msm_gpu *gpu)
+{
+	struct msm_gem_address_space *aspace = NULL;
+
+	if (!gpu)
+		return NULL;
+
+	/*
+	 * If the target doesn't support private address spaces then return
+	 * the global one
+	 */
+	if (gpu->funcs->create_private_address_space)
+		aspace = gpu->funcs->create_private_address_space(gpu);
+
+	if (IS_ERR_OR_NULL(aspace))
+		aspace = msm_gem_address_space_get(gpu->aspace);
+
+	return aspace;
+}
+
 int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, struct msm_gpu_config *config)
drivers/gpu/drm/msm/msm_gpu.h +5 -0
index 1f96ac0d9049..4052a18e18c2 100644
@@ -65,6 +65,8 @@ struct msm_gpu_funcs {
 	void (*gpu_set_freq)(struct msm_gpu *gpu, struct dev_pm_opp *opp);
 	struct msm_gem_address_space *(*create_address_space)
 		(struct msm_gpu *gpu, struct platform_device *pdev);
+	struct msm_gem_address_space *(*create_private_address_space)
+		(struct msm_gpu *gpu);
 };
 
 struct msm_gpu {
@@ -295,6 +297,9 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, struct msm_gpu_config *config);
 
+struct msm_gem_address_space *
+msm_gpu_create_private_address_space(struct msm_gpu *gpu);
+
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
 struct msm_gpu *adreno_load_gpu(struct drm_device *dev);
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
Reviewed-by on v1 15/19
#62 lore raw

Re: [PATCH 15/19] drm/msm: Add support for private address space instances

Quoted 8 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Add support for allocating private address space instances. Targets that support per-context pagetables should implement their own function to allocate private address spaces.

The default will return a pointer to the global address space.

Quoted 158 lines
---
 drivers/gpu/drm/msm/msm_drv.c     | 13 +++++++------
 drivers/gpu/drm/msm/msm_drv.h     |  5 +++++
 drivers/gpu/drm/msm/msm_gem_vma.c |  9 +++++++++
 drivers/gpu/drm/msm/msm_gpu.c     | 22 ++++++++++++++++++++++
 drivers/gpu/drm/msm/msm_gpu.h     |  5 +++++
 5 files changed, 48 insertions(+), 6 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +7 -6
index 01845a3b8d52..8e70d220bba8 100644
@@ -597,7 +597,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
-	ctx->aspace = priv->gpu ? priv->gpu->aspace : NULL;
+	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu);
 	file->driver_priv = ctx;
 
 	return 0;
@@ -780,18 +780,19 @@ static int msm_ioctl_gem_cpu_fini(struct drm_device *dev, void *data,
 }
 
 static int msm_ioctl_gem_info_iova(struct drm_device *dev,
-		struct drm_gem_object *obj, uint64_t *iova)
+		struct drm_file *file, struct drm_gem_object *obj,
+		uint64_t *iova)
 {
-	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_file_private *ctx = file->driver_priv;
 
-	if (!priv->gpu)
+	if (!ctx->aspace)
 		return -EINVAL;
 
 	/*
 	 * Don't pin the memory here - just get an address so that userspace can
 	 * be productive
 	 */
-	return msm_gem_get_iova(obj, priv->gpu->aspace, iova);
+	return msm_gem_get_iova(obj, ctx->aspace, iova);
 }
 
 static int msm_ioctl_gem_info(struct drm_device *dev, void *data,
@@ -830,7 +831,7 @@ static int msm_ioctl_gem_info(struct drm_device *dev, void *data,
 		args->value = msm_gem_mmap_offset(obj);
 		break;
 	case MSM_INFO_GET_IOVA:
-		ret = msm_ioctl_gem_info_iova(dev, obj, &args->value);
+		ret = msm_ioctl_gem_info_iova(dev, file, obj, &args->value);
 		break;
 	case MSM_INFO_SET_NAME:
 		/* length check should leave room for terminating null: */
drivers/gpu/drm/msm/msm_drv.h +5 -0
index 4561bfb5e745..2ca9c3c03845 100644
@@ -249,6 +249,10 @@ int msm_gem_map_vma(struct msm_gem_address_space *aspace,
 void msm_gem_close_vma(struct msm_gem_address_space *aspace,
 		struct msm_gem_vma *vma);
 
+
+struct msm_gem_address_space *
+msm_gem_address_space_get(struct msm_gem_address_space *aspace);
+
 void msm_gem_address_space_put(struct msm_gem_address_space *aspace);
 
 struct msm_gem_address_space *
@@ -434,6 +438,7 @@ static inline void __msm_file_private_destroy(struct kref *kref)
 	struct msm_file_private *ctx = container_of(kref,
 		struct msm_file_private, ref);
 
+	msm_gem_address_space_put(ctx->aspace);
 	kfree(ctx);
 }
 
drivers/gpu/drm/msm/msm_gem_vma.c +9 -0
index 5f6a11211b64..29cc1305cf37 100644
@@ -27,6 +27,15 @@ void msm_gem_address_space_put(struct msm_gem_address_space *aspace)
 		kref_put(&aspace->kref, msm_gem_address_space_destroy);
 }
 
+struct msm_gem_address_space *
+msm_gem_address_space_get(struct msm_gem_address_space *aspace)
+{
+	if (!IS_ERR_OR_NULL(aspace))
+		kref_get(&aspace->kref);
+
+	return aspace;
+}
+
 /* Actually unmap memory for the vma */
 void msm_gem_purge_vma(struct msm_gem_address_space *aspace,
 		struct msm_gem_vma *vma)
drivers/gpu/drm/msm/msm_gpu.c +22 -0
index e1a3cbe25a0c..951850804d77 100644
@@ -823,6 +823,28 @@ static int get_clocks(struct platform_device *pdev, struct msm_gpu *gpu)
 	return 0;
 }
 
+/* Return a new address space for a msm_drm_private instance */
+struct msm_gem_address_space *
+msm_gpu_create_private_address_space(struct msm_gpu *gpu)
+{
+	struct msm_gem_address_space *aspace = NULL;
+
+	if (!gpu)
+		return NULL;
+
+	/*
+	 * If the target doesn't support private address spaces then return
+	 * the global one
+	 */
+	if (gpu->funcs->create_private_address_space)
+		aspace = gpu->funcs->create_private_address_space(gpu);
+
+	if (IS_ERR_OR_NULL(aspace))
+		aspace = msm_gem_address_space_get(gpu->aspace);
+
+	return aspace;
+}
+
 int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, struct msm_gpu_config *config)
drivers/gpu/drm/msm/msm_gpu.h +5 -0
index 1f96ac0d9049..4052a18e18c2 100644
@@ -65,6 +65,8 @@ struct msm_gpu_funcs {
 	void (*gpu_set_freq)(struct msm_gpu *gpu, struct dev_pm_opp *opp);
 	struct msm_gem_address_space *(*create_address_space)
 		(struct msm_gpu *gpu, struct platform_device *pdev);
+	struct msm_gem_address_space *(*create_private_address_space)
+		(struct msm_gpu *gpu);
 };
 
 struct msm_gpu {
@@ -295,6 +297,9 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, struct msm_gpu_config *config);
 
+struct msm_gem_address_space *
+msm_gpu_create_private_address_space(struct msm_gpu *gpu);
+
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
 struct msm_gpu *adreno_load_gpu(struct drm_device *dev);
Signature · 1 line
2.26.2
#63 lore raw

PATCH 16/19 drm/msm/a6xx: Add support for per-instance pagetables

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Add support for using per-instance pagetables if all the dependencies are available.

---
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 70 +++++++++++++++++++++++++++
 drivers/gpu/drm/msm/adreno/a6xx_gpu.h |  1 +
 drivers/gpu/drm/msm/msm_ringbuffer.h  |  1 +
 3 files changed, 72 insertions(+)
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +70 -0
index 5eabb0109577..9653ac9b3cb8 100644
@@ -81,6 +81,56 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
 	OUT_RING(ring, upper_32_bits(iova));
 }
 
+static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
+		struct msm_ringbuffer *ring, struct msm_file_private *ctx)
+{
+	phys_addr_t ttbr;
+	u32 asid;
+	u64 memptr = rbmemptr(ring, ttbr0);
+
+	if (ctx == a6xx_gpu->cur_ctx)
+		return;
+
+	if (msm_iommu_pagetable_params(ctx->aspace->mmu, &ttbr, &asid))
+		return;
+
+	/* Execute the table update */
+	OUT_PKT7(ring, CP_SMMU_TABLE_UPDATE, 4);
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_0_TTBR0_LO(lower_32_bits(ttbr)));
+
+	/*
+	 * For now ignore the asid since the smmu driver uses a TLBIASID to
+	 * flush the TLB when we use iommu_flush_tlb_all() and the smmu driver
+	 * isn't aware that the asid changed.  Instead, keep the default asid
+	 * (0, same as the context bank) to make sure the TLB is properly
+	 * flushed.
+	 */
+	OUT_RING(ring,
+		CP_SMMU_TABLE_UPDATE_1_TTBR0_HI(upper_32_bits(ttbr)) |
+		CP_SMMU_TABLE_UPDATE_1_ASID(0));
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_2_CONTEXTIDR(0));
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_3_CONTEXTBANK(0));
+
+	/*
+	 * Write the new TTBR0 to the memstore. This is good for debugging.
+	 */
+	OUT_PKT7(ring, CP_MEM_WRITE, 4);
+	OUT_RING(ring, CP_MEM_WRITE_0_ADDR_LO(lower_32_bits(memptr)));
+	OUT_RING(ring, CP_MEM_WRITE_1_ADDR_HI(upper_32_bits(memptr)));
+	OUT_RING(ring, lower_32_bits(ttbr));
+	OUT_RING(ring, (0 << 16) | upper_32_bits(ttbr));
+
+	/*
+	 * And finally, trigger a uche flush to be sure there isn't anything
+	 * lingering in that part of the GPU
+	 */
+
+	OUT_PKT7(ring, CP_EVENT_WRITE, 1);
+	OUT_RING(ring, 0x31);
+
+	a6xx_gpu->cur_ctx = ctx;
+}
+
 static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
@@ -90,6 +140,8 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 	struct msm_ringbuffer *ring = submit->ring;
 	unsigned int i;
 
+	a6xx_set_pagetable(a6xx_gpu, ring, submit->queue->ctx);
+
 	get_stats_counter(ring, REG_A6XX_RBBM_PERFCTR_CP_0_LO,
 		rbmemptr_stats(ring, index, cpcycles_start));
 
@@ -696,6 +748,8 @@ static int a6xx_hw_init(struct msm_gpu *gpu)
 	/* Always come up on rb 0 */
 	a6xx_gpu->cur_ring = gpu->rb[0];
 
+	a6xx_gpu->cur_ctx = NULL;
+
 	/* Enable the SQE_to start the CP engine */
 	gpu_write(gpu, REG_A6XX_CP_SQE_CNTL, 1);
 
@@ -1008,6 +1062,21 @@ static unsigned long a6xx_gpu_busy(struct msm_gpu *gpu)
 	return (unsigned long)busy_time;
 }
 
+static struct msm_gem_address_space *
+a6xx_create_private_address_space(struct msm_gpu *gpu)
+{
+	struct msm_gem_address_space *aspace = NULL;
+	struct msm_mmu *mmu;
+
+	mmu = msm_iommu_pagetable_create(gpu->aspace->mmu);
+
+	if (!IS_ERR(mmu))
+		aspace = msm_gem_address_space_create(mmu,
+			"gpu", 0x100000000ULL, 0x1ffffffffULL);
+
+	return aspace;
+}
+
 static const struct adreno_gpu_funcs funcs = {
 	.base = {
 		.get_param = adreno_get_param,
@@ -1031,6 +1100,7 @@ static const struct adreno_gpu_funcs funcs = {
 		.gpu_state_put = a6xx_gpu_state_put,
 #endif
 		.create_address_space = adreno_iommu_create_address_space,
+		.create_private_address_space = a6xx_create_private_address_space,
 	},
 	.get_timestamp = a6xx_get_timestamp,
 };
drivers/gpu/drm/msm/adreno/a6xx_gpu.h +1 -0
index 03ba60d5b07f..da22d7549d9b 100644
@@ -19,6 +19,7 @@ struct a6xx_gpu {
 	uint64_t sqe_iova;
 
 	struct msm_ringbuffer *cur_ring;
+	struct msm_file_private *cur_ctx;
 
 	struct a6xx_gmu gmu;
 };
drivers/gpu/drm/msm/msm_ringbuffer.h +1 -0
index 7764373d0ed2..0987d6bf848c 100644
@@ -31,6 +31,7 @@ struct msm_rbmemptrs {
 	volatile uint32_t fence;
 
 	volatile struct msm_gpu_submit_stats stats[MSM_GPU_SUBMIT_STATS_COUNT];
+	volatile u64 ttbr0;
 };
 
 struct msm_ringbuffer {
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#64 lore raw

Re: [PATCH 16/19] drm/msm/a6xx: Add support for per-instance pagetables

Quoted 49 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Add support for using per-instance pagetables if all the dependencies are available.

---
  drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 70 +++++++++++++++++++++++++++
  drivers/gpu/drm/msm/adreno/a6xx_gpu.h |  1 +
  drivers/gpu/drm/msm/msm_ringbuffer.h  |  1 +
  3 files changed, 72 insertions(+)
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +28 -0
index 5eabb0109577..9653ac9b3cb8 100644
@@ -81,6 +81,56 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
  	OUT_RING(ring, upper_32_bits(iova));
  }
 
+static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
+		struct msm_ringbuffer *ring, struct msm_file_private *ctx)
+{
+	phys_addr_t ttbr;
+	u32 asid;
+	u64 memptr = rbmemptr(ring, ttbr0);
+
+	if (ctx == a6xx_gpu->cur_ctx)
+		return;
+
+	if (msm_iommu_pagetable_params(ctx->aspace->mmu, &ttbr, &asid))
+		return;
+
+	/* Execute the table update */
+	OUT_PKT7(ring, CP_SMMU_TABLE_UPDATE, 4);
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_0_TTBR0_LO(lower_32_bits(ttbr)));
+
+	/*
+	 * For now ignore the asid since the smmu driver uses a TLBIASID to
+	 * flush the TLB when we use iommu_flush_tlb_all() and the smmu driver
+	 * isn't aware that the asid changed.  Instead, keep the default asid
+	 * (0, same as the context bank) to make sure the TLB is properly
+	 * flushed.
+	 */
+	OUT_RING(ring,
+		CP_SMMU_TABLE_UPDATE_1_TTBR0_HI(upper_32_bits(ttbr)) |
+		CP_SMMU_TABLE_UPDATE_1_ASID(0));
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_2_CONTEXTIDR(0));

drivers/gpu/drm/msm/adreno/a6xx_gpu.c static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter, · line 121

Quoted 10 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +10 -0
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_3_CONTEXTBANK(0));
+
+	/*
+	 * Write the new TTBR0 to the memstore. This is good for debugging.
+	 */
+	OUT_PKT7(ring, CP_MEM_WRITE, 4);
+	OUT_RING(ring, CP_MEM_WRITE_0_ADDR_LO(lower_32_bits(memptr)));
+	OUT_RING(ring, CP_MEM_WRITE_1_ADDR_HI(upper_32_bits(memptr)));
+	OUT_RING(ring, lower_32_bits(ttbr));
+	OUT_RING(ring, (0 << 16) | upper_32_bits(ttbr));

why (0 << 16) is required here?

drivers/gpu/drm/msm/adreno/a6xx_gpu.c static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter, · line 129

Quoted 8 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +8 -0
+
+	/*
+	 * And finally, trigger a uche flush to be sure there isn't anything
+	 * lingering in that part of the GPU
+	 */
+
+	OUT_PKT7(ring, CP_EVENT_WRITE, 1);
+	OUT_RING(ring, 0x31);

This may be unnecessary, but no harm in keeping it. SMMU_TABLE_UPDATE is supposed to do a UCHE flush.

-Akhil

Quoted 73 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +24 -0
+
+	a6xx_gpu->cur_ctx = ctx;
+}
+
  static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
  {
  	unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
@@ -90,6 +140,8 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
  	struct msm_ringbuffer *ring = submit->ring;
  	unsigned int i;
 
+	a6xx_set_pagetable(a6xx_gpu, ring, submit->queue->ctx);
+
  	get_stats_counter(ring, REG_A6XX_RBBM_PERFCTR_CP_0_LO,
  		rbmemptr_stats(ring, index, cpcycles_start));
 
@@ -696,6 +748,8 @@ static int a6xx_hw_init(struct msm_gpu *gpu)
  	/* Always come up on rb 0 */
  	a6xx_gpu->cur_ring = gpu->rb[0];
 
+	a6xx_gpu->cur_ctx = NULL;
+
  	/* Enable the SQE_to start the CP engine */
  	gpu_write(gpu, REG_A6XX_CP_SQE_CNTL, 1);
 
@@ -1008,6 +1062,21 @@ static unsigned long a6xx_gpu_busy(struct msm_gpu *gpu)
  	return (unsigned long)busy_time;
  }
 
+static struct msm_gem_address_space *
+a6xx_create_private_address_space(struct msm_gpu *gpu)
+{
+	struct msm_gem_address_space *aspace = NULL;
+	struct msm_mmu *mmu;
+
+	mmu = msm_iommu_pagetable_create(gpu->aspace->mmu);
+
+	if (!IS_ERR(mmu))
+		aspace = msm_gem_address_space_create(mmu,
+			"gpu", 0x100000000ULL, 0x1ffffffffULL);
+
+	return aspace;
+}
+
  static const struct adreno_gpu_funcs funcs = {
  	.base = {
  		.get_param = adreno_get_param,
@@ -1031,6 +1100,7 @@ static const struct adreno_gpu_funcs funcs = {
  		.gpu_state_put = a6xx_gpu_state_put,
  #endif
  		.create_address_space = adreno_iommu_create_address_space,
+		.create_private_address_space = a6xx_create_private_address_space,
  	},
  	.get_timestamp = a6xx_get_timestamp,
  };
drivers/gpu/drm/msm/adreno/a6xx_gpu.h +1 -0
index 03ba60d5b07f..da22d7549d9b 100644
@@ -19,6 +19,7 @@ struct a6xx_gpu {
  	uint64_t sqe_iova;
 
  	struct msm_ringbuffer *cur_ring;
+	struct msm_file_private *cur_ctx;
 
  	struct a6xx_gmu gmu;
  };
drivers/gpu/drm/msm/msm_ringbuffer.h +0 -0
index 7764373d0ed2..0987d6bf848c 100644

drivers/gpu/drm/msm/msm_ringbuffer.h struct msm_rbmemptrs { · line 37

Quoted 7 lines
drivers/gpu/drm/msm/msm_ringbuffer.h +1 -0
@@ -31,6 +31,7 @@ struct msm_rbmemptrs {
  	volatile uint32_t fence;
 
  	volatile struct msm_gpu_submit_stats stats[MSM_GPU_SUBMIT_STATS_COUNT];
+	volatile u64 ttbr0;
  };
 
  struct msm_ringbuffer {

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

#65 lore raw

Re: [PATCH 16/19] drm/msm/a6xx: Add support for per-instance pagetables

Quoted 54 lines · Akhil P Oommen
Quoted 52 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Add support for using per-instance pagetables if all the dependencies are available.

---
  drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 70 +++++++++++++++++++++++++++
  drivers/gpu/drm/msm/adreno/a6xx_gpu.h |  1 +
  drivers/gpu/drm/msm/msm_ringbuffer.h  |  1 +
  3 files changed, 72 insertions(+)
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +31 -0
index 5eabb0109577..9653ac9b3cb8 100644
@@ -81,6 +81,56 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
      OUT_RING(ring, upper_32_bits(iova));
  }
 
+static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
+             struct msm_ringbuffer *ring, struct msm_file_private *ctx)
+{
+     phys_addr_t ttbr;
+     u32 asid;
+     u64 memptr = rbmemptr(ring, ttbr0);
+
+     if (ctx == a6xx_gpu->cur_ctx)
+             return;
+
+     if (msm_iommu_pagetable_params(ctx->aspace->mmu, &ttbr, &asid))
+             return;
+
+     /* Execute the table update */
+     OUT_PKT7(ring, CP_SMMU_TABLE_UPDATE, 4);
+     OUT_RING(ring, CP_SMMU_TABLE_UPDATE_0_TTBR0_LO(lower_32_bits(ttbr)));
+
+     /*
+      * For now ignore the asid since the smmu driver uses a TLBIASID to
+      * flush the TLB when we use iommu_flush_tlb_all() and the smmu driver
+      * isn't aware that the asid changed.  Instead, keep the default asid
+      * (0, same as the context bank) to make sure the TLB is properly
+      * flushed.
+      */
+     OUT_RING(ring,
+             CP_SMMU_TABLE_UPDATE_1_TTBR0_HI(upper_32_bits(ttbr)) |
+             CP_SMMU_TABLE_UPDATE_1_ASID(0));
+     OUT_RING(ring, CP_SMMU_TABLE_UPDATE_2_CONTEXTIDR(0));
+     OUT_RING(ring, CP_SMMU_TABLE_UPDATE_3_CONTEXTBANK(0));
+
+     /*
Quoted 8 lines

drivers/gpu/drm/msm/adreno/a6xx_gpu.c static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter, · line 121

Quoted 7 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +7 -0
+      * Write the new TTBR0 to the memstore. This is good for debugging.
+      */
+     OUT_PKT7(ring, CP_MEM_WRITE, 4);
+     OUT_RING(ring, CP_MEM_WRITE_0_ADDR_LO(lower_32_bits(memptr)));
+     OUT_RING(ring, CP_MEM_WRITE_1_ADDR_HI(upper_32_bits(memptr)));
+     OUT_RING(ring, lower_32_bits(ttbr));
+     OUT_RING(ring, (0 << 16) | upper_32_bits(ttbr));

why (0 << 16) is required here?

That would normally be (asid << 16), except due to the way we currently flush tlb, we need to use asid==0..

(Although msm_iommu_pagetable_params() should return asid==0, I fixed that in msm_iommu_pagetable_create() and didn't notice that Jordan had fixed that here instead)

BR,
-R

Quoted 86 lines

drivers/gpu/drm/msm/adreno/a6xx_gpu.c static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter, · line 129

Quoted 8 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +8 -0
+
+     /*
+      * And finally, trigger a uche flush to be sure there isn't anything
+      * lingering in that part of the GPU
+      */
+
+     OUT_PKT7(ring, CP_EVENT_WRITE, 1);
+     OUT_RING(ring, 0x31);

This may be unnecessary, but no harm in keeping it. SMMU_TABLE_UPDATE is supposed to do a UCHE flush.

-Akhil

Quoted 74 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +24 -0
+
+     a6xx_gpu->cur_ctx = ctx;
+}
+
  static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
  {
      unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
@@ -90,6 +140,8 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
      struct msm_ringbuffer *ring = submit->ring;
      unsigned int i;
 
+     a6xx_set_pagetable(a6xx_gpu, ring, submit->queue->ctx);
+
      get_stats_counter(ring, REG_A6XX_RBBM_PERFCTR_CP_0_LO,
              rbmemptr_stats(ring, index, cpcycles_start));
 
@@ -696,6 +748,8 @@ static int a6xx_hw_init(struct msm_gpu *gpu)
      /* Always come up on rb 0 */
      a6xx_gpu->cur_ring = gpu->rb[0];
 
+     a6xx_gpu->cur_ctx = NULL;
+
      /* Enable the SQE_to start the CP engine */
      gpu_write(gpu, REG_A6XX_CP_SQE_CNTL, 1);
 
@@ -1008,6 +1062,21 @@ static unsigned long a6xx_gpu_busy(struct msm_gpu *gpu)
      return (unsigned long)busy_time;
  }
 
+static struct msm_gem_address_space *
+a6xx_create_private_address_space(struct msm_gpu *gpu)
+{
+     struct msm_gem_address_space *aspace = NULL;
+     struct msm_mmu *mmu;
+
+     mmu = msm_iommu_pagetable_create(gpu->aspace->mmu);
+
+     if (!IS_ERR(mmu))
+             aspace = msm_gem_address_space_create(mmu,
+                     "gpu", 0x100000000ULL, 0x1ffffffffULL);
+
+     return aspace;
+}
+
  static const struct adreno_gpu_funcs funcs = {
      .base = {
              .get_param = adreno_get_param,
@@ -1031,6 +1100,7 @@ static const struct adreno_gpu_funcs funcs = {
              .gpu_state_put = a6xx_gpu_state_put,
  #endif
              .create_address_space = adreno_iommu_create_address_space,
+             .create_private_address_space = a6xx_create_private_address_space,
      },
      .get_timestamp = a6xx_get_timestamp,
  };
drivers/gpu/drm/msm/adreno/a6xx_gpu.h +1 -0
index 03ba60d5b07f..da22d7549d9b 100644
@@ -19,6 +19,7 @@ struct a6xx_gpu {
      uint64_t sqe_iova;
 
      struct msm_ringbuffer *cur_ring;
+     struct msm_file_private *cur_ctx;
 
      struct a6xx_gmu gmu;
  };
drivers/gpu/drm/msm/msm_ringbuffer.h +0 -0
index 7764373d0ed2..0987d6bf848c 100644
@@ -31,6 +31,7 @@ struct msm_rbmemptrs {
      volatile uint32_t fence;
 
Quoted 7 lines

drivers/gpu/drm/msm/msm_ringbuffer.h struct msm_rbmemptrs { · line 37

Quoted 6 lines
drivers/gpu/drm/msm/msm_ringbuffer.h +1 -0
      volatile struct msm_gpu_submit_stats stats[MSM_GPU_SUBMIT_STATS_COUNT];
+     volatile u64 ttbr0;
  };
 
  struct msm_ringbuffer {

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

#66 lore raw

Re: [PATCH 16/19] drm/msm/a6xx: Add support for per-instance pagetables

In reply to Akhil P Oommen
Quoted 52 lines · Akhil P Oommen
Quoted 51 lines · Rob Clark

From: Jordan Crouse <[email protected]>

Add support for using per-instance pagetables if all the dependencies are available.

---
 drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 70 +++++++++++++++++++++++++++
 drivers/gpu/drm/msm/adreno/a6xx_gpu.h |  1 +
 drivers/gpu/drm/msm/msm_ringbuffer.h  |  1 +
 3 files changed, 72 insertions(+)
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +31 -0
index 5eabb0109577..9653ac9b3cb8 100644
@@ -81,6 +81,56 @@ static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter,
 	OUT_RING(ring, upper_32_bits(iova));
 }
+static void a6xx_set_pagetable(struct a6xx_gpu *a6xx_gpu,
+		struct msm_ringbuffer *ring, struct msm_file_private *ctx)
+{
+	phys_addr_t ttbr;
+	u32 asid;
+	u64 memptr = rbmemptr(ring, ttbr0);
+
+	if (ctx == a6xx_gpu->cur_ctx)
+		return;
+
+	if (msm_iommu_pagetable_params(ctx->aspace->mmu, &ttbr, &asid))
+		return;
+
+	/* Execute the table update */
+	OUT_PKT7(ring, CP_SMMU_TABLE_UPDATE, 4);
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_0_TTBR0_LO(lower_32_bits(ttbr)));
+
+	/*
+	 * For now ignore the asid since the smmu driver uses a TLBIASID to
+	 * flush the TLB when we use iommu_flush_tlb_all() and the smmu driver
+	 * isn't aware that the asid changed.  Instead, keep the default asid
+	 * (0, same as the context bank) to make sure the TLB is properly
+	 * flushed.
+	 */
+	OUT_RING(ring,
+		CP_SMMU_TABLE_UPDATE_1_TTBR0_HI(upper_32_bits(ttbr)) |
+		CP_SMMU_TABLE_UPDATE_1_ASID(0));
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_2_CONTEXTIDR(0));
+	OUT_RING(ring, CP_SMMU_TABLE_UPDATE_3_CONTEXTBANK(0));
+
+	/*
Quoted 8 lines

drivers/gpu/drm/msm/adreno/a6xx_gpu.c static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter, · line 121

Quoted 7 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +7 -0
+	 * Write the new TTBR0 to the memstore. This is good for debugging.
+	 */
+	OUT_PKT7(ring, CP_MEM_WRITE, 4);
+	OUT_RING(ring, CP_MEM_WRITE_0_ADDR_LO(lower_32_bits(memptr)));
+	OUT_RING(ring, CP_MEM_WRITE_1_ADDR_HI(upper_32_bits(memptr)));
+	OUT_RING(ring, lower_32_bits(ttbr));
+	OUT_RING(ring, (0 << 16) | upper_32_bits(ttbr));

why (0 << 16) is required here?

Because that is the ASID we are using and we would want the debug TTBR0 to match the hardware as closely as possible.

Quoted 10 lines

drivers/gpu/drm/msm/adreno/a6xx_gpu.c static void get_stats_counter(struct msm_ringbuffer *ring, u32 counter, · line 129

Quoted 8 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +8 -0
+
+	/*
+	 * And finally, trigger a uche flush to be sure there isn't anything
+	 * lingering in that part of the GPU
+	 */
+
+	OUT_PKT7(ring, CP_EVENT_WRITE, 1);
+	OUT_RING(ring, 0x31);

This may be unnecessary, but no harm in keeping it. SMMU_TABLE_UPDATE is supposed to do a UCHE flush.

Correct but I think it is wise to try to match the downstream sequence as much as possible.

Jordan

Quoted 73 lines

-Akhil

Quoted 71 lines
drivers/gpu/drm/msm/adreno/a6xx_gpu.c +24 -0
+
+	a6xx_gpu->cur_ctx = ctx;
+}
+
 static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 {
 	unsigned int index = submit->seqno % MSM_GPU_SUBMIT_STATS_COUNT;
@@ -90,6 +140,8 @@ static void a6xx_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit)
 	struct msm_ringbuffer *ring = submit->ring;
 	unsigned int i;
+	a6xx_set_pagetable(a6xx_gpu, ring, submit->queue->ctx);
+
 	get_stats_counter(ring, REG_A6XX_RBBM_PERFCTR_CP_0_LO,
 		rbmemptr_stats(ring, index, cpcycles_start));
@@ -696,6 +748,8 @@ static int a6xx_hw_init(struct msm_gpu *gpu)
 	/* Always come up on rb 0 */
 	a6xx_gpu->cur_ring = gpu->rb[0];
+	a6xx_gpu->cur_ctx = NULL;
+
 	/* Enable the SQE_to start the CP engine */
 	gpu_write(gpu, REG_A6XX_CP_SQE_CNTL, 1);
@@ -1008,6 +1062,21 @@ static unsigned long a6xx_gpu_busy(struct msm_gpu *gpu)
 	return (unsigned long)busy_time;
 }
+static struct msm_gem_address_space *
+a6xx_create_private_address_space(struct msm_gpu *gpu)
+{
+	struct msm_gem_address_space *aspace = NULL;
+	struct msm_mmu *mmu;
+
+	mmu = msm_iommu_pagetable_create(gpu->aspace->mmu);
+
+	if (!IS_ERR(mmu))
+		aspace = msm_gem_address_space_create(mmu,
+			"gpu", 0x100000000ULL, 0x1ffffffffULL);
+
+	return aspace;
+}
+
 static const struct adreno_gpu_funcs funcs = {
 	.base = {
 		.get_param = adreno_get_param,
@@ -1031,6 +1100,7 @@ static const struct adreno_gpu_funcs funcs = {
 		.gpu_state_put = a6xx_gpu_state_put,
 #endif
 		.create_address_space = adreno_iommu_create_address_space,
+		.create_private_address_space = a6xx_create_private_address_space,
 	},
 	.get_timestamp = a6xx_get_timestamp,
 };
drivers/gpu/drm/msm/adreno/a6xx_gpu.h +1 -0
index 03ba60d5b07f..da22d7549d9b 100644
@@ -19,6 +19,7 @@ struct a6xx_gpu {
 	uint64_t sqe_iova;
 	struct msm_ringbuffer *cur_ring;
+	struct msm_file_private *cur_ctx;
 	struct a6xx_gmu gmu;
 };
drivers/gpu/drm/msm/msm_ringbuffer.h +1 -0
index 7764373d0ed2..0987d6bf848c 100644
@@ -31,6 +31,7 @@ struct msm_rbmemptrs {
 	volatile uint32_t fence;
 	volatile struct msm_gpu_submit_stats stats[MSM_GPU_SUBMIT_STATS_COUNT];
+	volatile u64 ttbr0;
 };
 struct msm_ringbuffer {
 
Signature · 6 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
#67 lore raw

PATCH 17/19 arm: dts: qcom: sm845: Set the compatible string for the GPU SMMU

In reply to Jordan Crouse

From: Jordan Crouse <[email protected]>

Set the qcom,adreno-smmu compatible string for the GPU SMMU to enable split pagetables and per-instance pagetables for drm/msm.

---
 arch/arm64/boot/dts/qcom/sdm845.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/boot/dts/qcom/sdm845.dtsi +1 -1
index 2884577dcb77..6a9adaa401a9 100644
@@ -4058,7 +4058,7 @@ opp-257000000 {
 		};
 
 		adreno_smmu: iommu@5040000 {
-			compatible = "qcom,sdm845-smmu-v2", "qcom,smmu-v2";
+			compatible = "qcom,adreno-smmu", "qcom,smmu-v2";
 			reg = <0 0x5040000 0 0x10000>;
 			#iommu-cells = <1>;
 			#global-interrupts = <2>;
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#68 lore raw

PATCH 18/19 iommu/arm-smmu: add a way for implementations to influence SCTLR

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

For the Adreno GPU's SMMU, we want SCTLR.HUPCF set to ensure that pending translations are not terminated on iova fault. Otherwise a terminated CP read could hang the GPU by returning invalid command-stream data.

---
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 6 ++++++
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 3 +++
 drivers/iommu/arm/arm-smmu/arm-smmu.h      | 3 +++
 3 files changed, 12 insertions(+)
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +6 -0
index 5640d9960610..2aa6249050ff 100644
@@ -127,6 +127,12 @@ static int qcom_adreno_smmu_init_context(struct arm_smmu_domain *smmu_domain,
 	    (smmu_domain->cfg.fmt == ARM_SMMU_CTX_FMT_AARCH64))
 		pgtbl_cfg->quirks |= IO_PGTABLE_QUIRK_ARM_TTBR1;
 
+	/*
+	 * On the GPU device we want to process subsequent transactions after a
+	 * fault to keep the GPU from hanging
+	 */
+	smmu_domain->cfg.sctlr_set |= ARM_SMMU_SCTLR_HUPCF;
+
 	/*
 	 * Initialize private interface with GPU:
 	 */
drivers/iommu/arm/arm-smmu/arm-smmu.c +3 -0
index e63a480d7f71..bbec5793faf8 100644
@@ -617,6 +617,9 @@ void arm_smmu_write_context_bank(struct arm_smmu_device *smmu, int idx)
 	if (IS_ENABLED(CONFIG_CPU_BIG_ENDIAN))
 		reg |= ARM_SMMU_SCTLR_E;
 
+	reg |= cfg->sctlr_set;
+	reg &= ~cfg->sctlr_clr;
+
 	arm_smmu_cb_write(smmu, idx, ARM_SMMU_CB_SCTLR, reg);
 }
 
drivers/iommu/arm/arm-smmu/arm-smmu.h +3 -0
index cd75a33967bb..2df3a70a8a41 100644
@@ -144,6 +144,7 @@ enum arm_smmu_cbar_type {
 #define ARM_SMMU_CB_SCTLR		0x0
 #define ARM_SMMU_SCTLR_S1_ASIDPNE	BIT(12)
 #define ARM_SMMU_SCTLR_CFCFG		BIT(7)
+#define ARM_SMMU_SCTLR_HUPCF		BIT(8)
 #define ARM_SMMU_SCTLR_CFIE		BIT(6)
 #define ARM_SMMU_SCTLR_CFRE		BIT(5)
 #define ARM_SMMU_SCTLR_E		BIT(4)
@@ -341,6 +342,8 @@ struct arm_smmu_cfg {
 		u16			asid;
 		u16			vmid;
 	};
+	u32				sctlr_set;    /* extra bits to set in SCTLR */
+	u32				sctlr_clr;    /* bits to mask in SCTLR */
 	enum arm_smmu_cbar_type		cbar;
 	enum arm_smmu_context_fmt	fmt;
 };
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
Reviewed-by on v1 18/19
#69 lore raw

Re: [PATCH 18/19] iommu/arm-smmu: add a way for implementations to influence SCTLR

Quoted 7 lines · Rob Clark

From: Rob Clark <[email protected]>

For the Adreno GPU's SMMU, we want SCTLR.HUPCF set to ensure that pending translations are not terminated on iova fault. Otherwise a terminated CP read could hang the GPU by returning invalid command-stream data.

Quoted 62 lines
---
 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 6 ++++++
 drivers/iommu/arm/arm-smmu/arm-smmu.c      | 3 +++
 drivers/iommu/arm/arm-smmu/arm-smmu.h      | 3 +++
 3 files changed, 12 insertions(+)
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c +6 -0
index 5640d9960610..2aa6249050ff 100644
@@ -127,6 +127,12 @@ static int qcom_adreno_smmu_init_context(struct arm_smmu_domain *smmu_domain,
 	    (smmu_domain->cfg.fmt == ARM_SMMU_CTX_FMT_AARCH64))
 		pgtbl_cfg->quirks |= IO_PGTABLE_QUIRK_ARM_TTBR1;
 
+	/*
+	 * On the GPU device we want to process subsequent transactions after a
+	 * fault to keep the GPU from hanging
+	 */
+	smmu_domain->cfg.sctlr_set |= ARM_SMMU_SCTLR_HUPCF;
+
 	/*
 	 * Initialize private interface with GPU:
 	 */
drivers/iommu/arm/arm-smmu/arm-smmu.c +3 -0
index e63a480d7f71..bbec5793faf8 100644
@@ -617,6 +617,9 @@ void arm_smmu_write_context_bank(struct arm_smmu_device *smmu, int idx)
 	if (IS_ENABLED(CONFIG_CPU_BIG_ENDIAN))
 		reg |= ARM_SMMU_SCTLR_E;
 
+	reg |= cfg->sctlr_set;
+	reg &= ~cfg->sctlr_clr;
+
 	arm_smmu_cb_write(smmu, idx, ARM_SMMU_CB_SCTLR, reg);
 }
 
drivers/iommu/arm/arm-smmu/arm-smmu.h +3 -0
index cd75a33967bb..2df3a70a8a41 100644
@@ -144,6 +144,7 @@ enum arm_smmu_cbar_type {
 #define ARM_SMMU_CB_SCTLR		0x0
 #define ARM_SMMU_SCTLR_S1_ASIDPNE	BIT(12)
 #define ARM_SMMU_SCTLR_CFCFG		BIT(7)
+#define ARM_SMMU_SCTLR_HUPCF		BIT(8)
 #define ARM_SMMU_SCTLR_CFIE		BIT(6)
 #define ARM_SMMU_SCTLR_CFRE		BIT(5)
 #define ARM_SMMU_SCTLR_E		BIT(4)
@@ -341,6 +342,8 @@ struct arm_smmu_cfg {
 		u16			asid;
 		u16			vmid;
 	};
+	u32				sctlr_set;    /* extra bits to set in SCTLR */
+	u32				sctlr_clr;    /* bits to mask in SCTLR */
 	enum arm_smmu_cbar_type		cbar;
 	enum arm_smmu_context_fmt	fmt;
 };
Signature · 1 line
2.26.2
#70 lore raw

PATCH 19/19 drm/msm: show process names in gem_describe

In reply to Jordan Crouse

From: Rob Clark <[email protected]>

In $debugfs/gem we already show any vma(s) associated with an object. Also show process names if the vma's address space is a per-process address space.

---
 drivers/gpu/drm/msm/msm_drv.c     |  2 +-
 drivers/gpu/drm/msm/msm_gem.c     | 25 +++++++++++++++++++++----
 drivers/gpu/drm/msm/msm_gem.h     |  5 +++++
 drivers/gpu/drm/msm/msm_gem_vma.c |  1 +
 drivers/gpu/drm/msm/msm_gpu.c     |  8 +++++---
 drivers/gpu/drm/msm/msm_gpu.h     |  2 +-
 6 files changed, 34 insertions(+), 9 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +1 -1
index 8e70d220bba8..8d5c4f98c332 100644
@@ -597,7 +597,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
-	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu);
+	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu, current);
 	file->driver_priv = ctx;
 
 	return 0;
drivers/gpu/drm/msm/msm_gem.c +21 -4
index 3cb7aeb93fd3..76a6c5271e57 100644
@@ -842,11 +842,28 @@ void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
 
 		seq_puts(m, "      vmas:");
 
-		list_for_each_entry(vma, &msm_obj->vmas, list)
-			seq_printf(m, " [%s: %08llx,%s,inuse=%d]",
-				vma->aspace != NULL ? vma->aspace->name : NULL,
-				vma->iova, vma->mapped ? "mapped" : "unmapped",
+		list_for_each_entry(vma, &msm_obj->vmas, list) {
+			const char *name, *comm;
+			if (vma->aspace) {
+				struct msm_gem_address_space *aspace = vma->aspace;
+				struct task_struct *task =
+					get_pid_task(aspace->pid, PIDTYPE_PID);
+				if (task) {
+					comm = kstrdup(task->comm, GFP_KERNEL);
+				} else {
+					comm = NULL;
+				}
+				name = aspace->name;
+			} else {
+				name = comm = NULL;
+			}
+			seq_printf(m, " [%s%s%s: aspace=%p, %08llx,%s,inuse=%d]",
+				name, comm ? ":" : "", comm ? comm : "",
+				vma->aspace, vma->iova,
+				vma->mapped ? "mapped" : "unmapped",
 				vma->inuse);
+			kfree(comm);
+		}
 
 		seq_puts(m, "\n");
 	}
drivers/gpu/drm/msm/msm_gem.h +5 -0
index 9c573c4269cb..7b1c7a5f8eef 100644
@@ -24,6 +24,11 @@ struct msm_gem_address_space {
 	spinlock_t lock; /* Protects drm_mm node allocation/removal */
 	struct msm_mmu *mmu;
 	struct kref kref;
+
+	/* For address spaces associated with a specific process, this
+	 * will be non-NULL:
+	 */
+	struct pid *pid;
 };
 
 struct msm_gem_vma {
drivers/gpu/drm/msm/msm_gem_vma.c +1 -0
index 29cc1305cf37..80a8a266d68f 100644
@@ -17,6 +17,7 @@ msm_gem_address_space_destroy(struct kref *kref)
 	drm_mm_takedown(&aspace->mm);
 	if (aspace->mmu)
 		aspace->mmu->funcs->destroy(aspace->mmu);
+	put_pid(aspace->pid);
 	kfree(aspace);
 }
 
drivers/gpu/drm/msm/msm_gpu.c +5 -3
index 951850804d77..ac8961187a73 100644
@@ -825,10 +825,9 @@ static int get_clocks(struct platform_device *pdev, struct msm_gpu *gpu)
 
 /* Return a new address space for a msm_drm_private instance */
 struct msm_gem_address_space *
-msm_gpu_create_private_address_space(struct msm_gpu *gpu)
+msm_gpu_create_private_address_space(struct msm_gpu *gpu, struct task_struct *task)
 {
 	struct msm_gem_address_space *aspace = NULL;
-
 	if (!gpu)
 		return NULL;
 
@@ -836,8 +835,11 @@ msm_gpu_create_private_address_space(struct msm_gpu *gpu)
 	 * If the target doesn't support private address spaces then return
 	 * the global one
 	 */
-	if (gpu->funcs->create_private_address_space)
+	if (gpu->funcs->create_private_address_space) {
 		aspace = gpu->funcs->create_private_address_space(gpu);
+		if (!IS_ERR(aspace))
+			aspace->pid = get_pid(task_pid(task));
+	}
 
 	if (IS_ERR_OR_NULL(aspace))
 		aspace = msm_gem_address_space_get(gpu->aspace);
drivers/gpu/drm/msm/msm_gpu.h +1 -1
index 4052a18e18c2..59f26bd0fe42 100644
@@ -298,7 +298,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		const char *name, struct msm_gpu_config *config);
 
 struct msm_gem_address_space *
-msm_gpu_create_private_address_space(struct msm_gpu *gpu);
+msm_gpu_create_private_address_space(struct msm_gpu *gpu, struct task_struct *task);
 
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
Signature · 6 lines
2.26.2

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author Reviewed-by on v1 19/19
#71 lore raw

Re: [Freedreno] [PATCH 19/19] drm/msm: show process names in gem_describe

Quoted 5 lines · Rob Clark

From: Rob Clark <[email protected]>

In $debugfs/gem we already show any vma(s) associated with an object. Also show process names if the vma's address space is a per-process address space.

Quoted 137 lines
---
 drivers/gpu/drm/msm/msm_drv.c     |  2 +-
 drivers/gpu/drm/msm/msm_gem.c     | 25 +++++++++++++++++++++----
 drivers/gpu/drm/msm/msm_gem.h     |  5 +++++
 drivers/gpu/drm/msm/msm_gem_vma.c |  1 +
 drivers/gpu/drm/msm/msm_gpu.c     |  8 +++++---
 drivers/gpu/drm/msm/msm_gpu.h     |  2 +-
 6 files changed, 34 insertions(+), 9 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +1 -1
index 8e70d220bba8..8d5c4f98c332 100644
@@ -597,7 +597,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
-	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu);
+	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu, current);
 	file->driver_priv = ctx;
 
 	return 0;
drivers/gpu/drm/msm/msm_gem.c +21 -4
index 3cb7aeb93fd3..76a6c5271e57 100644
@@ -842,11 +842,28 @@ void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
 
 		seq_puts(m, "      vmas:");
 
-		list_for_each_entry(vma, &msm_obj->vmas, list)
-			seq_printf(m, " [%s: %08llx,%s,inuse=%d]",
-				vma->aspace != NULL ? vma->aspace->name : NULL,
-				vma->iova, vma->mapped ? "mapped" : "unmapped",
+		list_for_each_entry(vma, &msm_obj->vmas, list) {
+			const char *name, *comm;
+			if (vma->aspace) {
+				struct msm_gem_address_space *aspace = vma->aspace;
+				struct task_struct *task =
+					get_pid_task(aspace->pid, PIDTYPE_PID);
+				if (task) {
+					comm = kstrdup(task->comm, GFP_KERNEL);
+				} else {
+					comm = NULL;
+				}
+				name = aspace->name;
+			} else {
+				name = comm = NULL;
+			}
+			seq_printf(m, " [%s%s%s: aspace=%p, %08llx,%s,inuse=%d]",
+				name, comm ? ":" : "", comm ? comm : "",
+				vma->aspace, vma->iova,
+				vma->mapped ? "mapped" : "unmapped",
 				vma->inuse);
+			kfree(comm);
+		}
 
 		seq_puts(m, "\n");
 	}
drivers/gpu/drm/msm/msm_gem.h +5 -0
index 9c573c4269cb..7b1c7a5f8eef 100644
@@ -24,6 +24,11 @@ struct msm_gem_address_space {
 	spinlock_t lock; /* Protects drm_mm node allocation/removal */
 	struct msm_mmu *mmu;
 	struct kref kref;
+
+	/* For address spaces associated with a specific process, this
+	 * will be non-NULL:
+	 */
+	struct pid *pid;
 };
 
 struct msm_gem_vma {
drivers/gpu/drm/msm/msm_gem_vma.c +1 -0
index 29cc1305cf37..80a8a266d68f 100644
@@ -17,6 +17,7 @@ msm_gem_address_space_destroy(struct kref *kref)
 	drm_mm_takedown(&aspace->mm);
 	if (aspace->mmu)
 		aspace->mmu->funcs->destroy(aspace->mmu);
+	put_pid(aspace->pid);
 	kfree(aspace);
 }
 
drivers/gpu/drm/msm/msm_gpu.c +5 -3
index 951850804d77..ac8961187a73 100644
@@ -825,10 +825,9 @@ static int get_clocks(struct platform_device *pdev, struct msm_gpu *gpu)
 
 /* Return a new address space for a msm_drm_private instance */
 struct msm_gem_address_space *
-msm_gpu_create_private_address_space(struct msm_gpu *gpu)
+msm_gpu_create_private_address_space(struct msm_gpu *gpu, struct task_struct *task)
 {
 	struct msm_gem_address_space *aspace = NULL;
-
 	if (!gpu)
 		return NULL;
 
@@ -836,8 +835,11 @@ msm_gpu_create_private_address_space(struct msm_gpu *gpu)
 	 * If the target doesn't support private address spaces then return
 	 * the global one
 	 */
-	if (gpu->funcs->create_private_address_space)
+	if (gpu->funcs->create_private_address_space) {
 		aspace = gpu->funcs->create_private_address_space(gpu);
+		if (!IS_ERR(aspace))
+			aspace->pid = get_pid(task_pid(task));
+	}
 
 	if (IS_ERR_OR_NULL(aspace))
 		aspace = msm_gem_address_space_get(gpu->aspace);
drivers/gpu/drm/msm/msm_gpu.h +1 -1
index 4052a18e18c2..59f26bd0fe42 100644
@@ -298,7 +298,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		const char *name, struct msm_gpu_config *config);
 
 struct msm_gem_address_space *
-msm_gpu_create_private_address_space(struct msm_gpu *gpu);
+msm_gpu_create_private_address_space(struct msm_gpu *gpu, struct task_struct *task);
 
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
Signature · 6 lines
2.26.2

_______________________________________________
Freedreno mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/freedreno
Signature · 6 lines
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
Reviewed-by on v1 19/19
#72 lore raw

Re: [PATCH 19/19] drm/msm: show process names in gem_describe

In reply to Rob Clark
Quoted 6 lines · Rob Clark

From: Rob Clark <[email protected]>

In $debugfs/gem we already show any vma(s) associated with an object. Also show process names if the vma's address space is a per-process address space.

Quoted 133 lines
---
 drivers/gpu/drm/msm/msm_drv.c     |  2 +-
 drivers/gpu/drm/msm/msm_gem.c     | 25 +++++++++++++++++++++----
 drivers/gpu/drm/msm/msm_gem.h     |  5 +++++
 drivers/gpu/drm/msm/msm_gem_vma.c |  1 +
 drivers/gpu/drm/msm/msm_gpu.c     |  8 +++++---
 drivers/gpu/drm/msm/msm_gpu.h     |  2 +-
 6 files changed, 34 insertions(+), 9 deletions(-)
drivers/gpu/drm/msm/msm_drv.c +1 -1
index 8e70d220bba8..8d5c4f98c332 100644
@@ -597,7 +597,7 @@ static int context_init(struct drm_device *dev, struct drm_file *file)
 	kref_init(&ctx->ref);
 	msm_submitqueue_init(dev, ctx);
 
-	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu);
+	ctx->aspace = msm_gpu_create_private_address_space(priv->gpu, current);
 	file->driver_priv = ctx;
 
 	return 0;
drivers/gpu/drm/msm/msm_gem.c +21 -4
index 3cb7aeb93fd3..76a6c5271e57 100644
@@ -842,11 +842,28 @@ void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
 
 		seq_puts(m, "      vmas:");
 
-		list_for_each_entry(vma, &msm_obj->vmas, list)
-			seq_printf(m, " [%s: %08llx,%s,inuse=%d]",
-				vma->aspace != NULL ? vma->aspace->name : NULL,
-				vma->iova, vma->mapped ? "mapped" : "unmapped",
+		list_for_each_entry(vma, &msm_obj->vmas, list) {
+			const char *name, *comm;
+			if (vma->aspace) {
+				struct msm_gem_address_space *aspace = vma->aspace;
+				struct task_struct *task =
+					get_pid_task(aspace->pid, PIDTYPE_PID);
+				if (task) {
+					comm = kstrdup(task->comm, GFP_KERNEL);
+				} else {
+					comm = NULL;
+				}
+				name = aspace->name;
+			} else {
+				name = comm = NULL;
+			}
+			seq_printf(m, " [%s%s%s: aspace=%p, %08llx,%s,inuse=%d]",
+				name, comm ? ":" : "", comm ? comm : "",
+				vma->aspace, vma->iova,
+				vma->mapped ? "mapped" : "unmapped",
 				vma->inuse);
+			kfree(comm);
+		}
 
 		seq_puts(m, "\n");
 	}
drivers/gpu/drm/msm/msm_gem.h +5 -0
index 9c573c4269cb..7b1c7a5f8eef 100644
@@ -24,6 +24,11 @@ struct msm_gem_address_space {
 	spinlock_t lock; /* Protects drm_mm node allocation/removal */
 	struct msm_mmu *mmu;
 	struct kref kref;
+
+	/* For address spaces associated with a specific process, this
+	 * will be non-NULL:
+	 */
+	struct pid *pid;
 };
 
 struct msm_gem_vma {
drivers/gpu/drm/msm/msm_gem_vma.c +1 -0
index 29cc1305cf37..80a8a266d68f 100644
@@ -17,6 +17,7 @@ msm_gem_address_space_destroy(struct kref *kref)
 	drm_mm_takedown(&aspace->mm);
 	if (aspace->mmu)
 		aspace->mmu->funcs->destroy(aspace->mmu);
+	put_pid(aspace->pid);
 	kfree(aspace);
 }
 
drivers/gpu/drm/msm/msm_gpu.c +5 -3
index 951850804d77..ac8961187a73 100644
@@ -825,10 +825,9 @@ static int get_clocks(struct platform_device *pdev, struct msm_gpu *gpu)
 
 /* Return a new address space for a msm_drm_private instance */
 struct msm_gem_address_space *
-msm_gpu_create_private_address_space(struct msm_gpu *gpu)
+msm_gpu_create_private_address_space(struct msm_gpu *gpu, struct task_struct *task)
 {
 	struct msm_gem_address_space *aspace = NULL;
-
 	if (!gpu)
 		return NULL;
 
@@ -836,8 +835,11 @@ msm_gpu_create_private_address_space(struct msm_gpu *gpu)
 	 * If the target doesn't support private address spaces then return
 	 * the global one
 	 */
-	if (gpu->funcs->create_private_address_space)
+	if (gpu->funcs->create_private_address_space) {
 		aspace = gpu->funcs->create_private_address_space(gpu);
+		if (!IS_ERR(aspace))
+			aspace->pid = get_pid(task_pid(task));
+	}
 
 	if (IS_ERR_OR_NULL(aspace))
 		aspace = msm_gem_address_space_get(gpu->aspace);
drivers/gpu/drm/msm/msm_gpu.h +1 -1
index 4052a18e18c2..59f26bd0fe42 100644
@@ -298,7 +298,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		const char *name, struct msm_gpu_config *config);
 
 struct msm_gem_address_space *
-msm_gpu_create_private_address_space(struct msm_gpu *gpu);
+msm_gpu_create_private_address_space(struct msm_gpu *gpu, struct task_struct *task);
 
 void msm_gpu_cleanup(struct msm_gpu *gpu);
 
Signature · 1 line
2.26.2