diff mbox series

[RFC,112/162] drm/i915/guc: put all guc objects in lmem when available

Message ID 20201127120718.454037-113-matthew.auld@intel.com (mailing list archive)
State New, archived
Headers show
Series DG1 + LMEM enabling | expand

Commit Message

Matthew Auld Nov. 27, 2020, 12:06 p.m. UTC
From: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>

The firmware binary has to be loaded from lmem and the recommendation is
to put all other objects in there as well. Note that we don't fall back
to system memory if the allocation in lmem fails because all objects are
allocated during driver load and if we have issues with lmem at that point
something is seriously wrong with the system, so no point in trying to
handle it.

Cc: Matthew Auld <matthew.auld@intel.com>
Cc: Abdiel Janulgue <abdiel.janulgue@linux.intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Vinay Belgaumkar <vinay.belgaumkar@intel.com>
Cc: Radoslaw Szwichtenberg <radoslaw.szwichtenberg@intel.com>
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com> #v1
Cc: Vinay Belgaumkar <vinay.belgaumkar@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
---
 drivers/gpu/drm/i915/gem/i915_gem_lmem.c  | 41 +++++++++++++++++++++++
 drivers/gpu/drm/i915/gem/i915_gem_lmem.h  |  8 +++++
 drivers/gpu/drm/i915/gt/uc/intel_guc.c    |  9 ++++-
 drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c | 11 ++++--
 drivers/gpu/drm/i915/gt/uc/intel_huc.c    | 14 ++++++--
 drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c  | 35 ++++++++++++++++---
 6 files changed, 107 insertions(+), 11 deletions(-)
diff mbox series

Patch

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_lmem.c b/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
index e56874e54fde..71c07e1f6f26 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
@@ -215,6 +215,21 @@  i915_gem_object_lmem_io_map_page_atomic(struct drm_i915_gem_object *obj,
 	return io_mapping_map_atomic_wc(&obj->mm.region->iomap, offset);
 }
 
+void __iomem *
+i915_gem_object_lmem_io_map(struct drm_i915_gem_object *obj,
+			    unsigned long n,
+			    unsigned long size)
+{
+	resource_size_t offset;
+
+	GEM_BUG_ON(!i915_gem_object_is_contiguous(obj));
+
+	offset = i915_gem_object_get_dma_address(obj, n);
+	offset -= obj->mm.region->region.start;
+
+	return io_mapping_map_wc(&obj->mm.region->iomap, offset, size);
+}
+
 bool i915_gem_object_is_lmem(struct drm_i915_gem_object *obj)
 {
 	struct intel_memory_region *region = obj->mm.region;
@@ -229,6 +244,32 @@  bool i915_gem_object_is_devmem(struct drm_i915_gem_object *obj)
 	return region && region->is_devmem;
 }
 
+struct drm_i915_gem_object *
+i915_gem_object_create_lmem_from_data(struct drm_i915_private *i915,
+				      const void *data, size_t size)
+{
+	struct drm_i915_gem_object *obj;
+	void *map;
+
+	obj = i915_gem_object_create_lmem(i915,
+					  round_up(size, PAGE_SIZE),
+					  I915_BO_ALLOC_CONTIGUOUS);
+	if (IS_ERR(obj))
+		return obj;
+
+	map = i915_gem_object_pin_map_unlocked(obj, I915_MAP_WC);
+	if (IS_ERR(map)) {
+		i915_gem_object_put(obj);
+		return map;
+	}
+
+	memcpy(map, data, size);
+
+	i915_gem_object_unpin_map(obj);
+
+	return obj;
+}
+
 struct drm_i915_gem_object *
 i915_gem_object_create_lmem(struct drm_i915_private *i915,
 			    resource_size_t size,
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_lmem.h b/drivers/gpu/drm/i915/gem/i915_gem_lmem.h
index a1b6a10050bf..e11e0545e39c 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_lmem.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_lmem.h
@@ -14,6 +14,10 @@  struct intel_memory_region;
 
 extern const struct drm_i915_gem_object_ops i915_gem_lmem_obj_ops;
 
+void __iomem *
+i915_gem_object_lmem_io_map(struct drm_i915_gem_object *obj,
+			    unsigned long n,
+			    unsigned long size);
 void __iomem *i915_gem_object_lmem_io_map_page(struct drm_i915_gem_object *obj,
 					       unsigned long n);
 void __iomem *
@@ -23,6 +27,10 @@  i915_gem_object_lmem_io_map_page_atomic(struct drm_i915_gem_object *obj,
 bool i915_gem_object_is_lmem(struct drm_i915_gem_object *obj);
 bool i915_gem_object_is_devmem(struct drm_i915_gem_object *obj);
 
+struct drm_i915_gem_object *
+i915_gem_object_create_lmem_from_data(struct drm_i915_private *i915,
+				      const void *data, size_t size);
+
 struct drm_i915_gem_object *
 i915_gem_object_create_lmem(struct drm_i915_private *i915,
 			    resource_size_t size,
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.c b/drivers/gpu/drm/i915/gt/uc/intel_guc.c
index b54b9de31c3e..703726825c50 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.c
@@ -3,6 +3,7 @@ 
  * Copyright © 2014-2019 Intel Corporation
  */
 
+#include "gem/i915_gem_lmem.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_irq.h"
 #include "gt/intel_gt_pm_irq.h"
@@ -650,7 +651,13 @@  struct i915_vma *intel_guc_allocate_vma(struct intel_guc *guc, u32 size)
 	u64 flags;
 	int ret;
 
-	obj = i915_gem_object_create_shmem(gt->i915, size);
+	if (HAS_LMEM(gt->i915))
+		obj = i915_gem_object_create_lmem(gt->i915, size,
+						  I915_BO_ALLOC_CPU_CLEAR |
+						  I915_BO_ALLOC_CONTIGUOUS);
+	else
+		obj = i915_gem_object_create_shmem(gt->i915, size);
+
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c
index f9d0907ea1a5..8790052f1562 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c
@@ -41,7 +41,7 @@  static void guc_prepare_xfer(struct intel_uncore *uncore)
 }
 
 /* Copy RSA signature from the fw image to HW for verification */
-static void guc_xfer_rsa(struct intel_uc_fw *guc_fw,
+static int guc_xfer_rsa(struct intel_uc_fw *guc_fw,
 			 struct intel_uncore *uncore)
 {
 	u32 rsa[UOS_RSA_SCRATCH_COUNT];
@@ -49,10 +49,13 @@  static void guc_xfer_rsa(struct intel_uc_fw *guc_fw,
 	int i;
 
 	copied = intel_uc_fw_copy_rsa(guc_fw, rsa, sizeof(rsa));
-	GEM_BUG_ON(copied < sizeof(rsa));
+	if (copied < sizeof(rsa))
+		return -ENOMEM;
 
 	for (i = 0; i < UOS_RSA_SCRATCH_COUNT; i++)
 		intel_uncore_write(uncore, UOS_RSA_SCRATCH(i), rsa[i]);
+
+	return 0;
 }
 
 /*
@@ -142,7 +145,9 @@  int intel_guc_fw_upload(struct intel_guc *guc)
 	 * by the DMA engine in one operation, whereas the RSA signature is
 	 * loaded via MMIO.
 	 */
-	guc_xfer_rsa(&guc->fw, uncore);
+	ret = guc_xfer_rsa(&guc->fw, uncore);
+	if (ret)
+		goto out;
 
 	/*
 	 * Current uCode expects the code to be loaded at 8k; locations below
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_huc.c b/drivers/gpu/drm/i915/gt/uc/intel_huc.c
index 56d2144dc6a0..c70bd024f1e1 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_huc.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_huc.c
@@ -87,17 +87,25 @@  static int intel_huc_rsa_data_create(struct intel_huc *huc)
 									vma->obj, true));
 	if (IS_ERR(vaddr)) {
 		i915_vma_unpin_and_release(&vma, 0);
-		return PTR_ERR(vaddr);
+		err = PTR_ERR(vaddr);
+		goto unpin_out;
 	}
 
 	copied = intel_uc_fw_copy_rsa(&huc->fw, vaddr, vma->size);
-	GEM_BUG_ON(copied < huc->fw.rsa_size);
-
 	i915_gem_object_unpin_map(vma->obj);
 
+	if (copied < huc->fw.rsa_size) {
+		err = -ENOMEM;
+		goto unpin_out;
+	}
+
 	huc->rsa_data = vma;
 
 	return 0;
+
+unpin_out:
+	i915_vma_unpin_and_release(&vma, 0);
+	return err;
 }
 
 static void intel_huc_rsa_data_destroy(struct intel_huc *huc)
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
index b05076d190cc..795eca2bd5b4 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
@@ -7,6 +7,7 @@ 
 #include <linux/firmware.h>
 #include <drm/drm_print.h>
 
+#include "gem/i915_gem_lmem.h"
 #include "intel_uc_fw.h"
 #include "intel_uc_fw_abi.h"
 #include "i915_drv.h"
@@ -371,7 +372,11 @@  int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw)
 	if (uc_fw->type == INTEL_UC_FW_TYPE_GUC)
 		uc_fw->private_data_size = css->private_data_size;
 
-	obj = i915_gem_object_create_shmem_from_data(i915, fw->data, fw->size);
+	if (HAS_LMEM(i915))
+		obj = i915_gem_object_create_lmem_from_data(i915, fw->data, fw->size);
+	else
+		obj = i915_gem_object_create_shmem_from_data(i915, fw->data, fw->size);
+
 	if (IS_ERR(obj)) {
 		err = PTR_ERR(obj);
 		goto fail;
@@ -420,14 +425,19 @@  static void uc_fw_bind_ggtt(struct intel_uc_fw *uc_fw)
 		.pages = obj->mm.pages,
 		.vm = &ggtt->vm,
 	};
+	u32 pte_flags = 0;
 
 	GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
 	GEM_BUG_ON(dummy.node.size > ggtt->uc_fw.size);
 
 	/* uc_fw->obj cache domains were not controlled across suspend */
-	drm_clflush_sg(dummy.pages);
+	if (i915_gem_object_has_struct_page(obj))
+		drm_clflush_sg(dummy.pages);
+
+	if (i915_gem_object_is_lmem(obj))
+		pte_flags |= PTE_LM;
 
-	ggtt->vm.insert_entries(&ggtt->vm, &dummy, I915_CACHE_NONE, 0);
+	ggtt->vm.insert_entries(&ggtt->vm, &dummy, I915_CACHE_NONE, pte_flags);
 }
 
 static void uc_fw_unbind_ggtt(struct intel_uc_fw *uc_fw)
@@ -592,7 +602,24 @@  size_t intel_uc_fw_copy_rsa(struct intel_uc_fw *uc_fw, void *dst, u32 max_len)
 
 	GEM_BUG_ON(!intel_uc_fw_is_available(uc_fw));
 
-	return sg_pcopy_to_buffer(pages->sgl, pages->nents, dst, size, offset);
+	if (i915_gem_object_is_lmem(uc_fw->obj)) {
+		unsigned long page_idx = offset >> PAGE_SHIFT;
+		unsigned int page_off = offset_in_page(offset);
+		void __iomem *vaddr;
+
+		vaddr = i915_gem_object_lmem_io_map(uc_fw->obj,
+						    page_idx,
+						    page_off + size);
+		if (!vaddr)
+			return 0;
+
+		memcpy(dst, vaddr + page_off, size);
+		io_mapping_unmap(vaddr);
+		return size;
+	} else {
+		return sg_pcopy_to_buffer(pages->sgl, pages->nents,
+					  dst, size, offset);
+	}
 }
 
 /**