intel-gfx.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Matthew Auld <matthew.auld@intel.com>
To: intel-gfx@lists.freedesktop.org
Subject: [PATCH 14/21] drm/i915: support 64K pages for the 48b PPGTT
Date: Fri, 29 Sep 2017 17:10:25 +0100	[thread overview]
Message-ID: <20170929161032.24394-15-matthew.auld@intel.com> (raw)
In-Reply-To: <20170929161032.24394-1-matthew.auld@intel.com>

Support inserting 64K pages into the 48b PPGTT.

v2: check for 64K scratch

v3: we should only have to re-adjust maybe_64K at every sg interval

Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem_gtt.c | 31 +++++++++++++++++++++++++++++++
 drivers/gpu/drm/i915/i915_gem_gtt.h |  7 +++++++
 2 files changed, 38 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 37029a42c701..629684c23a0e 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -1070,6 +1070,7 @@ static void gen8_ppgtt_insert_huge_entries(struct i915_vma *vma,
 		struct i915_page_directory_pointer *pdp = pdps[idx.pml4e];
 		struct i915_page_directory *pd = pdp->page_directory[idx.pdpe];
 		unsigned int page_size;
+		bool maybe_64K = false;
 		gen8_pte_t encode = pte_encode;
 		gen8_pte_t *vaddr;
 		u16 index, max;
@@ -1091,6 +1092,13 @@ static void gen8_ppgtt_insert_huge_entries(struct i915_vma *vma,
 			max = GEN8_PTES;
 			page_size = I915_GTT_PAGE_SIZE;
 
+			if (!index &&
+			    vma->page_sizes.sg & I915_GTT_PAGE_SIZE_64K &&
+			    IS_ALIGNED(iter->dma, I915_GTT_PAGE_SIZE_64K) &&
+			    (IS_ALIGNED(rem, I915_GTT_PAGE_SIZE_64K) ||
+			     rem >= (max - index) << PAGE_SHIFT))
+				maybe_64K = true;
+
 			vaddr = kmap_atomic_px(pt);
 		}
 
@@ -1110,12 +1118,35 @@ static void gen8_ppgtt_insert_huge_entries(struct i915_vma *vma,
 				iter->dma = sg_dma_address(iter->sg);
 				iter->max = iter->dma + rem;
 
+				if (maybe_64K && index < max &&
+				    !(IS_ALIGNED(iter->dma, I915_GTT_PAGE_SIZE_64K) &&
+				      (IS_ALIGNED(rem, I915_GTT_PAGE_SIZE_64K) ||
+				       rem >= (max - index) << PAGE_SHIFT)))
+					maybe_64K = false;
+
 				if (unlikely(!IS_ALIGNED(iter->dma, page_size)))
 					break;
 			}
 		} while (rem >= page_size && index < max);
 
 		kunmap_atomic(vaddr);
+
+		/*
+		 * Is it safe to mark the 2M block as 64K? -- Either we have
+		 * filled whole page-table with 64K entries, or filled part of
+		 * it and have reached the end of the sg table and we have
+		 * enough padding.
+		 */
+		if (maybe_64K &&
+		    (index == max ||
+		     (i915_vm_has_scratch_64K(vma->vm) &&
+		      !iter->sg && IS_ALIGNED(vma->node.start +
+					      vma->node.size,
+					      I915_GTT_PAGE_SIZE_2M)))) {
+			vaddr = kmap_atomic_px(pd);
+			vaddr[idx.pde] |= GEN8_PDE_IPS_64K;
+			kunmap_atomic(vaddr);
+		}
 	} while (iter->sg);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index 0a31dc369c28..475e4cf042be 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -153,6 +153,7 @@ typedef u64 gen8_ppgtt_pml4e_t;
 #define GEN8_PPAT_GET_AGE(x) ((x) & (3 << 4))
 #define CHV_PPAT_GET_SNOOP(x) ((x) & (1 << 6))
 
+#define GEN8_PDE_IPS_64K BIT(11)
 #define GEN8_PDE_PS_2M   BIT(7)
 
 struct sg_table;
@@ -351,6 +352,12 @@ i915_vm_is_48bit(const struct i915_address_space *vm)
 	return (vm->total - 1) >> 32;
 }
 
+static inline bool
+i915_vm_has_scratch_64K(struct i915_address_space *vm)
+{
+	return vm->scratch_page.order == get_order(I915_GTT_PAGE_SIZE_64K);
+}
+
 /* The Graphics Translation Table is the way in which GEN hardware translates a
  * Graphics Virtual Address into a Physical Address. In addition to the normal
  * collateral associated with any va->pa translations GEN hardware also has a
-- 
2.13.5

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  parent reply	other threads:[~2017-09-29 16:11 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-09-29 16:10 [PATCH 00/21] huge gtt pages Matthew Auld
2017-09-29 16:10 ` [PATCH 01/21] mm/shmem: introduce shmem_file_setup_with_mnt Matthew Auld
2017-09-29 16:10 ` [PATCH 02/21] drm/i915: introduce simple gemfs Matthew Auld
2017-10-02 11:55   ` Joonas Lahtinen
2017-09-29 16:10 ` [PATCH 03/21] drm/i915/gemfs: enable THP Matthew Auld
2017-09-29 16:18   ` Chris Wilson
2017-10-02 12:05   ` Joonas Lahtinen
2017-09-29 16:10 ` [PATCH 04/21] drm/i915: introduce page_sizes field to dev_info Matthew Auld
2017-09-29 16:10 ` [PATCH 05/21] drm/i915: push set_pages down to the callers Matthew Auld
2017-09-29 16:10 ` [PATCH 06/21] drm/i915: introduce page_size members Matthew Auld
2017-09-29 21:31   ` Chris Wilson
2017-10-03 16:32     ` Chris Wilson
2017-09-29 16:10 ` [PATCH 07/21] drm/i915: introduce vm set_pages/clear_pages Matthew Auld
2017-10-02 12:10   ` Joonas Lahtinen
2017-09-29 16:10 ` [PATCH 08/21] drm/i915: align the vma start to the largest gtt page size Matthew Auld
2017-10-02 12:16   ` Joonas Lahtinen
2017-09-29 16:10 ` [PATCH 09/21] drm/i915: align 64K objects to 2M Matthew Auld
2017-10-02 12:18   ` Joonas Lahtinen
2017-09-29 16:10 ` [PATCH 10/21] drm/i915: enable IPS bit for 64K pages Matthew Auld
2017-10-02 12:21   ` Joonas Lahtinen
2017-09-29 16:10 ` [PATCH 11/21] drm/i915: disable GTT cache for 2M pages Matthew Auld
2017-10-02 12:31   ` Joonas Lahtinen
2017-10-02 13:52     ` Matthew Auld
2017-10-03 12:32       ` Joonas Lahtinen
2017-10-09 12:07     ` Ville Syrjälä
2017-10-09 13:04       ` Matthew Auld
2017-09-29 16:10 ` [PATCH 12/21] drm/i915: support 2M pages for the 48b PPGTT Matthew Auld
2017-09-29 16:10 ` [PATCH 13/21] drm/i915: add support for 64K scratch page Matthew Auld
2017-09-29 16:10 ` Matthew Auld [this message]
2017-09-29 16:10 ` [PATCH 15/21] drm/i915: accurate page size tracking for the ppgtt Matthew Auld
2017-09-29 16:10 ` [PATCH 16/21] drm/i915/debugfs: include some gtt page size metrics Matthew Auld
2017-09-29 16:10 ` [PATCH 17/21] drm/i915/selftests: huge page tests Matthew Auld
2017-09-29 16:22   ` Chris Wilson
2017-10-02 22:11   ` kbuild test robot
2017-10-03  9:49   ` Chris Wilson
2017-09-29 16:10 ` [PATCH 18/21] drm/i915/selftests: mix huge pages Matthew Auld
2017-09-29 16:10 ` [PATCH 19/21] drm/i915: disable platform support for vGPU huge gtt pages Matthew Auld
2017-09-29 16:10 ` [PATCH 20/21] drm/i915: enable platform support for 64K pages Matthew Auld
2017-09-29 16:10 ` [PATCH 21/21] drm/i915: enable platform support for 2M pages Matthew Auld
2017-09-29 16:59 ` ✓ Fi.CI.BAT: success for huge gtt pages (rev10) Patchwork
2017-09-29 18:32 ` ✓ Fi.CI.IGT: " Patchwork
  -- strict thread matches above, loose matches on Subject: below --
2017-10-06 14:50 [PATCH 00/21] huge gtt pages Matthew Auld
2017-10-06 14:50 ` [PATCH 14/21] drm/i915: support 64K pages for the 48b PPGTT Matthew Auld
2017-10-05 15:18 [PATCH 00/21] huge gtt pages Matthew Auld
2017-10-05 15:19 ` [PATCH 14/21] drm/i915: support 64K pages for the 48b PPGTT Matthew Auld
2017-09-22 17:32 [PATCH 00/21] huge gtt pages Matthew Auld
2017-09-22 17:32 ` [PATCH 14/21] drm/i915: support 64K pages for the 48b PPGTT Matthew Auld
2017-09-23  8:53   ` Chris Wilson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170929161032.24394-15-matthew.auld@intel.com \
    --to=matthew.auld@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).