All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dave Gordon <david.s.gordon@intel.com>
To: intel-gfx@lists.freedesktop.org
Subject: [PATCH v4 2/4] drm/i915: optimise i915_gem_object_map() for small objects
Date: Fri, 20 May 2016 01:31:28 +0100	[thread overview]
Message-ID: <1463704290-3491-2-git-send-email-david.s.gordon@intel.com> (raw)
In-Reply-To: <1463704290-3491-1-git-send-email-david.s.gordon@intel.com>

We're using this function for ringbuffers and other "small" objects, so
it's worth avoiding an extra malloc()/free() cycle if the page array is
small enough to put on the stack. Here we've chosen an arbitrary cutoff
of 32 (4k) pages, which is big enough for a ringbuffer (4 pages) or a
context image (currently up to 22 pages).

v5:
    change name of local array [Chris Wilson]

Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/i915_gem.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index ac5fbf2..5797fa2 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2404,7 +2404,8 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj)
 	unsigned long n_pages = obj->base.size >> PAGE_SHIFT;
 	struct sg_table *sgt = obj->pages;
 	struct sg_page_iter sg_iter;
-	struct page **pages;
+	struct page *stack_pages[32];
+	struct page **pages = stack_pages;
 	unsigned long i = 0;
 	void *addr;
 
@@ -2412,9 +2413,12 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj)
 	if (n_pages == 1)
 		return kmap(sg_page(sgt->sgl));
 
-	pages = drm_malloc_gfp(n_pages, sizeof(*pages), GFP_TEMPORARY);
-	if (pages == NULL)
-		return NULL;
+	if (n_pages > ARRAY_SIZE(stack_pages)) {
+		/* Too big for stack -- allocate temporary array instead */
+		pages = drm_malloc_gfp(n_pages, sizeof(*pages), GFP_TEMPORARY);
+		if (pages == NULL)
+			return NULL;
+	}
 
 	for_each_sg_page(sgt->sgl, &sg_iter, sgt->nents, 0)
 		pages[i++] = sg_page_iter_page(&sg_iter);
@@ -2424,7 +2428,8 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj)
 
 	addr = vmap(pages, n_pages, 0, PAGE_KERNEL);
 
-	drm_free_large(pages);
+	if (pages != stack_pages)
+		drm_free_large(pages);
 
 	return addr;
 }
-- 
1.9.1

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

  reply	other threads:[~2016-05-20  0:31 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-05-20  0:31 [PATCH v4 1/4] drm/i915: refactor i915_gem_object_pin_map() Dave Gordon
2016-05-20  0:31 ` Dave Gordon [this message]
2016-05-20  0:31 ` [PATCH v4 3/4] Introduce & use new lightweight SGL iterators Dave Gordon
2016-05-20  8:15   ` Chris Wilson
2016-05-20  9:56     ` Chris Wilson
2016-05-20  0:31 ` [PATCH v4 4/4] Try inlining sg_next() Dave Gordon
2016-05-20  7:35   ` Chris Wilson
2016-05-20  9:18 ` ✗ Ro.CI.BAT: failure for series starting with [v4,1/4] drm/i915: refactor i915_gem_object_pin_map() Patchwork

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=1463704290-3491-2-git-send-email-david.s.gordon@intel.com \
    --to=david.s.gordon@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.