[Intel-gfx] [PATCH 48/70] drm/i915: Introduce an internal allocator for disposable private objects
Chris Wilson
chris at chris-wilson.co.uk
Tue Apr 7 08:21:12 PDT 2015
Quite a few of our objects used for internal hardware programming do not
benefit from being swappable or from being zero initialised. As such
they do not benefit from using a shmemfs backing storage and since they
are internal and never directly exposed to the user, we do not need to
worry about providing a filp. For these we can use an
drm_i915_gem_object wrapper around a sg_table of plain struct page. They
are not swapped backed and not automatically pinned. If they are reaped
by the shrinker, the pages are released and the contents discarded. For
the internal use case, this is fine as for example, ringbuffers are
pinned from being written by a request to be read by the hardware. Once
they are idle, they can be discarded entirely. As such they are a good
match for execlist ringbuffers and a small varierty of other internal
objects.
Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
drivers/gpu/drm/i915/Makefile | 1 +
drivers/gpu/drm/i915/i915_drv.h | 5 +
drivers/gpu/drm/i915/i915_gem_batch_pool.c | 25 ++---
drivers/gpu/drm/i915/i915_gem_internal.c | 149 +++++++++++++++++++++++++++
drivers/gpu/drm/i915/i915_gem_render_state.c | 2 +-
drivers/gpu/drm/i915/intel_ringbuffer.c | 10 +-
6 files changed, 168 insertions(+), 24 deletions(-)
create mode 100644 drivers/gpu/drm/i915/i915_gem_internal.c
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index a69002e2257d..0054a058477d 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -27,6 +27,7 @@ i915-y += i915_cmd_parser.o \
i915_gem_evict.o \
i915_gem_execbuffer.o \
i915_gem_gtt.o \
+ i915_gem_internal.o \
i915_gem.o \
i915_gem_shrinker.o \
i915_gem_stolen.o \
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 97f5d266b17c..c710a9ea1458 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -3027,6 +3027,11 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_device *dev,
u32 gtt_offset,
u32 size);
+/* i915_gem_internal.c */
+struct drm_i915_gem_object *
+i915_gem_object_create_internal(struct drm_device *dev,
+ unsigned size);
+
/* i915_gem_shrinker.c */
unsigned long i915_gem_shrink(struct drm_i915_private *dev_priv,
long target,
diff --git a/drivers/gpu/drm/i915/i915_gem_batch_pool.c b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
index 088020a1afc4..e68c439bceda 100644
--- a/drivers/gpu/drm/i915/i915_gem_batch_pool.c
+++ b/drivers/gpu/drm/i915/i915_gem_batch_pool.c
@@ -101,7 +101,7 @@ i915_gem_batch_pool_get(struct i915_gem_batch_pool *pool,
struct drm_i915_gem_object *obj = NULL;
struct drm_i915_gem_object *tmp;
struct list_head *list;
- int n;
+ int n, ret;
WARN_ON(!mutex_is_locked(&pool->dev->struct_mutex));
@@ -127,13 +127,6 @@ i915_gem_batch_pool_get(struct i915_gem_batch_pool *pool,
BUG_ON(tmp->last_write_req);
}
- /* While we're looping, do some clean up */
- if (tmp->madv == __I915_MADV_PURGED) {
- list_del(&tmp->batch_pool_link);
- drm_gem_object_unreference(&tmp->base);
- continue;
- }
-
if (tmp->base.size >= size) {
obj = tmp;
break;
@@ -141,20 +134,16 @@ i915_gem_batch_pool_get(struct i915_gem_batch_pool *pool,
}
if (obj == NULL) {
- int ret;
-
- obj = i915_gem_alloc_object(pool->dev, size);
+ obj = i915_gem_object_create_internal(pool->dev, size);
if (obj == NULL)
return ERR_PTR(-ENOMEM);
-
- ret = i915_gem_object_get_pages(obj);
- if (ret)
- return ERR_PTR(ret);
-
- obj->madv = I915_MADV_DONTNEED;
}
- list_move_tail(&obj->batch_pool_link, list);
+ ret = i915_gem_object_get_pages(obj);
+ if (ret)
+ return ERR_PTR(ret);
+
i915_gem_object_pin_pages(obj);
+ list_move_tail(&obj->batch_pool_link, list);
return obj;
}
diff --git a/drivers/gpu/drm/i915/i915_gem_internal.c b/drivers/gpu/drm/i915/i915_gem_internal.c
new file mode 100644
index 000000000000..583908392ff5
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_internal.c
@@ -0,0 +1,149 @@
+/*
+ * Copyright © 2015 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include <drm/drmP.h>
+#include <drm/i915_drm.h>
+#include "i915_drv.h"
+
+static void __i915_gem_object_free_pages(struct sg_table *st)
+{
+ struct sg_page_iter sg_iter;
+
+ for_each_sg_page(st->sgl, &sg_iter, st->nents, 0)
+ page_cache_release(sg_page_iter_page(&sg_iter));
+
+ sg_free_table(st);
+ kfree(st);
+}
+
+static int i915_gem_object_get_pages_internal(struct drm_i915_gem_object *obj)
+{
+ const unsigned npages = obj->base.size / PAGE_SIZE;
+ struct sg_table *st;
+ struct scatterlist *sg;
+ unsigned long last_pfn = 0; /* suppress gcc warning */
+ gfp_t gfp;
+ int i;
+
+ st = kmalloc(sizeof(*st), GFP_KERNEL);
+ if (st == NULL)
+ return -ENOMEM;
+
+ if (sg_alloc_table(st, npages, GFP_KERNEL)) {
+ kfree(st);
+ return -ENOMEM;
+ }
+
+ sg = st->sgl;
+ st->nents = 0;
+
+ gfp = GFP_KERNEL;
+ gfp |= __GFP_NORETRY | __GFP_NOWARN | __GFP_NO_KSWAPD;
+ gfp &= ~(__GFP_IO | __GFP_WAIT);
+ for (i = 0; i < npages; i++) {
+ struct page *page;
+
+ page = alloc_page(gfp);
+ if (page == NULL) {
+ i915_gem_shrink_all(to_i915(obj->base.dev));
+ page = alloc_page(GFP_KERNEL);
+ if (page == NULL)
+ goto err;
+ }
+
+ /* XXX page allocator needs to check for SNB bugs */
+
+#ifdef CONFIG_SWIOTLB
+ if (swiotlb_nr_tbl()) {
+ st->nents++;
+ sg_set_page(sg, page, PAGE_SIZE, 0);
+ sg = sg_next(sg);
+ continue;
+ }
+#endif
+ if (!i || page_to_pfn(page) != last_pfn + 1) {
+ if (i)
+ sg = sg_next(sg);
+ st->nents++;
+ sg_set_page(sg, page, PAGE_SIZE, 0);
+ } else {
+ sg->length += PAGE_SIZE;
+ }
+ last_pfn = page_to_pfn(page);
+ }
+#ifdef CONFIG_SWIOTLB
+ if (!swiotlb_nr_tbl())
+#endif
+ sg_mark_end(sg);
+ obj->pages = st;
+ obj->madv = I915_MADV_DONTNEED;
+
+ return 0;
+
+err:
+ sg_mark_end(sg);
+ __i915_gem_object_free_pages(st);
+ return -ENOMEM;
+}
+
+static void i915_gem_object_put_pages_internal(struct drm_i915_gem_object *obj)
+{
+ __i915_gem_object_free_pages(obj->pages);
+
+ obj->dirty = 0;
+ obj->madv = I915_MADV_WILLNEED;
+}
+
+static const struct drm_i915_gem_object_ops i915_gem_object_internal_ops = {
+ .get_pages = i915_gem_object_get_pages_internal,
+ .put_pages = i915_gem_object_put_pages_internal,
+};
+
+
+/**
+ * Creates a new object that wraps some internal memory for private use.
+ * This object is not backed by swappable storage, and as such its contents
+ * are volatile and only valid whilst pinned. If the object is reaped by the
+ * shrinker, its pages and data will be discarded. Equally, it is not a full
+ * GEM object and so not valid for access from userspace. This makes it useful
+ * for hardware interfaces like ringbuffers (which are pinned from the time
+ * the request is written to the time the hardware stops accessing it), but
+ * not for contexts (which need to be preserved when not active for later
+ * reuse).
+ */
+struct drm_i915_gem_object *
+i915_gem_object_create_internal(struct drm_device *dev,
+ unsigned size)
+{
+ struct drm_i915_gem_object *obj;
+
+ obj = i915_gem_object_alloc(dev);
+ if (obj == NULL)
+ return NULL;
+
+ drm_gem_private_object_init(dev, &obj->base, size);
+ i915_gem_object_init(obj, &i915_gem_object_internal_ops);
+
+ return obj;
+}
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
index 521548a08578..4bb91cdadec9 100644
--- a/drivers/gpu/drm/i915/i915_gem_render_state.c
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -57,7 +57,7 @@ static int render_state_init(struct render_state *so, struct drm_device *dev)
if (so->rodata->batch_items * 4 > 4096)
return -EINVAL;
- so->obj = i915_gem_alloc_object(dev, 4096);
+ so->obj = i915_gem_object_create_internal(dev, 4096);
if (so->obj == NULL)
return -ENOMEM;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index b6b2e076fed4..441adc8fa535 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -669,7 +669,7 @@ intel_init_pipe_control(struct intel_engine_cs *ring)
WARN_ON(ring->scratch.obj);
- ring->scratch.obj = i915_gem_alloc_object(ring->dev, 4096);
+ ring->scratch.obj = i915_gem_object_create_internal(ring->dev, 4096);
if (ring->scratch.obj == NULL) {
DRM_ERROR("Failed to allocate seqno page\n");
ret = -ENOMEM;
@@ -1845,7 +1845,7 @@ static int init_status_page(struct intel_engine_cs *ring)
unsigned flags;
int ret;
- obj = i915_gem_alloc_object(ring->dev, 4096);
+ obj = i915_gem_object_create_internal(ring->dev, 4096);
if (obj == NULL) {
DRM_ERROR("Failed to allocate status page\n");
return -ENOMEM;
@@ -1971,7 +1971,7 @@ int intel_alloc_ringbuffer_obj(struct drm_device *dev,
if (!HAS_LLC(dev))
obj = i915_gem_object_create_stolen(dev, ringbuf->size);
if (obj == NULL)
- obj = i915_gem_alloc_object(dev, ringbuf->size);
+ obj = i915_gem_object_create_internal(dev, ringbuf->size);
if (obj == NULL)
return -ENOMEM;
@@ -2446,7 +2446,7 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
if (INTEL_INFO(dev)->gen >= 8) {
if (i915_semaphore_is_enabled(dev)) {
- obj = i915_gem_alloc_object(dev, 4096);
+ obj = i915_gem_object_create_internal(dev, 4096);
if (obj == NULL) {
DRM_ERROR("Failed to allocate semaphore bo. Disabling semaphores\n");
i915.semaphores = 0;
@@ -2552,7 +2552,7 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
/* Workaround batchbuffer to combat CS tlb bug. */
if (HAS_BROKEN_CS_TLB(dev)) {
- obj = i915_gem_alloc_object(dev, I830_WA_SIZE);
+ obj = i915_gem_object_create_internal(dev, I830_WA_SIZE);
if (obj == NULL) {
DRM_ERROR("Failed to allocate batch bo\n");
return -ENOMEM;
--
2.1.4
More information about the Intel-gfx
mailing list