2008-08-01 07:07:16

by Eric Anholt

Subject: [PATCH] PCI: Add pci_read_base() API

From: Matthew Wilcox <[email protected]>

Some devices have a BAR at a non-standard address. The pci_read_base()
API allows us to probe these BARs and fill in a resource for it as if
they were standard BARs.

Signed-off-by: Matthew Wilcox <[email protected]>
---
drivers/pci/probe.c | 47 ++++++++++++++++++++++++++++++++++++++++-------
include/linux/pci.h | 10 ++++++++++
2 files changed, 50 insertions(+), 7 deletions(-)

diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 7098dfb..1518c4f 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -181,13 +181,6 @@ static u64 pci_size(u64 base, u64 maxbase, u64 mask)
return size;
}

-enum pci_bar_type {
- pci_bar_unknown, /* Standard PCI BAR probe */
- pci_bar_io, /* An io port BAR */
- pci_bar_mem32, /* A 32-bit memory BAR */
- pci_bar_mem64, /* A 64-bit memory BAR */
-};
-
static inline enum pci_bar_type decode_bar(struct resource *res, u32 bar)
{
if ((bar & PCI_BASE_ADDRESS_SPACE) == PCI_BASE_ADDRESS_SPACE_IO) {
@@ -300,6 +293,46 @@ static int __pci_read_base(struct pci_dev *dev, enum pci_bar_type type,
goto out;
}

+/**
+ * pci_read_base - Read a BAR from a specified location
+ * @dev: The PCI device to read
+ * @type: The type of BAR to read
+ * @res: A struct resource to be filled in
+ * @reg: The address in PCI config space to read the BAR from.
+ *
+ * Some devices have BARs in unusual places. This function lets a driver ask
+ * the PCI subsystem to read it and place it in the resource tree. If it is
+ * like a ROM BAR with an enable in bit 0, the caller should specify a @type
+ * of io, mem32 or mem64. If it's like a normal BAR with memory type in the
+ * low bits, specify unknown, even if the caller knows what kind of BAR it is.
+ *
+ * Returns -ENXIO if the BAR was not successfully read. If the BAR is read,
+ * but no suitable parent resource can be found for the BAR, this function
+ * returns -ENODEV. If the resource cannot be inserted into the resource tree,
+ * it will return -EBUSY. Note that the resource is still 'live' for these
+ * last two cases; the caller should set res->flags to 0 if this is not wanted.
+ */
+int pci_read_base(struct pci_dev *dev, enum pci_bar_type type,
+ struct resource *res, unsigned int reg)
+{
+ struct pci_bus_region region;
+ struct resource *parent;
+
+ __pci_read_base(dev, type, res, reg);
+ if (!res->flags)
+ return -ENXIO;
+
+ region.start = res->start;
+ region.end = res->end;
+ pcibios_bus_to_resource(dev, res, &region);
+
+ parent = pci_find_parent_resource(dev, res);
+ if (!parent)
+ return -ENODEV;
+ return request_resource(parent, res);
+}
+EXPORT_SYMBOL_GPL(pci_read_base);
+
static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom)
{
unsigned int pos, reg;
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 825be38..fa7e10a 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -312,6 +312,16 @@ struct pci_bus {
#define pci_bus_b(n) list_entry(n, struct pci_bus, node)
#define to_pci_bus(n) container_of(n, struct pci_bus, dev)

+enum pci_bar_type {
+ pci_bar_unknown, /* Standard PCI BAR probe */
+ pci_bar_io, /* An io port BAR */
+ pci_bar_mem32, /* A 32-bit memory BAR */
+ pci_bar_mem64, /* A 64-bit memory BAR */
+};
+
+int pci_read_base(struct pci_dev *dev, enum pci_bar_type type,
+ struct resource *res, unsigned int reg);
+
/*
* Error values that may be returned by PCI functions.
*/
--
1.5.6.3

2008-08-01 07:06:53

by Eric Anholt

Subject: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

From: Keith Packard <[email protected]>

GEM needs to create shmem files and get pages related to a shmem file, and
using this pair of functions is the easiest way to do that.

Signed-off-by: Eric Anholt <[email protected]>
---
include/linux/mm.h | 11 +++++++++++
mm/shmem.c | 17 ++++-------------
2 files changed, 15 insertions(+), 13 deletions(-)

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 866a3db..6e54210 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -715,6 +715,17 @@ static inline int shmem_lock(struct file *file, int lock,
#endif
struct file *shmem_file_setup(char *name, loff_t size, unsigned long flags);

+/* Flag allocation requirements to shmem_getpage and shmem_swp_alloc */
+enum sgp_type {
+ SGP_READ, /* don't exceed i_size, don't allocate page */
+ SGP_CACHE, /* don't exceed i_size, may allocate page */
+ SGP_DIRTY, /* like SGP_CACHE, but set new page dirty */
+ SGP_WRITE, /* may exceed i_size, may allocate page */
+};
+
+int shmem_getpage(struct inode *inode, unsigned long idx,
+ struct page **pagep, enum sgp_type sgp, int *type);
+
int shmem_zero_setup(struct vm_area_struct *);

#ifndef CONFIG_MMU
diff --git a/mm/shmem.c b/mm/shmem.c
index c1e5a3b..7166ff3 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -77,14 +77,6 @@
/* Pretend that each entry is of this size in directory's i_size */
#define BOGO_DIRENT_SIZE 20

-/* Flag allocation requirements to shmem_getpage and shmem_swp_alloc */
-enum sgp_type {
- SGP_READ, /* don't exceed i_size, don't allocate page */
- SGP_CACHE, /* don't exceed i_size, may allocate page */
- SGP_DIRTY, /* like SGP_CACHE, but set new page dirty */
- SGP_WRITE, /* may exceed i_size, may allocate page */
-};
-
#ifdef CONFIG_TMPFS
static unsigned long shmem_default_max_blocks(void)
{
@@ -97,9 +89,6 @@ static unsigned long shmem_default_max_inodes(void)
}
#endif

-static int shmem_getpage(struct inode *inode, unsigned long idx,
- struct page **pagep, enum sgp_type sgp, int *type);
-
static inline struct page *shmem_dir_alloc(gfp_t gfp_mask)
{
/*
@@ -1177,8 +1166,8 @@ static inline struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo)
* vm. If we swap it in we mark it dirty since we also free the swap
* entry since a page cannot live in both the swap and page cache
*/
-static int shmem_getpage(struct inode *inode, unsigned long idx,
- struct page **pagep, enum sgp_type sgp, int *type)
+int shmem_getpage(struct inode *inode, unsigned long idx,
+ struct page **pagep, enum sgp_type sgp, int *type)
{
struct address_space *mapping = inode->i_mapping;
struct shmem_inode_info *info = SHMEM_I(inode);
@@ -1431,6 +1420,7 @@ failed:
}
return error;
}
+EXPORT_SYMBOL(shmem_getpage);

static int shmem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
{
@@ -2582,6 +2572,7 @@ put_memory:
shmem_unacct_size(flags, size);
return ERR_PTR(error);
}
+EXPORT_SYMBOL(shmem_file_setup);

/**
* shmem_zero_setup - setup a shared anonymous mapping
--
1.5.6.3

2008-08-01 07:07:44

by Eric Anholt

Subject: [PATCH] drm: Add GEM ("graphics execution manager") to i915 driver.

GEM allows the creation of persistent buffer objects accessible by the
graphics device through new ioctls for managing execution of commands on the
device. The userland API is almost entirely driver-specific to ensure that
any driver building on this model can easily map the interface to individual
driver requirements.

GEM is used by the 2d driver for managing its internal state allocations and
will be used for pixmap storage to reduce memory consumption and enable
zero-copy GLX_EXT_texture_from_pixmap, and in the 3d driver is used to enable
GL_EXT_framebuffer_object and GL_ARB_pixel_buffer_object.

Signed-off-by: Eric Anholt <[email protected]>
---
drivers/gpu/drm/Makefile | 5 +-
drivers/gpu/drm/drm_agpsupport.c | 51 +-
drivers/gpu/drm/drm_cache.c | 76 +
drivers/gpu/drm/drm_drv.c | 4 +
drivers/gpu/drm/drm_fops.c | 6 +
drivers/gpu/drm/drm_gem.c | 420 ++++++
drivers/gpu/drm/drm_memory.c | 2 +
drivers/gpu/drm/drm_mm.c | 5 +-
drivers/gpu/drm/drm_proc.c | 135 ++-
drivers/gpu/drm/drm_stub.c | 10 +
drivers/gpu/drm/i915/Makefile | 6 +-
drivers/gpu/drm/i915/i915_dma.c | 94 +-
drivers/gpu/drm/i915/i915_drv.c | 8 +-
drivers/gpu/drm/i915/i915_drv.h | 253 ++++-
drivers/gpu/drm/i915/i915_gem.c | 2497 ++++++++++++++++++++++++++++++++
drivers/gpu/drm/i915/i915_gem_debug.c | 201 +++
drivers/gpu/drm/i915/i915_gem_proc.c | 292 ++++
drivers/gpu/drm/i915/i915_gem_tiling.c | 318 ++++
drivers/gpu/drm/i915/i915_irq.c | 8 +-
include/drm/drm.h | 31 +
include/drm/drmP.h | 151 ++
include/drm/i915_drm.h | 332 +++++
22 files changed, 4861 insertions(+), 44 deletions(-)
create mode 100644 drivers/gpu/drm/drm_cache.c
create mode 100644 drivers/gpu/drm/drm_gem.c
create mode 100644 drivers/gpu/drm/i915/i915_gem.c
create mode 100644 drivers/gpu/drm/i915/i915_gem_debug.c
create mode 100644 drivers/gpu/drm/i915/i915_gem_proc.c
create mode 100644 drivers/gpu/drm/i915/i915_gem_tiling.c

diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index e9f9a97..74da994 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -4,8 +4,9 @@

ccflags-y := -Iinclude/drm

-drm-y := drm_auth.o drm_bufs.o drm_context.o drm_dma.o drm_drawable.o \
- drm_drv.o drm_fops.o drm_ioctl.o drm_irq.o \
+drm-y := drm_auth.o drm_bufs.o drm_cache.o \
+ drm_context.o drm_dma.o drm_drawable.o \
+ drm_drv.o drm_fops.o drm_gem.o drm_ioctl.o drm_irq.o \
drm_lock.o drm_memory.o drm_proc.o drm_stub.o drm_vm.o \
drm_agpsupport.o drm_scatter.o ati_pcigart.o drm_pci.o \
drm_sysfs.o drm_hashtab.o drm_sman.o drm_mm.o
diff --git a/drivers/gpu/drm/drm_agpsupport.c b/drivers/gpu/drm/drm_agpsupport.c
index aefa5ac..2639be2 100644
--- a/drivers/gpu/drm/drm_agpsupport.c
+++ b/drivers/gpu/drm/drm_agpsupport.c
@@ -33,6 +33,7 @@

#include "drmP.h"
#include <linux/module.h>
+#include <asm/agp.h>

#if __OS_HAS_AGP

@@ -452,4 +453,52 @@ int drm_agp_unbind_memory(DRM_AGP_MEM * handle)
return agp_unbind_memory(handle);
}

-#endif /* __OS_HAS_AGP */
+/**
+ * Binds a collection of pages into AGP memory at the given offset, returning
+ * the AGP memory structure containing them.
+ *
+ * No reference is held on the pages during this time -- it is up to the
+ * caller to handle that.
+ */
+DRM_AGP_MEM *
+drm_agp_bind_pages(struct drm_device *dev,
+ struct page **pages,
+ unsigned long num_pages,
+ uint32_t gtt_offset)
+{
+ DRM_AGP_MEM *mem;
+ int ret, i;
+
+ DRM_DEBUG("\n");
+
+ mem = drm_agp_allocate_memory(dev->agp->bridge, num_pages,
+ AGP_USER_MEMORY);
+ if (mem == NULL) {
+ DRM_ERROR("Failed to allocate memory for %ld pages\n",
+ num_pages);
+ return NULL;
+ }
+
+ for (i = 0; i < num_pages; i++)
+ mem->memory[i] = phys_to_gart(page_to_phys(pages[i]));
+ mem->page_count = num_pages;
+
+ mem->is_flushed = true;
+ ret = drm_agp_bind_memory(mem, gtt_offset / PAGE_SIZE);
+ if (ret != 0) {
+ DRM_ERROR("Failed to bind AGP memory: %d\n", ret);
+ agp_free_memory(mem);
+ return NULL;
+ }
+
+ return mem;
+}
+EXPORT_SYMBOL(drm_agp_bind_pages);
+
+void drm_agp_chipset_flush(struct drm_device *dev)
+{
+ agp_flush_chipset(dev->agp->bridge);
+}
+EXPORT_SYMBOL(drm_agp_chipset_flush);
+
+#endif /* __OS_HAS_AGP */
diff --git a/drivers/gpu/drm/drm_cache.c b/drivers/gpu/drm/drm_cache.c
new file mode 100644
index 0000000..9475f7d
--- /dev/null
+++ b/drivers/gpu/drm/drm_cache.c
@@ -0,0 +1,76 @@
+/**************************************************************************
+ *
+ * Copyright (c) 2006-2007 Tungsten Graphics, Inc., Cedar Park, TX., USA
+ * All Rights Reserved.
+ *
+ * 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, sub license, 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 NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS 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.
+ *
+ **************************************************************************/
+/*
+ * Authors: Thomas Hellström <thomas-at-tungstengraphics-dot-com>
+ */
+
+#include "drmP.h"
+
+#if defined(CONFIG_X86)
+static void
+drm_clflush_page(struct page *page)
+{
+ uint8_t *page_virtual;
+ unsigned int i;
+
+ if (unlikely(page == NULL))
+ return;
+
+ page_virtual = kmap_atomic(page, KM_USER0);
+ for (i = 0; i < PAGE_SIZE; i += boot_cpu_data.x86_clflush_size)
+ clflush(page_virtual + i);
+ kunmap_atomic(page_virtual, KM_USER0);
+}
+#endif
+
+static void
+drm_clflush_ipi_handler(void *null)
+{
+ wbinvd();
+}
+
+void
+drm_clflush_pages(struct page *pages[], unsigned long num_pages)
+{
+
+#if defined(CONFIG_X86)
+ if (cpu_has_clflush) {
+ unsigned long i;
+
+ mb();
+ for (i = 0; i < num_pages; ++i)
+ drm_clflush_page(*pages++);
+ mb();
+
+ return;
+ }
+#endif
+
+ if (on_each_cpu(drm_clflush_ipi_handler, NULL, 1) != 0)
+ DRM_ERROR("Timed out waiting for cache flush.\n");
+}
+EXPORT_SYMBOL(drm_clflush_pages);
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index 452c2d8..f934414 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -117,6 +117,10 @@ static struct drm_ioctl_desc drm_ioctls[] = {
DRM_IOCTL_DEF(DRM_IOCTL_WAIT_VBLANK, drm_wait_vblank, 0),

DRM_IOCTL_DEF(DRM_IOCTL_UPDATE_DRAW, drm_update_drawable_info, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
+
+ DRM_IOCTL_DEF(DRM_IOCTL_GEM_CLOSE, drm_gem_close_ioctl, 0),
+ DRM_IOCTL_DEF(DRM_IOCTL_GEM_FLINK, drm_gem_flink_ioctl, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_IOCTL_GEM_OPEN, drm_gem_open_ioctl, DRM_AUTH),
};

#define DRM_CORE_IOCTL_COUNT ARRAY_SIZE( drm_ioctls )
diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c
index 851a53f..c0a309a 100644
--- a/drivers/gpu/drm/drm_fops.c
+++ b/drivers/gpu/drm/drm_fops.c
@@ -256,6 +256,9 @@ static int drm_open_helper(struct inode *inode, struct file *filp,

INIT_LIST_HEAD(&priv->lhead);

+ if (dev->driver->driver_features & DRIVER_GEM)
+ drm_gem_open(dev, priv);
+
if (dev->driver->open) {
ret = dev->driver->open(dev, priv);
if (ret < 0)
@@ -400,6 +403,9 @@ int drm_release(struct inode *inode, struct file *filp)
dev->driver->reclaim_buffers(dev, file_priv);
}

+ if (dev->driver->driver_features & DRIVER_GEM)
+ drm_gem_release(dev, file_priv);
+
drm_fasync(-1, filp, 0);

mutex_lock(&dev->ctxlist_mutex);
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
new file mode 100644
index 0000000..434155b
--- /dev/null
+++ b/drivers/gpu/drm/drm_gem.c
@@ -0,0 +1,420 @@
+/*
+ * Copyright © 2008 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.
+ *
+ * Authors:
+ * Eric Anholt <[email protected]>
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/mm.h>
+#include <linux/uaccess.h>
+#include <linux/fs.h>
+#include <linux/file.h>
+#include <linux/module.h>
+#include <linux/mman.h>
+#include <linux/pagemap.h>
+#include "drmP.h"
+
+/** @file drm_gem.c
+ *
+ * This file provides some of the base ioctls and library routines for
+ * the graphics memory manager implemented by each device driver.
+ *
+ * Because various devices have different requirements in terms of
+ * synchronization and migration strategies, implementing that is left up to
+ * the driver, and all that the general API provides should be generic --
+ * allocating objects, reading/writing data with the cpu, freeing objects.
+ * Even there, platform-dependent optimizations for reading/writing data with
+ * the CPU mean we'll likely hook those out to driver-specific calls. However,
+ * the DRI2 implementation wants to have at least allocate/mmap be generic.
+ *
+ * The goal was to have swap-backed object allocation managed through
+ * struct file. However, file descriptors as handles to a struct file have
+ * two major failings:
+ * - Process limits prevent more than 1024 or so being used at a time by
+ * default.
+ * - Inability to allocate high fds will aggravate the X Server's select()
+ * handling, and likely that of many GL client applications as well.
+ *
+ * This led to a plan of using our own integer IDs (called handles, following
+ * DRM terminology) to mimic fds, and implement the fd syscalls we need as
+ * ioctls. The objects themselves will still include the struct file so
+ * that we can transition to fds if the required kernel infrastructure shows
+ * up at a later date, and as our interface with shmfs for memory allocation.
+ */
+
+/**
+ * Initialize the GEM device fields
+ */
+
+int
+drm_gem_init(struct drm_device *dev)
+{
+ spin_lock_init(&dev->object_name_lock);
+ idr_init(&dev->object_name_idr);
+ atomic_set(&dev->object_count, 0);
+ atomic_set(&dev->object_memory, 0);
+ atomic_set(&dev->pin_count, 0);
+ atomic_set(&dev->pin_memory, 0);
+ atomic_set(&dev->gtt_count, 0);
+ atomic_set(&dev->gtt_memory, 0);
+ return 0;
+}
+
+/**
+ * Allocate a GEM object of the specified size with shmfs backing store
+ */
+struct drm_gem_object *
+drm_gem_object_alloc(struct drm_device *dev, size_t size)
+{
+ struct drm_gem_object *obj;
+
+ BUG_ON((size & (PAGE_SIZE - 1)) != 0);
+
+ obj = kcalloc(1, sizeof(*obj), GFP_KERNEL);
+
+ obj->dev = dev;
+ obj->filp = shmem_file_setup("drm mm object", size, 0);
+ if (IS_ERR(obj->filp)) {
+ kfree(obj);
+ return NULL;
+ }
+
+ kref_init(&obj->refcount);
+ kref_init(&obj->handlecount);
+ obj->size = size;
+ if (dev->driver->gem_init_object != NULL &&
+ dev->driver->gem_init_object(obj) != 0) {
+ fput(obj->filp);
+ kfree(obj);
+ return NULL;
+ }
+ atomic_inc(&dev->object_count);
+ atomic_add(obj->size, &dev->object_memory);
+ return obj;
+}
+EXPORT_SYMBOL(drm_gem_object_alloc);
+
+/**
+ * Removes the mapping from handle to filp for this object.
+ */
+static int
+drm_gem_handle_delete(struct drm_file *filp, int handle)
+{
+ struct drm_device *dev;
+ struct drm_gem_object *obj;
+
+ /* This is gross. The idr system doesn't let us try a delete and
+ * return an error code. It just spews if you fail at deleting.
+ * So, we have to grab a lock around finding the object and then
+ * doing the delete on it and dropping the refcount, or the user
+ * could race us to double-decrement the refcount and cause a
+ * use-after-free later. Given the frequency of our handle lookups,
+ * we may want to use ida for number allocation and a hash table
+ * for the pointers, anyway.
+ */
+ spin_lock(&filp->table_lock);
+
+ /* Check if we currently have a reference on the object */
+ obj = idr_find(&filp->object_idr, handle);
+ if (obj == NULL) {
+ spin_unlock(&filp->table_lock);
+ return -EINVAL;
+ }
+ dev = obj->dev;
+
+ /* Release reference and decrement refcount. */
+ idr_remove(&filp->object_idr, handle);
+ spin_unlock(&filp->table_lock);
+
+ mutex_lock(&dev->struct_mutex);
+ drm_gem_object_handle_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+
+ return 0;
+}
+
+/**
+ * Create a handle for this object. This adds a handle reference
+ * to the object, which includes a regular reference count. Callers
+ * will likely want to dereference the object afterwards.
+ */
+int
+drm_gem_handle_create(struct drm_file *file_priv,
+ struct drm_gem_object *obj,
+ int *handlep)
+{
+ int ret;
+
+ /*
+ * Get the user-visible handle using idr.
+ */
+again:
+ /* ensure there is space available to allocate a handle */
+ if (idr_pre_get(&file_priv->object_idr, GFP_KERNEL) == 0)
+ return -ENOMEM;
+
+ /* do the allocation under our spinlock */
+ spin_lock(&file_priv->table_lock);
+ ret = idr_get_new_above(&file_priv->object_idr, obj, 1, handlep);
+ spin_unlock(&file_priv->table_lock);
+ if (ret == -EAGAIN)
+ goto again;
+
+ if (ret != 0)
+ return ret;
+
+ drm_gem_object_handle_reference(obj);
+ return 0;
+}
+EXPORT_SYMBOL(drm_gem_handle_create);
+
+/** Returns a reference to the object named by the handle. */
+struct drm_gem_object *
+drm_gem_object_lookup(struct drm_device *dev, struct drm_file *filp,
+ int handle)
+{
+ struct drm_gem_object *obj;
+
+ spin_lock(&filp->table_lock);
+
+ /* Check if we currently have a reference on the object */
+ obj = idr_find(&filp->object_idr, handle);
+ if (obj == NULL) {
+ spin_unlock(&filp->table_lock);
+ return NULL;
+ }
+
+ drm_gem_object_reference(obj);
+
+ spin_unlock(&filp->table_lock);
+
+ return obj;
+}
+EXPORT_SYMBOL(drm_gem_object_lookup);
+
+/**
+ * Releases the handle to an mm object.
+ */
+int
+drm_gem_close_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_gem_close *args = data;
+ int ret;
+
+ if (!(dev->driver->driver_features & DRIVER_GEM))
+ return -ENODEV;
+
+ ret = drm_gem_handle_delete(file_priv, args->handle);
+
+ return ret;
+}
+
+/**
+ * Create a global name for an object, returning the name.
+ *
+ * Note that the name does not hold a reference; when the object
+ * is freed, the name goes away.
+ */
+int
+drm_gem_flink_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_gem_flink *args = data;
+ struct drm_gem_object *obj;
+ int ret;
+
+ if (!(dev->driver->driver_features & DRIVER_GEM))
+ return -ENODEV;
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL)
+ return -EINVAL;
+
+again:
+ if (idr_pre_get(&dev->object_name_idr, GFP_KERNEL) == 0)
+ return -ENOMEM;
+
+ spin_lock(&dev->object_name_lock);
+ if (obj->name) {
+ spin_unlock(&dev->object_name_lock);
+ return -EEXIST;
+ }
+ ret = idr_get_new_above(&dev->object_name_idr, obj, 1,
+ &obj->name);
+ spin_unlock(&dev->object_name_lock);
+ if (ret == -EAGAIN)
+ goto again;
+
+ if (ret != 0) {
+ mutex_lock(&dev->struct_mutex);
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+ }
+
+ /*
+ * Leave the reference from the lookup around as the
+ * name table now holds one
+ */
+ args->name = (uint64_t) obj->name;
+
+ return 0;
+}
+
+/**
+ * Open an object using the global name, returning a handle and the size.
+ *
+ * This handle (of course) holds a reference to the object, so the object
+ * will not go away until the handle is deleted.
+ */
+int
+drm_gem_open_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_gem_open *args = data;
+ struct drm_gem_object *obj;
+ int ret;
+ int handle;
+
+ if (!(dev->driver->driver_features & DRIVER_GEM))
+ return -ENODEV;
+
+ spin_lock(&dev->object_name_lock);
+ obj = idr_find(&dev->object_name_idr, (int) args->name);
+ if (obj)
+ drm_gem_object_reference(obj);
+ spin_unlock(&dev->object_name_lock);
+ if (!obj)
+ return -ENOENT;
+
+ ret = drm_gem_handle_create(file_priv, obj, &handle);
+ mutex_lock(&dev->struct_mutex);
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ if (ret)
+ return ret;
+
+ args->handle = handle;
+ args->size = obj->size;
+
+ return 0;
+}
+
+/**
+ * Called at device open time, sets up the structure for handling refcounting
+ * of mm objects.
+ */
+void
+drm_gem_open(struct drm_device *dev, struct drm_file *file_private)
+{
+ idr_init(&file_private->object_idr);
+ spin_lock_init(&file_private->table_lock);
+}
+
+/**
+ * Called at device close to release the file's
+ * handle references on objects.
+ */
+static int
+drm_gem_object_release_handle(int id, void *ptr, void *data)
+{
+ struct drm_gem_object *obj = ptr;
+
+ drm_gem_object_handle_unreference(obj);
+
+ return 0;
+}
+
+/**
+ * Called at close time when the filp is going away.
+ *
+ * Releases any remaining references on objects by this filp.
+ */
+void
+drm_gem_release(struct drm_device *dev, struct drm_file *file_private)
+{
+ mutex_lock(&dev->struct_mutex);
+ idr_for_each(&file_private->object_idr,
+ &drm_gem_object_release_handle, NULL);
+
+ idr_destroy(&file_private->object_idr);
+ mutex_unlock(&dev->struct_mutex);
+}
+
+/**
+ * Called after the last reference to the object has been lost.
+ *
+ * Frees the object
+ */
+void
+drm_gem_object_free(struct kref *kref)
+{
+ struct drm_gem_object *obj = (struct drm_gem_object *) kref;
+ struct drm_device *dev = obj->dev;
+
+ BUG_ON(!mutex_is_locked(&dev->struct_mutex));
+
+ if (dev->driver->gem_free_object != NULL)
+ dev->driver->gem_free_object(obj);
+
+ fput(obj->filp);
+ atomic_dec(&dev->object_count);
+ atomic_sub(obj->size, &dev->object_memory);
+ kfree(obj);
+}
+EXPORT_SYMBOL(drm_gem_object_free);
+
+/**
+ * Called after the last handle to the object has been closed
+ *
+ * Removes any name for the object. Note that this must be
+ * called before drm_gem_object_free or we'll be touching
+ * freed memory
+ */
+void
+drm_gem_object_handle_free(struct kref *kref)
+{
+ struct drm_gem_object *obj = container_of(kref,
+ struct drm_gem_object,
+ handlecount);
+ struct drm_device *dev = obj->dev;
+
+ /* Remove any name for this object */
+ spin_lock(&dev->object_name_lock);
+ if (obj->name) {
+ idr_remove(&dev->object_name_idr, obj->name);
+ spin_unlock(&dev->object_name_lock);
+ /*
+ * The object name held a reference to this object, drop
+ * that now.
+ */
+ drm_gem_object_unreference(obj);
+ } else
+ spin_unlock(&dev->object_name_lock);
+
+}
+EXPORT_SYMBOL(drm_gem_object_handle_free);
+
diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 0177012..803bc9e 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -133,6 +133,7 @@ int drm_free_agp(DRM_AGP_MEM * handle, int pages)
{
return drm_agp_free_memory(handle) ? 0 : -EINVAL;
}
+EXPORT_SYMBOL(drm_free_agp);

/** Wrapper around agp_bind_memory() */
int drm_bind_agp(DRM_AGP_MEM * handle, unsigned int start)
@@ -145,6 +146,7 @@ int drm_unbind_agp(DRM_AGP_MEM * handle)
{
return drm_agp_unbind_memory(handle);
}
+EXPORT_SYMBOL(drm_unbind_agp);

#else /* __OS_HAS_AGP */
static inline void *agp_remap(unsigned long offset, unsigned long size,
diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c
index dcff9e9..217ad7d 100644
--- a/drivers/gpu/drm/drm_mm.c
+++ b/drivers/gpu/drm/drm_mm.c
@@ -169,6 +169,7 @@ struct drm_mm_node *drm_mm_get_block(struct drm_mm_node * parent,

return child;
}
+EXPORT_SYMBOL(drm_mm_get_block);

/*
* Put a block. Merge with the previous and / or next block if they are free.
@@ -217,6 +218,7 @@ void drm_mm_put_block(struct drm_mm_node * cur)
drm_free(cur, sizeof(*cur), DRM_MEM_MM);
}
}
+EXPORT_SYMBOL(drm_mm_put_block);

struct drm_mm_node *drm_mm_search_free(const struct drm_mm * mm,
unsigned long size,
@@ -265,6 +267,7 @@ int drm_mm_clean(struct drm_mm * mm)

return (head->next->next == head);
}
+EXPORT_SYMBOL(drm_mm_search_free);

int drm_mm_init(struct drm_mm * mm, unsigned long start, unsigned long size)
{
@@ -273,7 +276,7 @@ int drm_mm_init(struct drm_mm * mm, unsigned long start, unsigned long size)

return drm_mm_create_tail_node(mm, start, size);
}
-
+EXPORT_SYMBOL(drm_mm_init);

void drm_mm_takedown(struct drm_mm * mm)
{
diff --git a/drivers/gpu/drm/drm_proc.c b/drivers/gpu/drm/drm_proc.c
index 93b1e04..d490db4 100644
--- a/drivers/gpu/drm/drm_proc.c
+++ b/drivers/gpu/drm/drm_proc.c
@@ -49,6 +49,10 @@ static int drm_queues_info(char *buf, char **start, off_t offset,
int request, int *eof, void *data);
static int drm_bufs_info(char *buf, char **start, off_t offset,
int request, int *eof, void *data);
+static int drm_gem_name_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data);
+static int drm_gem_object_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data);
#if DRM_DEBUG_CODE
static int drm_vma_info(char *buf, char **start, off_t offset,
int request, int *eof, void *data);
@@ -60,13 +64,16 @@ static int drm_vma_info(char *buf, char **start, off_t offset,
static struct drm_proc_list {
const char *name; /**< file name */
int (*f) (char *, char **, off_t, int, int *, void *); /**< proc callback*/
+ u32 driver_features; /**< Required driver features for this entry */
} drm_proc_list[] = {
- {"name", drm_name_info},
- {"mem", drm_mem_info},
- {"vm", drm_vm_info},
- {"clients", drm_clients_info},
- {"queues", drm_queues_info},
- {"bufs", drm_bufs_info},
+ {"name", drm_name_info, 0},
+ {"mem", drm_mem_info, 0},
+ {"vm", drm_vm_info, 0},
+ {"clients", drm_clients_info, 0},
+ {"queues", drm_queues_info, 0},
+ {"bufs", drm_bufs_info, 0},
+ {"gem_names", drm_gem_name_info, DRIVER_GEM},
+ {"gem_objects", drm_gem_object_info, DRIVER_GEM},
#if DRM_DEBUG_CODE
{"vma", drm_vma_info},
#endif
@@ -90,8 +97,9 @@ static struct drm_proc_list {
int drm_proc_init(struct drm_minor *minor, int minor_id,
struct proc_dir_entry *root)
{
+ struct drm_device *dev = minor->dev;
struct proc_dir_entry *ent;
- int i, j;
+ int i, j, ret;
char name[64];

sprintf(name, "%d", minor_id);
@@ -102,23 +110,42 @@ int drm_proc_init(struct drm_minor *minor, int minor_id,
}

for (i = 0; i < DRM_PROC_ENTRIES; i++) {
+ u32 features = drm_proc_list[i].driver_features;
+
+ if (features != 0 &&
+ (dev->driver->driver_features & features) != features)
+ continue;
+
ent = create_proc_entry(drm_proc_list[i].name,
S_IFREG | S_IRUGO, minor->dev_root);
if (!ent) {
DRM_ERROR("Cannot create /proc/dri/%s/%s\n",
name, drm_proc_list[i].name);
- for (j = 0; j < i; j++)
- remove_proc_entry(drm_proc_list[i].name,
- minor->dev_root);
- remove_proc_entry(name, root);
- minor->dev_root = NULL;
- return -1;
+ ret = -1;
+ goto fail;
}
ent->read_proc = drm_proc_list[i].f;
ent->data = minor;
}

+ if (dev->driver->proc_init) {
+ ret = dev->driver->proc_init(minor);
+ if (ret) {
+ DRM_ERROR("DRM: Driver failed to initialize "
+ "/proc/dri.\n");
+ goto fail;
+ }
+ }
+
return 0;
+ fail:
+
+ for (j = 0; j < i; j++)
+ remove_proc_entry(drm_proc_list[i].name,
+ minor->dev_root);
+ remove_proc_entry(name, root);
+ minor->dev_root = NULL;
+ return ret;
}

/**
@@ -133,12 +160,16 @@ int drm_proc_init(struct drm_minor *minor, int minor_id,
*/
int drm_proc_cleanup(struct drm_minor *minor, struct proc_dir_entry *root)
{
+ struct drm_device *dev = minor->dev;
int i;
char name[64];

if (!root || !minor->dev_root)
return 0;

+ if (dev->driver->proc_cleanup)
+ dev->driver->proc_cleanup(minor);
+
for (i = 0; i < DRM_PROC_ENTRIES; i++)
remove_proc_entry(drm_proc_list[i].name, minor->dev_root);
sprintf(name, "%d", minor->index);
@@ -480,6 +511,84 @@ static int drm_clients_info(char *buf, char **start, off_t offset,
return ret;
}

+struct drm_gem_name_info_data {
+ int len;
+ char *buf;
+ int eof;
+};
+
+static int drm_gem_one_name_info(int id, void *ptr, void *data)
+{
+ struct drm_gem_object *obj = ptr;
+ struct drm_gem_name_info_data *nid = data;
+
+ DRM_INFO("name %d size %d\n", obj->name, obj->size);
+ if (nid->eof)
+ return 0;
+
+ nid->len += sprintf(&nid->buf[nid->len],
+ "%6d%9d%8d%9d\n",
+ obj->name, obj->size,
+ atomic_read(&obj->handlecount.refcount),
+ atomic_read(&obj->refcount.refcount));
+ if (nid->len > DRM_PROC_LIMIT) {
+ nid->eof = 1;
+ return 0;
+ }
+ return 0;
+}
+
+static int drm_gem_name_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ struct drm_gem_name_info_data nid;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ nid.len = sprintf(buf, " name size handles refcount\n");
+ nid.buf = buf;
+ nid.eof = 0;
+ idr_for_each(&dev->object_name_idr, drm_gem_one_name_info, &nid);
+
+ *start = &buf[offset];
+ *eof = 0;
+ if (nid.len > request + offset)
+ return request;
+ *eof = 1;
+ return nid.len - offset;
+}
+
+static int drm_gem_object_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ int len = 0;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ *start = &buf[offset];
+ *eof = 0;
+ DRM_PROC_PRINT("%d objects\n", atomic_read(&dev->object_count));
+ DRM_PROC_PRINT("%d object bytes\n", atomic_read(&dev->object_memory));
+ DRM_PROC_PRINT("%d pinned\n", atomic_read(&dev->pin_count));
+ DRM_PROC_PRINT("%d pin bytes\n", atomic_read(&dev->pin_memory));
+ DRM_PROC_PRINT("%d gtt bytes\n", atomic_read(&dev->gtt_memory));
+ DRM_PROC_PRINT("%d gtt total\n", dev->gtt_total);
+ if (len > request + offset)
+ return request;
+ *eof = 1;
+ return len - offset;
+}
+
#if DRM_DEBUG_CODE

static int drm__vma_info(char *buf, char **start, off_t offset, int request,
diff --git a/drivers/gpu/drm/drm_stub.c b/drivers/gpu/drm/drm_stub.c
index c2f584f..82f4657 100644
--- a/drivers/gpu/drm/drm_stub.c
+++ b/drivers/gpu/drm/drm_stub.c
@@ -152,6 +152,15 @@ static int drm_fill_in_dev(struct drm_device * dev, struct pci_dev *pdev,
goto error_out_unreg;
}

+ if (driver->driver_features & DRIVER_GEM) {
+ retcode = drm_gem_init(dev);
+ if (retcode) {
+ DRM_ERROR("Cannot initialize graphics execution "
+ "manager (GEM)\n");
+ goto error_out_unreg;
+ }
+ }
+
return 0;

error_out_unreg:
@@ -317,6 +326,7 @@ int drm_put_dev(struct drm_device * dev)
int drm_put_minor(struct drm_minor **minor_p)
{
struct drm_minor *minor = *minor_p;
+
DRM_DEBUG("release secondary minor %d\n", minor->index);

if (minor->type == DRM_MINOR_LEGACY)
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index a9e6046..a6319d1 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -3,7 +3,11 @@
# Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.

ccflags-y := -Iinclude/drm
-i915-y := i915_drv.o i915_dma.o i915_irq.o i915_mem.o
+i915-y := i915_drv.o i915_dma.o i915_irq.o i915_mem.o \
+ i915_gem.o \
+ i915_gem_debug.o \
+ i915_gem_proc.o \
+ i915_gem_tiling.o

i915-$(CONFIG_COMPAT) += i915_ioc32.o

diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index b3c4ac9..1556852 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -170,24 +170,31 @@ static int i915_initialize(struct drm_device * dev, drm_i915_init_t * init)
dev_priv->sarea_priv = (drm_i915_sarea_t *)
((u8 *) dev_priv->sarea->handle + init->sarea_priv_offset);

- dev_priv->ring.Start = init->ring_start;
- dev_priv->ring.End = init->ring_end;
- dev_priv->ring.Size = init->ring_size;
- dev_priv->ring.tail_mask = dev_priv->ring.Size - 1;
+ if (init->ring_size != 0) {
+ if (dev_priv->ring.ring_obj != NULL) {
+ i915_dma_cleanup(dev);
+ DRM_ERROR("Client tried to initialize ringbuffer in "
+ "GEM mode\n");
+ return -EINVAL;
+ }

- dev_priv->ring.map.offset = init->ring_start;
- dev_priv->ring.map.size = init->ring_size;
- dev_priv->ring.map.type = 0;
- dev_priv->ring.map.flags = 0;
- dev_priv->ring.map.mtrr = 0;
+ dev_priv->ring.Size = init->ring_size;
+ dev_priv->ring.tail_mask = dev_priv->ring.Size - 1;

- drm_core_ioremap(&dev_priv->ring.map, dev);
+ dev_priv->ring.map.offset = init->ring_start;
+ dev_priv->ring.map.size = init->ring_size;
+ dev_priv->ring.map.type = 0;
+ dev_priv->ring.map.flags = 0;
+ dev_priv->ring.map.mtrr = 0;

- if (dev_priv->ring.map.handle == NULL) {
- i915_dma_cleanup(dev);
- DRM_ERROR("can not ioremap virtual address for"
- " ring buffer\n");
- return -ENOMEM;
+ drm_core_ioremap(&dev_priv->ring.map, dev);
+
+ if (dev_priv->ring.map.handle == NULL) {
+ i915_dma_cleanup(dev);
+ DRM_ERROR("can not ioremap virtual address for"
+ " ring buffer\n");
+ return -ENOMEM;
+ }
}

dev_priv->ring.virtual_start = dev_priv->ring.map.handle;
@@ -377,9 +384,10 @@ static int i915_emit_cmds(struct drm_device * dev, int __user * buffer, int dwor
return 0;
}

-static int i915_emit_box(struct drm_device * dev,
- struct drm_clip_rect __user * boxes,
- int i, int DR1, int DR4)
+int
+i915_emit_box(struct drm_device *dev,
+ struct drm_clip_rect __user *boxes,
+ int i, int DR1, int DR4)
{
drm_i915_private_t *dev_priv = dev->dev_private;
struct drm_clip_rect box;
@@ -681,6 +689,9 @@ static int i915_getparam(struct drm_device *dev, void *data,
case I915_PARAM_LAST_DISPATCH:
value = READ_BREADCRUMB(dev_priv);
break;
+ case I915_PARAM_HAS_GEM:
+ value = 1;
+ break;
default:
DRM_ERROR("Unknown parameter %d\n", param->param);
return -EINVAL;
@@ -784,6 +795,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
memset(dev_priv, 0, sizeof(drm_i915_private_t));

dev->dev_private = (void *)dev_priv;
+ dev_priv->dev = dev;

/* Add register map (needed for suspend/resume) */
base = drm_get_resource_start(dev, mmio_bar);
@@ -793,6 +805,8 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
_DRM_KERNEL | _DRM_DRIVER,
&dev_priv->mmio_map);

+ i915_gem_load(dev);
+
/* Init HWS */
if (!I915_NEED_GFX_HWS(dev)) {
ret = i915_init_phys_hws(dev);
@@ -833,6 +847,25 @@ int i915_driver_unload(struct drm_device *dev)
return 0;
}

+int i915_driver_open(struct drm_device *dev, struct drm_file *file_priv)
+{
+ struct drm_i915_file_private *i915_file_priv;
+
+ DRM_DEBUG("\n");
+ i915_file_priv = (struct drm_i915_file_private *)
+ drm_alloc(sizeof(*i915_file_priv), DRM_MEM_FILES);
+
+ if (!i915_file_priv)
+ return -ENOMEM;
+
+ file_priv->driver_priv = i915_file_priv;
+
+ i915_file_priv->mm.last_gem_seqno = 0;
+ i915_file_priv->mm.last_gem_throttle_seqno = 0;
+
+ return 0;
+}
+
void i915_driver_lastclose(struct drm_device * dev)
{
drm_i915_private_t *dev_priv = dev->dev_private;
@@ -840,6 +873,8 @@ void i915_driver_lastclose(struct drm_device * dev)
if (!dev_priv)
return;

+ i915_gem_lastclose(dev);
+
if (dev_priv->agp_heap)
i915_mem_takedown(&(dev_priv->agp_heap));

@@ -852,6 +887,13 @@ void i915_driver_preclose(struct drm_device * dev, struct drm_file *file_priv)
i915_mem_release(dev, file_priv, dev_priv->agp_heap);
}

+void i915_driver_postclose(struct drm_device *dev, struct drm_file *file_priv)
+{
+ struct drm_i915_file_private *i915_file_priv = file_priv->driver_priv;
+
+ drm_free(i915_file_priv, sizeof(*i915_file_priv), DRM_MEM_FILES);
+}
+
struct drm_ioctl_desc i915_ioctls[] = {
DRM_IOCTL_DEF(DRM_I915_INIT, i915_dma_init, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
DRM_IOCTL_DEF(DRM_I915_FLUSH, i915_flush_ioctl, DRM_AUTH),
@@ -870,6 +912,22 @@ struct drm_ioctl_desc i915_ioctls[] = {
DRM_IOCTL_DEF(DRM_I915_GET_VBLANK_PIPE, i915_vblank_pipe_get, DRM_AUTH ),
DRM_IOCTL_DEF(DRM_I915_VBLANK_SWAP, i915_vblank_swap, DRM_AUTH),
DRM_IOCTL_DEF(DRM_I915_HWS_ADDR, i915_set_status_page, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_I915_GEM_INIT, i915_gem_init_ioctl, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_I915_GEM_EXECBUFFER, i915_gem_execbuffer, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_I915_GEM_PIN, i915_gem_pin_ioctl, DRM_AUTH|DRM_ROOT_ONLY),
+ DRM_IOCTL_DEF(DRM_I915_GEM_UNPIN, i915_gem_unpin_ioctl, DRM_AUTH|DRM_ROOT_ONLY),
+ DRM_IOCTL_DEF(DRM_I915_GEM_BUSY, i915_gem_busy_ioctl, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_I915_GEM_THROTTLE, i915_gem_throttle_ioctl, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_I915_GEM_ENTERVT, i915_gem_entervt_ioctl, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_I915_GEM_LEAVEVT, i915_gem_leavevt_ioctl, DRM_AUTH),
+ DRM_IOCTL_DEF(DRM_I915_GEM_CREATE, i915_gem_create_ioctl, 0),
+ DRM_IOCTL_DEF(DRM_I915_GEM_PREAD, i915_gem_pread_ioctl, 0),
+ DRM_IOCTL_DEF(DRM_I915_GEM_PWRITE, i915_gem_pwrite_ioctl, 0),
+ DRM_IOCTL_DEF(DRM_I915_GEM_MMAP, i915_gem_mmap_ioctl, 0),
+ DRM_IOCTL_DEF(DRM_I915_GEM_SET_DOMAIN, i915_gem_set_domain_ioctl, 0),
+ DRM_IOCTL_DEF(DRM_I915_GEM_SW_FINISH, i915_gem_sw_finish_ioctl, 0),
+ DRM_IOCTL_DEF(DRM_I915_GEM_SET_TILING, i915_gem_set_tiling, 0),
+ DRM_IOCTL_DEF(DRM_I915_GEM_GET_TILING, i915_gem_get_tiling, 0),
};

int i915_max_ioctl = DRM_ARRAY_SIZE(i915_ioctls);
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 6c99aab..ea4aaa0 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -542,11 +542,13 @@ static struct drm_driver driver = {
.driver_features =
DRIVER_USE_AGP | DRIVER_REQUIRE_AGP | /* DRIVER_USE_MTRR |*/
DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | DRIVER_IRQ_VBL |
- DRIVER_IRQ_VBL2,
+ DRIVER_IRQ_VBL2 | DRIVER_GEM,
.load = i915_driver_load,
.unload = i915_driver_unload,
+ .open = i915_driver_open,
.lastclose = i915_driver_lastclose,
.preclose = i915_driver_preclose,
+ .postclose = i915_driver_postclose,
.suspend = i915_suspend,
.resume = i915_resume,
.device_is_agp = i915_driver_device_is_agp,
@@ -559,6 +561,10 @@ static struct drm_driver driver = {
.reclaim_buffers = drm_core_reclaim_buffers,
.get_map_ofs = drm_core_get_map_ofs,
.get_reg_ofs = drm_core_get_reg_ofs,
+ .proc_init = i915_gem_proc_init,
+ .proc_cleanup = i915_gem_proc_cleanup,
+ .gem_init_object = i915_gem_init_object,
+ .gem_free_object = i915_gem_free_object,
.ioctls = i915_ioctls,
.fops = {
.owner = THIS_MODULE,
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 8daf0d8..35b2a4c 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -39,7 +39,7 @@

#define DRIVER_NAME "i915"
#define DRIVER_DESC "Intel Graphics"
-#define DRIVER_DATE "20060119"
+#define DRIVER_DATE "20080730"

/* Interface history:
*
@@ -55,16 +55,23 @@
#define DRIVER_MINOR 6
#define DRIVER_PATCHLEVEL 0

+#define WATCH_COHERENCY 0
+#define WATCH_BUF 0
+#define WATCH_EXEC 0
+#define WATCH_LRU 0
+#define WATCH_RELOC 0
+#define WATCH_INACTIVE 0
+#define WATCH_PWRITE 0
+
typedef struct _drm_i915_ring_buffer {
int tail_mask;
- unsigned long Start;
- unsigned long End;
unsigned long Size;
u8 *virtual_start;
int head;
int tail;
int space;
drm_local_map_t map;
+ struct drm_gem_object *ring_obj;
} drm_i915_ring_buffer_t;

struct mem_block {
@@ -83,6 +90,8 @@ typedef struct _drm_i915_vbl_swap {
} drm_i915_vbl_swap_t;

typedef struct drm_i915_private {
+ struct drm_device *dev;
+
drm_local_map_t *sarea;
drm_local_map_t *mmio_map;

@@ -95,6 +104,7 @@ typedef struct drm_i915_private {
unsigned long counter;
unsigned int status_gfx_addr;
drm_local_map_t hws_map;
+ struct drm_gem_object *hws_obj;

unsigned int cpp;
int back_offset;
@@ -104,7 +114,6 @@ typedef struct drm_i915_private {

wait_queue_head_t irq_queue;
atomic_t irq_received;
- atomic_t irq_emitted;
/** Protects user_irq_refcount and irq_mask_reg */
spinlock_t user_irq_lock;
/** Refcount for i915_user_irq_get() versus i915_user_irq_put(). */
@@ -210,8 +219,174 @@ typedef struct drm_i915_private {
u8 saveDACMASK;
u8 saveDACDATA[256*3]; /* 256 3-byte colors */
u8 saveCR[37];
+
+ struct {
+ struct drm_mm gtt_space;
+
+ /**
+ * List of objects currently involved in rendering from the
+ * ringbuffer.
+ *
+ * A reference is held on the buffer while on this list.
+ */
+ struct list_head active_list;
+
+ /**
+ * List of objects which are not in the ringbuffer but which
+ * still have a write_domain which needs to be flushed before
+ * unbinding.
+ *
+ * A reference is held on the buffer while on this list.
+ */
+ struct list_head flushing_list;
+
+ /**
+ * LRU list of objects which are not in the ringbuffer and
+ * are ready to unbind, but are still in the GTT.
+ *
+ * A reference is not held on the buffer while on this list,
+ * as merely being GTT-bound shouldn't prevent its being
+ * freed, and we'll pull it off the list in the free path.
+ */
+ struct list_head inactive_list;
+
+ /**
+ * List of breadcrumbs associated with GPU requests currently
+ * outstanding.
+ */
+ struct list_head request_list;
+
+ /**
+ * We leave the user IRQ off as much as possible,
+ * but this means that requests will finish and never
+ * be retired once the system goes idle. Set a timer to
+ * fire periodically while the ring is running. When it
+ * fires, go retire requests.
+ */
+ struct delayed_work retire_work;
+
+ uint32_t next_gem_seqno;
+
+ /**
+ * Waiting sequence number, if any
+ */
+ uint32_t waiting_gem_seqno;
+
+ /**
+ * Last seq seen at irq time
+ */
+ uint32_t irq_gem_seqno;
+
+ /**
+ * Flag if the X Server, and thus DRM, is not currently in
+ * control of the device.
+ *
+ * This is set between LeaveVT and EnterVT. It needs to be
+ * replaced with a semaphore. It also needs to be
+ * transitioned away from for kernel modesetting.
+ */
+ int suspended;
+
+ /**
+ * Flag if the hardware appears to be wedged.
+ *
+ * This is set when attempts to idle the device timeout.
+ * It prevents command submission from occuring and makes
+ * every pending request fail
+ */
+ int wedged;
+
+ /** Bit 6 swizzling required for X tiling */
+ uint32_t bit_6_swizzle_x;
+ /** Bit 6 swizzling required for Y tiling */
+ uint32_t bit_6_swizzle_y;
+ } mm;
} drm_i915_private_t;

+/** driver private structure attached to each drm_gem_object */
+struct drm_i915_gem_object {
+ struct drm_gem_object *obj;
+
+ /** Current space allocated to this object in the GTT, if any. */
+ struct drm_mm_node *gtt_space;
+
+ /** This object's place on the active/flushing/inactive lists */
+ struct list_head list;
+
+ /**
+ * This is set if the object is on the active or flushing lists
+ * (has pending rendering), and is not set if it's on inactive (ready
+ * to be unbound).
+ */
+ int active;
+
+ /**
+ * This is set if the object has been written to since last bound
+ * to the GTT
+ */
+ int dirty;
+
+ /** AGP memory structure for our GTT binding. */
+ DRM_AGP_MEM *agp_mem;
+
+ struct page **page_list;
+
+ /**
+ * Current offset of the object in GTT space.
+ *
+ * This is the same as gtt_space->start
+ */
+ uint32_t gtt_offset;
+
+ /** Boolean whether this object has a valid gtt offset. */
+ int gtt_bound;
+
+ /** How many users have pinned this object in GTT space */
+ int pin_count;
+
+ /** Breadcrumb of last rendering to the buffer. */
+ uint32_t last_rendering_seqno;
+
+ /** Current tiling mode for the object. */
+ uint32_t tiling_mode;
+
+ /**
+ * Flagging of which individual pages are valid in GEM_DOMAIN_CPU when
+ * GEM_DOMAIN_CPU is not in the object's read domain.
+ */
+ uint8_t *page_cpu_valid;
+};
+
+/**
+ * Request queue structure.
+ *
+ * The request queue allows us to note sequence numbers that have been emitted
+ * and may be associated with active buffers to be retired.
+ *
+ * By keeping this list, we can avoid having to do questionable
+ * sequence-number comparisons on buffer last_rendering_seqnos, and associate
+ * an emission time with seqnos for tracking how far ahead of the GPU we are.
+ */
+struct drm_i915_gem_request {
+ /** GEM sequence number associated with this request. */
+ uint32_t seqno;
+
+ /** Time at which this request was emitted, in jiffies. */
+ unsigned long emitted_jiffies;
+
+ /** Cache domains that were flushed at the start of the request. */
+ uint32_t flush_domains;
+
+ struct list_head list;
+};
+
+struct drm_i915_file_private {
+ struct {
+ uint32_t last_gem_seqno;
+ uint32_t last_gem_throttle_seqno;
+ } mm;
+};
+
extern struct drm_ioctl_desc i915_ioctls[];
extern int i915_max_ioctl;

@@ -219,18 +394,26 @@ extern int i915_max_ioctl;
extern void i915_kernel_lost_context(struct drm_device * dev);
extern int i915_driver_load(struct drm_device *, unsigned long flags);
extern int i915_driver_unload(struct drm_device *);
+extern int i915_driver_open(struct drm_device *dev, struct drm_file *file_priv);
extern void i915_driver_lastclose(struct drm_device * dev);
extern void i915_driver_preclose(struct drm_device *dev,
struct drm_file *file_priv);
+extern void i915_driver_postclose(struct drm_device *dev,
+ struct drm_file *file_priv);
extern int i915_driver_device_is_agp(struct drm_device * dev);
extern long i915_compat_ioctl(struct file *filp, unsigned int cmd,
unsigned long arg);
+extern int i915_emit_box(struct drm_device *dev,
+ struct drm_clip_rect __user *boxes,
+ int i, int DR1, int DR4);

/* i915_irq.c */
extern int i915_irq_emit(struct drm_device *dev, void *data,
struct drm_file *file_priv);
extern int i915_irq_wait(struct drm_device *dev, void *data,
struct drm_file *file_priv);
+void i915_user_irq_get(struct drm_device *dev);
+void i915_user_irq_put(struct drm_device *dev);

extern int i915_driver_vblank_wait(struct drm_device *dev, unsigned int *sequence);
extern int i915_driver_vblank_wait2(struct drm_device *dev, unsigned int *sequence);
@@ -257,6 +440,67 @@ extern int i915_mem_destroy_heap(struct drm_device *dev, void *data,
extern void i915_mem_takedown(struct mem_block **heap);
extern void i915_mem_release(struct drm_device * dev,
struct drm_file *file_priv, struct mem_block *heap);
+/* i915_gem.c */
+int i915_gem_init_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_create_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_pread_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_mmap_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_execbuffer(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_pin_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_unpin_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_busy_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_throttle_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_entervt_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_leavevt_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_set_tiling(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int i915_gem_get_tiling(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+void i915_gem_load(struct drm_device *dev);
+int i915_gem_proc_init(struct drm_minor *minor);
+void i915_gem_proc_cleanup(struct drm_minor *minor);
+int i915_gem_init_object(struct drm_gem_object *obj);
+void i915_gem_free_object(struct drm_gem_object *obj);
+int i915_gem_object_pin(struct drm_gem_object *obj, uint32_t alignment);
+void i915_gem_object_unpin(struct drm_gem_object *obj);
+void i915_gem_lastclose(struct drm_device *dev);
+uint32_t i915_get_gem_seqno(struct drm_device *dev);
+void i915_gem_retire_requests(struct drm_device *dev);
+void i915_gem_retire_work_handler(struct work_struct *work);
+void i915_gem_clflush_object(struct drm_gem_object *obj);
+
+/* i915_gem_tiling.c */
+void i915_gem_detect_bit_6_swizzle(struct drm_device *dev);
+
+/* i915_gem_debug.c */
+void i915_gem_dump_object(struct drm_gem_object *obj, int len,
+ const char *where, uint32_t mark);
+#if WATCH_INACTIVE
+void i915_verify_inactive(struct drm_device *dev, char *file, int line);
+#else
+#define i915_verify_inactive(dev, file, line)
+#endif
+void i915_gem_object_check_coherency(struct drm_gem_object *obj, int handle);
+void i915_gem_dump_object(struct drm_gem_object *obj, int len,
+ const char *where, uint32_t mark);
+void i915_dump_lru(struct drm_device *dev, const char *where);

#define I915_READ(reg) DRM_READ32(dev_priv->mmio_map, (reg))
#define I915_WRITE(reg,val) DRM_WRITE32(dev_priv->mmio_map, (reg), (val))
@@ -309,6 +553,7 @@ extern void i915_mem_release(struct drm_device * dev,
*/
#define READ_HWSP(dev_priv, reg) (((volatile u32*)(dev_priv->hw_status_page))[reg])
#define READ_BREADCRUMB(dev_priv) READ_HWSP(dev_priv, 5)
+#define I915_GEM_HWS_INDEX 0x10

extern int i915_wait_ring(struct drm_device * dev, int n, const char *caller);

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
new file mode 100644
index 0000000..7dc6209
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -0,0 +1,2497 @@
+/*
+ * Copyright © 2008 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.
+ *
+ * Authors:
+ * Eric Anholt <[email protected]>
+ *
+ */
+
+#include "drmP.h"
+#include "drm.h"
+#include "i915_drm.h"
+#include "i915_drv.h"
+#include <linux/swap.h>
+
+static int
+i915_gem_object_set_domain(struct drm_gem_object *obj,
+ uint32_t read_domains,
+ uint32_t write_domain);
+static int
+i915_gem_object_set_domain_range(struct drm_gem_object *obj,
+ uint64_t offset,
+ uint64_t size,
+ uint32_t read_domains,
+ uint32_t write_domain);
+static int
+i915_gem_set_domain(struct drm_gem_object *obj,
+ struct drm_file *file_priv,
+ uint32_t read_domains,
+ uint32_t write_domain);
+static int i915_gem_object_get_page_list(struct drm_gem_object *obj);
+static void i915_gem_object_free_page_list(struct drm_gem_object *obj);
+static int i915_gem_object_wait_rendering(struct drm_gem_object *obj);
+
+int
+i915_gem_init_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_init *args = data;
+
+ mutex_lock(&dev->struct_mutex);
+
+ if (args->gtt_start >= args->gtt_end ||
+ (args->gtt_start & (PAGE_SIZE - 1)) != 0 ||
+ (args->gtt_end & (PAGE_SIZE - 1)) != 0) {
+ mutex_unlock(&dev->struct_mutex);
+ return -EINVAL;
+ }
+
+ drm_mm_init(&dev_priv->mm.gtt_space, args->gtt_start,
+ args->gtt_end - args->gtt_start);
+
+ dev->gtt_total = (uint32_t) (args->gtt_end - args->gtt_start);
+
+ mutex_unlock(&dev->struct_mutex);
+
+ return 0;
+}
+
+
+/**
+ * Creates a new mm object and returns a handle to it.
+ */
+int
+i915_gem_create_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_create *args = data;
+ struct drm_gem_object *obj;
+ int handle, ret;
+
+ args->size = roundup(args->size, PAGE_SIZE);
+
+ /* Allocate the new object */
+ obj = drm_gem_object_alloc(dev, args->size);
+ if (obj == NULL)
+ return -ENOMEM;
+
+ ret = drm_gem_handle_create(file_priv, obj, &handle);
+ mutex_lock(&dev->struct_mutex);
+ drm_gem_object_handle_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+
+ if (ret)
+ return ret;
+
+ args->handle = handle;
+
+ return 0;
+}
+
+/**
+ * Reads data from the object referenced by handle.
+ *
+ * On error, the contents of *data are undefined.
+ */
+int
+i915_gem_pread_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_pread *args = data;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ ssize_t read;
+ loff_t offset;
+ int ret;
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL)
+ return -EBADF;
+ obj_priv = obj->driver_private;
+
+ /* Bounds check source.
+ *
+ * XXX: This could use review for overflow issues...
+ */
+ if (args->offset > obj->size || args->size > obj->size ||
+ args->offset + args->size > obj->size) {
+ drm_gem_object_unreference(obj);
+ return -EINVAL;
+ }
+
+ mutex_lock(&dev->struct_mutex);
+
+ ret = i915_gem_object_set_domain_range(obj, args->offset, args->size,
+ I915_GEM_DOMAIN_CPU, 0);
+ if (ret != 0) {
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ }
+
+ offset = args->offset;
+
+ read = vfs_read(obj->filp, (char __user *)(uintptr_t)args->data_ptr,
+ args->size, &offset);
+ if (read != args->size) {
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ if (read < 0)
+ return read;
+ else
+ return -EINVAL;
+ }
+
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+
+ return 0;
+}
+
+static int
+i915_gem_gtt_pwrite(struct drm_device *dev, struct drm_gem_object *obj,
+ struct drm_i915_gem_pwrite *args,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ ssize_t remain;
+ loff_t offset;
+ char __user *user_data;
+ char *vaddr;
+ int i, o, l;
+ int ret = 0;
+ unsigned long pfn;
+ unsigned long unwritten;
+
+ user_data = (char __user *) (uintptr_t) args->data_ptr;
+ remain = args->size;
+ if (!access_ok(VERIFY_READ, user_data, remain))
+ return -EFAULT;
+
+
+ mutex_lock(&dev->struct_mutex);
+ ret = i915_gem_object_pin(obj, 0);
+ if (ret) {
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+ }
+ ret = i915_gem_set_domain(obj, file_priv,
+ I915_GEM_DOMAIN_GTT, I915_GEM_DOMAIN_GTT);
+ if (ret)
+ goto fail;
+
+ obj_priv = obj->driver_private;
+ offset = obj_priv->gtt_offset + args->offset;
+ obj_priv->dirty = 1;
+
+ while (remain > 0) {
+ /* Operation in this page
+ *
+ * i = page number
+ * o = offset within page
+ * l = bytes to copy
+ */
+ i = offset >> PAGE_SHIFT;
+ o = offset & (PAGE_SIZE-1);
+ l = remain;
+ if ((o + l) > PAGE_SIZE)
+ l = PAGE_SIZE - o;
+
+ pfn = (dev->agp->base >> PAGE_SHIFT) + i;
+
+#ifdef DRM_KMAP_ATOMIC_PROT_PFN
+ /* kmap_atomic can't map IO pages on non-HIGHMEM kernels
+ */
+ vaddr = kmap_atomic_prot_pfn(pfn, KM_USER0,
+ __pgprot(__PAGE_KERNEL));
+#if WATCH_PWRITE
+ DRM_INFO("pwrite i %d o %d l %d pfn %ld vaddr %p\n",
+ i, o, l, pfn, vaddr);
+#endif
+ unwritten = __copy_from_user_inatomic_nocache(vaddr + o,
+ user_data, l);
+ kunmap_atomic(vaddr, KM_USER0);
+
+ if (unwritten)
+#endif
+ {
+ vaddr = ioremap(pfn << PAGE_SHIFT, PAGE_SIZE);
+#if WATCH_PWRITE
+ DRM_INFO("pwrite slow i %d o %d l %d "
+ "pfn %ld vaddr %p\n",
+ i, o, l, pfn, vaddr);
+#endif
+ if (vaddr == NULL) {
+ ret = -EFAULT;
+ goto fail;
+ }
+ unwritten = __copy_from_user(vaddr + o, user_data, l);
+#if WATCH_PWRITE
+ DRM_INFO("unwritten %ld\n", unwritten);
+#endif
+ iounmap(vaddr);
+ if (unwritten) {
+ ret = -EFAULT;
+ goto fail;
+ }
+ }
+
+ remain -= l;
+ user_data += l;
+ offset += l;
+ }
+#if WATCH_PWRITE && 1
+ i915_gem_clflush_object(obj);
+ i915_gem_dump_object(obj, args->offset + args->size, __func__, ~0);
+ i915_gem_clflush_object(obj);
+#endif
+
+fail:
+ i915_gem_object_unpin(obj);
+ mutex_unlock(&dev->struct_mutex);
+
+ return ret;
+}
+
+int
+i915_gem_shmem_pwrite(struct drm_device *dev, struct drm_gem_object *obj,
+ struct drm_i915_gem_pwrite *args,
+ struct drm_file *file_priv)
+{
+ int ret;
+ loff_t offset;
+ ssize_t written;
+
+ mutex_lock(&dev->struct_mutex);
+
+ ret = i915_gem_set_domain(obj, file_priv,
+ I915_GEM_DOMAIN_CPU, I915_GEM_DOMAIN_CPU);
+ if (ret) {
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+ }
+
+ offset = args->offset;
+
+ written = vfs_write(obj->filp,
+ (char __user *)(uintptr_t) args->data_ptr,
+ args->size, &offset);
+ if (written != args->size) {
+ mutex_unlock(&dev->struct_mutex);
+ if (written < 0)
+ return written;
+ else
+ return -EINVAL;
+ }
+
+ mutex_unlock(&dev->struct_mutex);
+
+ return 0;
+}
+
+/**
+ * Writes data to the object referenced by handle.
+ *
+ * On error, the contents of the buffer that were to be modified are undefined.
+ */
+int
+i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_pwrite *args = data;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ int ret = 0;
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL)
+ return -EBADF;
+ obj_priv = obj->driver_private;
+
+ /* Bounds check destination.
+ *
+ * XXX: This could use review for overflow issues...
+ */
+ if (args->offset > obj->size || args->size > obj->size ||
+ args->offset + args->size > obj->size) {
+ drm_gem_object_unreference(obj);
+ return -EINVAL;
+ }
+
+ /* We can only do the GTT pwrite on untiled buffers, as otherwise
+ * it would end up going through the fenced access, and we'll get
+ * different detiling behavior between reading and writing.
+ * pread/pwrite currently are reading and writing from the CPU
+ * perspective, requiring manual detiling by the client.
+ */
+ if (obj_priv->tiling_mode == I915_TILING_NONE &&
+ dev->gtt_total != 0)
+ ret = i915_gem_gtt_pwrite(dev, obj, args, file_priv);
+ else
+ ret = i915_gem_shmem_pwrite(dev, obj, args, file_priv);
+
+#if WATCH_PWRITE
+ if (ret)
+ DRM_INFO("pwrite failed %d\n", ret);
+#endif
+
+ drm_gem_object_unreference(obj);
+
+ return ret;
+}
+
+/**
+ * Called when user space prepares to use an object
+ */
+int
+i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_set_domain *args = data;
+ struct drm_gem_object *obj;
+ int ret;
+
+ if (!(dev->driver->driver_features & DRIVER_GEM))
+ return -ENODEV;
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL)
+ return -EBADF;
+
+ mutex_lock(&dev->struct_mutex);
+ ret = i915_gem_set_domain(obj, file_priv,
+ args->read_domains, args->write_domain);
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+}
+
+/**
+ * Called when user space has done writes to this buffer
+ */
+int
+i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_sw_finish *args = data;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ int ret = 0;
+
+ if (!(dev->driver->driver_features & DRIVER_GEM))
+ return -ENODEV;
+
+ mutex_lock(&dev->struct_mutex);
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL) {
+ mutex_unlock(&dev->struct_mutex);
+ return -EBADF;
+ }
+
+#if WATCH_BUF
+ DRM_INFO("%s: sw_finish %d (%p)\n",
+ __func__, args->handle, obj);
+#endif
+ obj_priv = obj->driver_private;
+
+ /* Pinned buffers may be scanout, so flush the cache */
+ if ((obj->write_domain & I915_GEM_DOMAIN_CPU) && obj_priv->pin_count) {
+ i915_gem_clflush_object(obj);
+ drm_agp_chipset_flush(dev);
+ }
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+}
+
+/**
+ * Maps the contents of an object, returning the address it is mapped
+ * into.
+ *
+ * While the mapping holds a reference on the contents of the object, it doesn't
+ * imply a ref on the object itself.
+ */
+int
+i915_gem_mmap_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_mmap *args = data;
+ struct drm_gem_object *obj;
+ loff_t offset;
+ unsigned long addr;
+
+ if (!(dev->driver->driver_features & DRIVER_GEM))
+ return -ENODEV;
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL)
+ return -EBADF;
+
+ offset = args->offset;
+
+ down_write(&current->mm->mmap_sem);
+ addr = do_mmap(obj->filp, 0, args->size,
+ PROT_READ | PROT_WRITE, MAP_SHARED,
+ args->offset);
+ up_write(&current->mm->mmap_sem);
+ mutex_lock(&dev->struct_mutex);
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ if (IS_ERR((void *)addr))
+ return addr;
+
+ args->addr_ptr = (uint64_t) addr;
+
+ return 0;
+}
+
+static void
+i915_gem_object_free_page_list(struct drm_gem_object *obj)
+{
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int page_count = obj->size / PAGE_SIZE;
+ int i;
+
+ if (obj_priv->page_list == NULL)
+ return;
+
+
+ for (i = 0; i < page_count; i++)
+ if (obj_priv->page_list[i] != NULL) {
+ if (obj_priv->dirty)
+ set_page_dirty(obj_priv->page_list[i]);
+ mark_page_accessed(obj_priv->page_list[i]);
+ page_cache_release(obj_priv->page_list[i]);
+ }
+ obj_priv->dirty = 0;
+
+ drm_free(obj_priv->page_list,
+ page_count * sizeof(struct page *),
+ DRM_MEM_DRIVER);
+ obj_priv->page_list = NULL;
+}
+
+static void
+i915_gem_object_move_to_active(struct drm_gem_object *obj)
+{
+ struct drm_device *dev = obj->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+
+ /* Add a reference if we're newly entering the active list. */
+ if (!obj_priv->active) {
+ drm_gem_object_reference(obj);
+ obj_priv->active = 1;
+ }
+ /* Move from whatever list we were on to the tail of execution. */
+ list_move_tail(&obj_priv->list,
+ &dev_priv->mm.active_list);
+}
+
+
+static void
+i915_gem_object_move_to_inactive(struct drm_gem_object *obj)
+{
+ struct drm_device *dev = obj->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+ if (obj_priv->pin_count != 0)
+ list_del_init(&obj_priv->list);
+ else
+ list_move_tail(&obj_priv->list, &dev_priv->mm.inactive_list);
+
+ if (obj_priv->active) {
+ obj_priv->active = 0;
+ drm_gem_object_unreference(obj);
+ }
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+}
+
+/**
+ * Creates a new sequence number, emitting a write of it to the status page
+ * plus an interrupt, which will trigger i915_user_interrupt_handler.
+ *
+ * Must be called with struct_lock held.
+ *
+ * Returned sequence numbers are nonzero on success.
+ */
+static uint32_t
+i915_add_request(struct drm_device *dev, uint32_t flush_domains)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_request *request;
+ uint32_t seqno;
+ int was_empty;
+ RING_LOCALS;
+
+ request = drm_calloc(1, sizeof(*request), DRM_MEM_DRIVER);
+ if (request == NULL)
+ return 0;
+
+ /* Grab the seqno we're going to make this request be, and bump the
+ * next (skipping 0 so it can be the reserved no-seqno value).
+ */
+ seqno = dev_priv->mm.next_gem_seqno;
+ dev_priv->mm.next_gem_seqno++;
+ if (dev_priv->mm.next_gem_seqno == 0)
+ dev_priv->mm.next_gem_seqno++;
+
+ BEGIN_LP_RING(4);
+ OUT_RING(MI_STORE_DWORD_INDEX);
+ OUT_RING(I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
+ OUT_RING(seqno);
+
+ OUT_RING(MI_USER_INTERRUPT);
+ ADVANCE_LP_RING();
+
+ DRM_DEBUG("%d\n", seqno);
+
+ request->seqno = seqno;
+ request->emitted_jiffies = jiffies;
+ request->flush_domains = flush_domains;
+ was_empty = list_empty(&dev_priv->mm.request_list);
+ list_add_tail(&request->list, &dev_priv->mm.request_list);
+
+ if (was_empty)
+ schedule_delayed_work(&dev_priv->mm.retire_work, HZ);
+ return seqno;
+}
+
+/**
+ * Command execution barrier
+ *
+ * Ensures that all commands in the ring are finished
+ * before signalling the CPU
+ */
+uint32_t
+i915_retire_commands(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ uint32_t cmd = MI_FLUSH | MI_NO_WRITE_FLUSH;
+ uint32_t flush_domains = 0;
+ RING_LOCALS;
+
+ /* The sampler always gets flushed on i965 (sigh) */
+ if (IS_I965G(dev))
+ flush_domains |= I915_GEM_DOMAIN_SAMPLER;
+ BEGIN_LP_RING(2);
+ OUT_RING(cmd);
+ OUT_RING(0); /* noop */
+ ADVANCE_LP_RING();
+ return flush_domains;
+}
+
+/**
+ * Moves buffers associated only with the given active seqno from the active
+ * to inactive list, potentially freeing them.
+ */
+static void
+i915_gem_retire_request(struct drm_device *dev,
+ struct drm_i915_gem_request *request)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+
+ if (request->flush_domains != 0) {
+ struct drm_i915_gem_object *obj_priv, *next;
+
+ /* First clear any buffers that were only waiting for a flush
+ * matching the one just retired.
+ */
+
+ list_for_each_entry_safe(obj_priv, next,
+ &dev_priv->mm.flushing_list, list) {
+ struct drm_gem_object *obj = obj_priv->obj;
+
+ if (obj->write_domain & request->flush_domains) {
+ obj->write_domain = 0;
+ i915_gem_object_move_to_inactive(obj);
+ }
+ }
+
+ }
+
+ /* Move any buffers on the active list that are no longer referenced
+ * by the ringbuffer to the flushing/inactive lists as appropriate.
+ */
+ while (!list_empty(&dev_priv->mm.active_list)) {
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+
+ obj_priv = list_first_entry(&dev_priv->mm.active_list,
+ struct drm_i915_gem_object,
+ list);
+ obj = obj_priv->obj;
+
+ /* If the seqno being retired doesn't match the oldest in the
+ * list, then the oldest in the list must still be newer than
+ * this seqno.
+ */
+ if (obj_priv->last_rendering_seqno != request->seqno)
+ return;
+#if WATCH_LRU
+ DRM_INFO("%s: retire %d moves to inactive list %p\n",
+ __func__, request->seqno, obj);
+#endif
+
+ if (obj->write_domain != 0) {
+ list_move_tail(&obj_priv->list,
+ &dev_priv->mm.flushing_list);
+ } else {
+ i915_gem_object_move_to_inactive(obj);
+ }
+ }
+}
+
+/**
+ * Returns true if seq1 is later than seq2.
+ */
+static int
+i915_seqno_passed(uint32_t seq1, uint32_t seq2)
+{
+ return (int32_t)(seq1 - seq2) >= 0;
+}
+
+uint32_t
+i915_get_gem_seqno(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+
+ return READ_HWSP(dev_priv, I915_GEM_HWS_INDEX);
+}
+
+/**
+ * This function clears the request list as sequence numbers are passed.
+ */
+void
+i915_gem_retire_requests(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ uint32_t seqno;
+
+ seqno = i915_get_gem_seqno(dev);
+
+ while (!list_empty(&dev_priv->mm.request_list)) {
+ struct drm_i915_gem_request *request;
+ uint32_t retiring_seqno;
+
+ request = list_first_entry(&dev_priv->mm.request_list,
+ struct drm_i915_gem_request,
+ list);
+ retiring_seqno = request->seqno;
+
+ if (i915_seqno_passed(seqno, retiring_seqno) ||
+ dev_priv->mm.wedged) {
+ i915_gem_retire_request(dev, request);
+
+ list_del(&request->list);
+ drm_free(request, sizeof(*request), DRM_MEM_DRIVER);
+ } else
+ break;
+ }
+}
+
+void
+i915_gem_retire_work_handler(struct work_struct *work)
+{
+ drm_i915_private_t *dev_priv;
+ struct drm_device *dev;
+
+ dev_priv = container_of(work, drm_i915_private_t,
+ mm.retire_work.work);
+ dev = dev_priv->dev;
+
+ mutex_lock(&dev->struct_mutex);
+ i915_gem_retire_requests(dev);
+ if (!list_empty(&dev_priv->mm.request_list))
+ schedule_delayed_work(&dev_priv->mm.retire_work, HZ);
+ mutex_unlock(&dev->struct_mutex);
+}
+
+/**
+ * Waits for a sequence number to be signaled, and cleans up the
+ * request and object lists appropriately for that event.
+ */
+int
+i915_wait_request(struct drm_device *dev, uint32_t seqno)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ int ret = 0;
+
+ BUG_ON(seqno == 0);
+
+ if (!i915_seqno_passed(i915_get_gem_seqno(dev), seqno)) {
+ dev_priv->mm.waiting_gem_seqno = seqno;
+ i915_user_irq_get(dev);
+ ret = wait_event_interruptible(dev_priv->irq_queue,
+ i915_seqno_passed(i915_get_gem_seqno(dev),
+ seqno) ||
+ dev_priv->mm.wedged);
+ i915_user_irq_put(dev);
+ dev_priv->mm.waiting_gem_seqno = 0;
+ }
+ if (dev_priv->mm.wedged)
+ ret = -EIO;
+
+ if (ret)
+ DRM_ERROR("%s returns %d (awaiting %d at %d)\n",
+ __func__, ret, seqno, i915_get_gem_seqno(dev));
+
+ /* Directly dispatch request retiring. While we have the work queue
+ * to handle this, the waiter on a request often wants an associated
+ * buffer to have made it to the inactive list, and we would need
+ * a separate wait queue to handle that.
+ */
+ if (ret == 0)
+ i915_gem_retire_requests(dev);
+
+ return ret;
+}
+
+static void
+i915_gem_flush(struct drm_device *dev,
+ uint32_t invalidate_domains,
+ uint32_t flush_domains)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ uint32_t cmd;
+ RING_LOCALS;
+
+#if WATCH_EXEC
+ DRM_INFO("%s: invalidate %08x flush %08x\n", __func__,
+ invalidate_domains, flush_domains);
+#endif
+
+ if (flush_domains & I915_GEM_DOMAIN_CPU)
+ drm_agp_chipset_flush(dev);
+
+ if ((invalidate_domains | flush_domains) & ~(I915_GEM_DOMAIN_CPU |
+ I915_GEM_DOMAIN_GTT)) {
+ /*
+ * read/write caches:
+ *
+ * I915_GEM_DOMAIN_RENDER is always invalidated, but is
+ * only flushed if MI_NO_WRITE_FLUSH is unset. On 965, it is
+ * also flushed at 2d versus 3d pipeline switches.
+ *
+ * read-only caches:
+ *
+ * I915_GEM_DOMAIN_SAMPLER is flushed on pre-965 if
+ * MI_READ_FLUSH is set, and is always flushed on 965.
+ *
+ * I915_GEM_DOMAIN_COMMAND may not exist?
+ *
+ * I915_GEM_DOMAIN_INSTRUCTION, which exists on 965, is
+ * invalidated when MI_EXE_FLUSH is set.
+ *
+ * I915_GEM_DOMAIN_VERTEX, which exists on 965, is
+ * invalidated with every MI_FLUSH.
+ *
+ * TLBs:
+ *
+ * On 965, TLBs associated with I915_GEM_DOMAIN_COMMAND
+ * and I915_GEM_DOMAIN_CPU in are invalidated at PTE write and
+ * I915_GEM_DOMAIN_RENDER and I915_GEM_DOMAIN_SAMPLER
+ * are flushed at any MI_FLUSH.
+ */
+
+ cmd = MI_FLUSH | MI_NO_WRITE_FLUSH;
+ if ((invalidate_domains|flush_domains) &
+ I915_GEM_DOMAIN_RENDER)
+ cmd &= ~MI_NO_WRITE_FLUSH;
+ if (!IS_I965G(dev)) {
+ /*
+ * On the 965, the sampler cache always gets flushed
+ * and this bit is reserved.
+ */
+ if (invalidate_domains & I915_GEM_DOMAIN_SAMPLER)
+ cmd |= MI_READ_FLUSH;
+ }
+ if (invalidate_domains & I915_GEM_DOMAIN_INSTRUCTION)
+ cmd |= MI_EXE_FLUSH;
+
+#if WATCH_EXEC
+ DRM_INFO("%s: queue flush %08x to ring\n", __func__, cmd);
+#endif
+ BEGIN_LP_RING(2);
+ OUT_RING(cmd);
+ OUT_RING(0); /* noop */
+ ADVANCE_LP_RING();
+ }
+}
+
+/**
+ * Ensures that all rendering to the object has completed and the object is
+ * safe to unbind from the GTT or access from the CPU.
+ */
+static int
+i915_gem_object_wait_rendering(struct drm_gem_object *obj)
+{
+ struct drm_device *dev = obj->dev;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int ret;
+
+ /* If there are writes queued to the buffer, flush and
+ * create a new seqno to wait for.
+ */
+ if (obj->write_domain & ~(I915_GEM_DOMAIN_CPU|I915_GEM_DOMAIN_GTT)) {
+ uint32_t write_domain = obj->write_domain;
+#if WATCH_BUF
+ DRM_INFO("%s: flushing object %p from write domain %08x\n",
+ __func__, obj, write_domain);
+#endif
+ i915_gem_flush(dev, 0, write_domain);
+ obj->write_domain = 0;
+
+ i915_gem_object_move_to_active(obj);
+ obj_priv->last_rendering_seqno = i915_add_request(dev,
+ write_domain);
+ BUG_ON(obj_priv->last_rendering_seqno == 0);
+#if WATCH_LRU
+ DRM_INFO("%s: flush moves to exec list %p\n", __func__, obj);
+#endif
+ }
+ /* If there is rendering queued on the buffer being evicted, wait for
+ * it.
+ */
+ if (obj_priv->active) {
+#if WATCH_BUF
+ DRM_INFO("%s: object %p wait for seqno %08x\n",
+ __func__, obj, obj_priv->last_rendering_seqno);
+#endif
+ ret = i915_wait_request(dev, obj_priv->last_rendering_seqno);
+ if (ret != 0)
+ return ret;
+ }
+
+ return 0;
+}
+
+/**
+ * Unbinds an object from the GTT aperture.
+ */
+static int
+i915_gem_object_unbind(struct drm_gem_object *obj)
+{
+ struct drm_device *dev = obj->dev;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int ret = 0;
+
+#if WATCH_BUF
+ DRM_INFO("%s:%d %p\n", __func__, __LINE__, obj);
+ DRM_INFO("gtt_space %p\n", obj_priv->gtt_space);
+#endif
+ if (obj_priv->gtt_space == NULL)
+ return 0;
+
+ if (obj_priv->pin_count != 0) {
+ DRM_ERROR("Attempting to unbind pinned buffer\n");
+ return -EINVAL;
+ }
+
+ /* Wait for any rendering to complete
+ */
+ ret = i915_gem_object_wait_rendering(obj);
+ if (ret) {
+ DRM_ERROR("wait_rendering failed: %d\n", ret);
+ return ret;
+ }
+
+ /* Move the object to the CPU domain to ensure that
+ * any possible CPU writes while it's not in the GTT
+ * are flushed when we go to remap it. This will
+ * also ensure that all pending GPU writes are finished
+ * before we unbind.
+ */
+ ret = i915_gem_object_set_domain(obj, I915_GEM_DOMAIN_CPU,
+ I915_GEM_DOMAIN_CPU);
+ if (ret) {
+ DRM_ERROR("set_domain failed: %d\n", ret);
+ return ret;
+ }
+
+ if (obj_priv->agp_mem != NULL) {
+ drm_unbind_agp(obj_priv->agp_mem);
+ drm_free_agp(obj_priv->agp_mem, obj->size / PAGE_SIZE);
+ obj_priv->agp_mem = NULL;
+ }
+
+ BUG_ON(obj_priv->active);
+
+ i915_gem_object_free_page_list(obj);
+
+ if (obj_priv->gtt_space) {
+ atomic_dec(&dev->gtt_count);
+ atomic_sub(obj->size, &dev->gtt_memory);
+
+ drm_mm_put_block(obj_priv->gtt_space);
+ obj_priv->gtt_space = NULL;
+ }
+
+ /* Remove ourselves from the LRU list if present. */
+ if (!list_empty(&obj_priv->list))
+ list_del_init(&obj_priv->list);
+
+ return 0;
+}
+
+static int
+i915_gem_evict_something(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ int ret = 0;
+
+ for (;;) {
+ /* If there's an inactive buffer available now, grab it
+ * and be done.
+ */
+ if (!list_empty(&dev_priv->mm.inactive_list)) {
+ obj_priv = list_first_entry(&dev_priv->mm.inactive_list,
+ struct drm_i915_gem_object,
+ list);
+ obj = obj_priv->obj;
+ BUG_ON(obj_priv->pin_count != 0);
+#if WATCH_LRU
+ DRM_INFO("%s: evicting %p\n", __func__, obj);
+#endif
+ BUG_ON(obj_priv->active);
+
+ /* Wait on the rendering and unbind the buffer. */
+ ret = i915_gem_object_unbind(obj);
+ break;
+ }
+
+ /* If we didn't get anything, but the ring is still processing
+ * things, wait for one of those things to finish and hopefully
+ * leave us a buffer to evict.
+ */
+ if (!list_empty(&dev_priv->mm.request_list)) {
+ struct drm_i915_gem_request *request;
+
+ request = list_first_entry(&dev_priv->mm.request_list,
+ struct drm_i915_gem_request,
+ list);
+
+ ret = i915_wait_request(dev, request->seqno);
+ if (ret)
+ break;
+
+ /* if waiting caused an object to become inactive,
+ * then loop around and wait for it. Otherwise, we
+ * assume that waiting freed and unbound something,
+ * so there should now be some space in the GTT
+ */
+ if (!list_empty(&dev_priv->mm.inactive_list))
+ continue;
+ break;
+ }
+
+ /* If we didn't have anything on the request list but there
+ * are buffers awaiting a flush, emit one and try again.
+ * When we wait on it, those buffers waiting for that flush
+ * will get moved to inactive.
+ */
+ if (!list_empty(&dev_priv->mm.flushing_list)) {
+ obj_priv = list_first_entry(&dev_priv->mm.flushing_list,
+ struct drm_i915_gem_object,
+ list);
+ obj = obj_priv->obj;
+
+ i915_gem_flush(dev,
+ obj->write_domain,
+ obj->write_domain);
+ i915_add_request(dev, obj->write_domain);
+
+ obj = NULL;
+ continue;
+ }
+
+ DRM_ERROR("inactive empty %d request empty %d "
+ "flushing empty %d\n",
+ list_empty(&dev_priv->mm.inactive_list),
+ list_empty(&dev_priv->mm.request_list),
+ list_empty(&dev_priv->mm.flushing_list));
+ /* If we didn't do any of the above, there's nothing to be done
+ * and we just can't fit it in.
+ */
+ return -ENOMEM;
+ }
+ return ret;
+}
+
+static int
+i915_gem_object_get_page_list(struct drm_gem_object *obj)
+{
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int page_count, i;
+ struct address_space *mapping;
+ struct inode *inode;
+ struct page *page;
+ int ret;
+
+ if (obj_priv->page_list)
+ return 0;
+
+ /* Get the list of pages out of our struct file. They'll be pinned
+ * at this point until we release them.
+ */
+ page_count = obj->size / PAGE_SIZE;
+ BUG_ON(obj_priv->page_list != NULL);
+ obj_priv->page_list = drm_calloc(page_count, sizeof(struct page *),
+ DRM_MEM_DRIVER);
+ if (obj_priv->page_list == NULL) {
+ DRM_ERROR("Faled to allocate page list\n");
+ return -ENOMEM;
+ }
+
+ inode = obj->filp->f_path.dentry->d_inode;
+ mapping = inode->i_mapping;
+ for (i = 0; i < page_count; i++) {
+ page = find_get_page(mapping, i);
+ if (page == NULL || !PageUptodate(page)) {
+ if (page) {
+ page_cache_release(page);
+ page = NULL;
+ }
+ ret = shmem_getpage(inode, i, &page, SGP_DIRTY, NULL);
+
+ if (ret) {
+ DRM_ERROR("shmem_getpage failed: %d\n", ret);
+ i915_gem_object_free_page_list(obj);
+ return ret;
+ }
+ unlock_page(page);
+ }
+ obj_priv->page_list[i] = page;
+ }
+ return 0;
+}
+
+/**
+ * Finds free space in the GTT aperture and binds the object there.
+ */
+static int
+i915_gem_object_bind_to_gtt(struct drm_gem_object *obj, unsigned alignment)
+{
+ struct drm_device *dev = obj->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ struct drm_mm_node *free_space;
+ int page_count, ret;
+
+ if (alignment == 0)
+ alignment = PAGE_SIZE;
+ if (alignment & (PAGE_SIZE - 1)) {
+ DRM_ERROR("Invalid object alignment requested %u\n", alignment);
+ return -EINVAL;
+ }
+
+ search_free:
+ free_space = drm_mm_search_free(&dev_priv->mm.gtt_space,
+ obj->size, alignment, 0);
+ if (free_space != NULL) {
+ obj_priv->gtt_space = drm_mm_get_block(free_space, obj->size,
+ alignment);
+ if (obj_priv->gtt_space != NULL) {
+ obj_priv->gtt_space->private = obj;
+ obj_priv->gtt_offset = obj_priv->gtt_space->start;
+ }
+ }
+ if (obj_priv->gtt_space == NULL) {
+ /* If the gtt is empty and we're still having trouble
+ * fitting our object in, we're out of memory.
+ */
+#if WATCH_LRU
+ DRM_INFO("%s: GTT full, evicting something\n", __func__);
+#endif
+ if (list_empty(&dev_priv->mm.inactive_list) &&
+ list_empty(&dev_priv->mm.flushing_list) &&
+ list_empty(&dev_priv->mm.active_list)) {
+ DRM_ERROR("GTT full, but LRU list empty\n");
+ return -ENOMEM;
+ }
+
+ ret = i915_gem_evict_something(dev);
+ if (ret != 0) {
+ DRM_ERROR("Failed to evict a buffer %d\n", ret);
+ return ret;
+ }
+ goto search_free;
+ }
+
+#if WATCH_BUF
+ DRM_INFO("Binding object of size %d at 0x%08x\n",
+ obj->size, obj_priv->gtt_offset);
+#endif
+ ret = i915_gem_object_get_page_list(obj);
+ if (ret) {
+ drm_mm_put_block(obj_priv->gtt_space);
+ obj_priv->gtt_space = NULL;
+ return ret;
+ }
+
+ page_count = obj->size / PAGE_SIZE;
+ /* Create an AGP memory structure pointing at our pages, and bind it
+ * into the GTT.
+ */
+ obj_priv->agp_mem = drm_agp_bind_pages(dev,
+ obj_priv->page_list,
+ page_count,
+ obj_priv->gtt_offset);
+ if (obj_priv->agp_mem == NULL) {
+ i915_gem_object_free_page_list(obj);
+ drm_mm_put_block(obj_priv->gtt_space);
+ obj_priv->gtt_space = NULL;
+ return -ENOMEM;
+ }
+ atomic_inc(&dev->gtt_count);
+ atomic_add(obj->size, &dev->gtt_memory);
+
+ /* Assert that the object is not currently in any GPU domain. As it
+ * wasn't in the GTT, there shouldn't be any way it could have been in
+ * a GPU cache
+ */
+ BUG_ON(obj->read_domains & ~(I915_GEM_DOMAIN_CPU|I915_GEM_DOMAIN_GTT));
+ BUG_ON(obj->write_domain & ~(I915_GEM_DOMAIN_CPU|I915_GEM_DOMAIN_GTT));
+
+ return 0;
+}
+
+void
+i915_gem_clflush_object(struct drm_gem_object *obj)
+{
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+
+ /* If we don't have a page list set up, then we're not pinned
+ * to GPU, and we can ignore the cache flush because it'll happen
+ * again at bind time.
+ */
+ if (obj_priv->page_list == NULL)
+ return;
+
+ drm_clflush_pages(obj_priv->page_list, obj->size / PAGE_SIZE);
+}
+
+/*
+ * Set the next domain for the specified object. This
+ * may not actually perform the necessary flushing/invaliding though,
+ * as that may want to be batched with other set_domain operations
+ *
+ * This is (we hope) the only really tricky part of gem. The goal
+ * is fairly simple -- track which caches hold bits of the object
+ * and make sure they remain coherent. A few concrete examples may
+ * help to explain how it works. For shorthand, we use the notation
+ * (read_domains, write_domain), e.g. (CPU, CPU) to indicate the
+ * a pair of read and write domain masks.
+ *
+ * Case 1: the batch buffer
+ *
+ * 1. Allocated
+ * 2. Written by CPU
+ * 3. Mapped to GTT
+ * 4. Read by GPU
+ * 5. Unmapped from GTT
+ * 6. Freed
+ *
+ * Let's take these a step at a time
+ *
+ * 1. Allocated
+ * Pages allocated from the kernel may still have
+ * cache contents, so we set them to (CPU, CPU) always.
+ * 2. Written by CPU (using pwrite)
+ * The pwrite function calls set_domain (CPU, CPU) and
+ * this function does nothing (as nothing changes)
+ * 3. Mapped by GTT
+ * This function asserts that the object is not
+ * currently in any GPU-based read or write domains
+ * 4. Read by GPU
+ * i915_gem_execbuffer calls set_domain (COMMAND, 0).
+ * As write_domain is zero, this function adds in the
+ * current read domains (CPU+COMMAND, 0).
+ * flush_domains is set to CPU.
+ * invalidate_domains is set to COMMAND
+ * clflush is run to get data out of the CPU caches
+ * then i915_dev_set_domain calls i915_gem_flush to
+ * emit an MI_FLUSH and drm_agp_chipset_flush
+ * 5. Unmapped from GTT
+ * i915_gem_object_unbind calls set_domain (CPU, CPU)
+ * flush_domains and invalidate_domains end up both zero
+ * so no flushing/invalidating happens
+ * 6. Freed
+ * yay, done
+ *
+ * Case 2: The shared render buffer
+ *
+ * 1. Allocated
+ * 2. Mapped to GTT
+ * 3. Read/written by GPU
+ * 4. set_domain to (CPU,CPU)
+ * 5. Read/written by CPU
+ * 6. Read/written by GPU
+ *
+ * 1. Allocated
+ * Same as last example, (CPU, CPU)
+ * 2. Mapped to GTT
+ * Nothing changes (assertions find that it is not in the GPU)
+ * 3. Read/written by GPU
+ * execbuffer calls set_domain (RENDER, RENDER)
+ * flush_domains gets CPU
+ * invalidate_domains gets GPU
+ * clflush (obj)
+ * MI_FLUSH and drm_agp_chipset_flush
+ * 4. set_domain (CPU, CPU)
+ * flush_domains gets GPU
+ * invalidate_domains gets CPU
+ * wait_rendering (obj) to make sure all drawing is complete.
+ * This will include an MI_FLUSH to get the data from GPU
+ * to memory
+ * clflush (obj) to invalidate the CPU cache
+ * Another MI_FLUSH in i915_gem_flush (eliminate this somehow?)
+ * 5. Read/written by CPU
+ * cache lines are loaded and dirtied
+ * 6. Read written by GPU
+ * Same as last GPU access
+ *
+ * Case 3: The constant buffer
+ *
+ * 1. Allocated
+ * 2. Written by CPU
+ * 3. Read by GPU
+ * 4. Updated (written) by CPU again
+ * 5. Read by GPU
+ *
+ * 1. Allocated
+ * (CPU, CPU)
+ * 2. Written by CPU
+ * (CPU, CPU)
+ * 3. Read by GPU
+ * (CPU+RENDER, 0)
+ * flush_domains = CPU
+ * invalidate_domains = RENDER
+ * clflush (obj)
+ * MI_FLUSH
+ * drm_agp_chipset_flush
+ * 4. Updated (written) by CPU again
+ * (CPU, CPU)
+ * flush_domains = 0 (no previous write domain)
+ * invalidate_domains = 0 (no new read domains)
+ * 5. Read by GPU
+ * (CPU+RENDER, 0)
+ * flush_domains = CPU
+ * invalidate_domains = RENDER
+ * clflush (obj)
+ * MI_FLUSH
+ * drm_agp_chipset_flush
+ */
+static int
+i915_gem_object_set_domain(struct drm_gem_object *obj,
+ uint32_t read_domains,
+ uint32_t write_domain)
+{
+ struct drm_device *dev = obj->dev;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ uint32_t invalidate_domains = 0;
+ uint32_t flush_domains = 0;
+ int ret;
+
+#if WATCH_BUF
+ DRM_INFO("%s: object %p read %08x -> %08x write %08x -> %08x\n",
+ __func__, obj,
+ obj->read_domains, read_domains,
+ obj->write_domain, write_domain);
+#endif
+ /*
+ * If the object isn't moving to a new write domain,
+ * let the object stay in multiple read domains
+ */
+ if (write_domain == 0)
+ read_domains |= obj->read_domains;
+ else
+ obj_priv->dirty = 1;
+
+ /*
+ * Flush the current write domain if
+ * the new read domains don't match. Invalidate
+ * any read domains which differ from the old
+ * write domain
+ */
+ if (obj->write_domain && obj->write_domain != read_domains) {
+ flush_domains |= obj->write_domain;
+ invalidate_domains |= read_domains & ~obj->write_domain;
+ }
+ /*
+ * Invalidate any read caches which may have
+ * stale data. That is, any new read domains.
+ */
+ invalidate_domains |= read_domains & ~obj->read_domains;
+ if ((flush_domains | invalidate_domains) & I915_GEM_DOMAIN_CPU) {
+#if WATCH_BUF
+ DRM_INFO("%s: CPU domain flush %08x invalidate %08x\n",
+ __func__, flush_domains, invalidate_domains);
+#endif
+ /*
+ * If we're invaliding the CPU cache and flushing a GPU cache,
+ * then pause for rendering so that the GPU caches will be
+ * flushed before the cpu cache is invalidated
+ */
+ if ((invalidate_domains & I915_GEM_DOMAIN_CPU) &&
+ (flush_domains & ~(I915_GEM_DOMAIN_CPU |
+ I915_GEM_DOMAIN_GTT))) {
+ ret = i915_gem_object_wait_rendering(obj);
+ if (ret)
+ return ret;
+ }
+ i915_gem_clflush_object(obj);
+ }
+
+ if ((write_domain | flush_domains) != 0)
+ obj->write_domain = write_domain;
+
+ /* If we're invalidating the CPU domain, clear the per-page CPU
+ * domain list as well.
+ */
+ if (obj_priv->page_cpu_valid != NULL &&
+ (obj->read_domains & I915_GEM_DOMAIN_CPU) &&
+ ((read_domains & I915_GEM_DOMAIN_CPU) == 0)) {
+ memset(obj_priv->page_cpu_valid, 0, obj->size / PAGE_SIZE);
+ }
+ obj->read_domains = read_domains;
+
+ dev->invalidate_domains |= invalidate_domains;
+ dev->flush_domains |= flush_domains;
+#if WATCH_BUF
+ DRM_INFO("%s: read %08x write %08x invalidate %08x flush %08x\n",
+ __func__,
+ obj->read_domains, obj->write_domain,
+ dev->invalidate_domains, dev->flush_domains);
+#endif
+ return 0;
+}
+
+/**
+ * Set the read/write domain on a range of the object.
+ *
+ * Currently only implemented for CPU reads, otherwise drops to normal
+ * i915_gem_object_set_domain().
+ */
+static int
+i915_gem_object_set_domain_range(struct drm_gem_object *obj,
+ uint64_t offset,
+ uint64_t size,
+ uint32_t read_domains,
+ uint32_t write_domain)
+{
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int ret, i;
+
+ if (obj->read_domains & I915_GEM_DOMAIN_CPU)
+ return 0;
+
+ if (read_domains != I915_GEM_DOMAIN_CPU ||
+ write_domain != 0)
+ return i915_gem_object_set_domain(obj,
+ read_domains, write_domain);
+
+ /* Wait on any GPU rendering to the object to be flushed. */
+ if (obj->write_domain & ~(I915_GEM_DOMAIN_CPU | I915_GEM_DOMAIN_GTT)) {
+ ret = i915_gem_object_wait_rendering(obj);
+ if (ret)
+ return ret;
+ }
+
+ if (obj_priv->page_cpu_valid == NULL) {
+ obj_priv->page_cpu_valid = drm_calloc(1, obj->size / PAGE_SIZE,
+ DRM_MEM_DRIVER);
+ }
+
+ /* Flush the cache on any pages that are still invalid from the CPU's
+ * perspective.
+ */
+ for (i = offset / PAGE_SIZE; i < (offset + size - 1) / PAGE_SIZE; i++) {
+ if (obj_priv->page_cpu_valid[i])
+ continue;
+
+ drm_clflush_pages(obj_priv->page_list + i, 1);
+
+ obj_priv->page_cpu_valid[i] = 1;
+ }
+
+ return 0;
+}
+
+/**
+ * Once all of the objects have been set in the proper domain,
+ * perform the necessary flush and invalidate operations.
+ *
+ * Returns the write domains flushed, for use in flush tracking.
+ */
+static uint32_t
+i915_gem_dev_set_domain(struct drm_device *dev)
+{
+ uint32_t flush_domains = dev->flush_domains;
+
+ /*
+ * Now that all the buffers are synced to the proper domains,
+ * flush and invalidate the collected domains
+ */
+ if (dev->invalidate_domains | dev->flush_domains) {
+#if WATCH_EXEC
+ DRM_INFO("%s: invalidate_domains %08x flush_domains %08x\n",
+ __func__,
+ dev->invalidate_domains,
+ dev->flush_domains);
+#endif
+ i915_gem_flush(dev,
+ dev->invalidate_domains,
+ dev->flush_domains);
+ dev->invalidate_domains = 0;
+ dev->flush_domains = 0;
+ }
+
+ return flush_domains;
+}
+
+/**
+ * Pin an object to the GTT and evaluate the relocations landing in it.
+ */
+static int
+i915_gem_object_pin_and_relocate(struct drm_gem_object *obj,
+ struct drm_file *file_priv,
+ struct drm_i915_gem_exec_object *entry)
+{
+ struct drm_device *dev = obj->dev;
+ struct drm_i915_gem_relocation_entry reloc;
+ struct drm_i915_gem_relocation_entry __user *relocs;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int i, ret;
+ uint32_t last_reloc_offset = -1;
+ void *reloc_page = NULL;
+
+ /* Choose the GTT offset for our buffer and put it there. */
+ ret = i915_gem_object_pin(obj, (uint32_t) entry->alignment);
+ if (ret)
+ return ret;
+
+ entry->offset = obj_priv->gtt_offset;
+
+ relocs = (struct drm_i915_gem_relocation_entry __user *)
+ (uintptr_t) entry->relocs_ptr;
+ /* Apply the relocations, using the GTT aperture to avoid cache
+ * flushing requirements.
+ */
+ for (i = 0; i < entry->relocation_count; i++) {
+ struct drm_gem_object *target_obj;
+ struct drm_i915_gem_object *target_obj_priv;
+ uint32_t reloc_val, reloc_offset, *reloc_entry;
+ int ret;
+
+ ret = copy_from_user(&reloc, relocs + i, sizeof(reloc));
+ if (ret != 0) {
+ i915_gem_object_unpin(obj);
+ return ret;
+ }
+
+ target_obj = drm_gem_object_lookup(obj->dev, file_priv,
+ reloc.target_handle);
+ if (target_obj == NULL) {
+ i915_gem_object_unpin(obj);
+ return -EBADF;
+ }
+ target_obj_priv = target_obj->driver_private;
+
+ /* The target buffer should have appeared before us in the
+ * exec_object list, so it should have a GTT space bound by now.
+ */
+ if (target_obj_priv->gtt_space == NULL) {
+ DRM_ERROR("No GTT space found for object %d\n",
+ reloc.target_handle);
+ drm_gem_object_unreference(target_obj);
+ i915_gem_object_unpin(obj);
+ return -EINVAL;
+ }
+
+ if (reloc.offset > obj->size - 4) {
+ DRM_ERROR("Relocation beyond object bounds: "
+ "obj %p target %d offset %d size %d.\n",
+ obj, reloc.target_handle,
+ (int) reloc.offset, (int) obj->size);
+ drm_gem_object_unreference(target_obj);
+ i915_gem_object_unpin(obj);
+ return -EINVAL;
+ }
+ if (reloc.offset & 3) {
+ DRM_ERROR("Relocation not 4-byte aligned: "
+ "obj %p target %d offset %d.\n",
+ obj, reloc.target_handle,
+ (int) reloc.offset);
+ drm_gem_object_unreference(target_obj);
+ i915_gem_object_unpin(obj);
+ return -EINVAL;
+ }
+
+ if (reloc.write_domain && target_obj->pending_write_domain &&
+ reloc.write_domain != target_obj->pending_write_domain) {
+ DRM_ERROR("Write domain conflict: "
+ "obj %p target %d offset %d "
+ "new %08x old %08x\n",
+ obj, reloc.target_handle,
+ (int) reloc.offset,
+ reloc.write_domain,
+ target_obj->pending_write_domain);
+ drm_gem_object_unreference(target_obj);
+ i915_gem_object_unpin(obj);
+ return -EINVAL;
+ }
+
+#if WATCH_RELOC
+ DRM_INFO("%s: obj %p offset %08x target %d "
+ "read %08x write %08x gtt %08x "
+ "presumed %08x delta %08x\n",
+ __func__,
+ obj,
+ (int) reloc.offset,
+ (int) reloc.target_handle,
+ (int) reloc.read_domains,
+ (int) reloc.write_domain,
+ (int) target_obj_priv->gtt_offset,
+ (int) reloc.presumed_offset,
+ reloc.delta);
+#endif
+
+ target_obj->pending_read_domains |= reloc.read_domains;
+ target_obj->pending_write_domain |= reloc.write_domain;
+
+ /* If the relocation already has the right value in it, no
+ * more work needs to be done.
+ */
+ if (target_obj_priv->gtt_offset == reloc.presumed_offset) {
+ drm_gem_object_unreference(target_obj);
+ continue;
+ }
+
+ /* Now that we're going to actually write some data in,
+ * make sure that any rendering using this buffer's contents
+ * is completed.
+ */
+ i915_gem_object_wait_rendering(obj);
+
+ /* As we're writing through the gtt, flush
+ * any CPU writes before we write the relocations
+ */
+ if (obj->write_domain & I915_GEM_DOMAIN_CPU) {
+ i915_gem_clflush_object(obj);
+ drm_agp_chipset_flush(dev);
+ obj->write_domain = 0;
+ }
+
+ /* Map the page containing the relocation we're going to
+ * perform.
+ */
+ reloc_offset = obj_priv->gtt_offset + reloc.offset;
+ if (reloc_page == NULL ||
+ (last_reloc_offset & ~(PAGE_SIZE - 1)) !=
+ (reloc_offset & ~(PAGE_SIZE - 1))) {
+ if (reloc_page != NULL)
+ iounmap(reloc_page);
+
+ reloc_page = ioremap(dev->agp->base +
+ (reloc_offset & ~(PAGE_SIZE - 1)),
+ PAGE_SIZE);
+ last_reloc_offset = reloc_offset;
+ if (reloc_page == NULL) {
+ drm_gem_object_unreference(target_obj);
+ i915_gem_object_unpin(obj);
+ return -ENOMEM;
+ }
+ }
+
+ reloc_entry = (uint32_t *)((char *)reloc_page +
+ (reloc_offset & (PAGE_SIZE - 1)));
+ reloc_val = target_obj_priv->gtt_offset + reloc.delta;
+
+#if WATCH_BUF
+ DRM_INFO("Applied relocation: %p@0x%08x %08x -> %08x\n",
+ obj, (unsigned int) reloc.offset,
+ readl(reloc_entry), reloc_val);
+#endif
+ writel(reloc_val, reloc_entry);
+
+ /* Write the updated presumed offset for this entry back out
+ * to the user.
+ */
+ reloc.presumed_offset = target_obj_priv->gtt_offset;
+ ret = copy_to_user(relocs + i, &reloc, sizeof(reloc));
+ if (ret != 0) {
+ drm_gem_object_unreference(target_obj);
+ i915_gem_object_unpin(obj);
+ return ret;
+ }
+
+ drm_gem_object_unreference(target_obj);
+ }
+
+ if (reloc_page != NULL)
+ iounmap(reloc_page);
+
+#if WATCH_BUF
+ if (0)
+ i915_gem_dump_object(obj, 128, __func__, ~0);
+#endif
+ return 0;
+}
+
+/** Dispatch a batchbuffer to the ring
+ */
+static int
+i915_dispatch_gem_execbuffer(struct drm_device *dev,
+ struct drm_i915_gem_execbuffer *exec,
+ uint64_t exec_offset)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_clip_rect __user *boxes = (struct drm_clip_rect __user *)
+ (uintptr_t) exec->cliprects_ptr;
+ int nbox = exec->num_cliprects;
+ int i = 0, count;
+ uint32_t exec_start, exec_len;
+ RING_LOCALS;
+
+ exec_start = (uint32_t) exec_offset + exec->batch_start_offset;
+ exec_len = (uint32_t) exec->batch_len;
+
+ if ((exec_start | exec_len) & 0x7) {
+ DRM_ERROR("alignment\n");
+ return -EINVAL;
+ }
+
+ if (!exec_start)
+ return -EINVAL;
+
+ count = nbox ? nbox : 1;
+
+ for (i = 0; i < count; i++) {
+ if (i < nbox) {
+ int ret = i915_emit_box(dev, boxes, i,
+ exec->DR1, exec->DR4);
+ if (ret)
+ return ret;
+ }
+
+ if (IS_I830(dev) || IS_845G(dev)) {
+ BEGIN_LP_RING(4);
+ OUT_RING(MI_BATCH_BUFFER);
+ OUT_RING(exec_start | MI_BATCH_NON_SECURE);
+ OUT_RING(exec_start + exec_len - 4);
+ OUT_RING(0);
+ ADVANCE_LP_RING();
+ } else {
+ BEGIN_LP_RING(2);
+ if (IS_I965G(dev)) {
+ OUT_RING(MI_BATCH_BUFFER_START |
+ (2 << 6) |
+ MI_BATCH_NON_SECURE_I965);
+ OUT_RING(exec_start);
+ } else {
+ OUT_RING(MI_BATCH_BUFFER_START |
+ (2 << 6));
+ OUT_RING(exec_start | MI_BATCH_NON_SECURE);
+ }
+ ADVANCE_LP_RING();
+ }
+ }
+
+ /* XXX breadcrumb */
+ return 0;
+}
+
+/* Throttle our rendering by waiting until the ring has completed our requests
+ * emitted over 20 msec ago.
+ *
+ * This should get us reasonable parallelism between CPU and GPU but also
+ * relatively low latency when blocking on a particular request to finish.
+ */
+static int
+i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file_priv)
+{
+ struct drm_i915_file_private *i915_file_priv = file_priv->driver_priv;
+ int ret = 0;
+ uint32_t seqno;
+
+ mutex_lock(&dev->struct_mutex);
+ seqno = i915_file_priv->mm.last_gem_throttle_seqno;
+ i915_file_priv->mm.last_gem_throttle_seqno =
+ i915_file_priv->mm.last_gem_seqno;
+ if (seqno)
+ ret = i915_wait_request(dev, seqno);
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+}
+
+int
+i915_gem_execbuffer(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_file_private *i915_file_priv = file_priv->driver_priv;
+ struct drm_i915_gem_execbuffer *args = data;
+ struct drm_i915_gem_exec_object *exec_list = NULL;
+ struct drm_gem_object **object_list = NULL;
+ struct drm_gem_object *batch_obj;
+ int ret, i, pinned = 0;
+ uint64_t exec_offset;
+ uint32_t seqno, flush_domains;
+
+#if WATCH_EXEC
+ DRM_INFO("buffers_ptr %d buffer_count %d len %08x\n",
+ (int) args->buffers_ptr, args->buffer_count, args->batch_len);
+#endif
+
+ /* Copy in the exec list from userland */
+ exec_list = drm_calloc(sizeof(*exec_list), args->buffer_count,
+ DRM_MEM_DRIVER);
+ object_list = drm_calloc(sizeof(*object_list), args->buffer_count,
+ DRM_MEM_DRIVER);
+ if (exec_list == NULL || object_list == NULL) {
+ DRM_ERROR("Failed to allocate exec or object list "
+ "for %d buffers\n",
+ args->buffer_count);
+ ret = -ENOMEM;
+ goto pre_mutex_err;
+ }
+ ret = copy_from_user(exec_list,
+ (struct drm_i915_relocation_entry __user *)
+ (uintptr_t) args->buffers_ptr,
+ sizeof(*exec_list) * args->buffer_count);
+ if (ret != 0) {
+ DRM_ERROR("copy %d exec entries failed %d\n",
+ args->buffer_count, ret);
+ goto pre_mutex_err;
+ }
+
+ mutex_lock(&dev->struct_mutex);
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+
+ if (dev_priv->mm.wedged) {
+ DRM_ERROR("Execbuf while wedged\n");
+ mutex_unlock(&dev->struct_mutex);
+ return -EIO;
+ }
+
+ if (dev_priv->mm.suspended) {
+ DRM_ERROR("Execbuf while VT-switched.\n");
+ mutex_unlock(&dev->struct_mutex);
+ return -EBUSY;
+ }
+
+ /* Zero the gloabl flush/invalidate flags. These
+ * will be modified as each object is bound to the
+ * gtt
+ */
+ dev->invalidate_domains = 0;
+ dev->flush_domains = 0;
+
+ /* Look up object handles and perform the relocations */
+ for (i = 0; i < args->buffer_count; i++) {
+ object_list[i] = drm_gem_object_lookup(dev, file_priv,
+ exec_list[i].handle);
+ if (object_list[i] == NULL) {
+ DRM_ERROR("Invalid object handle %d at index %d\n",
+ exec_list[i].handle, i);
+ ret = -EBADF;
+ goto err;
+ }
+
+ object_list[i]->pending_read_domains = 0;
+ object_list[i]->pending_write_domain = 0;
+ ret = i915_gem_object_pin_and_relocate(object_list[i],
+ file_priv,
+ &exec_list[i]);
+ if (ret) {
+ DRM_ERROR("object bind and relocate failed %d\n", ret);
+ goto err;
+ }
+ pinned = i + 1;
+ }
+
+ /* Set the pending read domains for the batch buffer to COMMAND */
+ batch_obj = object_list[args->buffer_count-1];
+ batch_obj->pending_read_domains = I915_GEM_DOMAIN_COMMAND;
+ batch_obj->pending_write_domain = 0;
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+
+ for (i = 0; i < args->buffer_count; i++) {
+ struct drm_gem_object *obj = object_list[i];
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+
+ if (obj_priv->gtt_space == NULL) {
+ /* We evicted the buffer in the process of validating
+ * our set of buffers in. We could try to recover by
+ * kicking them everything out and trying again from
+ * the start.
+ */
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ /* make sure all previous memory operations have passed */
+ ret = i915_gem_object_set_domain(obj,
+ obj->pending_read_domains,
+ obj->pending_write_domain);
+ if (ret)
+ goto err;
+ }
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+
+ /* Flush/invalidate caches and chipset buffer */
+ flush_domains = i915_gem_dev_set_domain(dev);
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+
+#if WATCH_COHERENCY
+ for (i = 0; i < args->buffer_count; i++) {
+ i915_gem_object_check_coherency(object_list[i],
+ exec_list[i].handle);
+ }
+#endif
+
+ exec_offset = exec_list[args->buffer_count - 1].offset;
+
+#if WATCH_EXEC
+ i915_gem_dump_object(object_list[args->buffer_count - 1],
+ args->batch_len,
+ __func__,
+ ~0);
+#endif
+
+ /* Exec the batchbuffer */
+ ret = i915_dispatch_gem_execbuffer(dev, args, exec_offset);
+ if (ret) {
+ DRM_ERROR("dispatch failed %d\n", ret);
+ goto err;
+ }
+
+ /*
+ * Ensure that the commands in the batch buffer are
+ * finished before the interrupt fires
+ */
+ flush_domains |= i915_retire_commands(dev);
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+
+ /*
+ * Get a seqno representing the execution of the current buffer,
+ * which we can wait on. We would like to mitigate these interrupts,
+ * likely by only creating seqnos occasionally (so that we have
+ * *some* interrupts representing completion of buffers that we can
+ * wait on when trying to clear up gtt space).
+ */
+ seqno = i915_add_request(dev, flush_domains);
+ BUG_ON(seqno == 0);
+ i915_file_priv->mm.last_gem_seqno = seqno;
+ for (i = 0; i < args->buffer_count; i++) {
+ struct drm_gem_object *obj = object_list[i];
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+
+ i915_gem_object_move_to_active(obj);
+ obj_priv->last_rendering_seqno = seqno;
+#if WATCH_LRU
+ DRM_INFO("%s: move to exec list %p\n", __func__, obj);
+#endif
+ }
+#if WATCH_LRU
+ i915_dump_lru(dev, __func__);
+#endif
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+
+ /* Copy the new buffer offsets back to the user's exec list. */
+ ret = copy_to_user((struct drm_i915_relocation_entry __user *)
+ (uintptr_t) args->buffers_ptr,
+ exec_list,
+ sizeof(*exec_list) * args->buffer_count);
+ if (ret)
+ DRM_ERROR("failed to copy %d exec entries "
+ "back to user (%d)\n",
+ args->buffer_count, ret);
+err:
+ if (object_list != NULL) {
+ for (i = 0; i < pinned; i++)
+ i915_gem_object_unpin(object_list[i]);
+
+ for (i = 0; i < args->buffer_count; i++)
+ drm_gem_object_unreference(object_list[i]);
+ }
+ mutex_unlock(&dev->struct_mutex);
+
+pre_mutex_err:
+ drm_free(object_list, sizeof(*object_list) * args->buffer_count,
+ DRM_MEM_DRIVER);
+ drm_free(exec_list, sizeof(*exec_list) * args->buffer_count,
+ DRM_MEM_DRIVER);
+
+ return ret;
+}
+
+int
+i915_gem_object_pin(struct drm_gem_object *obj, uint32_t alignment)
+{
+ struct drm_device *dev = obj->dev;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int ret;
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+ if (obj_priv->gtt_space == NULL) {
+ ret = i915_gem_object_bind_to_gtt(obj, alignment);
+ if (ret != 0) {
+ DRM_ERROR("Failure to bind: %d", ret);
+ return ret;
+ }
+ }
+ obj_priv->pin_count++;
+
+ /* If the object is not active and not pending a flush,
+ * remove it from the inactive list
+ */
+ if (obj_priv->pin_count == 1) {
+ atomic_inc(&dev->pin_count);
+ atomic_add(obj->size, &dev->pin_memory);
+ if (!obj_priv->active &&
+ (obj->write_domain & ~(I915_GEM_DOMAIN_CPU |
+ I915_GEM_DOMAIN_GTT)) == 0 &&
+ !list_empty(&obj_priv->list))
+ list_del_init(&obj_priv->list);
+ }
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+
+ return 0;
+}
+
+void
+i915_gem_object_unpin(struct drm_gem_object *obj)
+{
+ struct drm_device *dev = obj->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+ obj_priv->pin_count--;
+ BUG_ON(obj_priv->pin_count < 0);
+ BUG_ON(obj_priv->gtt_space == NULL);
+
+ /* If the object is no longer pinned, and is
+ * neither active nor being flushed, then stick it on
+ * the inactive list
+ */
+ if (obj_priv->pin_count == 0) {
+ if (!obj_priv->active &&
+ (obj->write_domain & ~(I915_GEM_DOMAIN_CPU |
+ I915_GEM_DOMAIN_GTT)) == 0)
+ list_move_tail(&obj_priv->list,
+ &dev_priv->mm.inactive_list);
+ atomic_dec(&dev->pin_count);
+ atomic_sub(obj->size, &dev->pin_memory);
+ }
+ i915_verify_inactive(dev, __FILE__, __LINE__);
+}
+
+int
+i915_gem_pin_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_pin *args = data;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ int ret;
+
+ mutex_lock(&dev->struct_mutex);
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL) {
+ DRM_ERROR("Bad handle in i915_gem_pin_ioctl(): %d\n",
+ args->handle);
+ mutex_unlock(&dev->struct_mutex);
+ return -EBADF;
+ }
+ obj_priv = obj->driver_private;
+
+ ret = i915_gem_object_pin(obj, args->alignment);
+ if (ret != 0) {
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+ }
+
+ /* XXX - flush the CPU caches for pinned objects
+ * as the X server doesn't manage domains yet
+ */
+ if (obj->write_domain & I915_GEM_DOMAIN_CPU) {
+ i915_gem_clflush_object(obj);
+ drm_agp_chipset_flush(dev);
+ obj->write_domain = 0;
+ }
+ args->offset = obj_priv->gtt_offset;
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+
+ return 0;
+}
+
+int
+i915_gem_unpin_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_pin *args = data;
+ struct drm_gem_object *obj;
+
+ mutex_lock(&dev->struct_mutex);
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL) {
+ DRM_ERROR("Bad handle in i915_gem_unpin_ioctl(): %d\n",
+ args->handle);
+ mutex_unlock(&dev->struct_mutex);
+ return -EBADF;
+ }
+
+ i915_gem_object_unpin(obj);
+
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ return 0;
+}
+
+int
+i915_gem_busy_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_busy *args = data;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+
+ mutex_lock(&dev->struct_mutex);
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL) {
+ DRM_ERROR("Bad handle in i915_gem_busy_ioctl(): %d\n",
+ args->handle);
+ mutex_unlock(&dev->struct_mutex);
+ return -EBADF;
+ }
+
+ obj_priv = obj->driver_private;
+ args->busy = obj_priv->active;
+
+ drm_gem_object_unreference(obj);
+ mutex_unlock(&dev->struct_mutex);
+ return 0;
+}
+
+int
+i915_gem_throttle_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ return i915_gem_ring_throttle(dev, file_priv);
+}
+
+int i915_gem_init_object(struct drm_gem_object *obj)
+{
+ struct drm_i915_gem_object *obj_priv;
+
+ obj_priv = drm_calloc(1, sizeof(*obj_priv), DRM_MEM_DRIVER);
+ if (obj_priv == NULL)
+ return -ENOMEM;
+
+ /*
+ * We've just allocated pages from the kernel,
+ * so they've just been written by the CPU with
+ * zeros. They'll need to be clflushed before we
+ * use them with the GPU.
+ */
+ obj->write_domain = I915_GEM_DOMAIN_CPU;
+ obj->read_domains = I915_GEM_DOMAIN_CPU;
+
+ obj->driver_private = obj_priv;
+ obj_priv->obj = obj;
+ INIT_LIST_HEAD(&obj_priv->list);
+ return 0;
+}
+
+void i915_gem_free_object(struct drm_gem_object *obj)
+{
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+
+ while (obj_priv->pin_count > 0)
+ i915_gem_object_unpin(obj);
+
+ i915_gem_object_unbind(obj);
+
+ drm_free(obj_priv->page_cpu_valid, 1, DRM_MEM_DRIVER);
+ drm_free(obj->driver_private, 1, DRM_MEM_DRIVER);
+}
+
+static int
+i915_gem_set_domain(struct drm_gem_object *obj,
+ struct drm_file *file_priv,
+ uint32_t read_domains,
+ uint32_t write_domain)
+{
+ struct drm_device *dev = obj->dev;
+ int ret;
+ uint32_t flush_domains;
+
+ BUG_ON(!mutex_is_locked(&dev->struct_mutex));
+
+ ret = i915_gem_object_set_domain(obj, read_domains, write_domain);
+ if (ret)
+ return ret;
+ flush_domains = i915_gem_dev_set_domain(obj->dev);
+
+ if (flush_domains & ~(I915_GEM_DOMAIN_CPU|I915_GEM_DOMAIN_GTT))
+ (void) i915_add_request(dev, flush_domains);
+
+ return 0;
+}
+
+/** Unbinds all objects that are on the given buffer list. */
+static int
+i915_gem_evict_from_list(struct drm_device *dev, struct list_head *head)
+{
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ int ret;
+
+ while (!list_empty(head)) {
+ obj_priv = list_first_entry(head,
+ struct drm_i915_gem_object,
+ list);
+ obj = obj_priv->obj;
+
+ if (obj_priv->pin_count != 0) {
+ DRM_ERROR("Pinned object in unbind list\n");
+ mutex_unlock(&dev->struct_mutex);
+ return -EINVAL;
+ }
+
+ ret = i915_gem_object_unbind(obj);
+ if (ret != 0) {
+ DRM_ERROR("Error unbinding object in LeaveVT: %d\n",
+ ret);
+ mutex_unlock(&dev->struct_mutex);
+ return ret;
+ }
+ }
+
+
+ return 0;
+}
+
+static int
+i915_gem_idle(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ uint32_t seqno, cur_seqno, last_seqno;
+ int stuck;
+
+ if (dev_priv->mm.suspended)
+ return 0;
+
+ /* Hack! Don't let anybody do execbuf while we don't control the chip.
+ * We need to replace this with a semaphore, or something.
+ */
+ dev_priv->mm.suspended = 1;
+
+ i915_kernel_lost_context(dev);
+
+ /* Flush the GPU along with all non-CPU write domains
+ */
+ i915_gem_flush(dev, ~(I915_GEM_DOMAIN_CPU|I915_GEM_DOMAIN_GTT),
+ ~(I915_GEM_DOMAIN_CPU|I915_GEM_DOMAIN_GTT));
+ seqno = i915_add_request(dev, ~(I915_GEM_DOMAIN_CPU |
+ I915_GEM_DOMAIN_GTT));
+
+ if (seqno == 0) {
+ mutex_unlock(&dev->struct_mutex);
+ return -ENOMEM;
+ }
+
+ dev_priv->mm.waiting_gem_seqno = seqno;
+ last_seqno = 0;
+ stuck = 0;
+ for (;;) {
+ cur_seqno = i915_get_gem_seqno(dev);
+ if (i915_seqno_passed(cur_seqno, seqno))
+ break;
+ if (last_seqno == cur_seqno) {
+ if (stuck++ > 100) {
+ DRM_ERROR("hardware wedged\n");
+ dev_priv->mm.wedged = 1;
+ DRM_WAKEUP(&dev_priv->irq_queue);
+ break;
+ }
+ }
+ msleep(10);
+ last_seqno = cur_seqno;
+ }
+ dev_priv->mm.waiting_gem_seqno = 0;
+
+ i915_gem_retire_requests(dev);
+
+ /* Active and flushing should now be empty as we've
+ * waited for a sequence higher than any pending execbuffer
+ */
+ BUG_ON(!list_empty(&dev_priv->mm.active_list));
+ BUG_ON(!list_empty(&dev_priv->mm.flushing_list));
+
+ /* Request should now be empty as we've also waited
+ * for the last request in the list
+ */
+ BUG_ON(!list_empty(&dev_priv->mm.request_list));
+
+ /* Move all buffers out of the GTT. */
+ i915_gem_evict_from_list(dev, &dev_priv->mm.inactive_list);
+
+ BUG_ON(!list_empty(&dev_priv->mm.active_list));
+ BUG_ON(!list_empty(&dev_priv->mm.flushing_list));
+ BUG_ON(!list_empty(&dev_priv->mm.inactive_list));
+ BUG_ON(!list_empty(&dev_priv->mm.request_list));
+ return 0;
+}
+
+static int
+i915_gem_init_hws(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ int ret;
+
+ /* If we need a physical address for the status page, it's already
+ * initialized at driver load time.
+ */
+ if (!I915_NEED_GFX_HWS(dev))
+ return 0;
+
+ obj = drm_gem_object_alloc(dev, 4096);
+ if (obj == NULL) {
+ DRM_ERROR("Failed to allocate status page\n");
+ return -ENOMEM;
+ }
+ obj_priv = obj->driver_private;
+
+ ret = i915_gem_object_pin(obj, 4096);
+ if (ret != 0) {
+ drm_gem_object_unreference(obj);
+ return ret;
+ }
+
+ dev_priv->status_gfx_addr = obj_priv->gtt_offset;
+ dev_priv->hws_map.offset = dev->agp->base + obj_priv->gtt_offset;
+ dev_priv->hws_map.size = 4096;
+ dev_priv->hws_map.type = 0;
+ dev_priv->hws_map.flags = 0;
+ dev_priv->hws_map.mtrr = 0;
+
+ drm_core_ioremap(&dev_priv->hws_map, dev);
+ if (dev_priv->hws_map.handle == NULL) {
+ DRM_ERROR("Failed to map status page.\n");
+ memset(&dev_priv->hws_map, 0, sizeof(dev_priv->hws_map));
+ drm_gem_object_unreference(obj);
+ return -EINVAL;
+ }
+ dev_priv->hws_obj = obj;
+ dev_priv->hw_status_page = dev_priv->hws_map.handle;
+ memset(dev_priv->hw_status_page, 0, PAGE_SIZE);
+ I915_WRITE(HWS_PGA, dev_priv->status_gfx_addr);
+ DRM_DEBUG("hws offset: 0x%08x\n", dev_priv->status_gfx_addr);
+
+ return 0;
+}
+
+static int
+i915_gem_init_ringbuffer(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+ int ret;
+
+ ret = i915_gem_init_hws(dev);
+ if (ret != 0)
+ return ret;
+
+ obj = drm_gem_object_alloc(dev, 128 * 1024);
+ if (obj == NULL) {
+ DRM_ERROR("Failed to allocate ringbuffer\n");
+ return -ENOMEM;
+ }
+ obj_priv = obj->driver_private;
+
+ ret = i915_gem_object_pin(obj, 4096);
+ if (ret != 0) {
+ drm_gem_object_unreference(obj);
+ return ret;
+ }
+
+ /* Set up the kernel mapping for the ring. */
+ dev_priv->ring.Size = obj->size;
+ dev_priv->ring.tail_mask = obj->size - 1;
+
+ dev_priv->ring.map.offset = dev->agp->base + obj_priv->gtt_offset;
+ dev_priv->ring.map.size = obj->size;
+ dev_priv->ring.map.type = 0;
+ dev_priv->ring.map.flags = 0;
+ dev_priv->ring.map.mtrr = 0;
+
+ drm_core_ioremap(&dev_priv->ring.map, dev);
+ if (dev_priv->ring.map.handle == NULL) {
+ DRM_ERROR("Failed to map ringbuffer.\n");
+ memset(&dev_priv->ring, 0, sizeof(dev_priv->ring));
+ drm_gem_object_unreference(obj);
+ return -EINVAL;
+ }
+ dev_priv->ring.ring_obj = obj;
+ dev_priv->ring.virtual_start = dev_priv->ring.map.handle;
+
+ /* Stop the ring if it's running. */
+ I915_WRITE(PRB0_CTL, 0);
+ I915_WRITE(PRB0_HEAD, 0);
+ I915_WRITE(PRB0_TAIL, 0);
+ I915_WRITE(PRB0_START, 0);
+
+ /* Initialize the ring. */
+ I915_WRITE(PRB0_START, obj_priv->gtt_offset);
+ I915_WRITE(PRB0_CTL,
+ ((obj->size - 4096) & RING_NR_PAGES) |
+ RING_NO_REPORT |
+ RING_VALID);
+
+ /* Update our cache of the ring state */
+ i915_kernel_lost_context(dev);
+
+ return 0;
+}
+
+static void
+i915_gem_cleanup_ringbuffer(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+
+ if (dev_priv->ring.ring_obj == NULL)
+ return;
+
+ drm_core_ioremapfree(&dev_priv->ring.map, dev);
+
+ i915_gem_object_unpin(dev_priv->ring.ring_obj);
+ drm_gem_object_unreference(dev_priv->ring.ring_obj);
+ dev_priv->ring.ring_obj = NULL;
+ memset(&dev_priv->ring, 0, sizeof(dev_priv->ring));
+
+ if (dev_priv->hws_obj != NULL) {
+ i915_gem_object_unpin(dev_priv->hws_obj);
+ drm_gem_object_unreference(dev_priv->hws_obj);
+ dev_priv->hws_obj = NULL;
+ memset(&dev_priv->hws_map, 0, sizeof(dev_priv->hws_map));
+
+ /* Write high address into HWS_PGA when disabling. */
+ I915_WRITE(HWS_PGA, 0x1ffff000);
+ }
+}
+
+int
+i915_gem_entervt_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ int ret;
+
+ if (dev_priv->mm.wedged) {
+ DRM_ERROR("Reenabling wedged hardware, good luck\n");
+ dev_priv->mm.wedged = 0;
+ }
+
+ ret = i915_gem_init_ringbuffer(dev);
+ if (ret != 0)
+ return ret;
+
+ mutex_lock(&dev->struct_mutex);
+ BUG_ON(!list_empty(&dev_priv->mm.active_list));
+ BUG_ON(!list_empty(&dev_priv->mm.flushing_list));
+ BUG_ON(!list_empty(&dev_priv->mm.inactive_list));
+ BUG_ON(!list_empty(&dev_priv->mm.request_list));
+ dev_priv->mm.suspended = 0;
+ mutex_unlock(&dev->struct_mutex);
+ return 0;
+}
+
+int
+i915_gem_leavevt_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ int ret;
+
+ mutex_lock(&dev->struct_mutex);
+ ret = i915_gem_idle(dev);
+ if (ret == 0)
+ i915_gem_cleanup_ringbuffer(dev);
+ mutex_unlock(&dev->struct_mutex);
+
+ return 0;
+}
+
+void
+i915_gem_lastclose(struct drm_device *dev)
+{
+ int ret;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+
+ mutex_lock(&dev->struct_mutex);
+
+ if (dev_priv->ring.ring_obj != NULL) {
+ ret = i915_gem_idle(dev);
+ if (ret)
+ DRM_ERROR("failed to idle hardware: %d\n", ret);
+
+ i915_gem_cleanup_ringbuffer(dev);
+ }
+
+ mutex_unlock(&dev->struct_mutex);
+}
+
+void
+i915_gem_load(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+
+ INIT_LIST_HEAD(&dev_priv->mm.active_list);
+ INIT_LIST_HEAD(&dev_priv->mm.flushing_list);
+ INIT_LIST_HEAD(&dev_priv->mm.inactive_list);
+ INIT_LIST_HEAD(&dev_priv->mm.request_list);
+ INIT_DELAYED_WORK(&dev_priv->mm.retire_work,
+ i915_gem_retire_work_handler);
+ dev_priv->mm.next_gem_seqno = 1;
+
+ i915_gem_detect_bit_6_swizzle(dev);
+}
diff --git a/drivers/gpu/drm/i915/i915_gem_debug.c b/drivers/gpu/drm/i915/i915_gem_debug.c
new file mode 100644
index 0000000..131c088
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_debug.c
@@ -0,0 +1,201 @@
+/*
+ * Copyright © 2008 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.
+ *
+ * Authors:
+ * Keith Packard <[email protected]>
+ *
+ */
+
+#include "drmP.h"
+#include "drm.h"
+#include "i915_drm.h"
+#include "i915_drv.h"
+
+#if WATCH_INACTIVE
+void
+i915_verify_inactive(struct drm_device *dev, char *file, int line)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+
+ list_for_each_entry(obj_priv, &dev_priv->mm.inactive_list, list) {
+ obj = obj_priv->obj;
+ if (obj_priv->pin_count || obj_priv->active ||
+ (obj->write_domain & ~(I915_GEM_DOMAIN_CPU |
+ I915_GEM_DOMAIN_GTT)))
+ DRM_ERROR("inactive %p (p %d a %d w %x) %s:%d\n",
+ obj,
+ obj_priv->pin_count, obj_priv->active,
+ obj->write_domain, file, line);
+ }
+}
+#endif /* WATCH_INACTIVE */
+
+
+#if WATCH_BUF | WATCH_EXEC | WATCH_PWRITE
+static void
+i915_gem_dump_page(struct page *page, uint32_t start, uint32_t end,
+ uint32_t bias, uint32_t mark)
+{
+ uint32_t *mem = kmap_atomic(page, KM_USER0);
+ int i;
+ for (i = start; i < end; i += 4)
+ DRM_INFO("%08x: %08x%s\n",
+ (int) (bias + i), mem[i / 4],
+ (bias + i == mark) ? " ********" : "");
+ kunmap_atomic(mem, KM_USER0);
+ /* give syslog time to catch up */
+ msleep(1);
+}
+
+void
+i915_gem_dump_object(struct drm_gem_object *obj, int len,
+ const char *where, uint32_t mark)
+{
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int page;
+
+ DRM_INFO("%s: object at offset %08x\n", where, obj_priv->gtt_offset);
+ for (page = 0; page < (len + PAGE_SIZE-1) / PAGE_SIZE; page++) {
+ int page_len, chunk, chunk_len;
+
+ page_len = len - page * PAGE_SIZE;
+ if (page_len > PAGE_SIZE)
+ page_len = PAGE_SIZE;
+
+ for (chunk = 0; chunk < page_len; chunk += 128) {
+ chunk_len = page_len - chunk;
+ if (chunk_len > 128)
+ chunk_len = 128;
+ i915_gem_dump_page(obj_priv->page_list[page],
+ chunk, chunk + chunk_len,
+ obj_priv->gtt_offset +
+ page * PAGE_SIZE,
+ mark);
+ }
+ }
+}
+#endif
+
+#if WATCH_LRU
+void
+i915_dump_lru(struct drm_device *dev, const char *where)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv;
+
+ DRM_INFO("active list %s {\n", where);
+ list_for_each_entry(obj_priv, &dev_priv->mm.active_list,
+ list)
+ {
+ DRM_INFO(" %p: %08x\n", obj_priv,
+ obj_priv->last_rendering_seqno);
+ }
+ DRM_INFO("}\n");
+ DRM_INFO("flushing list %s {\n", where);
+ list_for_each_entry(obj_priv, &dev_priv->mm.flushing_list,
+ list)
+ {
+ DRM_INFO(" %p: %08x\n", obj_priv,
+ obj_priv->last_rendering_seqno);
+ }
+ DRM_INFO("}\n");
+ DRM_INFO("inactive %s {\n", where);
+ list_for_each_entry(obj_priv, &dev_priv->mm.inactive_list, list) {
+ DRM_INFO(" %p: %08x\n", obj_priv,
+ obj_priv->last_rendering_seqno);
+ }
+ DRM_INFO("}\n");
+}
+#endif
+
+
+#if WATCH_COHERENCY
+void
+i915_gem_object_check_coherency(struct drm_gem_object *obj, int handle)
+{
+ struct drm_device *dev = obj->dev;
+ struct drm_i915_gem_object *obj_priv = obj->driver_private;
+ int page;
+ uint32_t *gtt_mapping;
+ uint32_t *backing_map = NULL;
+ int bad_count = 0;
+
+ DRM_INFO("%s: checking coherency of object %p@0x%08x (%d, %dkb):\n",
+ __func__, obj, obj_priv->gtt_offset, handle,
+ obj->size / 1024);
+
+ gtt_mapping = ioremap(dev->agp->base + obj_priv->gtt_offset,
+ obj->size);
+ if (gtt_mapping == NULL) {
+ DRM_ERROR("failed to map GTT space\n");
+ return;
+ }
+
+ for (page = 0; page < obj->size / PAGE_SIZE; page++) {
+ int i;
+
+ backing_map = kmap_atomic(obj_priv->page_list[page], KM_USER0);
+
+ if (backing_map == NULL) {
+ DRM_ERROR("failed to map backing page\n");
+ goto out;
+ }
+
+ for (i = 0; i < PAGE_SIZE / 4; i++) {
+ uint32_t cpuval = backing_map[i];
+ uint32_t gttval = readl(gtt_mapping +
+ page * 1024 + i);
+
+ if (cpuval != gttval) {
+ DRM_INFO("incoherent CPU vs GPU at 0x%08x: "
+ "0x%08x vs 0x%08x\n",
+ (int)(obj_priv->gtt_offset +
+ page * PAGE_SIZE + i * 4),
+ cpuval, gttval);
+ if (bad_count++ >= 8) {
+ DRM_INFO("...\n");
+ goto out;
+ }
+ }
+ }
+ kunmap_atomic(backing_map, KM_USER0);
+ backing_map = NULL;
+ }
+
+ out:
+ if (backing_map != NULL)
+ kunmap_atomic(backing_map, KM_USER0);
+ iounmap(gtt_mapping);
+
+ /* give syslog time to catch up */
+ msleep(1);
+
+ /* Directly flush the object, since we just loaded values with the CPU
+ * from the backing pages and we don't want to disturb the cache
+ * management that we're trying to observe.
+ */
+
+ i915_gem_clflush_object(obj);
+}
+#endif
diff --git a/drivers/gpu/drm/i915/i915_gem_proc.c b/drivers/gpu/drm/i915/i915_gem_proc.c
new file mode 100644
index 0000000..15d4160
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_proc.c
@@ -0,0 +1,292 @@
+/*
+ * Copyright © 2008 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.
+ *
+ * Authors:
+ * Eric Anholt <[email protected]>
+ * Keith Packard <[email protected]>
+ *
+ */
+
+#include "drmP.h"
+#include "drm.h"
+#include "i915_drm.h"
+#include "i915_drv.h"
+
+static int i915_gem_active_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv;
+ int len = 0;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ *start = &buf[offset];
+ *eof = 0;
+ DRM_PROC_PRINT("Active:\n");
+ list_for_each_entry(obj_priv, &dev_priv->mm.active_list,
+ list)
+ {
+ struct drm_gem_object *obj = obj_priv->obj;
+ if (obj->name) {
+ DRM_PROC_PRINT(" %p(%d): %08x %08x %d\n",
+ obj, obj->name,
+ obj->read_domains, obj->write_domain,
+ obj_priv->last_rendering_seqno);
+ } else {
+ DRM_PROC_PRINT(" %p: %08x %08x %d\n",
+ obj,
+ obj->read_domains, obj->write_domain,
+ obj_priv->last_rendering_seqno);
+ }
+ }
+ if (len > request + offset)
+ return request;
+ *eof = 1;
+ return len - offset;
+}
+
+static int i915_gem_flushing_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv;
+ int len = 0;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ *start = &buf[offset];
+ *eof = 0;
+ DRM_PROC_PRINT("Flushing:\n");
+ list_for_each_entry(obj_priv, &dev_priv->mm.flushing_list,
+ list)
+ {
+ struct drm_gem_object *obj = obj_priv->obj;
+ if (obj->name) {
+ DRM_PROC_PRINT(" %p(%d): %08x %08x %d\n",
+ obj, obj->name,
+ obj->read_domains, obj->write_domain,
+ obj_priv->last_rendering_seqno);
+ } else {
+ DRM_PROC_PRINT(" %p: %08x %08x %d\n", obj,
+ obj->read_domains, obj->write_domain,
+ obj_priv->last_rendering_seqno);
+ }
+ }
+ if (len > request + offset)
+ return request;
+ *eof = 1;
+ return len - offset;
+}
+
+static int i915_gem_inactive_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_object *obj_priv;
+ int len = 0;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ *start = &buf[offset];
+ *eof = 0;
+ DRM_PROC_PRINT("Inactive:\n");
+ list_for_each_entry(obj_priv, &dev_priv->mm.inactive_list,
+ list)
+ {
+ struct drm_gem_object *obj = obj_priv->obj;
+ if (obj->name) {
+ DRM_PROC_PRINT(" %p(%d): %08x %08x %d\n",
+ obj, obj->name,
+ obj->read_domains, obj->write_domain,
+ obj_priv->last_rendering_seqno);
+ } else {
+ DRM_PROC_PRINT(" %p: %08x %08x %d\n", obj,
+ obj->read_domains, obj->write_domain,
+ obj_priv->last_rendering_seqno);
+ }
+ }
+ if (len > request + offset)
+ return request;
+ *eof = 1;
+ return len - offset;
+}
+
+static int i915_gem_request_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_i915_gem_request *gem_request;
+ int len = 0;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ *start = &buf[offset];
+ *eof = 0;
+ DRM_PROC_PRINT("Request:\n");
+ list_for_each_entry(gem_request, &dev_priv->mm.request_list,
+ list)
+ {
+ DRM_PROC_PRINT(" %d @ %d %08x\n",
+ gem_request->seqno,
+ (int) (jiffies - gem_request->emitted_jiffies),
+ gem_request->flush_domains);
+ }
+ if (len > request + offset)
+ return request;
+ *eof = 1;
+ return len - offset;
+}
+
+static int i915_gem_seqno_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ int len = 0;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ *start = &buf[offset];
+ *eof = 0;
+ DRM_PROC_PRINT("Current sequence: %d\n", i915_get_gem_seqno(dev));
+ DRM_PROC_PRINT("Waiter sequence: %d\n",
+ dev_priv->mm.waiting_gem_seqno);
+ DRM_PROC_PRINT("IRQ sequence: %d\n", dev_priv->mm.irq_gem_seqno);
+ if (len > request + offset)
+ return request;
+ *eof = 1;
+ return len - offset;
+}
+
+
+static int i915_interrupt_info(char *buf, char **start, off_t offset,
+ int request, int *eof, void *data)
+{
+ struct drm_minor *minor = (struct drm_minor *) data;
+ struct drm_device *dev = minor->dev;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ int len = 0;
+
+ if (offset > DRM_PROC_LIMIT) {
+ *eof = 1;
+ return 0;
+ }
+
+ *start = &buf[offset];
+ *eof = 0;
+ DRM_PROC_PRINT("Interrupt enable: %08x\n",
+ I915_READ(IER));
+ DRM_PROC_PRINT("Interrupt identity: %08x\n",
+ I915_READ(IIR));
+ DRM_PROC_PRINT("Interrupt mask: %08x\n",
+ I915_READ(IMR));
+ DRM_PROC_PRINT("Pipe A stat: %08x\n",
+ I915_READ(PIPEASTAT));
+ DRM_PROC_PRINT("Pipe B stat: %08x\n",
+ I915_READ(PIPEBSTAT));
+ DRM_PROC_PRINT("Interrupts received: %d\n",
+ atomic_read(&dev_priv->irq_received));
+ DRM_PROC_PRINT("Current sequence: %d\n",
+ i915_get_gem_seqno(dev));
+ DRM_PROC_PRINT("Waiter sequence: %d\n",
+ dev_priv->mm.waiting_gem_seqno);
+ DRM_PROC_PRINT("IRQ sequence: %d\n",
+ dev_priv->mm.irq_gem_seqno);
+ if (len > request + offset)
+ return request;
+ *eof = 1;
+ return len - offset;
+}
+
+static struct drm_proc_list {
+ /** file name */
+ const char *name;
+ /** proc callback*/
+ int (*f) (char *, char **, off_t, int, int *, void *);
+} i915_gem_proc_list[] = {
+ {"i915_gem_active", i915_gem_active_info},
+ {"i915_gem_flushing", i915_gem_flushing_info},
+ {"i915_gem_inactive", i915_gem_inactive_info},
+ {"i915_gem_request", i915_gem_request_info},
+ {"i915_gem_seqno", i915_gem_seqno_info},
+ {"i915_gem_interrupt", i915_interrupt_info},
+};
+
+#define I915_GEM_PROC_ENTRIES ARRAY_SIZE(i915_gem_proc_list)
+
+int i915_gem_proc_init(struct drm_minor *minor)
+{
+ struct proc_dir_entry *ent;
+ int i, j;
+
+ for (i = 0; i < I915_GEM_PROC_ENTRIES; i++) {
+ ent = create_proc_entry(i915_gem_proc_list[i].name,
+ S_IFREG | S_IRUGO, minor->dev_root);
+ if (!ent) {
+ DRM_ERROR("Cannot create /proc/dri/.../%s\n",
+ i915_gem_proc_list[i].name);
+ for (j = 0; j < i; j++)
+ remove_proc_entry(i915_gem_proc_list[i].name,
+ minor->dev_root);
+ return -1;
+ }
+ ent->read_proc = i915_gem_proc_list[i].f;
+ ent->data = minor;
+ }
+ return 0;
+}
+
+void i915_gem_proc_cleanup(struct drm_minor *minor)
+{
+ int i;
+
+ if (!minor->dev_root)
+ return;
+
+ for (i = 0; i < I915_GEM_PROC_ENTRIES; i++)
+ remove_proc_entry(i915_gem_proc_list[i].name, minor->dev_root);
+}
diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c
new file mode 100644
index 0000000..3af834d
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_tiling.c
@@ -0,0 +1,318 @@
+/*
+ * Copyright © 2008 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.
+ *
+ * Authors:
+ * Eric Anholt <[email protected]>
+ *
+ */
+
+#include "drmP.h"
+#include "drm.h"
+#include "i915_drm.h"
+#include "i915_drv.h"
+
+/** @file i915_gem_tiling.c
+ *
+ * Support for managing tiling state of buffer objects.
+ *
+ * The idea behind tiling is to increase cache hit rates by rearranging
+ * pixel data so that a group of pixel accesses are in the same cacheline.
+ * Performance improvement from doing this on the back/depth buffer are on
+ * the order of 30%.
+ *
+ * Intel architectures make this somewhat more complicated, though, by
+ * adjustments made to addressing of data when the memory is in interleaved
+ * mode (matched pairs of DIMMS) to improve memory bandwidth.
+ * For interleaved memory, the CPU sends every sequential 64 bytes
+ * to an alternate memory channel so it can get the bandwidth from both.
+ *
+ * The GPU also rearranges its accesses for increased bandwidth to interleaved
+ * memory, and it matches what the CPU does for non-tiled. However, when tiled
+ * it does it a little differently, since one walks addresses not just in the
+ * X direction but also Y. So, along with alternating channels when bit
+ * 6 of the address flips, it also alternates when other bits flip -- Bits 9
+ * (every 512 bytes, an X tile scanline) and 10 (every two X tile scanlines)
+ * are common to both the 915 and 965-class hardware.
+ *
+ * The CPU also sometimes XORs in higher bits as well, to improve
+ * bandwidth doing strided access like we do so frequently in graphics. This
+ * is called "Channel XOR Randomization" in the MCH documentation. The result
+ * is that the CPU is XORing in either bit 11 or bit 17 to bit 6 of its address
+ * decode.
+ *
+ * All of this bit 6 XORing has an effect on our memory management,
+ * as we need to make sure that the 3d driver can correctly address object
+ * contents.
+ *
+ * If we don't have interleaved memory, all tiling is safe and no swizzling is
+ * required.
+ *
+ * When bit 17 is XORed in, we simply refuse to tile at all. Bit
+ * 17 is not just a page offset, so as we page an objet out and back in,
+ * individual pages in it will have different bit 17 addresses, resulting in
+ * each 64 bytes being swapped with its neighbor!
+ *
+ * Otherwise, if interleaved, we have to tell the 3d driver what the address
+ * swizzling it needs to do is, since it's writing with the CPU to the pages
+ * (bit 6 and potentially bit 11 XORed in), and the GPU is reading from the
+ * pages (bit 6, 9, and 10 XORed in), resulting in a cumulative bit swizzling
+ * required by the CPU of XORing in bit 6, 9, 10, and potentially 11, in order
+ * to match what the GPU expects.
+ */
+
+/**
+ * Detects bit 6 swizzling of address lookup between IGD access and CPU
+ * access through main memory.
+ */
+void
+i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
+{
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct pci_dev *bridge;
+ uint32_t swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
+ uint32_t swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
+ int mchbar_offset;
+ enum pci_bar_type mchbar_type;
+ char __iomem *mchbar;
+ int ret;
+
+ bridge = pci_get_bus_and_slot(0, PCI_DEVFN(0, 0));
+ if (bridge == NULL) {
+ DRM_ERROR("Couldn't get bridge device\n");
+ return;
+ }
+
+ ret = pci_enable_device(bridge);
+ if (ret != 0) {
+ DRM_ERROR("pci_enable_device failed: %d\n", ret);
+ return;
+ }
+
+ if (IS_I965G(dev) || IS_G33(dev)) {
+ mchbar_type = pci_bar_mem64;
+ mchbar_offset = 0x48;
+ } else {
+ mchbar_type = pci_bar_mem32;
+ mchbar_offset = 0x44;
+ }
+
+ /* Use resource 2 for our BAR that's stashed in a nonstandard location,
+ * since the bridge would only ever use standard BARs 0-1 (though it
+ * doesn't anyway)
+ */
+ ret = pci_read_base(bridge, mchbar_type, &bridge->resource[2],
+ mchbar_offset);
+ /* pci_read_base may return these two error codes indicating that it
+ * couldn't quite set up the resource. We don't really care, as long
+ * as it's valid for ioremapping, which is the case.
+ */
+ if (ret != 0 && ret != -EBUSY && ret != -ENODEV) {
+ DRM_ERROR("pci_read_base failed: %d\n", ret);
+ return;
+ }
+
+ mchbar = ioremap(pci_resource_start(bridge, 2),
+ pci_resource_len(bridge, 2));
+ if (mchbar == NULL) {
+ DRM_ERROR("Couldn't map MCHBAR to determine tile swizzling\n");
+ return;
+ }
+
+ if (IS_I965G(dev) && !IS_I965GM(dev)) {
+ uint32_t chdecmisc;
+
+ /* On the 965, channel interleave appears to be determined by
+ * the flex bit. If flex is set, then the ranks (sides of a
+ * DIMM) of memory will be "stacked" (physical addresses walk
+ * through one rank then move on to the next, flipping channels
+ * or not depending on rank configuration). The GPU in this
+ * case does exactly the same addressing as the CPU.
+ *
+ * Unlike the 945, channel randomization based does not
+ * appear to be available.
+ *
+ * XXX: While the G965 doesn't appear to do any interleaving
+ * when the DIMMs are not exactly matched, the G4x chipsets
+ * might be for "L-shaped" configurations, and will need to be
+ * detected.
+ *
+ * L-shaped configuration:
+ *
+ * +-----+
+ * | |
+ * |DIMM2| <-- non-interleaved
+ * +-----+
+ * +-----+ +-----+
+ * | | | |
+ * |DIMM0| |DIMM1| <-- interleaved area
+ * +-----+ +-----+
+ */
+ chdecmisc = readb(mchbar + CHDECMISC);
+
+ if (chdecmisc == 0xff) {
+ DRM_ERROR("Couldn't read from MCHBAR. "
+ "Disabling tiling.\n");
+ } else if (chdecmisc & CHDECMISC_FLEXMEMORY) {
+ swizzle_x = I915_BIT_6_SWIZZLE_NONE;
+ swizzle_y = I915_BIT_6_SWIZZLE_NONE;
+ } else {
+ swizzle_x = I915_BIT_6_SWIZZLE_9_10;
+ swizzle_y = I915_BIT_6_SWIZZLE_9;
+ }
+ } else if (IS_I9XX(dev)) {
+ uint32_t dcc;
+
+ /* On 915-945 and GM965, channel interleave by the CPU is
+ * determined by DCC. The CPU will alternate based on bit 6
+ * in interleaved mode, and the GPU will then also alternate
+ * on bit 6, 9, and 10 for X, but the CPU may also optionally
+ * alternate based on bit 17 (XOR not disabled and XOR
+ * bit == 17).
+ */
+ dcc = readl(mchbar + DCC);
+ switch (dcc & DCC_ADDRESSING_MODE_MASK) {
+ case DCC_ADDRESSING_MODE_SINGLE_CHANNEL:
+ case DCC_ADDRESSING_MODE_DUAL_CHANNEL_ASYMMETRIC:
+ swizzle_x = I915_BIT_6_SWIZZLE_NONE;
+ swizzle_y = I915_BIT_6_SWIZZLE_NONE;
+ break;
+ case DCC_ADDRESSING_MODE_DUAL_CHANNEL_INTERLEAVED:
+ if (IS_I915G(dev) || IS_I915GM(dev) ||
+ dcc & DCC_CHANNEL_XOR_DISABLE) {
+ swizzle_x = I915_BIT_6_SWIZZLE_9_10;
+ swizzle_y = I915_BIT_6_SWIZZLE_9;
+ } else if (IS_I965GM(dev)) {
+ /* GM965 only does bit 11-based channel
+ * randomization
+ */
+ swizzle_x = I915_BIT_6_SWIZZLE_9_10_11;
+ swizzle_y = I915_BIT_6_SWIZZLE_9_11;
+ } else {
+ /* Bit 17 or perhaps other swizzling */
+ swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
+ swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
+ }
+ break;
+ }
+ if (dcc == 0xffffffff) {
+ DRM_ERROR("Couldn't read from MCHBAR. "
+ "Disabling tiling.\n");
+ swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
+ swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
+ }
+ } else {
+ /* As far as we know, the 865 doesn't have these bit 6
+ * swizzling issues.
+ */
+ swizzle_x = I915_BIT_6_SWIZZLE_NONE;
+ swizzle_y = I915_BIT_6_SWIZZLE_NONE;
+ }
+
+ iounmap(mchbar);
+
+ dev_priv->mm.bit_6_swizzle_x = swizzle_x;
+ dev_priv->mm.bit_6_swizzle_y = swizzle_y;
+}
+
+/**
+ * Sets the tiling mode of an object, returning the required swizzling of
+ * bit 6 of addresses in the object.
+ */
+int
+i915_gem_set_tiling(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_set_tiling *args = data;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL)
+ return -EINVAL;
+ obj_priv = obj->driver_private;
+
+ mutex_lock(&dev->struct_mutex);
+
+ if (args->tiling_mode == I915_TILING_NONE) {
+ obj_priv->tiling_mode = I915_TILING_NONE;
+ args->swizzle_mode = I915_BIT_6_SWIZZLE_NONE;
+ } else {
+ if (args->tiling_mode == I915_TILING_X)
+ args->swizzle_mode = dev_priv->mm.bit_6_swizzle_x;
+ else
+ args->swizzle_mode = dev_priv->mm.bit_6_swizzle_y;
+ /* If we can't handle the swizzling, make it untiled. */
+ if (args->swizzle_mode == I915_BIT_6_SWIZZLE_UNKNOWN) {
+ args->tiling_mode = I915_TILING_NONE;
+ args->swizzle_mode = I915_BIT_6_SWIZZLE_NONE;
+ }
+ }
+ obj_priv->tiling_mode = args->tiling_mode;
+
+ mutex_unlock(&dev->struct_mutex);
+
+ drm_gem_object_unreference(obj);
+
+ return 0;
+}
+
+/**
+ * Returns the current tiling mode and required bit 6 swizzling for the object.
+ */
+int
+i915_gem_get_tiling(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_i915_gem_get_tiling *args = data;
+ drm_i915_private_t *dev_priv = dev->dev_private;
+ struct drm_gem_object *obj;
+ struct drm_i915_gem_object *obj_priv;
+
+ obj = drm_gem_object_lookup(dev, file_priv, args->handle);
+ if (obj == NULL)
+ return -EINVAL;
+ obj_priv = obj->driver_private;
+
+ mutex_lock(&dev->struct_mutex);
+
+ args->tiling_mode = obj_priv->tiling_mode;
+ switch (obj_priv->tiling_mode) {
+ case I915_TILING_X:
+ args->swizzle_mode = dev_priv->mm.bit_6_swizzle_x;
+ break;
+ case I915_TILING_Y:
+ args->swizzle_mode = dev_priv->mm.bit_6_swizzle_y;
+ break;
+ case I915_TILING_NONE:
+ args->swizzle_mode = I915_BIT_6_SWIZZLE_NONE;
+ break;
+ default:
+ DRM_ERROR("unknown tiling mode\n");
+ }
+
+ mutex_unlock(&dev->struct_mutex);
+
+ drm_gem_object_unreference(obj);
+
+ return 0;
+}
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 24d11ed..a64b9bf 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -281,8 +281,10 @@ irqreturn_t i915_driver_irq_handler(DRM_IRQ_ARGS)

dev_priv->sarea_priv->last_dispatch = READ_BREADCRUMB(dev_priv);

- if (iir & I915_USER_INTERRUPT)
+ if (iir & I915_USER_INTERRUPT) {
+ dev_priv->mm.irq_gem_seqno = i915_get_gem_seqno(dev);
DRM_WAKEUP(&dev_priv->irq_queue);
+ }

if (iir & (I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT |
I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT)) {
@@ -343,7 +345,7 @@ static int i915_emit_irq(struct drm_device * dev)
return dev_priv->counter;
}

-static void i915_user_irq_get(struct drm_device *dev)
+void i915_user_irq_get(struct drm_device *dev)
{
drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;

@@ -353,7 +355,7 @@ static void i915_user_irq_get(struct drm_device *dev)
spin_unlock(&dev_priv->user_irq_lock);
}

-static void i915_user_irq_put(struct drm_device *dev)
+void i915_user_irq_put(struct drm_device *dev)
{
drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;

diff --git a/include/drm/drm.h b/include/drm/drm.h
index 38d3c6b..dc347ef 100644
--- a/include/drm/drm.h
+++ b/include/drm/drm.h
@@ -573,6 +573,34 @@ struct drm_set_version {
int drm_dd_minor;
};

+/** DRM_IOCTL_GEM_CLOSE ioctl argument type */
+struct drm_gem_close {
+ /** Handle of the object to be closed. */
+ uint32_t handle;
+ uint32_t pad;
+};
+
+/** DRM_IOCTL_GEM_FLINK ioctl argument type */
+struct drm_gem_flink {
+ /** Handle for the object being named */
+ uint32_t handle;
+
+ /** Returned global name */
+ uint32_t name;
+};
+
+/** DRM_IOCTL_GEM_OPEN ioctl argument type */
+struct drm_gem_open {
+ /** Name of object being opened */
+ uint32_t name;
+
+ /** Returned handle for the object */
+ uint32_t handle;
+
+ /** Returned size of the object */
+ uint64_t size;
+};
+
#define DRM_IOCTL_BASE 'd'
#define DRM_IO(nr) _IO(DRM_IOCTL_BASE,nr)
#define DRM_IOR(nr,type) _IOR(DRM_IOCTL_BASE,nr,type)
@@ -587,6 +615,9 @@ struct drm_set_version {
#define DRM_IOCTL_GET_CLIENT DRM_IOWR(0x05, struct drm_client)
#define DRM_IOCTL_GET_STATS DRM_IOR( 0x06, struct drm_stats)
#define DRM_IOCTL_SET_VERSION DRM_IOWR(0x07, struct drm_set_version)
+#define DRM_IOCTL_GEM_CLOSE DRM_IOW (0x09, struct drm_gem_close)
+#define DRM_IOCTL_GEM_FLINK DRM_IOWR(0x0a, struct drm_gem_flink)
+#define DRM_IOCTL_GEM_OPEN DRM_IOWR(0x0b, struct drm_gem_open)

#define DRM_IOCTL_SET_UNIQUE DRM_IOW( 0x10, struct drm_unique)
#define DRM_IOCTL_AUTH_MAGIC DRM_IOW( 0x11, struct drm_auth)
diff --git a/include/drm/drmP.h b/include/drm/drmP.h
index 1c1b13e..a540db8 100644
--- a/include/drm/drmP.h
+++ b/include/drm/drmP.h
@@ -104,6 +104,7 @@ struct drm_device;
#define DRIVER_DMA_QUEUE 0x200
#define DRIVER_FB_DMA 0x400
#define DRIVER_IRQ_VBL2 0x800
+#define DRIVER_GEM 0x1000

/***********************************************************************/
/** \name Begin the DRM... */
@@ -387,6 +388,10 @@ struct drm_file {
struct drm_minor *minor;
int remove_auth_on_close;
unsigned long lock_count;
+ /** Mapping of mm object handles to object pointers. */
+ struct idr object_idr;
+ /** Lock for synchronization of access to object_idr. */
+ spinlock_t table_lock;
struct file *filp;
void *driver_priv;
};
@@ -558,6 +563,56 @@ struct drm_ati_pcigart_info {
};

/**
+ * This structure defines the drm_mm memory object, which will be used by the
+ * DRM for its buffer objects.
+ */
+struct drm_gem_object {
+ /** Reference count of this object */
+ struct kref refcount;
+
+ /** Handle count of this object. Each handle also holds a reference */
+ struct kref handlecount;
+
+ /** Related drm device */
+ struct drm_device *dev;
+
+ /** File representing the shmem storage */
+ struct file *filp;
+
+ /**
+ * Size of the object, in bytes. Immutable over the object's
+ * lifetime.
+ */
+ size_t size;
+
+ /**
+ * Global name for this object, starts at 1. 0 means unnamed.
+ * Access is covered by the object_name_lock in the related drm_device
+ */
+ int name;
+
+ /**
+ * Memory domains. These monitor which caches contain read/write data
+ * related to the object. When transitioning from one set of domains
+ * to another, the driver is called to ensure that caches are suitably
+ * flushed and invalidated
+ */
+ uint32_t read_domains;
+ uint32_t write_domain;
+
+ /**
+ * While validating an exec operation, the
+ * new read/write domain values are computed here.
+ * They will be transferred to the above values
+ * at the point that any cache flushing occurs
+ */
+ uint32_t pending_read_domains;
+ uint32_t pending_write_domain;
+
+ void *driver_private;
+};
+
+/**
* DRM driver structure. This structure represent the common code for
* a family of cards. There will one drm_device for each card present
* in this family
@@ -614,6 +669,18 @@ struct drm_driver {
void (*set_version) (struct drm_device *dev,
struct drm_set_version *sv);

+ int (*proc_init)(struct drm_minor *minor);
+ void (*proc_cleanup)(struct drm_minor *minor);
+
+ /**
+ * Driver-specific constructor for drm_gem_objects, to set up
+ * obj->driver_private.
+ *
+ * Returns 0 on success.
+ */
+ int (*gem_init_object) (struct drm_gem_object *obj);
+ void (*gem_free_object) (struct drm_gem_object *obj);
+
int major;
int minor;
int patchlevel;
@@ -771,6 +838,22 @@ struct drm_device {
spinlock_t drw_lock;
struct idr drw_idr;
/*@} */
+
+ /** \name GEM information */
+ /*@{ */
+ spinlock_t object_name_lock;
+ struct idr object_name_idr;
+ atomic_t object_count;
+ atomic_t object_memory;
+ atomic_t pin_count;
+ atomic_t pin_memory;
+ atomic_t gtt_count;
+ atomic_t gtt_memory;
+ uint32_t gtt_total;
+ uint32_t invalidate_domains; /* domains pending invalidation */
+ uint32_t flush_domains; /* domains pending flush */
+ /*@} */
+
};

static __inline__ int drm_core_check_feature(struct drm_device *dev,
@@ -867,6 +950,10 @@ extern void *drm_realloc(void *oldpt, size_t oldsize, size_t size, int area);
extern DRM_AGP_MEM *drm_alloc_agp(struct drm_device *dev, int pages, u32 type);
extern int drm_free_agp(DRM_AGP_MEM * handle, int pages);
extern int drm_bind_agp(DRM_AGP_MEM * handle, unsigned int start);
+extern DRM_AGP_MEM *drm_agp_bind_pages(struct drm_device *dev,
+ struct page **pages,
+ unsigned long num_pages,
+ uint32_t gtt_offset);
extern int drm_unbind_agp(DRM_AGP_MEM * handle);

/* Misc. IOCTL support (drm_ioctl.h) */
@@ -929,6 +1016,9 @@ extern int drm_getmagic(struct drm_device *dev, void *data,
extern int drm_authmagic(struct drm_device *dev, void *data,
struct drm_file *file_priv);

+/* Cache management (drm_cache.c) */
+void drm_clflush_pages(struct page *pages[], unsigned long num_pages);
+
/* Locking IOCTL support (drm_lock.h) */
extern int drm_lock(struct drm_device *dev, void *data,
struct drm_file *file_priv);
@@ -1026,6 +1116,7 @@ extern DRM_AGP_MEM *drm_agp_allocate_memory(struct agp_bridge_data *bridge, size
extern int drm_agp_free_memory(DRM_AGP_MEM * handle);
extern int drm_agp_bind_memory(DRM_AGP_MEM * handle, off_t start);
extern int drm_agp_unbind_memory(DRM_AGP_MEM * handle);
+extern void drm_agp_chipset_flush(struct drm_device *dev);

/* Stub support (drm_stub.h) */
extern int drm_get_dev(struct pci_dev *pdev, const struct pci_device_id *ent,
@@ -1088,6 +1179,66 @@ extern unsigned long drm_mm_tail_space(struct drm_mm *mm);
extern int drm_mm_remove_space_from_tail(struct drm_mm *mm, unsigned long size);
extern int drm_mm_add_space_to_tail(struct drm_mm *mm, unsigned long size);

+/* Graphics Execution Manager library functions (drm_gem.c) */
+int drm_gem_init(struct drm_device *dev);
+void drm_gem_object_free(struct kref *kref);
+struct drm_gem_object *drm_gem_object_alloc(struct drm_device *dev,
+ size_t size);
+void drm_gem_object_handle_free(struct kref *kref);
+
+static inline void
+drm_gem_object_reference(struct drm_gem_object *obj)
+{
+ kref_get(&obj->refcount);
+}
+
+static inline void
+drm_gem_object_unreference(struct drm_gem_object *obj)
+{
+ if (obj == NULL)
+ return;
+
+ kref_put(&obj->refcount, drm_gem_object_free);
+}
+
+int drm_gem_handle_create(struct drm_file *file_priv,
+ struct drm_gem_object *obj,
+ int *handlep);
+
+static inline void
+drm_gem_object_handle_reference(struct drm_gem_object *obj)
+{
+ drm_gem_object_reference(obj);
+ kref_get(&obj->handlecount);
+}
+
+static inline void
+drm_gem_object_handle_unreference(struct drm_gem_object *obj)
+{
+ if (obj == NULL)
+ return;
+
+ /*
+ * Must bump handle count first as this may be the last
+ * ref, in which case the object would disappear before we
+ * checked for a name
+ */
+ kref_put(&obj->handlecount, drm_gem_object_handle_free);
+ drm_gem_object_unreference(obj);
+}
+
+struct drm_gem_object *drm_gem_object_lookup(struct drm_device *dev,
+ struct drm_file *filp,
+ int handle);
+int drm_gem_close_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int drm_gem_flink_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int drm_gem_open_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+void drm_gem_open(struct drm_device *dev, struct drm_file *file_private);
+void drm_gem_release(struct drm_device *dev, struct drm_file *file_private);
+
extern void drm_core_ioremap(struct drm_map *map, struct drm_device *dev);
extern void drm_core_ioremap_wc(struct drm_map *map, struct drm_device *dev);
extern void drm_core_ioremapfree(struct drm_map *map, struct drm_device *dev);
diff --git a/include/drm/i915_drm.h b/include/drm/i915_drm.h
index 05c66cf..59d08fc 100644
--- a/include/drm/i915_drm.h
+++ b/include/drm/i915_drm.h
@@ -143,6 +143,22 @@ typedef struct _drm_i915_sarea {
#define DRM_I915_GET_VBLANK_PIPE 0x0e
#define DRM_I915_VBLANK_SWAP 0x0f
#define DRM_I915_HWS_ADDR 0x11
+#define DRM_I915_GEM_INIT 0x13
+#define DRM_I915_GEM_EXECBUFFER 0x14
+#define DRM_I915_GEM_PIN 0x15
+#define DRM_I915_GEM_UNPIN 0x16
+#define DRM_I915_GEM_BUSY 0x17
+#define DRM_I915_GEM_THROTTLE 0x18
+#define DRM_I915_GEM_ENTERVT 0x19
+#define DRM_I915_GEM_LEAVEVT 0x1a
+#define DRM_I915_GEM_CREATE 0x1b
+#define DRM_I915_GEM_PREAD 0x1c
+#define DRM_I915_GEM_PWRITE 0x1d
+#define DRM_I915_GEM_MMAP 0x1e
+#define DRM_I915_GEM_SET_DOMAIN 0x1f
+#define DRM_I915_GEM_SW_FINISH 0x20
+#define DRM_I915_GEM_SET_TILING 0x21
+#define DRM_I915_GEM_GET_TILING 0x22

#define DRM_IOCTL_I915_INIT DRM_IOW( DRM_COMMAND_BASE + DRM_I915_INIT, drm_i915_init_t)
#define DRM_IOCTL_I915_FLUSH DRM_IO ( DRM_COMMAND_BASE + DRM_I915_FLUSH)
@@ -160,6 +176,20 @@ typedef struct _drm_i915_sarea {
#define DRM_IOCTL_I915_SET_VBLANK_PIPE DRM_IOW( DRM_COMMAND_BASE + DRM_I915_SET_VBLANK_PIPE, drm_i915_vblank_pipe_t)
#define DRM_IOCTL_I915_GET_VBLANK_PIPE DRM_IOR( DRM_COMMAND_BASE + DRM_I915_GET_VBLANK_PIPE, drm_i915_vblank_pipe_t)
#define DRM_IOCTL_I915_VBLANK_SWAP DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_VBLANK_SWAP, drm_i915_vblank_swap_t)
+#define DRM_IOCTL_I915_GEM_PIN DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_PIN, struct drm_i915_gem_pin)
+#define DRM_IOCTL_I915_GEM_UNPIN DRM_IOW(DRM_COMMAND_BASE + DRM_I915_GEM_UNPIN, struct drm_i915_gem_unpin)
+#define DRM_IOCTL_I915_GEM_BUSY DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_BUSY, struct drm_i915_gem_busy)
+#define DRM_IOCTL_I915_GEM_THROTTLE DRM_IO ( DRM_COMMAND_BASE + DRM_I915_GEM_THROTTLE)
+#define DRM_IOCTL_I915_GEM_ENTERVT DRM_IO(DRM_COMMAND_BASE + DRM_I915_GEM_ENTERVT)
+#define DRM_IOCTL_I915_GEM_LEAVEVT DRM_IO(DRM_COMMAND_BASE + DRM_I915_GEM_LEAVEVT)
+#define DRM_IOCTL_I915_GEM_CREATE DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_CREATE, struct drm_i915_gem_create)
+#define DRM_IOCTL_I915_GEM_PREAD DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_PREAD, struct drm_i915_gem_pread)
+#define DRM_IOCTL_I915_GEM_PWRITE DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_PWRITE, struct drm_i915_gem_pwrite)
+#define DRM_IOCTL_I915_GEM_MMAP DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_MMAP, struct drm_i915_gem_mmap)
+#define DRM_IOCTL_I915_GEM_SET_DOMAIN DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_SET_DOMAIN, struct drm_i915_gem_set_domain)
+#define DRM_IOCTL_I915_GEM_SW_FINISH DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_SW_FINISH, struct drm_i915_gem_sw_finish)
+#define DRM_IOCTL_I915_GEM_SET_TILING DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_GEM_SET_TILING, struct drm_i915_gem_set_tiling)
+#define DRM_IOCTL_I915_GEM_GET_TILING DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_GEM_GET_TILING, struct drm_i915_gem_get_tiling)

/* Allow drivers to submit batchbuffers directly to hardware, relying
* on the security mechanisms provided by hardware.
@@ -200,6 +230,7 @@ typedef struct drm_i915_irq_wait {
#define I915_PARAM_IRQ_ACTIVE 1
#define I915_PARAM_ALLOW_BATCHBUFFER 2
#define I915_PARAM_LAST_DISPATCH 3
+#define I915_PARAM_HAS_GEM 5

typedef struct drm_i915_getparam {
int param;
@@ -267,4 +298,305 @@ typedef struct drm_i915_hws_addr {
uint64_t addr;
} drm_i915_hws_addr_t;

+struct drm_i915_gem_init {
+ /**
+ * Beginning offset in the GTT to be managed by the DRM memory
+ * manager.
+ */
+ uint64_t gtt_start;
+ /**
+ * Ending offset in the GTT to be managed by the DRM memory
+ * manager.
+ */
+ uint64_t gtt_end;
+};
+
+struct drm_i915_gem_create {
+ /**
+ * Requested size for the object.
+ *
+ * The (page-aligned) allocated size for the object will be returned.
+ */
+ uint64_t size;
+ /**
+ * Returned handle for the object.
+ *
+ * Object handles are nonzero.
+ */
+ uint32_t handle;
+ uint32_t pad;
+};
+
+struct drm_i915_gem_pread {
+ /** Handle for the object being read. */
+ uint32_t handle;
+ uint32_t pad;
+ /** Offset into the object to read from */
+ uint64_t offset;
+ /** Length of data to read */
+ uint64_t size;
+ /**
+ * Pointer to write the data into.
+ *
+ * This is a fixed-size type for 32/64 compatibility.
+ */
+ uint64_t data_ptr;
+};
+
+struct drm_i915_gem_pwrite {
+ /** Handle for the object being written to. */
+ uint32_t handle;
+ uint32_t pad;
+ /** Offset into the object to write to */
+ uint64_t offset;
+ /** Length of data to write */
+ uint64_t size;
+ /**
+ * Pointer to read the data from.
+ *
+ * This is a fixed-size type for 32/64 compatibility.
+ */
+ uint64_t data_ptr;
+};
+
+struct drm_i915_gem_mmap {
+ /** Handle for the object being mapped. */
+ uint32_t handle;
+ uint32_t pad;
+ /** Offset in the object to map. */
+ uint64_t offset;
+ /**
+ * Length of data to map.
+ *
+ * The value will be page-aligned.
+ */
+ uint64_t size;
+ /**
+ * Returned pointer the data was mapped at.
+ *
+ * This is a fixed-size type for 32/64 compatibility.
+ */
+ uint64_t addr_ptr;
+};
+
+struct drm_i915_gem_set_domain {
+ /** Handle for the object */
+ uint32_t handle;
+
+ /** New read domains */
+ uint32_t read_domains;
+
+ /** New write domain */
+ uint32_t write_domain;
+};
+
+struct drm_i915_gem_sw_finish {
+ /** Handle for the object */
+ uint32_t handle;
+};
+
+struct drm_i915_gem_relocation_entry {
+ /**
+ * Handle of the buffer being pointed to by this relocation entry.
+ *
+ * It's appealing to make this be an index into the mm_validate_entry
+ * list to refer to the buffer, but this allows the driver to create
+ * a relocation list for state buffers and not re-write it per
+ * exec using the buffer.
+ */
+ uint32_t target_handle;
+
+ /**
+ * Value to be added to the offset of the target buffer to make up
+ * the relocation entry.
+ */
+ uint32_t delta;
+
+ /** Offset in the buffer the relocation entry will be written into */
+ uint64_t offset;
+
+ /**
+ * Offset value of the target buffer that the relocation entry was last
+ * written as.
+ *
+ * If the buffer has the same offset as last time, we can skip syncing
+ * and writing the relocation. This value is written back out by
+ * the execbuffer ioctl when the relocation is written.
+ */
+ uint64_t presumed_offset;
+
+ /**
+ * Target memory domains read by this operation.
+ */
+ uint32_t read_domains;
+
+ /**
+ * Target memory domains written by this operation.
+ *
+ * Note that only one domain may be written by the whole
+ * execbuffer operation, so that where there are conflicts,
+ * the application will get -EINVAL back.
+ */
+ uint32_t write_domain;
+};
+
+/** @{
+ * Intel memory domains
+ *
+ * Most of these just align with the various caches in
+ * the system and are used to flush and invalidate as
+ * objects end up cached in different domains.
+ */
+/** CPU cache */
+#define I915_GEM_DOMAIN_CPU 0x00000001
+/** Render cache, used by 2D and 3D drawing */
+#define I915_GEM_DOMAIN_RENDER 0x00000002
+/** Sampler cache, used by texture engine */
+#define I915_GEM_DOMAIN_SAMPLER 0x00000004
+/** Command queue, used to load batch buffers */
+#define I915_GEM_DOMAIN_COMMAND 0x00000008
+/** Instruction cache, used by shader programs */
+#define I915_GEM_DOMAIN_INSTRUCTION 0x00000010
+/** Vertex address cache */
+#define I915_GEM_DOMAIN_VERTEX 0x00000020
+/** GTT domain - aperture and scanout */
+#define I915_GEM_DOMAIN_GTT 0x00000040
+/** @} */
+
+struct drm_i915_gem_exec_object {
+ /**
+ * User's handle for a buffer to be bound into the GTT for this
+ * operation.
+ */
+ uint32_t handle;
+
+ /** Number of relocations to be performed on this buffer */
+ uint32_t relocation_count;
+ /**
+ * Pointer to array of struct drm_i915_gem_relocation_entry containing
+ * the relocations to be performed in this buffer.
+ */
+ uint64_t relocs_ptr;
+
+ /** Required alignment in graphics aperture */
+ uint64_t alignment;
+
+ /**
+ * Returned value of the updated offset of the object, for future
+ * presumed_offset writes.
+ */
+ uint64_t offset;
+};
+
+struct drm_i915_gem_execbuffer {
+ /**
+ * List of buffers to be validated with their relocations to be
+ * performend on them.
+ *
+ * This is a pointer to an array of struct drm_i915_gem_validate_entry.
+ *
+ * These buffers must be listed in an order such that all relocations
+ * a buffer is performing refer to buffers that have already appeared
+ * in the validate list.
+ */
+ uint64_t buffers_ptr;
+ uint32_t buffer_count;
+
+ /** Offset in the batchbuffer to start execution from. */
+ uint32_t batch_start_offset;
+ /** Bytes used in batchbuffer from batch_start_offset */
+ uint32_t batch_len;
+ uint32_t DR1;
+ uint32_t DR4;
+ uint32_t num_cliprects;
+ /** This is a struct drm_clip_rect *cliprects */
+ uint64_t cliprects_ptr;
+};
+
+struct drm_i915_gem_pin {
+ /** Handle of the buffer to be pinned. */
+ uint32_t handle;
+ uint32_t pad;
+
+ /** alignment required within the aperture */
+ uint64_t alignment;
+
+ /** Returned GTT offset of the buffer. */
+ uint64_t offset;
+};
+
+struct drm_i915_gem_unpin {
+ /** Handle of the buffer to be unpinned. */
+ uint32_t handle;
+ uint32_t pad;
+};
+
+struct drm_i915_gem_busy {
+ /** Handle of the buffer to check for busy */
+ uint32_t handle;
+
+ /** Return busy status (1 if busy, 0 if idle) */
+ uint32_t busy;
+};
+
+#define I915_TILING_NONE 0
+#define I915_TILING_X 1
+#define I915_TILING_Y 2
+
+#define I915_BIT_6_SWIZZLE_NONE 0
+#define I915_BIT_6_SWIZZLE_9 1
+#define I915_BIT_6_SWIZZLE_9_10 2
+#define I915_BIT_6_SWIZZLE_9_11 3
+#define I915_BIT_6_SWIZZLE_9_10_11 4
+/* Not seen by userland */
+#define I915_BIT_6_SWIZZLE_UNKNOWN 5
+
+struct drm_i915_gem_set_tiling {
+ /** Handle of the buffer to have its tiling state updated */
+ uint32_t handle;
+
+ /**
+ * Tiling mode for the object (I915_TILING_NONE, I915_TILING_X,
+ * I915_TILING_Y).
+ *
+ * This value is to be set on request, and will be updated by the
+ * kernel on successful return with the actual chosen tiling layout.
+ *
+ * The tiling mode may be demoted to I915_TILING_NONE when the system
+ * has bit 6 swizzling that can't be managed correctly by GEM.
+ *
+ * Buffer contents become undefined when changing tiling_mode.
+ */
+ uint32_t tiling_mode;
+
+ /**
+ * Stride in bytes for the object when in I915_TILING_X or
+ * I915_TILING_Y.
+ */
+ uint32_t stride;
+
+ /**
+ * Returned address bit 6 swizzling required for CPU access through
+ * mmap mapping.
+ */
+ uint32_t swizzle_mode;
+};
+
+struct drm_i915_gem_get_tiling {
+ /** Handle of the buffer to get tiling state for. */
+ uint32_t handle;
+
+ /**
+ * Current tiling mode for the object (I915_TILING_NONE, I915_TILING_X,
+ * I915_TILING_Y).
+ */
+ uint32_t tiling_mode;
+
+ /**
+ * Returned address bit 6 swizzling required for CPU access through
+ * mmap mapping.
+ */
+ uint32_t swizzle_mode;
+};
+
#endif /* _I915_DRM_H_ */
--
1.5.6.3

2008-08-01 07:38:18

by Eric Anholt

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Thu, 2008-07-31 at 23:58 -0700, Eric Anholt wrote:
> From: Keith Packard <[email protected]>
>
> GEM needs to create shmem files and get pages related to a shmem file, and
> using this pair of functions is the easiest way to do that.

Looks like I need to spend more time practicing my git-send-email
incantations. Here's the belated introduction to my patch series of

0001-PCI-Add-pci_read_base-API.patch
0002-Export-shmem_file_setup-and-shmem_getpage-for-DRM-GE.patch
0003-drm-Add-GEM-graphics-execution-manager-to-i915.patch

This patch series brings a long-awaited kernel memory manager to the i915
driver. This will allow us to do correct composited OpenGL, speed up
OpenGL-based compositing, and enable framebuffer objects and other "new"
OpenGL extensions. This patchset is also being built on to enable kernel
modesetting for a non-root, flicker-free X Server.

This patch series relies on a series of cleanups and fixes that have already
been queued by airlied. That tree can be browsed at:
http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=shortlog;h=drm-next

The main concern we expect to hear about the GEM API is the set of ioctls that
look suspiciously like file-related syscalls. We've got small integer handles
to things that we can read/write/mmap, flink and open and close, etc. Our
initial plan was to use fds, but we ran into two problems as noted in drm_gem.c:

- Process limits prevent more than 1024 or so being used at a time by
default, and we're looking at tens of thousands of these objects used by
a single client.
- Inability to allocate high fds will aggravate the X Server's select()
handling, and likely that of many GL client applications as well.

Overall, even if we cooked up a patch to make our GEM files get fds allocated
above some lower limit, interfering with clients fd namespace for these
driver-internal objects sounded like a recipe for more pain than we were ready
to sign up for. There are also concerns that for other drivers following the
rough GEM model, they'll need to supply extra hints with mmap/pread/pwrite for
acceptable performance ("access me through this aperture, please, because I
know what operation I'm going to do next"). We've already discussed a
potential mmap flag, even for the relatively simple Intel driver.

Should there be any mail-related failures on my part, the full tree on top of
2.6.27rc1, including what airlied has queued, is at:
git://people.freedesktop.org/~anholt/linux-2.6 on the drm-gem-merge branch
http://cgit.freedesktop.org/~anholt/linux-2.6/log/?h=drm-gem-merge

--
Eric Anholt
[email protected] [email protected]

Attachments:

2008-08-01 10:58:03

by Hugh Dickins

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Thu, 31 Jul 2008, Eric Anholt wrote:
> From: Keith Packard <[email protected]>
>
> GEM needs to create shmem files and get pages related to a shmem file, and
> using this pair of functions is the easiest way to do that.

It's fine to be getting the GEM end working using this easiest way,
and I don't think we'll have a problem with EXPORT_SYMBOL_GPL() on
shmem_file_setup() or something very like it - it just happens that
its current callers are not in loadable modules.

But shmem_getpage() is very much an internal helper function for
mm/shmem.c, and I'm pretty sure we should keep it and its sgp_types
that way: I'll take a look at how you're using it later on, and see
what makes sense to offer instead.

And I don't think you should assume that all GEM users must have
CONFIG_SHMEM=y: we should aim for an interface that mm/tiny_shmem.c
(redirecting to ramfs) can also support. I can believe that GEM and
TINY won't have much intersection, but let's not force that by using
shmem_getpage().

Hugh

>
> Signed-off-by: Eric Anholt <[email protected]>
> ---
> include/linux/mm.h | 11 +++++++++++
> mm/shmem.c | 17 ++++-------------
> 2 files changed, 15 insertions(+), 13 deletions(-)
>
> diff --git a/include/linux/mm.h b/include/linux/mm.h
> index 866a3db..6e54210 100644
> --- a/include/linux/mm.h
> +++ b/include/linux/mm.h
> @@ -715,6 +715,17 @@ static inline int shmem_lock(struct file *file, int lock,
> #endif
> struct file *shmem_file_setup(char *name, loff_t size, unsigned long flags);
>
> +/* Flag allocation requirements to shmem_getpage and shmem_swp_alloc */
> +enum sgp_type {
> + SGP_READ, /* don't exceed i_size, don't allocate page */
> + SGP_CACHE, /* don't exceed i_size, may allocate page */
> + SGP_DIRTY, /* like SGP_CACHE, but set new page dirty */
> + SGP_WRITE, /* may exceed i_size, may allocate page */
> +};
> +
> +int shmem_getpage(struct inode *inode, unsigned long idx,
> + struct page **pagep, enum sgp_type sgp, int *type);
> +
> int shmem_zero_setup(struct vm_area_struct *);
>
> #ifndef CONFIG_MMU
> diff --git a/mm/shmem.c b/mm/shmem.c
> index c1e5a3b..7166ff3 100644
> --- a/mm/shmem.c
> +++ b/mm/shmem.c
> @@ -77,14 +77,6 @@
> /* Pretend that each entry is of this size in directory's i_size */
> #define BOGO_DIRENT_SIZE 20
>
> -/* Flag allocation requirements to shmem_getpage and shmem_swp_alloc */
> -enum sgp_type {
> - SGP_READ, /* don't exceed i_size, don't allocate page */
> - SGP_CACHE, /* don't exceed i_size, may allocate page */
> - SGP_DIRTY, /* like SGP_CACHE, but set new page dirty */
> - SGP_WRITE, /* may exceed i_size, may allocate page */
> -};
> -
> #ifdef CONFIG_TMPFS
> static unsigned long shmem_default_max_blocks(void)
> {
> @@ -97,9 +89,6 @@ static unsigned long shmem_default_max_inodes(void)
> }
> #endif
>
> -static int shmem_getpage(struct inode *inode, unsigned long idx,
> - struct page **pagep, enum sgp_type sgp, int *type);
> -
> static inline struct page *shmem_dir_alloc(gfp_t gfp_mask)
> {
> /*
> @@ -1177,8 +1166,8 @@ static inline struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo)
> * vm. If we swap it in we mark it dirty since we also free the swap
> * entry since a page cannot live in both the swap and page cache
> */
> -static int shmem_getpage(struct inode *inode, unsigned long idx,
> - struct page **pagep, enum sgp_type sgp, int *type)
> +int shmem_getpage(struct inode *inode, unsigned long idx,
> + struct page **pagep, enum sgp_type sgp, int *type)
> {
> struct address_space *mapping = inode->i_mapping;
> struct shmem_inode_info *info = SHMEM_I(inode);
> @@ -1431,6 +1420,7 @@ failed:
> }
> return error;
> }
> +EXPORT_SYMBOL(shmem_getpage);
>
> static int shmem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
> {
> @@ -2582,6 +2572,7 @@ put_memory:
> shmem_unacct_size(flags, size);
> return ERR_PTR(error);
> }
> +EXPORT_SYMBOL(shmem_file_setup);
>
> /**
> * shmem_zero_setup - setup a shared anonymous mapping
> --
> 1.5.6.3

2008-08-01 15:49:19

by Randy Dunlap

Subject: Re: [PATCH] drm: Add GEM ("graphics execution manager") to i915 driver.

On Thu, 31 Jul 2008 23:58:39 -0700 Eric Anholt wrote:

> GEM allows the creation of persistent buffer objects accessible by the
> graphics device through new ioctls for managing execution of commands on the
> device. The userland API is almost entirely driver-specific to ensure that
> any driver building on this model can easily map the interface to individual
> driver requirements.
>
> GEM is used by the 2d driver for managing its internal state allocations and
> will be used for pixmap storage to reduce memory consumption and enable
> zero-copy GLX_EXT_texture_from_pixmap, and in the 3d driver is used to enable
> GL_EXT_framebuffer_object and GL_ARB_pixel_buffer_object.

"the 2d driver" ... "the 3d driver".
Just curious: Is there only one of each of these?


> diff --git a/drivers/gpu/drm/drm_agpsupport.c b/drivers/gpu/drm/drm_agpsupport.c
> index aefa5ac..2639be2 100644
> --- a/drivers/gpu/drm/drm_agpsupport.c
> +++ b/drivers/gpu/drm/drm_agpsupport.c
> @@ -33,6 +33,7 @@
>
> #include "drmP.h"
> #include <linux/module.h>
> +#include <asm/agp.h>
>
> #if __OS_HAS_AGP
>
> @@ -452,4 +453,52 @@ int drm_agp_unbind_memory(DRM_AGP_MEM * handle)
> return agp_unbind_memory(handle);
> }
>
> -#endif /* __OS_HAS_AGP */
> +/**

In the kernel source tree, "/**" means "beginning of kernel-doc notation",
so please don't use it when kernel-doc isn't being used.
(in multiple places/files)


> + * Binds a collection of pages into AGP memory at the given offset, returning
> + * the AGP memory structure containing them.
> + *
> + * No reference is held on the pages during this time -- it is up to the
> + * caller to handle that.
> + */
> +DRM_AGP_MEM *
> +drm_agp_bind_pages(struct drm_device *dev,
> + struct page **pages,
> + unsigned long num_pages,
> + uint32_t gtt_offset)

and the preferred function format is more like:

DRM_AGP_MEM *drm_agp_bind_page(struct drm_device *dev,
...
(many places)


> +{
> + DRM_AGP_MEM *mem;
> + int ret, i;
> +
> + DRM_DEBUG("\n");
> +
> + mem = drm_agp_allocate_memory(dev->agp->bridge, num_pages,
> + AGP_USER_MEMORY);
> + if (mem == NULL) {
> + DRM_ERROR("Failed to allocate memory for %ld pages\n",
> + num_pages);
> + return NULL;
> + }
> +
> + for (i = 0; i < num_pages; i++)
> + mem->memory[i] = phys_to_gart(page_to_phys(pages[i]));
> + mem->page_count = num_pages;
> +
> + mem->is_flushed = true;
> + ret = drm_agp_bind_memory(mem, gtt_offset / PAGE_SIZE);
> + if (ret != 0) {
> + DRM_ERROR("Failed to bind AGP memory: %d\n", ret);
> + agp_free_memory(mem);
> + return NULL;
> + }
> +
> + return mem;
> +}
> +EXPORT_SYMBOL(drm_agp_bind_pages);





> diff --git a/include/drm/drmP.h b/include/drm/drmP.h
> index 1c1b13e..a540db8 100644
> --- a/include/drm/drmP.h
> +++ b/include/drm/drmP.h
> @@ -104,6 +104,7 @@ struct drm_device;
> #define DRIVER_DMA_QUEUE 0x200
> #define DRIVER_FB_DMA 0x400
> #define DRIVER_IRQ_VBL2 0x800
> +#define DRIVER_GEM 0x1000
>
> /***********************************************************************/
> /** \name Begin the DRM... */

What uses/processes this comment notation, please?


> @@ -771,6 +838,22 @@ struct drm_device {
> spinlock_t drw_lock;
> struct idr drw_idr;
> /*@} */
> +
> + /** \name GEM information */
> + /*@{ */

and this one?

> + spinlock_t object_name_lock;
> + struct idr object_name_idr;
> + atomic_t object_count;
> + atomic_t object_memory;
> + atomic_t pin_count;
> + atomic_t pin_memory;
> + atomic_t gtt_count;
> + atomic_t gtt_memory;
> + uint32_t gtt_total;
> + uint32_t invalidate_domains; /* domains pending invalidation */
> + uint32_t flush_domains; /* domains pending flush */
> + /*@} */
> +
> };


---
~Randy
Linux Plumbers Conference, 17-19 September 2008, Portland, Oregon USA
http://linuxplumbersconf.org/

2008-08-01 18:12:11

by Keith Packard

Subject: Re: [PATCH] drm: Add GEM ("graphics execution manager") to i915 driver.

On Fri, 2008-08-01 at 08:44 -0700, Randy Dunlap wrote:

> "the 2d driver" ... "the 3d driver".
> Just curious: Is there only one of each of these?

We only need one X.org 2D and one Mesa 3D driver for each card, but DRI
supports as many rendering APIs as you like, including XvMC which offers
accelerated video decode on top of DRI as well.

So, the answer is that yes, at this point, we have only the two Intel
integrated graphics drivers using GEM, but also no, the interface can
support as many drivers as you want.

I hope that GEM will encourage more graphics API research as it allows
multiple user-space APIs to talk to the same underlying hardware
resources.

--
[email protected]

Attachments:

2008-08-01 18:18:24

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Fri, 2008-08-01 at 11:57 +0100, Hugh Dickins wrote:

> It's fine to be getting the GEM end working using this easiest way,
> and I don't think we'll have a problem with EXPORT_SYMBOL_GPL() on
> shmem_file_setup() or something very like it - it just happens that
> its current callers are not in loadable modules.

Cool. Using shmem saved a huge amount of code in DRM and allows the GEM
objects to be paged to disk if necessary. We haven't added the shrinker
API calls to finish this work, but that shouldn't take much additional
code.

> But shmem_getpage() is very much an internal helper function for
> mm/shmem.c, and I'm pretty sure we should keep it and its sgp_types
> that way: I'll take a look at how you're using it later on, and see
> what makes sense to offer instead.

We use it to get a struct page within a shmem object; we don't need the
sgp types exposed as we only ever use SGP_DIRTY. We could use a narrower
interface like:

int shmem_file_page(struct file *file, int idx, struct page **page)
{
struct inode *inode = file->f_path.dentry->d_inode;

return shmem_getpage(inode, idx, page, SGP_DIRTY, NULL);
}
EXPORT_SYMBOL(shmem_file_page);


> And I don't think you should assume that all GEM users must have
> CONFIG_SHMEM=y: we should aim for an interface that mm/tiny_shmem.c
> (redirecting to ramfs) can also support.

Seems like the simpler interface should be supportable from ramfs:

int shmem_file_page(struct file *file, int idx, struct page **page)
{
struct inode *inode = file->f_path.dentry->d_inode;
struct address_space *mapping = inode->i_mapping;
struct page *filepage;

filepage = find_lock_page(mapping, idx);
if (!filepage)
return -ENOSPC;
*page = filepage;
return 0;
}

(no, I didn't try this)

> I can believe that GEM and
> TINY won't have much intersection, but let's not force that by using
> shmem_getpage().

I don't know; I'm running without swap on several netbooks these days,
and I can easily believe that TINY would be a reasonable option there.
Thanks for the suggestions.

--
[email protected]

Attachments:

2008-08-01 20:51:02

by Christoph Hellwig

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Thu, Jul 31, 2008 at 11:58:38PM -0700, Eric Anholt wrote:
> From: Keith Packard <[email protected]>
>
> GEM needs to create shmem files and get pages related to a shmem file, and
> using this pair of functions is the easiest way to do that.

Nope. Let the userspace protion create a file in shmfs instead of
adding fugly kernel interfaces hdiing this fact.

2008-08-01 23:02:17

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Fri, 2008-08-01 at 16:50 -0400, Christoph Hellwig wrote:

> Nope. Let the userspace protion create a file in shmfs instead of
> adding fugly kernel interfaces hdiing this fact.

I can't create that many files; I need thousands of them.

--
[email protected]

Attachments:

2008-08-03 12:50:07

by John Stoffel

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

>>>>> "Keith" == Keith Packard <[email protected]> writes:

Keith> On Fri, 2008-08-01 at 16:50 -0400, Christoph Hellwig wrote:
>> Nope. Let the userspace protion create a file in shmfs instead of
>> adding fugly kernel interfaces hdiing this fact.

Keith> I can't create that many files; I need thousands of them.

Why? If you need thousands of files, won't the overhead of managing
them dynamically start to be a big load as well?

I assume (sorry for not looking at the code in depth) that you're
trying to setup specific regions of memory with various attributes for
the DRI/DRM/GEM/TTF access to Video cards?

Just seeing your statement that you wanted to add ioctls made me
shudder and try to visuallize a better way to do this.

I realize you're trying to make the drivers generic so that the *BSD
port is simple too, but... thousands of files (per X server? per
xclient) just seems like the wrong level.

Maybe you just need to open the *entire* card with one FD and then
have your own VFS like abstraction in there, which doesn't require
lots of filehandles?

Dunno... just trying to figure out what you're try to mediate here
with ioctl() and how it could be improved/simplified.


John

2008-08-03 17:52:38

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Sun, 2008-08-03 at 08:49 -0400, John Stoffel wrote:

> Why? If you need thousands of files, won't the overhead of managing
> them dynamically start to be a big load as well?

We're not having any trouble with upwards of 40000 shmem file objects at
this point. Most of these are around 4 pages in size, although they
range up to around 16MB. A typical 3D game will use around 100MB or so
of this kind of data at one time.

> I assume (sorry for not looking at the code in depth) that you're
> trying to setup specific regions of memory with various attributes for
> the DRI/DRM/GEM/TTF access to Video cards?

Not really; these are objects used in the rendering process, vertices,
command buffers, constant buffers, textures, programs, rendering
surfaces and various other state buffers. Lots of these are write-once
(from the CPU) and read-many (from the GPU), like textures, programs and
vertices. Some of these are effectively streamed from the CPU to the
GPU.

Each sequence of rendering commands requires that a set of these objects
be pinned to physical memory and bound to the graphics translation table
within the GPU.

Oh, and I allocate enough that I need to make them pageable as well;
under memory pressure, I can pull objects back out of the GTT binding
table and unpin the pages, letting shmem write them to disk.

> Just seeing your statement that you wanted to add ioctls made me
> shudder and try to visuallize a better way to do this.

I could use fds if the kernel supported applications needing many
thousand of them, and also some way to keep these FDs from polluting the
fd space used by select(2) (yeah, I know, don't use select).

> I realize you're trying to make the drivers generic so that the *BSD
> port is simple too, but... thousands of files (per X server? per
> xclient) just seems like the wrong level.

I'm not worried about BSD; I want to build the right API for Linux at
this point.

And, yes, thousands of objects per 3D application. Each X pixmap will
end up in one of these objects as well, and so something like Firefox
will likely allocate several hundred. To keep the rendering engine busy,
we'll have a couple hundred command buffers allocated as well, of 16KB
apiece.

> Maybe you just need to open the *entire* card with one FD and then
> have your own VFS like abstraction in there, which doesn't require
> lots of filehandles?

That's what I'm doing at present; the card is opened with a single fd
and then these shmem objects are allocated within an ioctl from there.
What I don't want to do is use a single linear address space for these
objects; it's just useless overhead.

> Dunno... just trying to figure out what you're try to mediate here
> with ioctl() and how it could be improved/simplified.

Suggestions welcome; I just need a few thousand multi-page allocations,
which the kernel deals with quite easily these days.

--
[email protected]

Attachments:

2008-08-03 23:37:15

by Ingo Oeser

Subject: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

Hi Keith,

On Sunday 03 August 2008, Keith Packard wrote:
> > Just seeing your statement that you wanted to add ioctls made me
> > shudder and try to visuallize a better way to do this.
>
> I could use fds if the kernel supported applications needing many
> thousand of them, and also some way to keep these FDs from polluting the
> fd space used by select(2) (yeah, I know, don't use select).

You may just have identified a scaling problem here.
I see no difference to many thousand TCP/IP connections in your description.

What actions on many thousand fds are supported poorly or not at all?
Are you only concerned about the memory requirements?

Please elaborate or point me to a place where you did already :-)


Best Regards

Ingo Oeser

2008-08-04 00:19:34

by Keith Packard

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Mon, 2008-08-04 at 01:35 +0200, Ingo Oeser wrote:

> What actions on many thousand fds are supported poorly or not at all?
> Are you only concerned about the memory requirements?

I didn't notice that the change in the maximum number of fds per process
from 1024 to 1024*1024 back in February. That makes it possible,
although requiring root privs, to allocate enough fds for this to work.

The other issue is that several important applications (including the X
server) use select instead of poll, and they have a small maximum number
of fds that they support. It seems like this could be worked around by
dup2'ing the shmem fds up a ways.

> Please elaborate or point me to a place where you did already :-)

I should have looked at fs/file.c.

--
[email protected]

Attachments:

2008-08-04 01:55:33

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Fri, 2008-08-01 at 16:50 -0400, Christoph Hellwig wrote:

> Nope. Let the userspace protion create a file in shmfs instead of
> adding fugly kernel interfaces hdiing this fact.

shmem_file_setup is already public, just not exposed to kernel modules.
I suppose we could have user space allocate the shmem file (either via
tmpfs or sysv ipc). tmpfs suffers from the maxfd issue, while sysv ipc
runs up against the SHMMAX value.

The other interface we use, shmem_getpage, doesn't seem supported
through existing interfaces though. I don't want to map the file into
any address space, I just need a pointer to the physical pages which
underlie it. I'll take those pages and hand them to the graphics engine.
shmem_getpage performs precisely the operation needed here.

--
[email protected]

Attachments:

2008-08-04 08:37:14

by Alan

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

> The other issue is that several important applications (including the X
> server) use select instead of poll, and they have a small maximum number
> of fds that they support. It seems like this could be worked around by
> dup2'ing the shmem fds up a ways.

That would work yes. Switching to poll() would probably be even smarter,
or if you have a large number of fds being scanned take a look at epoll
which is likely to be far more efficient but wouldn't be available on so
many systems - poll is at least standard.

2008-08-04 09:02:46

by Nick Piggin

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Monday 04 August 2008 11:54, Keith Packard wrote:
> On Fri, 2008-08-01 at 16:50 -0400, Christoph Hellwig wrote:
> > Nope. Let the userspace protion create a file in shmfs instead of
> > adding fugly kernel interfaces hdiing this fact.
>
> shmem_file_setup is already public, just not exposed to kernel modules.

This is not an argument. shmem_file_setup is public because that's
how ipc is implemented, not because it makes sense to be used
anywhere else. Lots of other things in mm/ are public that should
never be used outside that directory for example.


> I suppose we could have user space allocate the shmem file (either via
> tmpfs or sysv ipc). tmpfs suffers from the maxfd issue, while sysv ipc
> runs up against the SHMMAX value.

This is how I'd suggested it work as well. I think a little bit
more effort should be spent looking at making this work.


> The other interface we use, shmem_getpage, doesn't seem supported
> through existing interfaces though. I don't want to map the file into
> any address space, I just need a pointer to the physical pages which
> underlie it. I'll take those pages and hand them to the graphics engine.
> shmem_getpage performs precisely the operation needed here.

Mapping the file into an address space might be a way to make it
work (using get_user_pages to get the struct page). splice might
also work. read_mapping_page or similar could also be something to
look at. But using shmem_getpage seems wrong because it circumvents
the vfs API.

If you genuinely have problems that can't be fit into existing
APIs without significant modification, and that is specific just to
your app, then we could always look at making special cases for you.
But it would be nice if we generically solve problems you have with
processes manipulating thousands of files.

2008-08-04 10:26:55

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 2008-08-04 at 19:02 +1000, Nick Piggin wrote:

> This is how I'd suggested it work as well. I think a little bit
> more effort should be spent looking at making this work.

What I may be able to do is create a file, then hand it to my driver and
close the fd. That would avoid any ulimit or low-fd issues.

> Mapping the file into an address space might be a way to make it
> work (using get_user_pages to get the struct page). splice might
> also work. read_mapping_page or similar could also be something to
> look at. But using shmem_getpage seems wrong because it circumvents
> the vfs API.

It seems fairly ugly to map the object to user space just to get page
pointers; the expense of constructing that mapping will be entirely
wasted most of the time.

Would it be imprudent to use pagecache_write_begin/pagecache_write_end
here? For shmem, that appears to point at functions which will do what I
need. Of course, it will cause extra page-outs as each page will be
marked dirty, even if the GPU never writes them.

While shmem offers good semantics for graphics objects, it doesn't seem
like it is unique in any way, and it seems like it should be possible to
do this operation on any file system.

> If you genuinely have problems that can't be fit into existing
> APIs without significant modification, and that is specific just to
> your app, then we could always look at making special cases for you.
> But it would be nice if we generically solve problems you have with
> processes manipulating thousands of files.

There are some unique aspects to this operation which don't really have
parallels in other environments.

I'm doing memory management for a co-processor which uses the same pages
as the CPU. So, I need to allocate many pages that are just handed to
the GPU and never used by the CPU at all. Most rendering buffers are of
this form -- if you ever need to access them from the CPU, you've done
something terribly wrong.

Then there are textures which are constructed by the CPU (usually) and
handed over to the GPU for the entire lifetime of the application. These
are numerous enough that we need to be able to page them to disk; the
kernel driver will fault them back in when the GPU needs them again.

On the other hand, there are command and vertex buffers which are
constructed in user space and passed to the GPU for execution. These
operate just like any bulk-data transfer, and, in fact, I'm using the
pwrite API to transmit this data. For these buffers, the entire key is
to make sure you respect the various caches to keep them from getting
trashed.

--
[email protected]

Attachments:

2008-08-04 10:44:08

by Nick Piggin

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Monday 04 August 2008 20:26, Keith Packard wrote:
> On Mon, 2008-08-04 at 19:02 +1000, Nick Piggin wrote:
> > This is how I'd suggested it work as well. I think a little bit
> > more effort should be spent looking at making this work.
>
> What I may be able to do is create a file, then hand it to my driver and
> close the fd. That would avoid any ulimit or low-fd issues.

Sure.


> > Mapping the file into an address space might be a way to make it
> > work (using get_user_pages to get the struct page). splice might
> > also work. read_mapping_page or similar could also be something to
> > look at. But using shmem_getpage seems wrong because it circumvents
> > the vfs API.
>
> It seems fairly ugly to map the object to user space just to get page
> pointers; the expense of constructing that mapping will be entirely
> wasted most of the time.

True. It would make it possible for the userspace program to pass in
anonymous pages, but maybe not a big deal if you're using files and
shmem based management.


> Would it be imprudent to use pagecache_write_begin/pagecache_write_end
> here? For shmem, that appears to point at functions which will do what I
> need. Of course, it will cause extra page-outs as each page will be
> marked dirty, even if the GPU never writes them.
>
> While shmem offers good semantics for graphics objects, it doesn't seem
> like it is unique in any way, and it seems like it should be possible to
> do this operation on any file system.

pagecache_write_begin/pagecache_write_end should be reasonable, but you
have to be careful of the semantics of it. For example, you can't really
read anything from the page inside the calls because the filesystem may
not bring it up to date.

read_mapping_page might help there.


> > If you genuinely have problems that can't be fit into existing
> > APIs without significant modification, and that is specific just to
> > your app, then we could always look at making special cases for you.
> > But it would be nice if we generically solve problems you have with
> > processes manipulating thousands of files.
>
> There are some unique aspects to this operation which don't really have
> parallels in other environments.
>
> I'm doing memory management for a co-processor which uses the same pages
> as the CPU. So, I need to allocate many pages that are just handed to
> the GPU and never used by the CPU at all. Most rendering buffers are of
> this form -- if you ever need to access them from the CPU, you've done
> something terribly wrong.
>
> Then there are textures which are constructed by the CPU (usually) and
> handed over to the GPU for the entire lifetime of the application. These
> are numerous enough that we need to be able to page them to disk; the
> kernel driver will fault them back in when the GPU needs them again.
>
> On the other hand, there are command and vertex buffers which are
> constructed in user space and passed to the GPU for execution. These
> operate just like any bulk-data transfer, and, in fact, I'm using the
> pwrite API to transmit this data. For these buffers, the entire key is
> to make sure you respect the various caches to keep them from getting
> trashed.

Right, that's your specific implementation, but for some cases the
memory management can map or be implemented using generic primitives.
Using pagecache for your memory for example should work nicely. Making
it shmem specific and using internal APIs seems like a negative step
until you really have a good reason to.

2008-08-04 11:46:18

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 2008-08-04 at 20:43 +1000, Nick Piggin wrote:

> True. It would make it possible for the userspace program to pass in
> anonymous pages, but maybe not a big deal if you're using files and
> shmem based management.

We considered using anonymous pages, but as the user-mapping is not a
feature, it seemed like it wasn't the right model. Plus, many of these
objects need to be shared across multiple processes, so anonymous pages
would be a pain there.

> pagecache_write_begin/pagecache_write_end should be reasonable, but you
> have to be careful of the semantics of it. For example, you can't really
> read anything from the page inside the calls because the filesystem may
> not bring it up to date.

Ok, that's useful information which isn't clear from the docs.

> read_mapping_page might help there.

That does look a lot more like what I want, as it returns an unlocked
page. And, makes my code look cleaner to boot:

inode = obj->filp->f_path.dentry->d_inode;
mapping = inode->i_mapping;
for (i = 0; i < page_count; i++) {
page = read_mapping_page(mapping, i, NULL);
if (IS_ERR(page)) {
ret = PTR_ERR(page);
DRM_ERROR("read_mapping_page failed: %d\n", ret);
i915_gem_object_free_page_list(obj);
return ret;
}
obj_priv->page_list[i] = page;
}

Does this look like it conforms to the vfs api? It appears to work when
using shmem at least.

> Right, that's your specific implementation, but for some cases the
> memory management can map or be implemented using generic primitives.
> Using pagecache for your memory for example should work nicely. Making
> it shmem specific and using internal APIs seems like a negative step
> until you really have a good reason to.

Yup, I'm liking the general file mechanism, I used shmem only because
that seemed like the obvious file system you'd want underneath these
objects.

--
[email protected]

Attachments:

2008-08-04 13:51:23

by Arjan van de Ven

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Mon, 4 Aug 2008 09:19:30 +0100
Alan Cox <[email protected]> wrote:

> > The other issue is that several important applications (including
> > the X server) use select instead of poll, and they have a small
> > maximum number of fds that they support. It seems like this could
> > be worked around by dup2'ing the shmem fds up a ways.
>
> That would work yes. Switching to poll() would probably be even
> smarter, or if you have a large number of fds being scanned take a
> look at epoll which is likely to be far more efficient but wouldn't
> be available on so many systems - poll is at least standard.
> --

the hard part is that DRI is a library that gets linked into existing
applications... which already are using select ;-(

Fixing X is one thing and not impossible.. fixing all existing games in
the field is quite another


--
If you want to reach me at my work email, use [email protected]
For development, discussion and tips for power savings,
visit http://www.lesswatts.org

2008-08-04 14:29:20

by Alan

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Mon, 4 Aug 2008 06:51:01 -0700
> the hard part is that DRI is a library that gets linked into existing
> applications... which already are using select ;-(
>
> Fixing X is one thing and not impossible.. fixing all existing games in
> the field is quite another

X libraries provide the event loop so that isn't really as complex as it
seems. Shuffling the handles around for the benefit of the odd legacy app
isn't something I'd disagree with but DRI and X in the general case can
determine the event loop system used by the application.

Alan

2008-08-04 16:38:33

by Arjan van de Ven

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Mon, 4 Aug 2008 15:11:46 +0100
Alan Cox <[email protected]> wrote:

> On Mon, 4 Aug 2008 06:51:01 -0700
> > the hard part is that DRI is a library that gets linked into
> > existing applications... which already are using select ;-(
> >
> > Fixing X is one thing and not impossible.. fixing all existing
> > games in the field is quite another
>
> X libraries provide the event loop

they do?
I could have sworn things like glib etc do their own...
but I'm no desktop guy so I can totally be wrong as well

2008-08-04 16:59:17

by Keith Packard

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Mon, 2008-08-04 at 15:11 +0100, Alan Cox wrote:

> X libraries provide the event loop so that isn't really as complex as it
> seems.

Uh, no, X doesn't. Gtk+ and Qt provide event loops that applications may
use, but still many choose to roll their own.

> Shuffling the handles around for the benefit of the odd legacy app
> isn't something I'd disagree with but DRI and X in the general case can
> determine the event loop system used by the application.

Yeah, it would be nice if we could just fix all of the existing
applications. When you get that working, let me know and I'll take you
for a ride in my flying car.

--
[email protected]

Attachments:

2008-08-04 17:09:29

by Hugh Dickins

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 4 Aug 2008, Keith Packard wrote:
> On Mon, 2008-08-04 at 20:43 +1000, Nick Piggin wrote:
> > read_mapping_page might help there.
>
> That does look a lot more like what I want, as it returns an unlocked
> page. And, makes my code look cleaner to boot:
>
> inode = obj->filp->f_path.dentry->d_inode;
> mapping = inode->i_mapping;
> for (i = 0; i < page_count; i++) {
> page = read_mapping_page(mapping, i, NULL);
> if (IS_ERR(page)) {
> ret = PTR_ERR(page);
> DRM_ERROR("read_mapping_page failed: %d\n", ret);
> i915_gem_object_free_page_list(obj);
> return ret;
> }
> obj_priv->page_list[i] = page;
> }
>
> Does this look like it conforms to the vfs api? It appears to work when
> using shmem at least.

Yes, that's a good suggestion from Nick, should work with shmem/tmpfs
(with *proviso below) and others, and big improvement to your generality.

Whether such usage conforms to VFS API I'm not so sure: as I understand
it, it's really for internal use by a filesystem - if it's going to be
used beyond that, we ought to add a check that the filesystem it's used
upon really has a ->readpage method (and I'd rather we add such a check
than you do it at your end, in case we change the implementation later
to use something other than a ->readpage method - Nick, you'll be
nauseated to hear I was looking to see if ->fault with a pseudo-vma
could do it). But if the layering police are happy with this, I am.

The *proviso is that for tmpfs itself this actually isn't as convenient
as directly using shmem_getpage: because this way passes a page in to
shmem_getpage, when maybe the page wanted is already here but currently
marked as swapcache rather than filecache. It's an awkward extension
that allows tmpfs to support ->readpage at all. But that route is in
use and well-tested, and only an inefficiency when swapping, so should
not cause you any problems.

(I have been agonizing over the way __read_cache_page, like your
original i915_gem_object_get_page_list, uses find_get_page: whereas
shmem_getpage uses find_lock_page. I remember the latter is important,
but don't quite remember all the why. I'm believing that it's really
only the ->fault usage where it becomes vital: things get confused, on
2.4 more than 2.6, if a swapcache tmpfs page is mapped into userspace.)

You don't examine page->mapping at all: good, there's a race by which
it could go to NULL after you acquired the page by read_mapping_page:
not by file truncation, but by swapping out at the wrong instant.
Don't worry, you have the right page, just don't rely on page->mapping.

(Sorry if I'm rather talking aloud to myself here.)

Hugh

2008-08-04 17:26:06

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 2008-08-04 at 18:09 +0100, Hugh Dickins wrote:

> Whether such usage conforms to VFS API I'm not so sure: as I understand
> it, it's really for internal use by a filesystem

Sure, but presumably it could even be used by a layered file system?

> - if it's going to be
> used beyond that, we ought to add a check that the filesystem it's used
> upon really has a ->readpage method (and I'd rather we add such a check
> than you do it at your end, in case we change the implementation later
> to use something other than a ->readpage method - Nick, you'll be
> nauseated to hear I was looking to see if ->fault with a pseudo-vma
> could do it). But if the layering police are happy with this, I am.

It seems like I should put a check into my code that is kernel version
dependent so that I can't oops if someone tries to use a filesystem that
doesn't have ->readpage.

> But that route is in
> use and well-tested, and only an inefficiency when swapping, so should
> not cause you any problems.

Yeah, swapping performance isn't my primary concern; I looked through
the read_mapping_page codepath and it looked exactly like my existing
code in the fast path, which is why I was able to just delete all of
that from my driver and just call read_mapping_page.

So, when I release the pages from the page cache, I'm currently calling
mark_page_accessed for all pages, and set_page_dirty for pages which may
have been written by the GPU. Are those calls still needed?

--
[email protected]

Attachments:

2008-08-04 18:40:17

by Hugh Dickins

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 4 Aug 2008, Keith Packard wrote:
> On Mon, 2008-08-04 at 18:09 +0100, Hugh Dickins wrote:
>
> > Whether such usage conforms to VFS API I'm not so sure: as I understand
> > it, it's really for internal use by a filesystem
>
> Sure, but presumably it could even be used by a layered file system?

Could, yes, but should? I wouldn't presume to answer with any authority.

> > - if it's going to be
> > used beyond that, we ought to add a check that the filesystem it's used
> > upon really has a ->readpage method (and I'd rather we add such a check
> > than you do it at your end, in case we change the implementation later
> > to use something other than a ->readpage method - Nick, you'll be
> > nauseated to hear I was looking to see if ->fault with a pseudo-vma
> > could do it). But if the layering police are happy with this, I am.
>
> It seems like I should put a check into my code that is kernel version
> dependent so that I can't oops if someone tries to use a filesystem that
> doesn't have ->readpage.

Well, I guess put the check on ->readpage into your code for now, and
by the time GEM gets into Linus's tree, we should have -EINVAL checks
on NULL filler() in __read_cache_page() and read_cache_page_async(),
so remove check at your end before final submission.

(You could leave it there, and strictly we ought to update GEM if we
make any change to our implementation; but it is the kind of detail
that gets overlooked - witness the way I failed to grasp the readahead
side-effects of adding ->readpage into tmpfs until recently. I'm just
afraid we'd break you unwittingly: better not, though easily fixed.)

I'm not sending the patch right now, waiting to see if this direction
wins general favour.

> So, when I release the pages from the page cache, I'm currently calling
> mark_page_accessed for all pages, and set_page_dirty for pages which may
> have been written by the GPU. Are those calls still needed?

I think you should drop the mark_page_accessed(): that's done in
read_cache_page_async() as part of the initial read_mapping_page().
But do it again when releasing if you think there's a good chance
that object will be wanted again shortly. set_page_dirty() if
modified by GPU, yes, that would still be needed.

For how long are these objects' pages pinned in memory like this?
I ask because Rik & Lee have patches in -mm, trying to avoid long
scans of LRUs cluttered with unevictable pages. I've no idea
whether you're adding a lot or a little to that problem.

Hugh

2008-08-04 19:21:11

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 2008-08-04 at 19:39 +0100, Hugh Dickins wrote:

> Well, I guess put the check on ->readpage into your code for now, and
> by the time GEM gets into Linus's tree, we should have -EINVAL checks
> on NULL filler() in __read_cache_page() and read_cache_page_async(),
> so remove check at your end before final submission.

Sounds good to me.

> I think you should drop the mark_page_accessed(): that's done in
> read_cache_page_async() as part of the initial read_mapping_page().
> But do it again when releasing if you think there's a good chance
> that object will be wanted again shortly. set_page_dirty() if
> modified by GPU, yes, that would still be needed.

I'll just remove it then; pages pulled from the GTT are unlikely to be
used soon.

> For how long are these objects' pages pinned in memory like this?

Forever? The only reason an object would get unpinned would be on
eviction from the GTT, and objects will only be evicted if we fill the
aperture with other stuff. When we add a register_shrinker callback,
we'll also unpin pages at that point.

A busy system should have the GTT entirely full, and that will be
somewhere between 256MB and 1GB.

> I ask because Rik & Lee have patches in -mm, trying to avoid long
> scans of LRUs cluttered with unevictable pages. I've no idea
> whether you're adding a lot or a little to that problem.

If the idea of 1GB of pinned memory doesn't scare you, then I don't see
a problem ;-)

--
[email protected]

Attachments:

2008-08-04 19:56:18

by Hugh Dickins

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 4 Aug 2008, Keith Packard wrote:
> On Mon, 2008-08-04 at 19:39 +0100, Hugh Dickins wrote:
> > For how long are these objects' pages pinned in memory like this?
>
> Forever? The only reason an object would get unpinned would be on
> eviction from the GTT, and objects will only be evicted if we fill the
> aperture with other stuff. When we add a register_shrinker callback,
> we'll also unpin pages at that point.
>
> A busy system should have the GTT entirely full, and that will be
> somewhere between 256MB and 1GB.

Okay, thanks for the warning. It sounds like the shrinker will be
important, but we'll also need to mark those pages as unevictable
while they're unshrunk. (Usually when drivers grab a large number
of pages, they're not on any LRU to begin with: you're being nice
by choosing swappable LRU pages - in the tmpfs case - but enough
to upset the balance while they're not swappable.) Offhand I can't
say what will be the appropriate way to do that, it's something we
need to revisit later (from your point of view it should amount to
one function call or flag set somewhere, not any grand redesign).

Hugh

2008-08-04 21:39:34

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 2008-08-04 at 20:55 +0100, Hugh Dickins wrote:

> Okay, thanks for the warning. It sounds like the shrinker will be
> important, but we'll also need to mark those pages as unevictable
> while they're unshrunk.

That seems like a sensible optimization; otherwise the system could
spend quite a bit of time wandering over these pages.

> (Usually when drivers grab a large number
> of pages, they're not on any LRU to begin with: you're being nice
> by choosing swappable LRU pages - in the tmpfs case - but enough
> to upset the balance while they're not swappable.)

It's not a matter of 'being nice', of course, it's a matter of keeping
the system running under heavy firefox load. Check out the memory usage
by your X server with numerous tabs open to image-heavy pages someday.
Not making those pageable would result in OOM visiting far too often.

Right now, we make these images pageable by copying them in and out of a
set of pages bound permanently to the GTT. The performance implications
of doing that are fairly severe, enough so that in many cases it's
better to just use the CPU for most rendering.

> Offhand I can't
> say what will be the appropriate way to do that, it's something we
> need to revisit later (from your point of view it should amount to
> one function call or flag set somewhere, not any grand redesign).

Ok, cool. I haven't added the shrinker support yet in any case, although
that doesn't look like a big job.

--
[email protected]

Attachments:

2008-08-04 21:55:58

by Ingo Oeser

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Monday 04 August 2008, Keith Packard wrote:
> On Mon, 2008-08-04 at 15:11 +0100, Alan Cox wrote:
>
> > X libraries provide the event loop so that isn't really as complex as it
> > seems.
>
> Uh, no, X doesn't. Gtk+ and Qt provide event loops that applications may
> use, but still many choose to roll their own.

Ok, how many need support for GEM? How many of them would change their
event loop, if they can get better performance?

Maybe have a slow path for legacy behavior?

Really, the sleeping part of of event loops is usually hidden
in some libraries and the applications have a big switch statement
somewhere to dispatch the reasons for wakeup.

> > Shuffling the handles around for the benefit of the odd legacy app
> > isn't something I'd disagree with but DRI and X in the general case can
> > determine the event loop system used by the application.
>
> Yeah, it would be nice if we could just fix all of the existing
> applications.

That is never required as long as only performance suffers,
not functionality.

And the applications doing every syscall themselves are used to the pain :-)


Best Regards

Ingo Oeser

2008-08-04 21:58:51

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 2008-08-04 at 19:02 +1000, Nick Piggin wrote:

> > I suppose we could have user space allocate the shmem file (either via
> > tmpfs or sysv ipc). tmpfs suffers from the maxfd issue, while sysv ipc
> > runs up against the SHMMAX value.
>
> This is how I'd suggested it work as well. I think a little bit
> more effort should be spent looking at making this work.

Well, I've spent a day thinking about using existing user-space APIs to
get at shmem files. While it's nice that we've discovered a
filesystem-independent mechanism to pin file pages, we haven't found
anything similar for creating the files. In particular, what I want is:

1) Anonymous files backed by swap
2) Freed when the last process using them exits
3) That never appear in the file system
4) Do not consume a low FD (yeah, I know, rewrite the desktop)

So, what I could do is

char template[] = "/dev/shm/drm-XXXXXX";
int fd;
fd = mkstemp (template);
unlink (template);
ftruncate (fd, size)
object = drm_create_an_object_for_a_file (fd);
close (fd);

While I haven't written any code yet, this should work and will even be
compatible with my current user-space API. I have to create a DRM object
for the file in any case, and so I don't need to hold onto the fd.
Releasing the fd also eliminates any ulimit issues.

The drm_create_an_object_for_a_file call could return another fd. But,
note that the original shmem fd has no real value to the application in
this case.

I can imagine other cases where mapping non-shmem files would make sense
though, in particular it's fairly easy to envision mapping an image file
to the GTT and having the graphics process decode and display it without
any additional copies. I think this demonstrates the potential utility
of the general file mapping operation.

But, I'd like to have you reconsider whether it makes sense for user
space to go through the above dance to create anonymous shared objects
when the kernel already supports precisely the desired semantics and
even exposes them to the ipc/shm implementation.

We'd offer two paths in DRM -- one that used an existing file and
created an object using that as backing store, and a second one that
created anonymous objects using shmem as backing store. Transient data
would use anonymous objects while applications could directly map
arbitrary file contents as well.

--
[email protected]

Attachments:

2008-08-04 22:22:24

by Dave Airlie

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Tue, Aug 5, 2008 at 7:58 AM, Keith Packard <[email protected]> wrote:
> On Mon, 2008-08-04 at 19:02 +1000, Nick Piggin wrote:
>
>> > I suppose we could have user space allocate the shmem file (either via
>> > tmpfs or sysv ipc). tmpfs suffers from the maxfd issue, while sysv ipc
>> > runs up against the SHMMAX value.
>>
>> This is how I'd suggested it work as well. I think a little bit
>> more effort should be spent looking at making this work.
>
> Well, I've spent a day thinking about using existing user-space APIs to
> get at shmem files. While it's nice that we've discovered a
> filesystem-independent mechanism to pin file pages, we haven't found
> anything similar for creating the files. In particular, what I want is:
>
> 1) Anonymous files backed by swap
> 2) Freed when the last process using them exits
> 3) That never appear in the file system
> 4) Do not consume a low FD (yeah, I know, rewrite the desktop)
>
> So, what I could do is
>
> char template[] = "/dev/shm/drm-XXXXXX";
> int fd;
> fd = mkstemp (template);
> unlink (template);
> ftruncate (fd, size)
> object = drm_create_an_object_for_a_file (fd);
> close (fd);
>
> While I haven't written any code yet, this should work and will even be
> compatible with my current user-space API. I have to create a DRM object
> for the file in any case, and so I don't need to hold onto the fd.
> Releasing the fd also eliminates any ulimit issues.
>
> The drm_create_an_object_for_a_file call could return another fd. But,
> note that the original shmem fd has no real value to the application in
> this case.
>
> I can imagine other cases where mapping non-shmem files would make sense
> though, in particular it's fairly easy to envision mapping an image file
> to the GTT and having the graphics process decode and display it without
> any additional copies. I think this demonstrates the potential utility
> of the general file mapping operation.
>
> But, I'd like to have you reconsider whether it makes sense for user
> space to go through the above dance to create anonymous shared objects
> when the kernel already supports precisely the desired semantics and
> even exposes them to the ipc/shm implementation.
>
> We'd offer two paths in DRM -- one that used an existing file and
> created an object using that as backing store, and a second one that
> created anonymous objects using shmem as backing store. Transient data
> would use anonymous objects while applications could directly map
> arbitrary file contents as well.

We also have a need to create in-kernel objects for things like fbcon.
If we cannot
create these without doing a major dance around the vfs then it'll be
rather ugly.

Dave.

2008-08-04 22:27:39

by Dave Airlie

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Tue, Aug 5, 2008 at 7:46 AM, Ingo Oeser <[email protected]> wrote:
> On Monday 04 August 2008, Keith Packard wrote:
>> On Mon, 2008-08-04 at 15:11 +0100, Alan Cox wrote:
>>
>> > X libraries provide the event loop so that isn't really as complex as it
>> > seems.
>>
>> Uh, no, X doesn't. Gtk+ and Qt provide event loops that applications may
>> use, but still many choose to roll their own.
>
> Ok, how many need support for GEM? How many of them would change their
> event loop, if they can get better performance?
>

GEM is under OpenGL libs so they don't request support for GEM.
However if the underlying OpenGL implemenation
suddenly starts doing something with a lot of fds it could potentially
confuse the crap out of these exisiting applications.

Dave.

2008-08-05 00:34:20

by Keith Packard

Subject: Re: files/process scaling problem? (was: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM)

On Mon, 2008-08-04 at 23:46 +0200, Ingo Oeser wrote:

> Ok, how many need support for GEM? How many of them would change their
> event loop, if they can get better performance?

GEM will underlie the OpenGL implementation that applications use; we
aren't planning on writing two OpenGL implementations to work around
some file descriptor issues. And, even if we want to use fds for every
GEM object, we have a fairly simple way of moving them out of the way of
select -- dup2. Of course, it would be nice to have some way to get the
kernel to help allocate an unused fd 'up high', but

> Really, the sleeping part of of event loops is usually hidden
> in some libraries and the applications have a big switch statement
> somewhere to dispatch the reasons for wakeup.

That's not historically true in desktop applications. Yes, most modern
open source applications are sensible and use a library-based event
loop, but we can't control what applications people use.

> That is never required as long as only performance suffers,
> not functionality.

Alas, GEM offers a huge increase in functionality; performance is really
just a modest side benefit.

In reality, as I want to avoid problems caused by ulimit, I suspect I'd
end up treating most of these objects as just a bag of pages and close
the related fd after passing them to the driver, effectively turning the
whole exercise into a mechanism for passing the struct file from shmem
to GEM through user mode instead of directly across the kernel API. I'm
not sure this is a win.

--
[email protected]

Attachments:

2008-08-05 02:43:55

by John Stoffel

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

>>>>> "Hugh" == Hugh Dickins <[email protected]> writes:

Hugh> On Mon, 4 Aug 2008, Keith Packard wrote:
>> On Mon, 2008-08-04 at 19:39 +0100, Hugh Dickins wrote:
>> > For how long are these objects' pages pinned in memory like this?
>>
>> Forever? The only reason an object would get unpinned would be on
>> eviction from the GTT, and objects will only be evicted if we fill the
>> aperture with other stuff. When we add a register_shrinker callback,
>> we'll also unpin pages at that point.
>>
>> A busy system should have the GTT entirely full, and that will be
>> somewhere between 256MB and 1GB.

Hugh> Okay, thanks for the warning. It sounds like the shrinker will
Hugh> be important, but we'll also need to mark those pages as
Hugh> unevictable while they're unshrunk.

What about the onboard memory of graphics cards? Isn't that where
Textures and such are stored as well? So once something is loaded to
the card, shouldn't you be able to free it in system memory? Or swap
it out ahead of time?

Hugh> (Usually when drivers grab a large number of pages, they're not
Hugh> on any LRU to begin with: you're being nice by choosing
Hugh> swappable LRU pages - in the tmpfs case - but enough to upset
Hugh> the balance while they're not swappable.) Offhand I can't say
Hugh> what will be the appropriate way to do that, it's something we
Hugh> need to revisit later (from your point of view it should amount
Hugh> to one function call or flag set somewhere, not any grand
Hugh> redesign).

I just wonder here, since GPUs are different from other drivers in
that (I assume) they are written too much more often than they are
read to, that they should hopefully be much more amenable to drop
behind semantics for pages they've been sent, to free system
resources.

Now I admit I am most probably smoking something which is clouding my
understanding, so feel free to ingore my comments.

John

2008-08-05 04:28:55

by Nick Piggin

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Tuesday 05 August 2008 03:09, Hugh Dickins wrote:
> On Mon, 4 Aug 2008, Keith Packard wrote:
> > On Mon, 2008-08-04 at 20:43 +1000, Nick Piggin wrote:
> > > read_mapping_page might help there.
> >
> > That does look a lot more like what I want, as it returns an unlocked
> > page. And, makes my code look cleaner to boot:
> >
> > inode = obj->filp->f_path.dentry->d_inode;
> > mapping = inode->i_mapping;
> > for (i = 0; i < page_count; i++) {
> > page = read_mapping_page(mapping, i, NULL);
> > if (IS_ERR(page)) {
> > ret = PTR_ERR(page);
> > DRM_ERROR("read_mapping_page failed: %d\n", ret);
> > i915_gem_object_free_page_list(obj);
> > return ret;
> > }
> > obj_priv->page_list[i] = page;
> > }
> >
> > Does this look like it conforms to the vfs api? It appears to work when
> > using shmem at least.
>
> Yes, that's a good suggestion from Nick, should work with shmem/tmpfs
> (with *proviso below) and others, and big improvement to your generality.
>
> Whether such usage conforms to VFS API I'm not so sure: as I understand
> it, it's really for internal use by a filesystem - if it's going to be
> used beyond that, we ought to add a check that the filesystem it's used
> upon really has a ->readpage method (and I'd rather we add such a check
> than you do it at your end, in case we change the implementation later
> to use something other than a ->readpage method - Nick, you'll be
> nauseated to hear I was looking to see if ->fault with a pseudo-vma
> could do it). But if the layering police are happy with this, I am.

Yes, readpage is optional... A check will be required.

I don't exactly know if this would be deemed the best way to do it,
but it seems *relatively* filesystem agnostic (with the readpage
test being the exception), and it seems like it is to address_space
mappings what get_user_pages is to virtual memory mappings. And
there is no shortage of drivers using get_user_pages.

At any rate, it seems a bit nicer than using shmem_getpage directly.

cc'ing linux-fsdevel.


> The *proviso is that for tmpfs itself this actually isn't as convenient
> as directly using shmem_getpage: because this way passes a page in to
> shmem_getpage, when maybe the page wanted is already here but currently
> marked as swapcache rather than filecache. It's an awkward extension
> that allows tmpfs to support ->readpage at all. But that route is in
> use and well-tested, and only an inefficiency when swapping, so should
> not cause you any problems.
>
> (I have been agonizing over the way __read_cache_page, like your
> original i915_gem_object_get_page_list, uses find_get_page: whereas
> shmem_getpage uses find_lock_page. I remember the latter is important,
> but don't quite remember all the why. I'm believing that it's really
> only the ->fault usage where it becomes vital: things get confused, on
> 2.4 more than 2.6, if a swapcache tmpfs page is mapped into userspace.)
>
> You don't examine page->mapping at all: good, there's a race by which
> it could go to NULL after you acquired the page by read_mapping_page:
> not by file truncation, but by swapping out at the wrong instant.
> Don't worry, you have the right page, just don't rely on page->mapping.
>
> (Sorry if I'm rather talking aloud to myself here.)
>
> Hugh

2008-08-05 04:29:20

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Mon, 2008-08-04 at 22:25 -0400, John Stoffel wrote:

> What about the onboard memory of graphics cards? Isn't that where
> Textures and such are stored as well? So once something is loaded to
> the card, shouldn't you be able to free it in system memory? Or swap
> it out ahead of time?

Right now, I'm working only on Intel integrated graphics, which doesn't
have any on-board memory. My thinking is that we'd best solve the
easiest case first before attempting the harder discrete graphics
driver. Plus, Intel pays me do do integrated graphics, so I have an
incentive.

Not that we haven't been thinking about how this plays with a discrete
card; we'll need to allocate backing store for any objects which are
placed in on-card memory in case we run out of on-card memory, and also
as a place to store objects when the system suspends. From the
perspective of shmem, it should be precisely the same; allocating
anonymous pages and handing them to the DRM driver to store video card
data.

> I just wonder here, since GPUs are different from other drivers in
> that (I assume) they are written too much more often than they are
> read to, that they should hopefully be much more amenable to drop
> behind semantics for pages they've been sent, to free system
> resources.

Yup, I think you understand the situation fairly well. I've got a big
pile of pages sitting idle in the GTT which could be pulled out and
given back to the system. I like to keep a bunch around as the cost of
getting pages out of the CPU cache and bound to the GTT is non-trivial,
but if there's any memory pressure, the cost to free them is quite low.

--
[email protected]

Attachments:

2008-08-05 04:43:30

by Nick Piggin

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Tuesday 05 August 2008 07:58, Keith Packard wrote:
> On Mon, 2008-08-04 at 19:02 +1000, Nick Piggin wrote:
> > > I suppose we could have user space allocate the shmem file (either via
> > > tmpfs or sysv ipc). tmpfs suffers from the maxfd issue, while sysv ipc
> > > runs up against the SHMMAX value.
> >
> > This is how I'd suggested it work as well. I think a little bit
> > more effort should be spent looking at making this work.
>
> Well, I've spent a day thinking about using existing user-space APIs to
> get at shmem files. While it's nice that we've discovered a
> filesystem-independent mechanism to pin file pages, we haven't found
> anything similar for creating the files. In particular, what I want is:
>
> 1) Anonymous files backed by swap
> 2) Freed when the last process using them exits
> 3) That never appear in the file system
> 4) Do not consume a low FD (yeah, I know, rewrite the desktop)
>
> So, what I could do is
>
> char template[] = "/dev/shm/drm-XXXXXX";
> int fd;
> fd = mkstemp (template);
> unlink (template);
> ftruncate (fd, size)
> object = drm_create_an_object_for_a_file (fd);
> close (fd);
>
> While I haven't written any code yet, this should work and will even be
> compatible with my current user-space API. I have to create a DRM object
> for the file in any case, and so I don't need to hold onto the fd.
> Releasing the fd also eliminates any ulimit issues.
>
> The drm_create_an_object_for_a_file call could return another fd. But,
> note that the original shmem fd has no real value to the application in
> this case.
>
> I can imagine other cases where mapping non-shmem files would make sense
> though, in particular it's fairly easy to envision mapping an image file
> to the GTT and having the graphics process decode and display it without
> any additional copies. I think this demonstrates the potential utility
> of the general file mapping operation.
>
> But, I'd like to have you reconsider whether it makes sense for user
> space to go through the above dance to create anonymous shared objects
> when the kernel already supports precisely the desired semantics and
> even exposes them to the ipc/shm implementation.

In my opinion, doing this little song and dance (which is a few lines
of quite well defined APIs, btw) in userspace is preferable to adding
a single line or exporting a single function in kernel space. Unless
there is a better reason than eliminating a few lines of userspace code.

I'm absolutely not against exporting a nice API for a swappable, object
based memory allocator using ipc or shm to the wider kernel (with a nice
API rather than just using shmem functions directly of course). But the
fact that most or all of this seems to be able to be done in userspace
just tells me that's where it should be prototyped first. It adds
nothing to maintainence costs of the kernel code, and might actually be
helpful to show some shortcomings of our user API definition or
implementation.

In the worst case it completely fails, the effort will still show much
better how and why it needs to be done in kernel.

2008-08-05 05:20:17

by Keith Packard

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On Tue, 2008-08-05 at 14:43 +1000, Nick Piggin wrote:

> In my opinion, doing this little song and dance (which is a few lines
> of quite well defined APIs, btw) in userspace is preferable to adding
> a single line or exporting a single function in kernel space. Unless
> there is a better reason than eliminating a few lines of userspace code.

Yeah, Dave Airlie just reminded me of the real reason for creating an
API within the kernel. We need this to support kernel mode setting where
we want to provide a frame buffer console and smoothly transition to an
X desktop. Ideally, the console would start before user mode was running
so that our treasured boot messages could be displayed.

> I'm absolutely not against exporting a nice API for a swappable, object
> based memory allocator using ipc or shm to the wider kernel (with a nice
> API rather than just using shmem functions directly of course).

Yeah, we really only need a single argument.

> In the worst case it completely fails, the effort will still show much
> better how and why it needs to be done in kernel.

Sure, let's see if Dave can address the need to allocate these objects
from within the kernel. Of course, that allocation could also be done
using existing kernel APIs.

--
[email protected]

Attachments:

2008-08-06 01:13:30

by Eric Anholt

Subject: Re: [PATCH] drm: Add GEM ("graphics execution manager") to i915 driver.

On Fri, 2008-08-01 at 08:44 -0700, Randy Dunlap wrote:
> On Thu, 31 Jul 2008 23:58:39 -0700 Eric Anholt wrote:
>
> > GEM allows the creation of persistent buffer objects accessible by the
> > graphics device through new ioctls for managing execution of commands on the
> > device. The userland API is almost entirely driver-specific to ensure that
> > any driver building on this model can easily map the interface to individual
> > driver requirements.
> >
> > GEM is used by the 2d driver for managing its internal state allocations and
> > will be used for pixmap storage to reduce memory consumption and enable
> > zero-copy GLX_EXT_texture_from_pixmap, and in the 3d driver is used to enable
> > GL_EXT_framebuffer_object and GL_ARB_pixel_buffer_object.
>
> "the 2d driver" ... "the 3d driver".
> Just curious: Is there only one of each of these?

"The 2D driver" means xf86-video-intel xorg driver in userland.
"The 3D driver" means i9[16]5_dri.so mesa driver in userland.

> > diff --git a/drivers/gpu/drm/drm_agpsupport.c b/drivers/gpu/drm/drm_agpsupport.c
> > index aefa5ac..2639be2 100644
> > --- a/drivers/gpu/drm/drm_agpsupport.c
> > +++ b/drivers/gpu/drm/drm_agpsupport.c
> > @@ -33,6 +33,7 @@
> >
> > #include "drmP.h"
> > #include <linux/module.h>
> > +#include <asm/agp.h>
> >
> > #if __OS_HAS_AGP
> >
> > @@ -452,4 +453,52 @@ int drm_agp_unbind_memory(DRM_AGP_MEM * handle)
> > return agp_unbind_memory(handle);
> > }
> >
> > -#endif /* __OS_HAS_AGP */
> > +/**
>
> In the kernel source tree, "/**" means "beginning of kernel-doc notation",
> so please don't use it when kernel-doc isn't being used.
> (in multiple places/files)

The codebase this is coming from uses doxygen. I hadn't checked to see
if all the doxygen was converted to kernel-doc.

--
Eric Anholt
[email protected] [email protected]

Attachments:

2008-08-06 16:22:14

by Stephane Marchesin

Subject: Re: [PATCH] Export shmem_file_setup and shmem_getpage for DRM-GEM

On 8/5/08, Keith Packard <[email protected]> wrote:
> On Mon, 2008-08-04 at 22:25 -0400, John Stoffel wrote:
>
> > What about the onboard memory of graphics cards? Isn't that where
> > Textures and such are stored as well? So once something is loaded to
> > the card, shouldn't you be able to free it in system memory? Or swap
> > it out ahead of time?
>
>
> Right now, I'm working only on Intel integrated graphics, which doesn't
> have any on-board memory. My thinking is that we'd best solve the
> easiest case first before attempting the harder discrete graphics
> driver. Plus, Intel pays me do do integrated graphics, so I have an
> incentive.
>

Right, but it sounds adventurous to merge this work upstream before it
is generic enough to work on discrete cards. Right now it only works
on intel integrated cards ; integrated cards are one business,
discrete are a completely different world and the issues afoot
completely different.

Should this go upstream, the kernel guys have to keep in mind and
accept the possibility that another memory manager might be needed at
some point.

Stephane

2008-08-06 17:25:10

by Arjan van de Ven