drm/linux-core/drm_bufs.c

1663 lines
44 KiB
C
Raw Normal View History

2003-05-26 18:37:33 -06:00
/**
* \file drm_bufs.c
2003-05-26 18:37:33 -06:00
* Generic buffer template
*
2003-05-26 18:37:33 -06:00
* \author Rickard E. (Rik) Faith <faith@valinux.com>
* \author Gareth Hughes <gareth@valinux.com>
*/
/*
2001-02-15 01:12:14 -07:00
* Created: Thu Nov 23 03:10:50 2000 by gareth@valinux.com
*
* Copyright 1999, 2000 Precision Insight, Inc., Cedar Park, Texas.
* Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California.
* 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, 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
* VA LINUX SYSTEMS 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.
*/
#include <linux/vmalloc.h>
2001-02-15 01:12:14 -07:00
#include "drmP.h"
unsigned long drm_get_resource_start(drm_device_t *dev, unsigned int resource)
{
return pci_resource_start(dev->pdev, resource);
}
EXPORT_SYMBOL(drm_get_resource_start);
unsigned long drm_get_resource_len(drm_device_t *dev, unsigned int resource)
{
return pci_resource_len(dev->pdev, resource);
}
EXPORT_SYMBOL(drm_get_resource_len);
2005-09-29 21:14:18 -06:00
static drm_map_list_t *drm_find_matching_map(drm_device_t *dev,
drm_local_map_t *map)
{
struct list_head *list;
list_for_each(list, &dev->maplist->head) {
drm_map_list_t *entry = list_entry(list, drm_map_list_t, head);
if (entry->map && map->type == entry->map->type &&
2005-09-29 21:39:02 -06:00
((entry->map->offset == map->offset) ||
(map->type == _DRM_SHM && map->flags==_DRM_CONTAINS_LOCK))) {
2005-09-29 21:14:18 -06:00
return entry;
2004-08-24 05:15:53 -06:00
}
}
return NULL;
}
static int drm_map_handle(drm_device_t *dev, drm_hash_item_t *hash,
unsigned long user_token, int hashed_handle)
{
int use_hashed_handle;
#if (BITS_PER_LONG == 64)
use_hashed_handle = ((user_token & 0xFFFFFFFF00000000UL) || hashed_handle);
#elif (BITS_PER_LONG == 32)
use_hashed_handle = hashed_handle;
#else
#error Unsupported long size. Neither 64 nor 32 bits.
#endif
if (!use_hashed_handle) {
int ret;
hash->key = user_token;
ret = drm_ht_insert_item(&dev->map_hash, hash);
if (ret != -EINVAL)
return ret;
}
return drm_ht_just_insert_please(&dev->map_hash, hash,
user_token, 32 - PAGE_SHIFT - 3,
PAGE_SHIFT, DRM_MAP_HASH_OFFSET);
}
2003-05-26 18:37:33 -06:00
/**
* Ioctl to specify a range of memory that is available for mapping by a non-root process.
*
* \param inode device inode.
* \param filp file pointer.
* \param cmd command.
* \param arg pointer to a drm_map structure.
* \return zero on success or a negative value on error.
*
* Adjusts the memory offset to its absolute value according to the mapping
* type. Adds the map to the map list drm_device::maplist. Adds MTRR's where
* applicable and if supported by the kernel.
*/
static int drm_addmap_core(drm_device_t * dev, unsigned int offset,
2005-09-29 21:14:18 -06:00
unsigned int size, drm_map_type_t type,
drm_map_flags_t flags, drm_map_list_t ** maplist)
2001-02-15 01:12:14 -07:00
{
drm_map_t *map;
2001-03-14 15:22:50 -07:00
drm_map_list_t *list;
drm_dma_handle_t *dmah;
unsigned long user_token;
int ret;
2001-02-15 01:12:14 -07:00
map = drm_alloc(sizeof(*map), DRM_MEM_MAPS);
if (!map)
2001-02-15 01:12:14 -07:00
return -ENOMEM;
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
map->offset = offset;
map->size = size;
map->flags = flags;
map->type = type;
2001-02-15 01:12:14 -07:00
2001-03-14 15:22:50 -07:00
/* Only allow shared memory to be removable since we only keep enough
* book keeping information about shared memory to allow for removal
* when processes fork.
*/
if ((map->flags & _DRM_REMOVABLE) && map->type != _DRM_SHM) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2001-03-14 15:22:50 -07:00
return -EINVAL;
}
DRM_DEBUG("offset = 0x%08lx, size = 0x%08lx, type = %d\n",
map->offset, map->size, map->type);
if ((map->offset & (~PAGE_MASK)) || (map->size & (~PAGE_MASK))) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2001-02-15 01:12:14 -07:00
return -EINVAL;
}
map->mtrr = -1;
2004-07-15 07:03:55 -06:00
map->handle = NULL;
2001-02-15 01:12:14 -07:00
switch (map->type) {
2001-02-15 01:12:14 -07:00
case _DRM_REGISTERS:
case _DRM_FRAME_BUFFER:
2005-08-20 01:38:11 -06:00
#if !defined(__sparc__) && !defined(__alpha__) && !defined(__ia64__) && !defined(__powerpc64__) && !defined(__x86_64__)
if (map->offset + (map->size - 1) < map->offset ||
map->offset < virt_to_phys(high_memory)) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
return -EINVAL;
}
2001-02-15 01:12:14 -07:00
#endif
2001-05-01 11:07:59 -06:00
#ifdef __alpha__
map->offset += dev->hose->mem_space->start;
2001-05-01 11:07:59 -06:00
#endif
/* Some drivers preinitialize some maps, without the X Server
* needing to be aware of it. Therefore, we just return success
* when the server tries to create a duplicate map.
*/
2005-09-29 21:14:18 -06:00
list = drm_find_matching_map(dev, map);
if (list != NULL) {
if (list->map->size != map->size) {
DRM_DEBUG("Matching maps of type %d with "
2005-09-29 21:14:18 -06:00
"mismatched sizes, (%ld vs %ld)\n",
map->type, map->size,
list->map->size);
list->map->size = map->size;
2004-08-24 05:15:53 -06:00
}
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2005-09-29 21:14:18 -06:00
*maplist = list;
return 0;
2001-02-15 01:12:14 -07:00
}
if (drm_core_has_MTRR(dev)) {
if (map->type == _DRM_FRAME_BUFFER ||
(map->flags & _DRM_WRITE_COMBINING)) {
map->mtrr = mtrr_add(map->offset, map->size,
MTRR_TYPE_WRCOMB, 1);
}
}
2005-09-29 21:39:02 -06:00
if (map->type == _DRM_REGISTERS)
map->handle = drm_ioremap(map->offset, map->size, dev);
break;
2001-02-15 01:12:14 -07:00
case _DRM_SHM:
2005-09-29 21:14:18 -06:00
list = drm_find_matching_map(dev, map);
if (list != NULL) {
if(list->map->size != map->size) {
2005-08-04 08:48:43 -06:00
DRM_DEBUG("Matching maps of type %d with "
"mismatched sizes, (%ld vs %ld)\n",
2005-09-29 21:14:18 -06:00
map->type, map->size, list->map->size);
list->map->size = map->size;
2005-08-04 08:48:43 -06:00
}
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2005-09-29 21:14:18 -06:00
*maplist = list;
2005-08-04 08:48:43 -06:00
return 0;
}
2001-03-14 15:22:50 -07:00
map->handle = vmalloc_32(map->size);
DRM_DEBUG("%lu %d %p\n",
map->size, drm_order(map->size), map->handle);
if (!map->handle) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2001-02-15 01:12:14 -07:00
return -ENOMEM;
}
map->offset = (unsigned long)map->handle;
if (map->flags & _DRM_CONTAINS_LOCK) {
/* Prevent a 2nd X Server from creating a 2nd lock */
if (dev->lock.hw_lock != NULL) {
vfree(map->handle);
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
return -EBUSY;
}
dev->sigdata.lock = dev->lock.hw_lock = map->handle; /* Pointer to lock */
2001-02-15 01:12:14 -07:00
}
break;
case _DRM_AGP: {
drm_agp_mem_t *entry;
int valid = 0;
if (!drm_core_has_AGP(dev)) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
return -EINVAL;
}
2001-05-01 11:07:59 -06:00
#ifdef __alpha__
map->offset += dev->hose->mem_space->start;
2001-05-01 11:07:59 -06:00
#endif
/* Note: dev->agp->base may actually be 0 when the DRM
* is not in control of AGP space. But if user space is
* it should already have added the AGP base itself.
*/
map->offset += dev->agp->base;
map->mtrr = dev->agp->agp_mtrr; /* for getmap */
/* This assumes the DRM is in total control of AGP space.
* It's not always the case as AGP can be in the control
* of user space (i.e. i810 driver). So this loop will get
* skipped and we double check that dev->agp->memory is
* actually set as well as being invalid before EPERM'ing
*/
for (entry = dev->agp->memory; entry; entry = entry->next) {
if ((map->offset >= entry->bound) &&
(map->offset + map->size <= entry->bound + entry->pages * PAGE_SIZE)) {
valid = 1;
break;
}
}
if (dev->agp->memory && !valid) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
return -EPERM;
2004-08-24 05:15:53 -06:00
}
DRM_DEBUG("AGP offset = 0x%08lx, size = 0x%08lx\n", map->offset, map->size);
2001-02-15 01:12:14 -07:00
break;
}
2001-05-01 11:07:59 -06:00
case _DRM_SCATTER_GATHER:
if (!dev->sg) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2001-05-01 11:07:59 -06:00
return -EINVAL;
}
map->offset += (unsigned long)dev->sg->virtual;
2001-05-01 11:07:59 -06:00
break;
case _DRM_CONSISTENT:
/* dma_addr_t is 64bit on i386 with CONFIG_HIGHMEM64G,
* As we're limiting the address to 2^32-1 (or less),
* casting it down to 32 bits is no problem, but we
* need to point to a 64bit variable first. */
dmah = drm_pci_alloc(dev, map->size, map->size, 0xffffffffUL);
if (!dmah) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
return -ENOMEM;
}
map->handle = dmah->vaddr;
map->offset = (unsigned long)dmah->busaddr;
kfree(dmah);
break;
2001-02-15 01:12:14 -07:00
default:
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2001-02-15 01:12:14 -07:00
return -EINVAL;
}
list = drm_alloc(sizeof(*list), DRM_MEM_MAPS);
if (!list) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2001-03-14 15:22:50 -07:00
return -EINVAL;
2001-02-15 01:12:14 -07:00
}
2001-03-14 15:22:50 -07:00
memset(list, 0, sizeof(*list));
list->map = map;
mutex_lock(&dev->struct_mutex);
2001-03-14 15:22:50 -07:00
list_add(&list->head, &dev->maplist->head);
/* Assign a 32-bit handle */
user_token = (map->type == _DRM_SHM) ? (unsigned long) map->handle :
map->offset;
ret = drm_map_handle(dev, &list->hash, user_token, 0);
if (ret) {
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
drm_free(list, sizeof(*list), DRM_MEM_MAPS);
mutex_unlock(&dev->struct_mutex);
return ret;
}
list->user_token = list->hash.key;
mutex_unlock(&dev->struct_mutex);
2005-09-29 21:14:18 -06:00
*maplist = list;
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
return 0;
}
2005-09-29 21:14:18 -06:00
int drm_addmap(drm_device_t * dev, unsigned int offset,
unsigned int size, drm_map_type_t type,
drm_map_flags_t flags, drm_local_map_t ** map_ptr)
{
drm_map_list_t *list;
int rc;
rc = drm_addmap_core(dev, offset, size, type, flags, &list);
if (!rc)
*map_ptr = list->map;
return rc;
}
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
EXPORT_SYMBOL(drm_addmap);
int drm_addmap_ioctl(struct inode *inode, struct file *filp,
unsigned int cmd, unsigned long arg)
{
drm_file_t *priv = filp->private_data;
drm_device_t *dev = priv->head->dev;
drm_map_t map;
2005-09-29 21:14:18 -06:00
drm_map_list_t *maplist;
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
drm_map_t __user *argp = (void __user *)arg;
int err;
if (!(filp->f_mode & 3))
return -EACCES; /* Require read/write */
if (copy_from_user(&map, argp, sizeof(map))) {
2001-02-15 01:12:14 -07:00
return -EFAULT;
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
}
if (!(capable(CAP_SYS_ADMIN) || map.type == _DRM_AGP))
return -EPERM;
err = drm_addmap_core(dev, map.offset, map.size, map.type, map.flags,
&maplist);
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
2005-09-29 21:14:18 -06:00
if (err)
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
return err;
2005-09-29 21:14:18 -06:00
if (copy_to_user(argp, maplist->map, sizeof(drm_map_t)))
return -EFAULT;
/* avoid a warning on 64-bit, this casting isn't very nice, but the API is set so too late */
if (put_user((void *)(unsigned long)maplist->user_token, &argp->handle))
return -EFAULT;
2001-02-15 01:12:14 -07:00
return 0;
}
2003-05-26 18:37:33 -06:00
/**
* Remove a map private from list and deallocate resources if the mapping
2001-03-14 15:22:50 -07:00
* isn't in use.
2003-05-26 18:37:33 -06:00
*
* \param inode device inode.
* \param filp file pointer.
* \param cmd command.
* \param arg pointer to a drm_map_t structure.
* \return zero on success or a negative value on error.
*
* Searches the map on drm_device::maplist, removes it from the list, see if
* its being used, and free any associate resource (such as MTRR's) if it's not
* being on use.
*
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
* \sa drm_addmap
2001-03-14 15:22:50 -07:00
*/
int drm_rmmap_locked(drm_device_t *dev, drm_local_map_t *map)
2001-03-14 15:22:50 -07:00
{
struct list_head *list;
drm_map_list_t *r_list = NULL;
drm_dma_handle_t dmah;
/* Find the list entry for the map and remove it */
2001-03-14 15:22:50 -07:00
list_for_each(list, &dev->maplist->head) {
r_list = list_entry(list, drm_map_list_t, head);
2001-03-14 15:22:50 -07:00
if (r_list->map == map) {
list_del(list);
drm_ht_remove_key(&dev->map_hash, r_list->user_token);
drm_free(list, sizeof(*list), DRM_MEM_MAPS);
break;
}
2001-03-14 15:22:50 -07:00
}
/* List has wrapped around to the head pointer, or it's empty and we
* didn't find anything.
2001-03-14 15:22:50 -07:00
*/
if (list == (&dev->maplist->head)) {
2001-03-14 15:22:50 -07:00
return -EINVAL;
}
switch (map->type) {
case _DRM_REGISTERS:
drm_ioremapfree(map->handle, map->size, dev);
/* FALLTHROUGH */
case _DRM_FRAME_BUFFER:
if (drm_core_has_MTRR(dev) && map->mtrr >= 0) {
int retcode;
retcode = mtrr_del(map->mtrr, map->offset, map->size);
DRM_DEBUG("mtrr_del=%d\n", retcode);
}
break;
case _DRM_SHM:
vfree(map->handle);
break;
case _DRM_AGP:
case _DRM_SCATTER_GATHER:
break;
case _DRM_CONSISTENT:
dmah.vaddr = map->handle;
dmah.busaddr = map->offset;
dmah.size = map->size;
__drm_pci_free(dev, &dmah);
break;
}
drm_free(map, sizeof(*map), DRM_MEM_MAPS);
2001-03-14 15:22:50 -07:00
return 0;
}
EXPORT_SYMBOL(drm_rmmap_locked);
2001-03-14 15:22:50 -07:00
int drm_rmmap(drm_device_t *dev, drm_local_map_t *map)
{
int ret;
mutex_lock(&dev->struct_mutex);
ret = drm_rmmap_locked(dev, map);
mutex_unlock(&dev->struct_mutex);
return ret;
2001-03-14 15:22:50 -07:00
}
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
EXPORT_SYMBOL(drm_rmmap);
/* The rmmap ioctl appears to be unnecessary. All mappings are torn down on
* the last close of the device, and this is necessary for cleanup when things
* exit uncleanly. Therefore, having userland manually remove mappings seems
* like a pointless exercise since they're going away anyway.
*
* One use case might be after addmap is allowed for normal users for SHM and
* gets used by drivers that the server doesn't need to care about. This seems
* unlikely.
*/
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
int drm_rmmap_ioctl(struct inode *inode, struct file *filp,
unsigned int cmd, unsigned long arg)
{
drm_file_t *priv = filp->private_data;
drm_device_t *dev = priv->head->dev;
drm_map_t request;
2005-06-29 07:00:29 -06:00
drm_local_map_t *map = NULL;
struct list_head *list;
int ret;
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
if (copy_from_user(&request, (drm_map_t __user *) arg, sizeof(request))) {
return -EFAULT;
}
mutex_lock(&dev->struct_mutex);
list_for_each(list, &dev->maplist->head) {
drm_map_list_t *r_list = list_entry(list, drm_map_list_t, head);
if (r_list->map &&
r_list->user_token == (unsigned long)request.handle &&
r_list->map->flags & _DRM_REMOVABLE) {
map = r_list->map;
break;
}
}
/* List has wrapped around to the head pointer, or its empty we didn't
* find anything.
*/
if (list == (&dev->maplist->head)) {
mutex_unlock(&dev->struct_mutex);
return -EINVAL;
}
2006-07-05 08:39:22 -06:00
if (!map) {
mutex_unlock(&dev->struct_mutex);
2005-06-29 07:00:29 -06:00
return -EINVAL;
2006-07-05 08:39:22 -06:00
}
2005-06-29 07:00:29 -06:00
/* Register and framebuffer maps are permanent */
if ((map->type == _DRM_REGISTERS) || (map->type == _DRM_FRAME_BUFFER)) {
mutex_unlock(&dev->struct_mutex);
return 0;
}
ret = drm_rmmap_locked(dev, map);
mutex_unlock(&dev->struct_mutex);
return ret;
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
}
2001-03-14 15:22:50 -07:00
2003-05-26 18:37:33 -06:00
/**
* Cleanup after an error on one of the addbufs() functions.
*
* \param dev DRM device.
2003-05-26 18:37:33 -06:00
* \param entry buffer entry where the error occurred.
*
* Frees any pages and buffers associated with the given entry.
*/
static void drm_cleanup_buf_error(drm_device_t * dev, drm_buf_entry_t * entry)
{
int i;
if (entry->seg_count) {
for (i = 0; i < entry->seg_count; i++) {
if (entry->seglist[i]) {
drm_pci_free(dev, entry->seglist[i]);
}
}
drm_free(entry->seglist,
entry->seg_count *
sizeof(*entry->seglist), DRM_MEM_SEGS);
entry->seg_count = 0;
}
if (entry->buf_count) {
for (i = 0; i < entry->buf_count; i++) {
if (entry->buflist[i].dev_private) {
drm_free(entry->buflist[i].dev_private,
entry->buflist[i].dev_priv_size,
DRM_MEM_BUFS);
}
}
drm_free(entry->buflist,
entry->buf_count *
sizeof(*entry->buflist), DRM_MEM_BUFS);
entry->buf_count = 0;
}
}
2004-08-24 05:15:53 -06:00
#if __OS_HAS_AGP
2003-05-26 18:37:33 -06:00
/**
* Add AGP buffers for DMA transfers
2003-05-26 18:37:33 -06:00
*
* \param dev drm_device_t to which the buffers are to be added.
* \param request pointer to a drm_buf_desc_t describing the request.
2003-05-26 18:37:33 -06:00
* \return zero on success or a negative number on failure.
*
2003-05-26 18:37:33 -06:00
* After some sanity checks creates a drm_buf structure for each buffer and
* reallocates the buffer list of the same size order to accommodate the new
* buffers.
*/
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
int drm_addbufs_agp(drm_device_t * dev, drm_buf_desc_t * request)
2001-02-15 01:12:14 -07:00
{
drm_device_dma_t *dma = dev->dma;
drm_buf_entry_t *entry;
drm_agp_mem_t *agp_entry;
2001-02-15 01:12:14 -07:00
drm_buf_t *buf;
unsigned long offset;
unsigned long agp_offset;
int count;
int order;
int size;
int alignment;
int page_order;
int total;
int byte_count;
int i, valid;
drm_buf_t **temp_buflist;
if (!dma)
return -EINVAL;
2001-02-15 01:12:14 -07:00
count = request->count;
order = drm_order(request->size);
2001-02-15 01:12:14 -07:00
size = 1 << order;
alignment = (request->flags & _DRM_PAGE_ALIGN)
? PAGE_ALIGN(size) : size;
2001-02-15 01:12:14 -07:00
page_order = order - PAGE_SHIFT > 0 ? order - PAGE_SHIFT : 0;
total = PAGE_SIZE << page_order;
byte_count = 0;
agp_offset = dev->agp->base + request->agp_start;
2001-02-15 01:12:14 -07:00
DRM_DEBUG("count: %d\n", count);
DRM_DEBUG("order: %d\n", order);
DRM_DEBUG("size: %d\n", size);
DRM_DEBUG("agp_offset: %lx\n", agp_offset);
DRM_DEBUG("alignment: %d\n", alignment);
DRM_DEBUG("page_order: %d\n", page_order);
DRM_DEBUG("total: %d\n", total);
2001-02-15 01:12:14 -07:00
if (order < DRM_MIN_ORDER || order > DRM_MAX_ORDER)
return -EINVAL;
if (dev->queue_count)
return -EBUSY; /* Not while in use */
2001-02-15 01:12:14 -07:00
/* Make sure buffers are located in AGP memory that we own */
valid = 0;
for (agp_entry = dev->agp->memory; agp_entry; agp_entry = agp_entry->next) {
if ((agp_offset >= agp_entry->bound) &&
(agp_offset + total * count <= agp_entry->bound + agp_entry->pages * PAGE_SIZE)) {
valid = 1;
break;
}
}
if (dev->agp->memory && !valid) {
DRM_DEBUG("zone invalid\n");
return -EINVAL;
}
spin_lock(&dev->count_lock);
if (dev->buf_use) {
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
return -EBUSY;
}
atomic_inc(&dev->buf_alloc);
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
mutex_lock(&dev->struct_mutex);
2001-02-15 01:12:14 -07:00
entry = &dma->bufs[order];
if (entry->buf_count) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM; /* May only call once for each order */
2001-02-15 01:12:14 -07:00
}
if (count < 0 || count > 4096) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -EINVAL;
}
entry->buflist = drm_alloc(count * sizeof(*entry->buflist),
DRM_MEM_BUFS);
if (!entry->buflist) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
2001-02-15 01:12:14 -07:00
return -ENOMEM;
}
memset(entry->buflist, 0, count * sizeof(*entry->buflist));
2001-02-15 01:12:14 -07:00
entry->buf_size = size;
entry->page_order = page_order;
offset = 0;
while (entry->buf_count < count) {
buf = &entry->buflist[entry->buf_count];
buf->idx = dma->buf_count + entry->buf_count;
buf->total = alignment;
buf->order = order;
buf->used = 0;
2001-02-15 01:12:14 -07:00
buf->offset = (dma->byte_count + offset);
2001-02-15 01:12:14 -07:00
buf->bus_address = agp_offset + offset;
buf->address = (void *)(agp_offset + offset);
buf->next = NULL;
2001-02-15 01:12:14 -07:00
buf->waiting = 0;
buf->pending = 0;
init_waitqueue_head(&buf->dma_wait);
buf->filp = NULL;
2001-02-15 01:12:14 -07:00
buf->dev_priv_size = dev->driver->dev_priv_size;
buf->dev_private = drm_alloc(buf->dev_priv_size, DRM_MEM_BUFS);
if (!buf->dev_private) {
/* Set count correctly so we free the proper amount. */
entry->buf_count = count;
drm_cleanup_buf_error(dev, entry);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
memset(buf->dev_private, 0, buf->dev_priv_size);
2001-02-15 01:12:14 -07:00
DRM_DEBUG("buffer %d @ %p\n", entry->buf_count, buf->address);
2001-02-15 01:12:14 -07:00
offset += alignment;
entry->buf_count++;
byte_count += PAGE_SIZE << page_order;
}
DRM_DEBUG("byte_count: %d\n", byte_count);
2001-02-15 01:12:14 -07:00
temp_buflist = drm_realloc(dma->buflist,
dma->buf_count * sizeof(*dma->buflist),
(dma->buf_count + entry->buf_count)
* sizeof(*dma->buflist), DRM_MEM_BUFS);
if (!temp_buflist) {
/* Free the entry because it isn't valid */
drm_cleanup_buf_error(dev, entry);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
dma->buflist = temp_buflist;
for (i = 0; i < entry->buf_count; i++) {
2001-02-15 01:12:14 -07:00
dma->buflist[i + dma->buf_count] = &entry->buflist[i];
}
dma->buf_count += entry->buf_count;
dma->seg_count += entry->seg_count;
dma->page_count += byte_count >> PAGE_SHIFT;
2001-02-15 01:12:14 -07:00
dma->byte_count += byte_count;
DRM_DEBUG("dma->buf_count : %d\n", dma->buf_count);
DRM_DEBUG("entry->buf_count : %d\n", entry->buf_count);
2001-02-15 01:12:14 -07:00
mutex_unlock(&dev->struct_mutex);
2001-02-15 01:12:14 -07:00
request->count = entry->buf_count;
request->size = size;
2001-02-15 01:12:14 -07:00
dma->flags = _DRM_DMA_USE_AGP;
atomic_dec(&dev->buf_alloc);
2001-02-15 01:12:14 -07:00
return 0;
}
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
EXPORT_SYMBOL(drm_addbufs_agp);
#endif /* __OS_HAS_AGP */
2001-02-15 01:12:14 -07:00
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
int drm_addbufs_pci(drm_device_t * dev, drm_buf_desc_t * request)
2001-02-15 01:12:14 -07:00
{
drm_device_dma_t *dma = dev->dma;
int count;
int order;
int size;
int total;
int page_order;
drm_buf_entry_t *entry;
drm_dma_handle_t *dmah;
2001-02-15 01:12:14 -07:00
drm_buf_t *buf;
int alignment;
unsigned long offset;
int i;
int byte_count;
int page_count;
unsigned long *temp_pagelist;
drm_buf_t **temp_buflist;
2001-02-15 01:12:14 -07:00
if (!drm_core_check_feature(dev, DRIVER_PCI_DMA))
return -EINVAL;
2004-08-24 05:15:53 -06:00
if (!dma)
return -EINVAL;
2001-02-15 01:12:14 -07:00
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
count = request->count;
order = drm_order(request->size);
2001-02-15 01:12:14 -07:00
size = 1 << order;
DRM_DEBUG("count=%d, size=%d (%d), order=%d, queue_count=%d\n",
request->count, request->size, size, order, dev->queue_count);
2001-02-15 01:12:14 -07:00
if (order < DRM_MIN_ORDER || order > DRM_MAX_ORDER)
return -EINVAL;
if (dev->queue_count)
return -EBUSY; /* Not while in use */
2001-02-15 01:12:14 -07:00
alignment = (request->flags & _DRM_PAGE_ALIGN)
? PAGE_ALIGN(size) : size;
2001-02-15 01:12:14 -07:00
page_order = order - PAGE_SHIFT > 0 ? order - PAGE_SHIFT : 0;
total = PAGE_SIZE << page_order;
spin_lock(&dev->count_lock);
if (dev->buf_use) {
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
return -EBUSY;
}
atomic_inc(&dev->buf_alloc);
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
mutex_lock(&dev->struct_mutex);
2001-02-15 01:12:14 -07:00
entry = &dma->bufs[order];
if (entry->buf_count) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
2001-02-15 01:12:14 -07:00
return -ENOMEM; /* May only call once for each order */
}
if (count < 0 || count > 4096) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -EINVAL;
}
entry->buflist = drm_alloc(count * sizeof(*entry->buflist),
DRM_MEM_BUFS);
if (!entry->buflist) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
2001-02-15 01:12:14 -07:00
return -ENOMEM;
}
memset(entry->buflist, 0, count * sizeof(*entry->buflist));
entry->seglist = drm_alloc(count * sizeof(*entry->seglist),
DRM_MEM_SEGS);
if (!entry->seglist) {
drm_free(entry->buflist,
count * sizeof(*entry->buflist), DRM_MEM_BUFS);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
2001-02-15 01:12:14 -07:00
return -ENOMEM;
}
memset(entry->seglist, 0, count * sizeof(*entry->seglist));
2001-02-15 01:12:14 -07:00
/* Keep the original pagelist until we know all the allocations
* have succeeded
*/
temp_pagelist = drm_alloc((dma->page_count + (count << page_order))
* sizeof(*dma->pagelist), DRM_MEM_PAGES);
if (!temp_pagelist) {
drm_free(entry->buflist,
count * sizeof(*entry->buflist), DRM_MEM_BUFS);
drm_free(entry->seglist,
count * sizeof(*entry->seglist), DRM_MEM_SEGS);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
memcpy(temp_pagelist,
dma->pagelist, dma->page_count * sizeof(*dma->pagelist));
DRM_DEBUG("pagelist: %d entries\n",
dma->page_count + (count << page_order));
2001-02-15 01:12:14 -07:00
entry->buf_size = size;
2001-02-15 01:12:14 -07:00
entry->page_order = page_order;
byte_count = 0;
page_count = 0;
while (entry->buf_count < count) {
dmah = drm_pci_alloc(dev, PAGE_SIZE << page_order, 0x1000, 0xfffffffful);
if (!dmah) {
/* Set count correctly so we free the proper amount. */
entry->buf_count = count;
entry->seg_count = count;
drm_cleanup_buf_error(dev, entry);
drm_free(temp_pagelist,
(dma->page_count + (count << page_order))
* sizeof(*dma->pagelist), DRM_MEM_PAGES);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
entry->seglist[entry->seg_count++] = dmah;
for (i = 0; i < (1 << page_order); i++) {
DRM_DEBUG("page %d @ 0x%08lx\n",
dma->page_count + page_count,
(unsigned long)dmah->vaddr + PAGE_SIZE * i);
temp_pagelist[dma->page_count + page_count++]
= (unsigned long)dmah->vaddr + PAGE_SIZE * i;
2001-02-15 01:12:14 -07:00
}
for (offset = 0;
offset + size <= total && entry->buf_count < count;
offset += alignment, ++entry->buf_count) {
buf = &entry->buflist[entry->buf_count];
buf->idx = dma->buf_count + entry->buf_count;
buf->total = alignment;
buf->order = order;
buf->used = 0;
buf->offset = (dma->byte_count + byte_count + offset);
buf->address = (void *)(dmah->vaddr + offset);
buf->bus_address = dmah->busaddr + offset;
buf->next = NULL;
2001-02-15 01:12:14 -07:00
buf->waiting = 0;
buf->pending = 0;
init_waitqueue_head(&buf->dma_wait);
buf->filp = NULL;
buf->dev_priv_size = dev->driver->dev_priv_size;
buf->dev_private = drm_alloc(buf->dev_priv_size,
DRM_MEM_BUFS);
if (!buf->dev_private) {
/* Set count correctly so we free the proper amount. */
entry->buf_count = count;
entry->seg_count = count;
drm_cleanup_buf_error(dev, entry);
drm_free(temp_pagelist,
(dma->page_count +
(count << page_order))
* sizeof(*dma->pagelist),
DRM_MEM_PAGES);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
memset(buf->dev_private, 0, buf->dev_priv_size);
DRM_DEBUG("buffer %d @ %p\n",
entry->buf_count, buf->address);
2001-02-15 01:12:14 -07:00
}
byte_count += PAGE_SIZE << page_order;
}
temp_buflist = drm_realloc(dma->buflist,
dma->buf_count * sizeof(*dma->buflist),
(dma->buf_count + entry->buf_count)
* sizeof(*dma->buflist), DRM_MEM_BUFS);
if (!temp_buflist) {
/* Free the entry because it isn't valid */
drm_cleanup_buf_error(dev, entry);
drm_free(temp_pagelist,
(dma->page_count + (count << page_order))
* sizeof(*dma->pagelist), DRM_MEM_PAGES);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
dma->buflist = temp_buflist;
for (i = 0; i < entry->buf_count; i++) {
2001-02-15 01:12:14 -07:00
dma->buflist[i + dma->buf_count] = &entry->buflist[i];
}
/* No allocations failed, so now we can replace the orginal pagelist
* with the new one.
*/
if (dma->page_count) {
drm_free(dma->pagelist,
dma->page_count * sizeof(*dma->pagelist),
DRM_MEM_PAGES);
}
dma->pagelist = temp_pagelist;
2001-02-15 01:12:14 -07:00
dma->buf_count += entry->buf_count;
dma->seg_count += entry->seg_count;
dma->page_count += entry->seg_count << page_order;
dma->byte_count += PAGE_SIZE * (entry->seg_count << page_order);
mutex_unlock(&dev->struct_mutex);
2001-02-15 01:12:14 -07:00
request->count = entry->buf_count;
request->size = size;
2001-02-15 01:12:14 -07:00
if (request->flags & _DRM_PCI_BUFFER_RO)
dma->flags = _DRM_DMA_USE_PCI_RO;
atomic_dec(&dev->buf_alloc);
2001-02-15 01:12:14 -07:00
return 0;
2001-02-15 01:12:14 -07:00
}
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
EXPORT_SYMBOL(drm_addbufs_pci);
2001-02-15 01:12:14 -07:00
2005-06-04 00:18:11 -06:00
static int drm_addbufs_sg(drm_device_t * dev, drm_buf_desc_t * request)
2001-05-01 11:07:59 -06:00
{
drm_device_dma_t *dma = dev->dma;
drm_buf_entry_t *entry;
drm_buf_t *buf;
unsigned long offset;
unsigned long agp_offset;
int count;
int order;
int size;
int alignment;
int page_order;
int total;
int byte_count;
int i;
drm_buf_t **temp_buflist;
if (!drm_core_check_feature(dev, DRIVER_SG))
return -EINVAL;
if (!dma)
return -EINVAL;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
count = request->count;
order = drm_order(request->size);
size = 1 << order;
alignment = (request->flags & _DRM_PAGE_ALIGN)
? PAGE_ALIGN(size) : size;
page_order = order - PAGE_SHIFT > 0 ? order - PAGE_SHIFT : 0;
total = PAGE_SIZE << page_order;
byte_count = 0;
agp_offset = request->agp_start;
DRM_DEBUG("count: %d\n", count);
DRM_DEBUG("order: %d\n", order);
DRM_DEBUG("size: %d\n", size);
DRM_DEBUG("agp_offset: %lu\n", agp_offset);
DRM_DEBUG("alignment: %d\n", alignment);
DRM_DEBUG("page_order: %d\n", page_order);
DRM_DEBUG("total: %d\n", total);
if (order < DRM_MIN_ORDER || order > DRM_MAX_ORDER)
return -EINVAL;
if (dev->queue_count)
return -EBUSY; /* Not while in use */
spin_lock(&dev->count_lock);
if (dev->buf_use) {
spin_unlock(&dev->count_lock);
return -EBUSY;
}
atomic_inc(&dev->buf_alloc);
spin_unlock(&dev->count_lock);
mutex_lock(&dev->struct_mutex);
entry = &dma->bufs[order];
if (entry->buf_count) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM; /* May only call once for each order */
}
2001-05-01 11:07:59 -06:00
if (count < 0 || count > 4096) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -EINVAL;
}
entry->buflist = drm_alloc(count * sizeof(*entry->buflist),
DRM_MEM_BUFS);
if (!entry->buflist) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
memset(entry->buflist, 0, count * sizeof(*entry->buflist));
2001-05-01 11:07:59 -06:00
entry->buf_size = size;
entry->page_order = page_order;
offset = 0;
while (entry->buf_count < count) {
buf = &entry->buflist[entry->buf_count];
buf->idx = dma->buf_count + entry->buf_count;
buf->total = alignment;
buf->order = order;
buf->used = 0;
buf->offset = (dma->byte_count + offset);
buf->bus_address = agp_offset + offset;
buf->address = (void *)(agp_offset + offset
+ (unsigned long)dev->sg->virtual);
buf->next = NULL;
buf->waiting = 0;
buf->pending = 0;
init_waitqueue_head(&buf->dma_wait);
buf->filp = NULL;
buf->dev_priv_size = dev->driver->dev_priv_size;
buf->dev_private = drm_alloc(buf->dev_priv_size, DRM_MEM_BUFS);
if (!buf->dev_private) {
/* Set count correctly so we free the proper amount. */
entry->buf_count = count;
drm_cleanup_buf_error(dev, entry);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
2001-05-01 11:07:59 -06:00
memset(buf->dev_private, 0, buf->dev_priv_size);
2001-05-01 11:07:59 -06:00
DRM_DEBUG("buffer %d @ %p\n", entry->buf_count, buf->address);
offset += alignment;
entry->buf_count++;
byte_count += PAGE_SIZE << page_order;
}
DRM_DEBUG("byte_count: %d\n", byte_count);
temp_buflist = drm_realloc(dma->buflist,
dma->buf_count * sizeof(*dma->buflist),
(dma->buf_count + entry->buf_count)
* sizeof(*dma->buflist), DRM_MEM_BUFS);
if (!temp_buflist) {
/* Free the entry because it isn't valid */
drm_cleanup_buf_error(dev, entry);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
dma->buflist = temp_buflist;
2001-05-01 11:07:59 -06:00
for (i = 0; i < entry->buf_count; i++) {
dma->buflist[i + dma->buf_count] = &entry->buflist[i];
}
2001-05-01 11:07:59 -06:00
dma->buf_count += entry->buf_count;
dma->seg_count += entry->seg_count;
dma->page_count += byte_count >> PAGE_SHIFT;
dma->byte_count += byte_count;
2001-05-01 11:07:59 -06:00
DRM_DEBUG("dma->buf_count : %d\n", dma->buf_count);
DRM_DEBUG("entry->buf_count : %d\n", entry->buf_count);
2001-05-01 11:07:59 -06:00
mutex_unlock(&dev->struct_mutex);
2001-05-01 11:07:59 -06:00
request->count = entry->buf_count;
request->size = size;
2001-05-01 11:07:59 -06:00
dma->flags = _DRM_DMA_USE_SG;
2001-05-01 11:07:59 -06:00
atomic_dec(&dev->buf_alloc);
return 0;
2001-05-01 11:07:59 -06:00
}
int drm_addbufs_fb(drm_device_t * dev, drm_buf_desc_t * request)
{
drm_device_dma_t *dma = dev->dma;
drm_buf_entry_t *entry;
drm_buf_t *buf;
unsigned long offset;
unsigned long agp_offset;
int count;
int order;
int size;
int alignment;
int page_order;
int total;
int byte_count;
int i;
drm_buf_t **temp_buflist;
if (!drm_core_check_feature(dev, DRIVER_FB_DMA))
return -EINVAL;
if (!dma)
return -EINVAL;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
count = request->count;
order = drm_order(request->size);
size = 1 << order;
alignment = (request->flags & _DRM_PAGE_ALIGN)
? PAGE_ALIGN(size) : size;
page_order = order - PAGE_SHIFT > 0 ? order - PAGE_SHIFT : 0;
total = PAGE_SIZE << page_order;
byte_count = 0;
agp_offset = request->agp_start;
DRM_DEBUG("count: %d\n", count);
DRM_DEBUG("order: %d\n", order);
DRM_DEBUG("size: %d\n", size);
DRM_DEBUG("agp_offset: %lu\n", agp_offset);
DRM_DEBUG("alignment: %d\n", alignment);
DRM_DEBUG("page_order: %d\n", page_order);
DRM_DEBUG("total: %d\n", total);
if (order < DRM_MIN_ORDER || order > DRM_MAX_ORDER)
return -EINVAL;
if (dev->queue_count)
return -EBUSY; /* Not while in use */
spin_lock(&dev->count_lock);
if (dev->buf_use) {
spin_unlock(&dev->count_lock);
return -EBUSY;
}
atomic_inc(&dev->buf_alloc);
spin_unlock(&dev->count_lock);
mutex_lock(&dev->struct_mutex);
entry = &dma->bufs[order];
if (entry->buf_count) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM; /* May only call once for each order */
}
if (count < 0 || count > 4096) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -EINVAL;
}
entry->buflist = drm_alloc(count * sizeof(*entry->buflist),
DRM_MEM_BUFS);
if (!entry->buflist) {
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
memset(entry->buflist, 0, count * sizeof(*entry->buflist));
entry->buf_size = size;
entry->page_order = page_order;
offset = 0;
while (entry->buf_count < count) {
buf = &entry->buflist[entry->buf_count];
buf->idx = dma->buf_count + entry->buf_count;
buf->total = alignment;
buf->order = order;
buf->used = 0;
buf->offset = (dma->byte_count + offset);
buf->bus_address = agp_offset + offset;
buf->address = (void *)(agp_offset + offset);
buf->next = NULL;
buf->waiting = 0;
buf->pending = 0;
init_waitqueue_head(&buf->dma_wait);
buf->filp = NULL;
buf->dev_priv_size = dev->driver->dev_priv_size;
buf->dev_private = drm_alloc(buf->dev_priv_size, DRM_MEM_BUFS);
if (!buf->dev_private) {
/* Set count correctly so we free the proper amount. */
entry->buf_count = count;
drm_cleanup_buf_error(dev, entry);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
memset(buf->dev_private, 0, buf->dev_priv_size);
DRM_DEBUG("buffer %d @ %p\n", entry->buf_count, buf->address);
offset += alignment;
entry->buf_count++;
byte_count += PAGE_SIZE << page_order;
}
DRM_DEBUG("byte_count: %d\n", byte_count);
temp_buflist = drm_realloc(dma->buflist,
dma->buf_count * sizeof(*dma->buflist),
(dma->buf_count + entry->buf_count)
* sizeof(*dma->buflist), DRM_MEM_BUFS);
if (!temp_buflist) {
/* Free the entry because it isn't valid */
drm_cleanup_buf_error(dev, entry);
mutex_unlock(&dev->struct_mutex);
atomic_dec(&dev->buf_alloc);
return -ENOMEM;
}
dma->buflist = temp_buflist;
for (i = 0; i < entry->buf_count; i++) {
dma->buflist[i + dma->buf_count] = &entry->buflist[i];
}
dma->buf_count += entry->buf_count;
dma->seg_count += entry->seg_count;
dma->page_count += byte_count >> PAGE_SHIFT;
dma->byte_count += byte_count;
DRM_DEBUG("dma->buf_count : %d\n", dma->buf_count);
DRM_DEBUG("entry->buf_count : %d\n", entry->buf_count);
mutex_unlock(&dev->struct_mutex);
request->count = entry->buf_count;
request->size = size;
dma->flags = _DRM_DMA_USE_FB;
atomic_dec(&dev->buf_alloc);
return 0;
}
Adds support for PCI cards to MGA DRM This patch adds serveral new ioctls and a new query to get_param query to support PCI MGA cards. Two ioctls were added to implement interrupt based waiting. With this change, the client-side driver no longer needs to map the primary DMA region or the MMIO region. Previously, end-of-frame waiting was done by busy waiting in the client-side driver until one of the MMIO registers (the current DMA pointer) matched a pointer to the end of primary DMA space. By using interrupts, the busy waiting and the extra mappings are removed. A third ioctl was added to bootstrap DMA. This ioctl, which is used by the X-server, moves a *LOT* of code from the X-server into the kernel. This allows the kernel to do whatever needs to be done to setup DMA buffers. The entire process and the locations of the buffers are hidden from user-mode. Additionally, a get_param query was added to differentiate between G4x0 cards and G550 cards. A gap was left in the numbering sequence so that, if needed, G450 cards could be distinguished from G400 cards. According to Ville Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering differently. This seems the most compatible way to let the client-side driver know which card it's own. Doing this very small change now eliminates the need to bump the DRM minor version twice. http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2 A number of ioctl handlers in linux-core were also modified so that they could be called in-kernel. In these cases, the in-kernel callable version kept the existing name (e.g., drm_agp_acquire) and the ioctl handler added _ioctl to the name (e.g., drm_agp_acquire_ioctl). This patch also replaces the drm_agp_do_release function with drm_agp_release. drm_agp_release (drm_core_agp_release in the previous patch) is very similar to drm_agp_do_release, and I saw no reason to have both. This commit *breaks the build* on BSD. Eric said that he would make the required updates to the BSD side soon. Xorg bug: 3259 Reviewed by: Eric Anholt
2005-06-14 16:34:11 -06:00
EXPORT_SYMBOL(drm_addbufs_fb);
2003-05-26 18:37:33 -06:00
/**
* Add buffers for DMA transfers (ioctl).
*
* \param inode device inode.
* \param filp file pointer.
* \param cmd command.
* \param arg pointer to a drm_buf_desc_t request.
* \return zero on success or a negative number on failure.
*
* According with the memory type specified in drm_buf_desc::flags and the
* build options, it dispatches the call either to addbufs_agp(),
* addbufs_sg() or addbufs_pci() for AGP, scatter-gather or consistent
* PCI memory respectively.
*/
int drm_addbufs(struct inode *inode, struct file *filp,
unsigned int cmd, unsigned long arg)
2001-02-15 01:12:14 -07:00
{
drm_buf_desc_t request;
2004-08-24 05:15:53 -06:00
drm_file_t *priv = filp->private_data;
drm_device_t *dev = priv->head->dev;
int ret;
if (!drm_core_check_feature(dev, DRIVER_HAVE_DMA))
2004-08-24 05:15:53 -06:00
return -EINVAL;
2001-02-15 01:12:14 -07:00
if (copy_from_user(&request, (drm_buf_desc_t __user *) arg,
sizeof(request)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
2004-08-24 05:15:53 -06:00
#if __OS_HAS_AGP
if (request.flags & _DRM_AGP_BUFFER)
ret = drm_addbufs_agp(dev, &request);
2001-02-15 01:12:14 -07:00
else
#endif
if (request.flags & _DRM_SG_BUFFER)
ret = drm_addbufs_sg(dev, &request);
else if (request.flags & _DRM_FB_BUFFER)
ret = drm_addbufs_fb(dev, &request);
else
ret = drm_addbufs_pci(dev, &request);
if (ret == 0) {
if (copy_to_user((void __user *) arg, &request,
sizeof(request))) {
ret = -EFAULT;
}
}
return ret;
2001-02-15 01:12:14 -07:00
}
2003-05-26 18:37:33 -06:00
/**
* Get information about the buffer mappings.
*
* This was originally mean for debugging purposes, or by a sophisticated
* client library to determine how best to use the available buffers (e.g.,
* large buffers can be used for image transfer).
*
* \param inode device inode.
* \param filp file pointer.
* \param cmd command.
* \param arg pointer to a drm_buf_info structure.
* \return zero on success or a negative number on failure.
*
* Increments drm_device::buf_use while holding the drm_device::count_lock
* lock, preventing of allocating more buffers after this call. Information
* about each requested buffer is then copied into user space.
*/
int drm_infobufs(struct inode *inode, struct file *filp,
unsigned int cmd, unsigned long arg)
2001-02-15 01:12:14 -07:00
{
drm_file_t *priv = filp->private_data;
drm_device_t *dev = priv->head->dev;
2001-02-15 01:12:14 -07:00
drm_device_dma_t *dma = dev->dma;
drm_buf_info_t request;
drm_buf_info_t __user *argp = (void __user *)arg;
2001-02-15 01:12:14 -07:00
int i;
int count;
if (!drm_core_check_feature(dev, DRIVER_HAVE_DMA))
2004-08-24 05:15:53 -06:00
return -EINVAL;
if (!dma)
return -EINVAL;
2001-02-15 01:12:14 -07:00
spin_lock(&dev->count_lock);
if (atomic_read(&dev->buf_alloc)) {
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
return -EBUSY;
}
++dev->buf_use; /* Can't allocate more after this call */
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
if (copy_from_user(&request, argp, sizeof(request)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
for (i = 0, count = 0; i < DRM_MAX_ORDER + 1; i++) {
if (dma->bufs[i].buf_count)
++count;
2001-02-15 01:12:14 -07:00
}
DRM_DEBUG("count = %d\n", count);
2001-02-15 01:12:14 -07:00
if (request.count >= count) {
for (i = 0, count = 0; i < DRM_MAX_ORDER + 1; i++) {
if (dma->bufs[i].buf_count) {
drm_buf_desc_t __user *to =
&request.list[count];
2001-02-15 01:12:14 -07:00
drm_buf_entry_t *from = &dma->bufs[i];
drm_freelist_t *list = &dma->bufs[i].freelist;
if (copy_to_user(&to->count,
&from->buf_count,
sizeof(from->buf_count)) ||
copy_to_user(&to->size,
&from->buf_size,
sizeof(from->buf_size)) ||
copy_to_user(&to->low_mark,
&list->low_mark,
sizeof(list->low_mark)) ||
copy_to_user(&to->high_mark,
&list->high_mark,
sizeof(list->high_mark)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
DRM_DEBUG("%d %d %d %d %d\n",
i,
dma->bufs[i].buf_count,
dma->bufs[i].buf_size,
dma->bufs[i].freelist.low_mark,
dma->bufs[i].freelist.high_mark);
2001-02-15 01:12:14 -07:00
++count;
}
}
}
request.count = count;
if (copy_to_user(argp, &request, sizeof(request)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
return 0;
}
2003-05-26 18:37:33 -06:00
/**
* Specifies a low and high water mark for buffer allocation
*
* \param inode device inode.
* \param filp file pointer.
* \param cmd command.
* \param arg a pointer to a drm_buf_desc structure.
* \return zero on success or a negative number on failure.
*
* Verifies that the size order is bounded between the admissible orders and
* updates the respective drm_device_dma::bufs entry low and high water mark.
*
* \note This ioctl is deprecated and mostly never used.
*/
int drm_markbufs(struct inode *inode, struct file *filp,
unsigned int cmd, unsigned long arg)
2001-02-15 01:12:14 -07:00
{
drm_file_t *priv = filp->private_data;
drm_device_t *dev = priv->head->dev;
2001-02-15 01:12:14 -07:00
drm_device_dma_t *dma = dev->dma;
drm_buf_desc_t request;
int order;
drm_buf_entry_t *entry;
if (!drm_core_check_feature(dev, DRIVER_HAVE_DMA))
2004-08-24 05:15:53 -06:00
return -EINVAL;
if (!dma)
return -EINVAL;
2001-02-15 01:12:14 -07:00
if (copy_from_user(&request,
(drm_buf_desc_t __user *) arg, sizeof(request)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
DRM_DEBUG("%d, %d, %d\n",
request.size, request.low_mark, request.high_mark);
order = drm_order(request.size);
if (order < DRM_MIN_ORDER || order > DRM_MAX_ORDER)
return -EINVAL;
2001-02-15 01:12:14 -07:00
entry = &dma->bufs[order];
if (request.low_mark < 0 || request.low_mark > entry->buf_count)
2001-02-15 01:12:14 -07:00
return -EINVAL;
if (request.high_mark < 0 || request.high_mark > entry->buf_count)
2001-02-15 01:12:14 -07:00
return -EINVAL;
entry->freelist.low_mark = request.low_mark;
2001-02-15 01:12:14 -07:00
entry->freelist.high_mark = request.high_mark;
return 0;
}
2003-05-26 18:37:33 -06:00
/**
* Unreserve the buffers in list, previously reserved using drmDMA.
2003-05-26 18:37:33 -06:00
*
* \param inode device inode.
* \param filp file pointer.
* \param cmd command.
* \param arg pointer to a drm_buf_free structure.
* \return zero on success or a negative number on failure.
*
2003-05-26 18:37:33 -06:00
* Calls free_buffer() for each used buffer.
* This function is primarily used for debugging.
*/
int drm_freebufs(struct inode *inode, struct file *filp,
unsigned int cmd, unsigned long arg)
2001-02-15 01:12:14 -07:00
{
drm_file_t *priv = filp->private_data;
drm_device_t *dev = priv->head->dev;
2001-02-15 01:12:14 -07:00
drm_device_dma_t *dma = dev->dma;
drm_buf_free_t request;
int i;
int idx;
drm_buf_t *buf;
if (!drm_core_check_feature(dev, DRIVER_HAVE_DMA))
2004-08-24 05:15:53 -06:00
return -EINVAL;
if (!dma)
return -EINVAL;
2001-02-15 01:12:14 -07:00
if (copy_from_user(&request,
(drm_buf_free_t __user *) arg, sizeof(request)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
DRM_DEBUG("%d\n", request.count);
for (i = 0; i < request.count; i++) {
if (copy_from_user(&idx, &request.list[i], sizeof(idx)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
if (idx < 0 || idx >= dma->buf_count) {
DRM_ERROR("Index %d (of %d max)\n",
idx, dma->buf_count - 1);
2001-02-15 01:12:14 -07:00
return -EINVAL;
}
buf = dma->buflist[idx];
if (buf->filp != filp) {
DRM_ERROR("Process %d freeing buffer not owned\n",
current->pid);
2001-02-15 01:12:14 -07:00
return -EINVAL;
}
drm_free_buffer(dev, buf);
2001-02-15 01:12:14 -07:00
}
return 0;
}
2003-05-26 18:37:33 -06:00
/**
* Maps all of the DMA buffers into client-virtual space (ioctl).
*
* \param inode device inode.
* \param filp file pointer.
* \param cmd command.
* \param arg pointer to a drm_buf_map structure.
* \return zero on success or a negative number on failure.
*
* Maps the AGP, SG or PCI buffer region with do_mmap(), and copies information
* about each buffer into user space. For PCI buffers, it calls do_mmap() with
* offset equal to 0, which drm_mmap() interpretes as PCI buffers and calls
* drm_mmap_dma().
2003-05-26 18:37:33 -06:00
*/
int drm_mapbufs(struct inode *inode, struct file *filp,
unsigned int cmd, unsigned long arg)
2001-02-15 01:12:14 -07:00
{
drm_file_t *priv = filp->private_data;
drm_device_t *dev = priv->head->dev;
2001-02-15 01:12:14 -07:00
drm_device_dma_t *dma = dev->dma;
drm_buf_map_t __user *argp = (void __user *)arg;
2001-02-15 01:12:14 -07:00
int retcode = 0;
const int zero = 0;
unsigned long virtual;
unsigned long address;
drm_buf_map_t request;
int i;
if (!drm_core_check_feature(dev, DRIVER_HAVE_DMA))
2004-08-24 05:15:53 -06:00
return -EINVAL;
if (!dma)
return -EINVAL;
2001-02-15 01:12:14 -07:00
spin_lock(&dev->count_lock);
if (atomic_read(&dev->buf_alloc)) {
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
return -EBUSY;
}
dev->buf_use++; /* Can't allocate more after this call */
spin_unlock(&dev->count_lock);
2001-02-15 01:12:14 -07:00
if (copy_from_user(&request, argp, sizeof(request)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
if (request.count >= dma->buf_count) {
if ((drm_core_has_AGP(dev) && (dma->flags & _DRM_DMA_USE_AGP))
|| (drm_core_check_feature(dev, DRIVER_SG)
&& (dma->flags & _DRM_DMA_USE_SG))
|| (drm_core_check_feature(dev, DRIVER_FB_DMA)
&& (dma->flags & _DRM_DMA_USE_FB))) {
2004-08-17 07:10:05 -06:00
drm_map_t *map = dev->agp_buffer_map;
unsigned long token = dev->agp_buffer_token;
2001-02-15 01:12:14 -07:00
if (!map) {
2001-02-15 01:12:14 -07:00
retcode = -EINVAL;
goto done;
}
down_write(&current->mm->mmap_sem);
virtual = do_mmap(filp, 0, map->size,
PROT_READ | PROT_WRITE,
MAP_SHARED,
token);
up_write(&current->mm->mmap_sem);
2001-02-15 01:12:14 -07:00
} else {
down_write(&current->mm->mmap_sem);
virtual = do_mmap(filp, 0, dma->byte_count,
PROT_READ | PROT_WRITE,
MAP_SHARED, 0);
up_write(&current->mm->mmap_sem);
2001-02-15 01:12:14 -07:00
}
if (virtual > -1024UL) {
2001-02-15 01:12:14 -07:00
/* Real error */
retcode = (signed long)virtual;
goto done;
}
request.virtual = (void __user *)virtual;
2001-02-15 01:12:14 -07:00
for (i = 0; i < dma->buf_count; i++) {
if (copy_to_user(&request.list[i].idx,
&dma->buflist[i]->idx,
sizeof(request.list[0].idx))) {
2001-02-15 01:12:14 -07:00
retcode = -EFAULT;
goto done;
}
if (copy_to_user(&request.list[i].total,
&dma->buflist[i]->total,
sizeof(request.list[0].total))) {
2001-02-15 01:12:14 -07:00
retcode = -EFAULT;
goto done;
}
if (copy_to_user(&request.list[i].used,
&zero, sizeof(zero))) {
2001-02-15 01:12:14 -07:00
retcode = -EFAULT;
goto done;
}
address = virtual + dma->buflist[i]->offset; /* *** */
if (copy_to_user(&request.list[i].address,
&address, sizeof(address))) {
2001-02-15 01:12:14 -07:00
retcode = -EFAULT;
goto done;
}
}
}
done:
2001-02-15 01:12:14 -07:00
request.count = dma->buf_count;
DRM_DEBUG("%d buffers, retcode = %d\n", request.count, retcode);
2001-02-15 01:12:14 -07:00
if (copy_to_user(argp, &request, sizeof(request)))
2001-02-15 01:12:14 -07:00
return -EFAULT;
return retcode;
}
/**
* Compute size order. Returns the exponent of the smaller power of two which
* is greater or equal to given number.
*
* \param size size.
* \return order.
*
* \todo Can be made faster.
*/
int drm_order(unsigned long size)
{
int order;
unsigned long tmp;
for (order = 0, tmp = size >> 1; tmp; tmp >>= 1, order++) ;
if (size & (size - 1))
++order;
return order;
}
EXPORT_SYMBOL(drm_order);
2005-06-28 07:02:20 -06:00