drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
/**
|
|
|
|
* \file drm_ioc32.c
|
|
|
|
*
|
|
|
|
* 32-bit ioctl compatibility routines for the DRM.
|
|
|
|
*
|
|
|
|
* \author Paul Mackerras <paulus@samba.org>
|
|
|
|
*
|
|
|
|
* Copyright (C) Paul Mackerras 2005.
|
|
|
|
* 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
|
|
|
|
* THE AUTHOR 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/compat.h>
|
|
|
|
|
|
|
|
#include "drmP.h"
|
|
|
|
#include "drm_core.h"
|
|
|
|
|
|
|
|
#define DRM_IOCTL_VERSION32 DRM_IOWR(0x00, drm_version32_t)
|
|
|
|
#define DRM_IOCTL_GET_UNIQUE32 DRM_IOWR(0x01, drm_unique32_t)
|
|
|
|
#define DRM_IOCTL_GET_MAP32 DRM_IOWR(0x04, drm_map32_t)
|
|
|
|
#define DRM_IOCTL_GET_CLIENT32 DRM_IOWR(0x05, drm_client32_t)
|
|
|
|
#define DRM_IOCTL_GET_STATS32 DRM_IOR( 0x06, drm_stats32_t)
|
|
|
|
|
|
|
|
#define DRM_IOCTL_SET_UNIQUE32 DRM_IOW( 0x10, drm_unique32_t)
|
|
|
|
#define DRM_IOCTL_ADD_MAP32 DRM_IOWR(0x15, drm_map32_t)
|
|
|
|
#define DRM_IOCTL_ADD_BUFS32 DRM_IOWR(0x16, drm_buf_desc32_t)
|
|
|
|
#define DRM_IOCTL_MARK_BUFS32 DRM_IOW( 0x17, drm_buf_desc32_t)
|
|
|
|
#define DRM_IOCTL_INFO_BUFS32 DRM_IOWR(0x18, drm_buf_info32_t)
|
|
|
|
#define DRM_IOCTL_MAP_BUFS32 DRM_IOWR(0x19, drm_buf_map32_t)
|
|
|
|
#define DRM_IOCTL_FREE_BUFS32 DRM_IOW( 0x1a, drm_buf_free32_t)
|
|
|
|
|
|
|
|
#define DRM_IOCTL_RM_MAP32 DRM_IOW( 0x1b, drm_map32_t)
|
|
|
|
|
|
|
|
#define DRM_IOCTL_SET_SAREA_CTX32 DRM_IOW( 0x1c, drm_ctx_priv_map32_t)
|
|
|
|
#define DRM_IOCTL_GET_SAREA_CTX32 DRM_IOWR(0x1d, drm_ctx_priv_map32_t)
|
|
|
|
|
|
|
|
#define DRM_IOCTL_RES_CTX32 DRM_IOWR(0x26, drm_ctx_res32_t)
|
|
|
|
#define DRM_IOCTL_DMA32 DRM_IOWR(0x29, drm_dma32_t)
|
|
|
|
|
|
|
|
#define DRM_IOCTL_AGP_ENABLE32 DRM_IOW( 0x32, drm_agp_mode32_t)
|
|
|
|
#define DRM_IOCTL_AGP_INFO32 DRM_IOR( 0x33, drm_agp_info32_t)
|
|
|
|
#define DRM_IOCTL_AGP_ALLOC32 DRM_IOWR(0x34, drm_agp_buffer32_t)
|
|
|
|
#define DRM_IOCTL_AGP_FREE32 DRM_IOW( 0x35, drm_agp_buffer32_t)
|
|
|
|
#define DRM_IOCTL_AGP_BIND32 DRM_IOW( 0x36, drm_agp_binding32_t)
|
|
|
|
#define DRM_IOCTL_AGP_UNBIND32 DRM_IOW( 0x37, drm_agp_binding32_t)
|
|
|
|
|
|
|
|
#define DRM_IOCTL_SG_ALLOC32 DRM_IOW( 0x38, drm_scatter_gather32_t)
|
|
|
|
#define DRM_IOCTL_SG_FREE32 DRM_IOW( 0x39, drm_scatter_gather32_t)
|
|
|
|
|
|
|
|
#define DRM_IOCTL_WAIT_VBLANK32 DRM_IOWR(0x3a, drm_wait_vblank32_t)
|
|
|
|
|
|
|
|
typedef struct drm_version_32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int version_major; /**< Major version */
|
|
|
|
int version_minor; /**< Minor version */
|
|
|
|
int version_patchlevel; /**< Patch level */
|
|
|
|
u32 name_len; /**< Length of name buffer */
|
|
|
|
u32 name; /**< Name of driver */
|
|
|
|
u32 date_len; /**< Length of date buffer */
|
|
|
|
u32 date; /**< User-space buffer to hold date */
|
|
|
|
u32 desc_len; /**< Length of desc buffer */
|
|
|
|
u32 desc; /**< User-space buffer to hold desc */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_version32_t;
|
|
|
|
|
|
|
|
static int compat_drm_version(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_version32_t v32;
|
|
|
|
drm_version_t __user *version;
|
|
|
|
int err;
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
if (copy_from_user(&v32, (void __user *)arg, sizeof(v32)))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
version = compat_alloc_user_space(sizeof(*version));
|
|
|
|
if (!access_ok(VERIFY_WRITE, version, sizeof(*version)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(v32.name_len, &version->name_len)
|
|
|
|
|| __put_user((void __user *)(unsigned long)v32.name,
|
|
|
|
&version->name)
|
|
|
|
|| __put_user(v32.date_len, &version->date_len)
|
|
|
|
|| __put_user((void __user *)(unsigned long)v32.date,
|
|
|
|
&version->date)
|
|
|
|
|| __put_user(v32.desc_len, &version->desc_len)
|
|
|
|
|| __put_user((void __user *)(unsigned long)v32.desc,
|
|
|
|
&version->desc))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_VERSION, (unsigned long)version);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(v32.version_major, &version->version_major)
|
|
|
|
|| __get_user(v32.version_minor, &version->version_minor)
|
|
|
|
|| __get_user(v32.version_patchlevel, &version->version_patchlevel)
|
|
|
|
|| __get_user(v32.name_len, &version->name_len)
|
|
|
|
|| __get_user(v32.date_len, &version->date_len)
|
|
|
|
|| __get_user(v32.desc_len, &version->desc_len))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
if (copy_to_user((void __user *)arg, &v32, sizeof(v32)))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_unique32 {
|
|
|
|
u32 unique_len; /**< Length of unique */
|
|
|
|
u32 unique; /**< Unique name for driver instantiation */
|
|
|
|
} drm_unique32_t;
|
|
|
|
|
|
|
|
static int compat_drm_getunique(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_unique32_t uq32;
|
|
|
|
drm_unique_t __user *u;
|
|
|
|
int err;
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
if (copy_from_user(&uq32, (void __user *)arg, sizeof(uq32)))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
u = compat_alloc_user_space(sizeof(*u));
|
|
|
|
if (!access_ok(VERIFY_WRITE, u, sizeof(*u)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(uq32.unique_len, &u->unique_len)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((void __user *)(unsigned long)uq32.unique,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&u->unique))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_GET_UNIQUE, (unsigned long)u);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(uq32.unique_len, &u->unique_len))
|
|
|
|
return -EFAULT;
|
2005-09-25 04:28:13 +00:00
|
|
|
if (copy_to_user((void __user *)arg, &uq32, sizeof(uq32)))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_setunique(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_unique32_t uq32;
|
|
|
|
drm_unique_t __user *u;
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
if (copy_from_user(&uq32, (void __user *)arg, sizeof(uq32)))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
u = compat_alloc_user_space(sizeof(*u));
|
|
|
|
if (!access_ok(VERIFY_WRITE, u, sizeof(*u)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(uq32.unique_len, &u->unique_len)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((void __user *)(unsigned long)uq32.unique,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&u->unique))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_SET_UNIQUE, (unsigned long)u);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_map32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
u32 offset; /**< Requested physical address (0 for SAREA)*/
|
|
|
|
u32 size; /**< Requested physical size (bytes) */
|
|
|
|
drm_map_type_t type; /**< Type of memory to map */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
drm_map_flags_t flags; /**< Flags */
|
2005-09-25 04:28:13 +00:00
|
|
|
u32 handle; /**< User-space: "Handle" to pass to mmap() */
|
|
|
|
int mtrr; /**< MTRR slot used */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_map32_t;
|
|
|
|
|
|
|
|
static int compat_drm_getmap(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_map32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_map32_t m32;
|
|
|
|
drm_map_t __user *map;
|
|
|
|
int idx, err;
|
|
|
|
void *handle;
|
|
|
|
|
|
|
|
if (get_user(idx, &argp->offset))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
map = compat_alloc_user_space(sizeof(*map));
|
|
|
|
if (!access_ok(VERIFY_WRITE, map, sizeof(*map)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(idx, &map->offset))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_GET_MAP, (unsigned long)map);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(m32.offset, &map->offset)
|
|
|
|
|| __get_user(m32.size, &map->size)
|
|
|
|
|| __get_user(m32.type, &map->type)
|
|
|
|
|| __get_user(m32.flags, &map->flags)
|
|
|
|
|| __get_user(handle, &map->handle)
|
|
|
|
|| __get_user(m32.mtrr, &map->mtrr))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
m32.handle = (unsigned long)handle;
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (copy_to_user(argp, &m32, sizeof(m32)))
|
|
|
|
return -EFAULT;
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_addmap(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_map32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_map32_t m32;
|
|
|
|
drm_map_t __user *map;
|
|
|
|
int err;
|
|
|
|
void *handle;
|
|
|
|
|
|
|
|
if (copy_from_user(&m32, argp, sizeof(m32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
map = compat_alloc_user_space(sizeof(*map));
|
|
|
|
if (!access_ok(VERIFY_WRITE, map, sizeof(*map)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(m32.offset, &map->offset)
|
|
|
|
|| __put_user(m32.size, &map->size)
|
|
|
|
|| __put_user(m32.type, &map->type)
|
|
|
|
|| __put_user(m32.flags, &map->flags))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_ADD_MAP, (unsigned long)map);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(m32.offset, &map->offset)
|
|
|
|
|| __get_user(m32.mtrr, &map->mtrr)
|
|
|
|
|| __get_user(handle, &map->handle))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
m32.handle = (unsigned long)handle;
|
|
|
|
if (m32.handle != (unsigned long)handle && printk_ratelimit())
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
printk(KERN_ERR "compat_drm_addmap truncated handle"
|
|
|
|
" %p for type %d offset %x\n",
|
|
|
|
handle, m32.type, m32.offset);
|
|
|
|
|
|
|
|
if (copy_to_user(argp, &m32, sizeof(m32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_rmmap(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_map32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_map_t __user *map;
|
|
|
|
u32 handle;
|
|
|
|
|
|
|
|
if (get_user(handle, &argp->handle))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
map = compat_alloc_user_space(sizeof(*map));
|
|
|
|
if (!access_ok(VERIFY_WRITE, map, sizeof(*map)))
|
|
|
|
return -EFAULT;
|
2005-09-25 04:28:13 +00:00
|
|
|
if (__put_user((void *)(unsigned long)handle, &map->handle))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_RM_MAP, (unsigned long)map);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_client32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int idx; /**< Which client desired? */
|
|
|
|
int auth; /**< Is client authenticated? */
|
|
|
|
u32 pid; /**< Process ID */
|
|
|
|
u32 uid; /**< User ID */
|
|
|
|
u32 magic; /**< Magic */
|
|
|
|
u32 iocs; /**< Ioctl count */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_client32_t;
|
|
|
|
|
|
|
|
static int compat_drm_getclient(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_client32_t c32;
|
|
|
|
drm_client32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_client_t __user *client;
|
|
|
|
int idx, err;
|
|
|
|
|
|
|
|
if (get_user(idx, &argp->idx))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
client = compat_alloc_user_space(sizeof(*client));
|
|
|
|
if (!access_ok(VERIFY_WRITE, client, sizeof(*client)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(idx, &client->idx))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_GET_CLIENT, (unsigned long)client);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(c32.auth, &client->auth)
|
|
|
|
|| __get_user(c32.pid, &client->pid)
|
|
|
|
|| __get_user(c32.uid, &client->uid)
|
|
|
|
|| __get_user(c32.magic, &client->magic)
|
|
|
|
|| __get_user(c32.iocs, &client->iocs))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (copy_to_user(argp, &c32, sizeof(c32)))
|
|
|
|
return -EFAULT;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_stats32 {
|
|
|
|
u32 count;
|
|
|
|
struct {
|
|
|
|
u32 value;
|
|
|
|
drm_stat_type_t type;
|
|
|
|
} data[15];
|
|
|
|
} drm_stats32_t;
|
|
|
|
|
|
|
|
static int compat_drm_getstats(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_stats32_t s32;
|
|
|
|
drm_stats32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_stats_t __user *stats;
|
|
|
|
int i, err;
|
|
|
|
|
|
|
|
stats = compat_alloc_user_space(sizeof(*stats));
|
|
|
|
if (!access_ok(VERIFY_WRITE, stats, sizeof(*stats)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_GET_STATS, (unsigned long)stats);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(s32.count, &stats->count))
|
|
|
|
return -EFAULT;
|
|
|
|
for (i = 0; i < 15; ++i)
|
|
|
|
if (__get_user(s32.data[i].value, &stats->data[i].value)
|
|
|
|
|| __get_user(s32.data[i].type, &stats->data[i].type))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (copy_to_user(argp, &s32, sizeof(s32)))
|
|
|
|
return -EFAULT;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_buf_desc32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int count; /**< Number of buffers of this size */
|
|
|
|
int size; /**< Size in bytes */
|
|
|
|
int low_mark; /**< Low water mark */
|
|
|
|
int high_mark; /**< High water mark */
|
|
|
|
int flags;
|
|
|
|
u32 agp_start; /**< Start address in the AGP aperture */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_buf_desc32_t;
|
|
|
|
|
|
|
|
static int compat_drm_addbufs(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_buf_desc32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_buf_desc_t __user *buf;
|
|
|
|
int err;
|
|
|
|
unsigned long agp_start;
|
|
|
|
|
|
|
|
buf = compat_alloc_user_space(sizeof(*buf));
|
|
|
|
if (!access_ok(VERIFY_WRITE, buf, sizeof(*buf))
|
|
|
|
|| !access_ok(VERIFY_WRITE, argp, sizeof(*argp)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (__copy_in_user(buf, argp, offsetof(drm_buf_desc32_t, agp_start))
|
|
|
|
|| __get_user(agp_start, &argp->agp_start)
|
|
|
|
|| __put_user(agp_start, &buf->agp_start))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_ADD_BUFS, (unsigned long)buf);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__copy_in_user(argp, buf, offsetof(drm_buf_desc32_t, agp_start))
|
|
|
|
|| __get_user(agp_start, &buf->agp_start)
|
|
|
|
|| __put_user(agp_start, &argp->agp_start))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_markbufs(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_buf_desc32_t b32;
|
|
|
|
drm_buf_desc32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_buf_desc_t __user *buf;
|
|
|
|
|
|
|
|
if (copy_from_user(&b32, argp, sizeof(b32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
buf = compat_alloc_user_space(sizeof(*buf));
|
|
|
|
if (!access_ok(VERIFY_WRITE, buf, sizeof(*buf)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (__put_user(b32.size, &buf->size)
|
|
|
|
|| __put_user(b32.low_mark, &buf->low_mark)
|
|
|
|
|| __put_user(b32.high_mark, &buf->high_mark))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_MARK_BUFS, (unsigned long)buf);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_buf_info32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int count; /**< Entries in list */
|
|
|
|
u32 list;
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_buf_info32_t;
|
|
|
|
|
|
|
|
static int compat_drm_infobufs(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_buf_info32_t req32;
|
|
|
|
drm_buf_info32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_buf_desc32_t __user *to;
|
|
|
|
drm_buf_info_t __user *request;
|
|
|
|
drm_buf_desc_t __user *list;
|
|
|
|
size_t nbytes;
|
|
|
|
int i, err;
|
|
|
|
int count, actual;
|
|
|
|
|
|
|
|
if (copy_from_user(&req32, argp, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
count = req32.count;
|
2005-09-25 04:28:13 +00:00
|
|
|
to = (drm_buf_desc32_t __user *) (unsigned long)req32.list;
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (count < 0)
|
|
|
|
count = 0;
|
|
|
|
if (count > 0
|
|
|
|
&& !access_ok(VERIFY_WRITE, to, count * sizeof(drm_buf_desc32_t)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
nbytes = sizeof(*request) + count * sizeof(drm_buf_desc_t);
|
|
|
|
request = compat_alloc_user_space(nbytes);
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, nbytes))
|
|
|
|
return -EFAULT;
|
|
|
|
list = (drm_buf_desc_t *) (request + 1);
|
|
|
|
|
|
|
|
if (__put_user(count, &request->count)
|
|
|
|
|| __put_user(list, &request->list))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_INFO_BUFS, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(actual, &request->count))
|
|
|
|
return -EFAULT;
|
|
|
|
if (count >= actual)
|
|
|
|
for (i = 0; i < actual; ++i)
|
|
|
|
if (__copy_in_user(&to[i], &list[i],
|
|
|
|
offsetof(drm_buf_desc_t, flags)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (__put_user(actual, &argp->count))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_buf_pub32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int idx; /**< Index into the master buffer list */
|
|
|
|
int total; /**< Buffer size */
|
|
|
|
int used; /**< Amount of buffer in use (for DMA) */
|
|
|
|
u32 address; /**< Address of buffer */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_buf_pub32_t;
|
|
|
|
|
|
|
|
typedef struct drm_buf_map32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int count; /**< Length of the buffer list */
|
|
|
|
u32 virtual; /**< Mmap'd area in user-virtual */
|
|
|
|
u32 list; /**< Buffer information */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_buf_map32_t;
|
|
|
|
|
|
|
|
static int compat_drm_mapbufs(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_buf_map32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_buf_map32_t req32;
|
|
|
|
drm_buf_pub32_t __user *list32;
|
|
|
|
drm_buf_map_t __user *request;
|
|
|
|
drm_buf_pub_t __user *list;
|
|
|
|
int i, err;
|
|
|
|
int count, actual;
|
|
|
|
size_t nbytes;
|
|
|
|
void __user *addr;
|
|
|
|
|
|
|
|
if (copy_from_user(&req32, argp, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
count = req32.count;
|
|
|
|
list32 = (void __user *)(unsigned long)req32.list;
|
|
|
|
|
|
|
|
if (count < 0)
|
|
|
|
return -EINVAL;
|
|
|
|
nbytes = sizeof(*request) + count * sizeof(drm_buf_pub_t);
|
|
|
|
request = compat_alloc_user_space(nbytes);
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, nbytes))
|
|
|
|
return -EFAULT;
|
|
|
|
list = (drm_buf_pub_t *) (request + 1);
|
|
|
|
|
|
|
|
if (__put_user(count, &request->count)
|
|
|
|
|| __put_user(list, &request->list))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_MAP_BUFS, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(actual, &request->count))
|
|
|
|
return -EFAULT;
|
|
|
|
if (count >= actual)
|
|
|
|
for (i = 0; i < actual; ++i)
|
|
|
|
if (__copy_in_user(&list32[i], &list[i],
|
|
|
|
offsetof(drm_buf_pub_t, address))
|
|
|
|
|| __get_user(addr, &list[i].address)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((unsigned long)addr,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&list32[i].address))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (__put_user(actual, &argp->count)
|
|
|
|
|| __get_user(addr, &request->virtual)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((unsigned long)addr, &argp->virtual))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_buf_free32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int count;
|
|
|
|
u32 list;
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_buf_free32_t;
|
|
|
|
|
|
|
|
static int compat_drm_freebufs(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_buf_free32_t req32;
|
|
|
|
drm_buf_free_t __user *request;
|
|
|
|
drm_buf_free32_t __user *argp = (void __user *)arg;
|
|
|
|
|
|
|
|
if (copy_from_user(&req32, argp, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(req32.count, &request->count)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((int __user *)(unsigned long)req32.list,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&request->list))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_FREE_BUFS, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_ctx_priv_map32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
unsigned int ctx_id; /**< Context requesting private mapping */
|
|
|
|
u32 handle; /**< Handle of map */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_ctx_priv_map32_t;
|
|
|
|
|
|
|
|
static int compat_drm_setsareactx(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_ctx_priv_map32_t req32;
|
|
|
|
drm_ctx_priv_map_t __user *request;
|
|
|
|
drm_ctx_priv_map32_t __user *argp = (void __user *)arg;
|
|
|
|
|
|
|
|
if (copy_from_user(&req32, argp, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(req32.ctx_id, &request->ctx_id)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((void *)(unsigned long)req32.handle,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&request->handle))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_SET_SAREA_CTX, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_getsareactx(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_ctx_priv_map_t __user *request;
|
|
|
|
drm_ctx_priv_map32_t __user *argp = (void __user *)arg;
|
|
|
|
int err;
|
|
|
|
unsigned int ctx_id;
|
|
|
|
void *handle;
|
|
|
|
|
|
|
|
if (!access_ok(VERIFY_WRITE, argp, sizeof(*argp))
|
|
|
|
|| __get_user(ctx_id, &argp->ctx_id))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(ctx_id, &request->ctx_id))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_GET_SAREA_CTX, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(handle, &request->handle)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((unsigned long)handle, &argp->handle))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_ctx_res32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int count;
|
|
|
|
u32 contexts;
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_ctx_res32_t;
|
|
|
|
|
|
|
|
static int compat_drm_resctx(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_ctx_res32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_ctx_res32_t res32;
|
|
|
|
drm_ctx_res_t __user *res;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (copy_from_user(&res32, argp, sizeof(res32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
res = compat_alloc_user_space(sizeof(*res));
|
|
|
|
if (!access_ok(VERIFY_WRITE, res, sizeof(*res)))
|
|
|
|
return -EFAULT;
|
|
|
|
if (__put_user(res32.count, &res->count)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((drm_ctx_t __user *) (unsigned long)res32.contexts,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&res->contexts))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_RES_CTX, (unsigned long)res);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(res32.count, &res->count)
|
|
|
|
|| __put_user(res32.count, &argp->count))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_dma32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int context; /**< Context handle */
|
|
|
|
int send_count; /**< Number of buffers to send */
|
|
|
|
u32 send_indices; /**< List of handles to buffers */
|
|
|
|
u32 send_sizes; /**< Lengths of data to send */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
drm_dma_flags_t flags; /**< Flags */
|
2005-09-25 04:28:13 +00:00
|
|
|
int request_count; /**< Number of buffers requested */
|
|
|
|
int request_size; /**< Desired size for buffers */
|
|
|
|
u32 request_indices; /**< Buffer information */
|
|
|
|
u32 request_sizes;
|
|
|
|
int granted_count; /**< Number of buffers granted */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_dma32_t;
|
|
|
|
|
|
|
|
static int compat_drm_dma(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_dma32_t d32;
|
2005-09-25 04:28:13 +00:00
|
|
|
drm_dma32_t __user *argp = (void __user *)arg;
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
drm_dma_t __user *d;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (copy_from_user(&d32, argp, sizeof(d32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
d = compat_alloc_user_space(sizeof(*d));
|
|
|
|
if (!access_ok(VERIFY_WRITE, d, sizeof(*d)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (__put_user(d32.context, &d->context)
|
|
|
|
|| __put_user(d32.send_count, &d->send_count)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((int __user *)(unsigned long)d32.send_indices,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&d->send_indices)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((int __user *)(unsigned long)d32.send_sizes,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&d->send_sizes)
|
|
|
|
|| __put_user(d32.flags, &d->flags)
|
|
|
|
|| __put_user(d32.request_count, &d->request_count)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((int __user *)(unsigned long)d32.request_indices,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&d->request_indices)
|
2005-09-25 04:28:13 +00:00
|
|
|
|| __put_user((int __user *)(unsigned long)d32.request_sizes,
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
&d->request_sizes))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_DMA, (unsigned long)d);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(d32.request_size, &d->request_size)
|
|
|
|
|| __get_user(d32.granted_count, &d->granted_count)
|
|
|
|
|| __put_user(d32.request_size, &argp->request_size)
|
|
|
|
|| __put_user(d32.granted_count, &argp->granted_count))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
#if __OS_HAS_AGP
|
|
|
|
typedef struct drm_agp_mode32 {
|
|
|
|
u32 mode; /**< AGP mode */
|
|
|
|
} drm_agp_mode32_t;
|
|
|
|
|
|
|
|
static int compat_drm_agp_enable(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_agp_mode32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_agp_mode32_t m32;
|
|
|
|
drm_agp_mode_t __user *mode;
|
|
|
|
|
|
|
|
if (get_user(m32.mode, &argp->mode))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
mode = compat_alloc_user_space(sizeof(*mode));
|
|
|
|
if (put_user(m32.mode, &mode->mode))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_AGP_ENABLE, (unsigned long)mode);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_agp_info32 {
|
2005-09-25 04:28:13 +00:00
|
|
|
int agp_version_major;
|
|
|
|
int agp_version_minor;
|
|
|
|
u32 mode;
|
|
|
|
u32 aperture_base; /* physical address */
|
|
|
|
u32 aperture_size; /* bytes */
|
|
|
|
u32 memory_allowed; /* bytes */
|
|
|
|
u32 memory_used;
|
|
|
|
|
|
|
|
/* PCI information */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
unsigned short id_vendor;
|
|
|
|
unsigned short id_device;
|
|
|
|
} drm_agp_info32_t;
|
|
|
|
|
|
|
|
static int compat_drm_agp_info(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_agp_info32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_agp_info32_t i32;
|
|
|
|
drm_agp_info_t __user *info;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
info = compat_alloc_user_space(sizeof(*info));
|
|
|
|
if (!access_ok(VERIFY_WRITE, info, sizeof(*info)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_AGP_INFO, (unsigned long)info);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(i32.agp_version_major, &info->agp_version_major)
|
|
|
|
|| __get_user(i32.agp_version_minor, &info->agp_version_minor)
|
|
|
|
|| __get_user(i32.mode, &info->mode)
|
|
|
|
|| __get_user(i32.aperture_base, &info->aperture_base)
|
|
|
|
|| __get_user(i32.aperture_size, &info->aperture_size)
|
|
|
|
|| __get_user(i32.memory_allowed, &info->memory_allowed)
|
|
|
|
|| __get_user(i32.memory_used, &info->memory_used)
|
|
|
|
|| __get_user(i32.id_vendor, &info->id_vendor)
|
|
|
|
|| __get_user(i32.id_device, &info->id_device))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (copy_to_user(argp, &i32, sizeof(i32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_agp_buffer32 {
|
|
|
|
u32 size; /**< In bytes -- will round to page boundary */
|
|
|
|
u32 handle; /**< Used for binding / unbinding */
|
|
|
|
u32 type; /**< Type of memory to allocate */
|
2005-09-25 04:28:13 +00:00
|
|
|
u32 physical; /**< Physical used by i810 */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
} drm_agp_buffer32_t;
|
|
|
|
|
|
|
|
static int compat_drm_agp_alloc(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_agp_buffer32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_agp_buffer32_t req32;
|
|
|
|
drm_agp_buffer_t __user *request;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (copy_from_user(&req32, argp, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request))
|
|
|
|
|| __put_user(req32.size, &request->size)
|
|
|
|
|| __put_user(req32.type, &request->type))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_AGP_ALLOC, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(req32.handle, &request->handle)
|
|
|
|
|| __get_user(req32.physical, &request->physical)
|
|
|
|
|| copy_to_user(argp, &req32, sizeof(req32))) {
|
2006-12-08 10:36:59 +00:00
|
|
|
drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_AGP_FREE, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -EFAULT;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_agp_free(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_agp_buffer32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_agp_buffer_t __user *request;
|
|
|
|
u32 handle;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request))
|
|
|
|
|| get_user(handle, &argp->handle)
|
|
|
|
|| __put_user(handle, &request->handle))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_AGP_FREE, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct drm_agp_binding32 {
|
|
|
|
u32 handle; /**< From drm_agp_buffer */
|
|
|
|
u32 offset; /**< In bytes -- will round to page boundary */
|
|
|
|
} drm_agp_binding32_t;
|
|
|
|
|
|
|
|
static int compat_drm_agp_bind(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_agp_binding32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_agp_binding32_t req32;
|
|
|
|
drm_agp_binding_t __user *request;
|
|
|
|
|
|
|
|
if (copy_from_user(&req32, argp, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request))
|
|
|
|
|| __put_user(req32.handle, &request->handle)
|
|
|
|
|| __put_user(req32.offset, &request->offset))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_AGP_BIND, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_agp_unbind(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_agp_binding32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_agp_binding_t __user *request;
|
|
|
|
u32 handle;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request))
|
|
|
|
|| get_user(handle, &argp->handle)
|
|
|
|
|| __put_user(handle, &request->handle))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_AGP_UNBIND, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
2005-09-25 04:28:13 +00:00
|
|
|
#endif /* __OS_HAS_AGP */
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
|
|
|
|
typedef struct drm_scatter_gather32 {
|
|
|
|
u32 size; /**< In bytes -- will round to page boundary */
|
|
|
|
u32 handle; /**< Used for mapping / unmapping */
|
|
|
|
} drm_scatter_gather32_t;
|
|
|
|
|
|
|
|
static int compat_drm_sg_alloc(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_scatter_gather32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_scatter_gather_t __user *request;
|
|
|
|
int err;
|
|
|
|
unsigned long x;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request))
|
|
|
|
|| !access_ok(VERIFY_WRITE, argp, sizeof(*argp))
|
|
|
|
|| __get_user(x, &argp->size)
|
|
|
|
|| __put_user(x, &request->size))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_SG_ALLOC, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
/* XXX not sure about the handle conversion here... */
|
|
|
|
if (__get_user(x, &request->handle)
|
|
|
|
|| __put_user(x >> PAGE_SHIFT, &argp->handle))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int compat_drm_sg_free(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_scatter_gather32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_scatter_gather_t __user *request;
|
|
|
|
unsigned long x;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request))
|
|
|
|
|| !access_ok(VERIFY_WRITE, argp, sizeof(*argp))
|
|
|
|
|| __get_user(x, &argp->handle)
|
|
|
|
|| __put_user(x << PAGE_SHIFT, &request->handle))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
return drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_SG_FREE, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
struct drm_wait_vblank_request32 {
|
|
|
|
drm_vblank_seq_type_t type;
|
|
|
|
unsigned int sequence;
|
|
|
|
u32 signal;
|
|
|
|
};
|
|
|
|
|
|
|
|
struct drm_wait_vblank_reply32 {
|
|
|
|
drm_vblank_seq_type_t type;
|
|
|
|
unsigned int sequence;
|
|
|
|
s32 tval_sec;
|
|
|
|
s32 tval_usec;
|
|
|
|
};
|
|
|
|
|
|
|
|
typedef union drm_wait_vblank32 {
|
|
|
|
struct drm_wait_vblank_request32 request;
|
|
|
|
struct drm_wait_vblank_reply32 reply;
|
|
|
|
} drm_wait_vblank32_t;
|
|
|
|
|
|
|
|
static int compat_drm_wait_vblank(struct file *file, unsigned int cmd,
|
|
|
|
unsigned long arg)
|
|
|
|
{
|
|
|
|
drm_wait_vblank32_t __user *argp = (void __user *)arg;
|
|
|
|
drm_wait_vblank32_t req32;
|
|
|
|
drm_wait_vblank_t __user *request;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (copy_from_user(&req32, argp, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
request = compat_alloc_user_space(sizeof(*request));
|
|
|
|
if (!access_ok(VERIFY_WRITE, request, sizeof(*request))
|
|
|
|
|| __put_user(req32.request.type, &request->request.type)
|
|
|
|
|| __put_user(req32.request.sequence, &request->request.sequence)
|
|
|
|
|| __put_user(req32.request.signal, &request->request.signal))
|
|
|
|
return -EFAULT;
|
|
|
|
|
2006-12-08 10:36:59 +00:00
|
|
|
err = drm_ioctl(file->f_path.dentry->d_inode, file,
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_IOCTL_WAIT_VBLANK, (unsigned long)request);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
|
|
|
if (__get_user(req32.reply.type, &request->reply.type)
|
|
|
|
|| __get_user(req32.reply.sequence, &request->reply.sequence)
|
|
|
|
|| __get_user(req32.reply.tval_sec, &request->reply.tval_sec)
|
|
|
|
|| __get_user(req32.reply.tval_usec, &request->reply.tval_usec))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
if (copy_to_user(argp, &req32, sizeof(req32)))
|
|
|
|
return -EFAULT;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
drm_ioctl_compat_t *drm_compat_ioctls[] = {
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_VERSION32)] = compat_drm_version,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_GET_UNIQUE32)] = compat_drm_getunique,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_GET_MAP32)] = compat_drm_getmap,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_GET_CLIENT32)] = compat_drm_getclient,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_GET_STATS32)] = compat_drm_getstats,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_SET_UNIQUE32)] = compat_drm_setunique,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_ADD_MAP32)] = compat_drm_addmap,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_ADD_BUFS32)] = compat_drm_addbufs,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_MARK_BUFS32)] = compat_drm_markbufs,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_INFO_BUFS32)] = compat_drm_infobufs,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_MAP_BUFS32)] = compat_drm_mapbufs,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_FREE_BUFS32)] = compat_drm_freebufs,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_RM_MAP32)] = compat_drm_rmmap,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_SET_SAREA_CTX32)] = compat_drm_setsareactx,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_GET_SAREA_CTX32)] = compat_drm_getsareactx,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_RES_CTX32)] = compat_drm_resctx,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_DMA32)] = compat_drm_dma,
|
|
|
|
#if __OS_HAS_AGP
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_AGP_ENABLE32)] = compat_drm_agp_enable,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_AGP_INFO32)] = compat_drm_agp_info,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_AGP_ALLOC32)] = compat_drm_agp_alloc,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_AGP_FREE32)] = compat_drm_agp_free,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_AGP_BIND32)] = compat_drm_agp_bind,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_AGP_UNBIND32)] = compat_drm_agp_unbind,
|
|
|
|
#endif
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_SG_ALLOC32)] = compat_drm_sg_alloc,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_SG_FREE32)] = compat_drm_sg_free,
|
|
|
|
[DRM_IOCTL_NR(DRM_IOCTL_WAIT_VBLANK32)] = compat_drm_wait_vblank,
|
|
|
|
};
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Called whenever a 32-bit process running under a 64-bit kernel
|
|
|
|
* performs an ioctl on /dev/drm.
|
|
|
|
*
|
|
|
|
* \param filp file pointer.
|
|
|
|
* \param cmd command.
|
|
|
|
* \param arg user argument.
|
|
|
|
* \return zero on success or negative number on failure.
|
|
|
|
*/
|
|
|
|
long drm_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
|
|
|
|
{
|
|
|
|
unsigned int nr = DRM_IOCTL_NR(cmd);
|
|
|
|
drm_ioctl_compat_t *fn;
|
|
|
|
int ret;
|
|
|
|
|
2006-08-07 10:07:43 +00:00
|
|
|
if (nr >= ARRAY_SIZE(drm_compat_ioctls))
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
return -ENOTTY;
|
|
|
|
|
|
|
|
fn = drm_compat_ioctls[nr];
|
|
|
|
|
|
|
|
lock_kernel(); /* XXX for now */
|
|
|
|
if (fn != NULL)
|
2005-09-25 04:28:13 +00:00
|
|
|
ret = (*fn) (filp, cmd, arg);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
else
|
2006-12-08 10:36:59 +00:00
|
|
|
ret = drm_ioctl(filp->f_path.dentry->d_inode, filp, cmd, arg);
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
unlock_kernel();
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
2005-09-25 04:28:13 +00:00
|
|
|
|
drm: 32/64-bit DRM ioctl compatibility patch
The patch is against a 2.6.11 kernel tree. I am running this with a
32-bit X server (compiled up from X.org CVS as of a couple of weeks
ago) and 32-bit DRI libraries and clients. All the userland stuff is
identical to what I am using under a 32-bit kernel on my G4 powerbook
(which is a 32-bit machine of course). I haven't tried compiling up a
64-bit X server or clients yet.
In the compatibility routines I have assumed that the kernel can
safely access user addresses after set_fs(KERNEL_DS). That is, where
an ioctl argument structure contains pointers to other structures, and
those other structures are already compatible between the 32-bit and
64-bit ABIs (i.e. they only contain things like chars, shorts or
ints), I just check the address with access_ok() and then pass it
through to the 64-bit ioctl code. I believe this approach may not
work on sparc64, but it does work on ppc64 and x86_64 at least.
One tricky area which may need to be revisited is the question of how
to handle the handles which we pass back to userspace to identify
mappings. These handles are generated in the ADDMAP ioctl and then
passed in as the offset value to mmap. However, offset values for
mmap seem to be generated in other ways as well, particularly for AGP
mappings.
The approach I have ended up with is to generate a fake 32-bit handle
only for _DRM_SHM mappings. The handles for other mappings (AGP, REG,
FB) are physical addresses which are already limited to 32 bits, and
generating fake handles for them created all sorts of problems in the
mmap/nopage code.
This patch has been updated to use the new compatibility ioctls.
From: Paul Mackerras <paulus@samba.org>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-06-23 11:29:18 +00:00
|
|
|
EXPORT_SYMBOL(drm_compat_ioctl);
|