2011-10-04 18:19:01 +08:00
|
|
|
/* exynos_drm_gem.h
|
|
|
|
|
*
|
|
|
|
|
* Copyright (c) 2011 Samsung Electronics Co., Ltd.
|
|
|
|
|
* Authoer: Inki Dae <inki.dae@samsung.com>
|
|
|
|
|
*
|
|
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a
|
|
|
|
|
* copy of this software and associated documentation files (the "Software"),
|
|
|
|
|
* to deal in the Software without restriction, including without limitation
|
|
|
|
|
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
|
|
|
|
|
* and/or sell copies of the Software, and to permit persons to whom the
|
|
|
|
|
* Software is furnished to do so, subject to the following conditions:
|
|
|
|
|
*
|
|
|
|
|
* The above copyright notice and this permission notice (including the next
|
|
|
|
|
* paragraph) shall be included in all copies or substantial portions of the
|
|
|
|
|
* Software.
|
|
|
|
|
*
|
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
|
|
|
|
* VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
|
|
|
|
|
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
|
|
|
|
|
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
|
|
|
|
|
* OTHER DEALINGS IN THE SOFTWARE.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#ifndef _EXYNOS_DRM_GEM_H_
|
|
|
|
|
#define _EXYNOS_DRM_GEM_H_
|
|
|
|
|
|
|
|
|
|
#define to_exynos_gem_obj(x) container_of(x,\
|
|
|
|
|
struct exynos_drm_gem_obj, base)
|
|
|
|
|
|
2012-04-03 20:27:58 +08:00
|
|
|
#define IS_NONCONTIG_BUFFER(f) (f & EXYNOS_BO_NONCONTIG)
|
|
|
|
|
|
2011-11-12 14:23:32 +08:00
|
|
|
/*
|
|
|
|
|
* exynos drm gem buffer structure.
|
|
|
|
|
*
|
|
|
|
|
* @kvaddr: kernel virtual address to allocated memory region.
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
* *userptr: user space address.
|
2011-11-12 14:23:32 +08:00
|
|
|
* @dma_addr: bus address(accessed by dma) to allocated memory region.
|
|
|
|
|
* - this address could be physical address without IOMMU and
|
|
|
|
|
* device address with IOMMU.
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
* @write: whether pages will be written to by the caller.
|
2012-12-07 16:51:27 +08:00
|
|
|
* @pages: Array of backing pages.
|
2012-03-16 17:47:05 +08:00
|
|
|
* @sgt: sg table to transfer page data.
|
2011-11-12 14:23:32 +08:00
|
|
|
* @size: size of allocated memory region.
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
* @pfnmap: indicate whether memory region from userptr is mmaped with
|
|
|
|
|
* VM_PFNMAP or not.
|
2011-11-12 14:23:32 +08:00
|
|
|
*/
|
|
|
|
|
struct exynos_drm_gem_buf {
|
|
|
|
|
void __iomem *kvaddr;
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
unsigned long userptr;
|
2011-11-12 14:23:32 +08:00
|
|
|
dma_addr_t dma_addr;
|
2012-10-20 22:53:42 +08:00
|
|
|
struct dma_attrs dma_attrs;
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
unsigned int write;
|
2012-12-07 16:51:27 +08:00
|
|
|
struct page **pages;
|
2012-03-16 17:47:05 +08:00
|
|
|
struct sg_table *sgt;
|
2011-11-12 14:23:32 +08:00
|
|
|
unsigned long size;
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
bool pfnmap;
|
2011-11-12 14:23:32 +08:00
|
|
|
};
|
|
|
|
|
|
2011-10-04 18:19:01 +08:00
|
|
|
/*
|
|
|
|
|
* exynos drm buffer structure.
|
|
|
|
|
*
|
|
|
|
|
* @base: a gem object.
|
|
|
|
|
* - a new handle to this gem object would be created
|
|
|
|
|
* by drm_gem_handle_create().
|
2011-11-12 14:23:32 +08:00
|
|
|
* @buffer: a pointer to exynos_drm_gem_buffer object.
|
|
|
|
|
* - contain the information to memory region allocated
|
|
|
|
|
* by user request or at framebuffer creation.
|
2011-10-04 18:19:01 +08:00
|
|
|
* continuous memory region allocated by user request
|
|
|
|
|
* or at framebuffer creation.
|
2012-06-29 15:28:17 +08:00
|
|
|
* @size: size requested from user, in bytes and this size is aligned
|
|
|
|
|
* in page unit.
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
* @vma: a pointer to vm_area.
|
2012-03-16 17:47:05 +08:00
|
|
|
* @flags: indicate memory type to allocated buffer and cache attruibute.
|
2011-10-04 18:19:01 +08:00
|
|
|
*
|
|
|
|
|
* P.S. this object would be transfered to user as kms_bo.handle so
|
|
|
|
|
* user can access the buffer through kms_bo.handle.
|
|
|
|
|
*/
|
|
|
|
|
struct exynos_drm_gem_obj {
|
2011-12-13 13:20:23 +08:00
|
|
|
struct drm_gem_object base;
|
|
|
|
|
struct exynos_drm_gem_buf *buffer;
|
2012-03-16 17:47:05 +08:00
|
|
|
unsigned long size;
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
struct vm_area_struct *vma;
|
2012-03-16 17:47:05 +08:00
|
|
|
unsigned int flags;
|
2011-10-04 18:19:01 +08:00
|
|
|
};
|
|
|
|
|
|
2012-04-23 20:01:28 +08:00
|
|
|
struct page **exynos_gem_get_pages(struct drm_gem_object *obj, gfp_t gfpmask);
|
|
|
|
|
|
2011-12-13 13:39:13 +08:00
|
|
|
/* destroy a buffer with gem object */
|
|
|
|
|
void exynos_drm_gem_destroy(struct exynos_drm_gem_obj *exynos_gem_obj);
|
|
|
|
|
|
2012-04-23 20:01:28 +08:00
|
|
|
/* create a private gem object and initialize it. */
|
|
|
|
|
struct exynos_drm_gem_obj *exynos_drm_gem_init(struct drm_device *dev,
|
|
|
|
|
unsigned long size);
|
|
|
|
|
|
2011-12-13 13:39:13 +08:00
|
|
|
/* create a new buffer with gem object */
|
2011-11-12 13:51:23 +08:00
|
|
|
struct exynos_drm_gem_obj *exynos_drm_gem_create(struct drm_device *dev,
|
2012-03-16 17:47:05 +08:00
|
|
|
unsigned int flags,
|
|
|
|
|
unsigned long size);
|
2011-10-04 18:19:01 +08:00
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* request gem object creation and buffer allocation as the size
|
|
|
|
|
* that it is calculated with framebuffer information such as width,
|
|
|
|
|
* height and bpp.
|
|
|
|
|
*/
|
|
|
|
|
int exynos_drm_gem_create_ioctl(struct drm_device *dev, void *data,
|
2011-12-13 13:20:23 +08:00
|
|
|
struct drm_file *file_priv);
|
2011-10-04 18:19:01 +08:00
|
|
|
|
2012-03-16 17:47:06 +08:00
|
|
|
/*
|
|
|
|
|
* get dma address from gem handle and this function could be used for
|
|
|
|
|
* other drivers such as 2d/3d acceleration drivers.
|
|
|
|
|
* with this function call, gem object reference count would be increased.
|
|
|
|
|
*/
|
2012-11-04 12:53:24 +08:00
|
|
|
dma_addr_t *exynos_drm_gem_get_dma_addr(struct drm_device *dev,
|
2012-03-16 17:47:06 +08:00
|
|
|
unsigned int gem_handle,
|
2012-11-04 12:53:24 +08:00
|
|
|
struct drm_file *filp);
|
2012-03-16 17:47:06 +08:00
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* put dma address from gem handle and this function could be used for
|
|
|
|
|
* other drivers such as 2d/3d acceleration drivers.
|
|
|
|
|
* with this function call, gem object reference count would be decreased.
|
|
|
|
|
*/
|
|
|
|
|
void exynos_drm_gem_put_dma_addr(struct drm_device *dev,
|
|
|
|
|
unsigned int gem_handle,
|
2012-11-04 12:53:24 +08:00
|
|
|
struct drm_file *filp);
|
2012-03-16 17:47:06 +08:00
|
|
|
|
2011-10-04 18:19:01 +08:00
|
|
|
/* get buffer offset to map to user space. */
|
|
|
|
|
int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
|
2011-12-13 13:20:23 +08:00
|
|
|
struct drm_file *file_priv);
|
2011-10-04 18:19:01 +08:00
|
|
|
|
2011-12-13 13:20:23 +08:00
|
|
|
/*
|
|
|
|
|
* mmap the physically continuous memory that a gem object contains
|
|
|
|
|
* to user space.
|
|
|
|
|
*/
|
|
|
|
|
int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
|
|
|
|
|
struct drm_file *file_priv);
|
2011-10-04 18:19:01 +08:00
|
|
|
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
/* map user space allocated by malloc to pages. */
|
|
|
|
|
int exynos_drm_gem_userptr_ioctl(struct drm_device *dev, void *data,
|
|
|
|
|
struct drm_file *file_priv);
|
|
|
|
|
|
2012-05-04 14:51:17 +08:00
|
|
|
/* get buffer information to memory region allocated by gem. */
|
|
|
|
|
int exynos_drm_gem_get_ioctl(struct drm_device *dev, void *data,
|
|
|
|
|
struct drm_file *file_priv);
|
|
|
|
|
|
2011-10-04 18:19:01 +08:00
|
|
|
/* initialize gem object. */
|
|
|
|
|
int exynos_drm_gem_init_object(struct drm_gem_object *obj);
|
|
|
|
|
|
|
|
|
|
/* free gem object. */
|
|
|
|
|
void exynos_drm_gem_free_object(struct drm_gem_object *gem_obj);
|
|
|
|
|
|
|
|
|
|
/* create memory region for drm framebuffer. */
|
|
|
|
|
int exynos_drm_gem_dumb_create(struct drm_file *file_priv,
|
2011-12-13 13:20:23 +08:00
|
|
|
struct drm_device *dev,
|
|
|
|
|
struct drm_mode_create_dumb *args);
|
2011-10-04 18:19:01 +08:00
|
|
|
|
|
|
|
|
/* map memory region for drm framebuffer to user space. */
|
|
|
|
|
int exynos_drm_gem_dumb_map_offset(struct drm_file *file_priv,
|
2011-12-13 13:20:23 +08:00
|
|
|
struct drm_device *dev, uint32_t handle,
|
|
|
|
|
uint64_t *offset);
|
2011-10-04 18:19:01 +08:00
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* destroy memory region allocated.
|
|
|
|
|
* - a gem handle and physical memory region pointed by a gem object
|
|
|
|
|
* would be released by drm_gem_handle_delete().
|
|
|
|
|
*/
|
|
|
|
|
int exynos_drm_gem_dumb_destroy(struct drm_file *file_priv,
|
2011-12-13 13:20:23 +08:00
|
|
|
struct drm_device *dev,
|
|
|
|
|
unsigned int handle);
|
|
|
|
|
|
|
|
|
|
/* page fault handler and mmap fault address(virtual) to physical memory. */
|
|
|
|
|
int exynos_drm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
|
|
|
|
|
|
|
|
|
|
/* set vm_flags and we can change the vm attribute to other one at here. */
|
|
|
|
|
int exynos_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma);
|
2011-10-04 18:19:01 +08:00
|
|
|
|
drm/exynos: add userptr feature for g2d module
This patch adds userptr feautre for G2D module.
The userptr means user space address allocated by malloc().
And the purpose of this feature is to make G2D's dma able
to access the user space region.
To user this feature, user should flag G2D_BUF_USRPTR to
offset variable of struct drm_exynos_g2d_cmd and fill
struct drm_exynos_g2d_userptr with user space address
and size for it and then should set a pointer to
drm_exynos_g2d_userptr object to data variable of struct
drm_exynos_g2d_cmd. The last bit of offset variable is used
to check if the cmdlist's buffer type is userptr or not.
If userptr, the g2d driver gets user space address and size
and then gets pages through get_user_pages().
(another case is counted as gem handle)
Below is sample codes:
static void set_cmd(struct drm_exynos_g2d_cmd *cmd,
unsigned long offset, unsigned long data)
{
cmd->offset = offset;
cmd->data = data;
}
static int solid_fill_test(int x, int y, unsigned long userptr)
{
struct drm_exynos_g2d_cmd cmd_gem[5];
struct drm_exynos_g2d_userptr g2d_userptr;
unsigned int gem_nr = 0;
...
g2d_userptr.userptr = userptr;
g2d_userptr.size = x * y * 4;
set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG |
G2D_BUF_USERPTR,
(unsigned long)&g2d_userptr);
...
}
int main(int argc, char **argv)
{
unsigned long addr;
...
addr = malloc(x * y * 4);
...
solid_fill_test(x, y, addr);
...
}
And next, the pages are mapped with iommu table and the device
address is set to cmdlist so that G2D's dma can access it.
As you may know, the pages from get_user_pages() are pinned.
In other words, they CAN NOT be migrated and also swapped out.
So the dma access would be safe.
But the use of userptr feature has performance overhead so
this patch also has memory pool to the userptr feature.
Please, assume that user sends cmdlist filled with userptr
and size every time to g2d driver, and the get_user_pages
funcion will be called every time.
The memory pool has maximum 64MB size and the userptr that
user had ever sent, is holded in the memory pool.
This meaning is that if the userptr from user is same as one
in the memory pool, device address to the userptr in the memory
pool is set to cmdlist.
And last, the pages from get_user_pages() will be freed once
user calls free() and the dma access is completed. Actually,
get_user_pages() takes 2 reference counts if the user process
has never accessed user region allocated by malloc(). Then, if
the user calls free(), the page reference count becomes 1 and
becomes 0 with put_page() call. And the reverse holds as well.
This means how the pages backed are used by dma and freed.
This patch is based on "drm/exynos: add iommu support for g2d",
https://patchwork.kernel.org/patch/1629481/
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
2012-11-04 21:48:52 +08:00
|
|
|
static inline int vma_is_io(struct vm_area_struct *vma)
|
|
|
|
|
{
|
|
|
|
|
return !!(vma->vm_flags & (VM_IO | VM_PFNMAP));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* get a copy of a virtual memory region. */
|
|
|
|
|
struct vm_area_struct *exynos_gem_get_vma(struct vm_area_struct *vma);
|
|
|
|
|
|
|
|
|
|
/* release a userspace virtual memory area. */
|
|
|
|
|
void exynos_gem_put_vma(struct vm_area_struct *vma);
|
|
|
|
|
|
|
|
|
|
/* get pages from user space. */
|
|
|
|
|
int exynos_gem_get_pages_from_userptr(unsigned long start,
|
|
|
|
|
unsigned int npages,
|
|
|
|
|
struct page **pages,
|
|
|
|
|
struct vm_area_struct *vma);
|
|
|
|
|
|
|
|
|
|
/* drop the reference to pages. */
|
|
|
|
|
void exynos_gem_put_pages_to_userptr(struct page **pages,
|
|
|
|
|
unsigned int npages,
|
|
|
|
|
struct vm_area_struct *vma);
|
|
|
|
|
|
|
|
|
|
/* map sgt with dma region. */
|
|
|
|
|
int exynos_gem_map_sgt_with_dma(struct drm_device *drm_dev,
|
|
|
|
|
struct sg_table *sgt,
|
|
|
|
|
enum dma_data_direction dir);
|
|
|
|
|
|
|
|
|
|
/* unmap sgt from dma region. */
|
|
|
|
|
void exynos_gem_unmap_sgt_from_dma(struct drm_device *drm_dev,
|
|
|
|
|
struct sg_table *sgt,
|
|
|
|
|
enum dma_data_direction dir);
|
|
|
|
|
|
2011-10-04 18:19:01 +08:00
|
|
|
#endif
|