KonstantinSeurer
/
mesa
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
mesa/src/glx/dri2_glx.c

942 lines
25 KiB
C
Raw Normal View History

DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
/*
* Copyright © 2008 Red Hat, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Soft-
* ware"), to deal in the Software without restriction, including without
* limitation the rights to use, copy, modify, merge, publish, distribute,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, provided that the above copyright
* notice(s) and this permission notice appear in all copies of the Soft-
* ware and that both the above copyright notice(s) and this permission
* notice appear in supporting documentation.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
* OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABIL-
* ITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OF THIRD PARTY
* RIGHTS. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN
* THIS NOTICE BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSE-
* QUENTIAL DAMAGES, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
* DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
* TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFOR-
* MANCE OF THIS SOFTWARE.
*
* Except as contained in this notice, the name of a copyright holder shall
* not be used in advertising or otherwise to promote the sale, use or
* other dealings in this Software without prior written authorization of
* the copyright holder.
*
* Authors:
* Kristian Høgsberg (krh@redhat.com)
*/
apple: Change ifdefs for DRI to be DRI && !APPLE Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
2010-04-02 09:35:19 +01:00
#if defined(GLX_DIRECT_RENDERING) && !defined(GLX_USE_APPLEGL)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
include <X11/Xlib.h>
2008-04-23 15:59:23 +01:00
#include <X11/Xlib.h>
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
#include <X11/extensions/Xfixes.h>
#include <X11/extensions/Xdamage.h>
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
#include "glapi.h"
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
#include "glxclient.h"
glx: Fix SwapBuffers regression introduced by 01923fb72d. After that commit, some dri2 protocol symbols were being checked from places that weren't including dri2proto.h, effectively disabling some valuable SwapBuffers codepaths.
2010-02-08 22:15:03 +00:00
#include <X11/extensions/dri2proto.h>
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
#include "xf86dri.h"
#include <dlfcn.h>
Update DRI2 implementation according to new specification.
2008-10-12 01:41:14 +01:00
#include <fcntl.h>
#include <unistd.h>
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
#include <sys/types.h>
#include <sys/mman.h>
#include "xf86drm.h"
#include "dri2.h"
glx: add dri_common.c ala xserver also drop driFilterModes which is unused in preparation of loading swrast_dri.so
2008-06-09 15:47:09 +01:00
#include "dri_common.h"
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
/* From xmlpool/options.h, user exposed so should be stable */
#define DRI_CONF_VBLANK_NEVER 0
#define DRI_CONF_VBLANK_DEF_INTERVAL_0 1
#define DRI_CONF_VBLANK_DEF_INTERVAL_1 2
#define DRI_CONF_VBLANK_ALWAYS_SYNC 3
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
#undef DRI2_MINOR
#define DRI2_MINOR 1
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_display
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
{
__GLXDRIdisplay base;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/*
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
** XFree86-DRI version information
*/
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
int driMajor;
int driMinor;
int driPatch;
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
int swapAvailable;
dri2: Event driven buffer validation. When a buffer invalidation event is received from the X server, the "invalidate" hook of the DRI2 flush extension is executed: A generic implementation (dri2InvalidateDrawable) is provided that just bumps the "pStamp" sequence number in __DRIdrawableRec. For old servers not supporting buffer invalidation events, the invalidate hook will be called before flushing the fake front/back buffer (that's typically once per frame -- not a lot worse than the situation we were in before). No effort has been made on preserving backwards compatibility with version 2 of the flush extension, but I think it's acceptable because AFAIK no released stack is making use of it. Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
2010-02-08 18:27:56 +00:00
int invalidateAvailable;
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
__glxHashTable *dri2Hash;
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
const __DRIextension *loader_extensions[4];
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
};
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
struct dri2_screen {
__GLXscreenConfigs base;
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
__DRIscreen *driScreen;
__GLXDRIscreen vtable;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
const __DRIdri2Extension *dri2;
const __DRIcoreExtension *core;
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
const __DRI2flushExtension *f;
const __DRI2configQueryExtension *config;
const __DRItexBufferExtension *texBuffer;
glx: Move driver_configs to DRI screen privates
2010-07-20 03:12:22 +01:00
const __DRIconfig **driver_configs;
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
void *driver;
int fd;
};
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_context
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
{
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
__GLXcontext base;
__GLXDRIcontext dri_vtable;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
__DRIcontext *driContext;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
};
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
{
__GLXDRIdrawable base;
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
__DRIdrawable *driDrawable;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
__DRIbuffer buffers[5];
int bufferCount;
int width, height;
int have_back;
int have_fake_front;
DRI2: add SwapInterval support Add support for the DRI2SwapInterval protocol request. This allows direct rendered clients to control their swap interval per the SGI_swap_control extension. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-11-10 21:28:01 +00:00
int swap_interval;
DRI2: Drop sarea, implement swap buffers in the X server.
2008-08-13 16:46:25 +01:00
};
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
static const struct glx_context_vtable dri2_context_vtable;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
glx: Move context destroy to context vtable
2010-07-23 04:45:18 +01:00
dri2_destroy_context(__GLXcontext *context)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_context *pcp = (struct dri2_context *) context;
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
struct dri2_screen *psc = (struct dri2_screen *) context->psc;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: Don't destroy context with XID 0 We use XID 0 to indicate the context has already been destroyed, but it's currently bound.
2010-07-24 03:05:21 +01:00
if (context->xid)
glx_send_destroy_context(psc->base.dpy, context->xid);
glx: Move context destroy to context vtable
2010-07-23 04:45:18 +01:00
if (context->extensions)
XFree((char *) context->extensions);
GarbageCollectDRIDrawables(context->psc);
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
(*psc->core->destroyContext) (pcp->driContext);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(pcp);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static Bool
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
dri2BindContext(__GLXcontext *context,
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
__GLXDRIdrawable *draw, __GLXDRIdrawable *read)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_context *pcp = (struct dri2_context *) context;
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
struct dri2_screen *psc = (struct dri2_screen *) pcp->base.psc;
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
struct dri2_drawable *pdr = (struct dri2_drawable *) draw;
struct dri2_drawable *prd = (struct dri2_drawable *) read;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
return (*psc->core->bindContext) (pcp->driContext,
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
pdr->driDrawable, prd->driDrawable);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
dri2UnbindContext(__GLXcontext *context)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_context *pcp = (struct dri2_context *) context;
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
struct dri2_screen *psc = (struct dri2_screen *) pcp->base.psc;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
(*psc->core->unbindContext) (pcp->driContext);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
static __GLXcontext *
glx: Refactor and simplify context creation This lets us better separate context creation between the different backends.
2010-07-23 21:15:31 +01:00
dri2_create_context(__GLXscreenConfigs *base,
const __GLcontextModes * mode,
GLXContext shareList, int renderType)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_context *pcp, *pcp_shared;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
struct dri2_screen *psc = (struct dri2_screen *) base;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
__GLXDRIconfigPrivate *config = (__GLXDRIconfigPrivate *) mode;
__DRIcontext *shared = NULL;
if (shareList) {
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
pcp_shared = (struct dri2_context *) shareList->driContext;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
shared = pcp_shared->driContext;
}
pcp = Xmalloc(sizeof *pcp);
if (pcp == NULL)
return NULL;
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
memset(pcp, 0, sizeof *pcp);
if (!glx_context_init(&pcp->base, &psc->base, mode)) {
Xfree(pcp);
return NULL;
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pcp->driContext =
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
(*psc->dri2->createNewContext) (psc->driScreen,
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
config->driConfig, shared, pcp);
if (pcp->driContext == NULL) {
Xfree(pcp);
return NULL;
}
glx: Allocate the __GLXcontext in the DRI drivers
2010-07-23 02:24:14 +01:00
pcp->base.vtable = &dri2_context_vtable;
pcp->base.driContext = &pcp->dri_vtable;
pcp->dri_vtable.bindContext = dri2BindContext;
pcp->dri_vtable.unbindContext = dri2UnbindContext;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return &pcp->base;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
dri2DestroyDrawable(__GLXDRIdrawable *base)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
struct dri2_screen *psc = (struct dri2_screen *) base->psc;
struct dri2_drawable *pdraw = (struct dri2_drawable *) base;
glx: Don't use __glXInitialize() when we might be holding __glXLock()
2010-07-20 02:00:09 +01:00
__GLXdisplayPrivate *dpyPriv = psc->base.display;
struct dri2_display *pdp = (struct dri2_display *)dpyPriv->dri2Display;
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
__glxHashDelete(pdp->dri2Hash, pdraw->base.xDrawable);
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
(*psc->core->destroyDrawable) (pdraw->driDrawable);
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
DRI2DestroyDrawable(psc->base.dpy, pdraw->base.xDrawable);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(pdraw);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static __GLXDRIdrawable *
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
dri2CreateDrawable(__GLXscreenConfigs *base, XID xDrawable,
GLXDrawable drawable, const __GLcontextModes * modes)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *pdraw;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
struct dri2_screen *psc = (struct dri2_screen *) base;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
__GLXDRIconfigPrivate *config = (__GLXDRIconfigPrivate *) modes;
DRI2: check for swapAvailable before using swap interval protocol This should have been part of the last change...
2010-04-19 22:19:48 +01:00
__GLXdisplayPrivate *dpyPriv;
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_display *pdp;
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
GLint vblank_mode = DRI_CONF_VBLANK_DEF_INTERVAL_1;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdraw = Xmalloc(sizeof(*pdraw));
if (!pdraw)
return NULL;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: zero out drawable structs after allocation
2010-07-24 02:45:05 +01:00
memset(pdraw, 0, sizeof *pdraw);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdraw->base.destroyDrawable = dri2DestroyDrawable;
pdraw->base.xDrawable = xDrawable;
pdraw->base.drawable = drawable;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
pdraw->base.psc = &psc->base;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdraw->bufferCount = 0;
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
pdraw->swap_interval = 1; /* default may be overridden below */
Initialize have_back. There is a user somewhere that tests it before its initial set. Signed-off-by: Eric Anholt <eric@anholt.net>
2010-04-27 22:16:46 +01:00
pdraw->have_back = 0;
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
if (psc->config)
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
psc->config->configQueryi(psc->driScreen,
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
"vblank_mode", &vblank_mode);
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
switch (vblank_mode) {
case DRI_CONF_VBLANK_NEVER:
case DRI_CONF_VBLANK_DEF_INTERVAL_0:
pdraw->swap_interval = 0;
break;
case DRI_CONF_VBLANK_DEF_INTERVAL_1:
case DRI_CONF_VBLANK_ALWAYS_SYNC:
default:
pdraw->swap_interval = 1;
break;
}
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
DRI2CreateDrawable(psc->base.dpy, xDrawable);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
dpyPriv = __glXInitialize(psc->base.dpy);
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
pdp = (struct dri2_display *)dpyPriv->dri2Display;;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/* Create a new drawable */
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
pdraw->driDrawable =
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
(*psc->dri2->createNewDrawable) (psc->driScreen,
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
config->driConfig, pdraw);
DRI2: Drop sarea, implement swap buffers in the X server.
2008-08-13 16:46:25 +01:00
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
if (!pdraw->driDrawable) {
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
DRI2DestroyDrawable(psc->base.dpy, xDrawable);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(pdraw);
return NULL;
}
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
if (__glxHashInsert(pdp->dri2Hash, xDrawable, pdraw)) {
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
(*psc->core->destroyDrawable) (pdraw->driDrawable);
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
DRI2DestroyDrawable(psc->base.dpy, xDrawable);
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
Xfree(pdraw);
return None;
}
glx/dri2: Fix build with dri2proto which doesn't define X_DRI2SwapInterval.
2010-04-23 10:14:13 +01:00
#ifdef X_DRI2SwapInterval
DRI2: synchronize swap interval with server at startup time In the direct rendered case, we need to tell the server our initial swap interval. If we don't, the local and server values will be out of sync, since the server and client defaults may be different (as they were before this patch).
2010-04-19 17:54:08 +01:00
/*
* Make sure server has the same swap interval we do for the new
* drawable.
*/
DRI2: check for swapAvailable before using swap interval protocol This should have been part of the last change...
2010-04-19 22:19:48 +01:00
if (pdp->swapAvailable)
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
DRI2SwapInterval(psc->base.dpy, xDrawable, pdraw->swap_interval);
glx/dri2: Fix build with dri2proto which doesn't define X_DRI2SwapInterval.
2010-04-23 10:14:13 +01:00
#endif
DRI2: check for swapAvailable before using swap interval protocol This should have been part of the last change...
2010-04-19 22:19:48 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return &pdraw->base;
DRI2: Drop sarea, implement swap buffers in the X server.
2008-08-13 16:46:25 +01:00
}
glx/dri2: Fix debug build with old dri2proto. Apparently the higher compiler optimization level in non-debug builds was eliminating the unused functions referencing the unresolved DRI2 symbols...
2010-03-30 10:50:13 +01:00
#ifdef X_DRI2GetMSC
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
static int
dri2DrawableGetMSC(__GLXscreenConfigs *psc, __GLXDRIdrawable *pdraw,
int64_t *ust, int64_t *msc, int64_t *sbc)
{
glx: Workaround mismatch in signedness between extensions and protocol The DRI2 protocol for ust, msc and sbc are unsigned but the extensions talk about int64_t. Do a little dance to make the compiler shut up.
2010-07-20 01:01:59 +01:00
CARD64 dri2_ust, dri2_msc, dri2_sbc;
int ret;
ret = DRI2GetMSC(psc->dpy, pdraw->xDrawable,
&dri2_ust, &dri2_msc, &dri2_sbc);
*ust = dri2_ust;
*msc = dri2_msc;
*sbc = dri2_sbc;
return ret;
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
}
glx/dri2: Fix debug build with old dri2proto. Apparently the higher compiler optimization level in non-debug builds was eliminating the unused functions referencing the unresolved DRI2 symbols...
2010-03-30 10:50:13 +01:00
#endif
#ifdef X_DRI2WaitMSC
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
static int
dri2WaitForMSC(__GLXDRIdrawable *pdraw, int64_t target_msc, int64_t divisor,
int64_t remainder, int64_t *ust, int64_t *msc, int64_t *sbc)
{
glx: Workaround mismatch in signedness between extensions and protocol The DRI2 protocol for ust, msc and sbc are unsigned but the extensions talk about int64_t. Do a little dance to make the compiler shut up.
2010-07-20 01:01:59 +01:00
CARD64 dri2_ust, dri2_msc, dri2_sbc;
int ret;
ret = DRI2WaitMSC(pdraw->psc->dpy, pdraw->xDrawable, target_msc, divisor,
remainder, &dri2_ust, &dri2_msc, &dri2_sbc);
*ust = dri2_ust;
*msc = dri2_msc;
*sbc = dri2_sbc;
return ret;
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
}
static int
dri2WaitForSBC(__GLXDRIdrawable *pdraw, int64_t target_sbc, int64_t *ust,
int64_t *msc, int64_t *sbc)
{
glx: Workaround mismatch in signedness between extensions and protocol The DRI2 protocol for ust, msc and sbc are unsigned but the extensions talk about int64_t. Do a little dance to make the compiler shut up.
2010-07-20 01:01:59 +01:00
CARD64 dri2_ust, dri2_msc, dri2_sbc;
int ret;
ret = DRI2WaitSBC(pdraw->psc->dpy, pdraw->xDrawable,
target_sbc, &dri2_ust, &dri2_msc, &dri2_sbc);
*ust = dri2_ust;
*msc = dri2_msc;
*sbc = dri2_sbc;
return ret;
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
}
glx/dri2: Fix debug build with old dri2proto. Apparently the higher compiler optimization level in non-debug builds was eliminating the unused functions referencing the unresolved DRI2 symbols...
2010-03-30 10:50:13 +01:00
#endif /* X_DRI2WaitMSC */
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
dri2CopySubBuffer(__GLXDRIdrawable *pdraw, int x, int y, int width, int height)
Update DRI2 implementation according to new specification.
2008-10-12 01:41:14 +01:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *priv = (struct dri2_drawable *) pdraw;
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
struct dri2_screen *psc = (struct dri2_screen *) pdraw->psc;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
XRectangle xrect;
XserverRegion region;
Update DRI2 implementation according to new specification.
2008-10-12 01:41:14 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/* Check we have the right attachments */
if (!priv->have_back)
return;
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
xrect.x = x;
xrect.y = priv->height - y - height;
xrect.width = width;
xrect.height = height;
Update DRI2 implementation according to new specification.
2008-10-12 01:41:14 +01:00
glx: add support for a reallyFlush() function before swap occurs.
2009-02-19 18:39:08 +00:00
#ifdef __DRI2_FLUSH
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
if (psc->f)
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
(*psc->f->flush) (priv->driDrawable);
glx: add support for a reallyFlush() function before swap occurs.
2009-02-19 18:39:08 +00:00
#endif
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
region = XFixesCreateRegion(psc->base.dpy, &xrect, 1);
DRI2CopyRegion(psc->base.dpy, pdraw->xDrawable, region,
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
DRI2BufferFrontLeft, DRI2BufferBackLeft);
dri2: Refresh the fake front contents after glXSwapBuffers(). Bug #19177. Reviewed by: Ian Romanick <ian.d.romanick@intel.com>
2009-06-23 00:33:29 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/* Refresh the fake front (if present) after we just damaged the real
* front.
*/
glx/dri2: Fix dri2CopySubBuffer() again. Only refresh the fake front buffer if there is one, and only destroy the region once. Fixes X11 protocol errors reported by 'mcgreg' on IRC.
2010-07-20 11:21:17 +01:00
if (priv->have_fake_front)
DRI2CopyRegion(psc->base.dpy, pdraw->xDrawable, region,
DRI2BufferFakeFrontLeft, DRI2BufferFrontLeft);
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
XFixesDestroyRegion(psc->base.dpy, region);
Update DRI2 implementation according to new specification.
2008-10-12 01:41:14 +01:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
dri2_copy_drawable(struct dri2_drawable *priv, int dest, int src)
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
{
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
XRectangle xrect;
XserverRegion region;
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
struct dri2_screen *psc = (struct dri2_screen *) priv->base.psc;
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
xrect.x = 0;
xrect.y = 0;
xrect.width = priv->width;
xrect.height = priv->height;
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
glx: add support for a reallyFlush() function before swap occurs.
2009-02-19 18:39:08 +00:00
#ifdef __DRI2_FLUSH
glx: Factor out common code from dri2WaitGL() and dri2WaitX()
2010-07-19 15:32:02 +01:00
if (psc->f)
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
(*psc->f->flush) (priv->driDrawable);
glx: add support for a reallyFlush() function before swap occurs.
2009-02-19 18:39:08 +00:00
#endif
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
region = XFixesCreateRegion(psc->base.dpy, &xrect, 1);
DRI2CopyRegion(psc->base.dpy, priv->base.xDrawable, region, dest, src);
XFixesDestroyRegion(psc->base.dpy, region);
glx: Factor out common code from dri2WaitGL() and dri2WaitX()
2010-07-19 15:32:02 +01:00
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
glx: Move WaitGL, WaitX, UseXFont to context vtable functions
2010-07-23 03:24:00 +01:00
dri2_wait_x(__GLXcontext *gc)
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
{
glx: Move WaitGL, WaitX, UseXFont to context vtable functions
2010-07-23 03:24:00 +01:00
struct dri2_drawable *priv = (struct dri2_drawable *)
glx: Drop screen argument to GetGLXDRIDrawable We'll just get it from the returned drawable when we need it.
2010-07-23 03:36:37 +01:00
GetGLXDRIDrawable(gc->currentDpy, gc->currentDrawable);
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
glx: Move WaitGL, WaitX, UseXFont to context vtable functions
2010-07-23 03:24:00 +01:00
if (priv == NULL || !priv->have_fake_front)
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return;
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
glx: Factor out common code from dri2WaitGL() and dri2WaitX()
2010-07-19 15:32:02 +01:00
dri2_copy_drawable(priv, DRI2BufferFakeFrontLeft, DRI2BufferFrontLeft);
}
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
glx: Factor out common code from dri2WaitGL() and dri2WaitX()
2010-07-19 15:32:02 +01:00
static void
glx: Move WaitGL, WaitX, UseXFont to context vtable functions
2010-07-23 03:24:00 +01:00
dri2_wait_gl(__GLXcontext *gc)
glx: Factor out common code from dri2WaitGL() and dri2WaitX()
2010-07-19 15:32:02 +01:00
{
glx: Move WaitGL, WaitX, UseXFont to context vtable functions
2010-07-23 03:24:00 +01:00
struct dri2_drawable *priv = (struct dri2_drawable *)
glx: Drop screen argument to GetGLXDRIDrawable We'll just get it from the returned drawable when we need it.
2010-07-23 03:36:37 +01:00
GetGLXDRIDrawable(gc->currentDpy, gc->currentDrawable);
glx: add support for a reallyFlush() function before swap occurs.
2009-02-19 18:39:08 +00:00
glx: Move WaitGL, WaitX, UseXFont to context vtable functions
2010-07-23 03:24:00 +01:00
if (priv == NULL || !priv->have_fake_front)
glx: Factor out common code from dri2WaitGL() and dri2WaitX()
2010-07-19 15:32:02 +01:00
return;
dri2_copy_drawable(priv, DRI2BufferFrontLeft, DRI2BufferFakeFrontLeft);
dri2: support glXWaitX & glXWaitGL by using fake front buffer.
2009-02-16 11:44:40 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
dri2: Event driven buffer validation. When a buffer invalidation event is received from the X server, the "invalidate" hook of the DRI2 flush extension is executed: A generic implementation (dri2InvalidateDrawable) is provided that just bumps the "pStamp" sequence number in __DRIdrawableRec. For old servers not supporting buffer invalidation events, the invalidate hook will be called before flushing the fake front/back buffer (that's typically once per frame -- not a lot worse than the situation we were in before). No effort has been made on preserving backwards compatibility with version 2 of the flush extension, but I think it's acceptable because AFAIK no released stack is making use of it. Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
2010-02-08 18:27:56 +00:00
dri2FlushFrontBuffer(__DRIdrawable *driDrawable, void *loaderPrivate)
DRI2: Provide an interface for drivers to flush front-buffer rendering Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-06 20:44:18 +01:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *pdraw = loaderPrivate;
dri2_glx: Put the invalidate b/c code back in The backwards compatibility code calls the DRI driver invalidate hook on swap buffer and flush front buffer. This lets the DRI driver rely on invalidate callbacks and drop the glViewport() hack, even if the server doesn't send invalidate events. This is essentially a revert of 2d00d16da7f5d2255cb37b48edaf4cbb9ca7e930, except that we now also pass the __DRI_USE_INVALIDATE extension even when the server doesn't have DRI2 invalidate events.
2010-05-18 19:45:10 +01:00
__GLXdisplayPrivate *priv = __glXInitialize(pdraw->base.psc->dpy);
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_display *pdp = (struct dri2_display *)priv->dri2Display;
glx: Don't try to swap a front buffer if we don't have one. Fixes glean glsl1 since 7b7845a076c933e096ac511b4184141ba194449a
2010-07-23 21:41:13 +01:00
GLXContext gc = __glXGetCurrentContext();
dri2_glx: Put the invalidate b/c code back in The backwards compatibility code calls the DRI driver invalidate hook on swap buffer and flush front buffer. This lets the DRI driver rely on invalidate callbacks and drop the glViewport() hack, even if the server doesn't send invalidate events. This is essentially a revert of 2d00d16da7f5d2255cb37b48edaf4cbb9ca7e930, except that we now also pass the __DRI_USE_INVALIDATE extension even when the server doesn't have DRI2 invalidate events.
2010-05-18 19:45:10 +01:00
/* Old servers don't send invalidate events */
if (!pdp->invalidateAvailable)
dri2InvalidateBuffers(priv->dpy, pdraw->base.drawable);
glx: Don't try to swap a front buffer if we don't have one. Fixes glean glsl1 since 7b7845a076c933e096ac511b4184141ba194449a
2010-07-23 21:41:13 +01:00
dri2_wait_gl(gc);
DRI2: Provide an interface for drivers to flush front-buffer rendering Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-06 20:44:18 +01:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
dri2DestroyScreen(__GLXscreenConfigs *base)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
struct dri2_screen *psc = (struct dri2_screen *) base;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/* Free the direct rendering per screen data */
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
(*psc->core->destroyScreen) (psc->driScreen);
glx: Move driver_configs to DRI screen privates
2010-07-20 03:12:22 +01:00
driDestroyConfigs(psc->driver_configs);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
close(psc->fd);
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
Xfree(psc);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
/**
* Process list of buffer received from the server
*
* Processes the list of buffers received in a reply from the server to either
* \c DRI2GetBuffers or \c DRI2GetBuffersWithFormat.
*/
static void
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
process_buffers(struct dri2_drawable * pdraw, DRI2Buffer * buffers,
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
unsigned count)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
int i;
pdraw->bufferCount = count;
pdraw->have_fake_front = 0;
pdraw->have_back = 0;
/* This assumes the DRI2 buffer attachment tokens matches the
* __DRIbuffer tokens. */
for (i = 0; i < count; i++) {
pdraw->buffers[i].attachment = buffers[i].attachment;
pdraw->buffers[i].name = buffers[i].name;
pdraw->buffers[i].pitch = buffers[i].pitch;
pdraw->buffers[i].cpp = buffers[i].cpp;
pdraw->buffers[i].flags = buffers[i].flags;
if (pdraw->buffers[i].attachment == __DRI_BUFFER_FAKE_FRONT_LEFT)
pdraw->have_fake_front = 1;
if (pdraw->buffers[i].attachment == __DRI_BUFFER_BACK_LEFT)
pdraw->have_back = 1;
}
Revert pointless reindents to avoid merge conflicts. Why are we reindenting code that's work in progress...
2008-10-15 04:07:42 +01:00
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
}
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
static int64_t
dri2SwapBuffers(__GLXDRIdrawable *pdraw, int64_t target_msc, int64_t divisor,
int64_t remainder)
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *priv = (struct dri2_drawable *) pdraw;
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
__GLXdisplayPrivate *dpyPriv = __glXInitialize(priv->base.psc->dpy);
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
struct dri2_screen *psc = (struct dri2_screen *) priv->base.psc;
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_display *pdp =
(struct dri2_display *)dpyPriv->dri2Display;
glx: Workaround mismatch in signedness between extensions and protocol The DRI2 protocol for ust, msc and sbc are unsigned but the extensions talk about int64_t. Do a little dance to make the compiler shut up.
2010-07-20 01:01:59 +01:00
CARD64 ret;
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
#ifdef __DRI2_FLUSH
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
if (psc->f)
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
(*psc->f->flush)(priv->driDrawable);
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
#endif
dri2_glx: Put the invalidate b/c code back in The backwards compatibility code calls the DRI driver invalidate hook on swap buffer and flush front buffer. This lets the DRI driver rely on invalidate callbacks and drop the glViewport() hack, even if the server doesn't send invalidate events. This is essentially a revert of 2d00d16da7f5d2255cb37b48edaf4cbb9ca7e930, except that we now also pass the __DRI_USE_INVALIDATE extension even when the server doesn't have DRI2 invalidate events.
2010-05-18 19:45:10 +01:00
/* Old servers don't send invalidate events */
if (!pdp->invalidateAvailable)
dri2InvalidateBuffers(dpyPriv->dpy, pdraw->drawable);
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
/* Old servers can't handle swapbuffers */
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
if (!pdp->swapAvailable) {
dri2CopySubBuffer(pdraw, 0, 0, priv->width, priv->height);
return 0;
}
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
glx: permit building with older protocol headers I'd like to be able to build mesa on current distro releases without having to upgrade from the standard dri2proto and glproto headers. With this change I'm able to build on ancient releases such as Ubuntu 9-10... In general, it would be nice to be able to build-test mesa to check for unintended breakages without having to follow the external dependencies of every group working on the codebase. Seems to introduce no changes to the build of libglapi.a when tested against new versions of the headers.
2010-02-04 12:46:21 +00:00
#ifdef X_DRI2SwapBuffers
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
DRI2SwapBuffers(psc->base.dpy, pdraw->xDrawable, target_msc, divisor,
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
remainder, &ret);
glx: permit building with older protocol headers I'd like to be able to build mesa on current distro releases without having to upgrade from the standard dri2proto and glproto headers. With this change I'm able to build on ancient releases such as Ubuntu 9-10... In general, it would be nice to be able to build-test mesa to check for unintended breakages without having to follow the external dependencies of every group working on the codebase. Seems to introduce no changes to the build of libglapi.a when tested against new versions of the headers.
2010-02-04 12:46:21 +00:00
#endif
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
return ret;
DRI2: add SwapBuffers support Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct and indirect rendering context. This request allows the display server to optimize back->front swaps (e.g. through page flipping) and allows us to more easily support other GLX features like swap interval and the OML sync extension in DRI2. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-07-17 08:30:29 +01:00
}
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
static __DRIbuffer *
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
dri2GetBuffers(__DRIdrawable * driDrawable,
int *width, int *height,
unsigned int *attachments, int count,
int *out_count, void *loaderPrivate)
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *pdraw = loaderPrivate;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
DRI2Buffer *buffers;
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
buffers = DRI2GetBuffers(pdraw->base.psc->dpy, pdraw->base.xDrawable,
width, height, attachments, count, out_count);
if (buffers == NULL)
return NULL;
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdraw->width = *width;
pdraw->height = *height;
process_buffers(pdraw, buffers, *out_count);
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(buffers);
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return pdraw->buffers;
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
}
static __DRIbuffer *
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
dri2GetBuffersWithFormat(__DRIdrawable * driDrawable,
int *width, int *height,
unsigned int *attachments, int count,
int *out_count, void *loaderPrivate)
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *pdraw = loaderPrivate;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
DRI2Buffer *buffers;
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
buffers = DRI2GetBuffersWithFormat(pdraw->base.psc->dpy,
pdraw->base.xDrawable,
width, height, attachments,
count, out_count);
if (buffers == NULL)
return NULL;
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdraw->width = *width;
pdraw->height = *height;
process_buffers(pdraw, buffers, *out_count);
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(buffers);
Revert pointless reindents to avoid merge conflicts. Why are we reindenting code that's work in progress...
2008-10-15 04:07:42 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return pdraw->buffers;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
glx/dri2: Fix debug build with old dri2proto. Apparently the higher compiler optimization level in non-debug builds was eliminating the unused functions referencing the unresolved DRI2 symbols...
2010-03-30 10:50:13 +01:00
#ifdef X_DRI2SwapInterval
glx: Move DRI1 specific extensions and code to DRI1 screen private
2010-07-19 21:39:53 +01:00
static int
DRI2: add SwapInterval support Add support for the DRI2SwapInterval protocol request. This allows direct rendered clients to control their swap interval per the SGI_swap_control extension. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-11-10 21:28:01 +00:00
dri2SetSwapInterval(__GLXDRIdrawable *pdraw, int interval)
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *priv = (struct dri2_drawable *) pdraw;
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
GLint vblank_mode = DRI_CONF_VBLANK_DEF_INTERVAL_1;
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
struct dri2_screen *psc = (struct dri2_screen *) priv->base.psc;
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
if (psc->config)
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
psc->config->configQueryi(psc->driScreen,
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
"vblank_mode", &vblank_mode);
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
switch (vblank_mode) {
case DRI_CONF_VBLANK_NEVER:
glx: Move DRI1 specific extensions and code to DRI1 screen private
2010-07-19 21:39:53 +01:00
return GLX_BAD_VALUE;
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
case DRI_CONF_VBLANK_ALWAYS_SYNC:
if (interval <= 0)
glx: Move DRI1 specific extensions and code to DRI1 screen private
2010-07-19 21:39:53 +01:00
return GLX_BAD_VALUE;
DRI2/GLX: check for vblank_mode in DRI2 GLX code Re-add support for the vblank_mode environment and configuration variable. Useful for benchmarking and app control.
2010-04-22 20:49:03 +01:00
break;
default:
break;
}
DRI2: add SwapInterval support Add support for the DRI2SwapInterval protocol request. This allows direct rendered clients to control their swap interval per the SGI_swap_control extension. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-11-10 21:28:01 +00:00
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
DRI2SwapInterval(priv->base.psc->dpy, priv->base.xDrawable, interval);
DRI2: add SwapInterval support Add support for the DRI2SwapInterval protocol request. This allows direct rendered clients to control their swap interval per the SGI_swap_control extension. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-11-10 21:28:01 +00:00
priv->swap_interval = interval;
glx: Move DRI1 specific extensions and code to DRI1 screen private
2010-07-19 21:39:53 +01:00
return 0;
DRI2: add SwapInterval support Add support for the DRI2SwapInterval protocol request. This allows direct rendered clients to control their swap interval per the SGI_swap_control extension. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-11-10 21:28:01 +00:00
}
glx: Workaround mismatch in signedness between extensions and protocol The DRI2 protocol for ust, msc and sbc are unsigned but the extensions talk about int64_t. Do a little dance to make the compiler shut up.
2010-07-20 01:01:59 +01:00
static int
DRI2: add SwapInterval support Add support for the DRI2SwapInterval protocol request. This allows direct rendered clients to control their swap interval per the SGI_swap_control extension. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-11-10 21:28:01 +00:00
dri2GetSwapInterval(__GLXDRIdrawable *pdraw)
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_drawable *priv = (struct dri2_drawable *) pdraw;
DRI2: add SwapInterval support Add support for the DRI2SwapInterval protocol request. This allows direct rendered clients to control their swap interval per the SGI_swap_control extension. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-11-10 21:28:01 +00:00
return priv->swap_interval;
}
glx/dri2: Fix debug build with old dri2proto. Apparently the higher compiler optimization level in non-debug builds was eliminating the unused functions referencing the unresolved DRI2 symbols...
2010-03-30 10:50:13 +01:00
#endif /* X_DRI2SwapInterval */
DRI2: Drop sarea, implement swap buffers in the X server.
2008-08-13 16:46:25 +01:00
static const __DRIdri2LoaderExtension dri2LoaderExtension = {
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
{__DRI_DRI2_LOADER, __DRI_DRI2_LOADER_VERSION},
dri2GetBuffers,
dri2FlushFrontBuffer,
dri2GetBuffersWithFormat,
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
};
static const __DRIdri2LoaderExtension dri2LoaderExtension_old = {
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
{__DRI_DRI2_LOADER, __DRI_DRI2_LOADER_VERSION},
dri2GetBuffers,
dri2FlushFrontBuffer,
NULL,
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
};
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
#ifdef __DRI_USE_INVALIDATE
static const __DRIuseInvalidateExtension dri2UseInvalidate = {
{ __DRI_USE_INVALIDATE, __DRI_USE_INVALIDATE_VERSION }
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
};
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
#endif
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
dri2: Event driven buffer validation. When a buffer invalidation event is received from the X server, the "invalidate" hook of the DRI2 flush extension is executed: A generic implementation (dri2InvalidateDrawable) is provided that just bumps the "pStamp" sequence number in __DRIdrawableRec. For old servers not supporting buffer invalidation events, the invalidate hook will be called before flushing the fake front/back buffer (that's typically once per frame -- not a lot worse than the situation we were in before). No effort has been made on preserving backwards compatibility with version 2 of the flush extension, but I think it's acceptable because AFAIK no released stack is making use of it. Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
2010-02-08 18:27:56 +00:00
_X_HIDDEN void
dri2InvalidateBuffers(Display *dpy, XID drawable)
{
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
__GLXDRIdrawable *pdraw =
dri2GetGlxDrawableFromXDrawableId(dpy, drawable);
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
struct dri2_screen *psc = (struct dri2_screen *) pdraw->psc;
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
struct dri2_drawable *pdp = (struct dri2_drawable *) pdraw;
dri2: Event driven buffer validation. When a buffer invalidation event is received from the X server, the "invalidate" hook of the DRI2 flush extension is executed: A generic implementation (dri2InvalidateDrawable) is provided that just bumps the "pStamp" sequence number in __DRIdrawableRec. For old servers not supporting buffer invalidation events, the invalidate hook will be called before flushing the fake front/back buffer (that's typically once per frame -- not a lot worse than the situation we were in before). No effort has been made on preserving backwards compatibility with version 2 of the flush extension, but I think it's acceptable because AFAIK no released stack is making use of it. Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
2010-02-08 18:27:56 +00:00
#if __DRI2_FLUSH_VERSION >= 3
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
if (pdraw && psc->f)
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
psc->f->invalidate(pdp->driDrawable);
dri2: Event driven buffer validation. When a buffer invalidation event is received from the X server, the "invalidate" hook of the DRI2 flush extension is executed: A generic implementation (dri2InvalidateDrawable) is provided that just bumps the "pStamp" sequence number in __DRIdrawableRec. For old servers not supporting buffer invalidation events, the invalidate hook will be called before flushing the fake front/back buffer (that's typically once per frame -- not a lot worse than the situation we were in before). No effort has been made on preserving backwards compatibility with version 2 of the flush extension, but I think it's acceptable because AFAIK no released stack is making use of it. Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
2010-02-08 18:27:56 +00:00
#endif
}
glx: Split tfp functions out to context vtable This introduces a new per-context vtable, which lets us clean up all the #ifdef's a bit and move the DRI2 specific implementation into dri2_glx.c.
2010-05-21 15:36:56 +01:00
static void
dri2_bind_tex_image(Display * dpy,
GLXDrawable drawable,
int buffer, const int *attrib_list)
{
GLXContext gc = __glXGetCurrentContext();
glx: Correctly look up the dri2 context pointer for SetTexBuffer. gc->driContext points at the second member of the dri2 context. The dri2 context is just a subclass of the GLX context. Fixes piglit tfp testcase.
2010-07-23 22:51:06 +01:00
struct dri2_context *pcp = (struct dri2_context *) gc;
glx: Drop screen argument to GetGLXDRIDrawable We'll just get it from the returned drawable when we need it.
2010-07-23 03:36:37 +01:00
__GLXDRIdrawable *base = GetGLXDRIDrawable(dpy, drawable);
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
__GLXdisplayPrivate *dpyPriv = __glXInitialize(dpy);
struct dri2_drawable *pdraw = (struct dri2_drawable *) base;
struct dri2_display *pdp =
(struct dri2_display *) dpyPriv->dri2Display;
struct dri2_screen *psc = (struct dri2_screen *) base->psc;
glx: Split tfp functions out to context vtable This introduces a new per-context vtable, which lets us clean up all the #ifdef's a bit and move the DRI2 specific implementation into dri2_glx.c.
2010-05-21 15:36:56 +01:00
if (pdraw != NULL) {
glx: Invalidate drawable in glXBindTexImageEXT() if we don't get events When we don't get invalidate events, we need to invalidate a drawable before using it for tfp to make sure we have uptodate buffers.
2010-05-21 16:51:33 +01:00
#if __DRI2_FLUSH_VERSION >= 3
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
if (!pdp->invalidateAvailable && psc->f)
psc->f->invalidate(pdraw->driDrawable);
glx: Invalidate drawable in glXBindTexImageEXT() if we don't get events When we don't get invalidate events, we need to invalidate a drawable before using it for tfp to make sure we have uptodate buffers.
2010-05-21 16:51:33 +01:00
#endif
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
if (psc->texBuffer->base.version >= 2 &&
psc->texBuffer->setTexBuffer2 != NULL) {
glx: Move __driContext field out of __GLXcontextRec
2010-07-21 18:55:30 +01:00
(*psc->texBuffer->setTexBuffer2) (pcp->driContext,
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
pdraw->base.textureTarget,
pdraw->base.textureFormat,
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
pdraw->driDrawable);
glx: Split tfp functions out to context vtable This introduces a new per-context vtable, which lets us clean up all the #ifdef's a bit and move the DRI2 specific implementation into dri2_glx.c.
2010-05-21 15:36:56 +01:00
}
else {
glx: Move __driContext field out of __GLXcontextRec
2010-07-21 18:55:30 +01:00
(*psc->texBuffer->setTexBuffer) (pcp->driContext,
glx: Move __DRIdrawable pointers to DRI drawable privates
2010-07-19 23:11:03 +01:00
pdraw->base.textureTarget,
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
pdraw->driDrawable);
glx: Split tfp functions out to context vtable This introduces a new per-context vtable, which lets us clean up all the #ifdef's a bit and move the DRI2 specific implementation into dri2_glx.c.
2010-05-21 15:36:56 +01:00
}
}
}
static void
dri2_release_tex_image(Display * dpy, GLXDrawable drawable, int buffer)
{
}
static const struct glx_context_vtable dri2_context_vtable = {
glx: Move context destroy to context vtable
2010-07-23 04:45:18 +01:00
dri2_destroy_context,
glx: Move WaitGL, WaitX, UseXFont to context vtable functions
2010-07-23 03:24:00 +01:00
dri2_wait_gl,
dri2_wait_x,
DRI_glXUseXFont,
glx: Split tfp functions out to context vtable This introduces a new per-context vtable, which lets us clean up all the #ifdef's a bit and move the DRI2 specific implementation into dri2_glx.c.
2010-05-21 15:36:56 +01:00
dri2_bind_tex_image,
dri2_release_tex_image,
};
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
static void
dri2BindExtensions(struct dri2_screen *psc, const __DRIextension **extensions)
{
int i;
__glXEnableDirectExtension(&psc->base, "GLX_SGI_video_sync");
__glXEnableDirectExtension(&psc->base, "GLX_SGI_swap_control");
__glXEnableDirectExtension(&psc->base, "GLX_MESA_swap_control");
glx: Move DRI CopySubBuffer extension to DRI1 code We do this in the X server for DRI2.
2010-07-20 03:40:22 +01:00
__glXEnableDirectExtension(&psc->base, "GLX_SGI_make_current_read");
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
/* FIXME: if DRI2 version supports it... */
__glXEnableDirectExtension(&psc->base, "INTEL_swap_event");
for (i = 0; extensions[i]; i++) {
if ((strcmp(extensions[i]->name, __DRI_TEX_BUFFER) == 0)) {
psc->texBuffer = (__DRItexBufferExtension *) extensions[i];
__glXEnableDirectExtension(&psc->base, "GLX_EXT_texture_from_pixmap");
}
if ((strcmp(extensions[i]->name, __DRI2_FLUSH) == 0)) {
psc->f = (__DRI2flushExtension *) extensions[i];
/* internal driver extension, no GL extension exposed */
}
if ((strcmp(extensions[i]->name, __DRI2_CONFIG_QUERY) == 0))
psc->config = (__DRI2configQueryExtension *) extensions[i];
}
}
glx: Refactor and simplify context creation This lets us better separate context creation between the different backends.
2010-07-23 21:15:31 +01:00
static const struct glx_screen_vtable dri2_screen_vtable = {
dri2_create_context
};
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
static __GLXscreenConfigs *
dri2CreateScreen(int screen, __GLXdisplayPrivate * priv)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
const __DRIconfig **driver_configs;
const __DRIextension **extensions;
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
const struct dri2_display *const pdp = (struct dri2_display *)
DRI2: Implement interface for drivers to access DRI2GetBuffersWithFormat Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> Reviewed-by: Kristian Høgsberg <krh@redhat.com>
2009-04-21 04:55:56 +01:00
priv->dri2Display;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
struct dri2_screen *psc;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
__GLXDRIscreen *psp;
char *driverName, *deviceName;
drm_magic_t magic;
int i;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
psc = Xmalloc(sizeof *psc);
if (psc == NULL)
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return NULL;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
memset(psc, 0, sizeof *psc);
if (!glx_screen_init(&psc->base, screen, priv))
return NULL;
if (!DRI2Connect(priv->dpy, RootWindow(priv->dpy, screen),
glx: fix possible memory leaks in dri2CreateScreen() Original patch fixed up by Brian Paul.
2010-01-29 15:36:47 +00:00
&driverName, &deviceName)) {
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
XFree(psc);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return NULL;
glx: fix possible memory leaks in dri2CreateScreen() Original patch fixed up by Brian Paul.
2010-01-29 15:36:47 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
psc->driver = driOpenDriver(driverName);
if (psc->driver == NULL) {
ErrorMessageF("driver pointer missing\n");
goto handle_error;
}
extensions = dlsym(psc->driver, __DRI_DRIVER_EXTENSIONS);
if (extensions == NULL) {
ErrorMessageF("driver exports no extensions (%s)\n", dlerror());
goto handle_error;
}
for (i = 0; extensions[i]; i++) {
if (strcmp(extensions[i]->name, __DRI_CORE) == 0)
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
psc->core = (__DRIcoreExtension *) extensions[i];
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
if (strcmp(extensions[i]->name, __DRI_DRI2) == 0)
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
psc->dri2 = (__DRIdri2Extension *) extensions[i];
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
}
if (psc->core == NULL || psc->dri2 == NULL) {
ErrorMessageF("core dri or dri2 extension not found\n");
goto handle_error;
}
psc->fd = open(deviceName, O_RDWR);
if (psc->fd < 0) {
ErrorMessageF("failed to open drm device: %s\n", strerror(errno));
glx: fix possible memory leaks in dri2CreateScreen() Original patch fixed up by Brian Paul.
2010-01-29 15:36:47 +00:00
goto handle_error;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
}
if (drmGetMagic(psc->fd, &magic)) {
ErrorMessageF("failed to get magic\n");
glx: fix possible memory leaks in dri2CreateScreen() Original patch fixed up by Brian Paul.
2010-01-29 15:36:47 +00:00
goto handle_error;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
}
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
if (!DRI2Authenticate(priv->dpy, RootWindow(priv->dpy, screen), magic)) {
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
ErrorMessageF("failed to authenticate magic %d\n", magic);
glx: fix possible memory leaks in dri2CreateScreen() Original patch fixed up by Brian Paul.
2010-01-29 15:36:47 +00:00
goto handle_error;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
}
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/* If the server does not support the protocol for
* DRI2GetBuffersWithFormat, don't supply that interface to the driver.
*/
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
psc->driScreen =
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
psc->dri2->createNewScreen(screen, psc->fd,
(const __DRIextension **)
&pdp->loader_extensions[0],
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
&driver_configs, psc);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
if (psc->driScreen == NULL) {
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
ErrorMessageF("failed to create dri screen\n");
glx: fix possible memory leaks in dri2CreateScreen() Original patch fixed up by Brian Paul.
2010-01-29 15:36:47 +00:00
goto handle_error;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
}
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
extensions = psc->core->getExtensions(psc->driScreen);
glx: Move DRI2 extensions to DRI2 screen private
2010-07-19 20:35:48 +01:00
dri2BindExtensions(psc, extensions);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
psc->base.configs =
driConvertConfigs(psc->core, psc->base.configs, driver_configs);
psc->base.visuals =
driConvertConfigs(psc->core, psc->base.visuals, driver_configs);
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
glx: Move driver_configs to DRI screen privates
2010-07-20 03:12:22 +01:00
psc->driver_configs = driver_configs;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
glx: Refactor and simplify context creation This lets us better separate context creation between the different backends.
2010-07-23 21:15:31 +01:00
psc->base.vtable = &dri2_screen_vtable;
glx: Move __driScreen into the dri screen privates
2010-07-19 21:45:23 +01:00
psp = &psc->vtable;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
psc->base.driScreen = psp;
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
psp->destroyScreen = dri2DestroyScreen;
psp->createDrawable = dri2CreateDrawable;
psp->swapBuffers = dri2SwapBuffers;
glx: permit building with older protocol headers I'd like to be able to build mesa on current distro releases without having to upgrade from the standard dri2proto and glproto headers. With this change I'm able to build on ancient releases such as Ubuntu 9-10... In general, it would be nice to be able to build-test mesa to check for unintended breakages without having to follow the external dependencies of every group working on the codebase. Seems to introduce no changes to the build of libglapi.a when tested against new versions of the headers.
2010-02-04 12:46:21 +00:00
psp->getDrawableMSC = NULL;
psp->waitForMSC = NULL;
psp->waitForSBC = NULL;
psp->setSwapInterval = NULL;
psp->getSwapInterval = NULL;
DRI2: don't send DRI2 requests the server doesn't know about At init time, NULL out the hooks that send the server new requests. DRI2 never supported these extensions, so this preserves old behavior on old servers with new Mesa.
2010-01-14 00:01:51 +00:00
if (pdp->driMinor >= 2) {
glx: permit building with older protocol headers I'd like to be able to build mesa on current distro releases without having to upgrade from the standard dri2proto and glproto headers. With this change I'm able to build on ancient releases such as Ubuntu 9-10... In general, it would be nice to be able to build-test mesa to check for unintended breakages without having to follow the external dependencies of every group working on the codebase. Seems to introduce no changes to the build of libglapi.a when tested against new versions of the headers.
2010-02-04 12:46:21 +00:00
#ifdef X_DRI2GetMSC
DRI2: don't send DRI2 requests the server doesn't know about At init time, NULL out the hooks that send the server new requests. DRI2 never supported these extensions, so this preserves old behavior on old servers with new Mesa.
2010-01-14 00:01:51 +00:00
psp->getDrawableMSC = dri2DrawableGetMSC;
glx: permit building with older protocol headers I'd like to be able to build mesa on current distro releases without having to upgrade from the standard dri2proto and glproto headers. With this change I'm able to build on ancient releases such as Ubuntu 9-10... In general, it would be nice to be able to build-test mesa to check for unintended breakages without having to follow the external dependencies of every group working on the codebase. Seems to introduce no changes to the build of libglapi.a when tested against new versions of the headers.
2010-02-04 12:46:21 +00:00
#endif
#ifdef X_DRI2WaitMSC
DRI2: don't send DRI2 requests the server doesn't know about At init time, NULL out the hooks that send the server new requests. DRI2 never supported these extensions, so this preserves old behavior on old servers with new Mesa.
2010-01-14 00:01:51 +00:00
psp->waitForMSC = dri2WaitForMSC;
psp->waitForSBC = dri2WaitForSBC;
glx: permit building with older protocol headers I'd like to be able to build mesa on current distro releases without having to upgrade from the standard dri2proto and glproto headers. With this change I'm able to build on ancient releases such as Ubuntu 9-10... In general, it would be nice to be able to build-test mesa to check for unintended breakages without having to follow the external dependencies of every group working on the codebase. Seems to introduce no changes to the build of libglapi.a when tested against new versions of the headers.
2010-02-04 12:46:21 +00:00
#endif
#ifdef X_DRI2SwapInterval
DRI2: don't send DRI2 requests the server doesn't know about At init time, NULL out the hooks that send the server new requests. DRI2 never supported these extensions, so this preserves old behavior on old servers with new Mesa.
2010-01-14 00:01:51 +00:00
psp->setSwapInterval = dri2SetSwapInterval;
psp->getSwapInterval = dri2GetSwapInterval;
DRI2: expose OML_sync_control if possible If the server supports the OML related protocol, enable support for the extension.
2010-03-05 19:17:19 +00:00
#endif
#if defined(X_DRI2GetMSC) && defined(X_DRI2WaitMSC) && defined(X_DRI2SwapInterval)
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
__glXEnableDirectExtension(&psc->base, "GLX_OML_sync_control");
glx: permit building with older protocol headers I'd like to be able to build mesa on current distro releases without having to upgrade from the standard dri2proto and glproto headers. With this change I'm able to build on ancient releases such as Ubuntu 9-10... In general, it would be nice to be able to build-test mesa to check for unintended breakages without having to follow the external dependencies of every group working on the codebase. Seems to introduce no changes to the build of libglapi.a when tested against new versions of the headers.
2010-02-04 12:46:21 +00:00
#endif
DRI2: don't send DRI2 requests the server doesn't know about At init time, NULL out the hooks that send the server new requests. DRI2 never supported these extensions, so this preserves old behavior on old servers with new Mesa.
2010-01-14 00:01:51 +00:00
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/* DRI2 suports SubBuffer through DRI2CopyRegion, so it's always
* available.*/
psp->copySubBuffer = dri2CopySubBuffer;
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
__glXEnableDirectExtension(&psc->base, "GLX_MESA_copy_sub_buffer");
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(driverName);
Xfree(deviceName);
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
return &psc->base;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
DRI2: add OML_sync_control support Add OML_sync_control support, along with a simple program for testing it. This means adding support for the DRI2GetMSC, DRI2WaitMSC and DRI2WaitSBC requests. Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
2009-09-16 07:23:09 +01:00
handle_error:
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(driverName);
Xfree(deviceName);
glx: Add screen privates for dri drivers and moved some fields there GLXscreenConfigs is badly named and a dumping ground for a lot of stuff. This patch creates private screen structs for the dri drivers and moves some of their fields over there.
2010-07-19 19:57:59 +01:00
XFree(psc);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
/* FIXME: clean up here */
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return NULL;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
/* Called from __glXFreeDisplayPrivate.
*/
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
static void
dri2DestroyDisplay(__GLXDRIdisplay * dpy)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
Xfree(dpy);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
_X_HIDDEN __GLXDRIdrawable *
dri2GetGlxDrawableFromXDrawableId(Display *dpy, XID id)
{
__GLXdisplayPrivate *d = __glXInitialize(dpy);
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_display *pdp = (struct dri2_display *) d->dri2Display;
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
__GLXDRIdrawable *pdraw;
if (__glxHashLookup(pdp->dri2Hash, id, (void *) &pdraw) == 0)
return pdraw;
return NULL;
}
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
/*
* Allocate, initialize and return a __DRIdisplayPrivate object.
* This is called from __glXInitialize() when we are given a new
* display pointer.
*/
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
_X_HIDDEN __GLXDRIdisplay *
dri2CreateDisplay(Display * dpy)
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
{
glx: Rename various DRI structs away from obnoxious __GLXfooRec convention Enough is enough.
2010-07-19 15:51:09 +01:00
struct dri2_display *pdp;
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
int eventBase, errorBase, i;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
if (!DRI2QueryExtension(dpy, &eventBase, &errorBase))
return NULL;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdp = Xmalloc(sizeof *pdp);
if (pdp == NULL)
return NULL;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
if (!DRI2QueryVersion(dpy, &pdp->driMajor, &pdp->driMinor)) {
Xfree(pdp);
return NULL;
}
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdp->driPatch = 0;
dri2: Event driven buffer validation. When a buffer invalidation event is received from the X server, the "invalidate" hook of the DRI2 flush extension is executed: A generic implementation (dri2InvalidateDrawable) is provided that just bumps the "pStamp" sequence number in __DRIdrawableRec. For old servers not supporting buffer invalidation events, the invalidate hook will be called before flushing the fake front/back buffer (that's typically once per frame -- not a lot worse than the situation we were in before). No effort has been made on preserving backwards compatibility with version 2 of the flush extension, but I think it's acceptable because AFAIK no released stack is making use of it. Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
2010-02-08 18:27:56 +00:00
pdp->swapAvailable = (pdp->driMinor >= 2);
pdp->invalidateAvailable = (pdp->driMinor >= 3);
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
pdp->base.destroyDisplay = dri2DestroyDisplay;
pdp->base.createScreen = dri2CreateScreen;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
i = 0;
if (pdp->driMinor < 1)
pdp->loader_extensions[i++] = &dri2LoaderExtension_old.base;
else
pdp->loader_extensions[i++] = &dri2LoaderExtension.base;
pdp->loader_extensions[i++] = &systemTimeExtension.base;
#ifdef __DRI_USE_INVALIDATE
dri2_glx: Put the invalidate b/c code back in The backwards compatibility code calls the DRI driver invalidate hook on swap buffer and flush front buffer. This lets the DRI driver rely on invalidate callbacks and drop the glViewport() hack, even if the server doesn't send invalidate events. This is essentially a revert of 2d00d16da7f5d2255cb37b48edaf4cbb9ca7e930, except that we now also pass the __DRI_USE_INVALIDATE extension even when the server doesn't have DRI2 invalidate events.
2010-05-18 19:45:10 +01:00
pdp->loader_extensions[i++] = &dri2UseInvalidate.base;
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
#endif
dri2_glx: Terminate loader extension list outside #ifdef
2010-05-18 19:02:57 +01:00
pdp->loader_extensions[i++] = NULL;
glx: Provide the __DRI_USE_INVALIDATE extension to the driver when we can When we have DRI2 protocol at least 2.3, we get an event from the server when the back buffers get invalidated. When that's the case let the driver know that it can rely on invalidate instead of the glViewport polling.
2010-05-11 14:42:57 +01:00
glx: Fix drawable lookup in DRI2 event handler DRI2 events are sent to the X drawable ID used to create the DRI2 drawable, not the GLX drawable ID. So when an event comes in, we need to look up the __GLXDRIdrawable by its X drawable ID, which needs a new hash table.
2010-07-19 14:37:07 +01:00
pdp->dri2Hash = __glxHashCreate();
if (pdp->dri2Hash == NULL) {
Xfree(pdp);
return NULL;
}
glx: indent -br -i3 -npcs --no-tabs Some manual intervention applied since XEXT_* and other macro magic fooled indent. Auto generated files were also skipped.
2009-08-12 11:41:22 +01:00
return &pdp->base;
DRI interface changes and DRI2 direct rendering support. Add DRI2 direct rendering support to libGL and add DRI2 client side protocol code. Extend the GLX 1.3 create drawable functions in glx_pbuffer.c to call into the DRI driver when possible. Introduce __DRIconfig, opaque struct that represents a DRI driver configuration. Get's rid of the open coded __GLcontextModes in the DRI driver interface and the context modes create and destroy functions that the loader was requires to provide. glcore.h is no longer part of the DRI driver interface. The DRI config is GL binding agnostic, that is, not specific to GLX, EGL or other bindings. The core API is now also an extension, and the driver exports a list of extensions as the symbol __driDriverExtensions, which the loader must dlsym() for. The list of extension will always include the DRI core extension, which allows creating and manipulating DRI screens, drawables and contexts. The DRI legacy extension, when available, provides alternative entry points for creating the DRI objects that work with the XF86DRI infrastructure. Change DRI2 client code to not use drm drawables or contexts. We never used drm_drawable_t's and the only use for drm_context_t was as a unique identifier when taking the lock. We now just allocate a unique lock ID out of the DRILock sarea block. Once we get rid of the lock entirely, we can drop this hack. Change the interface between dri_util.c and the drivers, so that the drivers now export the DriverAPI struct as driDriverAPI instead of the InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2 init screen function to see if DRI2 is supported by the driver.
2008-03-26 23:26:59 +00:00
}
#endif /* GLX_DIRECT_RENDERING */