[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gnash-dev] Hardware acceleration support
From: |
Adrian Panasiuk |
Subject: |
Re: [Gnash-dev] Hardware acceleration support |
Date: |
Fri, 12 Mar 2010 20:57:08 +0100 |
The guis are still trying to link against librender/libgnashagg.la,
librender/libgnashcairo.la and librender/libgnashogl.la - I've checked
that symlinking librender/libgnashrender.la to those names allows
building working binaries.
A single build for all GUIs is not implemented right now, right?
I'm attaching a patch to help build when VAAPI or XV is not enabled:
=== modified file 'gui/gtk.cpp'
--- gui/gtk.cpp 2010-03-02 23:26:48 +0000
+++ gui/gtk.cpp 2010-03-12 19:42:40 +0000
@@ -39,6 +39,7 @@
#include "va/va.h"
#include "va/va_backend.h"
#include "vaapi_utils.h"
+extern VAStatus va_getDriverName(VADisplay dpy, char **driver_name);
#endif
#ifdef HAVE_X11
@@ -78,8 +79,6 @@
# include <hildon/hildon.h>
#endif
-extern VAStatus va_getDriverName(VADisplay dpy, char **driver_name);
-
namespace gnash
{
=== modified file 'gui/gtk_canvas.cpp'
--- gui/gtk_canvas.cpp 2010-03-10 20:54:03 +0000
+++ gui/gtk_canvas.cpp 2010-03-12 19:47:05 +0000
@@ -233,8 +233,10 @@
while (!initialized_renderer) {
renderer = next_renderer;
hwaccel = next_hwaccel;
+#ifdef USE_VAAPI
// Global enable VA-API, if requested
gnash::vaapi_set_is_enabled(hwaccel == "vaapi");
+#endif
// Use the Cairo renderer. Cairo is also used by GTK2, so using
// Cairo makes much sense. Unfortunately, our implementation seems
// to have serious performance issues, although it does work.
@@ -257,6 +259,7 @@
// rendering performance issues.
}
else if (renderer == "agg") {
+#ifdef USE_VAAPI
// Use LibVva, which works on Nvidia, AT, or Intel 965 GPUs
// with AGG or OpenGL.
if (hwaccel == "vaapi") {
@@ -265,14 +268,19 @@
// if initializing fails.
next_hwaccel = "xv";
}
- else if (hwaccel == "xv") {
+ else
+#endif
+#ifdef HAVE_XV
+ if (hwaccel == "xv") {
// Use the X11 XV extension, which works on most GPUs.
canvas->glue.reset(new gnash::GtkAggXvGlue);
// Set the hardware acclerator to the next one to try
// if initializing fails.
next_hwaccel = "none";
}
- else {
+ else
+#endif
+ {
canvas->glue.reset(new gnash::GtkAggGlue);
}
}
- Re: [Gnash-dev] Hardware acceleration support, (continued)
- Re: [Gnash-dev] Hardware acceleration support, Benjamin Wolsey, 2010/03/02
- Re: [Gnash-dev] Hardware acceleration support, Benjamin Wolsey, 2010/03/02
- Re: [Gnash-dev] Hardware acceleration support, strk, 2010/03/06
- Re: [Gnash-dev] Hardware acceleration support, Rob Savoye, 2010/03/07
- Re: [Gnash-dev] Hardware acceleration support,
Adrian Panasiuk <=
- Re: [Gnash-dev] Hardware acceleration support, Rob Savoye, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, Gwenole Beauchesne, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, Rob Savoye, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, Gwenole Beauchesne, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, strk, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, Rob Savoye, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, strk, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, Rob Savoye, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, strk, 2010/03/14
- Re: [Gnash-dev] Hardware acceleration support, strk, 2010/03/14