Re: [Mesa-dev] Should we need one release function for DRI's GLX_EXT_texture_from_pixmap?
On Wed, 2011-01-05 at 08:28 -0500, Kristian Høgsberg wrote:
On Tue, Jan 4, 2011 at 11:10 PM, Zhao, Juan J juan.j.z...@intel.com wrote:
Hi all,
In the structure “__DRItexBufferExtensionRec”, we don’t have
one release interface now. But in our platform, we need to release some
resources.
Why we don’t need the release interface? Or should we add
one?
In the open source drivers, glXBindTexImageEXT is pretty much the same
as glBindTexture. The pixmaps stays bound until you bind another
texture or pixmap, at which point all the resources are released.
There is nothing for the open source drivers to do in release, so the
DRI extension never had a release function. If you need a release
function, just send a patch and we can add it. You'll need to bump
the extension version number and then add the call to release in the
dri loaders (libGL, AIGLX, egl_dri2) conditional on the extension
version.
Thanks a lot! :)
I add this interface. Would you please help to check it?
Kristian
--
*^_^*
Many thanks Best Regards
SSD-OTC Meego Middleware TV Team
Zhao Juan
From da404aef0ec3e1eba47c4754424020f2484f7510 Mon Sep 17 00:00:00 2001
From: JuanZhao juan.j.z...@intel.com
Date: Thu, 6 Jan 2011 14:59:43 -0500
Subject: [PATCH] Add release function for texture_from_pixmap extension
Some platform need to release texture image for texture_from_pixmap extension, add this interface for those platforms.
---
include/GL/internal/dri_interface.h | 12 ++
src/egl/drivers/dri2/egl_dri2.c | 40 +++---
src/glx/dri2_glx.c | 27 +++
3 files changed, 75 insertions(+), 4 deletions(-)
diff --git a/include/GL/internal/dri_interface.h b/include/GL/internal/dri_interface.h
index 9ee039b..84a43ff 100644
--- a/include/GL/internal/dri_interface.h
+++ b/include/GL/internal/dri_interface.h
@@ -251,6 +251,18 @@ struct __DRItexBufferExtensionRec {
GLint target,
GLint format,
__DRIdrawable *pDraw);
+#if __DRI_TEX_BUFFER_VERSION =3
+/**
+ * Method to release texture buffer in case some special platform
+ * need this.
+ *
+ * For GLX_EXT_texture_from_pixmap with AIGLX.
+ */
+void (*releaseTexBuffer)(__DRIcontext *pDRICtx,
+ GLint target,
+ GLint format,
+ __DRIdrawable *pDraw);
+#endif
};
/**
diff --git a/src/egl/drivers/dri2/egl_dri2.c b/src/egl/drivers/dri2/egl_dri2.c
index 6f40ab9..1319bd0 100644
--- a/src/egl/drivers/dri2/egl_dri2.c
+++ b/src/egl/drivers/dri2/egl_dri2.c
@@ -1983,10 +1983,42 @@ static EGLBoolean
dri2_release_tex_image(_EGLDriver *drv,
_EGLDisplay *disp, _EGLSurface *surf, EGLint buffer)
{
- (void) drv;
- (void) disp;
- (void) surf;
- (void) buffer;
+#if __DRI_TEX_BUFFER_VERSION = 3
+ struct dri2_egl_display *dri2_dpy = dri2_egl_display(disp);
+ struct dri2_egl_surface *dri2_surf = dri2_egl_surface(surf);
+ struct dri2_egl_context *dri2_ctx;
+ _EGLContext *ctx;
+ GLint format, target;
+
+ ctx = _eglGetCurrentContext();
+ dri2_ctx = dri2_egl_context(ctx);
+
+ if (!_eglReleaseTexImage(drv, disp, surf, buffer))
+ return EGL_FALSE;
+
+ switch (dri2_surf-base.TextureFormat) {
+ case EGL_TEXTURE_RGB:
+ format = __DRI_TEXTURE_FORMAT_RGB;
+ break;
+ case EGL_TEXTURE_RGBA:
+ format = __DRI_TEXTURE_FORMAT_RGBA;
+ break;
+ default:
+ assert(0);
+ }
+
+ switch (dri2_surf-base.TextureTarget) {
+ case EGL_TEXTURE_2D:
+ target = GL_TEXTURE_2D;
+ break;
+ default:
+ assert(0);
+ }
+
+ (*dri2_dpy-tex_buffer-releaseTexBuffer)(dri2_ctx-dri_context,
+ target, format,
+ dri2_surf-dri_drawable);
+#endif
return EGL_TRUE;
}
diff --git a/src/glx/dri2_glx.c b/src/glx/dri2_glx.c
index b0559b2..f96265b 100644
--- a/src/glx/dri2_glx.c
+++ b/src/glx/dri2_glx.c
@@ -719,6 +719,33 @@ dri2_bind_tex_image(Display * dpy,
static void
dri2_release_tex_image(Display * dpy, GLXDrawable drawable, int buffer)
{
+#if __DRI_TEX_BUFFER_VERSION = 3
+ struct glx_context *gc = __glXGetCurrentContext();
+ struct dri2_context *pcp = (struct dri2_context *) gc;
+ __GLXDRIdrawable *base = GetGLXDRIDrawable(dpy, drawable);
+ struct glx_display *dpyPriv = __glXInitialize(dpy);
+ struct dri2_drawable *pdraw = (struct dri2_drawable *) base;
+ struct dri2_display *pdp =
+ (struct dri2_display *) dpyPriv-dri2Display;
+ struct dri2_screen *psc;
+
+ if (pdraw != NULL) {
+ psc = (struct dri2_screen *) base-psc;
+
+#if __DRI2_FLUSH_VERSION = 3
+ if (!pdp-invalidateAvailable psc-f)
+ psc-f-invalidate(pdraw-driDrawable);
+#endif
+
+ if (psc-texBuffer-base.version = 3
+ psc-texBuffer-releaseTexBuffer != NULL) {
+ (*psc-texBuffer-releaseTexBuffer) (pcp-driContext,
+
[Mesa-dev] [Bug 32825] egl_glx driver completely broken in 7.9 branch [fix in master]
https://bugs.freedesktop.org/show_bug.cgi?id=32825 --- Comment #4 from nobled nob...@dreamwidth.org 2011-01-06 06:45:04 PST --- (In reply to comment #3) Did you try a clean build? I don't see why the fix would make 0x3034 have the value 0x3 or 0x4. Yeah, a fresh git checkout: git checkout 7.9 git clean -xdf autoreconf ./configure --disable-glw --disable-glut --disable-glu --disable-gallium-i915 --disable-gallium-i965 --disable-gallium-radeon --disable-gallium-svga --enable-xcb cd src/egl make I just needed to rebuild the egl_glx.so, so I just did that. That shouldn't make a difference, right? -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are the assignee for the bug. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 32825] egl_glx driver completely broken in 7.9 branch [fix in master]
https://bugs.freedesktop.org/show_bug.cgi?id=32825 --- Comment #5 from Chia-I Wu olva...@gmail.com 2011-01-06 06:54:49 PST --- You also need to rebuild libEGL (src/egl/main/). Could you see if it helps? -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are the assignee for the bug. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH 1/7] u_upload_mgr: new features
Looks great. Thanks. -Brian On Wed, Jan 5, 2011 at 4:41 PM, Marek Olšák mar...@gmail.com wrote: Please see the attached patch. It also documents util_copy_vertex_buffers, as you asked for in another email. Marek (re-sending this, the original email didn't make it to ML because of the 40kB limit) On Mon, Jan 3, 2011 at 4:43 PM, Brian Paul bri...@vmware.com wrote: Could you also add comments to all the u_upload_mgr.c functions? In particular, what are the possible values for the usage/bind parameters? A follow-on patch is fine. I know that comments have been missing all along. -Brian ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH mesa-demos 1/6] es1_info: convert indentString into a literal string
This fixes compilation with -Wformat -Werror=format-security. Some
distros like Mandriva enable this flag by default. Its purpose is to
improve security.
Another option for this patch would be to do
printf(%s, indentString), but converting indentString into a literal
also gives the compiler some hints to improve performance.
Signed-off-by: Paulo Zanoni pzan...@mandriva.com
---
By the way, combining this patch with a printf(%s, indentString) would
make the code even safer. The last patch of this series does this change, so
you can choose to apply it or not.
Using printf(string); is dangerous, might lead to bugs and even
security issues. If the string being printed contains the % character
one can do really dangerous things. Even if you think the string in
question might not be dangerous, future code changes might lead that
piece of code to bugs or security holes.
Some references:
http://wiki.mandriva.com/en/Development/Packaging/Problems#format_not_a_string_literal_and_no_format_arguments
http://wiki.debian.org/Hardening#DEBBUILDHARDENINGFORMAT.28gcc.2BAC8-g.2B-.2B--Wformat-Wformat-security.29
http://en.wikipedia.org/wiki/Format_string_attack
See also How To Write Shared Libraries Section 2.4.1, written by Ulrich
Drepper for more information on the difference between 'char *foo = bar'
and 'char foo[] = bar'
src/egl/opengles1/es1_info.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/src/egl/opengles1/es1_info.c b/src/egl/opengles1/es1_info.c
index 93816b5..38becc5 100644
--- a/src/egl/opengles1/es1_info.c
+++ b/src/egl/opengles1/es1_info.c
@@ -29,7 +29,7 @@
static void
print_extension_list(const char *ext)
{
- const char *indentString = ;
+ const char indentString[] = ;
const int indent = 4;
const int max = 79;
int width, i, j;
--
1.7.1
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH mesa-demos 2/6] Add missing binaries to .gitignore
Signed-off-by: Paulo Zanoni pzan...@mandriva.com --- src/demos/.gitignore |1 + src/egl/opengles1/.gitignore |1 + src/glsl/.gitignore |1 + src/gs/.gitignore|1 + src/tests/.gitignore |1 + src/trivial/.gitignore |4 6 files changed, 9 insertions(+), 0 deletions(-) create mode 100644 src/gs/.gitignore diff --git a/src/demos/.gitignore b/src/demos/.gitignore index bf71667..807033d 100644 --- a/src/demos/.gitignore +++ b/src/demos/.gitignore @@ -32,6 +32,7 @@ lodbias morph3d multiarb paltex +pixeltest pointblast projtex rain diff --git a/src/egl/opengles1/.gitignore b/src/egl/opengles1/.gitignore index 135e3de..3263ff2 100644 --- a/src/egl/opengles1/.gitignore +++ b/src/egl/opengles1/.gitignore @@ -1,4 +1,5 @@ bindtex +clear drawtex_x11 drawtex_screen es1_info diff --git a/src/glsl/.gitignore b/src/glsl/.gitignore index 2b7f84f..44c6249 100644 --- a/src/glsl/.gitignore +++ b/src/glsl/.gitignore @@ -1,4 +1,5 @@ array +bezier bitmap brick bump diff --git a/src/gs/.gitignore b/src/gs/.gitignore new file mode 100644 index 000..5df2e95 --- /dev/null +++ b/src/gs/.gitignore @@ -0,0 +1 @@ +gs-tri diff --git a/src/tests/.gitignore b/src/tests/.gitignore index cb3b5fe..b6d112e 100644 --- a/src/tests/.gitignore +++ b/src/tests/.gitignore @@ -59,6 +59,7 @@ mipgen mipmap_comp mipmap_comp_tests mipmap_limits +mipmap_tunnel mipmap_view multipal multitexarray diff --git a/src/trivial/.gitignore b/src/trivial/.gitignore index 1787f27..b14de32 100644 --- a/src/trivial/.gitignore +++ b/src/trivial/.gitignore @@ -1,4 +1,5 @@ clear +clear-color clear-fbo clear-fbo-scissor clear-fbo-tex @@ -34,6 +35,7 @@ line-userclip-clip line-userclip-nop line-userclip-nop-clip line-wide +line-xor lineloop lineloop-clip lineloop-elts @@ -96,6 +98,7 @@ tri-cull tri-cull-both tri-dlist tri-edgeflag +tri-edgeflag-array tri-fbo tri-fbo-tex tri-fbo-tex-mip @@ -119,6 +122,7 @@ tri-square tri-stencil tri-stipple tri-tex +tri-tex-1d tri-tex-3d tri-tri tri-unfilled -- 1.7.1 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH mesa-demos 3/6] opengles2/tri.c: remove trailing white spaces
So git won't complain when you apply the next patch. Also, code looks prettier this way. Signed-off-by: Paulo Zanoni pzan...@mandriva.com --- src/egl/opengles2/tri.c |4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/egl/opengles2/tri.c b/src/egl/opengles2/tri.c index 8981d8a..df2751b 100644 --- a/src/egl/opengles2/tri.c +++ b/src/egl/opengles2/tri.c @@ -1,5 +1,5 @@ /** - * + * * Copyright 2008 Tungsten Graphics, Inc., Cedar Park, Texas. * All Rights Reserved. * @@ -416,7 +416,7 @@ usage(void) printf( -display displayname set the display to run on\n); printf( -info display OpenGL renderer info\n); } - + int main(int argc, char *argv[]) -- 1.7.1 ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH mesa-demos 6/6] es1_info: change printf(indentString) to printf(%s, indentString)
IMHO, code looks safer this way.
Signed-off-by: Paulo Zanoni pzan...@mandriva.com
---
Please read comments from patch 0001.
src/egl/opengles1/es1_info.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/egl/opengles1/es1_info.c b/src/egl/opengles1/es1_info.c
index 38becc5..e33597f 100644
--- a/src/egl/opengles1/es1_info.c
+++ b/src/egl/opengles1/es1_info.c
@@ -38,7 +38,7 @@ print_extension_list(const char *ext)
return;
width = indent;
- printf(indentString);
+ printf(%s, indentString);
i = j = 0;
while (1) {
if (ext[j] == ' ' || ext[j] == 0) {
@@ -48,7 +48,7 @@ print_extension_list(const char *ext)
/* start a new line */
printf(\n);
width = indent;
-printf(indentString);
+printf(%s, indentString);
}
/* print the extension name between ext[i] and ext[j] */
while (i j) {
--
1.7.1
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 31940] [r300g] Crash in dri2_invalidate_drawable
https://bugs.freedesktop.org/show_bug.cgi?id=31940 --- Comment #14 from Pavel Ondračka dra...@centrum.cz 2011-01-06 08:49:45 PST --- Any chance the patch from comment 7 can make it into master or at least to mesa-dev for review? It fixes crash in Star Trek Armada 2 for me and I have not encountered any regressions so far. Maybe also reassigning to Mesa core would be appropriate? -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are the assignee for the bug. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Upcoming Mesa 7.9.1 and 7.10 releases
On Thu, 06 Jan 2011 11:32:54 +0100 Philipp Klaus Krause p...@spth.de wrote: Am 06.01.2011 07:14, schrieb Arthur Huillet: It hasn't been committed into master yet but I think the patch at https://bugs.freedesktop.org/show_bug.cgi?id=32579 should go into 7.9.1 and 7.10. Failing that, the problem might be present on e.g. the current Ubuntu forever. What about applications that create lots of narrow, long, but still small textures? For those we would still do tiling, and use 1 MB for each 1x1024 texture. How about changing if (width0 1024 height0 1024) to if (width0 * height0 1024 * 1024) Yes, this would work as well. I don't care too much about which variant is committed as long as Mesa's upcoming releases aren't affected by the problem I reported (with a use case of textures being more-or-less square). -- Greetings, A. Huillet ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 32879] New: egl_glx driver segfaults with NULL config
https://bugs.freedesktop.org/show_bug.cgi?id=32879
Summary: egl_glx driver segfaults with NULL config
Product: Mesa
Version: git
Platform: All
OS/Version: All
Status: NEW
Severity: normal
Priority: medium
Component: Other
AssignedTo: mesa-dev@lists.freedesktop.org
ReportedBy: nob...@dreamwidth.org
egl_dri2 seems to handle this fine, but egl_glx tries to dereference it in this
function:
static int
GLX_egl_config_index(_EGLConfig *conf)
{
struct GLX_egl_config *GLX_conf = GLX_egl_config(conf);
return GLX_conf-index;
}
So it segfaults when it tries to create a context on this line in
GLX_eglCreateContext:
GLX_ctx-context =
glXCreateNewContext(GLX_dpy-dpy,
GLX_dpy-fbconfigs[GLX_egl_config_index(conf)],
GLX_RGBA_TYPE,
GLX_ctx_shared ? GLX_ctx_shared-context : NULL,
GL_TRUE);
--
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assignee for the bug.
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Upcoming Mesa 7.9.1 and 7.10 releases
On Mon, Jan 3, 2011 at 8:27 PM, Ian Romanick i...@freedesktop.org wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 It looks like things are shaping up nicely for the 7.9.1 and 7.10 releases. I've cherry picked a bunch of stuff over to the branches, but there are a few commits that I'm a bit uncertain about. I'd appreciate it if the commiters could weigh in. If you are on direct CC, that means you have a commit on one or both lists. NOTE: There are a few recent commits that are on both lists. 7.9.1 list: commit 1d33e940d2050f3d9180019f6ffd57f6fc295507 Author: Kristian Høgsberg k...@bitplanet.net Date: Thu Oct 14 14:57:47 2010 -0400 Only install vtxfmt tables for OpenGL R GLES1 and GLES2 install their own exec pointers and don't need the Save table. Also, the SET_* macros use different indices for the different APIs so the offsets used in vtxfmt.c are actually wrong for the ES APIs. Yes, this one fixes vertex attributes for GLES2 and I'd recommend 9f0d7dd25941b2f327c7bedf7513eee871829073 and 98ee6739d97b5592a7dad0b453c78e180a51ad50 as well. Kristian ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [PATCH] r600g: Fixed SIN/COS/SCS for the case where the operand is a literal.
Signed-off-by: Tilman Sauerbeck til...@code-monkey.de
---
src/gallium/drivers/r600/r600_shader.c | 17 +++--
1 files changed, 15 insertions(+), 2 deletions(-)
diff --git a/src/gallium/drivers/r600/r600_shader.c
b/src/gallium/drivers/r600/r600_shader.c
index bb5038c..35a4cea 100644
--- a/src/gallium/drivers/r600/r600_shader.c
+++ b/src/gallium/drivers/r600/r600_shader.c
@@ -948,7 +948,7 @@ static int tgsi_setup_trig(struct r600_shader_ctx *ctx,
struct r600_bc_alu_src r600_src[3])
{
struct tgsi_full_instruction *inst =
ctx-parse.FullToken.FullInstruction;
- int r;
+ int r, src0_chan;
uint32_t lit_vals[4];
struct r600_bc_alu alu;
@@ -960,6 +960,19 @@ static int tgsi_setup_trig(struct r600_shader_ctx *ctx,
if (r)
return r;
+ src0_chan = tgsi_chan(inst-Src[0], 0);
+
+ /* We are going to feed two literals to the MAD below,
+* which means that if the first operand is a literal as well,
+* we need to copy its value manually.
+*/
+ if (r600_src[0].sel == V_SQ_ALU_SRC_LITERAL) {
+ unsigned index = inst-Src[0].Register.Index;
+
+ lit_vals[2] = ctx-literals[index * 4 + src0_chan];
+ src0_chan = 2;
+ }
+
lit_vals[0] = fui(1.0 /(3.1415926535 * 2));
lit_vals[1] = fui(0.5f);
@@ -972,7 +985,7 @@ static int tgsi_setup_trig(struct r600_shader_ctx *ctx,
alu.dst.write = 1;
alu.src[0] = r600_src[0];
- alu.src[0].chan = tgsi_chan(inst-Src[0], 0);
+ alu.src[0].chan = src0_chan;
alu.src[1].sel = V_SQ_ALU_SRC_LITERAL;
alu.src[1].chan = 0;
--
1.7.3.1
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Upcoming Mesa 7.9.1 and 7.10 releases
2011/1/6 Kristian Høgsberg k...@bitplanet.net: On Mon, Jan 3, 2011 at 8:27 PM, Ian Romanick i...@freedesktop.org wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 It looks like things are shaping up nicely for the 7.9.1 and 7.10 releases. I've cherry picked a bunch of stuff over to the branches, but there are a few commits that I'm a bit uncertain about. I'd appreciate it if the commiters could weigh in. If you are on direct CC, that means you have a commit on one or both lists. NOTE: There are a few recent commits that are on both lists. 7.9.1 list: commit 1d33e940d2050f3d9180019f6ffd57f6fc295507 Author: Kristian Høgsberg k...@bitplanet.net Date: Thu Oct 14 14:57:47 2010 -0400 Only install vtxfmt tables for OpenGL R GLES1 and GLES2 install their own exec pointers and don't need the Save table. Also, the SET_* macros use different indices for the different APIs so the offsets used in vtxfmt.c are actually wrong for the ES APIs. Yes, this one fixes vertex attributes for GLES2 and I'd recommend 9f0d7dd25941b2f327c7bedf7513eee871829073 and 98ee6739d97b5592a7dad0b453c78e180a51ad50 as well. a889f9ee5cccee031c1090a6ef92cba894b1d77c is another good fix, without it writes to gl_PointSize breaks everything. It's been in master for a while and fixes a bunch of GLES2 test cases. Kristian ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] [PATCH] r600g: Fixed SIN/COS/SCS for the case where the operand is a literal.
On Thu, Jan 6, 2011 at 4:30 PM, Tilman Sauerbeck til...@code-monkey.de wrote:
Signed-off-by: Tilman Sauerbeck til...@code-monkey.de
looks good to me. should probably be applied to 7.10 as well.
Reviewed-by: Alex Deucher alexdeuc...@gmail.com
Alex
---
src/gallium/drivers/r600/r600_shader.c | 17 +++--
1 files changed, 15 insertions(+), 2 deletions(-)
diff --git a/src/gallium/drivers/r600/r600_shader.c
b/src/gallium/drivers/r600/r600_shader.c
index bb5038c..35a4cea 100644
--- a/src/gallium/drivers/r600/r600_shader.c
+++ b/src/gallium/drivers/r600/r600_shader.c
@@ -948,7 +948,7 @@ static int tgsi_setup_trig(struct r600_shader_ctx *ctx,
struct r600_bc_alu_src r600_src[3])
{
struct tgsi_full_instruction *inst =
ctx-parse.FullToken.FullInstruction;
- int r;
+ int r, src0_chan;
uint32_t lit_vals[4];
struct r600_bc_alu alu;
@@ -960,6 +960,19 @@ static int tgsi_setup_trig(struct r600_shader_ctx *ctx,
if (r)
return r;
+ src0_chan = tgsi_chan(inst-Src[0], 0);
+
+ /* We are going to feed two literals to the MAD below,
+ * which means that if the first operand is a literal as well,
+ * we need to copy its value manually.
+ */
+ if (r600_src[0].sel == V_SQ_ALU_SRC_LITERAL) {
+ unsigned index = inst-Src[0].Register.Index;
+
+ lit_vals[2] = ctx-literals[index * 4 + src0_chan];
+ src0_chan = 2;
+ }
+
lit_vals[0] = fui(1.0 /(3.1415926535 * 2));
lit_vals[1] = fui(0.5f);
@@ -972,7 +985,7 @@ static int tgsi_setup_trig(struct r600_shader_ctx *ctx,
alu.dst.write = 1;
alu.src[0] = r600_src[0];
- alu.src[0].chan = tgsi_chan(inst-Src[0], 0);
+ alu.src[0].chan = src0_chan;
alu.src[1].sel = V_SQ_ALU_SRC_LITERAL;
alu.src[1].chan = 0;
--
1.7.3.1
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
Re: [Mesa-dev] Upcoming Mesa 7.9.1 and 7.10 releases
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 01/06/2011 01:09 PM, Kristian Høgsberg wrote: On Mon, Jan 3, 2011 at 8:27 PM, Ian Romanick i...@freedesktop.org wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 It looks like things are shaping up nicely for the 7.9.1 and 7.10 releases. I've cherry picked a bunch of stuff over to the branches, but there are a few commits that I'm a bit uncertain about. I'd appreciate it if the commiters could weigh in. If you are on direct CC, that means you have a commit on one or both lists. NOTE: There are a few recent commits that are on both lists. 7.9.1 list: commit 1d33e940d2050f3d9180019f6ffd57f6fc295507 Author: Kristian Høgsberg k...@bitplanet.net Date: Thu Oct 14 14:57:47 2010 -0400 Only install vtxfmt tables for OpenGL R GLES1 and GLES2 install their own exec pointers and don't need the Save table. Also, the SET_* macros use different indices for the different APIs so the offsets used in vtxfmt.c are actually wrong for the ES APIs. Yes, this one fixes vertex attributes for GLES2 and I'd recommend 9f0d7dd25941b2f327c7bedf7513eee871829073 and 98ee6739d97b5592a7dad0b453c78e180a51ad50 as well. It looks like all three of these patches depend on 81ccb3e2c, which isn't in 7.9. I don't think I want that much churn on the day before the release. These will have to wait for 7.9.2. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk0mdzEACgkQX1gOwKyEAw8YZgCeN8T84Db0ZDaNKMOAfJKrVHWe YyQAn0zXa0tP/EjAmCksgeME2iHI7ypX =h1mO -END PGP SIGNATURE- ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
[Mesa-dev] [Bug 32825] egl_glx driver completely broken in 7.9 branch [fix in master]
https://bugs.freedesktop.org/show_bug.cgi?id=32825 cs...@daudt.org changed: What|Removed |Added CC||cs...@daudt.org -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email --- You are receiving this mail because: --- You are the assignee for the bug. ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
