Bug in xorg-xserver - missing declaration!
Dr.-Ing. Dieter Jurzitza
dieter.jurzitza at t-online.de
Sun Feb 19 19:29:22 UTC 2017
Dear listmembers,
I found the problem and I think I fixed it. Please validate & give a feedback
OS: openSUSE Leap 42.2
Result: server crash
Due to the similarity of this problem and the one that had had been found (and
fixed!) by Michel Dänzer I was capable to provide a fix that solves the issue
(for me...)
Due to my limited capabilities with regard to C-Programming someone more
experienced than I am ought to look over that (and probably improve!)
Please consider the following patch:
--- glamor/glamor_dash.c.original 2016-03-11 19:43:16.000000000 +0100
+++ glamor/glamor_dash.c 2017-02-19 20:00:11.883218927 +0100
@@ -32,6 +32,7 @@
static const char dash_vs_exec[] =
" dash_offset = primitive.z / dash_length;\n"
+ " vec2 pos = vec2(0,0);\n"
GLAMOR_POS(gl_Position, primitive.xy);
static const char dash_fs_vars[] =
Thank you for looking into this. The source filename is xorg-x11-
server-7.6_1.18.3-10.1.src.rpm
Take care,
Dieter Jurzitza
Am Freitag, 17. Februar 2017, 15:58:53 schrieb dieter.jurzitza at t-online.de:
> OS: openSUSE Leap 42.2
> Condition: start tgif with the file attached
> Result: server crash
>
>
> When tgif is started on my machine with the attached file, the X-server
> crashes badly, ending at the login - screen.
>
> The logfile Xorg.log.old states:
>
>
> fred at djunix-office:~/x11-crash> cat errorlog2
> [ 10168.529] Failed to compile VS: 0:13(43): error: `pos' undeclared
> 0:13(14): error: operands to arithmetic operators must be numeric
> 0:13(13): error: operands to arithmetic operators must be numeric
>
> [ 10168.529] Program source:
> #version 130
> attribute vec3 primitive;
> varying float dash_offset;
> uniform vec2 fill_offset;
> uniform vec2 fill_size_inv;
> varying vec2 fill_pos;
> uniform float dash_length;
> uniform vec4 v_matrix;
> void main() {
> dash_offset = primitive.z / dash_length;
> gl_Position.xy = primitive.xy.xy * v_matrix.xz + v_matrix.yw;
> gl_Position.zw = vec2(0.0,1.0);
> fill_pos = (fill_offset + primitive.xy + pos) * fill_size_inv;
> }
> [ 10168.529] (EE)
> Fatal server error:
> [ 10168.529] (EE) GLSL compile failure
> [ 10168.529] (EE)
> [ 10168.529] (EE)
> Please consult the The X.Org Foundation support
> at http://wiki.x.org
> for help.
> [ 10168.529] (EE) Please also check the log file at "/var/log/Xorg.0.log"
> for additional information. [ 10168.529] (EE)
> [ 10168.529] (II) AIGLX: Suspending AIGLX clients for VT switch
> [ 10168.903] (EE) Server terminated with error (1). Closing log file.
>
> This resembled me to the bug fixed by Michel Dänzer (see Bug 97300)
> --- glamor/glamor_composite_glyphs.c
> +++ glamor/glamor_composite_glyphs.c
> @@ -199,7 +199,8 @@ static const glamor_facet
> glamor_facet_composite_glyphs_12 .vs_vars = ("attribute vec2 primitive;\n"
> "attribute vec2 source;\n"
> "varying vec2 glyph_pos;\n"),
> - .vs_exec = (GLAMOR_POS(gl_Position, primitive)
> + .vs_exec = (" vec2 pos = vec2(0,0);\n"
> + GLAMOR_POS(gl_Position, primitive.xy)
> " glyph_pos = source.xy * ATLAS_DIM_INV;\n"),
> .fs_vars = ("varying vec2 glyph_pos;\n"),
> .fs_exec = (" vec4 mask = texture2D(atlas, glyph_pos);\n"),
>
> however, this did not fix the issue. What is somewhat strange is the fact
> that the error states "primitive.xy.xy", please note the double
> referencing. I could track this down to the line 81 in the (suse -patched
> ...) file glamor_program.c, where the line with "fill_pos = ...." orignates
> from - but my knowledge on x11 and C does not bring me further.
>
> Mr. Anholt: Michel Dänzer told me to probably CC you, so I did ....
>
> Thank you very much for looking into this,
> best regards
>
>
>
>
>
> Dieter Jurzitza
--
-----------------------------------------------------------
Dr.-Ing. Dieter Jurzitza 76131 Karlsruhe
More information about the xorg-devel
mailing list