November 2008 - Xquartz-dev - lists.macosforge.org

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 21 Nov '08

21 Nov '08

Am 20.11.2008 um 22:08 schrieb Jeremy Huddleston: > Actually, try putting this in pbproxy.h: > > #include <AvailabilityMacros.h> > #if MAC_OS_X_VERSION_MIN_REQUIRED < 1050 > #if __LP64__ || NS_BUILD_32_LIKE_64 > typedef long NSInteger; > typedef unsigned long NSUInteger; > #else > typedef int NSInteger; > typedef unsigned int NSUInteger; > #endif > #endif It failed once: X11Controller.m: In function '-[X11Controller install_window_menu:]': X11Controller.m:156: error: 'NSUInteger' undeclared (first use in this function) X11Controller.m:156: error: (Each undeclared identifier is reported only once X11Controller.m:156: error: for each function it appears in.) X11Controller.m:156: error: parse error before ']' token X11Controller.m:166: error: parse error before ']' token make[3]: *** [X11Controller.lo] Error 1 because pbproxy.h is not used. This is the list of C header files include that are *not* taken from /usr or /System branches: . sanitizedCarbon.h . ../../include/dix-config.h . quartzCommon.h . X11Controller.h .. sanitizedCocoa.h .. xpr/x-list.h . X11Application.h . ../../include/opaque.h .. ../../include/globals.h ... ../../include/window.h .... ../../include/misc.h ..... ../../include/os.h ...... ../../include/misc.h .... ../../include/region.h ..... ../../include/regionstr.h ...... ../../include/miscstruct.h ....... ../../include/misc.h ....... ../../include/gc.h ........ ../../include/screenint.h ......... ../../include/misc.h ........ ../../include/pixmap.h ......... ../../include/misc.h ......... ../../include/screenint.h .... ../../include/screenint.h . darwin.h .. ../../include/inputstr.h ... ../../include/input.h .... ../../include/misc.h .... ../../include/screenint.h .... ../../include/window.h ... ../../include/window.h ... ../../include/dixstruct.h .... ../../include/dix.h ..... ../../include/gc.h ..... ../../include/window.h ..... ../../include/input.h .... ../../include/resource.h ..... ../../include/misc.h .... ../../include/cursor.h ..... ../../include/misc.h ..... ../../include/screenint.h ..... ../../include/window.h .... ../../include/gc.h .... ../../include/pixmap.h .. ../../include/scrnintstr.h ... ../../include/screenint.h ... ../../include/regionstr.h ... ../../include/bstore.h .... ../../include/bstorestr.h ..... ../../include/gc.h ..... ../../include/pixmap.h ..... ../../include/region.h ..... ../../include/window.h ... ../../include/colormap.h .... ../../include/screenint.h .... ../../include/window.h ... ../../include/cursor.h ... ../../include/validate.h .... ../../include/miscstruct.h .... ../../include/regionstr.h ... ../../include/dix.h .. threadSafety.h . darwinEvents.h . quartz.h .. quartzPasteboard.h . applewmExt.h -- Greetings Pete Genius may have its limitations, but stupidity is not thus handicapped. – Elbert Hubbard

2 1

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 21 Nov '08

21 Nov '08

Am 20.11.2008 um 23:06 schrieb Peter Dyballa: > indirect.c: In function 'setup_dispatch_table': > indirect.c:1268: error: 'glBlitFramebufferEXT' undeclared (first use > in this function) > indirect.c:1268: error: (Each undeclared identifier is reported only > once > indirect.c:1268: error: for each function it appears in.) > The problem starts at lines #57: 57 #include <OpenGL/OpenGL.h> 58 #include <OpenGL/CGLContext.h> 59 60 // X11 and X11's glx 61 #include <GL/gl.h> 62 #include <GL/glxproto.h> 63 #include <windowstr.h> 64 #include <resource.h> 65 #include <GL/glxint.h> 66 #include <GL/glxtokens.h> 67 #include <scrnintstr.h> 68 #include <glxserver.h> 69 #include <glxscreens.h> 70 #include <glxdrawable.h> 71 #include <glxcontext.h> 72 #include <glxext.h> 73 #include <glxutil.h> 74 #include <glxscreens.h> 75 #include <GL/internal/glcore.h> 76 #include "x-hash.h" 77 #include "x-list.h" /System/Library/Frameworks/OpenGL.framework/Headers/CGLContext.h is included on place of <OpenGL/CGLContext.h>. Followed by . /System/Library/Frameworks/OpenGL.framework/Headers/CGLContext.h .. /System/Library/Frameworks/OpenGL.framework/Headers/gliContext.h .. /System/Library/Frameworks/OpenGL.framework/Headers/gliDispatch.h ... /System/Library/Frameworks/OpenGL.framework/Headers/gl.h .... /System/Library/Frameworks/OpenGL.framework/Headers/glext.h . /usr/X11/include/GL/gl.h . /usr/X11/include/GL/glxproto.h .. /usr/X11/include/GL/glxmd.h So we see that /System/Library/Frameworks/OpenGL.framework/Headers/ glext.h is included from an Apple version of GL/gl.h, /System/Library/ Frameworks/OpenGL.framework/Headers/gl.h. This has on line #64: 64 #ifndef GL_GLEXT_LEGACY 65 #include <OpenGL/glext.h> 66 #endif /usr/X11/include/GL/gl.h, which includes (the proper) <GL/glext.h> (that has glBlitFramebufferEXT) has on line #2140: 2140 /* 2141 * Define this token if you want "old-style" header file behaviour (extensions 2142 * defined in gl.h). Otherwise, extensions will be included from glext.h. 2143 */ 2144 #if defined(GL_GLEXT_LEGACY) 2145 2146 /* All extensions that used to be here are now found in glext.h */ 2147 2148 #else /* GL_GLEXT_LEGACY */ 2149 2150 #include <GL/glext.h> 2151 2152 #endif /* GL_GLEXT_LEGACY */ GL_GLEXT_LEGACY is not defined in any /System/Library/Frameworks/ OpenGL.framework/Headers file, nowhere in the pre-proceessed indirect.c file, so <GL/glext.h> should be included. But all I can see in the pre-proceessed indirect.c file is: # 59 "indirect.c" 2 # 1 "/usr/X11/include/GL/gl.h" 1 # 62 "indirect.c" 2 # 1 "/usr/X11/include/GL/glxproto.h" 1 #define _GLX_glxproto_h_ # 34 "/usr/X11/include/GL/glxproto.h" # 1 "/usr/X11/include/GL/glxmd.h" 1 #define _GLX_glxmd_h_ # 42 "/usr/X11/include/GL/glxmd.h" typedef float FLOAT32; typedef double FLOAT64; # 52 "/usr/X11/include/GL/glxmd.h" #define F32 # 35 "/usr/X11/include/GL/glxproto.h" 2 and no further mentioning of /usr/X11/include/GL/gl.h. All data types are taken from /System/Library/Frameworks/OpenGL.framework/Headers/gl.h. This obviously is going to persist even if I switch to Mesa 7.0.4 and X11R7.4 since xserver and operating system stay the same. So it might be appropriate to #include <GL/glext.h> in case of Tiger in indirect.c after line #61 ... -- Greetings Pete Specifications are for the weak and timid!

1 0

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 21 Nov '08

21 Nov '08

Am 20.11.2008 um 23:06 schrieb Peter Dyballa: >> glcontextmodes.c: In function '_gl_copy_visual_to_context_mode': >> glcontextmodes.c:193: error: 'struct __GLcontextModesRec' has no >> member named 'bindToTextureRgb' >> glcontextmodes.c:194: error: 'struct __GLcontextModesRec' has no >> member named 'bindToTextureRgba' >> glcontextmodes.c:196: error: 'struct __GLcontextModesRec' has no >> member named 'bindToMipmapTexture' >> glcontextmodes.c:197: error: 'struct __GLcontextModesRec' has no >> member named 'bindToTextureTargets' >> glcontextmodes.c:200: error: 'struct __GLcontextModesRec' has no >> member named 'yInverted' >> >> and also in functions _gl_get_context_mode_data, >> _gl_context_modes_create, and _gl_context_modes_are_same. These errors could be resolved by "updating" the installed /usr/X11/ include/GL/internal/glcore.h with seven lines (one empty, one comment) from Mesa-7.0.2/include/GL/internal/glcore.h. A bug report on this was filed at https://bugs.freedesktop.org. -- Greetings Pete Competition is the great erode of profits.

1 0

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 21 Nov '08

21 Nov '08

Am 20.11.2008 um 19:37 schrieb Jeremy Huddleston: > What private scripts? > > Everything he referenced is on the developer info page on the wiki: So I presume the shell scripts are copies of the boxed commands on this page? -- Greetings Pete America believes in education: the average professor earns more money in a year than a professional athlete earns in a whole week. – Evan Esar

2 1

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 20 Nov '08

20 Nov '08

Am 20.11.2008 um 19:37 schrieb Jeremy Huddleston: >> threadSafety.c:34:22: error: execinfo.h: No such file or directory > > hmm... Maybe the whole message helps: threadSafety.c:34:22: error: execinfo.h: No such file or directory threadSafety.c: In function 'spewCallStack': threadSafety.c:41: warning: implicit declaration of function 'backtrace' threadSafety.c:41: warning: nested extern declaration of 'backtrace' threadSafety.c:42: warning: implicit declaration of function 'backtrace_symbols' threadSafety.c:42: warning: nested extern declaration of 'backtrace_symbols' threadSafety.c:42: warning: initialization makes pointer from integer without a cast -- Greetings Pete Isn't vi that text editor with two modes... one that beeps and one that corrupts your file? – Dan Jacobson, on comp.os.linux.advocacy

3 2

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 20 Nov '08

20 Nov '08

Am 20.11.2008 um 19:37 schrieb Jeremy Huddleston: > You need to install xfixesproto first. I should add that to the > dependencies in configure.ac ... The reason is ridiculous: I had no --prefix=... argument for configure! So some hard-code /usr/local path was used instead of what CFLAGS or CPPFLAGS suggested. When I first tried to compile X11R7.1 on Tiger Apple's libtool stuff could not be used. The build process did not come to the stage of compilation. I think it's older than X11R7 ... Since Leopard is not offering an X11 with transparency I can easily wait till Christmas 2009 to buy (then reasonably priced) and install it. Until then I'll have some time to test Apple software whether it's able to compile X11. -- Greetings Pete By filing this bug report you have challenged the honor of my family. Prepare to die!

2 1

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 20 Nov '08

20 Nov '08

Am 20.11.2008 um 19:37 schrieb Jeremy Huddleston: >> threadSafety.c:34:22: error: execinfo.h: No such file or directory > > hmm... It's here at the top of xserver/hw/xquartz/threadSafety.c: #ifdef HAVE_DIX_CONFIG_H #include <dix-config.h> #endif #include "threadSafety.h" #include "os.h" #include <execinfo.h> pthread_t APPKIT_THREAD_ID; pthread_t SERVER_THREAD_ID; Part of Leopard? I definitely do not have a C header file with this name, not even in Fink. -- Greetings Pete The world would be a better place if Larry Wall had been born in Iceland, or any other country where the native language actually has syntax. – Peter da Silva

1 0

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 20 Nov '08

20 Nov '08

Am 20.11.2008 um 16:01 schrieb Peter Dyballa: > There are also a lot more errors, like for example: > > x-selection.h:36:35: error: X11/extensions/Xfixes.h: No such file or > directory > x-selection.h:101: error: parse error before > 'XFixesSelectionNotifyEvent' > x-selection.h:102: fatal error: method definition not in > @implementation context > > (which I have in: > > ...sources/X11R7.2/lib/libXfixes/include/X11/extensions/Xfixes.h > /usr/X11R7.2/include/X11/extensions/Xfixes.h > > and although I have -I/usr/X11R7.2/include I'll add another -I/usr/ > X11/include!) With this addition the failure went away ... > > There are also glapi.h and glapitable.h missing. 1.5.99.1 and > 1.3.99.0 have both files in xserver/glx resp. xserver/GL/mesa/glapi. > Could be it's my fault that I have Mesa 7.2 prepared (I can go back > to Mesa 7.0.2). These C header were found when I left away --disable-glx from from configure's invocation. Letting enabling happen automatically introduced new errors, that may come from the fact that I was compiling with Mesa 7.2.0: glcontextmodes.c: In function '_gl_copy_visual_to_context_mode': glcontextmodes.c:193: error: 'struct __GLcontextModesRec' has no member named 'bindToTextureRgb' glcontextmodes.c:194: error: 'struct __GLcontextModesRec' has no member named 'bindToTextureRgba' glcontextmodes.c:196: error: 'struct __GLcontextModesRec' has no member named 'bindToMipmapTexture' glcontextmodes.c:197: error: 'struct __GLcontextModesRec' has no member named 'bindToTextureTargets' glcontextmodes.c:200: error: 'struct __GLcontextModesRec' has no member named 'yInverted' and also in functions _gl_get_context_mode_data, _gl_context_modes_create, and _gl_context_modes_are_same. The error with xserver/GL/indirect.c changed to: indirect.c: In function 'setup_dispatch_table': indirect.c:1268: error: 'glBlitFramebufferEXT' undeclared (first use in this function) indirect.c:1268: error: (Each undeclared identifier is reported only once indirect.c:1268: error: for each function it appears in.) indirect.c:1598: warning: assignment from incompatible pointer type indirect.c:1768: warning: assignment from incompatible pointer type indirect.c:1823: warning: assignment from incompatible pointer type indirect.c:1824: warning: assignment from incompatible pointer type indirect.c:1825: warning: assignment from incompatible pointer type Could be Mesa again. The error in xserver/hw/xquartz/pbproxy/x- selection.m changed to: x-selection.m: In function '-[x_selection x_active:]': x-selection.m:360: error: nested functions are disabled, use -fnested- functions to re-enable x-selection.m:360: error: syntax error before 'changeCount' x-selection.m:361: error: 'NSInteger' undeclared (first use in this function) x-selection.m:361: error: (Each undeclared identifier is reported only once x-selection.m:361: error: for each function it appears in.) x-selection.m:371: error: 'countNow' undeclared (first use in this function) x-selection.m:373: error: 'changeCount' undeclared (first use in this function) x-selection.m: In function '-[x_selection handle_image:pasteboard:]': x-selection.m:1228: warning: use '-fobjc-exceptions' to enable Objective-C exception syntax Multiple include guards may be useful for: /usr/include/assert.h The supposed non-Leopard errors are still the same. Later I'll retry with Mesa 7.0.2 – and NSInteger changed to a simple int? Or size_t? I still do not use --enable-integrated-xpbproxy, I am leaving it for later tries. I added -H to CFLAGS to find any Fink pollution. Later, maybe next week, I'll try to remove this or that Fink related element piece by piece ... -- Greetings Pete A lot of people mistake a short memory for a clear conscience. – Doug Larson

2 1

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 20 Nov '08

20 Nov '08

Am 19.11.2008 um 21:09 schrieb Jeremy Huddleston: > http://xquartz.macosforge.org/trac/wiki/DeveloperInfo? > version=17#MacPortsFinkGentooConflicts I think I am following these descriptions and recommendations. Adding -I /usr/X11/share/aclocal as first element of ACLOCAL resolved the XTrans issue. One issue from the modular build is not solved by this: autoreconf: Entering directory `.' autoreconf: configure.ac: not using Gettext autoreconf: running: aclocal -I /usr/X11/share/aclocal autoreconf: configure.ac: tracing autoreconf: running: libtoolize --copy libtoolize: `/usr/local/share/aclocal/libtool.m4' is serial 50, less than 52 in `aclocal.m4' Use `--force' to replace newer libtool files with this version. autoreconf: libtoolize failed with exit status: 1 ***** autogen.sh failed on xserver/ None of my my libtool.m4 files has a serial greater or equal to 52. Manually applying some force (autoreconf -fvi) helps, but modular build is not possible yet. I ran again into: X11Controller.m: In function '-[X11Controller install_window_menu:]': X11Controller.m:156: error: 'NSUInteger' undeclared (first use in this function) X11Controller.m:156: error: (Each undeclared identifier is reported only once X11Controller.m:156: error: for each function it appears in.) X11Controller.m:156: error: parse error before ']' token X11Controller.m:166: error: parse error before ']' token There are also a lot more errors, like for example: x-selection.h:36:35: error: X11/extensions/Xfixes.h: No such file or directory x-selection.h:101: error: parse error before 'XFixesSelectionNotifyEvent' x-selection.h:102: fatal error: method definition not in @implementation context (which I have in: ...sources/X11R7.2/lib/libXfixes/include/X11/extensions/Xfixes.h /usr/X11R7.2/include/X11/extensions/Xfixes.h and although I have -I/usr/X11R7.2/include I'll add another -I/usr/ X11/include!) quartzAudio.c: In function 'QuartzAudioInit': quartzAudio.c:340: error: 'AudioDeviceIOProcID' undeclared (first use in this function) quartzAudio.c:340: error: (Each undeclared identifier is reported only once quartzAudio.c:340: error: for each function it appears in.) quartzAudio.c:340: error: parse error before 'sInputIOProcID' quartzAudio.c:341: warning: implicit declaration of function 'AudioDeviceCreateIOProcID' quartzAudio.c:341: warning: nested extern declaration of 'AudioDeviceCreateIOProcID' quartzAudio.c:341: error: 'sInputIOProcID' undeclared (first use in this function) quartzKeyboard.c: In function 'QuartzReadSystemKeymap': quartzKeyboard.c:859: error: 'currentKeyLayoutRef' undeclared (first use in this function) quartzKeyboard.c:859: error: (Each undeclared identifier is reported only once quartzKeyboard.c:859: error: for each function it appears in.) quartzKeyboard.c:715: warning: unused variable 'currentKeyLayoutDataRef' threadSafety.c:34:22: error: execinfo.h: No such file or directory stub.c: In function ‘set_x11_path’: stub.c:71: error: ‘kLSUnknownCreator’ undeclared (first use in this function) stub.c:71: error: (Each undeclared identifier is reported only once stub.c:71: error: for each function it appears in.) stub.c:112: error: ‘kLSApplicationNotFoundErr’ undeclared (first use in this function) So it looks to be merely a Leopard project ... There are also glapi.h and glapitable.h missing. 1.5.99.1 and 1.3.99.0 have both files in xserver/glx resp. xserver/GL/mesa/glapi. Could be it's my fault that I have Mesa 7.2 prepared (I can go back to Mesa 7.0.2). For the afternoon I'll start another compilation with the two fixes (NSUInteger and -I/usr/X11/include) -- Greetings Pete We also sponsor National Invisible Chronic Illness Awareness Week annually in September. Join the millions

2 1

Re: [Xquartz-dev] 2.3.2_beta3 on Tiger in X11R7.2
by Peter Dyballa 20 Nov '08

20 Nov '08

Am 19.11.2008 um 21:09 schrieb George Peter Staplin: > $ sh gps_conf_debug.sh > $ bld_install.sh > > $ cat gps_conf_debug.sh George, your private scripts are not publicly available, or are they? -- Greetings Pete Upgraded, adj.: Didn't work the first time.

2 1