X.Org found in NetBSD-current.
Thanks a lot to all who helped, especially Matthias Scheler who did
repeated tests on Mac OS X and older versions of NetBSD to make sure the
support for those platforms wouldn't be broken (or at least, not fatally,
as I would still expect a few hiccups here and there, because there is
only so much one can test in such limited time).
On the infrastructure side, this branch brings pkgconfig-builtin.mk, in
order to write very easily new builtin.mk files. It can actually handle
more than just pkgconfig files, but it will provide a version if it finds
such a file. x11.builtin.mk has also been made more useful and now all
existing (and future!) native-X11-related builtin.mk files should include
it.
lots of bug fixes; the intent of my update was mostly to fix
problems with i915 and r300 dri drivers and to provide the best environment
possible for the xorg server update to the 1.4 branch.
Morever, many of our patches from pkgsrc, i.e. those removed in this commit,
are now included upstream.
Additionally, MesaDemos now builds.
NOTES:
- MesaLib now requires xf86driproto 2.0.4, as earlier versions install
dri_interface.h, which is not correct.
- glu now exports its symbols via libtool's -export-symfile.
The symfile is generated pre-build--see glu/Makefile.
- glxinfo/glxgears do not seem to need to be linked to libpthread, and this
seems to cause problems on NetBSD with the new drm code. If I am wrong
about this, please let me know. The only program that seems to need
libpthread is glthreads in MesaDemos.
graphics/MesaLib/patches/patch-aw: export glXGetProcAddressARB
graphics/MesaLib/Makefile: bump PKGREVISION
graphics/glu/Makefile: bump PKGREVISION
Please help out by giving this change as extensive testing as
possible before 2008Q2 is branched.
Changes from 6.4.2 are (obviously) far too numerous to mention! This
package brings with it support for the Direct Rendering Infrastructure
(DRI) on NetBSD-current; the kernel DRM was committed approximately
one year ago. Patches for NetBSD 4.0 are available in the mailing list
archives (I think), but are _not supported_.
Developers and other interested parties are encouraged to read the TODO
file, as well as my remarks in options.mk. There is still a significant
amount of work to be done. Users of NetBSD/i386 (-current) shouldn't
have much trouble getting mainstream cards (Intel i915 and Radeon 9600-
R430 are by far the most tested) to work.
I have taken great pains to ensure that this works properly, but inevitably
there will be problems--especially on heretofore untested (read: non-x86)
hardware. Your questions, comments, problem reports, and of course
bug fixes are welcome.
NOTE: There are currently problems (on NetBSD, at least) linking some
executables, e.g. compiz, due to unresolved symbols for libpthread
functions. This is due to libX11's thread stubs, which should not
be enabled on NetBSD >= 2.0; see the TODO For more information
and the URL of an interim patch.
Downgraded graphics/MesaLib to 6.4.2nb3
Downgraded graphics/glu to 6.4.2
Downgraded graphics/glut to 6.4.2
Downgraded graphics/glx-utils to 6.4.2
Downgraded graphics/MesaDemos to 6.4.2
Downgraded graphics/Mesa to 6.4.2
Approved byg agc@.
New features:
* Added a bluegene-osmesa build config
Bug fixes:
* Added a few missing OpenGL 2.0 API entrypoints
* Fixed glDrawPixels(GL_STENCIL_INDEX) pixel transfer
* GLSL bug fix: added vec2(vec4) constructor
* GLSL bug fix: .strq and .rgba writemasks didn't always work
* Stencil pixel map didn't always work for glDrawPixels
* Fixed polygon stipple bug in i915 driver
* Binding a zero-sized texture didn't disable texturing
* Queries of GL_INFO_LOG_LENGTH, GL_SHADER_SOURCE_LENGTH didn't include
the terminating zero
* glXChooseFBConfig() in Xlib driver didn't handle GLX_STEREO flag properly
* Fixed a GLSL function call bug
* glPointParameteriv(GL_DISTANCE_ATTENUATION_EXT) didn't work
* glGetAttribLocation() always returned 1
* Fixed a few memory-related bugs in GLU library
Changes 7.0.0:
New features:
* OpenGL 2.0 and 2.1 API support.
Bug fixes:
* Fixed a few fog-related bugs.
* Fixed broken GLSL mix() function.
* Fixed broken GLSL exp() functions.
* Fixed GLSL mod4(vec4, vec4) bug.
* Implemented GLSL asin(), acos(), atan() functions.
* Fixed an R300 driver bug that caused Xorg composite manager to crash
* Fixed R300 vertex program/matrix bug
* GLSL dFdx() and dFdy() work for fragment program inputs now (texcoords)
* Specifying an invalid texture unit as a sampler could lead to a crash
* The GLX protocol request for glXDestroyPBuffer() was incorrect
* ARB vp state.light[n].half value was incorrect
* Fixed a positional light source bug
* Fixed point size attenuation problem
* glPopAttrib didn't restore texture object's LOD bias
* Fixed a TLS / TEXTREL problem
The latter is used by packages that build libraries from the Mesa
sources, while the former is only if the package builds from the Mesa
sources. Modify the various Mesa packages to include the proper
Makefile, and add some documentation to the Makefiles.
As a side effect, this suppresses the "duplicate target" warning in
graphics/glx-utils that was caused by having the same distfile listed
in DISTFILES more than once.
and add a new helper target and script, "show-buildlink3", that outputs
a listing of the buildlink3.mk files included as well as the depth at
which they are included.
For example, "make show-buildlink3" in fonts/Xft2 displays:
zlib
fontconfig
iconv
zlib
freetype2
expat
freetype2
Xrender
renderproto
RECOMMENDED is removed. It becomes ABI_DEPENDS.
BUILDLINK_RECOMMENDED.foo becomes BUILDLINK_ABI_DEPENDS.foo.
BUILDLINK_DEPENDS.foo becomes BUILDLINK_API_DEPENDS.foo.
BUILDLINK_DEPENDS does not change.
IGNORE_RECOMMENDED (which defaulted to "no") becomes USE_ABI_DEPENDS
which defaults to "yes".
Added to obsolete.mk checking for IGNORE_RECOMMENDED.
I did not manually go through and fix any aesthetic tab/spacing issues.
I have tested the above patch on DragonFly building and packaging
subversion and pkglint and their many dependencies.
I have also tested USE_ABI_DEPENDS=no on my NetBSD workstation (where I
have used IGNORE_RECOMMENDED for a long time). I have been an active user
of IGNORE_RECOMMENDED since it was available.
As suggested, I removed the documentation sentences suggesting bumping for
"security" issues.
As discussed on tech-pkg.
I will commit to revbump, pkglint, pkg_install, createbuildlink separately.
Note that if you use wip, it will fail! I will commit to pkgsrc-wip
later (within day).
New:
- added OSMesaColorClamp() function/feature
- added wglGetExtensionStringARB() function
Bug fixes:
- fixed some problems when building on Windows
- GLw header files weren't installed by installmesa script (bug 5396)
- GL/glfbdev.h file was missing from tarballs
- fixed TNL initialization bug which could lead to crash (bug 5791)
New:
- Added a fast XOR line drawing function in Xlib driver
- Added support for GL_ARB_texture_mirrored_repeat to savage
driver (supported only on Savage4 hardware).
Changes:
- Mesa now packaged in three parts: Library, Demos and GLUT
Bug fixes:
- GLX_X_RENDERABLE token wasn't accepted by glXChooseFBConfig
- Some files were present multiple times in the 6.3.2 tarballs
- r200_vtxtmp_x86.S file was missing from 6.3.2 tarball (bug 4207)
- glxgears_fbconfig demo didn't work (bug 4237)
- fixed bug when bilinear sampling 2d textures with borders
- glXCreatePbuffer() could segfault instead of returning 0 (bug 4235)
- fixed undefined frexp and rand in X.org libGLcore.a (bug 4242)
- fixed a few problems with proxy color tables (bug 4270)
- fixed precision problem in Z clearing (bug 4395)
- glBitmap, glDraw/CopyPixels mistakenly generated selection hits
- fixed potential segfault caused by reading pixels outside
of renderbuffer bounds
- glGetTexLevelParameter didn't accept GL_TEXTURE_DEPTH_SIZE_ARB
- fixed memory corruption bug involving software alpha buffers
- glReadPixels clipped by window bounds was sometimes broken
- glDraw/CopyPixels of stencil data ignored the stencil write mask
- glReadPixels from a texture bound to a framebuffer object didn't work
- glIsRender/FramebufferEXT weren't totally correct
- fixed a number of point size attenuation/fade bugs
- fixed glFogCoord bug 4729
- GLX encoding for transpose matrix functions was broken
- fixed broken fragment program KIL and SWZ instructions
- fragment programs that wrote result.depth.z didn't work
Changes 6.4.1:
Bug fixes:
- redefining a vertex program string didn't take effect in TNL module
- fixed occasional segfault upon vertex/fragment parsing error
- vertex program LIT instruction didn't handle 0^0=1 correctly
- fragment program fog option didn't work with glDrawPixels, glBitmap
- USE_MGL_NAMESPACE didn't work for x86-64
- OSMesa demos were missing from previous release tarballs
- fixed problem with float->ushort conversion in glClear (bug 4992)
- popping of GL_EYE_PLANE texgen state was broken (bug 4996)
- popping of GL_SPOT_DIRECTION light state was broken (bug 5005)
- fixed occasional triangle color interpolation problem on VMS
- work around invalid free() call (bug 5131)
- fixed BSD X server compilation problem by including stdint.h
the X11 distribution using imake into mk/buildlink3/imake-check.mk.
imake-check.mk calls out to a helper shell script mk/buildlink3/imake-check
that generates the required Imakefiles and runs imake. Remove the
now extraneous builtin-imake.mk files as the builtin.mk files can now
contain the name of the imake symbol to check.
contain a command plus arguments. We use the standard idiom to extract
the first word of a list stored in a make variable:
VAR_CMD= ${VAR:C/^/_asdf_/1:M_asdf_*:S/^_asdf_//}
XXX This can be replaced with a more succinct use of variable modifiers
XXX when we update the bootstrap bmake to match the latest NetBSD make
XXX sources, i.e.:
XXX
XXX VAR_CMD= ${VAR:[1]}
Several changes are involved since they are all interrelated. These
changes affect about 1000 files.
The first major change is rewriting bsd.builtin.mk as well as all of
the builtin.mk files to follow the new example in bsd.builtin.mk.
The loop to include all of the builtin.mk files needed by the package
is moved from bsd.builtin.mk and into bsd.buildlink3.mk. bsd.builtin.mk
is now included by each of the individual builtin.mk files and provides
some common logic for all of the builtin.mk files. Currently, this
includes the computation for whether the native or pkgsrc version of
the package is preferred. This causes USE_BUILTIN.* to be correctly
set when one builtin.mk file includes another.
The second major change is teach the builtin.mk files to consider
files under ${LOCALBASE} to be from pkgsrc-controlled packages. Most
of the builtin.mk files test for the presence of built-in software by
checking for the existence of certain files, e.g. <pthread.h>, and we
now assume that if that file is under ${LOCALBASE}, then it must be
from pkgsrc. This modification is a nod toward LOCALBASE=/usr. The
exceptions to this new check are the X11 distribution packages, which
are handled specially as noted below.
The third major change is providing builtin.mk and version.mk files
for each of the X11 distribution packages in pkgsrc. The builtin.mk
file can detect whether the native X11 distribution is the same as
the one provided by pkgsrc, and the version.mk file computes the
version of the X11 distribution package, whether it's built-in or not.
The fourth major change is that the buildlink3.mk files for X11 packages
that install parts which are part of X11 distribution packages, e.g.
Xpm, Xcursor, etc., now use imake to query the X11 distribution for
whether the software is already provided by the X11 distribution.
This is more accurate than grepping for a symbol name in the imake
config files. Using imake required sprinkling various builtin-imake.mk
helper files into pkgsrc directories. These files are used as input
to imake since imake can't use stdin for that purpose.
The fifth major change is in how packages note that they use X11.
Instead of setting USE_X11, package Makefiles should now include
x11.buildlink3.mk instead. This causes the X11 package buildlink3
and builtin logic to be executed at the correct place for buildlink3.mk
and builtin.mk files that previously set USE_X11, and fixes packages
that relied on buildlink3.mk files to implicitly note that X11 is
needed. Package buildlink3.mk should also include x11.buildlink3.mk
when linking against the package libraries requires also linking
against the X11 libraries. Where it was obvious, redundant inclusions
of x11.buildlink3.mk have been removed.
New:
- enabled GL_ARB_texture_rectangle (same as GL_NV_texture_rectangle)
- updated Doxygen support (Jose Fonseca)
Changes:
- some GGI driver updates (Christoph Egger, bug 1025977)
Bug fixes:
- Omit GL_ARB_texture_non_power_of_two from list of OpenGL 1.5 features
- fixed a few compilation issues on IRIX
- fixed a matrix classification bug (reported by Wes Bethel)
- we weren't reseting the vertex/fragment program error state
before parsing (Dave Reveman)
- adjust texcoords for sampling texture rectangles (Dave Reveman)
- glGet*(GL_MAX_VERTEX_ATTRIBS_ARB) wasn't implemented
- repeated calls to glDeleteTexture(t) could lead to a crash
- fixed potential ref count bugs in VBOs and vertex/fragment programs
- spriteblast demo didn't handle window size changes correctly
- glTexSubImage didn't handle pixels=NULL correctly for PBOs
- fixed color index mode glDrawPixels bug (Karl Schultz)
Changes 6.1:
New:
- Revamped Makefile system
- glXUseRotatedXFont() utility (see xdemos/xuserotfont.c)
- internal driver interface changes related to texture object
allocation, vertex/fragment programs, BlendEquationSeparate, etc.
- option to walk triangle edges with double-precision floats
(Justin Novosad of Discreet) (see config.h file)
- support for AUX buffers in software GLX driver
- updated glext.h to version 24 and glxext.h to version 6
- new MESA_GLX_FORCE_ALPHA and MESA_GLX_DEPTH_BITS env vars
- updated BeOS support (Philippe Houdoin)
Changes:
- fragment fog interpolation is perspective corrected now
- new glTexImage code, much cleaner, may be a bit faster
Bug fixes:
- glArrayElement in display lists didn't handle generic vertex attribs
- glFogCoord didn't always work properly
- ARB_fragment_program fog options didn't work
- frag prog TEX instruction no longer incorrectly divides s,t,r by q
- ARB frag prog TEX and TEXP instructions now use LOD=0
- glTexEnviv in display lists didn't work
- glRasterPos didn't do texgen or apply texture matrix
- GL_DOUBLE-valued vertex arrays were broken in some cases
- fixed texture rectangle edge/border sampling bugs
- sampling an incomplete texture in a fragment program would segfault
- glTexImage was missing a few error checks
- fixed some minor glGetTexParameter glitches
- GL_INTENSITY was mistakenly accepted as a to glTexImage
- fragment program writes to RC/HC register were broken
- fixed a few glitches in GL_HP_occlusion_test extension
- glBeginQueryARB and glEndQueryARB didn't work inside display lists
- vertex program state references were broken
- fixed triangle color interpolation bug on AIX (Shane Blackett)
- fixed a number of minor memory leaks (bug #1002030)
in the process. (More information on tech-pkg.)
Bump PKGREVISION and BUILDLINK_DEPENDS of all packages using libtool and
installing .la files.
Bump PKGREVISION (only) of all packages depending directly on the above
via a buildlink3 include.
All library names listed by *.la files no longer need to be listed
in the PLIST, e.g., instead of:
lib/libfoo.a
lib/libfoo.la
lib/libfoo.so
lib/libfoo.so.0
lib/libfoo.so.0.1
one simply needs:
lib/libfoo.la
and bsd.pkg.mk will automatically ensure that the additional library
names are listed in the installed package +CONTENTS file.
Also make LIBTOOLIZE_PLIST default to "yes".
by moving the inclusion of buildlink3.mk files outside of the protected
region. This bug would be seen by users that have set PREFER_PKGSRC
or PREFER_NATIVE to non-default values.
BUILDLINK_PACKAGES should be ordered so that for any package in the
list, that package doesn't depend on any packages to the left of it
in the list. This ordering property is used to check for builtin
packages in the correct order. The problem was that including a
buildlink3.mk file for <pkg> correctly ensured that <pkg> was removed
from BUILDLINK_PACKAGES and appended to the end. However, since the
inclusion of any other buildlink3.mk files within that buildlink3.mk
was in a region that was protected against multiple inclusion, those
dependencies weren't also moved to the end of BUILDLINK_PACKAGES.
modify {MesaLib,glu}/{buildlink3,builtin}.mk accordingly. Also add the
dependency on {MesaLib,glu}>=${_MESA_REQD} to the builtin.mk files so
that we correctly check the built-in Mesa version against the dependency
requirements.
built-in or not into a separate builtin.mk file. The code to deal
checking for built-in software is much simpler to deal with in pkgsrc.
The buildlink3.mk file for a package will be of the usual format
regardless of the package, which makes it simpler for packagers to
update a package.
The builtin.mk file for a package must define a single yes/no variable
USE_BUILTIN.<pkg> that is used by bsd.buildlink3.mk to decide whether
to use the built-in software or to use the pkgsrc software.
as PREFER_PKGSRC. Preferences are determined by the most specific
instance of the package in either PREFER_PKGSRC or PREFER_NATIVE. If
a package is specified in neither or in both variables, then PREFER_PKGSRC
has precedence over PREFER_NATIVE.
values that MesaLib/buildlink3.mk converts into a single _MESA_REQD that
is the highest version required. If including the Mesa/buildlink3.mk
file, then require the same version of Mesa as required by glut.
