I had used pkg-config from the Makefile because I didn't want to screw
around with the non-autoconf build, but that doesn't work because the
PKG_CONFIG_PATH or TALLOC_LIBS/TALLOC_CFLAGS that people set at
configure time needs to be respected and may not be present at build
time.
Bug #29585
This pulls in multiple i965 driver fixes which will help ensure better
testing coverage during development, and also gets past the conflicts
of the src/mesa/shader -> src/mesa/program move.
Conflicts:
src/mesa/Makefile
src/mesa/main/shaderapi.c
src/mesa/main/shaderobj.h
Fix this build error (in MesaGLUT-7.6.1)...
glut_cmap.c:23:66: error: X11/Xmu/StdCmap.h: No such file or directory
...by not preventing the cflags that pkg-config finds for glut dependencies
(including 'xmu') from being used.
Defining GLUT_CFLAGS before running the pkg-config prevents the
cflags found by pkg-config from being used.
This patch lets GLUT_CFLAGS that configure & pkg-config work
so hard to set actually get used.
Also make sure the generated configs/autoconf defines GLUT_CFLAGS
used in (at least) src/glut/glx/Makefile.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Merge all targets into targets/egl/. The target produces
egl_gallium_<HW>.so for each pipe driver and st_<API>.so for each client
APIs. This enables us to further merge egl_gallium_<HW>.so into
egl_gallium.so later.
When building OSMesa and xlib GL, the resulting OSMesa would be linked
against libGL instead of the internal mesa libraries. However, when
building with -fvisibility=hidden, some of the internal functions used
in OSMesa could not be resolved through libGL.
Instead, always build OSMesa standalone without linking against libGL.
This has the advantage that OSMesa is always built the same way, but it
means that disk space is wasted when libGL is installed since both
libraries will contain the internal objects.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Tested-by: Tom Fogal <tfogal@alumni.unh.edu>
This slightly reduces reduces cairo-gl firefox-talos-gfx runtime on my
Ironlake:
before:
[ # ] backend test min(s) median(s) stddev. count
[ 0] gl firefox-talos-gfx 38.236 38.383 0.43% 5/6
after:
[ 0] gl firefox-talos-gfx 37.799 38.203 0.39% 6/6
It turns out the cost of caching these objects and looking them up in
the cache again is greater than the cost of just computing the object
again, particularly when the overhead of having a separate BO to pin
is removed.
(Those that are paying close attention will note that this is a
reversal of the path I was moving the driver in a couple of years ago.
The major thing that has changed is that back then all state was
recomputed when we wrapped the streaming state buffer, including
recompiling our precious programs. Now, we're uncaching just the
objects that are cheap to compute, and retaining caching of expensive
objects)
The remaining programs are ones I've had difficulty finding a build
environment for to make the build system or are unit tests that should
probably live next to their code instead. Hopefully people can bring
over the build for remaining pieces they care about.
With the omit list gone, there are not too many differences in building
core mesa and ES overlay. Remove the mesa/es and build both of them in
src/mesa/Makefile.
The new target installs client API modules to EGL_DRIVER_INSTALL_DIR.
They are used by st/egl.
The client APIs are built from OpenGL and OpenVG state trackers. For
this to work, st/vega is modified to produce a static library,
libvega.a, instead. st/es is also not needed any more. It is removed
and --with-state-trackers=es is replaced by --enable-gles-overlay.
As st/egl now has its own client API modules, this solves the ABI issue
between st/egl and client APIs, as long as the client API modules are
distributed with st/egl. Plus, this allows st/egl to support OpenGL
with non-Gallium libGL.so.
Define <API>_LIB, <API>_LIB_NAME, <API>_LIB_GLOB, and some other
variables in the configs. Fix a typo in glesv1_cm.pc.in where an
inexistent variable is used.
Rename vgFooBar to vegaFooBar and use vgapi as the dispatcher. This
makes sure there is always a current context when the internal functions
are called. And eglGetProcAddress is finally supported.
ES overlay is built with FEATURE_ES1 or FEATURE_ES2, and is built
without FEATURE_GL. Fix the build by always building OpenGL ES sources,
but test for FEATURE_ES1 or FEATURE_ES2. Also, define symbols that are
missing because FEATURE_GL is not defined.
allows the swrastg_dri.so to be built with llvmpipe, also links llvm
to all dri drivers
use --enable-gallium-llvm to use it.
Signed-off-by: Dave Airlie <airlied@redhat.com>
The variable X_LIBS from AC_PATH_XTRA contains only the -L searchdir
parameter and not the -lX11 to link to Xlib. Use X11 prefixed build vars
for linking with Xlib to avoid the conflict.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
This is the last nvfx unification patch.
nv[34]0_fragtex.c are moved to the common directory
nv[34]0_shader.h are renamed to nv[34]0_vertprog.h and moved to
the common directory
The separate nv30 and nv40 directories are removed from the build
system
Will be used to hold source files unified between nv30 and nv40.
Eventually all nv30 and nv40 code will be moved there and the
nv30 and nv40 directories will be removed.
Make sure we always test for XEXT version.
Make sure that the user has a recent version of libkms and libdrm installed.
Note that the st/xorg code requires so new versions to work but these are
needed to make a proper vmwgfx_drv.so driver which is the only real user.
Cherry picked from 46cf606cd35d6740b28fd26bd32dbdfdde2c7833
Conflicts:
configure.ac
Mesa now requires GCC 3.3.0 or later. See the following thread from
the mesa3d-dev mailing list for more details:
http://marc.info/?t=126748568900005&r=1&w=2
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
This moves the logic for how to align pitches, heights, and sizes of
objects to one central location. Fixes rendering with texture tiling
on i915. Note that current libdrm is required for the change for
I915_TILING_NONE pitch alignment.
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit c1a4f249f18a766488da0832fbf6700de93af2f8)
GLX was built specially and before Mesa core because libglapi.a could
not be built with IN_DRI_DRIVER defined. This is no longer the case
since 6e99e6ddbf.
It works fine in my (limited) testing with both direct and indirect
rendering. I also compare the outputs after preprocessing (gcc -E) with
or without this commit, and they are identical.
The egl_g3d state tracker has support for KMS, and the support is based
on the egl state tracker. As egl_g3d provides more features, it should
be better to keep only egl_g3d to unify the efforts.
Need to use $CC for a shell variable, not $(CC) for a make variable
which the shell interprets as running the command "CC".
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
mesa/es should be built before the es state trackers. This is done by
separating those that need to be built early from SRC_DIRS to CORE_DIRS.
The new variable is not exported, and will be prepended to SRC_DIRS.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
The new EGL drivers use egl_g3d and respect EGL_DISPLAYS. They are
named after the display supported and the DRM name.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
This new (intermediate) EGL state tracker is the base work for EGL
drivers that uses Gallium. It makes it easier to support new window
systems.
Currently, there is support only for X11. This driver supports multiple
APIs (OpenVG, OpenGL, ...) and supports hardware acceleration through
winsys/drm.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
This was a silent dependency before, but now we really need 1.4.11 for
INTEL_swap_event support, so check for it.
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Support the new DRI2 protocol request, DRI2SwapBuffers, in both direct
and indirect rendering context. This request allows the display server
to optimize back->front swaps (e.g. through page flipping) and allows us
to more easily support other GLX features like swap interval and the OML
sync extension in DRI2.
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
This adds missing pkg-config lookup for intel and moves the radeon
lookup into a case...esac so it's only looked up when one or more of
the radeon drivers are enabled.
This change enabled gallium dri drivers by default under the
configure build system. Xorg drivers are built automaticaly
if a Xorg dev enviroment is installed and the Xorg version
is higher then 1.6.0.
Using a hack in the configure script the gallium intel
drivers have 3 options. Off, nothing is built. On, the
driver and binaries are built. Auto, only the driver but
not the binaries and winsys is built.
Since the i915g driver builds everywhere its can enable
the driver per default, so we can get build coverage.
But building the binaries per default is a pain for
distributions and testers since they conflict on the
install target with the old mesa drivers. Which are more
stable/faster/better.
So this change gives us the best of both worlds.
* Add an extra argument to configure which allows for specifying
different DRI driver search paths to libGL (FDO #24766)
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Not all drivers require the latest version, but distros are good at
tracking it since we've got a good record as far as not regressing, and
it sure beats having a weird compile failure in someone else's driver
later on.
When cross compiling on a 64bit machine, gen_matypes.c is build
for the host machine (64bit) but must generates code for the target
machine (32bit). This causes wrong offsets all over the place and
crashes googleearth on my machine. Solution is to add -m32 when
cross compiling.
Attached patch is compatible with linux-x86-32 and autoconf based
builds.
When cross compiling on a 64bit machine, gen_matypes.c is build
for the host machine (64bit) but must generates code for the target
machine (32bit). This causes wrong offsets all over the place and
crashes googleearth on my machine. Solution is to add -m32 when
cross compiling.
Attached patch is compatible with linux-x86-32 and autoconf based
builds.
This adds two --with configure options for setting defines for
MAX_WIDTH and MAX_HEIGHT. It's conceivably just as easy to define
these in CFLAGS manually, but this way users don't need to know
about internal Mesa details.
Patch updated by BrianP to set DEFINES, not CFLAGS.
(cherry picked from master, commit 7085dce750)
This adds two --with configure options for setting defines for
MAX_WIDTH and MAX_HEIGHT. It's conceivably just as easy to define
these in CFLAGS manually, but this way users don't need to know
about internal Mesa details.
Patch updated by BrianP to set DEFINES, not CFLAGS.
Since the drivers we produce on systems where we use
configure depend on none stable kernel API the driver
deliverables should not be built by default in the
mesa 7.7 release.
People wishing to shoot them self in the foot have to
pull the trigger themself, we just hand them the gun.
This driver does no transformation of the gallium calls
going to the real driver, like the identity matrix. It is
intended to be the basis for transforming and/or debug
drivers like trace and rbug.
Authors of this patch are:
Michal Krol, orignal heavy lifting.
José Fonesca, object wrapping code stolen from trace.
Jakob Bornecrantz, put it all toghether and renamed a stuff.
Fix mklib to deal with NOPREFIX and use --enable-auto-image-base for cygwin
Teach configure.ac some basic facts about cygwin
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
Now that the dlopen wrappers are built into libmesa.a, we need to link
standalone libOSMesa with libdl to resolve dlopen and friends on
platforms that need it.
(cherry picked from commit 4795dd5950)
Now that the dlopen wrappers are built into libmesa.a, we need to link
standalone libOSMesa with libdl to resolve dlopen and friends on
platforms that need it.
The special feature of bin/minstall to copy symlinks is only ever needed
when installing libraries which may have .so symlinks. All the headers
and directories can use a normal install program.
These two modes are separated as $(INSTALL) and $(MINSTALL) to allow the
user (or autoconf) to override installing normal files as they please.
An autoconf check for the install program has been added and will be
used in preference to minstall when available.
Fixes bug 16053.
Here is a couple of fixes for GNU/Hurd:
- dri_interface.h: no libdrm support either.
- configure.ac:
- GNU/Hurd is a GNU OS with _GNU_SOURCE and PTHREADS.
- GNU needs a couple of flags like other OSes
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Move flags for linking standard C/C++ libraries from configure.ac to mklib
Use -norunpath flag when linking with Sun C++ compiler
Convert mklib -exports list into a linker mapfile
Set FINAL_LIBS correctly when -noprefix is used
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
This catches the linux-uclibc case and any others that were being set
prior to 98fcdf3f. Fixes bug 20345.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
This rewrites the sparc GLAPI code so that it's PIC friendly and works
with all of the TLS/PTHREADS/64-bit/32-bit combinations properly.
As a result we can turn SPARC asm back on. Currently it's only
enabled on Linux, as that's the only place where I can test this
stuff out.
For the moment the cliptest SPARC asm routines are disabled as they
are non-working. The problem is that they use register %g7 as a
temporary which is where the threading libraries store the thread
pointer on SPARC. I will fix that code up in a future change as it's
a pretty important routine to optimize.
Like x86 we do the runtime patch as a pthread once-invoked initializer
in init_glapi_relocs().
Unlike x86, however, our GLAPI stubs on SPARC are just two instruction
sequences that branch to a trampoline and put the GLAPI offset into a
register. The trampoline is what we run-time patch. The stubs thus
all look like:
glFoo:
ba __glapi_sparc_foo_stub
sethi GLAPI_OFFSET(glFOO) * PTR_SIZE, %g3
This actually makes generate_entrypoint() a lot simpler on SPARC. For
this case in generate_entrypoint() we generate stubs using a 'call'
instead of the 'ba' above to make sure it can reach.
In order to get a proper tail call going here, in the unpatched case,
we do several tricks. To get the current PC, for example, we save the
return address register into a temporary, do a call, save the return
address register written by the call to another temporary, then
restore the original return address register value. This is to
avoid having to allocate a stack frame.
This is necessary for PIC address formation.
This new GLAPI scheme lets us get rid of the ugly SPARC GLAPI hacks in
__glXInitialize() and one_time_init().
Signed-off-by: David S. Miller <davem@davemloft.net>
It is only the i965simple pipe driver that was broken
in the gallium-texture-transfere merge that is being
disabled, mothing more nothing less.
FYI, there never where working i965 hardware support
in gallium anyways.
Commit 90b2beb661 changed the assembly
variables. Without this change, the glapi assembly wasn't being built
when it was supposed to, resulting in missing symbols in libGL.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
This is not the proper approach to disabling assembly. If there are bugs,
they should be fixed or the assembly can default to off.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
On newer GCC releases, the compiler's headers have been split between
the include and include-fixed directories. Add both if the directories
exist.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Signed-off-by: Julien Cristau <jcristau@debian.org>
The pkg-config files have been filled in more thoroughly to allow users
to use mesa more effectively. By adding metadata to Requires.private,
Libs.private and Cflags, we can ensure that all the libraries and
headers will be found in all situations. However, the full substitutions
are only done when using the configure script.
This also fixes the glu pkg-config file to account for using GL or
OSMesa.
Fixes bug 18161.
Add an --enable-motif option, which will enable the Motif widgets in
libGLw and link it with libXm. The Motif installation information will
be gathered from the motif-config script (this comes with LessTif) or
fallback to the standard autoconf checks.
To allow the location of the Motif headers to be set from configure, the
default setting of -I/usr/include/Motif1.2 has been moved into
configs/default and then passed to the Makefile through the MOTIF_CFLAGS
variable.
Establish the shell that make will use from configure. This is exactly
how autoconf/automake operate, with the environment variable
CONFIG_SHELL respected to override the autoconf checks. In the usual
case where the user just executes `./configure', autoconf will pick a
shell from the current shell, sh, bash, ksh or sh5 that meets its base
criteria.
The special Solaris case of looking for a POSIX shell has been changed
to just set the SHELL variable since autoconf substitutes this already.
The EXTRA_CONFIG_LINES substitution is dropped as it should no longer be
needed.
The documentation in autoconf.html is much more explicit about how the
different configure options control the build. This adds a notice at the
end of the `./configure --help' output to tell the user about it.
Explicitly allow the argument to --with-dri-drivers to contain
comma-separated or space-separated drivers. A space-separated driver
list worked by chance before.
Currently the installation directories for libraries and headers are
resolved within the install commands. For instance, the libraries will
be installed to $(INSTALL_DIR)/$(LIB_DIR). This limits the flexibility
of the installation, such as when the libraries should be installed to a
subdirectory like /usr/lib/tls.
This adds the make variables $(INSTALL_LIB_DIR) and $(INSTALL_INC_DIR)
to define the locations that the libraries and headers are installed.
For the static configs, this resolves exactly as before to
$(INSTALL_DIR)/include and $(INSTALL_DIR)/$(LIB_DIR). For autoconf, they
are derived directly from the --libdir and --includedir settings.
The architecture- and platorm-specific settings were scattered
throughout the configure script. This moves them near the beginning
before any of the driver-specific settings.
I'll release a libdrm 2.3.1 without TTM apis included from a special
drm branch that should allow mesa 7.1 to build against it.
I've had to turn off DRI2 stuff.
Since the only valid consumer of the DRI drivers is the X.Org xserver,
this changes the default DRI driver directory to match xorg-server:
${libdir}/dri. The old default of /usr/X11R6/modules/dri was wrong for
nearly all current systems.
This commit adds an autoconf macro, MESA_PIC_FLAGS, which sets the
PIC flags according to platform and static/shared setting. The platform
specifics are taken straight from libtool.m4 and stripped down to just
the flags and platforms we cover in Mesa. This should hopefully make it
possible to use autoconf on non-GCC platforms.
The macro is added external to configure.ac in acinclude.m4 since it's
pretty bloated.
Note to BSDers: Previously, x86 defaulted to non-PIC on FreeBSD. I
didn't carry that preference into this macro. Instead, you can just use
--disable-pic where desired.
Two fixes to the asm configuration:
- Disable when the user is cross-compiling for x86 or x86_64 since it
requires running an executable compiled for the target host.
- If the user has specified --enable-32-bit on x86_64 or --enable-64-bit
on x86, respect that and choose the correct asm architecture.
The DRI modules can only be built against libdrm master (currently
version 2.3.1), so this should be enforced to save people from trying to
build against older versions.
Added a section at the top of the script to consolidate all required
versions.
The variable no_x was being set to yes when libX11 was not found through
pkg-config. This causes AC_PATH_XTRA to skip its search for the X11
libraries, which was not the intended effect. Also switched to using the
PKG_CHECK_EXISTS autoconf macro.
Added the make script version.mk to print the various version numbers
from configs/default. This is used to substitute the version in autoconf
rather than duplicating it in both places.
Since the autoconf config inherits from default, we don't need to
duplicate and substitute the MESA_* version numbers in configure.ac.
The version number is only needed in configure for the help text.
The autoconf macro AC_PATH_PROGS handles the case of searching for
multiple program names already, so we don't need to open code it. Also
changed the search to AC_CHECK_PROGS so that it doesn't set the full
path unless the user specified. Finally, report back the found value at
the end for what the user should run.
Add DRI2 direct rendering support to libGL and add DRI2 client side
protocol code. Extend the GLX 1.3 create drawable functions in
glx_pbuffer.c to call into the DRI driver when possible.
Introduce __DRIconfig, opaque struct that represents a DRI driver
configuration. Get's rid of the open coded __GLcontextModes in the
DRI driver interface and the context modes create and destroy
functions that the loader was requires to provide. glcore.h is no
longer part of the DRI driver interface. The DRI config is GL binding
agnostic, that is, not specific to GLX, EGL or other bindings.
The core API is now also an extension, and the driver exports a list
of extensions as the symbol __driDriverExtensions, which the loader
must dlsym() for. The list of extension will always include the DRI
core extension, which allows creating and manipulating DRI screens,
drawables and contexts. The DRI legacy extension, when available,
provides alternative entry points for creating the DRI objects that
work with the XF86DRI infrastructure.
Change DRI2 client code to not use drm drawables or contexts. We
never used drm_drawable_t's and the only use for drm_context_t was as
a unique identifier when taking the lock. We now just allocate a
unique lock ID out of the DRILock sarea block. Once we get rid of the
lock entirely, we can drop this hack.
Change the interface between dri_util.c and the drivers, so that the
drivers now export the DriverAPI struct as driDriverAPI instead of the
InitScreen entry point. This lets us avoid dlsym()'ing for the DRI2
init screen function to see if DRI2 is supported by the driver.
Rather than constructing the GCC include path from `-print-search-dirs',
we can get the path directly from `-print-file-name=include'. This is
used in the Linux kernel build, for example. If no output is returned
from the command, then we don't append a -I path the the makedepend
options.
The platform-specific overrides to the DRI drivers were being ignored
because the default was being set first. Instead, have the default be a
fallback after the platform checks.
The conditional in src/mesa/Makefile currently hardcodes the cases where
libOSMesa can be built on libGL. Likewise, the xlib case always includes
libOSMesa in the stand-alone target.
This changes the conditional to a loop over the DRIVER_DIRS variable.
This means that any driver configuration can enable or disable osmesa.
The current "stand-alone" rule is changed so that DRIVER_DIRS=x11 and
DRIVER_DIRS="x11 osmesa" are both respected.
The configure option is changed to --enable-gl-osmesa as this change
allows libOSMesa to be built upon any of the libGL-enabling drivers.
Two new configure options to add -m32 or -m64 to the CFLAGS and CXXFLAGS
when GCC is in use. By default, the user supplied options are
environment variables are respected, but these options are quick helps
for the common case of x86/x86_64 using GCC.
Changed the text for the default demos to be "auto" instead of "yes"
since that didn't make much sense in this context. Added an example
value for the --with-dri-drivers option as suggested by Brian.
nobled
Jon TURNEY
Dave Airlie
Eric Anholt
Ian Romanick
Li Peng
Chia-I Wu
Thomas Balling Sørensen
Jakob Bornecrantz
Kristian Høgsberg
John Hein
Jakob Bornecrantz
Corbin Simpson
Dan Nicholson
Younes Manton
Vinson Lee
Xavier Chantry
George Sapountzis
Luca Barbieri
Jeff Smith
Michel Dänzer
Keith Whitwell
Brian Paul
Chia-I Wu
Francisco Jerez
Alan Coopersmith
Jesse Barnes
José Fonseca
Fabio Pedretti
Sedat Dilek
Chow Loong Jin
Tom Fogal
Thierry Vignaud
Marc Dietrich
Alex Deucher
Ben Skeggs
Jerome Glisse
Samuel Thibault
Dave Airlie
Robert Noland
Joel Bosveld
David S. Miller
Siddhartha Chaudhuri
Julien Cristau
Kristian Høgsberg
RALOVICH, Kristóf
Florent Thoumie
George Sapountzis
Hasso Tepper
Adam Jackson