Commit graph

370 commits

Author SHA1 Message Date
Tim Angus
5cc1fb1c05 * Improved the robustity^Wrobustosity^Wrobustness of the OpenAL music and raw
stream systems
* Implemented S_AL_BufferEvict for more graceful behaviour when OpenAL runs
  out of memory
2005-12-03 00:19:27 +00:00
Tim Angus
0f8c177cf3 * Fix to an OpenAL "bug" 2005-12-02 00:14:24 +00:00
Tim Angus
796ba11301 * Ditch Mac OS 9 support 2005-12-01 21:36:16 +00:00
Tim Angus
1b314e8934 * Updated AL headers
* Added a Makefile option USE_LOCAL_HEADERS which can be disabled to use system
  headers if desired
2005-12-01 20:53:28 +00:00
Tim Angus
acb810b972 * Fix compilation when using SDL and dlopening OpenAL 2005-12-01 20:45:01 +00:00
Tim Angus
3161a34a4f * Fixed bug where if two looping sounds were added to a single entity they
would fight for a source and you'd get a horrible distorted sound that was
  each of the two sounds interleaved
* Changed the default of s_alPrecache to 1 -- having it off seems to cause
  problems for some and it doesn't actually improve things much
* Changed a few of the variable names in snd_openal.c to be more sane
2005-12-01 18:22:42 +00:00
Tim Angus
31b7243ef3 * Fix a warning on non-PPC builds 2005-12-01 18:01:43 +00:00
Ryan C. Gordon
5bc17ccf4b Fixed detection of Altivec on Mac OS X, and added attempt at general detection
with SIGILL/setjmp hackery for PowerPC Linux, etc.
2005-12-01 11:49:02 +00:00
Ryan C. Gordon
4bea1e4ff9 r_smp now works as expected with sdl_glimp.c on Mac OS X. 2005-12-01 11:16:36 +00:00
Ryan C. Gordon
820f41162e Another "long double" ... 2005-12-01 10:59:50 +00:00
Ludwig Nussel
b130d6be52 fix the stupid i386/x86/amd64/x86_64 weirdness 2005-12-01 09:54:14 +00:00
Ludwig Nussel
f5228d8745 don't use ImageMagick 2005-11-30 21:21:46 +00:00
Ludwig Nussel
dfa3d1d875 add "very high quality" option (patch by Pascal de Bruijn) 2005-11-30 20:28:21 +00:00
Tim Angus
ae39051f1e * Support RIFF files with zero length data chunks (yes they exist, and yes,
they're legal)
* Colourise the OpenAL warnings so they're somewhat more obvious
2005-11-30 01:57:53 +00:00
Ryan C. Gordon
4784374221 Make sure Q3_BIG_ENDIAN and Q3_LITTLE_ENDIAN are defined correctly for
PowerPC/x86 Mac OS X. Fixes Bugzilla #2510.
2005-11-29 03:31:31 +00:00
Ryan C. Gordon
7590843afe Fixed comment. 2005-11-28 11:20:35 +00:00
Ryan C. Gordon
6c0079508c First shot at SMP renderer with sdl_glimp. Does not work yet! 2005-11-28 10:56:27 +00:00
Ryan C. Gordon
4517e7a34b Fixed some compiler warnings in lcc...mostly "long double" stuff. 2005-11-28 09:53:09 +00:00
Ryan C. Gordon
807ba08558 Must explicitly define MACOS_X here or lcc won't build on the Mac. 2005-11-28 09:50:55 +00:00
Ludwig Nussel
be42cb3201 files to create a loki-setup based installer. run "make installer" 2005-11-27 15:34:16 +00:00
Ludwig Nussel
64025e68f0 fix warning about cast from integer to pointer 2005-11-27 15:01:56 +00:00
Tim Angus
68a59e1c93 * Moved win32 GL extension prototypes from tr_init.c to win_qgl.c 2005-11-26 17:51:26 +00:00
Tim Angus
382c6adb54 * Disable ccache by default. If you want it, add USE_CCACHE=1 to Makefile.local
* Remove -gfull from linux section in Makefile -- it's darwin only
* Cast away some warnings that surfaced from using "new" AL headers
* Various whitespace and consistency fixes
2005-11-26 15:01:28 +00:00
Ryan C. Gordon
877f360812 SDL and OpenAL headers. 2005-11-26 07:59:00 +00:00
Ryan C. Gordon
721b9a7d01 Mac OS X work...lots of little changes that touch a lot of random places.
Still work to be done, but this at least matches the PowerPC Linux status
 now.

MacOS-specific directory (and XCode project) is gone...this now uses SDL,
 OpenAL, and the Unix Makefiles.

--ryan.
2005-11-26 07:46:21 +00:00
Ryan C. Gordon
b20b86bbbe Don't byteswap when resampling audio...we byteswapped it previously in the
codecs.
2005-11-20 15:35:45 +00:00
Ludwig Nussel
108c051345 make internal functions static 2005-11-19 14:53:46 +00:00
Tim Angus
fee4381e18 * Remove the debug text that nobody seemed to notice 2005-11-19 02:14:34 +00:00
Tim Angus
82893aebc9 * Bleh.. whitespace 2005-11-16 20:25:36 +00:00
Tim Angus
387d8041ce * Partial implementation of FS_Seek for files in pk3s
* A couple of RIFF decoder tweaks/fixes
2005-11-16 20:23:37 +00:00
Tim Angus
736d34c989 * Small "bug" fix to MSG_WriteBits 2005-11-15 23:49:14 +00:00
Tim Angus
b5c7d9dcd7 * Implementation of r_dlightBacks from Shane Isley 2005-11-15 23:30:35 +00:00
Tim Angus
47c70ae076 * ... 2005-11-15 22:22:45 +00:00
Tim Angus
6be591daa1 * Don't link OpenAL against q3ded 2005-11-15 22:09:45 +00:00
Tim Angus
25785fe0e6 Potential fix to https://bugzilla.icculus.org/show_bug.cgi?id=2486 2005-11-15 17:02:45 +00:00
Tim Angus
0163f77f4e * Oops, missed a bit 2005-11-14 18:35:31 +00:00
Tim Angus
9d7bb06743 * Replace a redundant function in the wav decoder 2005-11-14 18:26:25 +00:00
Tim Angus
e8a953fea6 * Quiet some warnings on OS X 2005-11-13 23:05:00 +00:00
Tim Angus
668fdc102c * Unbreak the MinGW build more...
* Unbreak USE_OPENAL_DLOPEN=1
2005-11-13 22:16:24 +00:00
Tim Angus
33c1528673 * Unbreak the MinGW build 2005-11-13 21:23:16 +00:00
Ludwig Nussel
1c2722a9ba link against libopenal by default rather than dlopening it 2005-11-13 21:01:44 +00:00
Ludwig Nussel
17719e0c61 soname is libopenal.so.0 2005-11-13 21:01:07 +00:00
Tim Angus
84c4f21082 * OpenAL support, from BlackAura aka Stuart Dalton <badcdev@gmail.com>
+ An abstract codec system, simplifying support for new formats
  + Changes versus BlackAura's patch:
    o Consolidated the OpenAL parts into one file
    o Changed the function naming scheme to more closely resemble Q3
    o Changed the interface to fall back on the "base" sound system
      if loading OpenAL fails
  + This is enabled on Linux and MinGW for now, but should work on the
    other *nixs with appropriate additions to the Makefile
  + NOT enabled on OS X or MSVC Windows builds
  + Probably breaks the Windows build again
* Retabulated sdl_snd.c and made the messages less verbose since
  there do not seem to be many having problems with SDL sound now
2005-11-13 18:58:14 +00:00
Ludwig Nussel
79ceef93cc store temporary files in /tmp to save space in ~ 2005-11-13 18:33:28 +00:00
Ludwig Nussel
c95d0c29e4 use common macro for alignment 2005-11-13 17:51:21 +00:00
Ludwig Nussel
bfb24584f5 - add -Wimplicit -Wstrict-prototypes CFLAGS again
- remove -pipe. not needed IMO.
2005-11-11 15:07:15 +00:00
Tim Angus
e905b369c3 * Makefile target dependancy fix 2005-11-07 17:44:05 +00:00
Tim Angus
fd9c1b00bb * Fix TA font loader to use new endian defines 2005-11-06 21:18:21 +00:00
Ludwig Nussel
b2f39d7ddd use system memcpy and memset 2005-11-06 18:22:49 +00:00
Tim Angus
5201000f39 * Ignore in_dgamouse setting if dga isn't available
* Changed status of Solaris sparc
2005-11-06 17:31:48 +00:00
Tim Angus
bbf2e36b5b * I'm so anal 2005-11-06 16:51:28 +00:00
Tim Angus
f20cca46e8 * Install the tools on every build, so a distclean isn't necessary if the tool
source changes now
* Basically rewrote the lcc Makefile to be more sane
* Removed various bits of lcc that weren't built/needed
2005-11-06 16:50:58 +00:00
Ludwig Nussel
590988222f - crash solved 2005-11-06 16:32:42 +00:00
Ludwig Nussel
cc8b03c655 - don't spam the tty with that message 2005-11-06 16:24:49 +00:00
Ludwig Nussel
48ee88ad77 - move code for standalone compilation out of vm_x86_64.c
- reset vm->compiled if something goes wrong in VM_Compile
2005-11-06 13:50:28 +00:00
Ludwig Nussel
8a18ecb961 - fall back to interpreter if VM_Compile resets vm->compiled 2005-11-06 13:46:17 +00:00
Ludwig Nussel
87b12105bc - set umask to 0077 before writing the cdkey, the rest of the world
doesn't need to read it.
2005-11-06 13:45:20 +00:00
Tim Angus
4eaebe4a31 * Tidy-up of q_platform.h
* Introduced Q3_BIG_ENDIAN and Q3_LITTLE_ENDIAN #defines
* Changed some stricmp to Q_stricmp so that #define stricmp strcasecmp could be
  removed
2005-11-06 00:29:59 +00:00
Ludwig Nussel
dfe20a177c - put ALIGN macro into q_shared.h
- yet another alignment fix for botlib
2005-11-05 19:06:10 +00:00
Tim Angus
81c1b3128b * Enable use of asm on Solaris-x86 2005-11-05 18:41:37 +00:00
Tim Angus
a31e38986c * Patch from AJ <anthonyj@planetquake.com> which replaces a bunch of hard coded
constants with #define constants
2005-11-05 15:54:56 +00:00
Tim Angus
fce1a89fad * Fix a stupid typo 2005-11-05 15:07:05 +00:00
Tim Angus
fe09cd3217 * Patch from vapier that aids packaging
* Truth value of USE_SDL changed to 1 from true, for consistency
2005-11-05 01:06:09 +00:00
Tim Angus
d41de844d3 * Oops, missed a bit 2005-11-05 00:42:28 +00:00
Tim Angus
a87185a96f * Centralise architecture defines in q_platform.h 2005-11-05 00:29:40 +00:00
Tim Angus
e4b7a04d5c * Replaced a bunch of inline and __inline with ID_INLINE
* Replaced a bunch of __i386__ with id386
* General tidy up of asm preprocessor decisions
* Removed C_ONLY from the dedicated server build
2005-11-04 22:32:00 +00:00
Zachary Slater
857a5e264e Updated msvc project files from > Michael Zoech! 2005-11-04 02:35:30 +00:00
Ludwig Nussel
b0dfb2579b found the bug. alignment reenabled. 2005-11-03 22:32:39 +00:00
Ludwig Nussel
2984b5bea2 disable the alignment "fix" until its clear why it causes crashes 2005-11-03 22:26:59 +00:00
Tim Angus
2e3a08e4bd * Juggled CFLAGS a bit so that -fno-strict-aliasing is always on 2005-11-03 18:34:03 +00:00
Ludwig Nussel
1c94cfbcbd more alignment fixes 2005-11-02 21:46:05 +00:00
Ludwig Nussel
e3e3d8b9aa actually when calling from the vm to the engine there are more than 10
arguments
2005-11-02 21:45:48 +00:00
Ludwig Nussel
5f79f9958b pass only 10 arguments. vm_* doesn't use more anyways. 2005-11-02 20:58:27 +00:00
Ludwig Nussel
3ce5d5dccd pass only 10 arguments. vm_* doesn't use more anyways. 2005-11-02 20:56:13 +00:00
Ludwig Nussel
d29dfeeb75 round string length for pointer alignment 2005-11-02 20:36:30 +00:00
Ludwig Nussel
e121614406 allocate memory pointer aligned 2005-11-02 20:34:25 +00:00
Tim Angus
3c05bf1870 * Fies to the xcode project to update it with the recent changes 2005-11-02 18:32:03 +00:00
Tim Angus
2cc87dfc9d * Removed the last of the binary target stuff from lcc (hopefully) 2005-11-02 16:05:14 +00:00
Tim Angus
2d2df7682d * Removed the lcc self tests -- there are none that test bytecode 2005-11-02 15:28:16 +00:00
Tim Angus
8b63fdab58 * Fix for lcc on windows so that it can execute files with spaces in them 2005-11-02 02:08:29 +00:00
Tim Angus
87cd6f8a93 * Split off q_platform.h from q_shared.h
* Removed lcc PATH hack and replaced with something slightly less hacky
* Removed all platform specific hostfiles from lcc and replaced with bytecode.c
  (from ankon)
* Turned lcc option "-S" on permanently
* Improved q3cpp so that it recursively adds include dirs to its list
2005-11-01 22:09:15 +00:00
Ludwig Nussel
48683c91a8 use memcpy to fix use of misaligned data 2005-11-01 19:59:14 +00:00
Ludwig Nussel
f4dbf0db95 5+2==7 :-) 2005-11-01 19:47:47 +00:00
Ludwig Nussel
39f5333862 use memcpy to fix use of misaligned data 2005-11-01 19:44:36 +00:00
Tim Angus
acf1f46a2b * More Solaris work 2005-11-01 16:05:06 +00:00
Tim Angus
0a3e96af77 * Further work on the Solaris port from Vincent S. Cojot 2005-10-31 16:46:05 +00:00
Tim Angus
3647cca9f7 * Applied https://bugzilla.icculus.org/show_bug.cgi?id=2464
* Removed rule to build C++ (for splines) from the Makefile
2005-10-30 14:36:49 +00:00
Tim Angus
c41483634c * General decrufting:
* Removed Q3_STATIC and associated defines
* Removed MAC_STATIC
* Replaced __LCC__ with Q3_VM
* Removed bspc and splines directories
2005-10-29 23:13:09 +00:00
Tim Angus
63fdc74843 * Fix to the VM interpreter on the linux ppc dedicated server 2005-10-29 22:06:43 +00:00
Tim Angus
c5bdb034f2 * Beginnings of Solaris support from Vincent S. Cojot
* Note this patch also splits USE_SDL into USE_SDL_VIDEO and USE_SDL_AUDIO
2005-10-29 22:05:20 +00:00
Ludwig Nussel
58512aff09 use memcpy for OP_BLOCK_COPY 2005-10-29 17:42:12 +00:00
Ludwig Nussel
b6228f1408 add HAVE_VM_COMPILED variable so one can override it in Makefile.local 2005-10-29 17:41:17 +00:00
Tim Angus
05e8ab9538 * Added STATUS
* Updated TODO
* Moved ChangeLog to root
* Updated ChangeLog
* s/Foobar/Quake III Arena Source Code/
* Biggest patch EVAR. I wonder how many mail boxes this will fill...
2005-10-29 01:53:09 +00:00
Tim Angus
b72f59f524 * In hindsight this is probably wise too 2005-10-29 01:02:12 +00:00
Tim Angus
ffdc69f138 * Deleted code/unix/vm_x86.c - wtf was that about anyway?
* Removed HAVE_VM_NATIVE
* Removed DLL_ONLY
* Replace HAVE_VM_COMPILED with NO_VM_COMPILED -- this means the JIT compiler
  should be enabled on the OS X and VC builds now
* Remove the remainder of the freetype building stuff
2005-10-29 00:23:10 +00:00
Tim Angus
090f4d7266 * Compiler warning fix from ankon 2005-10-28 22:10:39 +00:00
Tim Angus
4c63d1ed40 * Fix to https://bugzilla.icculus.org/show_bug.cgi?id=2454 2005-10-28 21:09:50 +00:00
Tim Angus
b1b7f975f1 * Fix the x86 vm 2005-10-27 22:38:25 +00:00
Tim Angus
556b3307c5 * ...and the other common headers 2005-10-27 22:05:38 +00:00
Tim Angus
bc9ee5644b * Ignore __attribute__ when not using gcc 2005-10-27 21:49:36 +00:00
Ludwig Nussel
0bf8e3a8b0 add x86_64 vm. experimental, not enabled by default. you need as for it
to work.
2005-10-27 21:13:47 +00:00
Tim Angus
9af615f27c * Add notional support for various architectures from vapier. Whether or not
these architectures work in practice is another matter :).
2005-10-26 23:15:44 +00:00
Tim Angus
f455a27edd * Use platform headers to source various libc prototypes instead of unsafe
static ones (from vapier)
2005-10-26 23:08:56 +00:00
Tim Angus
0167152647 * Small changes to distclean target from vapier 2005-10-26 22:59:30 +00:00
Tim Angus
c02a7fce3c * Split lcc $AR into $AR and $ARFLAGS (from vapier) 2005-10-26 22:52:17 +00:00
Tim Angus
7b84b0295f * Fixed the OS X build as broken by revision 181 2005-10-26 20:52:53 +00:00
Tim Angus
3a2ad50406 * Moved various source files from their mod sdk locations to places more
appropriate for open source Q3
* This patch looks bigger than it really is, however it will probably break the
  VC and OS X builds (easy to fix though)
2005-10-26 19:45:21 +00:00
Tim Angus
3ff42330e5 * Fixed some warnings on the OS X build 2005-10-26 19:32:15 +00:00
Tim Angus
9f55ae2dce * Fixed a bug with QVM loading on big endian architectures 2005-10-26 17:09:50 +00:00
Tim Angus
d5af2e3d45 * Removed some Conscript 2005-10-25 13:34:47 +00:00
Tim Angus
6b5cf6e248 * Fixed unused variable warning 2005-10-23 17:23:23 +00:00
Tim Angus
69b81f2bd9 * Fix to https://bugzilla.icculus.org/show_bug.cgi?id=2401 2005-10-23 17:18:23 +00:00
Ludwig Nussel
e27a35a207 remove french keyboard hack. we cannot know what this does on other
layouts
2005-10-23 15:27:59 +00:00
Ludwig Nussel
80cf67e2ac map extra mouse buttons to K_AUX* 2005-10-23 15:25:23 +00:00
Tim Angus
f01ed7b3da * Tab completed commands no longer treated like cvars 2005-10-21 19:52:52 +00:00
Tim Angus
d6b2a1e527 * Deleted some exes that were in the repository 2005-10-21 18:07:25 +00:00
Tim Angus
f306b2895f * Tiny patch to print values of cvars when tab completing ala ET 2005-10-21 00:09:34 +00:00
Tim Angus
64f167f183 * Fix to potential out of range index bug from misanthropia 2005-10-20 16:19:43 +00:00
Tim Angus
769372e2f9 * Fix to buffer overflow in lightmap loading code from misanthropia 2005-10-20 16:14:05 +00:00
Tim Angus
653c5621ac * Fixed the fix for the shader overrun. If this isn't a case for unified diffs,
I don't know what is :).
2005-10-20 01:23:17 +00:00
Tim Angus
b9c9f86cd1 * Cap the maximum snaps a client can request at the server framerate rather
than a hard-coded 30 (from misantropia). I'm slightly nervous about this
  since it changes the behaviour versus 1.32b in that a client requesting e.g.
  25 snaps will actually get 20 (default sv_fps) with this patch applied. I'm
  not sure why anyone would want this though -- it may even destablise their
  inter/extrapolation. I'll leave the bug (2422) open for a while and if no one
  complains it can be closed.
2005-10-19 22:25:32 +00:00
Tim Angus
7d1b5b4acb * Fix to buffer overrun in shader parser from misanthropia 2005-10-19 22:15:51 +00:00
Tim Angus
80919a1ce3 * Silly little patch to convert Q3 style color codes to ANSI escape codes. Note
this doesn't work too well with baseq3 since it logs color strings with no
  color termination.
2005-10-19 01:51:19 +00:00
Zachary Slater
bcc229d5bb Minor edit, thanks ankon! 2005-10-18 02:46:20 +00:00
Tim Angus
e652200405 * Added option to q3asm to instruct it to create qvms compatible with Q3 1.32b 2005-10-16 01:13:51 +00:00
Tim Angus
354196b3a9 * Added a little hack to support the console key on french keyboards 2005-10-14 23:00:39 +00:00
Tim Angus
20c5bc38cf * Removed hard coded mouse acceleration in the unix build(s)
* Added CVAR_ROM cl_platformSensitivity to normalise the scale
  of cl_sensitivity across platforms
2005-10-14 22:58:28 +00:00
Tim Angus
535545684f * Adjusted VM_LoadQVM interface 2005-10-14 17:09:03 +00:00
Tim Angus
f7c82c1adc * Fixed some annoying code duplication 2005-10-13 23:08:45 +00:00
Tim Angus
35cb35ff61 * (Non-trivial) fix to the "opStack corrupted in compiled code" bug 2005-10-13 14:47:00 +00:00
Tim Angus
8295a83897 * Removed Q3ASM_TURBO defines from q3asm
* Removed q3asm qfiles.h and redirected it to qcommon qfiles.h
2005-10-13 01:55:31 +00:00
Six
ecd73786d5 Fix build problems under OS X 2005-10-11 01:52:29 +00:00
Tim Angus
3857ce21c4 * Replaced drawSurfs qsort algorithm with a radix sort
+ Performance is comparable, although generally slightly
    worse, but...
  + Radix is a stable sort algorithm, so overlapping
    coplanar drawSurfs (i.e. with the same sort key) no
    longer flicker indeterminately
  + Also removes the dubious comment: "FIXME: this was 
    lifted and modified from the microsoft lib source..."
2005-10-10 00:37:54 +00:00
Ludwig Nussel
b1ea2ed136 - get rid of vm_none
- if VM_PPC is set compiled vm support is enabled on ppc (default off atm)
2005-10-09 13:07:54 +00:00
Ludwig Nussel
44f5273e02 ugly: undefine memmove to avoid build failure with gcc fortify
extensions enabled
2005-10-09 13:06:33 +00:00
Ludwig Nussel
f62601ac53 don't call VM_Compile if it's not supported so the VM_Compile symbol
isn't needed
2005-10-09 13:05:44 +00:00
Tim Angus
f6cb6eee96 * q3lcc is no longer the default target in the Makefile
* Changed readme to suggest using the root Makefile instead of the one in
  code/unix
2005-10-08 14:36:15 +00:00
Six
8866ce3f8b Mac OS X needs Com_Memcpy and Com_Memset to be set to memcpy and memset. 2005-10-08 06:30:38 +00:00
Tim Angus
29db2dbbfd * Patch from Andreas Kohn to fix q3asm build on FreeBSD
* Removal of SYSTEM variable from the lcc makefile due to problems with FreeBSD
2005-10-07 01:03:49 +00:00
Tim Angus
3271b52bf5 * ATI skybox bug fix 2005-10-06 15:33:54 +00:00
Tim Angus
0b6beec202 * Usage of find on FreeBSD fix from Andreas Kohn 2005-10-06 15:16:12 +00:00
Tim Angus
707ee2fdbb * Append LCC directory to the PATH variable on all platforms 2005-10-05 21:46:03 +00:00
Tim Angus
151c9461c5 * Fixed a foul-up with the lcc Makefile 2005-10-05 21:07:47 +00:00
Tim Angus
260ca8d730 * A couple of small Makefile fixes 2005-10-05 18:13:34 +00:00
Tim Angus
490c84d907 * Applied q3asm-turbo patches from http://www.icculus.org/~phaethon/q3/q3asm-turbo/q3asm-turbo.html
* Added -m option to q3asm to write a map file (which is now disabled by default)
* q3asm now returns an error code on failure
2005-10-05 17:59:10 +00:00
Tim Angus
9d41e17e1f * Fixed some vm build warnings 2005-10-05 14:50:45 +00:00
Tim Angus
79ba2d974d * Fixes to the MinGW vm compiler
* General MinGW tweaks
2005-10-05 03:13:20 +00:00
Tim Angus
41cbd8c9fd * VM game modules now built from the main Makefile 2005-10-04 23:54:51 +00:00
Tim Angus
12b5f3bded * MinGW fixes to the tools Makefiles 2005-10-04 19:15:34 +00:00
Tim Angus
e414f46c4d * Added install/uninstall rules to the lcc and q3asm Makefiles
* Moved the q3asm dir to asm so that is doesn't clash with the binary when
  installed
2005-10-04 18:34:21 +00:00