10a81394ff
removed test rig crap
...
This was SVN commit r1207.
2004-10-03 23:28:57 +00:00
b6c1ad7126
had a look at CPU freq detect due to win32asm thread.
...
added comments + slight improvement (get system time first, then rdtsc)
=> good bit more accurate
This was SVN commit r1206.
2004-10-03 13:06:37 +00:00
df1b502bbd
Linux Compat and some updated PARANOIA code
...
This was SVN commit r1189.
2004-09-22 15:20:58 +00:00
c28c5a1beb
fix of VC2k5 fix (both pointers to tokenized string are now char*)
...
This was SVN commit r1186.
2004-09-22 00:46:12 +00:00
a1bd91d198
VC2k5 fix: don't keep converting char* -> const char* -> char*
...
This was SVN commit r1184.
2004-09-21 22:10:49 +00:00
0868b2ee6c
VC2k5 fix: strchr return value saved as const char* (cf Stu's CVS report #161 )
...
This was SVN commit r1183.
2004-09-21 18:43:58 +00:00
ff4af72da1
fix incorrect round up calculation
...
This was SVN commit r1182.
2004-09-21 16:59:50 +00:00
777f06143d
- A lot of small network fixes
...
- Connecting client(s) to a server and giving orders to a shared set of
entities now works - Yeah!
- Removed thread sync from CNetServer
This was SVN commit r1180.
2004-09-21 14:40:43 +00:00
8b2e918e05
cleanup + bulletproofing
...
This was SVN commit r1176.
2004-09-21 13:40:14 +00:00
dce92ce677
fix 2 stupid bugs causing crashes when loading from loose files - wasn't detecting all causes where transfers are unaligned => read rounded-up amount of data into original buffer as opposed to realign buffer => clobbered the heap.
...
This was SVN commit r1175.
2004-09-21 11:58:22 +00:00
7925a3fc4d
Avoid crashing when something minor that should never fail fails (though the cause of that failure is fairly major (but currently unknown))
...
This was SVN commit r1174.
2004-09-20 18:27:22 +00:00
ae2d42bbad
fix incorrect buffer free (now handled by ALBuffer); work in progress
...
This was SVN commit r1172.
2004-09-19 19:29:03 +00:00
85cf439f62
slight changes (memset IOSlot; if(.. < 0) -> CHECK_ERR)
...
This was SVN commit r1171.
2004-09-19 19:28:00 +00:00
b459e7e640
fixed debug_check_heap definition
...
This was SVN commit r1170.
2004-09-19 19:27:04 +00:00
a0f0215c15
fixed debug_check_heap definition
...
This was SVN commit r1167.
2004-09-19 18:45:04 +00:00
c3ed7df432
0 -> FALSE + add/removed debug code
...
This was SVN commit r1166.
2004-09-19 18:44:21 +00:00
1166c2a22c
workaround for stupid "void* -> unsigned cast" warning
...
This was SVN commit r1165.
2004-09-19 17:27:03 +00:00
cfe4562e6a
move low-level sound/music code from musicplayer to res/snd; also put oal.* in there
...
work in progress.
This was SVN commit r1162.
2004-09-19 15:57:20 +00:00
f6cfcd0555
slight change to interface (FileIO* vs simple void* output value) to simplify Zip interface (needed to change because now it has to store data of its own in IOs when doing partial reads)
...
This was SVN commit r1161.
2004-09-19 15:51:12 +00:00
5db49638cf
worked on partial zip reads. also changed to bring in line with file/vfs for slightly changed streaming interface
...
This was SVN commit r1160.
2004-09-19 15:49:46 +00:00
d3f452cd80
- add vfs_io_complete (wasn't in header) and vfs_size (convenience function)
...
- have vfs_load allocate the buffer, so mem_assign is no longer needed.
exposes buffer alignment requirement to VFS, but hey. simplifies things
a good bit.
- moved performance measuring crap to vfs_timed_io, a wrapper function
- various fixes for streaming code
This was SVN commit r1159.
2004-09-19 15:42:32 +00:00
b3a5bad377
fix 2 stupid h_mgr_shutdown bugs leading to failure to close some leaked handles at exit (see bug 26)
...
- was iterating over all HDATA slots up to last_in_use, but it is
inclusive
- when closing, wasn't resetting refcount, so h_free kept those with
refcount > 1 open
This was SVN commit r1158.
2004-09-19 13:46:38 +00:00
a872461a47
no longer 2 tex_write calls - just one that fills a TexInfo directly => no longer need mem_assign to wrap a pointer, only to dereference immediately thereafter
...
This was SVN commit r1157.
2004-09-19 13:43:53 +00:00
d46cf9cf1f
remove mem_assign and mem_assign_user (all call sites rewritten so that they're no longer needed)
...
still available in source, but shouldn't be necessary.
This was SVN commit r1156.
2004-09-19 13:41:49 +00:00
e4f160df27
fix headers: ogl_tex.h is user-level texture stuff, tex.h is only the texture loader.
...
this came up due to header change - some unnecessary stuff isn't
included by default anymore.
This was SVN commit r1155.
2004-09-19 11:57:51 +00:00
02ec0a7e62
add comment: tex_upload binds texture as a side effect
...
This was SVN commit r1154.
2004-09-19 11:40:42 +00:00
7c269d7735
better encapsulation of header (=> less dependencies)
...
This was SVN commit r1151.
2004-09-19 11:36:26 +00:00
5a8dd803a7
fix for gcc: philip reports it doesn't have upper_bound (although it's part of STL). replaced with equal_range
...
This was SVN commit r1150.
2004-09-19 11:23:12 +00:00
2d153bac2d
Increased paranoia, in case the i18n system crashes while translating the crash message
...
This was SVN commit r1139.
2004-09-06 11:39:27 +00:00
Gee
5c2838f7cf
Forgot to remove a temp log-output
...
This was SVN commit r1129.
2004-09-06 03:21:56 +00:00
Gee
6e07632775
Changed a function that the GUI uses to Unicode.
...
This was SVN commit r1127.
2004-09-06 02:24:34 +00:00
3b15f3f1e4
Text alignment alterations
...
This was SVN commit r1119.
2004-09-04 20:35:12 +00:00
78934b50c1
split up call to _CrtSetDbgFlag
...
This was SVN commit r1112.
2004-09-03 20:21:09 +00:00
f57e5ab737
Localised crash message (try ":loadLanguage("pseudogreek"), crash()" in the console in a non-Debug build)
...
This was SVN commit r1105.
2004-09-02 19:24:29 +00:00
72e730be3a
minor improvements / commenting
...
This was SVN commit r1098.
2004-09-02 02:47:11 +00:00
71b62dc484
more explanation of auto-init system
...
This was SVN commit r1097.
2004-09-02 02:46:50 +00:00
f05afa04a2
fix incorrect Tex_reload state handling (caused actual reloads to fail, i.e. display white texture)
...
This was SVN commit r1079.
2004-08-28 17:44:28 +00:00
b7ceaa0652
fix init segment names (must not be A or Z); see meeting log, 2004-08-28
...
This was SVN commit r1078.
2004-08-28 17:34:06 +00:00
c630437a9e
Removed -- it was no longer used (replaced by unifont.(cpp|h)) and causing conflicts with ps/Font.cpp
...
This was SVN commit r1077.
2004-08-28 13:35:48 +00:00
39a2589e2c
mk_handle -> cast_to_HANDLE (less confusion)
...
This was SVN commit r1076.
2004-08-28 11:14:23 +00:00
70d84beab2
Altered font handling -- now uses CFont in preference to unifont_* (yay, pointless OO) and allows font mapping ("default"=>"palatino12", etc) using .cfg files
...
This was SVN commit r1074.
2004-08-27 22:08:30 +00:00
1aae7dcead
more dox!
...
also found problem with partial reads from zip; not important since all
files are currently loaded in one go, but will attack tomorrow.
This was SVN commit r1065.
2004-08-27 00:33:20 +00:00
d607aad4db
had a look at mmap() due to gamedev post. cleaned it up; now bulletproof and armor-plated (I hope :)).
...
This was SVN commit r1064.
2004-08-27 00:29:46 +00:00
c01f1f00a8
Minorly better error handling
...
This was SVN commit r1053.
2004-08-24 21:07:20 +00:00
4ce93a9f12
remove misalign warning (no longer relevant)
...
This was SVN commit r1052.
2004-08-24 17:42:46 +00:00
df5ba1f690
simplification: only VFS allocates output buffer
...
This was SVN commit r1051.
2004-08-24 17:34:00 +00:00
13e5f5dc11
underwent review, armor-plating and documentation. "finished" (i hope) except for one last outstanding issue:
...
special-casing uncompressed files in archive. the impl is hackish ATM,
and requires further work on the file_io cache.
This was SVN commit r1050.
2004-08-24 17:29:54 +00:00
fb0b498de2
pthread_mutex update - now uses CRITICAL_SECTIONs. quite hacky; awaiting word from simon how he'd like it :)
...
This was SVN commit r1049.
2004-08-24 17:27:51 +00:00
59fd28fc93
GCC fixes: made zip_archive_open/close not inline
...
This was SVN commit r1033.
2004-08-21 12:54:25 +00:00
57fad52ef3
GCC non-fixes: marked broken code
...
This was SVN commit r1032.
2004-08-21 12:53:17 +00:00