This was necessary because everything is already allocated with the Xmalloc() functions, but a future commit will make blocks allocated with those functions no longer compatible with the system implementation of free(), which Bfree() wraps.
git-svn-id: https://svn.eduke32.com/eduke32@7705 1a8010ca-5511-0410-912e-c29ae57300e0
# Conflicts:
# source/build/src/build.cpp
# source/build/src/mdsprite.cpp
# source/build/src/polymer.cpp
# source/build/src/polymost.cpp
# source/build/src/texcache.cpp
# source/build/src/voxmodel.cpp
This may also fix a minor performance regression introduced in Polymer sometime after we switched to building as C++, caused by confusion between STRUCT_TRACKERS_ENABLED and UNTRACKED_STRUCTS. UNTRACKED_STRUCTS is now named UNTRACT_STRUCTS__ to dissuade use outside of buildtypes.h.
git-svn-id: https://svn.eduke32.com/eduke32@7695 1a8010ca-5511-0410-912e-c29ae57300e0
# Conflicts:
# source/build/src/polymer.cpp
This improves, but doesn't fix, the issue with sound popping when the player angle changes.
git-svn-id: https://svn.eduke32.com/eduke32@7676 1a8010ca-5511-0410-912e-c29ae57300e0
This removes some of the error handling for gamevar reads and writes and trades it for simpler code and a small-but-benchmarkable performance increase.
git-svn-id: https://svn.eduke32.com/eduke32@7674 1a8010ca-5511-0410-912e-c29ae57300e0
https://kristerw.blogspot.com/2016/02/how-undefined-signed-overflow-enables.html
Doing this as cleanly as possible involved demoting several function parameters concerning object sizes and counts from size_t to int--I'm fine with this change as the functions in question are not actually capable of handling input with sizes larger than what can be stored in a signed 32-bit integer, making the use of size_t here misleading at best.
git-svn-id: https://svn.eduke32.com/eduke32@7673 1a8010ca-5511-0410-912e-c29ae57300e0
# Conflicts:
# source/build/src/polymost.cpp
# source/build/src/texcache.cpp
# source/build/src/tilepacker.cpp
"Add 'bsubweapon' player structure. Same as 'subweapon', except that it writes a bit for each weapon. For example, 'ife player[].bsubweapon GROW_WEAPON 1' has the same result as 'ifand player[].subweapon 2048'."
git-svn-id: https://svn.eduke32.com/eduke32@7668 1a8010ca-5511-0410-912e-c29ae57300e0
I don't remember if there was any particular reason they were added as gamevars in the first place.
git-svn-id: https://svn.eduke32.com/eduke32@7664 1a8010ca-5511-0410-912e-c29ae57300e0
This is almost as fast as ditching the bounds check here entirely.
git-svn-id: https://svn.eduke32.com/eduke32@7657 1a8010ca-5511-0410-912e-c29ae57300e0
Also removes a redundant "break" statement in builds not using CON_USE_COMPUTED_GOTO.
git-svn-id: https://svn.eduke32.com/eduke32@7647 1a8010ca-5511-0410-912e-c29ae57300e0
This also moves the loop control stuff into #ifndef CON_DIRECT_THREADING_DISPATCH blocks, since we don't actually need a loop anymore when using the computed goto stuff.
git-svn-id: https://svn.eduke32.com/eduke32@7646 1a8010ca-5511-0410-912e-c29ae57300e0
It looks like these were simply missed when vm.pSprite was initially added.
git-svn-id: https://svn.eduke32.com/eduke32@7639 1a8010ca-5511-0410-912e-c29ae57300e0
This is for errors that simply check a condition (surrounded with EDUKE32_PREDICT_FALSE) and bail out with CON_ERRPRINTF() and abort_after_error(). Not everything that fits this description has been transitioned to use the new macro yet.
git-svn-id: https://svn.eduke32.com/eduke32@7638 1a8010ca-5511-0410-912e-c29ae57300e0
