registers AMD64 provides, this routine still needs to be written as self-
modifying code for maximum performance. The additional registers do allow
for further optimization over the x86 version by allowing all four pixels
to be in flight at the same time. The end result is that AMD64 ASM is about
2.18 times faster than AMD64 C and about 1.06 times faster than x86 ASM.
(For further comparison, AMD64 C and x86 C are practically the same for
this function.) Should I port any more assembly to AMD64, mvlineasm4 is the
most likely candidate, but it's not used enough at this point to bother.
Also, this may or may not work with Linux at the moment, since it doesn't
have the eh_handler metadata. Win64 is easier, since I just need to
structure the function prologue and epilogue properly and use some
assembler directives/macros to automatically generate the metadata. And
that brings up another point: You need YASM to assemble the AMD64 code,
because NASM doesn't support the Win64 metadata directives.
- Added an SSE version of DoBlending. This is strictly C intrinsics.
VC++ still throws around unneccessary register moves. GCC seems to be
pretty close to optimal, requiring only about 2 cycles/color. They're
both faster than my hand-written MMX routine, so I don't need to feel
bad about not hand-optimizing this for x64 builds.
- Removed an extra instruction from DoBlending_MMX, transposed two
instructions, and unrolled it once, shaving off about 80 cycles from the
time required to blend 256 palette entries. Why? Because I tried writing
a C version of the routine using compiler intrinsics and was appalled by
all the extra movq's VC++ added to the code. GCC was better, but still
generated extra instructions. I only wanted a C version because I can't
use inline assembly with VC++'s x64 compiler, and x64 assembly is a bit
of a pain. (It's a pain because Linux and Windows have different calling
conventions, and you need to maintain extra metadata for functions.) So,
the assembly version stays and the C version stays out.
- Removed all the pixel doubling r_detail modes, since the one platform they
were intended to assist (486) actually sees very little benefit from them.
- Rewrote CheckMMX in C and renamed it to CheckCPU.
- Fixed: CPUID function 0x80000005 is specified to return detailed L1 cache
only for AMD processors, so we must not use it on other architectures, or
we end up overwriting the L1 cache line size with 0 or some other number
we don't actually understand.
SVN r1134 (trunk)
- AActor::PreExplode is gone now that the last item that was using it has been converted.
- Converted the Sigil and the remaining things in a_strifeitems.cpp to DECORATE.
SVN r1132 (trunk)
- Changed PoisonCloud and Lightning special treatment in P_DamageMobj to use damage
types instead to keep dependencies on specific actor types out of the main engine code.
- Added Korax DECORATE conversion by Gez and a few others by Karate Chris.
SVN r1130 (trunk)
pieces.
- Added DECORATE conversions for Hexen's Fighter weapons by Karate Chris.
- Added aWeaponGiver class to generalize the standing AssaultGun.
- converted a_Strifeweapons.cpp to DECORATE, except for the Sigil.
SVN r1129 (trunk)
- Fixed: AStalker::CheckMeleeRange did not perform all checks of AActor::CheckMeleeRange.
I replaced this virtual override with a new flag MF5_NOVERTICALMELEERANGE so that
this feature can also be used by other actors.
- Converted Strife's Stalker to DECORATE.
- Converted ArtiTeleport to DECORATE.
SVN r1126 (trunk)
as well inherit directly from AActor.
- Converted Strife's Coin, Oracle, Macil and StrifeHumanoid to DECORATE. Also moved
the burning hand states to StrifePlayer where they really belong.
SVN r1125 (trunk)
redundant ammo multiplication code from P_DropItem and ADehackedPickup::TryPickup.
- Restricted native action function definitions to zdoom.pk3.
SVN r1123 (trunk)
SBARINfO update:
- Added: disablegrin, disableouch, disablepain, and disablerampage flags to
drawmugshot.
- Fixed: LowerHealthCap did not work properly.
- Fixed: Various bugs I noticed in the fullscreenoffsets code.
SVN r1122 (trunk)
occurences in the code that depended on it were changed accordingly.
Invulnerability colormaps are now being set by the items exclusively.
- Changed many checks for the friendly Minotaur to a new flag MF5_SUMMONEDMONSTER
so that it can hopefully be generalized to be usable elsewhere later.
- Added Gez's submission for converting the Minotaur to DECORATE.
SVN r1120 (trunk)
GCC defaults to extended precision instead, unlike Visual C++.
- Removed the useless src/Linux directory from the repository. This was a
holdover from the long-gone 1.22 build system.
SVN r1118 (trunk)
- Changed coordinate storage for EntityBoss so that it works properly even
when the pod is not used to spawn it.
- Converted Strife's Spectres and Entity to DECORATE.
SVN r1117 (trunk)
- Added: fullscreenoffsets flag for status bars. This changes the coordinate
system to be relative to the top left corner of the screen. This is useful
for full screen status bars.
- Changed: drawinventorybar will use the width of artibox or invcurs (strife)
to determine the spacing. Let me know if this breaks any released mods.
- Fixed: If a status bar height of 0 was specified in SBarInfo the wrong bar
would be shown.
- Fixed: If a static inventory bar was used the user still had to press invuse
in order to get rid of the "overlay".
- Fixed: forcescaled would not work if the height of the bar was 0.
- Added: keyslot to drawswitchableimage.
- Fixed: The transition effects for the log and keys popups were switched.
SVN r1116 (trunk)
- Added a check to Zipdir that excludes files with a .orig extension. These
can be left behind by patch.exe and create problems.
- fixed: Unmorphing from chicken caused a crash when reading non-existent
meta-data strings.
SVN r1112 (trunk)
the spawn function that only spawns one mace per level.
- Moved Mace respawning code into AInventory so that it works properly
for replacement actors.
SVN r1107 (trunk)
DECORATE so that the exported code pointers can be used.
- Separated Heretic's and Hexen's invulnerability items for stability
reasons.
SVN r1105 (trunk)
Karate Chris.
- Added several type checks to the weapon slot code.
- Changed: Players no longer respawn in instant death sectors with
the 'Respawn where died' flag on.
SVN r1101 (trunk)
- Fixed compilation of snes_spc on GCC <4.3.
- Fixed compilation of zstring.h on 32-bit MinGW, where SIZE_MAX
and UINT_MAX are not considered identical by the preprocessor.
SVN r1096 (trunk)
CMake 2.4, since the distros don't seem to consider 2.6 stable yet.
As a bonus, GTK+ is no longer a required dependency; now it's optional.
- Made dehsupp ignore CR characters, so it doesn't spew warnings on Linux.
SVN r1092 (trunk)
