Mercurial > sdl-ios-xcode
view src/thread/dc/SDL_syscond.c @ 1542:a8bf1aa21020
Fixed bug #15
SDL_blit_A.mmx-speed.patch.txt --
Speed improvements and a bugfix for the current GCC inline mmx
asm code:
- Changed some ops and removed some resulting useless ones.
- Added some instruction parallelism (some gain)
The resulting speed on my Xeon improved upto 35% depending on
the function (measured in fps).
- Fixed a bug where BlitRGBtoRGBSurfaceAlphaMMX() was
setting the alpha component on the destination surfaces (to
opaque-alpha) even when the surface had none.
SDL_blit_A.mmx-msvc.patch.txt --
MSVC mmx intrinsics version of the same GCC asm code.
MSVC compiler tries to parallelize the code and to avoid
register stalls, but does not always do a very good job.
Per-surface blending MSVC functions run quite a bit faster
than their pure-asm counterparts (upto 55% faster for 16bit
ones), but the per-pixel blending runs somewhat slower than asm.
- BlitRGBtoRGBSurfaceAlphaMMX and BlitRGBtoRGBPixelAlphaMMX (and all
variants) can now also handle formats other than (A)RGB8888. Formats
like RGBA8888 and some quite exotic ones are allowed -- like
RAGB8888, or actually anything having channels aligned on 8bit
boundary and full 8bit alpha (for per-pixel alpha blending).
The performance cost of this change is virtually 0 for per-surface
alpha blending (no extra ops inside the loop) and a single non-MMX
op inside the loop for per-pixel blending. In testing, the per-pixel
alpha blending takes a ~2% performance hit, but it still runs much
faster than the current code in CVS. If necessary, a separate function
with this functionality can be made.
This code requires Processor Pack for VC6.
author | Sam Lantinga <slouken@libsdl.org> |
---|---|
date | Wed, 15 Mar 2006 15:39:29 +0000 |
parents | d910939febfa |
children | 782fd950bd46 c121d94672cb a1b03ba2fcd0 |
line wrap: on
line source
/* SDL - Simple DirectMedia Layer Copyright (C) 1997-2006 Sam Lantinga This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA Sam Lantinga slouken@libsdl.org */ #include "SDL_config.h" /* An implementation of condition variables using semaphores and mutexes */ /* This implementation borrows heavily from the BeOS condition variable implementation, written by Christopher Tate and Owen Smith. Thanks! */ #include "SDL_thread.h" struct SDL_cond { SDL_mutex *lock; int waiting; int signals; SDL_sem *wait_sem; SDL_sem *wait_done; }; /* Create a condition variable */ SDL_cond * SDL_CreateCond(void) { SDL_cond *cond; cond = (SDL_cond *) SDL_malloc(sizeof(SDL_cond)); if ( cond ) { cond->lock = SDL_CreateMutex(); cond->wait_sem = SDL_CreateSemaphore(0); cond->wait_done = SDL_CreateSemaphore(0); cond->waiting = cond->signals = 0; if ( ! cond->lock || ! cond->wait_sem || ! cond->wait_done ) { SDL_DestroyCond(cond); cond = NULL; } } else { SDL_OutOfMemory(); } return(cond); } /* Destroy a condition variable */ void SDL_DestroyCond(SDL_cond *cond) { if ( cond ) { if ( cond->wait_sem ) { SDL_DestroySemaphore(cond->wait_sem); } if ( cond->wait_done ) { SDL_DestroySemaphore(cond->wait_done); } if ( cond->lock ) { SDL_DestroyMutex(cond->lock); } SDL_free(cond); } } /* Restart one of the threads that are waiting on the condition variable */ int SDL_CondSignal(SDL_cond *cond) { if ( ! cond ) { SDL_SetError("Passed a NULL condition variable"); return -1; } /* If there are waiting threads not already signalled, then signal the condition and wait for the thread to respond. */ SDL_LockMutex(cond->lock); if ( cond->waiting > cond->signals ) { ++cond->signals; SDL_SemPost(cond->wait_sem); SDL_UnlockMutex(cond->lock); SDL_SemWait(cond->wait_done); } else { SDL_UnlockMutex(cond->lock); } return 0; } /* Restart all threads that are waiting on the condition variable */ int SDL_CondBroadcast(SDL_cond *cond) { if ( ! cond ) { SDL_SetError("Passed a NULL condition variable"); return -1; } /* If there are waiting threads not already signalled, then signal the condition and wait for the thread to respond. */ SDL_LockMutex(cond->lock); if ( cond->waiting > cond->signals ) { int i, num_waiting; num_waiting = (cond->waiting - cond->signals); cond->signals = cond->waiting; for ( i=0; i<num_waiting; ++i ) { SDL_SemPost(cond->wait_sem); } /* Now all released threads are blocked here, waiting for us. Collect them all (and win fabulous prizes!) :-) */ SDL_UnlockMutex(cond->lock); for ( i=0; i<num_waiting; ++i ) { SDL_SemWait(cond->wait_done); } } else { SDL_UnlockMutex(cond->lock); } return 0; } /* Wait on the condition variable for at most 'ms' milliseconds. The mutex must be locked before entering this function! The mutex is unlocked during the wait, and locked again after the wait. Typical use: Thread A: SDL_LockMutex(lock); while ( ! condition ) { SDL_CondWait(cond); } SDL_UnlockMutex(lock); Thread B: SDL_LockMutex(lock); ... condition = true; ... SDL_UnlockMutex(lock); */ int SDL_CondWaitTimeout(SDL_cond *cond, SDL_mutex *mutex, Uint32 ms) { int retval; if ( ! cond ) { SDL_SetError("Passed a NULL condition variable"); return -1; } /* Obtain the protection mutex, and increment the number of waiters. This allows the signal mechanism to only perform a signal if there are waiting threads. */ SDL_LockMutex(cond->lock); ++cond->waiting; SDL_UnlockMutex(cond->lock); /* Unlock the mutex, as is required by condition variable semantics */ SDL_UnlockMutex(mutex); /* Wait for a signal */ if ( ms == SDL_MUTEX_MAXWAIT ) { retval = SDL_SemWait(cond->wait_sem); } else { retval = SDL_SemWaitTimeout(cond->wait_sem, ms); } /* Let the signaler know we have completed the wait, otherwise the signaler can race ahead and get the condition semaphore if we are stopped between the mutex unlock and semaphore wait, giving a deadlock. See the following URL for details: http://www-classic.be.com/aboutbe/benewsletter/volume_III/Issue40.html */ SDL_LockMutex(cond->lock); if ( cond->signals > 0 ) { /* If we timed out, we need to eat a condition signal */ if ( retval > 0 ) { SDL_SemWait(cond->wait_sem); } /* We always notify the signal thread that we are done */ SDL_SemPost(cond->wait_done); /* Signal handshake complete */ --cond->signals; } --cond->waiting; SDL_UnlockMutex(cond->lock); /* Lock the mutex, as is required by condition variable semantics */ SDL_LockMutex(mutex); return retval; } /* Wait on the condition variable forever */ int SDL_CondWait(SDL_cond *cond, SDL_mutex *mutex) { return SDL_CondWaitTimeout(cond, mutex, SDL_MUTEX_MAXWAIT); }