/** * OpenAL cross platform audio library * Copyright (C) 2011 by authors. * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Library General Public * License as published by the Free Software Foundation; either * version 2 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 * Library General Public License for more details. * * You should have received a copy of the GNU Library General Public * License along with this library; if not, write to the * Free Software Foundation, Inc., 59 Temple Place - Suite 330, * Boston, MA 02111-1307, USA. * Or go to http://www.gnu.org/copyleft/lgpl.html */ #include "config.h" #if defined(HAVE_GUIDDEF_H) || defined(HAVE_INITGUID_H) #define INITGUID #include #ifdef HAVE_GUIDDEF_H #include #else #include #endif DEFINE_GUID(KSDATAFORMAT_SUBTYPE_PCM, 0x00000001, 0x0000, 0x0010, 0x80,0x00, 0x00,0xaa,0x00,0x38,0x9b,0x71); DEFINE_GUID(KSDATAFORMAT_SUBTYPE_IEEE_FLOAT, 0x00000003, 0x0000, 0x0010, 0x80,0x00, 0x00,0xaa,0x00,0x38,0x9b,0x71); DEFINE_GUID(IID_IDirectSoundNotify, 0xb0210783, 0x89cd, 0x11d0, 0xaf,0x08, 0x00,0xa0,0xc9,0x25,0xcd,0x16); DEFINE_GUID(CLSID_MMDeviceEnumerator, 0xbcde0395, 0xe52f, 0x467c, 0x8e,0x3d, 0xc4,0x57,0x92,0x91,0x69,0x2e); DEFINE_GUID(IID_IMMDeviceEnumerator, 0xa95664d2, 0x9614, 0x4f35, 0xa7,0x46, 0xde,0x8d,0xb6,0x36,0x17,0xe6); DEFINE_GUID(IID_IAudioClient, 0x1cb9ad4c, 0xdbfa, 0x4c32, 0xb1,0x78, 0xc2,0xf5,0x68,0xa7,0x03,0xb2); DEFINE_GUID(IID_IAudioRenderClient, 0xf294acfc, 0x3146, 0x4483, 0xa7,0xbf, 0xad,0xdc,0xa7,0xc2,0x60,0xe2); #ifdef HAVE_MMDEVAPI #include DEFINE_DEVPROPKEY(DEVPKEY_Device_FriendlyName, 0xa45c254e, 0xdf1c, 0x4efd, 0x80,0x20, 0x67,0xd1,0x46,0xa8,0x50,0xe0, 14); #endif #endif #include #include #include #include #ifdef HAVE_DLFCN_H #include #endif #ifdef HAVE_CPUID_H #include #endif #ifdef HAVE_FLOAT_H #include #endif #ifdef HAVE_IEEEFP_H #include #endif #include "alMain.h" ALuint CPUCapFlags = 0; void FillCPUCaps(ALuint capfilter) { ALuint caps = 0; #if defined(__i386__) || defined(__x86_64__) || defined(_M_IX86) || defined(_M_X64) /* FIXME: We really should get this for all available CPUs in case different * CPUs have different caps (is that possible on one machine?). */ #ifdef HAVE_CPUID_H union { unsigned int regs[4]; char str[sizeof(unsigned int[4])]; } cpuinf[3]; if(!__get_cpuid(0, &cpuinf[0].regs[0], &cpuinf[0].regs[1], &cpuinf[0].regs[2], &cpuinf[0].regs[3])) ERR("Failed to get CPUID\n"); else { unsigned int maxfunc = cpuinf[0].regs[0]; unsigned int maxextfunc = 0; if(__get_cpuid(0x80000000, &cpuinf[0].regs[0], &cpuinf[0].regs[1], &cpuinf[0].regs[2], &cpuinf[0].regs[3])) maxextfunc = cpuinf[0].regs[0]; TRACE("Detected max CPUID function: 0x%x (ext. 0x%x)\n", maxfunc, maxextfunc); TRACE("Vendor ID: \"%.4s%.4s%.4s\"\n", cpuinf[0].str+4, cpuinf[0].str+12, cpuinf[0].str+8); if(maxextfunc >= 0x80000004 && __get_cpuid(0x80000002, &cpuinf[0].regs[0], &cpuinf[0].regs[1], &cpuinf[0].regs[2], &cpuinf[0].regs[3]) && __get_cpuid(0x80000003, &cpuinf[1].regs[0], &cpuinf[1].regs[1], &cpuinf[1].regs[2], &cpuinf[1].regs[3]) && __get_cpuid(0x80000004, &cpuinf[2].regs[0], &cpuinf[2].regs[1], &cpuinf[2].regs[2], &cpuinf[2].regs[3])) TRACE("Name: \"%.16s%.16s%.16s\"\n", cpuinf[0].str, cpuinf[1].str, cpuinf[2].str); if(maxfunc >= 1 && __get_cpuid(1, &cpuinf[0].regs[0], &cpuinf[0].regs[1], &cpuinf[0].regs[2], &cpuinf[0].regs[3])) { #ifdef bit_SSE if((cpuinf[0].regs[3]&bit_SSE)) caps |= CPU_CAP_SSE; #endif } } #endif #endif #ifdef HAVE_NEON /* Assume Neon support if compiled with it */ caps |= CPU_CAP_NEON; #endif TRACE("Got caps:%s%s%s\n", ((caps&CPU_CAP_SSE)?((capfilter&CPU_CAP_SSE)?" SSE":" (SSE)"):""), ((caps&CPU_CAP_NEON)?((capfilter&CPU_CAP_NEON)?" Neon":" (Neon)"):""), ((!caps)?" -none-":"")); CPUCapFlags = caps & capfilter; } void *al_malloc(size_t alignment, size_t size) { #if defined(HAVE_ALIGNED_ALLOC) size = (size+(alignment-1))&~(alignment-1); return aligned_alloc(alignment, size); #elif defined(HAVE_POSIX_MEMALIGN) void *ret; if(posix_memalign(&ret, alignment, size) == 0) return ret; return NULL; #elif defined(HAVE__ALIGNED_MALLOC) return _aligned_malloc(size, alignment); #else char *ret = malloc(size+alignment); if(ret != NULL) { *(ret++) = 0x00; while(((ALintptrEXT)ret&(alignment-1)) != 0) *(ret++) = 0x55; } return ret; #endif } void *al_calloc(size_t alignment, size_t size) { void *ret = al_malloc(alignment, size); if(ret) memset(ret, 0, size); return ret; } void al_free(void *ptr) { #if defined(HAVE_ALIGNED_ALLOC) || defined(HAVE_POSIX_MEMALIGN) free(ptr); #elif defined(HAVE__ALIGNED_MALLOC) _aligned_free(ptr); #else if(ptr != NULL) { char *finder = ptr; do { --finder; } while(*finder == 0x55); free(finder); } #endif } #if (defined(HAVE___CONTROL87_2) || defined(HAVE__CONTROLFP)) && (defined(__x86_64__) || defined(_M_X64)) /* Win64 doesn't allow us to set the precision control. */ #undef _MCW_PC #define _MCW_PC 0 #endif void SetMixerFPUMode(FPUCtl *ctl) { #if defined(__GNUC__) && (defined(__i386__) || defined(__x86_64__)) unsigned short fpuState; __asm__ __volatile__("fnstcw %0" : "=m" (*&fpuState)); ctl->state = fpuState; fpuState &= ~0x300; /* clear precision to single */ fpuState |= 0xC00; /* set round-to-zero */ __asm__ __volatile__("fldcw %0" : : "m" (*&fpuState)); #ifdef HAVE_SSE if((CPUCapFlags&CPU_CAP_SSE)) { int sseState; __asm__ __volatile__("stmxcsr %0" : "=m" (*&sseState)); ctl->sse_state = sseState; sseState |= 0x0C00; /* set round-to-zero */ sseState |= 0x8000; /* set flush-to-zero */ __asm__ __volatile__("ldmxcsr %0" : : "m" (*&sseState)); } #endif #elif defined(HAVE___CONTROL87_2) int mode; __control87_2(0, 0, &ctl->state, &ctl->sse_state); __control87_2(_RC_CHOP|_PC_24, _MCW_RC|_MCW_PC, &mode, NULL); #ifdef HAVE_SSE if((CPUCapFlags&CPU_CAP_SSE)) __control87_2(_RC_CHOP|_DN_FLUSH, _MCW_RC|_MCW_DN, NULL, &mode); #endif #elif defined(HAVE__CONTROLFP) ctl->state = _controlfp(0, 0); (void)_controlfp(_RC_CHOP|_PC_24, _MCW_RC|_MCW_PC); #elif defined(HAVE_FESETROUND) ctl->state = fegetround(); #ifdef FE_TOWARDZERO fesetround(FE_TOWARDZERO); #endif #endif } void RestoreFPUMode(const FPUCtl *ctl) { #if defined(__GNUC__) && (defined(__i386__) || defined(__x86_64__)) unsigned short fpuState = ctl->state; __asm__ __volatile__("fldcw %0" : : "m" (*&fpuState)); #ifdef HAVE_SSE if((CPUCapFlags&CPU_CAP_SSE)) __asm__ __volatile__("ldmxcsr %0" : : "m" (*&ctl->sse_state)); #endif #elif defined(HAVE___CONTROL87_2) int mode; __control87_2(ctl->state, _MCW_RC|_MCW_PC, &mode, NULL); #ifdef HAVE_SSE if((CPUCapFlags&CPU_CAP_SSE)) __control87_2(ctl->sse_state, _MCW_RC|_MCW_DN, NULL, &mode); #endif #elif defined(HAVE__CONTROLFP) _controlfp(ctl->state, _MCW_RC|_MCW_PC); #elif defined(HAVE_FESETROUND) fesetround(ctl->state); #endif } #ifdef _WIN32 void pthread_once(pthread_once_t *once, void (*callback)(void)) { LONG ret; while((ret=InterlockedExchange(once, 1)) == 1) sched_yield(); if(ret == 0) callback(); InterlockedExchange(once, 2); } int pthread_key_create(pthread_key_t *key, void (*callback)(void*)) { *key = TlsAlloc(); if(callback) InsertUIntMapEntry(&TlsDestructor, *key, callback); return 0; } int pthread_key_delete(pthread_key_t key) { InsertUIntMapEntry(&TlsDestructor, key, NULL); TlsFree(key); return 0; } void *pthread_getspecific(pthread_key_t key) { return TlsGetValue(key); } int pthread_setspecific(pthread_key_t key, void *val) { TlsSetValue(key, val); return 0; } void *LoadLib(const char *name) { return LoadLibraryA(name); } void CloseLib(void *handle) { FreeLibrary((HANDLE)handle); } void *GetSymbol(void *handle, const char *name) { void *ret; ret = (void*)GetProcAddress((HANDLE)handle, name); if(ret == NULL) ERR("Failed to load %s\n", name); return ret; } WCHAR *strdupW(const WCHAR *str) { const WCHAR *n; WCHAR *ret; size_t len; n = str; while(*n) n++; len = n - str; ret = calloc(sizeof(WCHAR), len+1); if(ret != NULL) memcpy(ret, str, sizeof(WCHAR)*len); return ret; } #else #include #ifdef HAVE_PTHREAD_NP_H #include #endif #include void InitializeCriticalSection(CRITICAL_SECTION *cs) { pthread_mutexattr_t attrib; int ret; ret = pthread_mutexattr_init(&attrib); assert(ret == 0); ret = pthread_mutexattr_settype(&attrib, PTHREAD_MUTEX_RECURSIVE); #ifdef HAVE_PTHREAD_NP_H if(ret != 0) ret = pthread_mutexattr_setkind_np(&attrib, PTHREAD_MUTEX_RECURSIVE); #endif assert(ret == 0); ret = pthread_mutex_init(cs, &attrib); assert(ret == 0); pthread_mutexattr_destroy(&attrib); } void DeleteCriticalSection(CRITICAL_SECTION *cs) { int ret; ret = pthread_mutex_destroy(cs); assert(ret == 0); } void EnterCriticalSection(CRITICAL_SECTION *cs) { int ret; ret = pthread_mutex_lock(cs); assert(ret == 0); } void LeaveCriticalSection(CRITICAL_SECTION *cs) { int ret; ret = pthread_mutex_unlock(cs); assert(ret == 0); } /* NOTE: This wrapper isn't quite accurate as it returns an ALuint, as opposed * to the expected DWORD. Both are defined as unsigned 32-bit types, however. * Additionally, Win32 is supposed to measure the time since Windows started, * as opposed to the actual time. */ ALuint timeGetTime(void) { #if _POSIX_TIMERS > 0 struct timespec ts; int ret = -1; #if defined(_POSIX_MONOTONIC_CLOCK) && (_POSIX_MONOTONIC_CLOCK >= 0) #if _POSIX_MONOTONIC_CLOCK == 0 static int hasmono = 0; if(hasmono > 0 || (hasmono == 0 && (hasmono=sysconf(_SC_MONOTONIC_CLOCK)) > 0)) #endif ret = clock_gettime(CLOCK_MONOTONIC, &ts); #endif if(ret != 0) ret = clock_gettime(CLOCK_REALTIME, &ts); assert(ret == 0); return ts.tv_nsec/1000000 + ts.tv_sec*1000; #else struct timeval tv; int ret; ret = gettimeofday(&tv, NULL); assert(ret == 0); return tv.tv_usec/1000 + tv.tv_sec*1000; #endif } void Sleep(ALuint t) { struct timespec tv, rem; tv.tv_nsec = (t*1000000)%1000000000; tv.tv_sec = t/1000; while(nanosleep(&tv, &rem) == -1 && errno == EINTR) tv = rem; } #ifdef HAVE_DLFCN_H void *LoadLib(const char *name) { const char *err; void *handle; dlerror(); handle = dlopen(name, RTLD_NOW); if((err=dlerror()) != NULL) handle = NULL; return handle; } void CloseLib(void *handle) { dlclose(handle); } void *GetSymbol(void *handle, const char *name) { const char *err; void *sym; dlerror(); sym = dlsym(handle, name); if((err=dlerror()) != NULL) { WARN("Failed to load %s: %s\n", name, err); sym = NULL; } return sym; } #endif #endif void al_print(const char *type, const char *func, const char *fmt, ...) { char str[256]; int i; i = snprintf(str, sizeof(str), "AL lib: %s %s: ", type, func); if(i > 0 && (unsigned int)i < sizeof(str)) { va_list ap; va_start(ap, fmt); vsnprintf(str+i, sizeof(str)-i, fmt, ap); va_end(ap); } str[sizeof(str)-1] = 0; fprintf(LogFile, "%s", str); fflush(LogFile); } void SetRTPriority(void) { ALboolean failed = AL_FALSE; #ifdef _WIN32 if(RTPrioLevel > 0) failed = !SetThreadPriority(GetCurrentThread(), THREAD_PRIORITY_TIME_CRITICAL); #elif defined(HAVE_PTHREAD_SETSCHEDPARAM) && !defined(__OpenBSD__) if(RTPrioLevel > 0) { struct sched_param param; /* Use the minimum real-time priority possible for now (on Linux this * should be 1 for SCHED_RR) */ param.sched_priority = sched_get_priority_min(SCHED_RR); failed = !!pthread_setschedparam(pthread_self(), SCHED_RR, ¶m); } #else /* Real-time priority not available */ failed = (RTPrioLevel>0); #endif if(failed) ERR("Failed to set priority level for thread\n"); } static void Lock(volatile ALenum *l) { while(ExchangeInt(l, AL_TRUE) == AL_TRUE) sched_yield(); } static void Unlock(volatile ALenum *l) { ExchangeInt(l, AL_FALSE); } void RWLockInit(RWLock *lock) { lock->read_count = 0; lock->write_count = 0; lock->read_lock = AL_FALSE; lock->read_entry_lock = AL_FALSE; lock->write_lock = AL_FALSE; } void ReadLock(RWLock *lock) { Lock(&lock->read_entry_lock); Lock(&lock->read_lock); if(IncrementRef(&lock->read_count) == 1) Lock(&lock->write_lock); Unlock(&lock->read_lock); Unlock(&lock->read_entry_lock); } void ReadUnlock(RWLock *lock) { if(DecrementRef(&lock->read_count) == 0) Unlock(&lock->write_lock); } void WriteLock(RWLock *lock) { if(IncrementRef(&lock->write_count) == 1) Lock(&lock->read_lock); Lock(&lock->write_lock); } void WriteUnlock(RWLock *lock) { Unlock(&lock->write_lock); if(DecrementRef(&lock->write_count) == 0) Unlock(&lock->read_lock); } void InitUIntMap(UIntMap *map, ALsizei limit) { map->array = NULL; map->size = 0; map->maxsize = 0; map->limit = limit; RWLockInit(&map->lock); } void ResetUIntMap(UIntMap *map) { WriteLock(&map->lock); free(map->array); map->array = NULL; map->size = 0; map->maxsize = 0; WriteUnlock(&map->lock); } ALenum InsertUIntMapEntry(UIntMap *map, ALuint key, ALvoid *value) { ALsizei pos = 0; WriteLock(&map->lock); if(map->size > 0) { ALsizei low = 0; ALsizei high = map->size - 1; while(low < high) { ALsizei mid = low + (high-low)/2; if(map->array[mid].key < key) low = mid + 1; else high = mid; } if(map->array[low].key < key) low++; pos = low; } if(pos == map->size || map->array[pos].key != key) { if(map->size == map->limit) { WriteUnlock(&map->lock); return AL_OUT_OF_MEMORY; } if(map->size == map->maxsize) { ALvoid *temp = NULL; ALsizei newsize; newsize = (map->maxsize ? (map->maxsize<<1) : 4); if(newsize >= map->maxsize) temp = realloc(map->array, newsize*sizeof(map->array[0])); if(!temp) { WriteUnlock(&map->lock); return AL_OUT_OF_MEMORY; } map->array = temp; map->maxsize = newsize; } if(pos < map->size) memmove(&map->array[pos+1], &map->array[pos], (map->size-pos)*sizeof(map->array[0])); map->size++; } map->array[pos].key = key; map->array[pos].value = value; WriteUnlock(&map->lock); return AL_NO_ERROR; } ALvoid *RemoveUIntMapKey(UIntMap *map, ALuint key) { ALvoid *ptr = NULL; WriteLock(&map->lock); if(map->size > 0) { ALsizei low = 0; ALsizei high = map->size - 1; while(low < high) { ALsizei mid = low + (high-low)/2; if(map->array[mid].key < key) low = mid + 1; else high = mid; } if(map->array[low].key == key) { ptr = map->array[low].value; if(low < map->size-1) memmove(&map->array[low], &map->array[low+1], (map->size-1-low)*sizeof(map->array[0])); map->size--; } } WriteUnlock(&map->lock); return ptr; } ALvoid *LookupUIntMapKey(UIntMap *map, ALuint key) { ALvoid *ptr = NULL; ReadLock(&map->lock); if(map->size > 0) { ALsizei low = 0; ALsizei high = map->size - 1; while(low < high) { ALsizei mid = low + (high-low)/2; if(map->array[mid].key < key) low = mid + 1; else high = mid; } if(map->array[low].key == key) ptr = map->array[low].value; } ReadUnlock(&map->lock); return ptr; }