diff options
Diffstat (limited to 'indra/llcommon')
| -rw-r--r-- | indra/llcommon/llassettype.cpp | 3 | ||||
| -rw-r--r-- | indra/llcommon/llassettype.h | 6 | ||||
| -rw-r--r-- | indra/llcommon/lldefs.h | 8 | ||||
| -rwxr-xr-x[-rw-r--r--] | indra/llcommon/llfasttimer.h | 135 | ||||
| -rw-r--r-- | indra/llcommon/llfasttimer_class.cpp | 172 | ||||
| -rw-r--r-- | indra/llcommon/llfasttimer_class.h | 9 | ||||
| -rw-r--r-- | indra/llcommon/llfoldertype.cpp | 3 | ||||
| -rw-r--r-- | indra/llcommon/llfoldertype.h | 6 | ||||
| -rw-r--r-- | indra/llcommon/llmd5.h | 2 | ||||
| -rw-r--r-- | indra/llcommon/llmemory.cpp | 19 | ||||
| -rw-r--r-- | indra/llcommon/llmemory.h | 83 | ||||
| -rw-r--r-- | indra/llcommon/llrefcount.cpp | 110 | ||||
| -rw-r--r-- | indra/llcommon/llrefcount.h | 20 | ||||
| -rw-r--r-- | indra/llcommon/llsdserialize.cpp | 188 | ||||
| -rw-r--r-- | indra/llcommon/llsdserialize.h | 4 | ||||
| -rw-r--r-- | indra/llcommon/llthread.cpp | 63 | ||||
| -rw-r--r-- | indra/llcommon/llthread.h | 22 | ||||
| -rw-r--r-- | indra/llcommon/stdenums.h | 3 |
18 files changed, 681 insertions, 175 deletions
diff --git a/indra/llcommon/llassettype.cpp b/indra/llcommon/llassettype.cpp index eb610f625a..145dddd543 100644 --- a/indra/llcommon/llassettype.cpp +++ b/indra/llcommon/llassettype.cpp @@ -93,8 +93,9 @@ LLAssetDictionary::LLAssetDictionary() addEntry(LLAssetType::AT_LINK, new AssetEntry("LINK", "link", "sym link", false, false, true)); addEntry(LLAssetType::AT_LINK_FOLDER, new AssetEntry("FOLDER_LINK", "link_f", "sym folder link", false, false, true)); + addEntry(LLAssetType::AT_MESH, new AssetEntry("MESH", "mesh", "mesh", false, false, false)); + addEntry(LLAssetType::AT_NONE, new AssetEntry("NONE", "-1", NULL, FALSE, FALSE, FALSE)); - addEntry(LLAssetType::AT_NONE, new AssetEntry("NONE", "-1", NULL, false, false, false)); }; // static diff --git a/indra/llcommon/llassettype.h b/indra/llcommon/llassettype.h index c5ff2364cc..74ccd00324 100644 --- a/indra/llcommon/llassettype.h +++ b/indra/llcommon/llassettype.h @@ -108,8 +108,10 @@ public: AT_LINK_FOLDER = 25, // Inventory folder link - - AT_COUNT = 26, + AT_MESH = 49, + // Mesh data in our proprietary SLM format + + AT_COUNT = 50, // +*********************************************************+ // | TO ADD AN ELEMENT TO THIS ENUM: | diff --git a/indra/llcommon/lldefs.h b/indra/llcommon/lldefs.h index 6b38de6500..5a4b8325f4 100644 --- a/indra/llcommon/lldefs.h +++ b/indra/llcommon/lldefs.h @@ -236,5 +236,13 @@ inline LLDATATYPE llclampb(const LLDATATYPE& a) return llmin(llmax(a, (LLDATATYPE)0), (LLDATATYPE)255); } +template <class LLDATATYPE> +inline void llswap(LLDATATYPE& lhs, LLDATATYPE& rhs) +{ + LLDATATYPE tmp = lhs; + lhs = rhs; + rhs = tmp; +} + #endif // LL_LLDEFS_H diff --git a/indra/llcommon/llfasttimer.h b/indra/llcommon/llfasttimer.h index 4ff93a553c..2b25f2fabb 100644..100755 --- a/indra/llcommon/llfasttimer.h +++ b/indra/llcommon/llfasttimer.h @@ -27,140 +27,9 @@ #ifndef LL_FASTTIMER_H #define LL_FASTTIMER_H +// Implementation of getCPUClockCount32() and getCPUClockCount64 are now in llfastertimer_class.cpp. + // pull in the actual class definition #include "llfasttimer_class.h" -// -// Important note: These implementations must be FAST! -// - -#if LL_WINDOWS -// -// Windows implementation of CPU clock -// - -// -// NOTE: put back in when we aren't using platform sdk anymore -// -// because MS has different signatures for these functions in winnt.h -// need to rename them to avoid conflicts -//#define _interlockedbittestandset _renamed_interlockedbittestandset -//#define _interlockedbittestandreset _renamed_interlockedbittestandreset -//#include <intrin.h> -//#undef _interlockedbittestandset -//#undef _interlockedbittestandreset - -//inline U32 LLFastTimer::getCPUClockCount32() -//{ -// U64 time_stamp = __rdtsc(); -// return (U32)(time_stamp >> 8); -//} -// -//// return full timer value, *not* shifted by 8 bits -//inline U64 LLFastTimer::getCPUClockCount64() -//{ -// return __rdtsc(); -//} - -// shift off lower 8 bits for lower resolution but longer term timing -// on 1Ghz machine, a 32-bit word will hold ~1000 seconds of timing -inline U32 LLFastTimer::getCPUClockCount32() -{ - U32 ret_val; - __asm - { - _emit 0x0f - _emit 0x31 - shr eax,8 - shl edx,24 - or eax, edx - mov dword ptr [ret_val], eax - } - return ret_val; -} - -// return full timer value, *not* shifted by 8 bits -inline U64 LLFastTimer::getCPUClockCount64() -{ - U64 ret_val; - __asm - { - _emit 0x0f - _emit 0x31 - mov eax,eax - mov edx,edx - mov dword ptr [ret_val+4], edx - mov dword ptr [ret_val], eax - } - return ret_val; -} -#endif - - -#if (LL_LINUX || LL_SOLARIS) && !(defined(__i386__) || defined(__amd64__)) -// -// Linux and Solaris implementation of CPU clock - non-x86. -// This is accurate but SLOW! Only use out of desperation. -// -// Try to use the MONOTONIC clock if available, this is a constant time counter -// with nanosecond resolution (but not necessarily accuracy) and attempts are -// made to synchronize this value between cores at kernel start. It should not -// be affected by CPU frequency. If not available use the REALTIME clock, but -// this may be affected by NTP adjustments or other user activity affecting -// the system time. -inline U64 LLFastTimer::getCPUClockCount64() -{ - struct timespec tp; - -#ifdef CLOCK_MONOTONIC // MONOTONIC supported at build-time? - if (-1 == clock_gettime(CLOCK_MONOTONIC,&tp)) // if MONOTONIC isn't supported at runtime then ouch, try REALTIME -#endif - clock_gettime(CLOCK_REALTIME,&tp); - - return (tp.tv_sec*LLFastTimer::sClockResolution)+tp.tv_nsec; -} - -inline U32 LLFastTimer::getCPUClockCount32() -{ - return (U32)(LLFastTimer::getCPUClockCount64() >> 8); -} -#endif // (LL_LINUX || LL_SOLARIS) && !(defined(__i386__) || defined(__amd64__)) - - -#if (LL_LINUX || LL_SOLARIS || LL_DARWIN) && (defined(__i386__) || defined(__amd64__)) -// -// Mac+Linux+Solaris FAST x86 implementation of CPU clock -inline U32 LLFastTimer::getCPUClockCount32() -{ - U64 x; - __asm__ volatile (".byte 0x0f, 0x31": "=A"(x)); - return (U32)(x >> 8); -} - -inline U64 LLFastTimer::getCPUClockCount64() -{ - U64 x; - __asm__ volatile (".byte 0x0f, 0x31": "=A"(x)); - return x; -} -#endif - - -#if ( LL_DARWIN && !(defined(__i386__) || defined(__amd64__))) -// -// Mac PPC (deprecated) implementation of CPU clock -// -// Just use gettimeofday implementation for now - -inline U32 LLFastTimer::getCPUClockCount32() -{ - return (U32)(get_clock_count()>>8); -} - -inline U64 LLFastTimer::getCPUClockCount64() -{ - return get_clock_count(); -} -#endif - #endif // LL_LLFASTTIMER_H diff --git a/indra/llcommon/llfasttimer_class.cpp b/indra/llcommon/llfasttimer_class.cpp index bce87ada96..bd594b06cf 100644 --- a/indra/llcommon/llfasttimer_class.cpp +++ b/indra/llcommon/llfasttimer_class.cpp @@ -35,10 +35,13 @@ #include <boost/bind.hpp> + #if LL_WINDOWS +#include "lltimer.h" #elif LL_LINUX || LL_SOLARIS #include <sys/time.h> #include <sched.h> +#include "lltimer.h" #elif LL_DARWIN #include <sys/time.h> #include "lltimer.h" // get_clock_count() @@ -61,6 +64,8 @@ BOOL LLFastTimer::sMetricLog = FALSE; LLMutex* LLFastTimer::sLogLock = NULL; std::queue<LLSD> LLFastTimer::sLogQueue; +#define USE_RDTSC 0 + #if LL_LINUX || LL_SOLARIS U64 LLFastTimer::sClockResolution = 1000000000; // Nanosecond resolution #else @@ -234,10 +239,23 @@ U64 LLFastTimer::countsPerSecond() // counts per second for the *32-bit* timer #else // windows or x86-mac or x86-linux or x86-solaris U64 LLFastTimer::countsPerSecond() // counts per second for the *32-bit* timer { +#if USE_RDTSC || !LL_WINDOWS //getCPUFrequency returns MHz and sCPUClockFrequency wants to be in Hz static U64 sCPUClockFrequency = U64(LLProcessorInfo().getCPUFrequency()*1000000.0); // we drop the low-order byte in our timers, so report a lower frequency +#else + // If we're not using RDTSC, each fasttimer tick is just a performance counter tick. + // Not redefining the clock frequency itself (in llprocessor.cpp/calculate_cpu_frequency()) + // since that would change displayed MHz stats for CPUs + static bool firstcall = true; + static U64 sCPUClockFrequency; + if (firstcall) + { + QueryPerformanceFrequency((LARGE_INTEGER*)&sCPUClockFrequency); + firstcall = false; + } +#endif return sCPUClockFrequency >> 8; } #endif @@ -482,6 +500,19 @@ void LLFastTimer::NamedTimer::resetFrame() { if (sLog) { //output current frame counts to performance log + + static S32 call_count = 0; + if (call_count % 100 == 0) + { + llinfos << "countsPerSecond (32 bit): " << countsPerSecond() << llendl; + llinfos << "get_clock_count (64 bit): " << get_clock_count() << llendl; + llinfos << "LLProcessorInfo().getCPUFrequency() " << LLProcessorInfo().getCPUFrequency() << llendl; + llinfos << "getCPUClockCount32() " << getCPUClockCount32() << llendl; + llinfos << "getCPUClockCount64() " << getCPUClockCount64() << llendl; + llinfos << "elapsed sec " << ((F64)getCPUClockCount64())/((F64)LLProcessorInfo().getCPUFrequency()*1000000.0) << llendl; + } + call_count++; + F64 iclock_freq = 1000.0 / countsPerSecond(); // good place to calculate clock frequency F64 total_time = 0; @@ -763,3 +794,144 @@ LLFastTimer::LLFastTimer(LLFastTimer::FrameState* state) ////////////////////////////////////////////////////////////////////////////// +// +// Important note: These implementations must be FAST! +// + + +#if LL_WINDOWS +// +// Windows implementation of CPU clock +// + +// +// NOTE: put back in when we aren't using platform sdk anymore +// +// because MS has different signatures for these functions in winnt.h +// need to rename them to avoid conflicts +//#define _interlockedbittestandset _renamed_interlockedbittestandset +//#define _interlockedbittestandreset _renamed_interlockedbittestandreset +//#include <intrin.h> +//#undef _interlockedbittestandset +//#undef _interlockedbittestandreset + +//inline U32 LLFastTimer::getCPUClockCount32() +//{ +// U64 time_stamp = __rdtsc(); +// return (U32)(time_stamp >> 8); +//} +// +//// return full timer value, *not* shifted by 8 bits +//inline U64 LLFastTimer::getCPUClockCount64() +//{ +// return __rdtsc(); +//} + +// shift off lower 8 bits for lower resolution but longer term timing +// on 1Ghz machine, a 32-bit word will hold ~1000 seconds of timing +#if USE_RDTSC +U32 LLFastTimer::getCPUClockCount32() +{ + U32 ret_val; + __asm + { + _emit 0x0f + _emit 0x31 + shr eax,8 + shl edx,24 + or eax, edx + mov dword ptr [ret_val], eax + } + return ret_val; +} + +// return full timer value, *not* shifted by 8 bits +U64 LLFastTimer::getCPUClockCount64() +{ + U64 ret_val; + __asm + { + _emit 0x0f + _emit 0x31 + mov eax,eax + mov edx,edx + mov dword ptr [ret_val+4], edx + mov dword ptr [ret_val], eax + } + return ret_val; +} + +std::string LLFastTimer::sClockType = "rdtsc"; + +#else +//LL_COMMON_API U64 get_clock_count(); // in lltimer.cpp +// These use QueryPerformanceCounter, which is arguably fine and also works on amd architectures. +U32 LLFastTimer::getCPUClockCount32() +{ + return (U32)(get_clock_count()>>8); +} + +U64 LLFastTimer::getCPUClockCount64() +{ + return get_clock_count(); +} + +std::string LLFastTimer::sClockType = "QueryPerformanceCounter"; +#endif + +#endif + + +#if (LL_LINUX || LL_SOLARIS) && !(defined(__i386__) || defined(__amd64__)) +// +// Linux and Solaris implementation of CPU clock - non-x86. +// This is accurate but SLOW! Only use out of desperation. +// +// Try to use the MONOTONIC clock if available, this is a constant time counter +// with nanosecond resolution (but not necessarily accuracy) and attempts are +// made to synchronize this value between cores at kernel start. It should not +// be affected by CPU frequency. If not available use the REALTIME clock, but +// this may be affected by NTP adjustments or other user activity affecting +// the system time. +U64 LLFastTimer::getCPUClockCount64() +{ + struct timespec tp; + +#ifdef CLOCK_MONOTONIC // MONOTONIC supported at build-time? + if (-1 == clock_gettime(CLOCK_MONOTONIC,&tp)) // if MONOTONIC isn't supported at runtime then ouch, try REALTIME +#endif + clock_gettime(CLOCK_REALTIME,&tp); + + return (tp.tv_sec*LLFastTimer::sClockResolution)+tp.tv_nsec; +} + +U32 LLFastTimer::getCPUClockCount32() +{ + return (U32)(LLFastTimer::getCPUClockCount64() >> 8); +} + +std::string LLFastTimer::sClockType = "clock_gettime"; + +#endif // (LL_LINUX || LL_SOLARIS) && !(defined(__i386__) || defined(__amd64__)) + + +#if (LL_LINUX || LL_SOLARIS || LL_DARWIN) && (defined(__i386__) || defined(__amd64__)) +// +// Mac+Linux+Solaris FAST x86 implementation of CPU clock +U32 LLFastTimer::getCPUClockCount32() +{ + U64 x; + __asm__ volatile (".byte 0x0f, 0x31": "=A"(x)); + return (U32)(x >> 8); +} + +U64 LLFastTimer::getCPUClockCount64() +{ + U64 x; + __asm__ volatile (".byte 0x0f, 0x31": "=A"(x)); + return x; +} + +std::string LLFastTimer::sClockType = "rdtsc"; +#endif + diff --git a/indra/llcommon/llfasttimer_class.h b/indra/llcommon/llfasttimer_class.h index eb9789682b..827747f0c6 100644 --- a/indra/llcommon/llfasttimer_class.h +++ b/indra/llcommon/llfasttimer_class.h @@ -31,12 +31,15 @@ #define FAST_TIMER_ON 1 #define TIME_FAST_TIMERS 0 +#define DEBUG_FAST_TIMER_THREADS 1 class LLMutex; #include <queue> #include "llsd.h" +LL_COMMON_API void assert_main_thread(); + class LL_COMMON_API LLFastTimer { public: @@ -176,6 +179,11 @@ public: U64 timer_end = getCPUClockCount64(); sTimerCycles += timer_end - timer_start; #endif +#if DEBUG_FAST_TIMER_THREADS +#if !LL_RELEASE + assert_main_thread(); +#endif +#endif } LL_FORCE_INLINE ~LLFastTimer() @@ -245,6 +253,7 @@ public: U32 mChildTime; }; static CurTimerData sCurTimerData; + static std::string sClockType; private: static U32 getCPUClockCount32(); diff --git a/indra/llcommon/llfoldertype.cpp b/indra/llcommon/llfoldertype.cpp index ebc79af412..c2cfb7286e 100644 --- a/indra/llcommon/llfoldertype.cpp +++ b/indra/llcommon/llfoldertype.cpp @@ -89,6 +89,9 @@ LLFolderDictionary::LLFolderDictionary() addEntry(LLFolderType::FT_CURRENT_OUTFIT, new FolderEntry("current", TRUE)); addEntry(LLFolderType::FT_OUTFIT, new FolderEntry("outfit", FALSE)); addEntry(LLFolderType::FT_MY_OUTFITS, new FolderEntry("my_otfts", TRUE)); + + addEntry(LLFolderType::FT_MESH, new FolderEntry("mesh", TRUE)); + addEntry(LLFolderType::FT_INBOX, new FolderEntry("inbox", TRUE)); addEntry(LLFolderType::FT_NONE, new FolderEntry("-1", FALSE)); diff --git a/indra/llcommon/llfoldertype.h b/indra/llcommon/llfoldertype.h index 936fbed17d..cb32cb075b 100644 --- a/indra/llcommon/llfoldertype.h +++ b/indra/llcommon/llfoldertype.h @@ -80,9 +80,11 @@ public: FT_OUTFIT = 47, FT_MY_OUTFITS = 48, - FT_INBOX = 49, + FT_MESH = 49, - FT_COUNT = 50, + FT_INBOX = 50, + + FT_COUNT = 51, FT_NONE = -1 }; diff --git a/indra/llcommon/llmd5.h b/indra/llcommon/llmd5.h index c8acbbe591..1526e6ac3c 100644 --- a/indra/llcommon/llmd5.h +++ b/indra/llcommon/llmd5.h @@ -103,7 +103,7 @@ public: void raw_digest(unsigned char *array) const; // provide 16-byte array for binary data void hex_digest(char *string) const; // provide 33-byte array for ascii-hex string - friend std::ostream& operator<< (std::ostream&, LLMD5 context); + friend LL_COMMON_API std::ostream& operator<< (std::ostream&, LLMD5 context); private: diff --git a/indra/llcommon/llmemory.cpp b/indra/llcommon/llmemory.cpp index a502d1a7eb..06ad7736d3 100644 --- a/indra/llcommon/llmemory.cpp +++ b/indra/llcommon/llmemory.cpp @@ -67,25 +67,6 @@ void LLMemory::freeReserve() reserveMem = NULL; } -void* ll_allocate (size_t size) -{ - if (size == 0) - { - llwarns << "Null allocation" << llendl; - } - void *p = malloc(size); - if (p == NULL) - { - LLMemory::freeReserve(); - llerrs << "Out of memory Error" << llendl; - } - return p; -} - -void ll_release (void *p) -{ - free(p); -} //---------------------------------------------------------------------------- diff --git a/indra/llcommon/llmemory.h b/indra/llcommon/llmemory.h index 9bf4248bb7..8d114f744b 100644 --- a/indra/llcommon/llmemory.h +++ b/indra/llcommon/llmemory.h @@ -1,39 +1,102 @@ -/** +/** * @file llmemory.h * @brief Memory allocation/deallocation header-stuff goes here. * * $LicenseInfo:firstyear=2002&license=viewerlgpl$ * Second Life Viewer Source Code * Copyright (C) 2010, Linden Research, Inc. - * + * * 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; * version 2.1 of the License only. - * + * * 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 Street, Fifth Floor, Boston, MA 02110-1301 USA - * + * * Linden Research, Inc., 945 Battery Street, San Francisco, CA 94111 USA * $/LicenseInfo$ */ #ifndef LLMEMORY_H #define LLMEMORY_H +#include <stdlib.h> + +// A not necessarily efficient, but general, aligned malloc http://stackoverflow.com/questions/196329/osx-lacks-memalign +#if 0 //DON'T use ll_aligned_foo now that we use tcmalloc everywhere (tcmalloc aligns automatically at appropriate intervals) +inline void* ll_aligned_malloc( size_t size, int align ) +{ + void* mem = malloc( size + (align - 1) + sizeof(void*) ); + char* aligned = ((char*)mem) + sizeof(void*); + aligned += align - ((uintptr_t)aligned & (align - 1)); + + ((void**)aligned)[-1] = mem; + return aligned; +} +inline void ll_aligned_free( void* ptr ) +{ + free( ((void**)ptr)[-1] ); +} -extern S32 gTotalDAlloc; -extern S32 gTotalDAUse; -extern S32 gDACount; +inline void* ll_aligned_malloc_16(size_t size) // returned hunk MUST be freed with ll_aligned_free_16(). +{ +#if defined(LL_WINDOWS) + return _mm_malloc(size, 16); +#elif defined(LL_DARWIN) + return malloc(size); // default osx malloc is 16 byte aligned. +#else + void *rtn; + if (LL_LIKELY(0 == posix_memalign(&rtn, 16, size))) + return rtn; + else // bad alignment requested, or out of memory + return NULL; +#endif +} -extern void* ll_allocate (size_t size); -extern void ll_release (void *p); +inline void ll_aligned_free_16(void *p) +{ +#if defined(LL_WINDOWS) + _mm_free(p); +#elif defined(LL_DARWIN) + return free(p); +#else + free(p); // posix_memalign() is compatible with heap deallocator +#endif +} + +inline void* ll_aligned_malloc_32(size_t size) // returned hunk MUST be freed with ll_aligned_free_32(). +{ +#if defined(LL_WINDOWS) + return _mm_malloc(size, 32); +#elif defined(LL_DARWIN) + return ll_aligned_malloc( size, 32 ); +#else + void *rtn; + if (LL_LIKELY(0 == posix_memalign(&rtn, 32, size))) + return rtn; + else // bad alignment requested, or out of memory + return NULL; +#endif +} + +inline void ll_aligned_free_32(void *p) +{ +#if defined(LL_WINDOWS) + _mm_free(p); +#elif defined(LL_DARWIN) + ll_aligned_free( p ); +#else + free(p); // posix_memalign() is compatible with heap deallocator +#endif +} +#endif class LL_COMMON_API LLMemory { diff --git a/indra/llcommon/llrefcount.cpp b/indra/llcommon/llrefcount.cpp index 55d0c85cbd..e1876599fc 100644 --- a/indra/llcommon/llrefcount.cpp +++ b/indra/llcommon/llrefcount.cpp @@ -29,9 +29,25 @@ #include "llerror.h" +#if LL_REF_COUNT_DEBUG +#include "llthread.h" +#include "llapr.h" +#endif + LLRefCount::LLRefCount(const LLRefCount& other) : mRef(0) { +#if LL_REF_COUNT_DEBUG + if(gAPRPoolp) + { + mMutexp = new LLMutex(gAPRPoolp) ; + } + else + { + mMutexp = NULL ; + } + mCrashAtUnlock = FALSE ; +#endif } LLRefCount& LLRefCount::operator=(const LLRefCount&) @@ -43,6 +59,17 @@ LLRefCount& LLRefCount::operator=(const LLRefCount&) LLRefCount::LLRefCount() : mRef(0) { +#if LL_REF_COUNT_DEBUG + if(gAPRPoolp) + { + mMutexp = new LLMutex(gAPRPoolp) ; + } + else + { + mMutexp = NULL ; + } + mCrashAtUnlock = FALSE ; +#endif } LLRefCount::~LLRefCount() @@ -51,4 +78,87 @@ LLRefCount::~LLRefCount() { llerrs << "deleting non-zero reference" << llendl; } + +#if LL_REF_COUNT_DEBUG + if(gAPRPoolp) + { + delete mMutexp ; + } +#endif } + +#if LL_REF_COUNT_DEBUG +void LLRefCount::ref() const +{ + if(mMutexp) + { + if(mMutexp->isLocked()) + { + mCrashAtUnlock = TRUE ; + llerrs << "the mutex is locked by the thread: " << mLockedThreadID + << " Current thread: " << LLThread::currentID() << llendl ; + } + + mMutexp->lock() ; + mLockedThreadID = LLThread::currentID() ; + + mRef++; + + if(mCrashAtUnlock) + { + while(1); //crash here. + } + mMutexp->unlock() ; + } + else + { + mRef++; + } +} + +S32 LLRefCount::unref() const +{ + if(mMutexp) + { + if(mMutexp->isLocked()) + { + mCrashAtUnlock = TRUE ; + llerrs << "the mutex is locked by the thread: " << mLockedThreadID + << " Current thread: " << LLThread::currentID() << llendl ; + } + + mMutexp->lock() ; + mLockedThreadID = LLThread::currentID() ; + + llassert(mRef >= 1); + if (0 == --mRef) + { + if(mCrashAtUnlock) + { + while(1); //crash here. + } + mMutexp->unlock() ; + + delete this; + return 0; + } + + if(mCrashAtUnlock) + { + while(1); //crash here. + } + mMutexp->unlock() ; + return mRef; + } + else + { + llassert(mRef >= 1); + if (0 == --mRef) + { + delete this; + return 0; + } + return mRef; + } +} +#endif diff --git a/indra/llcommon/llrefcount.h b/indra/llcommon/llrefcount.h index 19f008b15c..8eb5d53f3f 100644 --- a/indra/llcommon/llrefcount.h +++ b/indra/llcommon/llrefcount.h @@ -28,6 +28,11 @@ #include <boost/noncopyable.hpp> +#define LL_REF_COUNT_DEBUG 0 +#if LL_REF_COUNT_DEBUG +class LLMutex ; +#endif + //---------------------------------------------------------------------------- // RefCount objects should generally only be accessed by way of LLPointer<>'s // see llthread.h for LLThreadSafeRefCount @@ -43,12 +48,16 @@ protected: public: LLRefCount(); - void ref() const +#if LL_REF_COUNT_DEBUG + void ref() const ; + S32 unref() const ; +#else + inline void ref() const { mRef++; } - S32 unref() const + inline S32 unref() const { llassert(mRef >= 1); if (0 == --mRef) @@ -58,6 +67,7 @@ public: } return mRef; } +#endif //NOTE: when passing around a const LLRefCount object, this can return different results // at different types, since mRef is mutable @@ -68,6 +78,12 @@ public: private: mutable S32 mRef; + +#if LL_REF_COUNT_DEBUG + LLMutex* mMutexp ; + mutable U32 mLockedThreadID ; + mutable BOOL mCrashAtUnlock ; +#endif }; #endif diff --git a/indra/llcommon/llsdserialize.cpp b/indra/llcommon/llsdserialize.cpp index 10f460e8a6..f3cbfab77a 100644 --- a/indra/llcommon/llsdserialize.cpp +++ b/indra/llcommon/llsdserialize.cpp @@ -34,6 +34,12 @@ #include <iostream> #include "apr_base64.h" +#ifdef LL_STANDALONE +# include <zlib.h> +#else +# include "zlib/zlib.h" // for davep's dirty little zip functions +#endif + #if !LL_WINDOWS #include <netinet/in.h> // htonl & ntohl #endif @@ -1983,3 +1989,185 @@ std::ostream& operator<<(std::ostream& s, const LLSD& llsd) return s; } + +//dirty little zippers -- yell at davep if these are horrid + +//return a string containing gzipped bytes of binary serialized LLSD +// VERY inefficient -- creates several copies of LLSD block in memory +std::string zip_llsd(LLSD& data) +{ + std::stringstream llsd_strm; + + LLSDSerialize::toBinary(data, llsd_strm); + + const U32 CHUNK = 65536; + + z_stream strm; + strm.zalloc = Z_NULL; + strm.zfree = Z_NULL; + strm.opaque = Z_NULL; + + S32 ret = deflateInit(&strm, Z_BEST_COMPRESSION); + if (ret != Z_OK) + { + llwarns << "Failed to compress LLSD block." << llendl; + return std::string(); + } + + std::string source = llsd_strm.str(); + + U8 out[CHUNK]; + + strm.avail_in = source.size(); + strm.next_in = (U8*) source.data(); + U8* output = NULL; + + U32 cur_size = 0; + + U32 have = 0; + + do + { + strm.avail_out = CHUNK; + strm.next_out = out; + + ret = deflate(&strm, Z_FINISH); + if (ret == Z_OK || ret == Z_STREAM_END) + { //copy result into output + if (strm.avail_out >= CHUNK) + { + llerrs << "WTF?" << llendl; + } + + have = CHUNK-strm.avail_out; + output = (U8*) realloc(output, cur_size+have); + memcpy(output+cur_size, out, have); + cur_size += have; + } + else + { + free(output); + llwarns << "Failed to compress LLSD block." << llendl; + return std::string(); + } + } + while (ret == Z_OK); + + std::string::size_type size = cur_size; + + std::string result((char*) output, size); + deflateEnd(&strm); + free(output); + +#if 0 //verify results work with unzip_llsd + std::istringstream test(result); + LLSD test_sd; + if (!unzip_llsd(test_sd, test, result.size())) + { + llerrs << "Invalid compression result!" << llendl; + } +#endif + + return result; +} + +//decompress a block of LLSD from provided istream +// not very efficient -- creats a copy of decompressed LLSD block in memory +// and deserializes from that copy using LLSDSerialize +bool unzip_llsd(LLSD& data, std::istream& is, S32 size) +{ + U8* result = NULL; + U32 cur_size = 0; + z_stream strm; + + const U32 CHUNK = 65536; + + U8 *in = new U8[size]; + is.read((char*) in, size); + + U8 out[CHUNK]; + + strm.zalloc = Z_NULL; + strm.zfree = Z_NULL; + strm.opaque = Z_NULL; + strm.avail_in = size; + strm.next_in = in; + + S32 ret = inflateInit(&strm); + + if (ret != Z_OK) + { + llerrs << "WTF?" << llendl; + } + + do + { + strm.avail_out = CHUNK; + strm.next_out = out; + ret = inflate(&strm, Z_NO_FLUSH); + if (ret == Z_STREAM_ERROR) + { + inflateEnd(&strm); + free(result); + delete [] in; + return false; + } + + switch (ret) + { + case Z_NEED_DICT: + ret = Z_DATA_ERROR; + case Z_DATA_ERROR: + case Z_MEM_ERROR: + inflateEnd(&strm); + free(result); + delete [] in; + return false; + break; + } + + U32 have = CHUNK-strm.avail_out; + + result = (U8*) realloc(result, cur_size + have); + memcpy(result+cur_size, out, have); + cur_size += have; + + } while (ret == Z_OK); + + inflateEnd(&strm); + delete [] in; + + if (ret != Z_STREAM_END) + { + free(result); + return false; + } + + //result now points to the decompressed LLSD block + { + std::string res_str((char*) result, cur_size); + + std::string deprecated_header("<? LLSD/Binary ?>"); + + if (res_str.substr(0, deprecated_header.size()) == deprecated_header) + { + res_str = res_str.substr(deprecated_header.size()+1, cur_size); + } + cur_size = res_str.size(); + + std::istringstream istr(res_str); + + if (!LLSDSerialize::fromBinary(data, istr, cur_size)) + { + llwarns << "Failed to unzip LLSD block" << llendl; + free(result); + return false; + } + } + + free(result); + return true; +} + + + diff --git a/indra/llcommon/llsdserialize.h b/indra/llcommon/llsdserialize.h index 1f096b5254..99a3ea3cd4 100644 --- a/indra/llcommon/llsdserialize.h +++ b/indra/llcommon/llsdserialize.h @@ -790,4 +790,8 @@ public: } }; +//dirty little zip functions -- yell at davep +LL_COMMON_API std::string zip_llsd(LLSD& data); +LL_COMMON_API bool unzip_llsd(LLSD& data, std::istream& is, S32 size); + #endif // LL_LLSDSERIALIZE_H diff --git a/indra/llcommon/llthread.cpp b/indra/llcommon/llthread.cpp index 49d05ef411..b4617c5453 100644 --- a/indra/llcommon/llthread.cpp +++ b/indra/llcommon/llthread.cpp @@ -56,6 +56,21 @@ // //---------------------------------------------------------------------------- +#if !LL_DARWIN +U32 ll_thread_local sThreadID = 0; +#endif + +U32 LLThread::sIDIter = 0; + +LL_COMMON_API void assert_main_thread() +{ + static U32 s_thread_id = LLThread::currentID(); + if (LLThread::currentID() != s_thread_id) + { + llerrs << "Illegal execution outside main thread." << llendl; + } +} + // // Handed to the APR thread creation function // @@ -63,10 +78,14 @@ void *APR_THREAD_FUNC LLThread::staticRun(apr_thread_t *apr_threadp, void *datap { LLThread *threadp = (LLThread *)datap; +#if !LL_DARWIN + sThreadID = threadp->mID; +#endif + // Run the user supplied function threadp->run(); - llinfos << "LLThread::staticRun() Exiting: " << threadp->mName << llendl; + //llinfos << "LLThread::staticRun() Exiting: " << threadp->mName << llendl; // We're done with the run function, this thread is done executing now. threadp->mStatus = STOPPED; @@ -81,6 +100,8 @@ LLThread::LLThread(const std::string& name, apr_pool_t *poolp) : mAPRThreadp(NULL), mStatus(STOPPED) { + mID = ++sIDIter; + // Thread creation probably CAN be paranoid about APR being initialized, if necessary if (poolp) { @@ -121,7 +142,7 @@ void LLThread::shutdown() // First, set the flag that indicates that we're ready to die setQuitting(); - llinfos << "LLThread::~LLThread() Killing thread " << mName << " Status: " << mStatus << llendl; + //llinfos << "LLThread::~LLThread() Killing thread " << mName << " Status: " << mStatus << llendl; // Now wait a bit for the thread to exit // It's unclear whether I should even bother doing this - this destructor // should netver get called unless we're already stopped, really... @@ -143,7 +164,7 @@ void LLThread::shutdown() if (!isStopped()) { // This thread just wouldn't stop, even though we gave it time - llwarns << "LLThread::~LLThread() exiting thread before clean exit!" << llendl; + //llwarns << "LLThread::~LLThread() exiting thread before clean exit!" << llendl; // Put a stake in its heart. apr_thread_exit(mAPRThreadp, -1); return; @@ -283,7 +304,7 @@ void LLThread::wakeLocked() //============================================================================ LLMutex::LLMutex(apr_pool_t *poolp) : - mAPRMutexp(NULL) + mAPRMutexp(NULL), mCount(0), mLockingThread(NO_THREAD) { //if (poolp) //{ @@ -315,7 +336,18 @@ LLMutex::~LLMutex() void LLMutex::lock() { +#if LL_DARWIN + if (mLockingThread == LLThread::currentID()) +#else + if (mLockingThread == sThreadID) +#endif + { //redundant lock + mCount++; + return; + } + apr_thread_mutex_lock(mAPRMutexp); + #if MUTEX_DEBUG // Have to have the lock before we can access the debug info U32 id = LLThread::currentID(); @@ -323,10 +355,22 @@ void LLMutex::lock() llerrs << "Already locked in Thread: " << id << llendl; mIsLocked[id] = TRUE; #endif + +#if LL_DARWIN + mLockingThread = LLThread::currentID(); +#else + mLockingThread = sThreadID; +#endif } void LLMutex::unlock() { + if (mCount > 0) + { //not the root unlock + mCount--; + return; + } + #if MUTEX_DEBUG // Access the debug info while we have the lock U32 id = LLThread::currentID(); @@ -334,6 +378,8 @@ void LLMutex::unlock() llerrs << "Not locked in Thread: " << id << llendl; mIsLocked[id] = FALSE; #endif + + mLockingThread = NO_THREAD; apr_thread_mutex_unlock(mAPRMutexp); } @@ -351,6 +397,11 @@ bool LLMutex::isLocked() } } +U32 LLMutex::lockingThread() const +{ + return mLockingThread; +} + //============================================================================ LLCondition::LLCondition(apr_pool_t *poolp) : @@ -371,6 +422,10 @@ LLCondition::~LLCondition() void LLCondition::wait() { + if (!isLocked()) + { //mAPRMutexp MUST be locked before calling apr_thread_cond_wait + apr_thread_mutex_lock(mAPRMutexp); + } apr_thread_cond_wait(mAPRCondp, mAPRMutexp); } diff --git a/indra/llcommon/llthread.h b/indra/llcommon/llthread.h index f1c6cd75af..40291a2569 100644 --- a/indra/llcommon/llthread.h +++ b/indra/llcommon/llthread.h @@ -35,8 +35,17 @@ class LLThread; class LLMutex; class LLCondition; +#if LL_WINDOWS +#define ll_thread_local __declspec(thread) +#else +#define ll_thread_local __thread +#endif + class LL_COMMON_API LLThread { +private: + static U32 sIDIter; + public: typedef enum e_thread_status { @@ -77,6 +86,8 @@ public: apr_pool_t *getAPRPool() { return mAPRPoolp; } LLVolatileAPRPool* getLocalAPRFilePool() { return mLocalAPRFilePoolp ; } + U32 getID() const { return mID; } + private: BOOL mPaused; @@ -91,6 +102,7 @@ protected: apr_pool_t *mAPRPoolp; BOOL mIsLocalPool; EThreadStatus mStatus; + U32 mID; //a local apr_pool for APRFile operations in this thread. If it exists, LLAPRFile::sAPRFilePoolp should not be used. //Note: this pool is used by APRFile ONLY, do NOT use it for any other purposes. @@ -128,17 +140,27 @@ protected: class LL_COMMON_API LLMutex { public: + typedef enum + { + NO_THREAD = 0xFFFFFFFF + } e_locking_thread; + LLMutex(apr_pool_t *apr_poolp); // NULL pool constructs a new pool for the mutex virtual ~LLMutex(); void lock(); // blocks void unlock(); bool isLocked(); // non-blocking, but does do a lock/unlock so not free + U32 lockingThread() const; //get ID of locking thread protected: apr_thread_mutex_t *mAPRMutexp; + mutable U32 mCount; + mutable U32 mLockingThread; + apr_pool_t *mAPRPoolp; BOOL mIsLocalPool; + #if MUTEX_DEBUG std::map<U32, BOOL> mIsLocked; #endif diff --git a/indra/llcommon/stdenums.h b/indra/llcommon/stdenums.h index 9f86de124e..556eff8370 100644 --- a/indra/llcommon/stdenums.h +++ b/indra/llcommon/stdenums.h @@ -49,7 +49,8 @@ enum EDragAndDropType DAD_ANIMATION = 12, DAD_GESTURE = 13, DAD_LINK = 14, - DAD_COUNT = 15, // number of types in this enum + DAD_MESH = 15, + DAD_COUNT = 16, // number of types in this enum }; // Reasons for drags to be denied. |