Singularity-Viewer/SingularityViewer
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

WIP: Make curl thread code robust and flexible.

Browse Source 
Conflicts:

	indra/llmessage/llcurl.cpp
	indra/llmessage/llcurl.h
	indra/newview/app_settings/settings.xml
	indra/newview/llappviewer.cpp
	indra/newview/llmeshrepository.cpp

Resolved:

	indra/llmessage/llcurl.cpp:

	  Basically removed (not used anyway)

	indra/llmessage/llcurl.h:

	  Basically removed (just includes aiculr.h now)

	indra/newview/app_settings/settings.xml:

	  CurlUseMultipleThreads was remvoved.
	  CurlMaximumNumberOfHandles and CurlRequestTimeOut
	  are still in there, but unused at the moment.

	indra/newview/llappviewer.cpp:

	  CurlMaximumNumberOfHandles and CurlRequestTimeOut
	  are unused at the moment.

	indra/newview/llmeshrepository.cpp:

	  Lock mSignal always (is unlocked inside wait()).
	  Use mSignal lock to see if we are waiting; remove mWaiting.
	  Return false from the MeshFetch functions iff we have to retry
	  a HTTP fetch. Catch the error exception thrown by getByteRange
	  instead of using it's return value (always returns true
	  anyway).
This commit is contained in:
Aleric Inglewood
2012-06-27 12:57:07 +02:00
parent 221e3908b9
commit 69ca6cd5b2
47 changed files with 4173 additions and 2502 deletions
Download Patch File Download Diff File Expand all files Collapse all files

213
indra/newview/llmeshrepository.cpp
View File

@@ -35,7 +35,7 @@
#include "llappviewer.h"
#include "llbufferstream.h"
#include "llcallbacklist.h"
#include "llcurl.h"
#include "llcurlrequest.h"
#include "lldatapacker.h"
#include "llfasttimer.h"
#if MESH_IMPORT
@@ -465,7 +465,6 @@ public:
LLMeshRepoThread::LLMeshRepoThread()
: LLThread("mesh repo")
{
mWaiting = false;
mMutex = new LLMutex();
mHeaderMutex = new LLMutex();
mSignal = new LLCondition();
@@ -483,7 +482,7 @@ LLMeshRepoThread::~LLMeshRepoThread()
void LLMeshRepoThread::run()
{
mCurlRequest = new LLCurlRequest();
mCurlRequest = new AICurlInterface::Request;
#if MESH_IMPORT
LLCDResult res = LLConvexDecomposition::initThread();
if (res != LLCD_OK)
@@ -492,13 +491,10 @@ void LLMeshRepoThread::run()
}
#endif //MESH_IMPORT
mSignal->lock();
while (!LLApp::isQuitting())
{
mWaiting = true;
mSignal->wait();
mWaiting = false;
if (!LLApp::isQuitting())
// Left braces in order not to change the indentation.
{
static U32 count = 0;
@@ -511,38 +507,53 @@ void LLMeshRepoThread::run()
}
// NOTE: throttling intentionally favors LOD requests over header requests
while (!mLODReqQ.empty() && count < MAX_MESH_REQUESTS_PER_SECOND && sActiveLODRequests < (S32)sMaxConcurrentRequests)
{
if (mMutex)
{
mMutex->lock();
LODRequest req = mLODReqQ.front();
mLODReqQ.pop();
LLMeshRepository::sLODProcessing--;
mMutex->unlock();
if (!fetchMeshLOD(req.mMeshParams, req.mLOD, count))//failed, resubmit
try
{
fetchMeshLOD(req.mMeshParams, req.mLOD, count);
}
catch(AICurlNoEasyHandle const& error)
{
llwarns << "fetchMeshLOD() failed: " << error.what() << llendl;
mMutex->lock();
mLODReqQ.push(req) ;
LLMeshRepository::sLODProcessing++;
mLODReqQ.push(req);
mMutex->unlock();
break;
}
}
}
while (!mHeaderReqQ.empty() && count < MAX_MESH_REQUESTS_PER_SECOND && sActiveHeaderRequests < (S32)sMaxConcurrentRequests)
{
if (mMutex)
{
mMutex->lock();
HeaderRequest req = mHeaderReqQ.front();
mHeaderReqQ.pop();
mMutex->unlock();
if (!fetchMeshHeader(req.mMeshParams, count))//failed, resubmit
bool success = false;
try
{
success = fetchMeshHeader(req.mMeshParams, count);
}
catch(AICurlNoEasyHandle const& error)
{
llwarns << "fetchMeshHeader() failed: " << error.what() << llendl;
}
if (!success)
{
mMutex->lock();
mHeaderReqQ.push(req) ;
mMutex->unlock();
break;
}
}
}
@@ -552,7 +563,16 @@ void LLMeshRepoThread::run()
for (std::set<LLUUID>::iterator iter = mSkinRequests.begin(); iter != mSkinRequests.end(); ++iter)
{
LLUUID mesh_id = *iter;
if (!fetchMeshSkinInfo(mesh_id))
bool success = false;
try
{
success = fetchMeshSkinInfo(mesh_id);
}
catch(AICurlNoEasyHandle const& error)
{
llwarns << "fetchMeshSkinInfo(" << mesh_id << ") failed: " << error.what() << llendl;
}
if (!success)
{
incomplete.insert(mesh_id);
}
@@ -565,7 +585,16 @@ void LLMeshRepoThread::run()
for (std::set<LLUUID>::iterator iter = mDecompositionRequests.begin(); iter != mDecompositionRequests.end(); ++iter)
{
LLUUID mesh_id = *iter;
if (!fetchMeshDecomposition(mesh_id))
bool success = false;
try
{
success = fetchMeshDecomposition(mesh_id);
}
catch(AICurlNoEasyHandle const& error)
{
llwarns << "fetchMeshDecomposition(" << mesh_id << ") failed: " << error.what() << llendl;
}
if (!success)
{
incomplete.insert(mesh_id);
}
@@ -578,7 +607,16 @@ void LLMeshRepoThread::run()
for (std::set<LLUUID>::iterator iter = mPhysicsShapeRequests.begin(); iter != mPhysicsShapeRequests.end(); ++iter)
{
LLUUID mesh_id = *iter;
if (!fetchMeshPhysicsShape(mesh_id))
bool success = false;
try
{
success = fetchMeshPhysicsShape(mesh_id);
}
catch(AICurlNoEasyHandle const& error)
{
llwarns << "fetchMeshPhysicsShape(" << mesh_id << ") failed: " << error.what() << llendl;
}
if (!success)
{
incomplete.insert(mesh_id);
}
@@ -586,14 +624,11 @@ void LLMeshRepoThread::run()
mPhysicsShapeRequests = incomplete;
}
mCurlRequest->process();
}
mSignal->wait();
}
if (mSignal->isLocked())
{ //make sure to let go of the mutex associated with the given signal before shutting down
mSignal->unlock();
}
mSignal->unlock();
#if MESH_IMPORT
res = LLConvexDecomposition::quitThread();
@@ -681,21 +716,15 @@ std::string LLMeshRepoThread::constructUrl(LLUUID mesh_id)
bool LLMeshRepoThread::fetchMeshSkinInfo(const LLUUID& mesh_id)
{ //protected by mMutex
if (!mHeaderMutex)
{
return false;
}
mHeaderMutex->lock();
if (mMeshHeader.find(mesh_id) == mMeshHeader.end())
{ //we have no header info for this mesh, do nothing
{
// We have no header info for this mesh, try again later.
mHeaderMutex->unlock();
return false;
}
bool ret = true ;
U32 header_size = mMeshHeaderSize[mesh_id];
if (header_size > 0)
@@ -743,12 +772,10 @@ bool LLMeshRepoThread::fetchMeshSkinInfo(const LLUUID& mesh_id)
std::string http_url = constructUrl(mesh_id);
if (!http_url.empty())
{
ret = mCurlRequest->getByteRange(http_url, headers, offset, size,
new LLMeshSkinInfoResponder(mesh_id, offset, size));
if(ret)
{
LLMeshRepository::sHTTPRequestCount++;
}
// This might throw AICurlNoEasyHandle.
mCurlRequest->getByteRange(http_url, headers, offset, size,
new LLMeshSkinInfoResponder(mesh_id, offset, size));
LLMeshRepository::sHTTPRequestCount++;
}
}
}
@@ -758,26 +785,22 @@ bool LLMeshRepoThread::fetchMeshSkinInfo(const LLUUID& mesh_id)
}
//early out was not hit, effectively fetched
return ret;
return true;
}
//return false if failed to get header
bool LLMeshRepoThread::fetchMeshDecomposition(const LLUUID& mesh_id)
{ //protected by mMutex
if (!mHeaderMutex)
{
return false;
}
mHeaderMutex->lock();
if (mMeshHeader.find(mesh_id) == mMeshHeader.end())
{ //we have no header info for this mesh, do nothing
{
// We have no header info for this mesh, try again later.
mHeaderMutex->unlock();
return false;
}
U32 header_size = mMeshHeaderSize[mesh_id];
bool ret = true ;
if (header_size > 0)
{
@@ -824,12 +847,10 @@ bool LLMeshRepoThread::fetchMeshDecomposition(const LLUUID& mesh_id)
std::string http_url = constructUrl(mesh_id);
if (!http_url.empty())
{
ret = mCurlRequest->getByteRange(http_url, headers, offset, size,
new LLMeshDecompositionResponder(mesh_id, offset, size));
if(ret)
{
LLMeshRepository::sHTTPRequestCount++;
}
// This might throw AICurlNoEasyHandle.
mCurlRequest->getByteRange(http_url, headers, offset, size,
new LLMeshDecompositionResponder(mesh_id, offset, size));
LLMeshRepository::sHTTPRequestCount++;
}
}
}
@@ -839,26 +860,22 @@ bool LLMeshRepoThread::fetchMeshDecomposition(const LLUUID& mesh_id)
}
//early out was not hit, effectively fetched
return ret;
return true;
}
//return false if failed to get header
bool LLMeshRepoThread::fetchMeshPhysicsShape(const LLUUID& mesh_id)
{ //protected by mMutex
if (!mHeaderMutex)
{
return false;
}
mHeaderMutex->lock();
if (mMeshHeader.find(mesh_id) == mMeshHeader.end())
{ //we have no header info for this mesh, do nothing
{
// We have no header info for this mesh, retry later.
mHeaderMutex->unlock();
return false;
}
U32 header_size = mMeshHeaderSize[mesh_id];
bool ret = true ;
if (header_size > 0)
{
@@ -905,13 +922,10 @@ bool LLMeshRepoThread::fetchMeshPhysicsShape(const LLUUID& mesh_id)
std::string http_url = constructUrl(mesh_id);
if (!http_url.empty())
{
ret = mCurlRequest->getByteRange(http_url, headers, offset, size,
new LLMeshPhysicsShapeResponder(mesh_id, offset, size));
if(ret)
{
LLMeshRepository::sHTTPRequestCount++;
}
// This might throw AICurlNoEasyHandle.
mCurlRequest->getByteRange(http_url, headers, offset, size,
new LLMeshPhysicsShapeResponder(mesh_id, offset, size));
LLMeshRepository::sHTTPRequestCount++;
}
}
else
@@ -925,7 +939,7 @@ bool LLMeshRepoThread::fetchMeshPhysicsShape(const LLUUID& mesh_id)
}
//early out was not hit, effectively fetched
return ret;
return true;
}
//return false if failed to get header
@@ -944,14 +958,14 @@ bool LLMeshRepoThread::fetchMeshHeader(const LLVolumeParams& mesh_params, U32& c
LLMeshRepository::sCacheBytesRead += bytes;
file.read(buffer, bytes);
if (headerReceived(mesh_params, buffer, bytes))
{ //did not do an HTTP request, return false
{
// Already have header, no need to retry.
return true;
}
}
}
//either cache entry doesn't exist or is corrupt, request header from simulator
bool retval = true ;
std::vector<std::string> headers;
headers.push_back("Accept: application/octet-stream");
@@ -961,29 +975,19 @@ bool LLMeshRepoThread::fetchMeshHeader(const LLVolumeParams& mesh_params, U32& c
//grab first 4KB if we're going to bother with a fetch. Cache will prevent future fetches if a full mesh fits
//within the first 4KB
//NOTE -- this will break of headers ever exceed 4KB
retval = mCurlRequest->getByteRange(http_url, headers, 0, 4096, new LLMeshHeaderResponder(mesh_params));
if(retval)
{
LLMeshRepository::sHTTPRequestCount++;
}
// This might throw AICurlNoEasyHandle.
mCurlRequest->getByteRange(http_url, headers, 0, 4096, new LLMeshHeaderResponder(mesh_params));
LLMeshRepository::sHTTPRequestCount++;
count++;
}
return retval;
return true;
}
//return false if failed to get mesh lod.
bool LLMeshRepoThread::fetchMeshLOD(const LLVolumeParams& mesh_params, S32 lod, U32& count)
void LLMeshRepoThread::fetchMeshLOD(const LLVolumeParams& mesh_params, S32 lod, U32& count)
{ //protected by mMutex
if (!mHeaderMutex)
{
return false;
}
mHeaderMutex->lock();
bool retval = true;
LLUUID mesh_id = mesh_params.getSculptID();
U32 header_size = mMeshHeaderSize[mesh_id];
@@ -1019,7 +1023,7 @@ bool LLMeshRepoThread::fetchMeshLOD(const LLVolumeParams& mesh_params, S32 lod,
if (lodReceived(mesh_params, lod, buffer, size))
{
delete[] buffer;
return true;
return;
}
}
@@ -1033,13 +1037,10 @@ bool LLMeshRepoThread::fetchMeshLOD(const LLVolumeParams& mesh_params, S32 lod,
std::string http_url = constructUrl(mesh_id);
if (!http_url.empty())
{
retval = mCurlRequest->getByteRange(constructUrl(mesh_id), headers, offset, size,
// This might throw AICurlNoEasyHandle.
mCurlRequest->getByteRange(constructUrl(mesh_id), headers, offset, size,
new LLMeshLODResponder(mesh_params, lod, offset, size));
if(retval)
{
LLMeshRepository::sHTTPRequestCount++;
}
LLMeshRepository::sHTTPRequestCount++;
count++;
}
else
@@ -1056,8 +1057,6 @@ bool LLMeshRepoThread::fetchMeshLOD(const LLVolumeParams& mesh_params, S32 lod,
{
mHeaderMutex->unlock();
}
return retval;
}
bool LLMeshRepoThread::headerReceived(const LLVolumeParams& mesh_params, U8* data, S32 data_size)
@@ -1102,7 +1101,7 @@ bool LLMeshRepoThread::headerReceived(const LLVolumeParams& mesh_params, U8* dat
LLMutexLock lock(mHeaderMutex);
mMeshHeaderSize[mesh_id] = header_size;
mMeshHeader[mesh_id] = header;
}
}
//check for pending requests
pending_lod_map::iterator iter = mPendingLOD.find(mesh_params);
@@ -1599,7 +1598,7 @@ void LLMeshUploadThread::generateHulls()
void LLMeshUploadThread::doWholeModelUpload()
{
mCurlRequest = new LLCurlRequest();
mCurlRequest = new AICurlInterface::Request();
if (mWholeModelUploadURL.empty())
{
@@ -1614,6 +1613,7 @@ void LLMeshUploadThread::doWholeModelUpload()
LLSD body = full_model_data["asset_resources"];
dump_llsd_to_file(body,make_dump_name("whole_model_body_",dump_num));
LLCurlRequest::headers_t headers;
//FIXME: this might throw AICurlNoEasyHandle
mCurlRequest->post(mWholeModelUploadURL, headers, body,
new LLWholeModelUploadResponder(this, full_model_data, mUploadObserverHandle), mMeshUploadTimeOut);
do
@@ -1635,7 +1635,7 @@ void LLMeshUploadThread::requestWholeModelFee()
{
dump_num++;
mCurlRequest = new LLCurlRequest();
mCurlRequest = new AICurlInterface::Request;
generateHulls();
@@ -1645,6 +1645,7 @@ void LLMeshUploadThread::requestWholeModelFee()
mPendingUploads++;
LLCurlRequest::headers_t headers;
//FIXME: this might throw AICurlNoEasyHandle
mCurlRequest->post(mWholeModelFeeCapability, headers, model_data,
new LLWholeModelFeeResponder(this,model_data, mFeeObserverHandle), mMeshUploadTimeOut);
@@ -1665,11 +1666,6 @@ void LLMeshUploadThread::requestWholeModelFee()
void LLMeshRepoThread::notifyLoadedMeshes()
{//called via gMeshRepo.notifyLoadedMeshes(). mMutex already locked
if (!mMutex)
{
return;
}
while (!mLoadedQ.empty())
{
mMutex->lock();
@@ -2373,17 +2369,17 @@ void LLMeshRepository::notifyLoadedMeshes()
mInventoryQ.pop();
}
}
#endif //MESH_IMPORT
//call completed callbacks on finished decompositions
mDecompThread->notifyCompleted();
if (!mThread->mWaiting)
{ //curl thread is churning, wait for it to go idle
if (!mThread->mSignal->tryLock())
{
// Curl thread is churning, wait for it to go idle.
return;
}
mThread->mSignal->unlock();
static std::string region_name("never name a region this");
@@ -3329,6 +3325,7 @@ void LLPhysicsDecomp::run()
mStageID[stages[i].mName] = i;
}
mSignal->lock();
while (!mQuitting)
{
mSignal->wait();
@@ -3357,14 +3354,10 @@ void LLPhysicsDecomp::run()
}
}
}
mSignal->unlock();
decomp->quitThread();
if (mSignal->isLocked())
{ //let go of mSignal's associated mutex
mSignal->unlock();
}
mDone = true;
#endif //MESH_IMPORT
}