summaryrefslogtreecommitdiff
path: root/indra
AgeCommit message (Collapse)Author
2017-11-21MAINT-7684: Stop packaging or signing download_update.exe.Nat Goodspeed
2017-11-17MAINT-7977: Per Graham, unbind textures as well as deleting.Nat Goodspeed
2017-11-17MAINT-7977: Additional cleanup per code reviews.Nat Goodspeed
Introduce helper classes to manage paired initProfile() / finishProfile() calls and gBenchmarkProgram.bind() / unbind() calls. Make TextureHolder a class instead of a struct. Per Henri Beauchamp, since gpu_benchmark() takes a very early exit if (!gGLManager.mHasTimerQuery), subsequent tests of mHasTimerQuery are redundant. Remove. One of those tests controls the busted_finish bool, which can never become true. Remove that and all tests on it.
2017-11-17SL-704 - attachment surface area now computed in bulk rather than updated ↵Brad Payne (Vir Linden)
incrementally. Same dubious formula.
2017-11-17SL-763 - disabled in slightly different way, more commentsBrad Payne (Vir Linden)
2017-11-16MAINT-7977: Release the LLRenderTargets when done.Nat Goodspeed
A classic-C array doesn't destroy its individual elements, but a std::vector does. Use a std::vector<LLRenderTarget> for dest, so each LLRenderTarget will be destroyed. ~LLRenderTarget() calls its release() method.
2017-11-16MAINT-7977: If getVertexStrider() returns false, abandon benchmark.Nat Goodspeed
Ruslan tracked the observed crash to assignments (to create a dummy triangle) through an LLStrider<LLVector3> obtained from getVertexStrider(). When getVertexStrider() returns false, produce a warning and just skip the rest of the benchmark test. The one bit of explicit cleanup apparently required by that early exit is a call to LLImageGL::deleteTextures() to match the preceding generateTextures() call. Wrap both in a new TextureHolder class whose destructor takes care of cleanup. The only other references to the corresponding U32 array are a couple calls to LLTexUnit::bindManual(); add a bind() method to support that. Also fix apparent bug in the LL_DARWIN special case for "improbably high and likely incorrect": the code assigned -1.f (the "couldn't compute" value) to gbps, overlooking the fact that gbps is unconditionally recomputed below. In the "likely incorrect" stanza, simply return -1.f instead.
2017-11-16SL-763 - disabling the use_quantum optimization in LLMotionController until ↵Brad Payne (Vir Linden)
it can be fixed.
2017-11-16SL-704 - more cleanup of AXON commented codeBrad Payne (Vir Linden)
2017-11-16SL-704 - more cleanup, including releaseMeshData() for control avatars, ↵Brad Payne (Vir Linden)
which don't need mesh data.
2017-11-16SL-704 - first pass cleaning up comments. Also added ↵Brad Payne (Vir Linden)
LLControlAvatar::initInstance() as a potential hook for optimizations.
2017-11-16MAINT-7977 [Alex Ivy] Feature Table crashesruslantproductengine
In case of buff->getVertexStrider(v) return false it mean that glMapBufferRange() return NULL The next three lines can be the reason of this crash.
2017-11-16MAINT-7228 Vertex buffer allocation failure handlingandreykproductengine
2017-11-14MAINT-7228 Dummy triangle should use map instead of offset (in case of VBO)andreykproductengine
2017-11-14MAINT-7228 Vertex buffer allocation failure handlingandreykproductengine
2017-11-09no-op whitespace commit to unstick the build machineryBrad Payne (Vir Linden)
2017-11-09SL-786 - control avatar deletionBrad Payne (Vir Linden)
2017-11-09MAINT-7913, MAINT-7817 - fix for animesh attachment corruption after ↵Brad Payne (Vir Linden)
teleporting to a different simulator. Turns out the control avatar (skeleton) was getting caught up in the cleanup for the original region
2017-11-08Added debug setting AnimatedObjectIgnoreLimits for server testingBrad Payne (Vir Linden)
2017-11-08SL-831 - animesh objects are not required to have any tris. If they have no ↵Brad Payne (Vir Linden)
tris, we don't bother making a control avatar for them.
2017-11-07MAINT-7979 Debug info should contain correct stateandreykproductengine
2017-11-07MAINT-7966 Second Life viewer logs session ID in SecondLife.logMnikolenko Productengine
2017-11-07SL-830 - fix for previously selected animated objects disappearing when ↵Brad Payne (Vir Linden)
going static
2017-11-06SL-830 - more diagnostics and possible fixes around corrupted graphics stateBrad Payne (Vir Linden)
2017-11-03MAINT-7228 Substituted assert with errorandreykproductengine
2017-11-01mergeBrad Payne (Vir Linden)
2017-11-01SL-830, SL-831 - no linking if any of the roots is animated object. ↵Brad Payne (Vir Linden)
isAnimatedObject() just uses check against extended mesh flag of the root. On select, LLVOVolume marks whole linkset for update if animated object.
2017-11-01MAINT-7228 Vertex buffer allocation failure handlingandreykproductengine
2017-11-03MAINT-7974 Fixed LLProfile crash(error) caused by thread unsafe variableandreykproductengine
2017-11-06MAINT-7978 - Fixed allocation crash in doRead(), workers should never crashandreykproductengine
2017-11-06MAINT-7961 Search link in Recent tab of Inventory should open the Search ↵Mnikolenko Productengine
filters for Inventory
2017-10-31MAINT-2161 Fixed dae parsing issue - dae expects escaped namesandreykproductengine
2017-10-30no-op change to try to get build host unstuckBrad Payne (Vir Linden)
2017-10-30Automated merge with ssh://bitbucket.org/lindenlab/viewer64Nat Goodspeed
2017-10-27SL-731, SL-779 - more diagnostics for DebugAnimatedObjects, fixed a problem ↵Brad Payne (Vir Linden)
with recursiveMarkForUpdate
2017-10-27MAINT-7857 - shadow fix for animated objects.Brad Payne (Vir Linden)
2017-10-26Automated merge with ssh://bitbucket.org/lindenlab/viewer64Nat Goodspeed
2017-10-25SL-779 - refactoring related to rigged-mesh determination. animated objects ↵Brad Payne (Vir Linden)
must contain some triangles even if not at the root
2017-10-25MAINT-7911: Set build_data.json's Platform to win for both W32, W64.Nat Goodspeed
2017-10-25SL-785 - get non-mesh objects in animesh linksets to displayBrad Payne (Vir Linden)
2017-10-25SL-779 - partial support for animesh objects with non-mesh rootsBrad Payne (Vir Linden)
2017-10-24MAINT-7081: Only request HTTP2 with $VIEWERASSET override (testing)Nat Goodspeed
2017-10-24MAINT-7081: Mention libnghttp2*.dylib in yet another place.Nat Goodspeed
2017-10-24MAINT-7081: Merge from lindenlab/viewer64.Nat Goodspeed
2017-10-18MAINT-7812: NSIS check for /marker before writing marker file.Nat Goodspeed
The /marker switch is passed by the (new) VMP. If any user wants to explicitly pass the /marker switch to the installer, s/he shouldn't mind ending up with an nsis.winstall file in the download directory.
2017-10-12MAINT-7081 [FIXED] Access (write) violation / buffer overrun in ↵Kitty Barnett
LLTextureFetchWorker::doWork() The trouble lines are: U8 * buffer = (U8 *) ALLOCATE_MEM(LLImageBase::getPrivatePool(), total_size); if (cur_size > 0) { memcpy(buffer, mFormattedImage->getData(), cur_size); } If 'cur_size > mHttpReplyOffset + append_size' then 'total_size -= src_offset' will cause total_size to be smaller than cur_size causing a write access violation on the memcpy. Since the response is invalid it seemed best to make it follow the other failed partial condition. (transplanted from 737e28ec6b4d74f3ff915a4effc13d7b615a6a9b)
2017-10-27MAINT-1114 crash at LLVOAvatar updateGeometry()andreykproductengine
2017-10-27Merged from viewer-lynxandreykproductengine
2017-10-23MAINT-5261 The "Ping Interpolate the object position" seems to no longer workandreykproductengine
2017-10-23Backed out changeset: bfa432c3c7d5 (SL-775)andreykproductengine
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-26 03:02:43 +0000