| Age | Commit message (Collapse) | Author |
|---|
|
|
|
This file is no longer used.
|
|
This reverts commit 810a3d24c2e3671f926091c062b101bdec6a1517. (secondlife/jira-archive-internal#70482)
|
|
|
|
|
|
|
|
|
|
The extras cache was never fully implemented and thus grows forever.
This coupled with the server side "bug" that sends innumerable blank overrides leads users to collect 100MB+ SLEC files which in turn cause significant pauses post TP.
|
|
Inject XZ_DEFAULTS=-T0 into the environment. This will speed up archive generation.
|
|
* GTK remnants removal
Remove old GTK code and wire up some missing fltk logic in llfilepicker
This will leave linux_crash_logger as an empy stub which could be used as a skeleton in case crash logging gets implemented.
Otherwise linux_crash_logger and what is left of it in some cmake files can be deleted.
* Update SDL2 to latest version.
This revealed openal has also a dependebcy on sndio. Consequently the dependency was moved to OPENMAL.cmake and canot be dropped.
* Handle missing filetypes:
FFSAVE_GLTF,FFLOAD_GLTF, FFLOAD_MATERIAL, FFLOAD_MATERIAL_TEXTURE
|
|
generation.
|
|
This reverts commit d07cd115e3bdef8a2b1d2f6a5a507ddb67ff8cd9.
|
|
|
|
which interferred with getting benefits from LLSD array
|
|
|
|
|
|
|
|
|
|
|
|
CI: adopt xz compression, actions/*-artifact@v4
|
|
Release (Maint W) -> Maint X merge
|
|
png_read_info triggered a PngError, LLAppViewer::frame() handled it
instead of LLPngWrapper::readPng, and since status didn't
change viewer tried to decode image again and again and again.
|
|
# Conflicts:
# indra/llimage/llimageworker.cpp
# indra/llimage/llimageworker.h
# indra/newview/llcontrolavatar.cpp
# indra/newview/llfloaterprofiletexture.cpp
# indra/newview/lloutfitslist.cpp
# indra/newview/lloutfitslist.h
# indra/newview/lltexturefetch.cpp
|
|
Found by running with -fsanitze=thread
Suggestion to avoid accessing invalid memory:
In both cases memory will be allocated by can be accessed beyond bounds.
In LLPolyMesh it can be off by at least one (+x%2). Though I am not even sure if even in best case it always will be a multiple of 16.
In LLViewerJointMesh::updateFaceData the code tries to account for padding by, but the allocation in LLPolyMeshSharedData::allocateVertexData is done without any padding. Thus the sizes must not match.
Replacing the calls with memcpy as a quick fix to see if the error goes away fixed address sanitzer complaining.
It is up to debate if memcpy is a good replacement. LLVector4a::memcpyNonAliased16 was invented for performance. But on the other hand one could argue that nowadays every stdlib maintainer will very heavily optmize functions like memcpy themselves and could take advantage of CPU features the old LL implementation does not take into account.
AVX comes to mind. In any case did I not measure any of this.
|
|
* - Add an option to with ASAN
- Fix GCC -Werror option.
* - find_program should not set REQUIRED if per logic a fallback to ld.bfd is fine
- cmake idiomatic is to test Variables without dereferencing them
* Add -Wno-unknown-warning-option for clang or it will complain about some GCC only options
```
/home/runner/work/viewer/viewer/indra/llcommon/llsdutil.cpp:39:32: error: unknown warning group '-Wstringop-truncation', ignored [-Werror,-Wunknown-warning-option]
#pragma GCC diagnostic ignored "-Wstringop-truncation" // It's actually okay what happens here
```
|
|
|
|
Move towards packaging artifacts with xz, which offers higher compression ratios and faster decode time.
|
|
Remove unused fix-incredibuild.py
|
|
Remove BuildParams
|
|
CI: Remove python-version from matrix
|
|
|
|
Drop python version from matrix configuration as it's always 3.11.
|
|
Add LINK_WITH_MOLD option
|
|
Otherwise the viewer ends with a dangling pointer in the map
|
|
|
|
|
|
viewer#1163: Avoid terrain texture memory leak
|
|
|
|
* #1064 Fix for mirror surfaces going black when the roughness is set too high.
* #1130 Fix for the user's avatar not appearing in mouselook in a mirror.
* #1059 Disable mirrors on low settings per the feature table.
* #860 Modify the settings UI to better reflect that we only have two AA modes: off and FXAA.
* #1191 Add some settings for mirrors in preferences.
* #1185 Add support for runtime mirror resolution changing. Useful for quality settings, and auto-FPS.
|
|
secondlife/viewer#1149 (#1170)
|
|
* Fix mac build
* Mac build take 2
* Mac build take 3
|
|
This is to do with misunderstandings related to how .find()
works with multimaps. .find() will, in fact, return an iterator
to the first iterator it finds, and will iterate through all
elements in the multimap when incremented, not just items with
the same key.
Change code working with animation sources to be aware of this
fact, so unrelated animation sources do not have their animations
stopped.
|
|
Release (Maint W) -> Maint B merge
|
|
Release (Maint W) -> Maint YZ merge
|
|
# Conflicts:
# autobuild.xml
# indra/llcommon/llsys.cpp
|
|
# Conflicts:
# indra/llrender/llgl.cpp
# indra/newview/lloutfitslist.cpp
|
|
* #1126 GLTF Scene import initial prototype (working geometry import for some assets)
* #1126 WIP -- Expand support for more vertex formats, PoC material import, shadow support, scale support
* #1126 move GLTF implementation to newview/gltf
* #1126 Refactor attribute loading to be less copy/pasta for each combination of types
* #1126 Partially working object selection. Ability to have multiple scenes at once. Helpful message on how to use the preview button.
* #1126 Add bounding box debug display and untangle GLTF raycast from LLVOVolume raycast
* #1126 Working raycast on GLTF scenes.
* #1126 Remove some #pragma optimize offs
|
|
|
|
following promotion of secondlife/viewer #690
|
|
secondlife/viewer#759: Fix minimap loading for 2k PBR materials
|
