| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
Fix rendering on Parallels VMs with graphics set to "low"
|
|
This reverts commit 1c81813e79c4acc917d75b47117778dd277e5b1f.
|
|
Fix UI issues, improve warnings, and enhance process handling
|
|
|
|
|
|
In LLRender::syncMatrices(), treat uniform location 0 as valid when uploading the inverse model-view, inverse projection, and identity matrices. Only location -1 indicates that a uniform is absent. (This is a zero-based index.)
Fix Parallels GLSL compilation by:
- renaming the local waterClip() wrapper to avoid an overload conflict
- returning float literals (not ints) from a few functions declared to return float
|
|
|
|
This reverts commit 6b4e3f3288ec1e9917cecd862a7a52945e5b4db2, reversing
changes made to 99ab6316b4ae9058f22d9f57d21e795ca45797fd.
|
|
This reverts commit 3d465f6c966f0a2fcfd5d3c9effb6ce20c2754c8, reversing
changes made to f2c356fc7130ce5e38b994dcfa1e210609b458b9.
|
|
This reverts commit a937b237de3651e79cdb517f14381a5bdd4c844b.
|
|
|
|
Simplify key a bit.
|
|
|
|
Dejavu used fixed width numbers by default, tweaked fonts to do the same
for inter (when weight > 0).
|
|
* Add more controls for texture loading budgets. Should yield much faster loading within a given FPS target - should generally self regulate depending on your framerate.
* Harden texture pipeline against stalls and OOM. Generally makes texture loading faster, at the expense of some budgeting (which we weren't doing a great job at anyways).
|
|
|
|
Also make sure that probes don't stamp - that creates thrashing when distant probes render. We're at a low enough resolution on those things that higher mips are likely to go unnoticed, and we bake them frequently enough that we can pick up lower mips when we actually need them.
|
|
|
|
|
|
|
|
sanity check size, catch exceptions. Cache isn't mission critical.
|
|
Texture streaming rework
|
|
|
|
|
|
|
|
New "high res" bubble near the camera, minimum discard settings, and discard scaling.
|
|
This is a big one:
- Reworks the discard signal almost entirely. Now has a normalized 0..1 discard signal: distance x size x channel exponent, floored by staleness and background app state. Shaped by VRAM pressure.
- Textures can now scale down to the smallest GPU mip (1×1), independent of the codec's encoded mip
count.
- Terrain texture LOD now works. Useful for 2K textures and PBR on terrain. Based upon camera distance to nearest terrain patch.
- New texture quality setting. Low/Medium/High/Ultra
- Caps texture resolution on Low to 1024, and otherwise shifts the discard signal around. Makes distance based texture LOD work a lot more predictably.
- We now track last bind state for textures, and discard accordingly. We progressively discard based upon last bind time.
- Avatar textures get a residency boost to stay loaded in VRAM longer under pressure.
|
|
QA found no difference performance wise, I see a decrease in performance.
|
|
|
|
Unused fonts were excluded
|
|
delta_correction should have been substructed, not added
|
|
By Rye
|
|
WIP, beacsue variable fonts are not yeat included.
|
|
by caching string width
|
|
|
|
# Conflicts:
# indra/newview/skins/default/xui/en/panel_preferences_general.xml
|
|
|
|
|
|
|
|
Replace LLUUID and LLMaterialID container hashing functions with more collision resistant versions
Utilize boost::hash_combine for TEMaterialPair to generate good hash distribution
Generalize is_in_map and get_if_there for usage with all mapped types
|
|
|
|
|
|
Signed-off-by: Rye <rye@alchemyviewer.org>
|
|
allocations and utf8->utf16 conversions
Fix LLFile::tmpdir failing when temp directory is located in a unicode path on windows
Add Tracy profiler tagging to LLFile that's disabled by default
Replace LLFile::utf8StringToWstring with our fsyspath std::filesystem::path adapter
Introduce std::filesystem::path to llifstream/llofstream wrappers
Add move operators to fsyspath to match copy operator
Restore LLUniqueFile RAII class for direct c file io wrapping and reducing merge conflicts with downstream code that may depend upon it
Signed-off-by: Rye <rye@alchemyviewer.org>
|
|
# Conflicts:
# indra/llcrashlogger/llcrashlock.cpp
# indra/llrender/llshadermgr.cpp
# indra/test/CMakeLists.txt
|
|
vertexBatchPreTransformed()
|
|
This allows removing #include "llerror.h" from llpointer.h.
Also remove #include "llmutex.h" as a heavy way to get
<boost/functional/hash.hpp>.
That requires adding #include "llmutex.h" to llimage.h, llnotifications.h,
llwatchdog.cpp and llvolumemgr.cpp, which were inheriting it from llpointer.h.
|
