A mission's $Environment Map is used as a cubemap without checking. Naming
an ordinary 2D texture made OpenGL copy six faces out of a one-face image
and crash when the mission started (issue #2245). Check the loaded bitmap
with a new bm_is_cubemap(); if it is not a cubemap, warn, release it and
build the generated environment map, as for a mission that names none.
DX10 (BC7) DDS files carry no cube flag through the loader, so they are let
through unchecked, as before.
Co-authored-by: Dustinspace217 <[email protected]>
* Fix heap overrun in texture readback for the model transparency scan
gr_opengl_get_bitmap_from_texture() sized its readback buffer from the
texture cache slot dimensions and the bmpman frame count, but
glGetTexImage() returns whatever the bound GL texture actually contains.
When texture detail below maximum culls textures on upload, the two can
disagree, and the readback overruns the buffer (observed as heap
corruption when loading Zeus.pof with Texture Detail: Minimum). Size
the readback from the dimensions and layer count OpenGL itself reports
for the texture, and pad the staging buffer to 4x4 block granularity in
case a driver writes tiny compressed mips back in whole blocks.
The readback now also allocates the returned buffer and reports the
dimensions of the data it returns, because the second half of the bug
was on the consumer side: bitmap_lookup indexed the readback data by
the bitmap's own dimensions, so with any culling in effect the
transparency scan read mostly uninitialized memory. bitmap_lookup (and
its cache) now carries the actual readback dimensions, and it skips the
texture cleanly if the readback fails.
Also pairs the cache's vm_malloc allocations with vm_free instead of
free, and no longer ignores a failed tcache_set before reading back.
Co-Authored-By: Claude Fable 5 <[email protected]>
* Harden the transparency-scan texture readback
Return fully opaque from bitmap_lookup::get_channel_alpha() when the
bitmap has no alpha channel, since indexing channel 3 of a 3-channel
buffer reads out of bounds (currently unreachable because the only
caller checks bm_has_alpha_channel() first, but no longer a trap).
Per review, also assert at maximum texture detail, where no culling
occurs on upload, that the readback dimensions OpenGL reports match the
texture cache slot, so any future slot mismatch is caught immediately
in debug builds instead of surfacing as corruption.
Co-Authored-By: Claude Fable 5 <[email protected]>
* Fix inverted dimension check in opengl_export_render_target
The validation rejected the export only if both the width and the
height mismatched the texture cache slot, so a texture wrong in one
dimension passed through into opengl_get_texture, which trusts the
caller-supplied dimensions for its mipmap loop.
Co-Authored-By: Claude Fable 5 <[email protected]>
---------
Co-authored-by: Claude Fable 5 <[email protected]>
* avoid undefined behavior from UNREACHABLE on reachable code paths
UNREACHABLE expands to __assume(false) / __builtin_unreachable() in
release builds, which is undefined behavior if the code is ever actually
reached. Many UNREACHABLE sites only guard against unexpected-but-
possible runtime states (values from tables, mission/save files, network
packets, scripting, or user input), so reaching them should degrade
gracefully rather than invoke UB.
Replace UNREACHABLE with Assertion(false, ...) at all such sites. It
still pops an error in debug builds, but in release it is a harmless
no-op that simply falls through to the following code. Sites that are
genuinely unreachable by local invariants (e.g. a switch over a bounded
random number) are left as UNREACHABLE.
A few converted sites had no safe fall-through -- an uninitialized read,
null dereference, or null call would have followed -- so they also get a
minimal recovery: initialize the value, guard the dereference, return
early, or skip the offending item.
Where the converted site was a simple guard of the form
if (x) { Assertion(false, ...); return; }
hoist the check onto the invariant instead:
Assertion(!x, ...);
if (x) { return; }
This states the expectation directly and shows the offending value in
the debug popup, while keeping the same graceful runtime degradation.
Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
* addendum
After consultation with Lafiel, use UNREACHABLE in situations that are effectively impossible even if not provably impossible. UNREACHABLE = "this shouldn't ever happen"; Assertion = "if this ever happens, it's a coding error". Also fix a few other sites and change a couple to Warnings.
---------
Co-authored-by: Claude Opus 4.8 (1M context) <[email protected]>
* Initial KTX1/ETC2 support
* Conditional glad include
* remove unused function
* define glenum for non opengl build
* remove unused param
* fix qtbuild
* change ktx format to enum
* use nullptr
* ktx: support the same etc2 formats available in amd compressonator
* add comments and missing include
* correct flag type
* clarify variable name for clang
* cleanup
* update glad loader to GL4.3
Adds bcdec lib to convert compressed DDS textures to RGBA based on what the
system supports. BC7 if BPTC extension not supported, DXT? if S3TC extension
not supported.
Should support cubemaps and volume textures but those were not tested.
Addtional changes:
- remove indexed DDS support since it had bugs but didn't work anyway
(R8 should be used instead)
- fix issues with volume textures (nothing to test with)
- remove old compression support flags as they are now obsolete
Fixes#4920
This advances the versions for Clang to 16 and GCC to 13 along with
fixing any warnings these new versions are detecting now.
The new image versions also include Vulkan so the additional workflow
step for installing that is not necessary anymore.
This is a bit tricky because both animated and non-animated PNGs use the same extension. So for PNGs we need to defer the filename length check until we know for sure that it is an animated image.
Fixes#5387.
* add subfolder support to cfile
* sort files in VP archives before indexing them
* fix indexing of sub-subfolders
* remove unused 8B22K & 16B11K sound path types
* add more items to skip subfolder scanning
* fix windows build issues
* fix unit tests by allowing 0-byte files to be indexed
* fix file indexing on windows
* add test for cfile subfolders
* remove specific 'markup' and 'css' interface subfolders
They aren't directly referenced anywhere by pathtype and the new
subfolder code will continue to index and load files from them
* add player binds to list of always-root locations
* remove file-based localization support
An alternative is left in place for old fonts, as that was the only place
it appeared to be used.
This method of file localization conflicts with subfolder support however
so if it's needed in the future then changes to the subfolder code will
be necessary.
* add subfolder support to cf_find_file_location_ext()
* add subfolder tests for cf_find_file_location*()
* add shlwapi to needed win32 libs
* small bit of cleanup
* add subfolder support to cf_get_file_list()
* add tests for cfile subfolder lists
* report files that might be shadowed
* fix issue with lists when a filter is used
Also fixes discrepancy in older cf_get_file_list() when a sub path
is specified (should be used in all cases, not just glob).
* address review suggestions
Fixed environment map rendering
shift fresnel term to match ordering of ComputeGGX, document shader code.
Envmaps: Use tangentMatrix to build local coordinate system
Fixed geometry self-shading calcs to improve shading of rough surfaces.
Cite paper correctly
optimsed env sampling
Put envmap face generation a loop to make it easier to read
Precalculate points on circle for even faster envmap sampling..
Guard against NaNs
Revert to simpler envmap calcs, significantly faster on slower GPUs.
typo
We don't need the sample points any more.
Diffuse IBL shading
Fix gamma issues in envmap rendering
AO term on envmaps.
fix lab crash with retail data
Always render envmaps - PBR shouldn't care if a shine/specmap exists.
The file used by bmp_load()/bm_load_animation() is a hardcoded
interface image on standalone servers. If the value of dir_type
isn't ALL or INTERFACE then it can't find the image. For ship
textures this can cause primary weapon hits to not register on
the server.
Modify bm_load() so that it will only search in intended paths,
improving performance of both fast and slow loading of image
files. Now matches behavior of the other bm_load_*() functions.
* Expose disposing FBO's (with the option to keep the Bitmap Alive) to scripting
* Modernize nullptr
* Moved RenderTarget clearing to own public API
* Fixed Obsolete Comment
* Fix bad deletion of FBOs
* Resolve Memory Leak
* Change FBO cleanup to not reuse old ID's
* FIx oversight
* FIx oversight 2
The amount of output parameters is becoming quite large which makes it
harder to add new input parameters. For that reason I converted the
existing output parameters to a struct which cleans up the function
signature a lot.
Caching bitmaps between missions is no longer a problem for modern
system and this will improve the default load times when playing more
than one mission.
The TGA loading code pretends that it does pixel conversion but doesn't
actually do that. This changes ensures that this circumstance is
communicated to the caller of bm_lock which will allow that code to
handle the case properly.
This was reported by Bomb3rman on discord.
This can be used for mark parts of the code which should be unreachable
in normal execution. In debug mode, if such a statement is reached an
Error dialog is shown which aborts the execution. In release mode the
macro is replaced with a compiler specific expression that lets the
compiler know that this part is unreachable which allows it to optimize
the code better.
This changes how the bitmap slots are managed by the bmpman system.
Instead of allocating a fixed-size array when the game starts this will
manage multiple "blocks" of bitmap slots. When the engine runs out of
slots in the already allocated blocks it simply creates a new block and
allocates the new slot in that.
This removed the hard limit on how many bitmaps can be loaded at the
same time and should fix issues encountered by mods which use a lot of
bitmap slots (I'm looking at you Spoon).
This implements the idea described in #1503.
This is mostly intended for total conversions to allow them to set
themselves apart from FS2 based mods. It uses the new in-memory
filesystem to support a default icon in case no icon was specified by
the mod.
This allows to use more complex file structures in the executable
without requiring that those exist as real files. The rest of the code
does not need to handle the difference between a default file and a real
file anymore which means that it's now possible to embed image files
into the executable since bmpman only needs to use the standard CFile
API.
I added two tests which check if the default file can be found and
opened and if a mod file overrides the default file which should always
be possible.
This allows to render screen space decals into the deferred lighting
buffers no matter how complex the underlying geometry is.
Currently there is only limited support for using this in the engine.
Only weapon impact decals are supported at the moment for testing
purposes.
