* 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]>
* use std::move() for local variables that are copied but not reused
Addresses ~140 Coverity "Variable copied when it could be moved"
defects across 74 files. Every std::move() targets a local variable
whose value is consumed by push_back, emplace, assignment, lambda
capture, or a function call and is never read again afterward.
Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
* clang: use std::make_shared
* clang: use std::make_unique
* address feedback
---------
Co-authored-by: Claude Opus 4.6 (1M context) <[email protected]>
Since bitmaps, suns, and skyboxes store their coordinates as floating-point numbers, we can allow mission designers to specify floating-point numbers in the FRED UI. This PR modifies the dialogs in FRED and QtFRED to allow two decimal places for pitch, bank, and heading.
Implements #7170.
For type safety, use enum class rather than int.
Also make a few miscellaneous tweaks such as adding a new casting macro and making detail level settings a bit cleaner.
Although normalized vectors in FSO are supposed to stay normalized, occasionally small floating point errors can pop up. In most cases these do not cause a problem, but in certain cases these errors can rapidly accumulate and corrupt an otherwise well-behaved matrix. This can most easily be seen with daisy-chained dock sequences or docked ships with angled dockpoints.
This PR audits the usage of `vm_vector_2_matrix_norm` to re-normalize vectors in cases where floating point errors could accumulate. There are also a few new Assertions for extra vigilance.
Follow-up to #6596. Fixes#6634.
Most places in the code that use `vm_vector_2_matrix` do so on already-normalized vectors. Changing these to `vm_vector_2_matrix_norm` avoids redundant normalization.
Fix the inadvertent swapping of the two utility functions for parsing sounds. While we're at it, streamline a few lines, remove a few C-style casts, and remove an unused function prototype.
Follow-up to #6314. Fixes a crash encountered by Inferno on 24.2-RC1.
According to DahBlount:
> Most atan2 implementations for a long time had problems when the inputs were close to machine epsilon or when x was less than or equal to 0
And according to z64555:
> Modern versions of C++ should have an atan2() that works correctly when the inputs are at 0 or near epsilon of the floating point specification, meaning Volition's workaround shouldn't be needed anymore.
So this removes `atan2_safe` and changes all instances to `atan2`. Consequently, the angle range in FRED now shows as [-180, 180] as expected.
Fixes#3730.
* coverity - using `std::move` instead of copying, and change some parameters from value to reference
* coverity - fix or mark potential copy-paste mistakes
* coverity - various checking of parameter and return values
* coverity - miscellaneous fixes, and change some 'intentional' markers to 'false positive' markers
I took a close look at the code involved in parsing particle tables. There is no smoking gun, but there are several ways to polish the code which I have implemented here. The most likely culprits for the slowdown, IMO, are the repeated conversions between SCP_string and C-string, especially in `ParticleManager::getEffectByName()`; followed by the omission of `std::move` in several places.
* Remove old interpolate angles quick
* Remove old interpolate functions
These are no longer used.
* Add linear interpolate vm function
More clearing functions
* Add physics snapshot struct to silo interpolation
Two of these two functions require an object pointer, so they cannot be in physics_state.cpp
With doxygen documentation
* Try to placate Linux
* Address Feedback
* And update declaration, of course
* Attempt to appease linter
Change the casts used in the convenience #defines in floating.h. Since these #defines are used pretty widely, this should reduce the count of C-style casts in the codebase by quite a bit.
Also change the casts in floating.cpp while we're at it, and remove some obsolete or unused code.
This adds a new field to particle effects so that particle rotation can be configured on a per-type basis, not just on a blanket basis. Implements #5941.
The ai-fly-to-ship goal was added early in FSO's development but never worked right. Until recently it could crash; and even aside from the crash the AI would try to reach the target's exact position, causing repeated collisions. This folds ai-fly-to-ship into ai-stay-near-ship, since the behaviors are conceptually the same except for fly-to being satisfied as soon as the target is in range.
This also adds a new "escort" mode usable by both behaviors. The default Volition behavior is to swoop toward the target in a long, graceful arc; the new mode generates a goal point and flies directly to that point. These can be toggled via an optional sexp parameter.
There is also some minor cleanup and a new vec3d operator for convenience.
Needed for Scroll; depends on #5667.
See title. There were some cascade effects, but the changes are mostly contained to the model code. The calls to `gr_render_model` and `g3_render_primitives` use `const_cast` because constifying that API would require a whole other PR.
The `SCP_string_lcase_equal_to()()` and `SCP_string_lcase_less_than()()` functors are rather unwieldy to use in non-functor situations, so this adds conventional function equivalents.
