| Age | Commit message (Collapse) | Author |
|---|
|
|
|
This is a naive implementation based on the symmetry with `asin`.
Before, I'm not really sure what we were doing, but it was returning
wildly incorrect results.
|
|
Caught by ENABLE_ALL_THE_DEBUG_MACROS.
|
|
This isn't a complete conversion to ErrorOr<void>, but a good chunk.
The end goal here is to propagate buffer allocation failures to the
caller, and allow the use of TRY() with formatting functions.
|
|
These will allow us to start using TRY() with StringBuilder operations.
|
|
|
|
Also add slightly richer parse errors now that we can include a string
literal with returned errors.
This will allow us to use TRY() when working with JSON data.
|
|
|
|
This feels like it was a refactor transition kind of conversion. The
places that were relying on it can easily be changed to explicitly ask
for the ptr() or a new vaddr() method on Userspace<T*>.
FlatPtr can still implicitly convert to Userspace<T> because the
constructor is not explicit, but there's quite a few more places that
are relying on that conversion.
|
|
Returning a const BitmapView doesn't make much sense :^)
|
|
In AK::Function::CallableWrapper::init_and_swap(), clang-tidy wants us
to mark the destination argument as pointer to const, which doesn't work
because we use placement new to construct a move'd *this into.
|
|
cert-dcl50-cpp: No variadic functions, suppressed in RefCounted and
ThreadSafeRefCounted for implementing the magic one_ref_left and
will_be_destroyed functions.
cert-dcl58-cpp: No opening ::std, suppressed in the places we put names
in ::std to aid tools (move, forward, nullptr_t, align_val_t, etc).
|
|
|
|
Subclasses of StandardFormatter don't need to use the formatter argument
in their constructor, so move() it into the base class initializer.
|
|
Refactor to a visitor+functor pattern that does an explicit static_cast
to size_t after performing suitable range checks for each type.
|
|
... In files included by Kernel/Process.cpp and Kernel/Thread.cpp
|
|
|
|
Either not returning *this, or in the case of Variant, not checking for
self assignment. In AK::Atomic, we can't return *this due to the wrapper
semantics Atomic implements.
|
|
... In files included by Kernel/Process.cpp and Kernel/Thread.cpp
|
|
Using `l` for long double causes a clang-tidy warning, so use all caps
suffixes for all of the AK::abs() overloads for consistency. Also, avoid
leaking the internal __DEFINE_GENERIC_ABS macro.
|
|
|
|
We had these declared as returning char, which looks like a copy paste
error. Found by clang-tidy.
|
|
The definition of VERIFY_NOT_REACHED() as `assert(false)` causes the
tool to suggest converting it to a static_assert. Which doesn't make
any sense in context for what the macro is trying to do: crash the
program at runtime.
|
|
Add a basic clang-tidy configuration that enables checks from the
following categories:
- bugprone
- cert
- clang-analyzer
- concurrency
- misc
- performance
- portability
- readability
The noisiest rules that have conflicts with the project style guide or
accepted practices have been turned off.
There's absolutely more work to be done here before we could consider
setting any of these warnings as errors and enforcing them in CI, but
committing a project clang-tidy configuration should help the rules
become more visible and let other contributors take a crack at tweaking
rules and/or finding possible bugs.
Sadly the cpp-core-guidelines and modernize categories are very, very
noisy. If we want to enable rules from these categories, they would need
to be on a rule by rule basis.
|
|
Previously if code attempted to use the format specifier somewhere
(Ex: "%#4.3Lg"), the specifier would get dropped and we would just
print "g" instead of any value. Now at least we print a value.
|
|
|
|
|
|
|
|
|
|
This allows us to use TRY() and MUST() with them.
|
|
Same as Vector, ByteBuffer now also signals allocation failure by
returning an ENOMEM Error instead of a bool, allowing us to use the
TRY() and MUST() patterns.
|
|
Instead of signalling allocation failure with a bool return value
(false), we now use ErrorOr<void> and return ENOMEM as appropriate.
This allows us to use TRY() and MUST() with Vector. :^)
|
|
This is generally useful so let's move it to AK. Also it seems that we
don't need the temporary variable hack anymore, so let's lose that.
|
|
Preparation for using Error.h from Vector.h. This required moving some
things out of line.
|
|
All the read-only methods of Bitmap simply defer to BitmapView. Let's
make this relationship official by using class inheritance. This might
even shave off a few instructions, although any sufficiently optimizing
compiler probably already optimized them away.
|
|
|
|
We now use AK::Error and AK::ErrorOr<T> in both kernel and userspace!
This was a slightly tedious refactoring that took a long time, so it's
not unlikely that some bugs crept in.
Nevertheless, it does pass basic functionality testing, and it's just
real nice to finally see the same pattern in all contexts. :^)
|
|
This is preparation for using Error in the kernel instead of KResult.
|
|
Users of this API will want TRY().
|
|
This variant returns ErrorOr<NonnullOwnPtr<T>> instead of KResultOr.
Eventually the KResultOr variant should go away once the kernel adopts
Error and ErrorOr<T>.
|
|
This has been superseded by the more generally useful AK::Error :^)
|
|
|
|
Replace Result<T, E> with ErrorOr<T> and propagate the error to callers.
|
|
|
|
This feature confuses CLion's parser, so let's turn it off if we see the
__CLION_IDE_ macro as well.
|
|
ErrnoCode is not a thing outside __serenity__, so let's not make
assumptions about it existing.
|
|
This is an alternative to ErrorOr<T>::release_value() that can be used
when converting code to signal that we're releasing the value without
error propagation as a way to move forward now.
This makes these cases much easier to find later on, once more paths for
error propagation are available.
|
|
|
|
We already had this mechanism in the kernel. Let's have it in userspace
as well. This return an ErrorOr<NonnullRefPt<T>>. :^)
|
|
The goal with these is to eventually replace AK::Result, KResult and
KResultOr<T> with something that works (and makes sense) in both kernel
and userspace.
This first cut of Error can be made from an errno code, or from a string
literal (StringView)
|
