| Age | Commit message (Collapse) | Author |
|---|
|
|
|
They're mostly the same apart from some x86_64-specific parts.
|
|
|
|
This solves tearing issues and improves performance when updating the
VirtIOGPU framebuffers.
|
|
Depending on the driver, the second buffer may not be located right
after the first, e.g. it may be page aligned. This removes this
assumption and queries the driver for the appropriate offset.
|
|
Some devices may require DMA transfers to flush the updated buffer
areas prior to flipping. For those devices we track the areas that
require flushing prior to the next flip. For devices that do not
support flipping, but require flushing, we'll simply flush after
updating the front buffer.
This also adds a small optimization that skips these steps entirely for
a screen that doesn't have any updates that need to be rendered.
|
|
Specifically, explicitly specify the checked type, use the resulting
value instead of doing the same calculation twice, and break down
calculations to discrete operations to ensure no intermediary overflows
are missed.
|
|
Because the remainder variable will always be 0 unless a fault happened
we should not use it to decide if we have nothing left to memset when
finishing the fast path. This caused not all bytes to be zeroed
if the size was not an exact multiple of sizeof(size_t).
Fixes #8352
|
|
|
|
When retrieving and setting x86 MSRs two registers are required. The
existing setter and getter for the MSR class made this implementation
detail visible to the caller. This changes the setter and getter to
use u64 instead.
|
|
We can't safely update the other threads' FS selector. This shouldn't
be a problem in practice because allocate_tls() is only used by the
loader.
|
|
|
|
|
|
|
|
Let's put the PCI IDs as enums in the PCI namespace so they're free to
pollute that namespace, but it's also more easier to use them.
|
|
|
|
We use a switch-case statements to ensure we try to find the best
suitable driver for a specific graphics card. In case we don't find
such, we use the default statement to initialize the graphics card as a
generic VGA adapter, if the adapter is VGA compatible.
If we couldn't initialize the driver, we don't touch this adapter
anymore.
Also, GraphicsDevice should not be tied to a PCI::Address member, as it
can be theortically be used with other buses (e.g. ISA cards).
|
|
We were building with red-zone before, but were not accounting for it on
signal handler entries. This should fix that.
Also shorten the stack alignment calculations for this.
|
|
Right now we're using the FS segment for our per-CPU struct. On x86_64
there's an instruction to switch between a kernel and usermode GS
segment (swapgs) which we could use.
This patch doesn't update the rest of the code to use swapgs but it
prepares for that by using the GS segment instead of the FS segment.
|
|
These small changes fix the remaining warnings that come up during
kernel compilation with Clang. These specific fixes were for benign
things: unused lambda captures and braces around scalar initializers.
|
|
|
|
When creating uninitialized storage for variables, we need to make sure
that the alignment is correct. Fixes a KUBSAN failure when running
kernels compiled with Clang.
In `Syscalls/socket.cpp`, we can simply use local variables, as
`sockaddr_un` is a POD type.
Along with moving the `alignas` specifier to the correct member,
`AK::Optional`'s internal buffer has been made non-zeroed by default.
GCC emitted bogus uninitialized memory access warnings, so we now use
`__builtin_launder` to tell the compiler that we know what we are doing.
This might disable some optimizations, but judging by how GCC failed to
notice that the memory's initialization is dependent on `m_has_value`,
I'm not sure that's a bad thing.
|
|
|
|
The `#pragma GCC diagnostic` part is needed because the class has
virtual methods with the same name but different arguments, and Clang
tries to warn us that we are not actually overriding anything with
these.
Weirdly enough, GCC does not seem to care.
|
|
Fixes #8383.
|
|
|
|
|
|
Fixes #6948.
|
|
Now we use WeakPtrs to break Ref-counting cycle. Also, we call the
prepare_for_deletion method to ensure deleted objects are ready for
deletion. This is necessary to ensure we don't keep dead processes,
which would become zombies.
In addition to that, add some debug prints to aid debug in the future.
|
|
|
|
|
|
Let's just add the padding before we jump to interrupt_common_asm_entry.
|
|
Move FPUState allocation to Thread::try_create so that allocation
failure can be observed properly by the caller.
|
|
Also make AllocationHeader acquisition from pointers more verbose
|
|
Also let the compiler enforce the size and type restrictions
|
|
|
|
|
|
|
|
|
|
They're not yet properly supported.
|
|
These were already properly aligned (as far as I can tell).
|
|
|
|
Starting usermode threads doesn't currently work on x86_64. With this
stubbed out we can get text mode to boot though.
|
|
|
|
This changes the m_parts, m_dirname, m_basename, m_title and m_extension
member variables to StringViews onto the m_string String. It also
removes the m_is_absolute member in favour of computing if a path is
absolute in the is_absolute() getter. Due to this, the canonicalize()
method has been completely rewritten.
The parts() getter still returns a Vector<String>, although it is no
longer a const reference as m_parts is no longer a Vector<String>.
Rather, it is constructed from the StringViews in m_parts upon request.
The parts_view() getter has been added, which returns Vector<StringView>
const&. Most previous users of parts() have been changed to use
parts_view(), except where Strings are required.
Due to this change, it's is now no longer allow to create temporary
LexicalPath objects to call the dirname, basename, title, or extension
getters on them because the returned StringViews will point to possible
freed memory.
|
|
The LexicalPath instance methods dirname(), basename(), title() and
extension() will be changed to return StringView const& in a further
commit. Due to this, users creating temporary LexicalPath objects just
to call one of those getters will recieve a StringView const& pointing
to a possible freed buffer.
To avoid this, static methods for those APIs have been added, which will
return a String by value to avoid those problems. All cases where
temporary LexicalPath objects have been used as described above haven
been changed to use the static APIs.
|
|
There is a check in map_bus_region to make sure we don't pointlessly
remap the bus region if the previous mapping was for the same bus.
This is tracked with `m_mapped_bus`.
However, nothing was actually updating `m_mapped_bus`, and it is
initialised to 0. This means that if we start with a device on bus 0,
the read in data will be valid. If we map say bus 1 then bus 0 again,
the map for bus 0 will now be ignored and invalid data will be read in.
Fixed by updating `m_mapped_bus` with the currently mapped bus.
|
|
This can be accessed after init via lspci.
|
|
This hack allows self-test mode run-tests-and-shutdown.sh to give
TestProcFs a stat(2)-able /proc/self/fd/0. For some reason, when
stdin is a SerialDevice, /proc/self/fd/0 will be a symlink to the device
as expected, but, calling realpath or stat on /proc/self/fd/0 will error
out. realpath will give the string from Device::absolute_path() which
would be something like "device:4,64 (SerialDevice)". When VFS is trying
to resolve_path so that we can stat the file, it would bail out on this
fake-y path.
Change the fake path (that doesn't show up when you ls a device, nor
when checking the devices tab in SystemMonitor) from the major/minor
device number and class_name() to /dev/device_name(). There's probably
a very hairy yak standing behind this issue that was only discovered due
to the ProcFS rework.
|
|
This fixes the x86_64 kernel build. :^)
|
