| Age | Commit message (Collapse) | Author |
|---|
|
This flag doesn't conform to any POSIX standard nor is found in any OS
out there. The idea behind this mount flag is to ensure that only
non-regular files will be placed in a filesystem, which includes device
nodes, symbolic links, directories, FIFOs and sockets. Currently, the
only valid case for using this mount flag is for TmpFS instances, where
we want to mount a TmpFS but disallow any kind of regular file and only
allow other types of files on the filesystem.
|
|
Although this code worked quite well, it is considered to be a code
duplication with the TmpFS code which is more tested and works quite
well for a variety of cases. The only valid reason to keep this
filesystem was that it enforces that no regular files will be created at
all in the filesystem. Later on, we will re-introduce this feature in a
sane manner. Therefore, this can be safely removed after SystemServer no
longer uses this filesystem type anymore.
|
|
Later on we will remove the DevTmpFS code, so in order to support
mounting TmpFS instead, we need to be able to create device nodes on
the filesystem.
|
|
Also update Random.cpp such that it builds for aarch64.
|
|
And also implemented and update the related functions.
|
|
KUBSAN complained about a misaligned address when trying to construct
the Thread class.
|
|
ProcessProcFSTraits.cpp is also added the CMakeLists.
|
|
|
|
|
|
Instead of using absolute paths which is considered an abstraction layer
violation between the kernel and userspace, let's not hardcode the path
to children PID directories but instead we can use relative path links
to them.
|
|
With all these functions stubbed, Scheduler.cpp is now buidable!
|
|
This expresses the intent better, and we shouldn't be calling global
functions anyway.
|
|
|
|
A few more stubs have been added to support the building of this file
|
|
|
|
Slowly getting there...
|
|
Having this function return `nullptr` explicitly triggers the compiler's
inbuilt checker, as it knows the destination is null. Having this as a
static (scoped) variable for now circumvents this problem.
|
|
|
|
|
|
|
|
Decompose the current monolithic USBD Pipe interface into several
subclasses, one for each pair of endpoint type & direction. This is to
make it more clear what data and functionality belongs to which Pipe
type, and prevent nonsensical things like trying to execute a control
transfer on a non-control pipe. This is important, because the Pipe
class is the interface by which USB device drivers will interact with
the HCD, so the clearer and more explicit this interface is the better.
|
|
Allocate DMA buffer pages for use within the USBD Pipe class, and allow
for the user to specify the size of this buffer, rounding up to the
next page boundary.
|
|
Also remove the check for aarch64 in AK/Format.cpp, so now the format
functions will prepend the time since boot!
|
|
This sets up the RPi::Timer to trigger an interurpt every 4ms using one
of the comparators. The actual time is calculated by looking at the main
counter of the RPi::Timer using the Timer::update_time function.
A stub for Scheduler::timer_tick is also added, since the TimeManagement
code now calls the function.
|
|
This makes the RPi::Timer suitable for use in the TimeManagement code.
|
|
We never caught this bug in the HPET, since it takes ages for a 64-bit
counter to wrap around. Also remove an unnecessary if check.
|
|
|
|
This should really go into a generic Processor class, but there is no
such class yet, so a FIXME is added for that.
|
|
The code in this file is not architecture specific, so it can be moved
to the base Kernel directory.
|
|
|
|
Instead of just having a giant KBuffer that is not resizeable easily, we
use multiple AnonymousVMObjects in one Vector to store them.
The idea is to not have to do giant memcpy or memset each time we need
to allocate or de-allocate memory for TmpFS inodes, but instead, we can
allocate only the desired block range when trying to write to it.
Therefore, it is also possible to have data holes in the inode content
in case of skipping an entire set of one data block or more when writing
to the inode content, thus, making memory usage much more efficient.
To ensure we don't run out of virtual memory range, don't allocate a
Region in advance to each TmpFSInode, but instead try to allocate a
Region on IO operation, and then use that Region to map the VMObjects
in IO loop.
|
|
|
|
This adds a new arch-independent header which in turn includes the
correct header for the build architecture.
|
|
|
|
This makes it easier to differentiate between cases where certain
functionality is not implemented vs. cases where a code location
should really be unreachable.
|
|
It costs us nothing, and some utilities (such as the known file utility)
rely on the exposed file size (after doing lstat on it), to show
anything useful besides saying the file is "empty".
|
|
|
|
In a few places we check `!Processor::in_critical()` to validate
that the current processor doesn't hold any kernel spinlocks.
Instead lets provide it a first class name for readability.
I'll also be adding more of these, so I would rather add more
usages of a nice API instead of this implicit/assumed logic.
|
|
This node is not used anymore, so let's remove it.
|
|
This currently fails with:
error: #pragma once in main file [-Werror]
|
|
This commit adds read-only support for the FAT32 filesystem. It also
includes support for long file names.
|
|
|
|
This change ensures that the scheduler doesn't depend on a platform
specific or arch-specific code when it initializes itself, but rather we
ensure that in compile-time we will generate the appropriate code to
find the correct arch-specific current time methods.
|
|
... instead of hard-coding it in the uname syscall.
|
|
This should equate to roughly the same code but improve typechecking.
|
|
This is done by generating a Kernel/Version.h header with major version,
minor version, and git hash.
|
|
|
|
This was necessary until a few months ago because of b0rked toolchain
options.
|
|
There are no guarantees that the final syscall API will look like this
but at least for now this lets us compile the userland binaries.
|
|
Also remove UserOrKernelBuffer::{read, write} and __stack_chk_fail from
Dummy.cpp and init.cpp respectively.
|
