Commit Graph

147 Commits

Author SHA1 Message Date
Bananymous d4ea720239 Kernel: Don't crash the kernel if ext2 encounters disk error
This will most likely result in a corrupted filesystem, but crashing the
kernel is too much :D
2024-09-17 15:54:33 +03:00
Bananymous 7feb4c4ebd Kernel: VFS::file_from_absolute_path now used root_file() API 2024-09-15 02:38:25 +03:00
Bananymous 8956835d95 Kernel: Make sys_.*at use VFS relative path finding 2024-09-14 21:15:58 +03:00
Bananymous 3a352078de Kernel: Make VirtualFileSystem::File non copyable 2024-09-14 20:09:28 +03:00
Bananymous 6060b39548 Kernel: Implement relative file searching in VFS 2024-09-14 19:44:20 +03:00
Bananymous 2b52ea4c6f Kernel: Make Inode::can_access const 2024-09-14 19:39:41 +03:00
Bananymous 467ac6c365 Kernel/LibC: Implement SOCK_CLOEXEC and SOCK_NONBLOCK
This removes the need for fcntl after creating a socket :)
2024-09-11 21:59:11 +03:00
Bananymous c77ad5fb34 Kernel: Implement copy-on-write memory for file backed `mmap`s 2024-09-11 19:33:50 +03:00
Bananymous bac3219a01 Kernel: Fix `Pipe::can_read_impl()`
The logic was inversed, which made all select calls report incorrectly
for pipes. This made terminal emulator just freeze.
2024-08-10 18:18:25 +03:00
Bananymous b6c964c444 Kernel: Rewrite pipes
Pipes have now a fixed size buffer and pipe clone and close is working
again.
2024-08-09 16:50:19 +03:00
Bananymous 44f0ec601f Kernel: Expose /proc/meminfo and fix /proc/<pid>/meminfo
Thread was unconditionally calling <stack>->size() without validating
that the stack actually exists
2024-08-09 15:58:56 +03:00
Bananymous f8261c60c0 Kernel: Rewrite the whole scheduler and re-architecture SMP handling
Change Semaphore -> ThreadBlocker
  This was not a semaphore, I just named it one because I didn't know
  what semaphore was. I have meant to change this sooner, but it was in
  no way urgent :D

Implement SMP events. Processors can now be sent SMP events through
IPIs. SMP events can be sent either to a single processor or broadcasted
to every processor.

PageTable::{map_page,map_range,unmap_page,unmap_range}() now send SMP
event to invalidate TLB caches for the changed pages.

Scheduler no longer uses a global run queue. Each processor has its own
scheduler that keeps track of the load on the processor. Once every
second schedulers do load balancing. Schedulers have no access to other
processors' schedulers, they just see approximate loads. If scheduler
decides that it has too much load, it will send a thread to another
processor through a SMP event.

Schedulers are currently run using the timer interrupt on BSB. This
should be not the case, and each processor should use its LAPIC timer
for interrupts. There is no reason to broadcast SMP event to all
processors when BSB gets timer interrupt.

Old scheduler only achieved 20% idle load on qemu. That was probably a
very inefficient implementation. This new scheduler seems to average
around 1% idle load. This is much closer to what I would expect. On my
own laptop idle load seems to be only around 0.5% on each processor.
2024-07-22 00:33:50 +03:00
Bananymous 9d7f97ccd5 Kernel: Implement DevFileSystem::remove_device
This function cleanly removes the devices from the whole filesystem.

USB devices are now removed from the filesystem as soon as they are
destroyed.
2024-07-15 22:10:42 +03:00
Bananymous a97a574718 Kernel: Rewrite the whole input system
PS/2 code is now kind of messed up, but it works. Keyboards and mice are
now an abstract class that is automatically exposed to userspace. This
will make adding USB input much nicer.
2024-07-14 01:53:50 +03:00
Bananymous 31568fc5a1 Kernel: Rewrite Sockets to not be TmpInodes
TmpInodes just caused issues because TmpFS kept them alive. There was
really no reason for sockets to even be stored inside a TmpFS...
2024-06-27 00:35:19 +03:00
Bananymous bce16cdd6e Kernel: Fix how socket closing works
Sockets are now closed only when they are not referenced any more. This
allows child process to close socket and still keep it open for the
parent.
2024-06-19 10:39:44 +03:00
Bananymous 318ce5dec8 All: Fix a lot of compiler warnings from header files
While reworking build system, header files started to report warnings.
2024-06-18 23:02:10 +03:00
Bananymous cad55a4da5 Kernel/LibC: Implement getsockname for ipv4 sockets 2024-06-17 20:54:45 +03:00
Bananymous ea7fc7f6c4 Kernel: Implement read-only FAT12/16/32 driver with long name support
You can now mount FAT filesystems! This code might not work perfectly
but my quick testing seemed to work on all (FAT12/16/32) variants.
2024-06-14 01:04:12 +03:00
Bananymous 766439db6d Kernel: Start work on adding support for new filesystems
Old code tried to create ext2 filesystem from all devices.
2024-06-11 10:50:26 +03:00
Bananymous 0af74fccda Kernel/LibC: Rework dirent structure
dirent now contains statically sized d_name. This allows using
sizeof on the name and dirent properly, which some programs seem
to be using.
2024-05-22 20:19:59 +03:00
Bananymous 4d1f0e77f2 Kernel: Fix physical address size for x86_32
Having 32 bit address space does not mean physical address space
is also only 32 bits...
2024-03-26 20:16:20 +02:00
Bananymous 99e30a4d7d Kernel: Replace i386 with i686
I don't really want to be working with i386 since it doesn't support
compare exchange instruction
2024-03-26 02:48:26 +02:00
Bananymous 2352c86048 Kernel: i386 has 14 indirect blocks in TmpInode instead of 2
This allows keeping size of TmpInodeInfo as 128
2024-03-22 14:01:27 +02:00
Bananymous d94f6388b7 Kernel: Fix all broken locks from new mutexes 2024-02-28 22:45:34 +02:00
Bananymous 9594ee8e47 Kernel: Start making device numbers unique for each device 2024-02-22 15:53:48 +02:00
Bananymous f73e954b28 Kernel: Remove SpinLock from Pipe
Pipe already is using lock on the inode. If you read from pipe when
there was no data, pipe blocked indefinately since writes were blocked
by Inode::m_lock.
2024-02-14 17:21:32 +02:00
Bananymous 3fc1edede0 Kernel/LibC: Implement super basic select
This does not really even block but it works... :D
2024-02-12 17:26:33 +02:00
Bananymous ff49d8b84f Kernel: Cleanup OSI layer overlapping 2024-02-09 17:05:07 +02:00
Bananymous 49889858fa Kernel: Allow chmod on TmpSocketInode 2024-02-08 03:16:01 +02:00
Bananymous e7dd03e551 Kernel: Implement basic connection-mode unix domain sockets 2024-02-08 02:28:19 +02:00
Bananymous 132286895f Kernel: Implement Socket inodes for tmpfs 2024-02-07 15:57:45 +02:00
Bananymous 41cad88d6e Kernel/LibC: Implement dummy syscalls for accept, connect, listen 2024-02-07 15:57:45 +02:00
Bananymous e1ffbb710b Kernel/LibC: Implement basic ioctl for network addresses 2024-02-03 01:50:10 +02:00
Bananymous 0f154c3173 Kernel: Implement basic recvfrom 2024-02-02 13:50:00 +02:00
Bananymous ec2f21bb9f Kernel/LibC: Implement SYS_SENDTO 2024-02-02 03:16:01 +02:00
Bananymous ab150b458a Kernel/LibC: Implement basic socket binding 2024-02-02 01:31:58 +02:00
Bananymous dfe5a2d665 All: Cleanup all files
Add newline to end of files and remove whitespace from end of lines
2024-01-24 15:53:38 +02:00
Bananymous 2cee2a85e6 Kernel: DevFS now stores all devices
This allows removing hack in PCI that was required to keep NVMe
controller alive.
2024-01-24 14:33:50 +02:00
Bananymous 96d831c31a Kernel/LibC/Userspace: Implement chown and set tty owner on login 2024-01-02 22:19:38 +02:00
Bananymous 12474addda Kernel: Make Inodes use the new lock
Also remove old lock from TTY since it can just use the one Inode
already has.
2023-12-07 13:19:12 +02:00
Bananymous bc5e8add19 Kernel: Make Ext2 filesystem use BlockDevice instead of Partition 2023-11-21 15:20:24 +02:00
Bananymous 7a8fd6d04a Kernel: TmpFS doesn't mark any functions as final
I didn't think these would be overloaded, but they are
2023-11-21 15:19:34 +02:00
Bananymous b749963b62 Kernel: Add common {read,write}_blocks() api to BlockDevice 2023-11-21 15:19:07 +02:00
Bananymous cadb56d8ba Kernel: ProcFS inodes reflect processes ruid/rgid
setgid/setuid did not change the permissions of procfs inodes. This
made Shell launched by init not appear in meminfo.
2023-11-18 14:26:44 +02:00
Bananymous fda0dfec30 Kernel: Make TmpFS enforce max page count. 2023-11-07 16:13:21 +02:00
Bananymous d6ae1bcf36 Kernel: Remove now obsolete RamFS
Everything is using now the better TmpFS which uses physical pages
for page allocation instead of the static kmalloc memory.
2023-11-07 16:07:11 +02:00
Bananymous 5044810451 Kernel: Make DevFS use the new and better TmpFS instead of RamFS 2023-11-07 16:05:05 +02:00
Bananymous 147cd93ed3 Kernel: Add method to TmpFS for looping over all (cached) inodes 2023-11-07 16:04:34 +02:00
Bananymous 07b5920f3f Kernel: Lock TmpFS in all its methods 2023-11-07 16:03:52 +02:00