Bananymous
/
banan-os
1
1
Fork 1
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
2,603 Commits 1 Branch 0 Tags 14 MiB
Commit Graph

1496 Commits

Author SHA1 Message Date
Bananymous 267fdf9fa1 Kernel: Remove storage devices after unplugging them 2025-02-07 18:04:54 +02:00
Bananymous 26d2a4420e Kernel: Don't limit xHCI TRB transfer length to max packet size 
This made no sense and just used a ton of TRBs.

This patch also fixes some incorrectly set flags in data/status stage
TRBs.
 2025-02-06 23:18:14 +02:00
Bananymous c623ad7da0 Kernel: Use correct USB endpoint field to determine interval 
I was using the wrong endpoint descriptor field which caused intervals
to be messed up.
 2025-02-06 23:18:14 +02:00
Bananymous 7de689055c Kernel: Pass xHCI device information in structs 
This makes code more readable and extendable
 2025-02-06 23:18:14 +02:00
Bananymous 63b15a8855 Kernel: Rename USB initialize_endpoint -> configure_endpoint 
This makes more sense as the USB command is CONFIGURE_ENDPOINT

Also configure_endpoint can be called multiple times on the same
endpoint. There was no reason to limit this to only one call.
 2025-02-06 22:10:00 +02:00
Bananymous a2a7302964 Kernel: Make sure USB class driver is deinitialized before xhci device 2025-02-06 22:00:26 +02:00
Bananymous 6a5367dbe3 Kernel: Don't initialize USB alternate interfaces 2025-02-06 21:43:15 +02:00
Bananymous 6768a18475 Kernel: Don't parse xHCI custom slot_types 
xHCI spec mandates that slot_type for USB protocol is 0.
 2025-02-06 21:41:44 +02:00
Bananymous 242ed4a3c2 Kernel: Remove support for custom xHCI speed_ids 
This gets really weird with hubs and I don't think even linux handles
them.
 2025-02-06 21:38:30 +02:00
Bananymous f9b70d1b5b Kernel: Don't enter infinite loop on unexpected serial behaviour 2025-02-06 20:59:17 +02:00
Bananymous faa5252191 Kernel: Fix TTY ANSI CSI m and prevent crash :) 2025-02-06 20:58:03 +02:00
Bananymous b30a79c7fe Kernel: Temporary hack to "fix" munmap 
My current munmap implementation does not support partial unmaps and
GCC relies on this behaviour. This patch removes unmapping if the
address and len does not fully contain the mmap region.
 2025-02-01 22:25:22 +02:00
Bananymous d1c3d3d5aa Kernel: Fix ANSI CSI L 
With my new memcpy implementation this crashed. I have no idea how this
was not crashing before :D
 2025-01-28 18:41:53 +02:00
Bananymous 5938cc4086 Kernel: Support pselect sigmask 2025-01-28 17:17:59 +02:00
Bananymous 7c57d736c6 Kernel/LibC: Fix dirent functions 
dirent functions used to fail if a directory contained more than 128
files :D
 2025-01-28 17:15:11 +02:00
Bananymous cbe3f2a4ac BuildSystem: Cleanup and fix cmake files 
I now set library SONAME so i don't have to build them in build root.

Creating font now makes sure the install directory exists. This allows
building using make files.

LibC now links agains ligcc
 2025-01-28 17:11:21 +02:00
Bananymous 7a10e0e347 Kernel: Fix TmpFS inode linking 
TmpFS now looks for empty entries in directories instead of always
appending files to the end of directories. This makes kernel not crash
after process pid 126 is created :D
 2025-01-28 17:06:35 +02:00
Bananymous cbcb9f9044 Kernel: Warn on unsupport `clock_gettime` `clock_id`s 2025-01-25 17:31:17 +02:00
Bananymous 2a16a67aed Kernel: Ignore SA_RESTART flag on sigaction 
This is a big hack but a lot of ports set this flag. This seems really
annoying to implement properly
 2025-01-24 19:34:22 +02:00
Bananymous ee078fc638 Kernel: Fix fd status/descriptor flag handling 
I was not sharing status and offset on fork and I was sharing descriptor
flags on dup/dup2
 2025-01-17 20:12:04 +02:00
Bananymous 9893c90e74 Kernel: Remove SYS_DUP and implement it using fcntl F_DUPFD 2025-01-17 19:29:47 +02:00
Bananymous 82978da5e3 Kernel: Cleanup AML code 2025-01-17 16:20:00 +02:00
Bananymous 80ce9d546a Kernel: Don't error out on SYS_WAIT with W{CONTINUED,UNTRACED} 
As I don't support stopping and continuing processes these flags are
effecively no-ops
 2025-01-17 16:14:52 +02:00
Bananymous 34775633b2 Kernel: Clear DF on every interrupt handler 
I was missing this and it lead to UB with my new mem* functions that
were implemented with `rep`
 2025-01-17 16:11:13 +02:00
Bananymous 03b5c8e76e Kernel: Fix disk caching when `readonly` is specified 2025-01-16 16:16:38 +02:00
Bananymous 609067cefa Kernel: Cleanup internal CPUID functions declarations 2025-01-14 22:57:44 +02:00
Bananymous 9d497ad311 Kernel: Cleanup interrupt stub definitions 2025-01-14 22:55:06 +02:00
Bananymous ebecbb69ec LibC: Implement some mem* and str* functions in assembly 
This made them a lot faster on modern cpus with optimized rep stos and
rep movs
 2025-01-14 22:50:46 +02:00
Bananymous b129bab81a Kernel: Make mem* and str* functions weak symbols 2025-01-14 19:08:03 +02:00
Bananymous d189f00f38 Kernel: Null terminate tmpfs directory entries 2025-01-12 16:03:09 +02:00
Bananymous a5cf92b2ff Kernel: Use uint64_t instead of size_t for AML node conversion 
This allows 32 bit to compile again :)
 2024-12-30 06:27:49 +02:00
Bananymous 69137cddab Kernel: Implement TIOCSWINSZ for pseudo terminals 
I have no idea how I had forgotten this
 2024-12-21 03:22:48 +02:00
Bananymous d7b4f54983 Kernel: Check Control Method Battery _STA bit 4 
This bit indicates wheter the battery is actually present.
 2024-12-19 02:35:09 +02:00
Bananymous 00b0dcd306 Kernel: Reorder ACPI initialization method calls 2024-12-19 02:35:09 +02:00
Bananymous 53c356a940 Kernel: Implement AML MatchOp 2024-12-18 19:26:19 +02:00
Bananymous 3a6cdfff45 Kernel: Fix ACPI namespace initialization 
Now _REG, _STA, _INI are called in the order my laptop expects them to
be called. This was kinda weird because what uACPI was doing did not
work.

\_SB_.PCI0.LPC0.EC0_.BAT0._STA required \_SB_.PCI0.LPC0.EC0_._REG to be
called
\_SB_.PCI0.LPC0.EC0_._REG required \_SB_.PCI0._STA to be called

Now I call all the _REG methods of a device after calling _STA/_INI and
after performing the whole _STA/_INI sequence i call rest of missing
_REG functions
 2024-12-18 18:30:45 +02:00
Bananymous c26e347e91 Kernel: Fix AML FieldUnit conversion and NameString construction 2024-12-18 18:29:04 +02:00
Bananymous 1b94957b07 Kernel: Copy AML NameString base when moving it :) 2024-12-18 18:28:16 +02:00
Bananymous 4e364bd2f6 Kernel: Add support for ACPI Control Method Batteries 
The implementation is kinda weird but it exposes some battery
information to userspace!
 2024-12-18 14:06:32 +02:00
Bananymous b9fe564d78 Kernel: Check for GCC when ignoring Wmaybe-uninitialized 2024-12-18 01:45:36 +02:00
Bananymous 437fa45ca2 Kernel: Add helper to find ACPI devices with EISA id 2024-12-18 01:45:00 +02:00
Bananymous 85e215650c Kernel: Implement AML ToStringOp 2024-12-18 01:44:33 +02:00
Bananymous 7eb186dad4 Kernel: Add to_sv() helper for AML String 2024-12-18 01:43:46 +02:00
Bananymous 2259614640 Kernel: Remove unused function from ACPI namespace 2024-12-18 01:43:04 +02:00
Bananymous 869f4011a1 Kernel: Replace the old AML interpreter by a new, better one 
The old AML interpreter was trash and did not follow value/reference
semantics at all. It was also super slow, one of my machines taking over
7 seconds to parse ACPI namespace and call _INI and _STA.
 2024-12-17 03:36:33 +02:00
Bananymous 46d4e32510 Kenrel: Validate inode open mode on read/write syscalls 2024-12-15 20:49:12 +02:00
Bananymous 1815a913c3 Kernel: Remove unnecessary hash map lookups 2024-12-15 20:47:40 +02:00
Bananymous 12a37500b0 Kernel: Fix triple-fault in vitual tty 
This was happening when printing non-utf8 data while having virtual tty
as the debug console.
 2024-12-09 03:35:51 +02:00
Bananymous 2d0690ae2d Kernel: Cleanup most of syscalls dealing with files 2024-12-07 05:33:04 +02:00
Bananymous 71a2d0971f Kernel: Remove unnecessary reverse from VFS file search 2024-12-07 01:41:23 +02:00
Bananymous 3bf57e0748 Kernel: Add . and .. entries to /proc/<pid> directories 
This allows to run `cd ..` in one of these directories
 2024-12-07 01:37:40 +02:00
Bananymous 51b6329c86 Kernel: Make backspace \b and delete 0x7F 
This is what `vim` seems to expect and imo makes way more sense.
 2024-12-05 17:34:37 +02:00
Bananymous 72d9e4c1e7 Kernel: Remove accidental debug prints from fcntl syscall 2024-12-05 17:34:32 +02:00
Bananymous 49133dce48 Kernel: Make pipe check data every 100ms 
This prevents dead lock if read-end does not detect data before blocking
why write-end already notified.
 2024-12-03 21:54:27 +02:00
Bananymous 0a5aacfd87 Kernel: Allow open with O_CREAT bypass file permission checks 
call to open(path, O_CREAT|O_EXCL|O_RDWR, 0444) should open file as
read-write although file is created as read-only on filesystem.
 2024-12-03 16:17:34 +02:00
Bananymous 2dec3a6c95 Kernel: Add some sanity checks for existanse of threads 2024-12-03 16:16:50 +02:00
Bananymous fcc2814199 Kernel: Remove process from alarm list after it dies 2024-12-03 16:16:08 +02:00
Bananymous 0fab7ad63b Kernel: Fix SYS_FCNTL command handling :) 
I had misunderstood how these work
 2024-12-03 16:12:26 +02:00
Bananymous e6a2f55a59 Kernel: Closing file backed mmap wont expand the file size 2024-12-03 16:12:26 +02:00
Bananymous 67e9ca56ac Kernel: Create /dev/urandom -> /dev/random symlink 
Most software uses urandom instead of random so this allows it work
normally.
 2024-12-03 16:12:26 +02:00
Bananymous 713daf6cd3 Kernel/LibC: Add support for creating hardlinks 2024-12-03 16:12:26 +02:00
Bananymous 415b20f884 LibC: Remove errno EEXISTS 
I don't know why I had added this. It doesn't seem to be a thing on
either POSIX or linux. This was literally equivalent to EXIST
 2024-12-02 20:13:38 +02:00
Bananymous d58ca5f37a Kernel/LibC: Implement symlink{,at} 2024-12-02 20:13:38 +02:00
Bananymous a10ca47657 Kernel/LibC: Implement {,f}statvfs 2024-12-02 20:13:37 +02:00
Bananymous ad1f175a39 Kernel/LibC: Implement getppid 2024-12-02 20:13:37 +02:00
Bananymous 6ed1435aeb Kernel/LibC: Implement tcgetpgrp 2024-12-02 20:13:37 +02:00
Bananymous 6346e288ad LibC: Implement getrlimit 2024-12-02 20:13:37 +02:00
Bananymous 747c3b2a4b Kernel/LibC: Implement fsync 2024-12-02 20:13:37 +02:00
Bananymous c02de2580d Kernel: Add support for SIOCGIFNAME ioctl 2024-12-02 20:13:37 +02:00
Bananymous 4149748766 Kernel: Add missing include in ACPI headers 2024-12-01 21:32:34 +02:00
Bananymous 8dbbbc1a1a Kernel: Add command line option `readonly` 
This prevents calls to write_sectors_impl and all dirty pages are always
kept in RAM.
 2024-11-27 13:31:55 +02:00
Bananymous a6fc3cf7a6 Kernel: Clear AHCI pending interrupts 
I have no idea why I did not have this in. All devices were just
constatly spamming interrupts like crazy :D
 2024-11-26 01:39:11 +02:00
Bananymous 1941885cfd Kernel: Disable USB Mass Storage writing until I get it fixed 
This prevented real hw from running for over 10-20 seconds. When first
disk sync happens after 10 secods, USB storage would enter faulty state
and not allow any io operations.

I have no idea what the problem is. If I don't send Reset command to the
USB device writing seems to work fine. I also don't think its part of
the Reset process either as reading works just fine.
 2024-11-26 01:01:25 +02:00
Bananymous 88a86a9927 Kernel: Fix Thread destruction after sys_exit 
This had undefined behaviour as Thread's (Processes's) PageTable was
destroyed before Thread had the change to destroy its own stacks that
lived on the PageTable.
 2024-11-26 00:59:34 +02:00
Bananymous 48e030bca3 Kernel: Make PageTable non-copyable and non-movable 
Also PageTable destructor now verifies that is has allocated something
instead of assuming paddr of 0.
 2024-11-26 00:58:35 +02:00
Bananymous 793c0368f2 Kernel: Refactor USB mass storage code 
Also increment command timeout to 10 seconds so commands don't timeout
when they are not supposted to :)
 2024-11-23 01:24:32 +02:00
Bananymous 076001462e Kernel: Implement AML ConcatOp 2024-11-22 22:23:07 +02:00
Bananymous d2e1d8ec13 Kernel: Implement AML NotOp 2024-11-22 22:22:44 +02:00
Bananymous 30ceaa6bef Kernel: Implement AML ToHexStringOp 2024-11-22 22:22:27 +02:00
Bananymous 0247d47a3d Kernel: Make USBMassStorage send_scsi_command templated 
This allows passing ConstByteSpan when data will not be modified
 2024-11-22 22:21:19 +02:00
Bananymous 480368c878 Kernel: Cleanup and fix xHCI interrupt handling 
Always clear EventHandlerBusy bit and remove unnecessary nesting
 2024-11-22 22:19:53 +02:00
Bananymous c1d8790623 Kernel: Don't allow sending commands to halted xHCI 2024-11-22 22:19:29 +02:00
Bananymous fb1ffec5bb Kernel: Remove unused mutex from xHCI 
This was preventing multiple xHCI controllers to do parallel device
initialization
 2024-11-22 22:18:38 +02:00
Bananymous 1bdabe693f Kernel: Preserve RsvdP bits in xHCI controller initialization 2024-11-22 22:18:09 +02:00
Bananymous ce09f35275 Kernel: Fix PCI 64 bit MSI msg_data offset 2024-11-22 22:17:25 +02:00
Bananymous 40c13043b3 Kernel: Add support for scratchpad buffers in xHCI 2024-11-22 22:16:57 +02:00
Bananymous e620068416 Kernel: Allow USB devices to handle STALL conditions 2024-11-22 22:15:22 +02:00
Bananymous 8e624ca85a Kernel: Don't panic if uninitialized AHCI device gets an interrupt 2024-11-21 20:41:13 +02:00
Bananymous c7afd46016 Kernel: Don't log spurious interrupts 
Some PCs seem to be spamming a lot of int 7
 2024-11-21 20:40:29 +02:00
Bananymous 3573656244 Kernel: Make PARTUUID matching case insensitive 
Case sensitivity made it super annoying to work with :)
 2024-11-21 20:39:27 +02:00
Bananymous 1de6de975a Kernel: Rename cmdline UUID= to PARTUUID= 
This makes more sense and behaves more like existing software.
 2024-11-21 20:38:50 +02:00
Bananymous 865061b492 Kernel: Temporary fix to make tmpfs more stable 
whole TmpFS will have to be rewritten at some point :)
 2024-11-21 18:13:26 +02:00
Bananymous 70880636f4 Kernel: Implement basic USB Mass Storage support 
I finally decided to do this :D
 2024-11-21 18:08:37 +02:00
Bananymous 8054f6c618 Kernel: Wait 10 seconds to find root device 
If root device is initialized asynchronously it may not be ready when
main initialization thread is finished. This patch adds searching for
root device every 500 ms for 10 seconds.
 2024-11-21 17:56:58 +02:00
Bananymous 658ea68d04 Kernel: Check max packet size for all USB devices 
Spec says that this has to be done. Most cases can be probably skipped
as I used to do, but this is less error prone and adds practically no
overhead
 2024-11-21 13:46:42 +02:00
Bananymous 5750f87396 Kernel: Fix unaligned access in xHCI controller initialization 2024-11-21 13:46:09 +02:00
Bananymous 1253e2a458 Kernel: Add support for bulk endpoints and update endpoint API 
USB device now sets its own data buffers for IN/OUT endpoints. This
allows more customization and parallelism as data buffer does not have
to be shared.
 2024-11-21 13:44:21 +02:00
Bananymous 857b3e92f8 Kernel: Support initializing USB 3 devices 2024-11-21 13:40:16 +02:00
Bananymous 8bf14d542e Kernel: Move SCSI device rdev allocation out of ATA code 2024-11-21 13:36:59 +02:00
Bananymous 97718b4046 Kernel: Don't launch init process twice 
I had accidentrally left the old init process launch in when replacing
it with the new one.
 2024-11-19 20:43:38 +02:00