maksw 4add6789ba add a realtime scheduler class and priority-inheritance mutexes with timeouts and abandonment, add an HDA audio driver and a SYS_AUDIO_* syscall family, flesh out kernel32 with Win32 threads, events, mutexes, critical sections, and condition variables, add SYS_RESUME_THREAD for suspended thread creation, add KeInitializeMemoryTypes to reprogram MTRRs and install a canonical PAT for WB/WC/UC control, rework the display pipeline with a pitched present path and console output flush, bring back disk.vhd with mtools since vvfat fucking sucks, harden AHCI with serialized cache flush, USB mass storage with SCSI sense-data decoding, and xHCI with cleaner DMA teardown, fix the last stray for(;;) in HalStopProcessor to while(true), pile on smoke tests for audio, win32 sync primitives, and thread suspend, implement AHCI timeout, kill readahead since it was desperetaly trying to murder me, fix doom sound x2.

2026-08-11 18:38:36 +02:00

..

vmwaresvga.c

add a realtime scheduler class and priority-inheritance mutexes with timeouts and abandonment, add an HDA audio driver and a SYS_AUDIO_* syscall family, flesh out kernel32 with Win32 threads, events, mutexes, critical sections, and condition variables, add SYS_RESUME_THREAD for suspended thread creation, add KeInitializeMemoryTypes to reprogram MTRRs and install a canonical PAT for WB/WC/UC control, rework the display pipeline with a pitched present path and console output flush, bring back disk.vhd with mtools since vvfat fucking sucks, harden AHCI with serialized cache flush, USB mass storage with SCSI sense-data decoding, and xHCI with cleaner DMA teardown, fix the last stray for(;;) in HalStopProcessor to while(true), pile on smoke tests for audio, win32 sync primitives, and thread suspend, implement AHCI timeout, kill readahead since it was desperetaly trying to murder me, fix doom sound x2.

2026-08-11 18:38:36 +02:00

vmwaresvga.h

introduce a stable NTL_DDK driver ABI to replace direct HAL_API and kernel-header access in drivers, add code/kernel/io_ddk.c and code/public/ntlite/ddk/ headers, change DriverEntry to take (const NTL_DDK*, NTL_DRIVER_INIT*), update kmain to build the DDK once via IoInitializeDdk/IoGetDdk and hand it to both boot and runtime drivers, downgrade code/public/ntlite/ddi/io_driver.h to a legacy I/O model header pulling from ntlite/ddk instead of kernel internals, move code/third_party/lwip into code/drivers/net/lwip/third_party, fix up lwip_sources.txt and build.c include paths to match, split build.c's single CFLAGS macro into COMMON/KERNEL/USER/BOOT/DRIVER cflags.

2026-07-01 23:19:11 +02:00