Commit by sir_richard :: r45526 reactos/boot/ (16 files in 8 dirs):

[FREELDR]: Use WinLDR-style booting exclusively for ARM. Remove thousands of lines of old code where we basically did the same thing.
[FREELDR]: Build WinLDR for ARM too, and implement the required portability functions to get it working. Don't build "ros-style" loading on ARM.
[FREELDR]: Redo from scratch the entire paging setup for ARM. Instead of using 20MB of physical RAM, we now only use 1MB that's passed on to the kernel.
[FREELDR]: Fix loads of bugs in the ARM loader code (ha-ha!).
[FREELDR]: Cleanup compiler and linker flags in the rbuild files.
[ARMLLB]: Remove all deprecated video machine routines that are not used anymore.
If all went well, the ARM port should now boot on Versatile platforms up until the ARM3 initialization code.

Commit by sir_richard :: r45524 reactos/boot/freeldr/freeldr/ (10 files in 2 dirs):

[FREELDR]: Remove C++ style indirected "UI" component from FreeLDR ARM, for several reasons that ought to be considered on x86 too:
    - It redraws the screen in two different ways (zero memory and character-by-character fill) over 15 times for a single boot. You notice these things on a little ARM chip!
    - It keeps coping data from/to an "off-screen buffer" way too much, instead of only when the video needs to be synced.
    - It makes assumptions about all Video Buffers supporting Character-Mode VGA.
    - It can be configured to draw cyan/yellow ponies, instead of a normal NT boot menu.
    The ARM build now uses a "directui.c" implementation which fixes all these issues and is efficient and to the point. It presents a normal Windows boot selection menu on ARM.
[FREELDR]: Do not build any of the video routines, or the UI routines, on ARM.

Commit by sir_richard :: r45523 reactos/boot/freeldr/freeldr/ (4 files in 3 dirs):

[FREELDR]: No Linux support for ARM, no RS232 support for ARM, and no complex debug infrastructure for ARM either.

Commit by tkreuzer :: r45522 reactos/lib/sdk/crt/libcntpr.rbuild:

add setjmp/longjmp to libcntpr, just like in ms sdks

Commit by sir_richard :: r45521 reactos/boot/freeldr/freeldr/disk/ (disk.c partition.c):

[FREELDR]: Do not build Disk/Partition support for ARM, we use ram disks.

Commit by sir_richard :: r45520 reactos/boot/freeldr/freeldr/fs/ (ext2.c fs.c iso.c ntfs.c):

[FREELDR]: Do not build EXT2, ISO or NTFS support for ARM targets.

Commit by sir_richard :: r45519 reactos/boot/freeldr/freeldr/windows/winldr.c:

[FREELDR]: The WinLDR code forgot to update the File Path of boot driver entries from PA to VA as well, not just the registry path. It also tried to touch the PA LoaderBlock right after enabling the MMU. It should touch the VA LoaderBlock instead, since there is no guarantee that the address has been identity mapped (and hence the PA address interpreted as a VA address by the MMU is bogus).

Commit by evb :: r45518 reactos/hal/ (32 files in 6 dirs):

• New ARM HAL after 2 nights in the company of a madman.
• Many fixes, more implemented routines, support for ARMv7.
• Some Linux code, marked as GPL, from ARM Ltd in cache.s. All OS use this code.

Commit by sir_richard :: r45517 reactos/boot/freeldr/freeldr/mm/meminit.c:

[FREELDR]: The hack to set pages from 0x100 to 0x1FF as busy so that the heap allocator wouldn't use them always restored the pages back to Free (since they usually were free on x86). However, if the pages were already being used in the first place, this made them appear free and corrupted existing memory. Fixed the hack so that it restores the previous state of the memory pages.

Commit by sir_richard :: r45516 reactos/include/ndk/arm/mmtypes.h:

[NDK]: Use new ARMv6 PTE structures.

Commit by sir_richard :: r45515 reactos/ntoskrnl/ (mm/arm/page.c ntoskrnl-generic.rbuild):

[NTOS]: Tabula rasa (clean slate) on the old-MM page routines for ARM. Get rid of all of them for now.
[NTOS]: Implement MmInitGlobalKernelPageDirectory as the first ARM page function in the new file.
That's it for the kernel.

Commit by sir_richard :: r45514 reactos/ntoskrnl/include/ (3 files in 2 dirs):

[NTOS]: All calls to DbgPrint will go to DbgPrintEarly for now.
[NTOS]: More header fixes to match new definitions.

Commit by sir_richard :: r45513 reactos/ntoskrnl/ke/arm/kiinit.c:

[NTOS]: Make system startup match x86 code as closely as possible, instead of mixing MIPS/PPC code. Only rip out parts that are obviously x86-specific. Also fix code to use new KPCR definitions.
[NTOS]: Implement a DbgPrintEarly function that directly uses KDCOM to print to the serial console, instead of going through the Debug Service Interrupt.

Commit by sir_richard :: r45512 reactos/ntoskrnl/ke/arm/ (trapc.c usercall.c):

[NTOS]: Update code to use new KPCR structure.
[NTOS]: Make all trap code currently spin forever, because it has not been reviewed yet.

Commit by sir_richard :: r45511 reactos/include/reactos/arm/armddk.h:

[ARMDDK]: Use the same KPCR structure on ARM as on x86.
[ARMDDK]: Use the same addresses for KPCR and KUSER_SHARED_DATA on ARM as on x86. On ARM however, since there are no "segments", the KPCR is also mapped in user-mode, read-only, to have a static structure that can point to the current TEB.

Commit by sir_richard :: r45510 reactos/include/ndk/arm/ketypes.h:

[ARM DDK]: Use the closest possible approximation to the x86 KPCR/KPRCB (in fact, the public KPCR is identical, just the KIPCR is different, internally), instead of a mashup of PPC and MIPS structures.

Commit by sir_richard :: r45509 reactos/ntoskrnl/ke/arm/exp.c:

[NTOS]: Sync up KiDispatchException with x86 code (and going back to C comment formatting).

Commit by sir_richard :: r45508 reactos/ntoskrnl/ke/arm/cpu.c:

[NTOS]: Get rid of fixed TLB entries on ARM, I don't think that idea makes sense (unlike IA64 or MIPS where they are lot more flexible), since they only allow 1MB sections, not pages.

Commit by sir_richard :: r45507 reactos/ntoskrnl/ke/arm/boot.s:

[NTOS]: Forgot to setup an IRQ stack, the kernel was working by magic since in previous incarnations of the ARM port, 0xFFFFyyyy was mapped, so a stack of 0 worked (going downwards).
[NTOS]: Now FreeLDR just calls the KiSystemStartup routine correctly, without any "magic", so fix the code to use a1 instead of a2 (argument 1/2).

Commit by sir_richard :: r45506 reactos/include/reactos/arm/peripherals/ (pl011.h pl190.h sp804.h):

[ARM SDK]: Use real Versatile base addresses instead of the old hacked FreeLDR base addresses.