fs.mitm: add and use memlet module to temporarily steal applet memory while building romfs images. ...

Starting in 20.0.0, the browser needs more applet memory to function, so we can't steal as much any more.
Thus, we now steal 14 MB on 20.0.0+ instead of 40MB.

However, since this reduces memory available for custom system modules, we are adjusting to compensate.
ams.mitm's heap size has been reduced from 32MB to 12MB (recovering 20MB).
In addition, fs.mitm now uses a new mechanism for stealing memory from the applet pool while romfs is being built.

On net, we are compromising:
    * Custom sysmodules lose memory available to them.
        On 19.0.0/AMS 1.8.0, there was 30 MB available for custom sysmodules.
        Stealing 14 MB instead of 40 MB, we lose 26 MB of that. Reducing ams.mitm's usage will gain us back 20.
        Nintendo also appears to...use 4 extra MB, in 20.0.0, from my test homebrew.
        So on 20.0.0/AMS 1.9.0, there should be 20 MB available for custom sysmodules.
        On the bright side, on <20.0.0/AMS 1.9.0, I guess there will be 50 MB available for custom sysmodules now?
    * totk mods will lose the ability to...put every file in the romfs on sd card. There will be some unknown maximum filecount for totk mods.
        On the bright side, implementing the transient memory stealing should improve compatibility for some mods which strictly add files?

kern/ldr: add support for --x executables

kern/os: support CreateProcessFlag_EnableAliasRegionExtraSize

kern/svc: fix query/insecure names

os: add UnsafeMemory apis

os: update os namespace for 15.0.0, loader/ro: update to use csrng, spl: bump max sessions

os: pthread_yield is deprecated, don't use it on linux

os: silence warning building with gcc 12 on windows

os: fix whoops in SetMemoryPermission on windows

os: implement MemoryHeapManager, SetMemoryAttribute

os: fix various regressions since 1.3.1

os: fix minor bug (and simplify) MapProcessCodeMemory

ro/os: use os primitives for MapProcessCodeMemory

loader: use os apis for interacting with process memory

os: silence a maybe-uninit warning

os: refactor multi wait apis to better match Nintendo's latest implementation

os: broadcast, not signal, on release more than 1 sema

os: add DetachIoRegionHandle

ams: globally prefer R_RETURN to return for ams::Result

ams: use R_SUCCEED, R_THROW globally

fs: for my sanity, begin tracking version where code was last checked for accuracy

os/dd: add missing semicolon to new cache logic

kern: update for new hw maintenance semantics

kern: add InfoType_IsSvcPermitted

strat: only include windows.h in translation units where actually needed

fs: fix clang-build, os: StackGuardManager

os: add working stack logic for macOS

fix stack getting on linux, add test

abort/error: print backtrace, abuse templates, overhaul result/diag (macos not done yet)

add basic tests for os::Event/SystemEvent functionality