| Filename | Latest commit message | Latest commit date |
|---|---|---|
3b26575b4a
Three things, and the first one is the freeze.
**The loader jumped out from under itself.** A machine with a lot of RAM
stopped at "Starting the kernel" with the splash still up: no panic, no reset,
nothing. The firmware's own exception handler had the answer the whole time,
once a machine was found that printed it:
X64 Exception Type - 0E(#PF - Page-Fault)
ExceptionData - 0x10 I:1 P:0
RIP - 0000000140004DEE <- the loader's own code, at 5.4 GB
RAX - 0000000000001000 <- the page tables it had just installed
RCX - 00000000001000F0 <- the kernel it was about to jump to
I:1 P:0 is an instruction fetch from a page that is not present. The loader
built page tables covering four gigabytes, loaded cr3 -- and the firmware had
placed the loader image *above* four gigabytes, so the next instruction fetch
was unmapped. It reproduces exactly on `-m 8G` and never on `-m 1G`, which is
why every test here passed.
The PDPT is one page whether four entries are filled or all 512, so all 512
are: 512 GB of identity map, replaced by kvmalloc() a few hundred instructions
later.
**And 3 GB was never the limit, it was a circle.** PHYSTOP_MAX was 0xC0000000
because kvmalloc() mapped exactly four 1 GB pages, and kvmalloc mapped four
because that was the cap -- each holding the other up, throwing away thirteen
gigabytes of a sixteen gigabyte machine. kvmalloc now maps as many gigabytes as
the machine has RAM and tells ioremap where the direct map stopped, instead of
ioremap assuming it stopped at four.
Freeing memory becomes two passes, and the ordering is forced rather than
chosen: kinit1() runs *before* kvmalloc because kvmalloc calls kalloc for its
own page tables, and freeing a page means writing into it -- so kinit1 can only
touch what the loader mapped. kinit_high() adds the rest afterwards.
`phystop` and the new `lowtop` are now two different numbers: the DMA arena
stays below 4 GB because a device is not obliged to reach the top of memory,
and an RTL8139's buffer address register is thirty-two bits wide.
**And the 0 MB.** The System Information screen reported "0 MB usable" on a
machine with sixteen gigabytes in it, because memory_mb() had its own 16 KB
guess at how big a memory map gets, and returned zero when GetMemoryMap said
EFI_BUFFER_TOO_SMALL. That is the same bug as the previous commit's, in a
second copy, and this one failed silently. There is one buffer and one
map_read() now.
Verified at 1 GB, 8 GB and 16 GB: all three boot, no traps, no panics, socktest
and memfdtest green in each. 8 GB and 16 GB did not boot at all before -- and
16 GB now reports 17408 MB with "mem: 13312 MB above 4 GB added to the
allocator", where the old kernel would have used three. Both display
configurations still clean.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
|
||
| .. | ||
| efi.h | ||
| main.c | ||
| README.md | ||
| startup.nsh | ||
UEFI loader
The loader. This directory once existed only so the BIOS path could be moved
into boot/bios/ without pretending the split was hypothetical; it is now the
only path there is.
See ../README.md for the hand-off contract a loader has to satisfy. The
short version: the kernel wants long mode with paging on, its ELF segments
placed, a 24-byte-entry E820-shaped memory map, and framebuffer geometry — and
it does not care which firmware produced them.
What this will need, in rough order:
- A PE/COFF target. clang can emit one directly
(
--target=x86_64-unknown-windows -fuse-ld=lld-link), so no new toolchain: the same LLVM install already builds every other target here. - The handful of UEFI structures actually used — the system table, boot
services,
LOADED_IMAGE,SIMPLE_FILE_SYSTEM,GRAPHICS_OUTPUT. Written from the specification rather than by importing gnu-efi or EDK II, which is both a licence question and a large dependency for something this small. GetMemoryMaptranslated into the same 24-byte entriese820.calready parses. Not a second format for the kernel to learn.ExitBootServicesbefore touching the memory it described.- Graphics Output Protocol for the framebuffer, filling the same
VBE_INFO_ADDRstructure the kernel reads.
The image goes at EFI/BOOT/BOOTX64.EFI on a FAT ESP, which also means the
ISO carries a single El Torito EFI entry.