Tested by inserting a call into the shell, but we can't leave it in because
every test ends up clobbering the disk. So it's now time to think about
a testable interface for the disk.
dd if=/dev/zero of=data.img count=20160
echo '(+ 1 1)' |dd of=data.img conv=notrunc
./translate ex2.mu && qemu-system-i386 -hda disk.img -hdb data.img
Compare output with `xxd data.img |head`.
This works because primary and secondary drives on the primary bus share
the same ports. All we have to do is select the right drive by flipping
a bit.
Both LBA and CHS coordinates are now working for the primary disk on the
primary bus.
Failure modes I ran into:
- ATA ports are 16-bit values. Using instructions with 8-bit immediates
will yield strange results. (I had to debug this twice because I missed
poll-ata-primary-bus-primary-drive-regular-status-word the first time
around.)
Mu's toolchain has been found out here. bootstrap has good
errors but doesn't support the instructions I need in boot.subx. The
self-hosted phases support the instructions but provide no error-checking.
Might be worth starting to add error-checking as I encounter the need.
In this case, a vote for validating metadata sizes even if we don't
validate that instructions pass in the right metadata sizes.
- Can't poll readiness first thing. Maybe we need to always select the
drive first.
- Reading 8-bit values from a 16-bit port (data port 0x1f0) returns garbage.
Reading 32-bit values however works totally fine; go figure. (Maybe
it won't work on real hardware?)
https://forum.osdev.org/viewtopic.php?t=36415
- Passing in a 0 segment will never return data.
Requires the following commands:
dd if=/dev/zero of=data.img count=20160
./translate life.mu
qemu-system-i386 -hda disk.img -hdb data.img
Before running Qemu, data.img will contain all 0s.
After quitting Qemu, data.img will contain the first 512 bytes of disk.img.
I'd been assuming that the image would be identical if it worked, but I
need to actually validate this at each step:
./translate life.mu && qemu-system-i386 disk.img && diff disk.img 0
The plan is to keep the binary identical until I finish translating boot.subx.
Then I can remove obsolete padding.
Just one instruction translated so far. The rest is treated as data.
Currently programs only work because the toolchain treats boot.subx as
all data. As we turn it into code we'll move the '== data' segment boundary
further down.
Kartik K. Agaram