Massively simplified, by not trying to shoehorn differently-behaving
cases into the same one loop!
This may have introduced bugs (the test suite has caught three different
oversights, two of which via external projects, in fact!), but this also
makes the logic clearer and more streamlined, so that we are also less
likely to have any latent or future bugs.
In particular, previously, the first iteration of the loop could attempt
placement at an address not matching the section's constraints,
which made advancing to the next target address unnecessarily complicated
(https://github.com/gbdev/rgbds/pull/2064#discussion_r3985466596),
among other weirdness. The code ended up being defensive, and thus the
overall logic was murky.
I'm also expecting that this should provide a performance improvement due
to being essentially a form of loop-invariant code motion (and very likely
one a compiler couldn't have performed automatically), though I haven't
measured.
Fixes#2095
This code pre-dates the introduction of alignment offsets,
so it was never updated to take it into account.
That said, some of the code that I have factored out is a little awkward because
it will be used in an upcoming refactoring of the whole assignment loop,
since that turns out to be more byzantine than it ought to be due to
trying to handle a too disparate set of cases in a single loop.
Turns out the checks could fail spuriously if the starting alignment is non-zero,
e.g. `align 8,$C0` in HRAM. (Fixes #2113.)
This was tripped up by a test added to exercise the linker's section placement algorithm
on a similar aligment-related issue. :D
Note that this may be caused by the code in question pre-dating alignment offsets,
and we never checked. That said, these checks only work due to the regions' starting
addresses being no more aligned than their size (..if that makes sense?), so allowing
custom memory regions (#524) could violate that assumption and make those checks incorrect.
Assembling is never *supposed* to fail, but in this case I ended up adding
a test case that triggered a bug in the assembler, and that caused the rest
of the test to become garbled. This is thus more robust.
The only case where it is not generated by our own code
(which we ought to be `assume`ing is correct) is when the bank
is used-specified, but that is caught by an earlier check.
Moving both pieces of related code together, as it were.
This also ends up changing the behaviour of scrambling,
where sections could “leak” out of the scrambling's bank pool
and resume being placed normally.
We have discussed it offline, and decided that this was a bug.
The test suite has been updated accordingly, which even gives
the occasion to move this test from a special case to the main
generic loop!
This reverts a change introduced in fd78a9ae8, though it wasn't that commit's main point
so I'm feeling okay with undoing that.
This feels like an overkill change, using a static string is good enough for this
since we never modify this. I have considered using `string_view` instead, to have
the best of both worlds, but that's not NUL-terminated so our print functions
get a little grumpy.
This should remain equivalent for our current setups,
but play nicer with anyone wanting to experiment further.
It *may* break for someone building on an Intel Mac,
but this is a libpng limitation and thus out of our control.
The current setup fails with a "cannot read tree" error.
Also added flags to parallelise each fetch, which is relevant when submodules
are involved. (So, pretty much just starter-kit, haha.)
It would be nice to also parallelise the various fetches, but that would hit
the same remote multiple times, and could lead to rate-limiting...
The latter is not a Bash built-in,
and some environments turn out to produce non-conforming
return codes.
`type` is provided by Bash itself,
so we can rely on it better.
MSBuild is slow, and its output is hard to parse visually
(it's all interwoven, like Make without `--output-sync`).
Since we use CMake anyway, we don't care about the build system.
Automatically grabs and compiles libpng, and avoids having hacky lines in our Makefile
(the compiler specification *should* be orthogonal to the build target!)
Also move the MinGW package install to the `install_deps.sh` script,
to move logic off of the YAML.
- Turns out the variable wasn't being set, and thus the flags weren't being passed
- Passing the flags via those variables is basically a bad idea, and it
turns out to break libpng's config script
- The flags need to be passed during link as well, otherwise the binary
isn't linked as a fat one
ISSOtm
