gbdev/rgbds
1
0
Fork 0
mirror of https://github.com/gbdev/rgbds.git synced 2026-09-22 13:47:05 +00:00
Code Issues Packages Projects Releases Wiki Activity
3,708 Commits 3 Branches 52 Tags
5b9903f358356d06cfb53036dce9d35900ce18d8
Commit Graph
983 Commits
Author SHA1 Message Date
ISSOtm 5b9903f358 Mark a link test as failed if assembling fails 
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.
 2026-09-18 21:19:37 +02:00
ISSOtm ef709fade6 Refactor and fix sect packing's bank iteration 
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!
 2026-09-18 21:19:37 +02:00
Eldred Habert afe1f01d00 Fix linker script align 16 in floating address context (#2115) 
Turns out we were just using the wrong var lmao
 2026-09-17 20:46:55 -04:00
Rangi 8bf5957c83 Fix RGBLINK assertion failure when -s and -c invalidly match (#2109) 2026-09-18 01:34:14 +02:00
Rangi b4988afc8a Fix RGBLINK ignoring alignment constraints when placing empty sections 
Fixes #2095
 2026-09-17 11:16:18 -04:00
ISSOtm 6f69867d04 Make more RGBLINK tests use at-files 
Less special handling = good

Possibly this can be done to more tests, but this is what I found.
 2026-09-17 10:25:47 -04:00
ISSOtm 960f82126f Fix to actually allow one-past-the-end zero-sized sections in linker scripts 2026-09-17 10:19:47 -04:00
ISSOtm e39dbe1752 Remove no-op in link test script 
No, `printf` does not do that.
This works because making it read from stdin suppresses
any words after the first one.
 2026-09-17 09:35:32 -04:00
Rangi b2967ff509 Correct rgblink(5) linkerscript doc to match behavior of unbanked ROMX/WRAMX (#2103) 2026-09-17 05:32:15 +02:00
Rangi 004749ea75 Halt processing when any ambiguous-transparency colors exist (#2093) 2026-09-16 04:03:56 +02:00
Rangi f4c6221eb1 Fix underflow of ALIGN[16, offset] (#2090) 2026-09-15 20:22:33 -04:00
Rangi d81ab2fb23 Fix underflowing SECTION FRAGMENT fixed address (#2091) 2026-09-15 20:22:19 -04:00
Rangi b0f49a36b7 Fix -t/--tiny for ROMX in bank 1, and -w/--wramx for WRAMX in bank 1 (#2089) 2026-09-14 15:53:45 -04:00
Rangi ed1ac7309b Fix overflow with high fixed-point magnitudes 2026-09-14 14:27:16 -04:00
Rangi d082f24a2d Fix rounding of fixed-point constants where fractional digits should carry into integer part (#2088) 2026-09-14 13:52:39 -04:00
Rangi 7746ddbf2b Fix reverse-graphics check for whether the tilemap references too-high tile IDs in bank 1 (#2086) 2026-09-14 10:26:57 -04:00
Rangi f8a1da71c1 Only sort generated palette colors by PLTE order for indexed PNG images 
Non-indexed images can still have a PLTE chunk. We should use
`sortRgb` instead of `sortIndexed` for them, since their pixels' colors
may not all be present in the PLTE chunk and would be unsortable.
 2026-09-13 18:53:31 -04:00
Rangi 49b3f566b1 Fix handling of PNG files with RGB or RGBA color type that have a PLTE chunk 2026-09-13 16:09:49 -04:00
Rangi 82f58b95ac Fix reversing some images with an input tileset (#2081) 2026-09-13 21:00:58 +02:00
RangiandEldred Habert e9fc7b3135 Warn when tile or palette IDs cannot fit in the tilemap or attrmap respectively (#2077) 
Co-authored-by: Eldred Habert <[email protected]>
 2026-09-13 19:18:24 +02:00
Rangi 47bbd30f2c Fix a failing assertion when two sections have the same name (#2079) 2026-09-12 17:58:37 -04:00
Rangi 520cc2a80d Fix reversing a tilemapped image with -x trimmed tiles 2026-09-12 15:38:34 -04:00
Rangi e9c800b3be Fix reversing 1bpp images 
Fixes #2075
 2026-09-12 14:17:33 -04:00
Rangi 8a7019e1f8 Only register colors inside the -L/--slice region, including transparent color 
Fixes #2074
 2026-09-12 12:52:35 -04:00
Rangi 439a7a1828 Fix reversing multiple palettes with fewer than 4 colors per palette 
Fixes #2073
 2026-09-12 11:50:47 -04:00
Rangi 438b943a1a Make background tile IDs consistent whether or not deduplication is enabled 
Fixes #2072
 2026-09-11 20:45:52 -04:00
RangiandISSOtm 9e785495ff Fix crash when a background tile first occurs after both tile banks are full (#2070) 
Co-authored-by: ISSOtm <[email protected]>
 2026-09-12 01:04:31 +02:00
ISSOtm 9d69a08909 Determine the number of fetch/clone jobs ourselves 
Apparently our CI has some versions of Git too old to support zero to mean that.
 2026-09-11 16:31:19 -04:00
ISSOtm 970bb84450 Fix updating ext-test repos from existing copies 
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...
 2026-09-11 16:17:26 -04:00
Rangi a662cad7ce Fix infinite loop when an aligned HRAM section cannot be placed (#2064) 2026-09-11 05:21:26 +02:00
Rangi e85375f3e1 Avoid truncation warning from sp instruction test 2026-09-10 19:09:15 -04:00
Rangi 429a2325fb Add test case for reversing image with transparent color 2026-09-10 18:18:37 -04:00
Rangi 9c5e872f9c First color of -c palettes with transparent pixels is transparent $8000, not unspecified $ffff 2026-09-10 16:41:32 -04:00
Rangi 25240e9fc0 Fix but deprecate an implicit transparent color #0 2026-09-10 19:32:00 +02:00
Rangi 8a6b1946e3 Parse inline palette specs after getting CLI palette size limit (#2054) 2026-09-10 18:54:24 +02:00
Max Freedom Pollard 631ef003e7 Avoid signed overflow in RGBLINK's +, -, and * (#2060) 
RGBASM computes these with unsigned arithmetic and casts back, since
signed overflow is UB in C++, but RGBLINK's RPN evaluator used `int32_t`
directly, so `src/link/patch.cpp` tripped UBSan on lines 117, 121, and
124. Share the three operators through `opmath.cpp` so both evaluators
stay in step.
 2026-09-07 16:15:07 -04:00
Rangi fdd6cece30 Warn about signed 8-bit truncation for add sp, e8 and ld hl, sp + e8 2026-09-07 19:28:45 +02:00
Rangi 7d0cb000e8 Specify COLUMNS in test scripts for consistent output 2026-09-06 00:07:16 +02:00
Rangi 8988cdde46 Fix reversing images with fewer than 4 colors per palette 2026-09-03 13:39:24 +02:00
Rangi a2c52caca2 Disallow NUL characters in section names and assertion messages 
Allowing these in object files would lose anything after the '\0'
when RGBLINK reads the object.
 2026-08-22 19:51:03 -04:00
Rangi 21a4b85a4f Fix STRFMT stopping at NUL characters in format spec strings 2026-08-22 19:23:49 -04:00
Rangi 9fa5058add Correct error message for macro arg \<-INT_MIN> 2026-08-22 19:02:23 -04:00
Rangi e287ee2724 Fix C++ UB from negating INT_MIN with macro shift INT_MIN 2026-08-22 18:48:36 -04:00
Rangi 0888600cb7 Avoid OOM allocation error from invalid too-high bank numbers 2026-08-22 18:31:26 -04:00
Rangi 8bc7de35f9 Prevent SECTION FRAGMENT combined sizes from overflowing their uint16_t size 2026-08-22 18:24:17 -04:00
Rangi e7e671b249 Update external test commits 2026-08-01 11:24:42 -04:00
Rangi 0c772ea049 Update external testing CI with libbet's new way of running Python 
libbet now uses `py -3` if the `py` binary exists, instead of if
`COMSPEC` is defined. So we don't need to `unset COMSPEC`, we just
need to avoid Windows' `py` launcher from being in Cygwin's `PATH`.

It turns out that `C:\cygwin\bin\env.exe CYGWIN_NOWINPATH=1 ...`
was not effectively preventing Windows' `PATH` from being inherited
by Cygwin's `PATH`, because to take effect it needs to run Cygwin's
`/etc/profile` script. The simple way to do that is using a login
shell. So we don't need the `PREFIX`/`--prefix` from the previous
commit after all.
 2026-08-01 04:49:43 +02:00
Rangi ec5e6cbabb Handle -Wtruncation for link-time jr values 
Fixes #2028
 2026-07-21 14:51:58 -04:00
Rangi d9c0d37fd5 Explicitly allow JR offsets to wrap around in RGBLINK 
This reverts commit f27a8132d9.
 2026-07-20 19:23:11 -04:00
Rangi 9a15b723f2 Format CGB colors in error/verbose messages as "GB:rr,gg,bb" 
These should be more debuggable than packed "$xxxx" format.
 2026-07-14 10:34:27 -04:00