Post B2f4AbcA2nPSWBnu64 by techokami@woof.tech
 (DIR) More posts by techokami@woof.tech
 (DIR) Post #B2ckcxQORc1zYM4kts by asie@mk.asie.pl
       2026-01-24T22:38:37.768Z
       
       0 likes, 0 repeats
       
       the newest compiler for the Epson S1C33 CPU is currently binutils 2.10.1 / gcc 3.3.2 (the CD shipped with gcc 2.7.2). should I:
       
 (DIR) Post #B2ckuvxubolKEuKFo8 by asie@mk.asie.pl
       2026-01-24T22:41:55.185Z
       
       0 likes, 0 repeats
       
       if you vote for the hobbyist compiler option you are obligated to link the hobbyist compiler
       
 (DIR) Post #B2eXO6fwZfWm85P4E4 by asie@mk.asie.pl
       2026-01-25T08:44:12.016Z
       
       0 likes, 0 repeats
       
       I dived deeper into the binutils / GCC port.It seems to be a derivative of the V850 code (which is good, as that is still actively maintained) with decent comments explaining their changes and rationale. The reason EPSON stuck with these old versions, per notes from OpenMoko who also used this chip, is because they didn't want the hassle of tracking compiler bugs ("stability over new features"). But it should be viable to port upwards - at least binutils should be easy enough.
       
 (DIR) Post #B2eXO7afAmtixzwMdM by asie@mk.asie.pl
       2026-01-25T08:45:59.173Z
       
       0 likes, 0 repeats
       
       Makes me wonder if they would okay an FSF copyright assignment lol
       
 (DIR) Post #B2eXO8Depo5cuwBVPk by asie@mk.asie.pl
       2026-01-25T10:45:11.713Z
       
       0 likes, 0 repeats
       
       Makes me wonder how long they kept maintaining it. The OpenMoko patches are from late 2007, but manuals suggest they were putting out updates as late as 2011-2012.
       
 (DIR) Post #B2eXO90DvEwDLYuI76 by asie@mk.asie.pl
       2026-01-25T17:58:52.537Z
       
       0 likes, 0 repeats
       
       Porting GNU assembler targets is giving me eerie Minecraft modding flashbacks...
       
 (DIR) Post #B2eXOA2k3MXwZf5og4 by asie@mk.asie.pl
       2026-01-25T19:19:36.820Z
       
       0 likes, 0 repeats
       
       https://codeberg.org/WonderfulToolchain/binutils/commit/e1a1a69194331c7da6cfb26e3cc8d383dad3230aSeems to work enough for a start. Lots of warnings to be fixed, but the day is nearing its end. GCC will be more fun.
       
 (DIR) Post #B2ead20uUN1vsS1Cs4 by wolf480pl@mstdn.io
       2026-01-25T18:00:30Z
       
       0 likes, 0 repeats
       
       @asie are you replacing function pointers that are not meant to be replaced?
       
 (DIR) Post #B2f4AbcA2nPSWBnu64 by techokami@woof.tech
       2026-01-25T01:38:39Z
       
       0 likes, 0 repeats
       
       @asie Had to look this up, and surprisingly I found the manuals for this CPU https://global.epson.com/products_and_drivers/semicon/pdf/id000859.pdf and the ISA itself! https://global.epson.com/products_and_drivers/semicon/pdf/id000569.pdf
       
 (DIR) Post #B2f4AcrRPTE6NrxcYK by asie@mk.asie.pl
       2026-01-25T06:46:41.013Z
       
       0 likes, 0 repeats
       
       @techokami@woof.tech Oh, you found the English S1C33209 manual! Great - I only had the one for the core, and a Japanese one for the ISA (from the P/ECE bundled CD).
       
 (DIR) Post #B2f7L0hVvc3H52xcX2 by asie@mk.asie.pl
       2026-01-25T08:13:27.443Z
       
       0 likes, 0 repeats
       
       --- /dev/null+++ b/bfd/cpu-c33.c@@ -0,0 +1,109 @@+/* BFD support for the NEC V850 processor+   Copyright 1996, 1997, 1998 Free Software Foundation, Inc.oh I see what you did there EPSON. The V850 support in binutils/GCC is still maintained, so it would actually be pretty easy to just replicate-+/* C33: All registers are same in C33 and any MODE can be held.+        But the following line corresponds to V850. Just in case. */Mm-hmm.