Conversation

Notices

1. Embed this notice
   ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Wednesday, 30-Apr-2025 14:38:51 JST ✧✦Catherine✦✧

   this is the worst computer game i've ever played

   i haven't even managed to get the values of every core register yet

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Wednesday, 30-Apr-2025 15:53:14 JST ✧✦Catherine✦✧

     in reply to

     2 days after starting this i have finally managed to read one (1) dword from ROM

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Wednesday, 30-Apr-2025 16:11:26 JST ✧✦Catherine✦✧

     in reply to

     • Dirk Eibach

     @de i've dumped it via openocd before

     (i could also just dump the flash, but the code disassembles to a perfectly sensible reset handler, so...)

   • Embed this notice
     Dirk Eibach (de@noc.social)'s status on Wednesday, 30-Apr-2025 16:11:27 JST Dirk Eibach

     in reply to

     @whitequark How do you know it's correct?

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Wednesday, 30-Apr-2025 16:35:03 JST ✧✦Catherine✦✧

     in reply to

     • Brian Swetland

     @swetland I've done both and ARM7TDMI is closer to "an elaborate torture method" than "a debug interface"

   • Embed this notice
     Brian Swetland (swetland@chaos.social)'s status on Wednesday, 30-Apr-2025 16:35:04 JST Brian Swetland

     in reply to

     @whitequark Having done ARM7TDMI JTAG debug interface implementation first, back in the day, made the convoluted bits of SWD based debug really seem not all that bad by comparison.

     It's been long enough now that your posts on this are bringing back more nostalgia than trauma.

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Wednesday, 30-Apr-2025 16:38:23 JST ✧✦Catherine✦✧

     in reply to

     • Brian Swetland

     @swetland yeah ARM9 and ARM11 debug macrocells terrify me. I think I'll just skip anything with MMU

   • Embed this notice
     Brian Swetland (swetland@chaos.social)'s status on Wednesday, 30-Apr-2025 16:38:24 JST Brian Swetland

     in reply to

     @whitequark SWD has its warts, but it feels mostly (sometimes over-) designed... whereas the ARMTDMI stuff feels very, uh, tacked on after the fact is probably kindest... then ARM9 (not A9)... well I've forgotten the specifics, but I what remember is it being an even more convoluted evolution.

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Thursday, 01-May-2025 16:39:44 JST ✧✦Catherine✦✧

     in reply to

     my collection of upsetting facts grows

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Thursday, 01-May-2025 17:31:47 JST ✧✦Catherine✦✧

     in reply to

     my ARM7TDMI debug probe can consistently hit 1.022 MB/s (excluding protocol overhead, i.e. useful throughput) of reads!

     the theoretical maximum bandwidth is 1.290 MB/s, but the bottleneck in this case is actually the DUT itself: the core has internal synchronization overhead i know i'm bumping into since i poll DBGACK

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Thursday, 01-May-2025 17:49:54 JST ✧✦Catherine✦✧

     in reply to

     • abrasive

     @abrasive hehe~

   • Embed this notice
     abrasive (abrasive@digipres.club)'s status on Thursday, 01-May-2025 17:49:55 JST abrasive

     in reply to

     @whitequark that's unreasonably hot

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Thursday, 01-May-2025 17:55:10 JST ✧✦Catherine✦✧

     in reply to

     for some bizarre reason, cpython actually outperforms pypy here, by about 25%

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Thursday, 01-May-2025 18:01:14 JST ✧✦Catherine✦✧

     in reply to

     anyway, check this out: a complete high performance ARM7TDMI debug probe* https://github.com/whitequark/glasgow/blob/applet.debug.arm.arm7tdmi/software/glasgow/applet/debug/arm/arm7/__init__.py

     it took me 3 days or so and 180 lines of code to implement it. that's right, one hundred eighty

     * not a debugger; a debugger presents a UI, a debug probe can merely execute code and read/write data

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Thursday, 01-May-2025 18:08:00 JST ✧✦Catherine✦✧

     in reply to

     • ɗ𐐩ʃƕρʋ

     @deshipu probably but i honestly have too many things to be chasing down here already

   • Embed this notice
     ɗ𐐩ʃƕρʋ (deshipu@fosstodon.org)'s status on Thursday, 01-May-2025 18:08:01 JST ɗ𐐩ʃƕρʋ

     in reply to

     @whitequark sounds like something the pypy people would like to know about

   • Embed this notice
     a kilo of saucepans (rakslice) (rakslice@mastodon.social)'s status on Thursday, 01-May-2025 19:02:28 JST a kilo of saucepans (rakslice)

     in reply to

     @whitequark am I not understanding what LDR R0, [R0] means properly or is that last one just the usual rotation of data from unaligned reads on ARM7

   • Embed this notice
     ✧✦Catherine✦✧ (whitequark@mastodon.social)'s status on Thursday, 01-May-2025 19:28:41 JST ✧✦Catherine✦✧

     in reply to

     • a kilo of saucepans (rakslice)

     @rakslice the usual rotation of the data, but note that when you're in Debug mode and running at debug speed, the core ignores the pointer in R0 and asks you via JTAG what to put there

     so having it transposed is _weird_