| ghostmansd | lkcl, does the following update to svp64.py look correct to you? | 13:38 |
|---|---|---|
| ghostmansd | https://pastebin.com/5g9UhW5Z | 13:38 |
| ghostmansd | https://libre-soc.org/openpower/sv/setvl/ here vf is last, 25 | 13:39 |
| ghostmansd | also, other insns which mention this field use offset at 25 bit | 13:39 |
| ghostmansd | this is a totally minor, comment-only, but still that was confusing | 13:40 |
| lkcl | if i can get an internet connection that isn't slooooow urrr | 14:11 |
| lkcl | ok hang on let's check fields.txt | 14:12 |
| lkcl | https://git.libre-soc.org/?p=openpower-isa.git;a=blob;f=openpower/isatables/fields.text;h=d4b5075f2b3c16252c6686163c0147d2546e1971;hb=HEAD#l258 | 14:13 |
| lkcl | 258 # 1.6.28 SVL-FORM | 14:13 |
| lkcl | 259 |0 |6 |11 |16 |23 |24 |25 |26 |31 | | 14:13 |
| lkcl | 260 | PO | RT | RA | SVi |ms |vs |vf | XO |Rc | | 14:13 |
| lkcl | sigh | 14:13 |
| lkcl | now let's check PowerDecoder2 etc. | 14:13 |
| lkcl | which is... errr.. | 14:13 |
| lkcl | https://git.libre-soc.org/?p=openpower-isa.git;a=blob;f=openpower/isa/simplev.mdwn;h=8a21f038075e94668fcb344cff1260db8cc1162b;hb=HEAD#l22 | 14:18 |
| lkcl | arse. | 14:18 |
| lkcl | going to have to work out how to sort that *without* having to alter a dozen-or-more unit tests | 14:18 |
| ghostmansd[m] | I've just thought that in the future we might generate the fields automatically as well. We _do_ have some stuff which process fields.txt, right? | 14:22 |
| lkcl | yes of course | 14:23 |
| lkcl | power_decoder.py | 14:23 |
| lkcl | uses something called DecodeFields in power_fields.py | 14:24 |
| lkcl | https://git.libre-soc.org/?p=openpower-isa.git;a=blob;f=src/openpower/decoder/power_fields.py;hb=HEAD | 14:24 |
| lkcl | the "common fields" was a bit of a cheat, that came initially from microwatt. | 14:25 |
| lkcl | it shouldn't really be used | 14:25 |
| lkcl | but it's pretty frickin easy, but relies on whitespace positioning | 14:26 |
| lkcl | as well as the number of "/"s | 14:26 |
| lkcl | this would be wrong: | 14:26 |
| lkcl | 258 # 1.6.28 SVL-FORM | 14:27 |
| lkcl | 259 |0 |6 |11 |16 |23 |24 |25 |26 |31 | | 14:27 |
| lkcl | 260 | PO | RT | RA | SVi |ms |vs |vf | XO |Rc | | 14:27 |
| ghostmansd[m] | Ok I'll take a look | 14:27 |
| lkcl | because the "|" separating RT and RA is not lined up with its column | 14:27 |
| ghostmansd[m] | That's not for now, just a note to my future self | 14:27 |
| * lkcl nods | 14:27 | |
| lkcl | urr you know what? you know why i allowed sv.svstep? | 14:33 |
| lkcl | it's so that the register numbers aren't limited to r0-r31 | 14:33 |
| lkcl | okaay so i lied: sv.svstep *is* in, sv.setvl is not | 14:38 |
| lkcl | ghostmansd[m], okaaay it's just the naming you changed (effectively the comments) | 15:20 |
| lkcl | not the actual format. | 15:20 |
| lkcl | thank god for that | 15:20 |
| programmerjake | sv.setvl would make sense too for larger reg numbers...just no vector arguments. | 19:46 |
| programmerjake | lkcl: budget_sync.budget_graph.BudgetGraphPayeesParseError: Failed to parse cf_payees_list field of bug #813: unknown payee name: 'red' is not the name or an alias of any known person | 20:30 |
| programmerjake | idk the right info, can you add that to the config toml? | 20:30 |
| programmerjake | also, thx for the donation: https://lists.libre-soc.org/pipermail/libre-soc-bugs/2022-May/006333.html | 20:34 |
| lkcl | added | 20:41 |
| lkcl | yyeah, except it's a lot of wasted bits. | 20:42 |
| programmerjake | thx! | 20:42 |
| lkcl | i'd like to think of something else to do with them but leave it for now | 20:42 |
| lkcl | reserved if nothing else because there reaallly is no room, at all, for expansion | 20:43 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!