Tuesday, 2023-08-08

openpowerbot[irc] <programmerjake> lkcl: finished splitting out insns, pytest passes on my computer, all that's left is getting someone to double-check the split-insns branch before pushing to master00:25
openpowerbot[irc] <programmerjake> broken memory model + speculative execution fun: https://randomascii.wordpress.com/2018/01/07/finding-a-cpu-design-bug-in-the-xbox-360/01:15
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc01:20
*** gnucode <gnucode!~gnucode@user/jab> has quit IRC02:52
*** jn <jn!~quassel@user/jn/x-3390946> has quit IRC03:10
*** jn <jn!~quassel@2a0a-a541-726a-0-20d-b9ff-fe49-15fc.ipv6dyn.netcologne.de> has joined #libre-soc03:10
*** jn <jn!~quassel@2a0a-a541-726a-0-20d-b9ff-fe49-15fc.ipv6dyn.netcologne.de> has quit IRC03:10
*** jn <jn!~quassel@user/jn/x-3390946> has joined #libre-soc03:10
*** libredev <libredev!libredev@libredev.ircforever.org> has quit IRC07:13
*** libredev <libredev!libredev@libredev.ircforever.org> has joined #libre-soc07:25
*** leah <leah!~leah@libreboot/developer/leah> has quit IRC08:50
*** leah <leah!~leah@libreboot/developer/leah> has joined #libre-soc08:52
markos_lkcl, madd*rs need to be added to the 3rd list "openpower/isa/* mdwn files, but not otherwise implemented yet", you removed them completely from the list09:15
markos_I can do it but I wanted to let you know first09:16
markos_that's for #106809:16
openpowerbot[irc] <programmerjake> maddsubrs/maddrs/msubrs are implemented in the simulator, so they go under that list. the "in openpower/isa/*" list is only for things that we can't even simulate yet09:19
openpowerbot[irc] <programmerjake> (or a corresponding list for things that are ready for binutils but aren't part of that task)09:20
markos_right, there should be different tasks (with different budgets) for these lists09:20
lkclmarkos_, if they're not implemented in the simulator (not listed by the insndb database command) then they *do not* go into #1068.09:21
lkclit will need another round, later09:21
lkclthere is a massive re-organisation coming, involving PO9, and i do not want time wasted on placing instructions into PO22 that are going to have to be completely reordered09:21
lkcl> markos_> I can do it but I wanted to let you know first09:22
lkclyes please, but can you hold off for 5 mins, i just need to put "TODO" in front of the list of instructions09:22
lkcl>  lkcl: finished splitting out insns,09:22
lkclfantastic, thank you09:23
markos_I see you're still editing, I'm not touching it09:23
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-90-154-80-239.ip.moscow.rt.ru> has quit IRC09:25
openpowerbot[irc] <programmerjake> well, a repairman's coming over tomorrow at 9:30am, so gn, ttyl09:25
lkclmarkos_, done.09:25
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc09:25
lkclediting the same comment is not possible (it's not git). editing a *different* comment, not a problem. it is what it is09:26
lkclprogrammerjake: excellent https://bugs.libre-soc.org/show_bug.cgi?id=1068#c4009:53
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC10:46
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc10:46
*** lkcl <lkcl!lkcl@freebnc.bnc4you.xyz> has quit IRC10:52
*** awilfox <awilfox!~awilfox@kelsey.foxkit.us> has quit IRC10:52
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC10:56
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc10:57
*** lkcl <lkcl!lkcl@freebnc.bnc4you.xyz> has joined #libre-soc11:04
*** awilfox <awilfox!~awilfox@kelsey.foxkit.us> has joined #libre-soc11:04
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC11:09
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-90-154-80-239.ip.moscow.rt.ru> has joined #libre-soc11:09
*** octavius <octavius!~octavius@> has joined #libre-soc11:49
*** octavius <octavius!~octavius@> has quit IRC13:10
openpowerbot[irc] <sadoon[m]1> Hi. Does this work?13:34
openpowerbot[irc] <programmerjake> yes: https://libre-soc.org/irclog/latest.log.html#t2023-08-08T13:34:2513:36
openpowerbot[irc] <sadoon[m]1> Awesome13:36
openpowerbot[irc] <sadoon[m]1> Alright planning to speedrun debian today before the meeting!13:37
openpowerbot[irc] <sadoon[m]1> I'm lucky to have backed up my old ppc64/ppc build VM because I created a bunch of useful scripts to automate building there before I even joined libre-soc, will share them later as well13:38
*** octavius <octavius!~octavius@> has joined #libre-soc14:03
openpowerbot[irc] <sadoon[m]1> Build script is chugging along nicely!14:33
openpowerbot[irc] <sadoon[m]1> Hmm I could have setup a tmpfs for the build dirs but I kept it this way, should be more unviersal. my schroot is in a tmpfs though14:48
openpowerbot[irc] <sadoon[m]1> The script is on bugzilla 1130 now14:48
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-90-154-80-239.ip.moscow.rt.ru> has quit IRC15:47
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc15:48
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC15:58
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@broadband-90-154-80-239.ip.moscow.rt.ru> has joined #libre-soc15:59
*** choozy <choozy!~choozy@75-63-174-82.ftth.glasoperator.nl> has joined #libre-soc16:15
ghostmansd[m]lkcl, programmerjake, thank you guys for updates on 1068. I've almost completed it, one question remains: do we need to introduce aliases for fminmax now at this point? We have them for minmax.18:38
ghostmansd[m]If yes, it'd be great if we have some table which depicts all possible aliases for fminmax (I guess it depends on FMM operand, right?).18:39
ghostmansd[m]Also, since we are going to have more aliases... Perhaps we could come up with some format which binds the aliases to the instructions?18:40
ghostmansd[m]I'll add the fminmax instruction itself for now (along with new FMM operand), and will wait for your opinion on whether I need to add aliases manually or, preferably, invent some format for aliases binding.18:42
openpowerbot[irc] <programmerjake> the table of aliases is the same table describing FMM https://libre-soc.org/openpower/sv/rfc/ls013/18:42
openpowerbot[irc] <programmerjake> (some extrapolation required to Rc=1 and *s variants)18:43
openpowerbot[irc] <programmerjake> the list should be the same as the list of old fmin*/fmax* insns18:44
*** josuah <josuah!~josuah@> has quit IRC19:00
ghostmansd[m]Thanks programmerjake!19:04
ghostmansd[m]*s, what does it stand for?19:07
ghostmansd[m]I'd have guessed it's "signed", but it's floating and there's no "u" counterpart.19:08
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc19:24
openpowerbot[irc] <programmerjake> single19:40
openpowerbot[irc] <programmerjake> for 32-bit float19:41
openpowerbot[irc] <programmerjake> oh, actually, fminmax doesn't have *s variants, since they're redundant since the non-s variant works just fine on 32-bit floats too19:42
*** octavius <octavius!~igloo@> has joined #libre-soc19:43
octaviusGood evening chaps, I'm on the train back from London, so my connection may be intermittent. I'll try to join though19:44
*** octavius <octavius!~igloo@> has quit IRC19:46
*** gnucode <gnucode!~gnucode@user/jab> has quit IRC19:47
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc19:47
*** octavius <octavius!~igloo@> has joined #libre-soc19:56
octaviusOn second thought, I'm too tired. Apologies, will skip for today19:56
*** octavius <octavius!~igloo@> has quit IRC19:57
openpowerbot[irc] <sadoon[m]1> Safe trip octavius :)20:01
*** josuah <josuah!~josuah@> has joined #libre-soc21:43

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!