Monday, 2023-01-16

*** lkcl <lkcl!> has joined #libre-soc00:16
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc00:26
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC02:17
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc02:46
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC03:00
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc03:04
*** Ritish <Ritish!~Ritish@2401:4900:4aa8:c9ad:4a2f:f17a:3c9f:f189> has joined #libre-soc04:10
sadoon[m]Get well soon04:28
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC04:32
lkclsadoon[m], oh - the change in weather (pressure) is done - it's raining. the moment the pressure stabilised, no more pain04:51
lkclnext pain event: when it *stops* raining... urr...04:51
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc05:02
sadoon[m]Good to hear :)05:04
ghostmansd[m]lkcl, get well. That's OK. I will reorganize this crap later once things work. For now this is in branch, and serves no other purpose than collecting some stuff for internal algorithms. I'll think of better way when I'm done with the tests.05:45
ghostmansd[m]Please check the comment at 950, I have other questions too. As for ff for crand, I'll implement a custom disassembly for it which was broken when I generalized and unified assembly/disassembly.05:46
ghostmansd[m]lkcl, I also have another question. Considering that the updated SVP64Asm produces the same assembly it produced before, what else could go wrong with the Program that some tests fail?05:48
ghostmansd[m]I mean, if we have exactly the same bytes (IIRC called SVP64Asm.binfile) we had before for instruction... What else could go wrong?05:50
ghostmansd[m]Any ideas? I would be happy to avoid debugging the Program/execute()/call() layers.05:51
lkclghostmansd[m], 2nd question first: it's stateless so there really should be nothing.  if it's exactly the same05:59
lkclif however you've bypassed PowerDecoder2 then there's an untested codepath05:59
ghostmansd[m]I simply dumped what I had in Program.binfile... I don't think I ever touched PowerDecoder2.06:00
ghostmansd[m]I touched pysvp64asm massively, though. And, perhaps, it's re-used somewhere even after we compile the Program?06:01
ghostmansd[m]I mean, I have completely the same Program.binfile as master branch does... But still some tests break. I think only sv-related.06:02
ghostmansd[m]Word instructions seem to work.06:02
*** Ritish <Ritish!~Ritish@2401:4900:4aa8:c9ad:4a2f:f17a:3c9f:f189> has quit IRC07:42
openpowerbot[mattermost] <gayatri> lkcl, i have doubt on microwatt building. As i saw a microwatt created at "soc" while creating libresoc. I can see a folder name as "Microwatt" at Soc of libresoc.07:57
openpowerbot[mattermost] <gayatri> Now i a trying to build microwatt alone at debian07:58
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC08:16
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has joined #libre-soc08:17
openpowerbot[mattermost] <gayatri> lkcl as you suggested to "once you have microwatt working you can then attempt substitution of the libresoc core using external_core_top.v". For this can we replace the external microwatt core code at "external_core_top.v  of module core on line-54558" .08:18
openpowerbot[mattermost] <gayatri> can we go head like that08:20
*** Ritish <Ritish!~Ritish@2401:4900:4aa8:c9ad:ff93:271e:ee70:62ab> has joined #libre-soc09:51
*** ghostmansd[m] <ghostmansd[m]!~ghostmans@> has quit IRC10:50
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc10:52
*** ghostmansd <ghostmansd!> has quit IRC11:29
lkclgayatri: that's what external_core_top.v is for.11:52
lkclthe only thing is, you must rename the top module to match the module name.11:55
lkclit is something daft, like core_123456789_98731_512_123411:55
*** Ritish <Ritish!~Ritish@2401:4900:4aa8:c9ad:ff93:271e:ee70:62ab> has quit IRC12:33
lkclRitish, your key's added, please make sure never to give out the private key to anyone - notify me immediately if it ever goes outside of your exclusive control12:34
lkclyou can clone and write to the wiki repo;a=summary12:34
*** yambo <yambo!> has quit IRC12:35
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC14:48
*** octavius <octavius!> has joined #libre-soc14:50
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc15:49
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC16:45
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc17:59
*** ghostmansd <ghostmansd!> has joined #libre-soc18:26
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC18:31
*** lx0 <lx0!~lxo@gateway/tor-sasl/lxo> has joined #libre-soc18:59
*** octavius <octavius!> has quit IRC19:00
*** octavius <octavius!> has joined #libre-soc19:01
*** lxo <lxo!~lxo@gateway/tor-sasl/lxo> has quit IRC19:01
octaviusToday been going through the ls2 wiki page (and setting up another chroot). During building of the Libre-SOC external core, I'm getting an error coming from nMigen's (see the pastebin: , will document this issue if it's not just me). Has anyone seen this before? I've ONLY used the dev-scripts to install software.19:16
octaviusIs it due to nMigen being updated since ls2 was last used?19:17
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc19:17
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc19:19
*** lx0 <lx0!~lxo@gateway/tor-sasl/lxo> has quit IRC19:32
*** lx0 <lx0!~lxo@gateway/tor-sasl/lxo> has joined #libre-soc19:33
programmerjakeit looks like you got a version of nmigen from pypi or something, you need the version from
programmerjakethe version should be<something>19:42
programmerjakeyou have 0.219:43
octaviusOk interesting. Thanks Jacob!20:00
*** openpowerbot <openpowerbot!> has quit IRC20:12
*** openpowerbot <openpowerbot!> has joined #libre-soc20:12
octaviusThe hdl-dev-repos script deletes any tags found for nmigen (and the other soc repos too). Once the tag is deleted, running leaves a 0.1 version of nmigen20:13
octaviusFinished processing dependencies for nmigen==0.1.dev1205+g29dec3020:13
*** openpowerbot <openpowerbot!> has quit IRC20:13
programmerjakeoh? deleting tags is incorrect...20:14
*** openpowerbot <openpowerbot!> has joined #libre-soc20:14
programmerjakethere should always be at least a v0.2 tag20:14
*** gnucode <gnucode!~gnucode@user/jab> has quit IRC20:16
octaviuswhat about the rest of the nmigen-* repos (boards, soc, stdio), do they also need a particular tag?20:17
programmerjakeidk, maybe?20:19
programmerjakelemme look20:19
programmerjakeafaict no, no other repos need tags20:21
octaviusOk, then I'll try again, but make sure nmigen retains the 0.2 tag20:22
*** gnucode <gnucode!~gnucode@user/jab> has joined #libre-soc20:30
octaviusI've been able to generate the external_core_top.v successfully now.20:42
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC20:42
octaviusHowever during installation of gram (DRAM controller lib), I'm hitting an error with python setuptools:
programmerjakeyour version of setuptools_scm is too old, you need at least v3.3.0 since that's when it added support for fallback_version20:52
programmerjakeit was added in
*** ghostmansd[m] <ghostmansd[m]!> has joined #libre-soc22:02
octaviusAh, debian buster oldstable package is at v3.2.022:28
octaviusbodged it by installing a .whl copy from pypi for now22:35
*** ghostmansd[m] <ghostmansd[m]!> has quit IRC23:38

Generated by 2.17.1 by Marius Gedminas - find it at!