| *** tpb has joined #timvideos | 00:00 | |
| *** tactics__ has quit IRC | 01:54 | |
| *** futarisIRCcloud has joined #timvideos | 02:08 | |
| nrossi | mithro: if you are still around. Managed to get Vivado to install and vivado --help run entirely within the OE build. However it does have the draw back of being copied once (to store it in sstate). Have not had a chance to test ISE though | 03:02 |
|---|---|---|
| mithro | nrossi: Is there a way to "provide" vivido as part of the host/building system? I think RP had a suggestion? | 03:24 |
| nrossi | mithro: yes there are a couple of ways, probably the most effective is to have a recipe that provides some wrapper scripts in the recipe sysroot that just source the xilinx tools and call the associated executables | 03:36 |
| mithro | nrossi: That actually sounds reasonable pleasant because it seems like you could have *either* it is installed in the sysroot via your method above **or** you provide the fake wrappers which shell out to your host? | 03:56 |
| mithro | shorne: I see you just pushed a "OpenRISC GCC Toolchain 9.0.0-20181026" ? | 03:57 |
| nrossi | mithro: yep exactly, and you can select which one using a PREFERRED_PROVIDER setting | 03:57 |
| mithro | That is "interesting" -- Release: GCC 6.5Release date: October 26, 2018 | 03:58 |
| mithro | Is GCC 6.5 a "long term supported" release or something? | 03:59 |
| nrossi | mithro: isn't 6.5 the last of the 6.x tree? https://gcc.gnu.org/ml/gcc/2018-10/msg00221.html | 04:00 |
| tpb | Title: Jakub Jelinek - GCC 6.5 Released (at gcc.gnu.org) | 04:00 |
| mithro | nrossi: That is what I mean, is the GCC 6.x tree special in some way? | 04:00 |
| mithro | Oh interesting - "This is also the last release from the GCC 6 branch" | 04:02 |
| mithro | nrossi: I'm very interested in see if you put together all the stuff you have done, does is actually create something similar to litex-buildenv but "better" | 04:08 |
| futarisIRCcloud | nrossi: Me too. Where is your OE stuff? | 04:09 |
| nrossi | mithro: the xilinx tools stuff is just wip commits on the meta-hdl "wip" branch https://github.com/nathanrossi/meta-hdl/tree/wip | 04:13 |
| tpb | Title: GitHub - nathanrossi/meta-hdl at wip (at github.com) | 04:13 |
| nrossi | and the openrisc changes are on a oe-core branch here: https://github.com/nathanrossi/openembedded-core/tree/nrossi/or1k (i plan to submit some of that upstream after oe-core does their imminent release) | 04:15 |
| tpb | Title: GitHub - nathanrossi/openembedded-core at nrossi/or1k (at github.com) | 04:15 |
| *** Kripton has quit IRC | 04:52 | |
| *** Kripton has joined #timvideos | 04:54 | |
| mithro | nrossi: Could we maybe put a "burn down" or similar type of "thing needed to still be done" list in the doc at https://docs.google.com/document/d/1hBvcE83vDRn5EVzcv4Qv_dcGlDuXp5pNI-_yzrEOwxs/edit ? -- I'm a bit unsure about how close / far away we are from being able to replicate litex-buildenv style functionality..... | 05:18 |
| tpb | Title: Google Docs - create and edit documents online, for free. (at docs.google.com) | 05:18 |
| nrossi | mithro: sure, but out of query is the desire to have it capable of building the HDMI2USB project at first? | 05:20 |
| mithro | nrossi: The first targets would be - HDMI2USB firmware, Linux, bare metal MicroPython (and maybe Zephyr) | 05:21 |
| mithro | nrossi: The order isn't all that important I guess? | 05:21 |
| nrossi | mithro: no not really, just want to make sure I have something to target :) | 05:22 |
| mithro | MicroPython is kind of my current focus | 05:22 |
| mithro | And it does have the advantage of being able to be small enough to sensibly usable on the iCE40 devices which have the open toolchain while also working on the "big" boards too | 05:23 |
| mithro | nrossi: Did I share https://github.com/timvideos/litex-buildenv/wiki/HowTo-FuPy-on-a-Digilent-Arty-A7 ? | 05:23 |
| tpb | Title: HowTo FuPy on a Digilent Arty A7 · timvideos/litex-buildenv Wiki · GitHub (at github.com) | 05:23 |
| mithro | nrossi: So maybe it makes sense to target the MicroPython initially? | 05:24 |
| nrossi | mithro: sure i will target that first | 05:24 |
| mithro | nrossi: FYI - If it turns out that doing things this way with Yocto / OE works well enough that I can throw away conda+crappy shell scripts and do more actual interesting FPGA work, then I will owe you big time! | 05:27 |
| nrossi | mithro: :) we will see | 05:27 |
| cr1901_modern | nrossi: Does yocto/openembedded build on Windoze? | 05:28 |
| nrossi | cr1901_modern (IRC): it can target mingw. But for building on windows no. WSL though i believe has had some success | 05:28 |
| cr1901_modern | Hrm, that's prob gonna be an issue. The conda situation isn't great, but with the present setup it's possible to use litex-buildenv under an MSYS2 environment | 05:29 |
| cr1901_modern | if you're careful* | 05:30 |
| mithro | nrossi: What is the difference between "building" and "targeting" ? Do you mean you could build an SDK package which lets someone do "development" on windows, but not build the whole thing? | 05:30 |
| nrossi | mithro: yer, so you can get it to build the tools (gcc, qemu, iverilog, yosys, etc) that will run on windows. But running bitbake/OE/yocto itself on windows is just in general a no go | 05:31 |
| cr1901_modern | mithro: That's actually fine. I think the idea was to try to use travis for all deps nyway. | 05:32 |
| nrossi | Also as soon as you want to build things like the Linux kernel your just gonna have a bad time trying to do that one windows | 05:32 |
| cr1901_modern | build linux on Windows is most certainly not supported right now as-is anyway | 05:32 |
| cr1901_modern | it _may_ be possible from an msys2 prompt | 05:32 |
| nrossi | sure, but if thats a future goal to support with litex then you are likely unable to support that for windows | 05:33 |
| nrossi | also, whilst you might be able to build the kernel, building the rootfs and programs will prove challenging ;) | 05:34 |
| mithro | nrossi: qemu all the say! :-P | 05:35 |
| mithro | s/say/way/ | 05:35 |
| mithro | nrossi: What about Mac? | 05:35 |
| cr1901_modern | As long as one can _use_ litex-buildenv from Windows after the switch to yocto, then I don't really have a problem with building all the deps on linux/travis | 05:36 |
| nrossi | mithro: same a windows, and likely in a broken state as it is not well supported by yocto (meta-darwin) | 05:36 |
| mithro | We don't support Mac *at all* at the moment | 05:36 |
| nrossi | mithro: well, it is something that could be improved, there are other users out there that have asked about macos support with yocto | 05:38 |
| mithro | nrossi: Someone was playing with the idea of using wine to run ISE / Vivado on Mac -- I think they got it to install and compile stuff but never got around to integrating it into the build environment... | 05:39 |
| nrossi | mithro: interesting, although i imagine that it would be more functional to build open tools for macos no? | 05:40 |
| mithro | Yes :-) | 05:40 |
| mithro | It was more impressive just from actually making it do something other than just segfault (segfaulting seems to be the natural state for vendor released toolchains ;-) | 05:42 |
| nrossi | cr1901_modern (IRC): just for curiosity, your reason for using windows, choice? or work enforced? | 05:42 |
| cr1901_modern | choice | 05:42 |
| nrossi | mithro: yes, even on the supported distros apparently :) | 05:42 |
| *** rohitksingh has joined #timvideos | 05:57 | |
| *** mcarden_ has quit IRC | 06:07 | |
| *** mcarden has joined #timvideos | 06:07 | |
| *** rohitksingh has quit IRC | 06:11 | |
| nrossi | mithro: added a table to the doc for the lite-buildenv components and the state for it in OE | 06:13 |
| *** rohitksingh has joined #timvideos | 06:24 | |
| *** rohitksingh has quit IRC | 06:34 | |
| *** rqou has quit IRC | 06:39 | |
| *** rqou has joined #timvideos | 06:39 | |
| *** futarisIRCcloud has quit IRC | 06:57 | |
| *** futarisIRCcloud has joined #timvideos | 07:21 | |
| *** rohitksingh has joined #timvideos | 07:22 | |
| *** rohitksingh has quit IRC | 08:30 | |
| *** futarisIRCcloud has quit IRC | 09:37 | |
| *** rohitksingh has joined #timvideos | 10:53 | |
| *** rohitksingh has quit IRC | 11:08 | |
| *** mauz555 has joined #timvideos | 11:27 | |
| *** mauz555 has quit IRC | 14:52 | |
| *** mauz555 has joined #timvideos | 14:53 | |
| *** mauz555_ has joined #timvideos | 15:02 | |
| *** mauz555 has quit IRC | 15:03 | |
| *** mauz555 has joined #timvideos | 15:09 | |
| *** mauz555_ has quit IRC | 15:11 | |
| *** rohitksingh has joined #timvideos | 15:35 | |
| *** rohitksingh has quit IRC | 15:53 | |
| *** rohitksingh has joined #timvideos | 16:00 | |
| rohitksingh | cr1901_modern: looks like your patches in OpenOCD have been merged now! :) | 16:10 |
| cr1901_modern | rohitksingh: Holy smokes, so they are :o... | 16:13 |
| *** rohitksingh has quit IRC | 16:13 | |
| *** rohitksingh1 has joined #timvideos | 16:13 | |
| *** rohitksingh1 has joined #timvideos | 17:11 | |
| *** rohitksingh1 has quit IRC | 17:14 | |
| *** rohitksingh has joined #timvideos | 17:15 | |
| *** rohitksingh has quit IRC | 17:17 | |
| *** rohitksingh has joined #timvideos | 17:17 | |
| *** rohitksingh1 has joined #timvideos | 17:21 | |
| *** rohitksingh has joined #timvideos | 17:28 | |
| *** rohitksingh1 has quit IRC | 17:33 | |
| *** mauz555 has quit IRC | 17:37 | |
| *** rohitksingh has quit IRC | 17:48 | |
| *** rohitksingh has joined #timvideos | 17:51 | |
| *** rohitksingh has quit IRC | 17:56 | |
| *** rohitksingh has joined #timvideos | 17:56 | |
| *** rohitksingh has quit IRC | 18:26 | |
| *** rohitksingh has joined #timvideos | 18:28 | |
| *** Kripton has quit IRC | 18:47 | |
| *** Kripton has joined #timvideos | 18:57 | |
| *** tpb has joined #timvideos | 19:35 | |
| mithro | nrossi: Thanks! | 19:49 |
| *** rohitksingh has joined #timvideos | 19:52 | |
| *** rohitksingh has quit IRC | 20:41 | |
| CarlFK | mithro: you wrote this for me in .. 2012 or something: https://github.com/CarlFK/veyepar/blob/master/dj/scripts/fixunicode.py | 21:06 |
| tpb | Title: veyepar/fixunicode.py at master · CarlFK/veyepar · GitHub (at github.com) | 21:06 |
| CarlFK | in #positivepython techalchemy: wow that is super awesome, mind if i swipe it? | 21:07 |
| CarlFK | me: yes | 21:07 |
| CarlFK | here to pass on the thanks | 21:07 |
| mithro | CarlFK: Ha - I'm not sure I actually wrote that... | 21:08 |
| mithro | CarlFK: https://github.com/CarlFK/veyepar/commit/dc33c66bb154501988ed72ee32a920e8feebb835 | 21:09 |
| tpb | Title: Making fixunicode more robust and also fix stderr. · CarlFK/veyepar@dc33c66 · GitHub (at github.com) | 21:09 |
| CarlFK | 2014 - close enough | 21:09 |
| mithro | CarlFK: https://github.com/CarlFK/veyepar/commit/d38e4c6cc379697d2607dfe3fae69fa44fe105e4#diff-2031d030bedffc622b376f32b4fd0151 -- looks like that came from you? | 21:09 |
| tpb | Title: print unicode chars out to term · CarlFK/veyepar@d38e4c6 · GitHub (at github.com) | 21:09 |
| mithro | I just cleaned it up | 21:10 |
| CarlFK | I'm guessing you wrote it and I commited it | 21:10 |
| mithro | CarlFK: I think it came from https://github.com/mithro/rcfiles/commits/master/python/force_unicode.py ? | 21:11 |
| tpb | Title: History for python/force_unicode.py - mithro/rcfiles · GitHub (at github.com) | 21:11 |
| CarlFK | that explains how you wrote it as fast as you did ;) | 21:12 |
| cr1901_modern | https://github.com/enjoy-digital/litex/issues/117#issuecomment-433655547 Huh... is the SPI flash core really that small... | 21:18 |
| cr1901_modern | even a minimal UART is like 100 LUTs | 21:18 |
| *** sc00bz1 has quit IRC | 23:09 | |
| *** sc00bz has joined #timvideos | 23:14 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!