IRC logs for #openrisc Thursday, 2016-04-21

--- Log opened Thu Apr 21 00:00:31 2016
jeremybennettmithro: wallento: Happy to help with GDB as far as I can. IIRC I completely rewrote the openrisc bit, then _franck_/_franck__ added lots of goodness. So should be no significant legal issues/04:27
jeremybennett(my rewrite was the GDB 6.8 version, as documented here: http://www.embecosm.com/resources/appnotes/#EAN304:27
mithroolofk: the rebootstrap debian guys are blocked on you needing to patch GCC04:28
mithrojeremybennett: I'm more an end user :)04:29
mithrojeremybennett: should i be able to single step the or1k?04:29
mithro(Instruction step that is)04:32
_franck__mithro: you should04:33
olofkmithro: What kind of patch?04:43
olofkAh, you mean that the don't want to use a non-upstreamed version?04:44
olofkYes, we're banging our heads against the wall here04:44
mithroolofk: yes, they only want patches to be a couple of lines, preferably no patches at all04:47
olofkmithro: Well, we would like the same :)04:49
mithroYeah, stupid CAs05:05
olofk+ stupid people who won't sign them05:05
shorneolofk: maybe silly question, I see that wb_stream has a .core file in your git repo.  I created one in orpsoc-cores and it works there. but is it possible to source directly the core from your github repo?05:27
shornewb_streamer05:27
shorneIt seems the one in olofk/wb_streamer is in an older format, and also it looks like there are some dependency issues05:28
olofkshorne: Yes, there are two options05:34
olofkClone the repo and then...05:34
olofkAdd the path to the repo in fusesoc.conf05:34
olofkor Run fusesoc with --cores-root=/path/to/repo05:34
olofkRegarding fusesoc.conf, FuseSoC first checks if it exists in the current directory, and if not, in ~/.local/share/fusesoc05:35
olofkSo you can create a workspace directory, copy your fusesoc.conf from ~/.config/fusesoc/fusesoc.conf there, and add the path05:36
olofkoh... .config is the current basedir. Not .local/share05:36
olofks/current/correct05:37
* olofk needs coffee05:37
olofkshorne: And yes, there is a dependency problem. I need to migrate wb_streamer to use the new wb_bfm, but haven't gotten around to do that ye05:38
olofkt05:38
olofkThis only affects the testbench05:38
shorneok, I can do that05:47
shornethanks a lot05:59
shorneolofk: Its working, I guess you want to update to wb_bfm-1.0 ?06:26
shorneand maybe stream_utils-1.1?06:26
shornebut those both pull in different versions of vlog_tb_utils, which causes some issues of dual definitions06:30
olofkshorne: Yes, versioning is still a mess. I was hoping that some GSoC student would like to take a look at that :)07:14
olofkWe are working on it however, and things will improve soonish07:14
olofkBut for now, I suggest always use vlog_tb_utils-1.0. I've updated alll cores in orpsoc-cores to use that version07:15
olofkAh, now I remember. I got it to work with a newer version of wb_bfm, but I found a bug, so I would need to bump wb_bfm first, and then we're back in depdenceny hell :/07:16
mithroolofk: What do you do as a day job? _florent_ did some work with you on something?07:34
SMDwrkDoes mor1kx uarch differ lots compared to or1200? I need to decide which docs I should study. Thanks.09:12
shorneolofk: I got it working with wb_bfm-1.0, had to make 2 changes09:28
shorne1. remote include for _params.v09:28
shorne2. switch from MAX_BURST_LENGTH(1) to MAX_BURST_LEN(2)09:29
shornesomething in wb_bfm didnt alow MAX_BURST_LEN less than 2 because it was being used for some bus width calculations09:29
SMDwrkAnd where should I get recent documentation on openrisc? I've heard opencores is not in favor anymore10:00
olofkmithro: Working as a consulatant, mostly doing FPGA design10:40
olofkshorne: I got a fix for the last issue10:41
olofkSorry, no, there is a third issue that you might have missed10:41
olofkOr does the testbench pass?10:41
olofkshorne: Pushed a fix to https://github.com/olofk/wb_bfm10:43
olofkI had to do that to make the wb_streamer tb pass10:43
shorneolofk: the test bench was passed10:52
shornejust tested wb_stream_reader_tb10:52
-!- Netsplit *.net <-> *.split quits: blueCmd13:19
-!- Netsplit over, joins: blueCmd13:28
olofkshorne: With wb_bfm-1.0 ?15:05
shorneolofk: yes, but now I am using your latest. But it was passing before too20:55
shornehttps://gist.github.com/stffrdhrn/01bfdd40b3fd570c4ecfd3660f292bf620:55
shornetests were wb_stream_reader_tb though20:57
shorneoh, my orpsoc-cores was a bit old21:17
--- Log closed Fri Apr 22 00:00:32 2016

Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!