--- Log opened Fri Jun 24 00:00:07 2016 | ||
wallento | shorne: I don't know how to handle amending to #9 the best. I opened a PR (#11) with the revert commit. I was thinking we can just add the fixed commit to this one or amend, but I don't know how exactly :) | 02:18 |
---|---|---|
shorne | wallento: its ok, ill sort out | 05:51 |
shorne | trying to remember why I thought that was the best change at the time | 05:56 |
shorne | ok, so all my change does it change global contstructor/destructor symbol for (more sofisticated linkers) to use __libc_init_array instead of old _init | 07:01 |
shorne | if possible | 07:01 |
shorne | it shouldnt cause those warnings about array types | 07:02 |
shorne | looking into | 07:02 |
shorne | Interesting, my change is in newlib, these warnings are in gcc-phase1, which doesnt depend on newlib | 07:21 |
olofk | That's strange | 07:22 |
olofk | Can we check further back if the warnings were there before the __libc_init_array patch? | 07:22 |
shorne | yeah, looking at it | 07:24 |
shorne | just was checking a few thing, slowly :) | 07:24 |
shorne | Maybe this is a thing, the newlib travis build does or1k-gcc checkout. Which takes default branch I guess with is or1k | 07:38 |
shorne | I have been working with or1k-5.3.0 | 07:38 |
shorne | I dont see this issues with or1k-5.3.0 | 07:39 |
shorne | let me see with or1k | 07:39 |
shorne | But, I think we might want to make or1k-5.3.0 the default branch? | 07:39 |
shorne | or merge it into or1k | 07:39 |
_franck_ | anyone is going to ELCE ? | 07:40 |
shorne | yup... I get all those errors in gcc or1k branch but not or1k-5.3.0 (not even using --with-newlib) | 07:43 |
shorne | I dont think its related to my changes, unless its related to some recent binutils changes. | 07:45 |
shorne | but I dont think so | 07:45 |
wallento | thanks, shorne! | 08:10 |
wallento | we should move gcc to 5.3.0 | 08:10 |
wallento | anything against it? | 08:10 |
shorne | not from me, but anyone else using 5.3.0? | 08:10 |
shorne | no one has complained either way | 08:11 |
wallento | okay, lets wait a day or two | 08:14 |
wallento | blueCmd, olofk, stekern? | 08:14 |
blueCmd | go ahead I say. GCC is up to what, 6 now? | 08:15 |
blueCmd | it's like a year I rebased gcc for the 5 series | 08:16 |
blueCmd | since I* | 08:16 |
wallento | okay, thanks, I put up the recent 6 snapshots too | 08:21 |
wallento | but I think 5.3.0 is the favorable stable version | 08:21 |
shorne | brb, hopefully | 08:41 |
shorne | 0 | 09:17 |
shorne | wallento: it seems my openrisc@libracores subscription never got setup, I am sure I requested before, and just requested now. Anything I do wrong? | 09:37 |
wallento | You already sent there, right? Let me check | 09:38 |
wallento | You are not on the members list | 09:39 |
wallento | did you check your spam? | 09:39 |
wallento | I can subscribe you manually anyways | 09:39 |
shorne | checked now it worked this time, maybe I never subscribed the first time. | 09:42 |
shorne | it took a while for the mail to come back | 09:42 |
shorne | also, its going to "Promotions" folder strangely | 09:43 |
wallento | at least it worked now :) | 09:44 |
wallento | by the way, are you planning to come to ORCONF? | 09:44 |
shorne | I am not planning yet. I really like Italy though | 09:45 |
wallento | would be great to meet you there. but its quite a distance, right? | 09:51 |
shorne | Yes, from Tokyo its quite a way. It would be nice to meet everyone I am sure there will be some good discussions | 09:53 |
shorne | There is an Air France flight that looks reasonable | 09:55 |
olofk | wallento: I have no preferences to which gcc version we should base the or1k branch on. I only use Turbo Pascal :) | 18:48 |
olofk | Did you sort out all the patch stuff btw? | 18:49 |
--- Log closed Sat Jun 25 00:00:09 2016 |
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!