| --- Log opened Fri Dec 09 00:00:19 2016 | ||
| promach | http://openrisc.io/newlib/multicore.html how many FPGA LUT resources would the multicore implementation will take approximately ? | 02:01 |
|---|---|---|
| promach | is it related to optimsoc ? | 02:02 |
| olofk | promach: wallento is the optimsoc guy. I don't have any numbers unfortunately | 04:00 |
| wallento | hey | 04:01 |
| wallento | it is not related to optimsoc | 04:01 |
| wallento | the baseline OpenRISC implementation (mor1kx) has SMP support | 04:01 |
| wallento | The only reason we did not make the SMP support for newlib mainline is that people may argue about performance losses for single cores | 04:02 |
| olofk | True. But I figured wou would have some LUT numbers :) | 04:02 |
| wallento | not from my head, but it is not much more than the standard mor1kx | 04:02 |
| wallento | let me check if I currently have a project on my disk | 04:03 |
| olofk | wallento: Hmm.. aren't those parts of newlib upstream? The reentrant functions are at least defined in newlib and reading the other code gave me the impression that it was all there | 04:03 |
| wallento | yes, but it is still a define you need to set to build it | 04:04 |
| wallento | CFLAGS="-D__OR1K_MULTICORE__" | 04:04 |
| wallento | otherwise it builds the default | 04:04 |
| wallento | like here: https://github.com/openrisc/newlib/blob/or1k/libgloss/or1k/crt0.S#L254 | 04:05 |
| wallento | okay, it points to the weakest point in the port :) | 04:06 |
| wallento | but anyways.. | 04:06 |
| promach | will the multicore implementation able to fit in Zynq7010 (Parallella Desktop) ? | 04:07 |
| LoneTech | I admit I don't quite get what that does. it reminds me of the isa bus pnp atrocity. | 04:07 |
| olofk | promach: You would likely be able to fit a few cores in the z7010 | 04:22 |
| promach | olofk: let me build the multicore toolchain first. How would I select the "number of cores" afterwards ? | 04:28 |
| promach | I have this error | 06:13 |
| promach | cfns.gperf:101:1: error: ‘const char* libc_name_p(const char*, unsigned int)’ redeclared inline with ‘gnu_inline’ attribute | 06:13 |
| promach | while building gcc-stage-2 | 06:14 |
| promach | seems related to earlier gcc version | 06:14 |
| promach | my gcc version is 6.2.1 | 06:14 |
| promach | anyone have the same problem ? | 06:47 |
| olofk | Any FuseSoC users on Windows here? Need some help trying out a patch | 07:04 |
| promach | should I use the upstream gcc ? | 07:47 |
| olofk | promach: upstream gcc doesn't have openrisc support | 08:04 |
| bandvig | olofk: I'm. But I'll able to help you on Saturday / Sunday | 08:07 |
| promach | olofk: so, need to patch gcc ? | 08:16 |
| wallento | promach: definitely | 08:22 |
| promach | wallento: how would I trace the bug ? I have no experience with debugging gcc ? | 08:30 |
| olofk | ok. Screw documentation. I'm releasing FuseSoC 1.6 anyway | 09:47 |
| promach | RISC-V already has patch https://groups.google.com/a/groups.riscv.org/forum/#!msg/sw-dev/WAy9qJwKamI/Tx0K66dVAQAJ | 10:12 |
| promach | upstream has no openRISC support ? | 10:13 |
| olofk | promach: I don't think RISC-V is upstream yet, is it? | 10:22 |
| promach | I have not used RISC-V before | 10:24 |
| promach | anyone have problem building gcc-stage-2 with gcc-6.x ? | 20:35 |
| --- Log closed Fri Dec 09 22:37:16 2016 | ||
| --- Log opened Fri Dec 09 22:38:43 2016 | ||
| -!- Irssi: #openrisc: Total of 48 nicks [0 ops, 0 halfops, 0 voices, 48 normal] | 22:38 | |
| -!- Irssi: Join to #openrisc was synced in 19 secs | 22:38 | |
| promach | olofk, wallento: who is actually working on openRISC repo ? | 23:37 |
| promach | how to patch the gcc ? | 23:37 |
| --- Log closed Sat Dec 10 00:00:20 2016 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!