| stekern | Ok, llvm jump tables is now working; llvm is slightly slower (45.6 iter/sec) in coremark than gcc4.5.1 (47.7 iter/sec) and gcc4.8.0 (48.3 iter/sec) | 06:28 |
|---|---|---|
| Jia | does linux-or32 still need lzop? | 08:21 |
| jeremybennett | stekern: Good progress | 11:00 |
| juliusb | Jia: lzop? | 15:49 |
| juliusb | stekern: which bit about the sim didn't you get? how to use it? | 15:49 |
| juliusb | i haven't looked at exaclt how he's done it, but in minei had to hack in a few bits and pieces to make it start up correctly | 15:50 |
| stekern | no, how to properly use the 'cd' command (see ml) :) | 15:50 |
| stekern | I have all my build directories as build-<src_tree_name> | 15:51 |
| stekern | and I had cd'ed out from build-or1k-dejagnu and was suppose to cd into build-or1k-gcc, but I had cd'ed back into build-or1k-dejagnu... | 15:52 |
| stekern | after correcting my mistake it runs fine | 15:55 |
| juliusb | ah cool | 16:21 |
| stekern | # of unexpected failures 2354 | 16:31 |
| stekern | I don't know how that compares to peter's results | 16:32 |
| stekern | ah, that was libstdc++ | 16:35 |
| juliusb | mmm i got my version down to 13 unexepected failures | 17:01 |
| stekern | do you got the link to the result page nearby? | 17:03 |
| stekern | can't seem to find it now | 17:03 |
| jeremybennett | juliusb: Looking good. Was that C and C++? Any unexpected passes or unresolved (usually timeouts)? | 17:03 |
| stekern | ah, found it | 17:04 |
| stekern | http://opencores.org/or1k/Newlib_tool_chain_test_results | 17:04 |
| juliusb | I haven't touched that for a few weeks | 17:27 |
| juliusb | never got to the bottom of them all but nothing seems fatal | 17:27 |
| juliusb | or nothing seems impossible | 17:27 |
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!