| 13:05:56 | mjn3 | sjhill: you did the dynamic lib stuff for mips, right? |
| 13:10:11 | sjhill | mjn3: yes |
| 13:11:24 | mjn3 | did erik mention to you that the new tests in test/ldso segfault on his mips box? they're something i cobbled together to illustrate the problem i was having with perl, and when i ran them on his mips box... |
| 13:12:14 | sjhill | no, he didn't mention anything to me about it |
| 13:12:22 | mjn3 | the makefile needs cleaning up... they require thread support. |
| 13:12:49 | mjn3 | i sent him the tests to look at and he just kind of threw them into test/ldso. |
| 13:12:55 | sjhill | the 'test/ldso/makefile' needs cleaning? |
| 13:13:36 | sjhill | what needs thread support, mips? |
| 13:15:07 | mjn3 | no... |
| 13:15:27 | mjn3 | the fact that the makefile needs cleaning up is a tangent... sorry to confuse you |
| 13:16:39 | mjn3 | the perl problem is that the perl thread support shared libs aren't linked against libpthread although perl is, and when those libs are loaded the libc weak equivalents of the pthread functions are being used instead of the real functions from libpthread |
| 13:16:41 | mjn3 | but.... |
| 13:16:58 | mjn3 | when i ran dlttest on eriks mips box, it segfaulted |
| 13:17:19 | mjn3 | it was the segfault behavior that i wanted to alert you too |
| 13:17:23 | mjn3 | s/too/to/ |
| 13:23:44 | sjhill | ok, i have to do a bunch of uclibc work over the next few days, so i'll take a look |
| 13:25:42 | mjn3 | just thought you'd like to know, and wasn't sure erik had mentioned it to you |