bug#329: linking against host libs
Nate Riffe <[email protected]>
| Newsgroups | gmane.linux.bbc.bugs |
|---|---|
| Message-ID | <[email protected]> |
[14:08] <deejoe> anybody have a bead on the hackedbox bug? [15:56] <CrackMonkey> deejoe: inkblot is working on it, but it's slow going [15:58] <@inkblot> WHAT BUG [15:59] <CrackMonkey> the bug that hackedbox uses the wrong libstdc++ [16:00] <@inkblot> well, first of all, [16:00] <@inkblot> it's not a problem with hackedbox [16:03] <@inkblot> we just need a known-good build environment [16:03] <@inkblot> and the set of known-good build environments is currently very very small [16:03] <CrackMonkey> we can build in the chroot [16:06] <CrackMonkey> proposal: NEW UPSTREAM VERSION [16:07] <@inkblot> on the proposal to rev glibc (please respond "ja" or "nee"): [16:08] <@inkblot> does this mean we'll also need to rev gcc? [16:08] <@psykoyiko> the chroot has the same version of libstdc++ as gargoyle [16:08] <@inkblot> ja [16:08] <CrackMonkey> ja The fallout from my development toward bottled building is that it's not a problem which has ever been generally solved. At this point, our build system is for the most part up to the task of performing a bottled build, but the tools themselves are not. This means that for at least the short term, linking against host libraries is inevitable. In the short term, our best bet is version parity with the build host. -- --< ((\))< >----< [email protected] >----< http://www.movealong.org/ >-- "I just feel bad for the SUVs... forced to live all cooped up in the city like that." -- The Onion, Vol.39 Issue 07 pub 1024D/05A058E0 2002-03-07 Nate Riffe (06-Mar-2002) <[email protected]> Key fingerprint = 0DAC F5CB D182 3165 D757 C466 CD42 12A8 05A0 58E0