* On Mon, Jan 19, 2004 at 20:38:53 -0500, Jerry McBride wrote: > It seems odd that a few people are having problems with the 232r9 glibc. I > just rolled it out on three machines, 1 amdXP, 1 amdMobile an old k6-2 and a > k6-3 systems and no errors. All running 2.6.1 kernels, NPTL and the latest > gcc...
After the DEPEND/DEPEND problem[1] with glibc-2.3.2r9 and NPTL I emerged gcc-3.3.2-r5, trying to solve that problem (which it didn't). Now I see NPTL support has been removed from glibc-2.3.2r9, solving the DEPEND/DEPEND problem. My question is this: Will upgrading a "NPTLed glibc" to a "non NPTLed glibc" cause known problems (i. e. with applications that have been compiled with the "NPTLed glibc")? Are there any more users who broke their system upgrading this way? Is there someone who can tell us a bit more about it? Reading the reports from Eric and Dhruba at bugzilla[2] and on this mailinglist, I'm a bit scared of upgrading to glibc-2.3.2-r9 using a system set up as mine. I'd be very happy for some clarifications. ;) Thank you for your warnings in here, Dhruba! I wouldn't have recognized that issue and probably ran into the same problems. Regards, Jens Footnotes: ---------- [1] http://bugs.gentoo.org/show_bug.cgi?id=38622 [2] http://bugs.gentoo.org/show_bug.cgi?id=38619 PS: Wouldn't it have been a considerable solution to have the glibc ebuild correctly recognize a decent gcc (like 3.3.2-r5), maybe issuing some warnings while emerging with NPTL, or are there other (technical) reasons not to do this? -- The discerning person is always at a disadvantage. -- [EMAIL PROTECTED] mailing list
