Re: Re: Debina install
Dale Stein <[email protected]> Thu, 06 Jan 2005 18:13:49 -0600
| Newsgroups | gmane.linux.real-time.rtlinux.general |
|---|---|
| Message-ID | <[email protected]> |
That makes sense. Did a little checking and RH7.2 uses glibc-2.2.2, RH-8.0 uses glibc-2.2.93 and RH-9.0 uses glibc-2.3.2-11.9. I don't know if anyone has made a successful RTlinux build on RH-8, 9 or other destros which use ?? glibc version. I do know I can get the linux kernel and Rtlinux to compile with some of the newer gcc versions ie. 3.2 3.3, and newer glibc versions, but they would fail the regression test. I can get the the linux kernel and RTlinux to compile with gcc-2.95.4 and glibc-2.3.1, 2.3.2, 2.3.3 - but fails the regression test....... A glibc-2.2.(low subversion) and gcc-2.95.3 should work. I do know that changing glibc on a working linux system can cause bad bad problems. (sorry for taking so much on this list, it might just all be all hot air, but it works for me!) Dale Stein On Tue, 4 Jan 2005 02:35:00 -0500, Norm <[email protected]> wrote: > I have had inconsistent success using the 2.96.x compiler that ships with > RedHat 7.2 but never a single failure with 2.95.3. > > I remember reading in kernel mailing list messages that it is the > preferred > 2.x.y version of the compiler for the kernel. > > YMMV > > Norm > > ----- Original Message ----- > From: "Dale Stein" <[email protected]> > To: <[email protected]> > Sent: Monday, January 03, 2005 7:21 PM > Subject: [Rtl] Re: Debina install > > >> For a long time I have been unable to compile and get a successful >> regression test with >> rtlinux3.2-per3. I have used all of the kernels that were patchable with >> the rtlinux patch and nothing really worked. So a few months ago I >> made >> a old lfs v3.3 on a spare partition. >> This version uses gcc 2.95.3 (which is what I used before), kernel >> 2.4.22 >> and >> glibc-2.2.5. The only real difference is glibc, I think, well, there >> are >> many others, but >> glibc and gcc are hand in hand generating the final code. >> So I am wondering weather failed regression tests are the result of > the >> glibc version used, and not just the gcc version used? >> Dale Stein >> >> >> -- >> Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/ >> _______________________________________________ >> Rtl mailing list >> [email protected] >> http://www2.fsmlabs.com/mailman/listinfo/rtl > > _______________________________________________ > Rtl mailing list > [email protected] > http://www2.fsmlabs.com/mailman/listinfo/rtl -- Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/ _______________________________________________ Rtl mailing list [email protected] http://www2.fsmlabs.com/mailman/listinfo/rtl