Most recent kernel where this bug did not occur: CentOS 5 Distribution: CentOS Hardware Environment: Xeon Software Environment: CentOS 5 Problem Description: Hi All, Sorry for reporting a CentOS (RHEL) bug here. I had originally reported it on CentOS's site, but they did not take responsibility (said to report it "upstream", which I presume is you guys): http://bugs.centos.org/view.php?id=2386 When compiling a new kernel with "make menuconfig", the processor section does not adequately state which option is for the Xeon Dual Core processors. The closest notations are: Pentium-4/Celeron(P4-based)/Pentium-4 M/Xeon Pentium-III/Celeron(Coppermine)/Pentium-III Xeon Being that the Xeon Dual Core is based on the Pentium III, not the Pentium 4, Pentium III is probably the correct choice, BUT WHO CAN TELL! Steps to reproduce: make menuconfig
This is not really a problem, this is still a Xeon. You can compile with any option and it will actually try to find the right one for you. But I guess a little clarification wouldn't hurt. Maybe, we'll ask Intel.
yes, it's based on piii, but it itself has an optimization option in gcc. I don't know which kernel you are using, but latest kernel have a config option for 'Core 2/new Xeon', so at least it's not a upstream bug.
> yes, it's based on piii, but it itself has an optimization option in gcc. I > don't know which kernel you are using, but latest kernel have a config option > for 'Core 2/new Xeon', so at least it's not a upstream bug. If it is not in mainline kernel then I suggest to close the bug. /Sam
Even you didn't select the correct optimization option for gcc, it doesn't raelly matter to me. If there are optimization space, kernel usually will do it by itself, like write some code with mmx instruction. There isn't too many space left for gcc. I'll mark this as invalid as suggested.