> Am 18.04.2014 22:12, schrieb Andrew Savchenko: > > On Mon, 14 Apr 2014 22:11:05 -0400 Walter Dnes wrote: > >> I'm (re)installing Gentoo on an older AMD notebook. The output > >> from "less /proc/cpuinfo" includes... > >> > >> processor : 1 > >> vendor_id : AuthenticAMD > >> cpu family : 16 > >> model : 6 > >> model name : AMD Athlon(tm) II P320 Dual-Core Processor > >> stepping : 3 > >> microcode : 0x10000b6 > >> cpu MHz : 2100.000 > >> cache size : 512 KB > >> physical id : 0 > >> siblings : 2 > >> core id : 1 > >> cpu cores : 2 > >> apicid : 1 > >> initial apicid : 1 > >> fpu : yes > >> fpu_exception : yes > >> cpuid level : 5 > >> wp : yes > >> flags : fpu vme de pse tsc msr pae mce cx8 apic sep > >> mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall > >> nx mmxext fxsr_opt pdpe1gb rdtscp lm 3dnowext 3dnow constant_tsc > >> rep_good nopl nonstop_tsc extd_apicid pni monitor cx16 popcnt > >> lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a 3dnowprefetch > >> osvw ibs skinit wdt nodeid_msr hw_pstate npt lbrv svm_lock > >> nrip_save bogomips : 4189.59 TLB size : 1024 4K > >> pages > >> > >> Now for the questions > >> > >> * In "make menuconfig", I'm not sure which of 2 CPU options to > >> select in Processor type and features ---> > >> Processor family (*****************) ---> > >> > >> ( ) Athlon/Duron/K7 > >> ( ) Opteron/Athlon64/Hammer/K8 > > I'd like to recommend you this kernel gcc patch which enables > > -march=native support for kernel compilation: > > https://github.com/graysky2/kernel_gcc_patch > > > > Just select native and you will get both best performance and one > > less headache. > > or it creates code that is much slower or breaks in subtle and hard to > catch ways. > > The kernel devs are very astute when it comes to gcc options - I > wouldn't screw around with them. If an app is crashy, disturbing but not > a big problem. If the kernel decides to write the wrong stuff all over a > partition boundary you are screwed. > > > > Best regards, > > Andrew Savchenko > >
Hi Volker Armin Hemmann, Have you encountered -march=native related bugs? I haven't had any problems on mainstream amd/intel cpus. With regards to the kernel I would expect more problems from an aggressive -O flag than that of -march since -march just indicates the instruction set the compiler can use. That being said if the goal is stability I wouldn't be applying unofficial patches. Regards, Dragostin Yanev