that is used to store node id and coreid to the apic path. for core0 and core1, if core1 apic id path is created there bus_cpu_scan...
Because LNXI create the core1 apic id path in core0 of amd_sibling_init, in that case need to store nodeid and coreid too...
I hope you can look the my northbridge.c to create core1 apic id patch in bus_cpu_scan ...
So You can disable CPU_INIT_SERIALIZE for big SMP system...
YH
On 11/23/05, Ronald G Minnich rminnich@lanl.gov wrote:
yhlu wrote:
a_s1_core.diff and a_s1_mb.diff
That went very badly:
[rminnich@q LinuxBIOSv2]$ patch -p1 < /tmp/a_s1_core.diff patching file src/arch/i386/boot/acpi.c patching file src/arch/i386/include/arch/acpi.h Hunk #3 succeeded at 114 with fuzz 1. Hunk #4 succeeded at 134 (offset 1 line). Hunk #6 succeeded at 285 (offset 1 line). patching file src/config/Options.lb Hunk #1 succeeded at 778 (offset -4 lines). patching file src/include/device/path.h patching file src/northbridge/amd/amdk8/amdk8_acpi.c patching file src/northbridge/amd/amdk8/Config.lb patching file src/northbridge/amd/amdk8/get_sblk_pci1234.c patching file src/northbridge/amd/amdk8/northbridge.c Hunk #1 succeeded at 692 with fuzz 2 (offset -102 lines). Hunk #2 succeeded at 810 (offset 1 line). Hunk #3 succeeded at 720 with fuzz 2 (offset -104 lines). Hunk #4 succeeded at 841 (offset 4 lines). Hunk #5 FAILED at 1033. 1 out of 5 hunks FAILED -- saving rejects to file src/northbridge/amd/amdk8/northbridge.c.rej patching file src/southbridge/amd/amd8111/amd8111_acpi.c Hunk #2 succeeded at 123 (offset 1 line). patching file src/southbridge/amd/amd8132/amd8132_bridge.c patching file src/southbridge/amd/amd8132/Config.lb
I looked atthe HUNKS that failed on northbridge.c and see no obvious way to fix it.
If anyone else wants to take a look, be my guest; I'm done for the night. good luck.
ron