system.cpu.num for OpenBSD (needs testing)

[freak]

Hi,

based on the changes I already posted in http://www.zabbix.com/forum/showthread.php?t=8116 I added system.cpu.num for OpenBSD. I copied my changes to trunk now, since the codebase in svn seems to be the same in trunk and tags/1.4.2/

sysctl on OpenBSD 4.1 doesn't seem to be able to differentiate between online and max, correct me on that if I'm wrong. I don't have a SMP environment to test it anyways.

So system.cpu.num[whatever] will always return the same value as sysctl -a hw.ncpu:

./src/libs/zbxsysinfo/openbsd/cpu.c

--- trunk-sra/src/libs/zbxsysinfo/openbsd/cpu.c 2007-10-18 19:47:39.000000000 +0200
+++ trunk/src/libs/zbxsysinfo/openbsd/cpu.c     2007-10-18 18:26:27.000000000 +0200
@@ -23,23 +23,7 @@

 int    SYSTEM_CPU_NUM(const char *cmd, const char *param, unsigned flags, AGENT_RESULT *result)
 {
-       int ret = SYSINFO_RET_FAIL;
-       int mib[2];
-       long ncpu;
-       size_t len;
-
-       mib[0] = CTL_HW;
-       mib[1] = HW_NCPU;
-
-       len = sizeof(ncpu);
-
-       if (sysctl(mib, 2, &ncpu, (size_t *) &len, NULL, 0) == 0)
-       {
-               SET_UI64_RESULT(result, ncpu);
-               ret = SYSINFO_RET_OK;
-       }
-
-       return ret;
+       return SYSINFO_RET_FAIL;
 }

 static int get_cpu_data(unsigned long long *idle,

So, please let me know what you think and what you find testing it..

Sebastian

PS: I forgot the "return ret;" at the end.