On Oct 9, 2009, at 10:58 PM, David Johnson wrote:
Steven, Thanks for responding to my question. [1] I downloaded v5.0 using anon CVS from NetBSD.org[2] Following the NetBSD Guide, Chapter 31.3 Creating the kernel Configuration File, as a test, I simply copied the GENERIC kernel config file, as is to MYKERNEL with no changes in the same directory. I then attempted, as a sanity test, the config MYKERNEL command which is the next step to generate dependencies. The result was a bunch of lines of errors reported. This was my test case before venturing off into actually trying to change the config for real. Is anyone else able to confirm my findings or am I off in the weeds here?[3] MYKERNEL is in the same directory as GENERIC. I did not try running build.sh yet.
Again: what version of NetBSD are you trying this on? If you try to build a 5.0 kernel while running on 4.0 and using 4.0's config, I would *expect* errors. build.sh handles that for you automatically. Note the warning in 31.4.1 -- it applies to version differences as well as platform differences.
Thanks for any light you can shed on this problem, David Steven Bellovin wrote:On Oct 8, 2009, at 10:18 PM, David Johnson wrote:I am attempting to compile the i836 arch version 5.0 kernel in order to customize it to my group of machines.What version are you running on? Are you using build.sh? If not, you should; it will compile a 5.0 config for you.The other obvious question: is MYKERNEL in the same directory? There are often included files in GENERIC.I copied GENERIC to MYKERNEL and attempted to config it with the followingconfig MYKERNELand received a bunch of errors which prevented further progress. The kernel building Howto explains that these must be cleaned up before going to the next step.I would have thought that the existing GENERIC kernel config file would be suitable for reuse. Am I doing something wrong orwhat?If I have failed to note a previous response to this question, please excuse the faux pas.. I'm just getting started with this and am prone to stumble here and there.thanks, David Birmingham <electronsciences.vcf>--Steve Bellovin, http://www.cs.columbia.edu/~smb<electronsciences.vcf>
--Steve Bellovin, http://www.cs.columbia.edu/~smb