On 04.07.2017 22:00, maya%netbsd.org@localhost wrote: > On Tue, Jul 04, 2017 at 08:19:57PM +0200, Kamil Rytarowski wrote: >> I will reevaluate it once I will work on i386 port. As of no it >> introduces extra complications. >> > > (The reason I suggest this is that it was changed a few months ago) > > Is the header not C++-safe? it's including sys/mutex.h for some > reason... > Yes, there is a list of around six headers and required global define _KERNTYPES. Other mainstream systems tend to have different vm layout for i386 user on i386 kernel and i386 user on amd64 kernel... at the moment I don't want to dig into it. My preference would be to use a dynamic sysctl. There is one for VM_MAX_ADDRESS, and a missing one for VM_MAXUSER_ADDRESS and VM_MAXUSER_ADDRESS32 for netbsd32.
Attachment:
signature.asc
Description: OpenPGP digital signature