Re: Build failure in usr.sbin/bhyvectl; sources at main-n263112-ad513b4dba3e

On 24 May 2023, at 13:25, David Wolfskill <david@catwhisker.org> wrote:
> 
> This is from an in-place source update from main-n263073-634a770a5e16 to
> main-n263112-ad513b4dba3e:
> 
> ...
> Building /common/S4/obj/usr/src/amd64.amd64/lib/libc/tests/stdio/scanf_test
> Building /common/S4/obj/usr/src/amd64.amd64/usr.bin/ncurses/dump_entry.o
> Building /common/S4/obj/usr/src/amd64.amd64/usr.bin/mkimg/tests/img-1x1-4096-mbr.vhdx
> Building /common/S4/obj/usr/src/amd64.amd64/cddl/usr.sbin/zfsd/zfsd.full
> /usr/src/usr.sbin/bhyvectl/bhyvectl.c:2389:24: error: incompatible pointer types passing 'struct vm_exit *' to parameter of type 'struct vm_run *' [-Werror,-Wincompatible-pointer-types]
>                error = vm_run(vcpu, &vmexit);
>                                     ^~~~~~~
> /common/S4/obj/usr/src/amd64.amd64/tmp/usr/include/vmmapi.h:158:46: note: passing argument to parameter 'vmrun' here
> int     vm_run(struct vcpu *vcpu, struct vm_run *vmrun);
>                                                 ^
> 1 error generated.
> 
> 
> Given that yesterday's update was uneventful, and that
> src/usr.sbin/bhyvectl/bhyvectl.c has not changed in several days,
> I'm guessing that perhaps a recent change to a header, possibly
> involving vmexit, may be at issue here.

It appears to be broken since "vmm: Avoid embedding cpuset_t ioctl ABIs":
https://cgit.freebsd.org/src/commit/?id=e17eca327633efc511450310afb5e4a662724e3d

See also https://ci.freebsd.org/job/FreeBSD-main-amd64-build/, where it shows an error since https://ci.freebsd.org/job/FreeBSD-main-amd64-build/26653/

-Dimitry