Re: i386 9.99.17 build fails for NET4501 kernel

[Maxime Villard <max%m00nbsd.net@localhost>]

Le 14/10/2019 à 01:01, Chavdar Ivanov a écrit :
> My amd64 build also failed with
> ...
> src/tests/lib/libnvmm/h_mem_assist.c:178:2: error: missing initializer
> for field 'off' of 'const struct test'
> [-Werror=missing-field-initializers]
> ...
> (and many more).

Ah yes, I was doing local builds of it, not the in-src ones.

> On Sun, 13 Oct 2019 at 23:48, matthew green <mrg%eterna.com.au@localhost> wrote:
> > Robert Swindells writes:
> > > "John D. Baker" <jdbaker%consolidated.net@localhost>
> > > > Building the NET4501 kernel with 9.99.17 sources fails with:
> > > >
> > > > /x/current/src/sys/arch/x86/x86/vm_machdep.c: In function 'cpu_lwp_fork':
> > > > /x/current/src/sys/arch/x86/x86/vm_machdep.c:134:21: error: variable 'pcb2' might be clobbered by 'longjmp' or 'vfork' [-Werror=clobbered]
> > > >   struct pcb *pcb1, *pcb2;
> > > >                      ^~~~
> > >
> > > My custom amd64 kernels fail too, there was a failure report from
> > > someone else but it doesn't seem to be in the archive yet.
> > >
> > > The amd64 GENERIC config does build though.
> >
> > i don't know what is up yet, but maxv changed this file recently.
> > i wonder if there was/is something special about either SVS here,
> > or fpu_lwp_fork() it's triggering on.  i don't see any attributes
> > that should apply here.. guessing SVS as it is the only conditional
> > compile there that should matter, but i still don't see anything
> > that actually points to it.
> >
> > Maxime?

I don't understand what this error means. There is nothing related to
longjmp in this code or any code I've touched, and the FPU code does
not rely on that. As well I don't see what difference SVS could make
since it doesn't rely on that either.

No idea, looks like a compiler bug.