Jeremie Courreges-Anglas writes:
> Maybe something like this? The idea is to point people to ubsan_minimal
> instead of assuming that ubsan isn't supported at all. I can't see
> a better place to document this.
>
> For more context: UBSan is supposed to catch undefined behavior at
> runtime and
On Tue, Feb 15 2022, Jeremie Courreges-Anglas wrote:
[...]
>>> Index: clang-local.1
>>> ===
>>> RCS file: /home/cvs/src/share/man/man1/clang-local.1,v
>>> retrieving revision 1.22
>>> diff -u -p -p -u -r1.22 clang-local.1
>>> --- cl
Hi Florian,
just for completeness-sake, I've retested with snapshot #337* Wed Feb 16
09:33:31 MST 2022 that just hit ftp.eu.openbsd.org and would like to confirm
that your change works as expected. No NULs around anymore, /etc/myname is
clean, smtpd is happy too.
big thanks again!
JV
* mornin
The following patch makes audio(4) event filters MP-safe.
The knote lists are already serialized by audio_lock. By adding klist
init calls, the lock becomes visible to the lock assertion in knote().
The new f_modify and f_process callbacks cover the knote handling
with audio_lock. This also makes
On Tue, Feb 15, 2022 at 10:58:02PM +0100, Mark Kettenis wrote:
> > +int
> > +mpfiic_i2c_send_stop(void *cookie, int flags)
> > +{
> > + struct mpfiic_softc *sc = cookie;
> > +
> > + sc->sc_start_sent = 0;
> > +
> > + HWRITE4(sc, I2C_CTRL, sc->sc_ctrl | I2C_CTRL_STO);
> > +
> > + /* Let a fe
