Hi Denis, would you please comment also on the mail series with subject: [RFC PATCH v6 x/1] lte: protocol and authentication for default ctx which is the core atom for the lte series discussed here, so that I don't resubmit the series just for spam.
Also, I would really appreciate a comment on the series: [RFC PATCH x/1] atmodem/lte: Gemalto vendor specific extension and possibly the introduction of germaltomodem/voicecall. Series: [PATCH v3 x/2] Gemalto: support for voicecall thank you, Giacinto On Mon, Oct 15, 2018 at 7:30 PM Giacinto Cifelli <gciof...@gmail.com> wrote: > > Hi Denis & Jonas, > > On Mon, Oct 15, 2018 at 7:27 PM Denis Kenzior <denk...@gmail.com> wrote: > > > > Hi Jonas, > > > > >> + > > > > > > The above isn't wrong, but I question the need for the atomic > > > constructs. ofono isn't a thread-safe library. > > > > > > I see that there other uses of atomic accessors in the codebase, but I > > > believe these are remnants of a past approach with other ambitions... > > > Denis will set me right if I'm wrong about this because this mostly > > > predates my involvement with ofono. :) > > > > oFono was never thread safe, and as you point out, atomic operations are > > not really necessary. But glib uses atomics for reference counting, so > > we stuck to that. Also Marcel at some point evangelized that g_atomic > > is the one true way (TM) of implementing reference counting. > > > > With ell we don't use atomics at all. > > > > So given the above, I find both approaches acceptable. > > > > Regards, > > -Denis > > Thank you both. I think I will keep this as it is then, and change the rest. > I will re-submit the series later. > > Regards, > Giacinto _______________________________________________ ofono mailing list ofono@ofono.org https://lists.ofono.org/mailman/listinfo/ofono
