linux-kernel - Re: [PATCH 1/2] rtc/ia64: remove legacy efirtc driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAK8P3a28aRv5TpRn1xhjZFO9n=+CBG6V276Roqm=bE_eQZkw_w@mail.gmail.com>
Date:   Fri, 25 Oct 2019 12:41:49 +0200
From:   Arnd Bergmann <arnd@...db.de>
To:     "Luck, Tony" <tony.luck@...el.com>
Cc:     Alexandre Belloni <alexandre.belloni@...tlin.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        "linux-rtc@...r.kernel.org" <linux-rtc@...r.kernel.org>,
        Alessandro Zummo <a.zummo@...ertech.it>,
        "linux-ia64@...r.kernel.org" <linux-ia64@...r.kernel.org>,
        "Yu, Fenghua" <fenghua.yu@...el.com>,
        Stephane Eranian <eranian@...gle.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 1/2] rtc/ia64: remove legacy efirtc driver

On Thu, Oct 24, 2019 at 6:57 PM Luck, Tony <tony.luck@...el.com> wrote:
>
> > arch/ia64 has a read_persistent_clock64() function, so it ends up reading
> > the system time regardless of the RTC driver or CONFIG_RTC_HCTOSYS.
> >
> > As ia64 sets neither ARCH_HIBERNATION_POSSIBLE nor
> > ARCH_SUSPEND_POSSIBLE, so we could just remove the
> > read_persistent_clock64() and efi_gettimeofday(), relying instead
> > on user space (/sbin/hwclock) or CONFIG_RTC_HCTOSYS.
>
> Seems weird. ia64 has always assumed from day 1 that it is running
> on a UEFI capable platorm (well at day 1 it was called "EFI", the "U"
> came later).
>
> So read_persistent_clock64() just calls EFI directly to get the time.
>
> Seems simpler than worrying about having the right drivers and CONFIG
> bits set.

It would just be a little more consistent. Most architectures cannot
implement  read_persistent_clock64()  or CONFIG_RTC_HCTOSYS
when the RTC driver is a loadable module, so distros normally have to
set the system time after loading modules already. If some architectures
have a reliable platform interface that allows setting the time at early
boot, that doesn't mean we have to rely on that.

       Arnd