linux-kernel - Re: [PATCH 3/3] RTC: Update seconds time programming logic

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

Date:	Wed, 20 Apr 2016 14:02:21 +0200
From:	Alexandre Belloni <alexandre.belloni@...e-electrons.com>
To:	Anurag Kumar Vulisha <anurag.kumar.vulisha@...inx.com>
Cc:	Alessandro Zummo <a.zummo@...ertech.it>,
	Soren Brinkmann <sorenb@...inx.com>,
	Michal Simek <michals@...inx.com>,
	"rtc-linux@...glegroups.com" <rtc-linux@...glegroups.com>,
	"linux-arm-kernel@...ts.infradead.org" 
	<linux-arm-kernel@...ts.infradead.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Punnaiah Choudary Kalluri <punnaia@...inx.com>,
	Anirudha Sarangi <anirudh@...inx.com>,
	Srikanth Vemula <svemula@...inx.com>,
	Srinivas Goud <sgoud@...inx.com>
Subject: Re: [PATCH 3/3] RTC: Update seconds time programming logic

On 20/04/2016 at 10:31:06 +0000, Anurag Kumar Vulisha wrote :
> > Yeas, I understood that. But my question was whether the interrupt handling
> > was necessary at all.
> > Instead of waiting for an interrupt to set time_updated, can't you simply read
> > RTC_INT_STS and check for the RTC_INT_SEC bit in
> > xlnx_rtc_read_time() ?
> >
> > Something like:
> >
> > status = readl(xrtcdev->reg_base + RTC_INT_STS) if (status & RTC_INT_SEC)
> >       rtc_time64_to_tm(readl(xrtcdev->reg_base + RTC_CUR_TM), tm);
> > else
> >       rtc_time64_to_tm(readl(xrtcdev->reg_base + RTC_SET_TM_RD) - 1,
> > tm);
> >
> > It all depends on whether the RTC_INT_SEC bit in RTC_INT_STS is being
> > updated even when it is not enabled as an interrupt.
> >
> 
> The above said logic will work if we doesn't clear the RTC_INT_STS register after the
> RTC_INT_SEC bit is set, this happens only if interrupts are not enabled. If interrupts
> are enabled we will be clearing the RTC_INT_STS every time in the interrupt handler.
> And moreover we need to return time from RTC_SET_TM_RD only if time is requested
> within 1 sec span after programming the time only , so this is required only for one time.
> Since we are clearing the RTC_INT_STS in our interrupt handler, we might end up in giving
> the wrong time to the user when requested.So I think this logic might not work.
> Please correct me if am wrong.
> 

Simply stop clearing RTC_INT_SEC from RTC_INT_STS in the interrupt
handler.

You can also remove
	if (status & RTC_INT_SEC)
		rtc_update_irq(xrtcdev->rtc, 1, RTC_IRQF | RTC_UF);

as it will not be used. Update interrupts are handled by the core using
timers anyway. And as you can see, there was no code enabling
RTC_INT_SEC in RTC_INT_EN before your patch.

-- 
Alexandre Belloni, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com