linux-kernel - Re: [PATCH v16 3/7] soc: mediatek: SVS: introduce MTK SVS engine

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]

Message-ID: <f64f2419217c0c8e0e269db0b480abb08040adb4.camel@mediatek.com>
Date:   Fri, 14 May 2021 13:58:01 +0800
From:   Roger Lu <roger.lu@...iatek.com>
To:     Guenter Roeck <linux@...ck-us.net>
CC:     Matthias Brugger <matthias.bgg@...il.com>,
        Enric Balletbo Serra <eballetbo@...il.com>,
        Kevin Hilman <khilman@...nel.org>,
        Rob Herring <robh+dt@...nel.org>,
        Nicolas Boichat <drinkcat@...gle.com>,
        Stephen Boyd <sboyd@...nel.org>,
        Philipp Zabel <p.zabel@...gutronix.de>,
        Fan Chen <fan.chen@...iatek.com>,
        HenryC Chen <HenryC.Chen@...iatek.com>,
        YT Lee <yt.lee@...iatek.com>,
        Xiaoqing Liu <Xiaoqing.Liu@...iatek.com>,
        "Charles Yang" <Charles.Yang@...iatek.com>,
        Angus Lin <Angus.Lin@...iatek.com>,
        "Mark Rutland" <mark.rutland@....com>, Nishanth Menon <nm@...com>,
        <devicetree@...r.kernel.org>,
        <linux-arm-kernel@...ts.infradead.org>,
        <linux-mediatek@...ts.infradead.org>,
        <linux-kernel@...r.kernel.org>, <linux-pm@...r.kernel.org>,
        <Project_Global_Chrome_Upstream_Group@...iatek.com>
Subject: Re: [PATCH v16 3/7] soc: mediatek: SVS: introduce MTK SVS engine

On Thu, 2021-05-13 at 20:33 -0700, Guenter Roeck wrote:
> On 5/13/21 8:10 PM, Roger Lu wrote:
> > Hi Guenter,
> > 
> > Sorry for the late reply and thanks for the notice.
> > 
> > On Wed, 2021-05-05 at 21:51 -0700, Guenter Roeck wrote:
> > > On Wed, Apr 28, 2021 at 02:54:36PM +0800, Roger Lu wrote:
> > > > The Smart Voltage Scaling(SVS) engine is a piece of hardware
> > > > which calculates suitable SVS bank voltages to OPP voltage
> > > > table.
> > > > Then, DVFS driver could apply those SVS bank voltages to
> > > > PMIC/Buck
> > > > when receiving OPP_EVENT_ADJUST_VOLTAGE.
> > > > 
> > > > Signed-off-by: Roger Lu <roger.lu@...iatek.com>
> > > > ---
> > > >   drivers/soc/mediatek/Kconfig   |   10 +
> > > >   drivers/soc/mediatek/Makefile  |    1 +
> > > >   drivers/soc/mediatek/mtk-svs.c | 1723
> > > > ++++++++++++++++++++++++++++++++
> > > >   3 files changed, 1734 insertions(+)
> > > >   create mode 100644 drivers/soc/mediatek/mtk-svs.c
> > > > 
> > > 
> > > [ ... ]
> > > 
> > > > +
> > > > +	svsp_irq = irq_of_parse_and_map(svsp->dev->of_node, 0);
> > > > +	ret = devm_request_threaded_irq(svsp->dev, svsp_irq,
> > > > NULL,
> > > > svs_isr,
> > > > +					svsp->irqflags, svsp-
> > > > >name,
> > > > svsp);
> > > 
> > > 0-day reports:
> > > 
> > > drivers/soc/mediatek/mtk-svs.c:1663:7-32: ERROR:
> > > 	Threaded IRQ with no primary handler requested without
> > > IRQF_ONESHOT
> > > 
> > > I would be a bit concerned about this. There is no primary (hard)
> > > interrupt handler, meaning the hard interrupt may be re-enabled
> > > after
> > > the default hard interrupt handler runs. This might result in
> > > endless
> > > interrupts.
> > 
> > Oh, we add IRQF_ONESHOT in "svs_get_svs_mt8183_platform_data()" for
> > threaded irq. So, please kindly let us know if we need to set more
> > flags or any other potential risks we should be aware. Thanks in
> > advance.
> > 
> 
> After reviewing the code, I think this was actually a false alarm,
> at least if svsp->irqflags always includes IRQF_ONESHOT.
> The code is kind of unusual, though. Unless I am missing something,
> svsp->irqflags is only set in one place and it is always set
> to IRQF_TRIGGER_LOW | IRQF_ONESHOT. If there is a remote chance
> that the flag is ever different, it would have been better (and less
> confusing) to specify IRQF_ONESHOT directly when requesting the
> interrupt (because it is always needed, no matter which SOC).
> If the flags are always the same, there is no reason for having
> the svsp->irqflags variable in the first place.

Thanks for the advice. We still need svsp->irqflags for customizing
different SoC requirement as [1] and I'll specify IRQF_ONESHOT directly
as [2] in the next version.

[1]
mt8183 setting: svsp->irqflags = IRQF_TRIGGER_LOW | IRQF_ONESHOT;
mt8192 setting: svsp->irqflags = IRQF_TRIGGER_HIGH | IRQF_ONESHOT;
https://lore.kernel.org/patchwork/patch/1419309/ (mt8192)

[2]
ret = devm_request_threaded_irq(svsp->dev, svsp_irq, NULL, svs_isr,
				svsp->irqflags | IRQF_ONESHOT,
				svsp->name, svsp);
> 
> Thanks,
> Guenter