linux-kernel - RE: [PATCH v3 2/2] misc: Add iop driver for Sunplus SP7021

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: <f4d45bcf7ddd45ceaf8ab06e088c69b9@sphcmbx02.sunplus.com.tw>
Date:   Sat, 18 Dec 2021 17:30:52 +0000
From:   Tony Huang 黃懷厚 <tony.huang@...plus.com>
To:     gregkh <gregkh@...uxfoundation.org>
CC:     Arnd Bergmann <arnd@...db.de>,
        Tony Huang <tonyhuang.sunplus@...il.com>,
        Rob Herring <robh+dt@...nel.org>,
        DTML <devicetree@...r.kernel.org>,
        "Linux Kernel Mailing List" <linux-kernel@...r.kernel.org>,
        Derek Kiernan <derek.kiernan@...inx.com>,
        Dragan Cvetic <dragan.cvetic@...inx.com>,
        Wells Lu 呂芳騰 <wells.lu@...plus.com>
Subject: RE: [PATCH v3 2/2] misc: Add iop driver for Sunplus SP7021

Dear gregkh, arnd:


> >
> > > > > wrote:
> > > > > >
> > > > > > IOP (IO Processor) embedded inside SP7021 which is used as
> > > > > > Processor for I/O control, RTC wake-up and cooperation with
> > > > > > CPU & PMC in power management purpose.
> > > > > > The IOP core is DQ8051, so also named IOP8051, it supports
> > > > > > dedicated JTAG debug pins which share with SP7021.
> > > > > > In standby mode operation, the power spec reach 400uA.
> > > > > >
> > > > > > Signed-off-by: Tony Huang <tonyhuang.sunplus@...il.com>
> > > > >
> > > > > Thanks for the improvements, this again looks better than the
> > > > > previous
> > > version.
> > > > > I still have some minor comments, and there are a couple of
> > > > > details I have commented on before that would need to be
> > > > > addressed, but let's focus on the one main issue for now:
> > > > >
> > > > > The driver still doesn't actually /do/ anything: you load the
> > > > > firmware when the driver is loaded, and you shut it down when
> > > > > the driver is removed, but otherwise there is no way to interact
> > > > > with the iop. You had the miscdevice earlier, and you still
> > > > > register that, but there are no file_operations associated with
> > > > > it, so it still doesn't
> > > have any effect.
> > > > >
> > > > > In the original version you had a couple of user-side
> > > > > interfaces, for which Greg and I commented that they were not
> > > > > using the correct abstractions, and you still list them in the
> > > > > changelog text as "I/O control, RTC wake-up and cooperation with
> > > > > CPU & PMC in power
> > > management".
> > > > >
> > > > > If you want to make any progress with adding the driver, I'd say
> > > > > you should implement at least two of those high-level interfaces
> > > > > that interact with the respective kernel subsystems in order to
> > > > > show that the
> > > abstraction works.
> > > > >
> > > >
> > > > Q:"with respective kernel subsystems in order to show that the
> > > > abstraction
> > > works."
> > > > May I ask you about repective kernel subsystem.
> > > > If I use the file_operation method Provide user can read and write
> > > > IOP(8051)'s register.
> > > > Is this a repective kernel subsystem?
> > > > if not
> > > > There are other driver code can give me reference
> > > >
> > >
> > > I still do not understand what the goal of this driver is.
> > >
> >
> > When the poweroff command is executed.
> 
> What exactly do you mean by this?  The power off command that is sent to
> the kernel from userspace?
> 

I type "poweroff" in command line.

> > 1.The 8051 has a register to control the power-on and power-off of the
> system(Linux kernel).
> >  If you turn off the power through the 8051 register(DEF_PWR_EN_0=0),
> > The current measured by the circuit board is 0.4mA only. In order to save
> power.
> > 2.The power is not turned off through the 8051 register.
> >  The current measured on the circuit board is 33mA 3.When the system
> > linux kerenl is powered off. /driver/rtc, /driver/gpio cannot operate.
> >   8051 is still alive and operational
> >   8051 has RTC register. When the time is up, 8051 powers on the system
> >   The 8051 can detect GPIO0~7 pins, and GPIO pin high/low can be used as
> a power-on judgment mechanism for the system.
> 
> So what do you need to do?  When the kernel needs to shut down the system,
> send a message to this hardware device?  If so, great, just tie into the normal
> powerdown sequence, this is nothing new.
> 

Yes, 									
poweroff command ---> run misc/sunplus_ioc.c/sp_iop_platform_driver_shutdown()---->switch standby.bin for 8051--->setting 8051 PMC register---> sent a message to 8051			
Q:Why need switch standby.bin for 8051?
A: 8051 has two bin files, normal.bin and standby.bin in rootfs.
Normally, 8051 executes normal.bin code. Normal.bin code size can exceed 16K.
When system linux kernel is shutdown, 8051 is to execute standby.bin code.
Standby.bin code cannot exceed 16K bytes. Because 8051 Icache size is 16k
8051 runs with standby.bin code in the Icache before the system(linux kernel) is poweroff.
So I need misc/sunplus_iop.c interface to switch bin code and sends message to 8051.

Thanks