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: <20200527140027.elc4vkbrzjrq4gqh@mobilestation>
Date:   Wed, 27 May 2020 17:00:27 +0300
From:   Serge Semin <Sergey.Semin@...kalelectronics.ru>
To:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
CC:     Serge Semin <fancer.lancer@...il.com>,
        <devicetree-compiler@...r.kernel.org>,
        Alexey Malahov <Alexey.Malahov@...kalelectronics.ru>,
        Thomas Bogendoerfer <tsbogend@...ha.franken.de>,
        Wolfram Sang <wsa@...-dreams.de>,
        Jarkko Nikula <jarkko.nikula@...ux.intel.com>,
        Frank Rowand <frowand.list@...il.com>,
        Rob Herring <robh+dt@...nel.org>, <devicetree@...r.kernel.org>,
        <linux-i2c@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] check: Add 10bit/slave i2c reg flags support

On Wed, May 27, 2020 at 04:36:56PM +0300, Andy Shevchenko wrote:
> On Wed, May 27, 2020 at 03:25:25PM +0300, Serge Semin wrote:
> > Recently the I2C-controllers slave interface support was added to the
> > kernel I2C subsystem. In this case Linux can be used as, for example,
> > a I2C EEPROM machine. See [1] for details. Other than instantiating
> > the EEPROM-slave device from user-space there is a way to declare the
> > device in dts. In this case firstly the I2C bus controller must support
> > the slave interface. Secondly I2C-slave sub-node of that controller
> > must have "reg"-property with flag I2C_OWN_SLAVE_ADDRESS set (flag is
> > declared in [2]). That flag is declared as (1 << 30), which when set
> > makes dtc unhappy about too big address set for a I2C-slave:
> > 
> > Warning (i2c_bus_reg): /example-2/i2c@...0000/eeprom@64: I2C bus unit address format error, expected "40000064"
> > Warning (i2c_bus_reg): /example-2/i2c@...0000/eeprom@64:reg: I2C address must be less than 10-bits, got "0x40000064"
> > 
> > Similar problem would have happened if we had set the 10-bit address
> > flag I2C_TEN_BIT_ADDRESS in the "reg"-property.
> > 
> > In order to fix the problem we suggest to alter the I2C-bus reg-check
> > algorithm, so one would be aware of the upper bits set. Normally if no
> > flag specified, the 7-bit address is expected in the "reg"-property.
> > If I2C_TEN_BIT_ADDRESS is set, then the 10-bit address check will be
> > performed. The I2C_OWN_SLAVE_ADDRESS flag will be just ignored.
> > 
> > [1] kernel/Documentation/i2c/slave-interface.rst
> > [2] kernel/include/dt-bindings/i2c/i2c.h
> 
> ...
> 

> > +		addr = reg & 0x3FFFFFFFU;
> > +		if ((reg & (1 << 31)) && addr > 0x3ff)
> >  			FAIL_PROP(c, dti, node, prop, "I2C address must be less than 10-bits, got \"0x%x\"",
> > -				  reg);
> > -
> > +				  addr);
> > +		else if (!(reg & (1 << 31)) && addr > 0x7f)
> > +			FAIL_PROP(c, dti, node, prop, "I2C address must be less than 7-bits, got \"0x%x\"",
> > +				  addr);
> 
> 1 << 31 is UB.

Good point. Thanks.

-Sergey

> 
> -- 
> With Best Regards,
> Andy Shevchenko
> 
> 

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ