linux-kernel - Re: [ANNOUNCE] Kconfiglib menuconfig implementation

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:   Tue, 1 May 2018 23:07:03 +0200
From:   Ulf Magnusson <ulfalizer@...il.com>
To:     Randy Dunlap <rdunlap@...radead.org>
Cc:     Linux Kbuild mailing list <linux-kbuild@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [ANNOUNCE] Kconfiglib menuconfig implementation

On Tue, May 1, 2018 at 9:47 PM, Ulf Magnusson <ulfalizer@...il.com> wrote:
> On Tue, May 1, 2018 at 9:12 PM, Ulf Magnusson <ulfalizer@...il.com> wrote:
>> On Tue, May 1, 2018 at 7:56 PM, Randy Dunlap <rdunlap@...radead.org> wrote:
>>> On 04/30/2018 05:57 PM, Ulf Magnusson wrote:
>>>> Hello,
>>>>
>>>> Kconfiglib (https://github.com/ulfalizer/Kconfiglib) now has a
>>>> terminal menuconfig implementation, implemented in plain curses
>>>> (which is in the Python standard library).
>>>>
>>>> The interface should feel familiar to people used to mconf. It has
>>>> some features that mconf lacks:
>>>>
>>>>   - Seamless resizing
>>>>
>>>>   - Unicode support
>>>>
>>>>   - Runs on Windows (via 'pip install windows-curses', which uses
>>>>     PDCurses)
>>>>
>>>>   - Improved information displays:
>>>>
>>>>     * All expressions are split into readable chunks
>>>>
>>>>     * Menus and comments have information displays
>>>>
>>>>   - Relatively easy-to-read and easy-to-tweak code.
>>>>
>>>>     Kconfiglib automatically invalidates symbols as needed, and
>>>>     values can never get stale, which helps.
>>>>
>>>> Some upcoming features are mouse support and a search feature that
>>>> can jump directly to the definition of a symbol. The jump-to feature
>>>> will use a "show-all" mode in case the symbol isn't visible.
>>>>
>>>> See the Kconfiglib GitHub page for screenshots. The menuconfig
>>>> implementation is at
>>>> https://github.com/ulfalizer/Kconfiglib/blob/master/menuconfig.py.
>>>> The docstring at the top has some more information.
>>>
>>> Hi,
>>>
>>> I'm probably missing some python additive (I hope it's that easy), but
>>> menuconfig.py is not liking the "degree" symbol in drivers/net/can/peak_canfd/Kconfig:
>>>
>>> config CAN_PEAK_PCIEFD
>>>         depends on PCI
>>>         tristate "PEAK-System PCAN-PCIe FD cards"
>>>         ---help---
>>>           This driver adds support for the PEAK-System PCI Express FD
>>>           CAN-FD cards family.
>>>           These 1x or 2x CAN-FD channels cards offer CAN 2.0 a/b as well as
>>>           CAN-FD access to the CAN bus. Besides the nominal bitrate of up to
>>>           1 Mbit/s, the data bytes of CAN-FD frames can be transmitted with
>>>           up to 12 Mbit/s. A galvanic isolation of the CAN ports protects the
>>>           electronics of the card and the respective computer against
>>>           disturbances of up to 500 Volts. The PCAN-PCI Express FD can be
>>>           operated with ambient temperatures in a range of -40 to +85 °C.
>>>
>>>
>>> kconfiglib.KconfigSyntaxError:
>>> Malformed ascii in drivers/net/can/peak_canfd/Kconfig
>>> Context: b't temperatures in a range of -40 to +85 \xc2\xb0C.\n'
>>> Problematic data: b'\xc2'
>>> Reason: ordinal not in range(128)
>>>
>>>
>>> thanks,
>>> --
>>> ~Randy
>>
>> Thanks for trying it out!
>>
>> You're probably running in the C locale, which implies an ASCII
>> encoding. That has caused enough trouble that the Python devs decided
>> to automatically convert it to UTF-8 in Python 3.7:
>> https://www.python.org/dev/peps/pep-0538/. LC_CTYPE=C.UTF-8 should fix
>> it.
>>
>> It's a bit silly to have it crash for something like that though. I
>> could force UTF-8 instead of respecting the locale (though it feels
>> neater to respect settings), or tell Python to ignore decoding errors.
>> Should probably do something at least...
>>
>> Cheers,
>> Ulf
>
> Leaning towards just forcing UTF-8. It's what you want in 99% of
> cases, and ignoring decoding errors would be unsafe for Unicode string
> values.
>
> Could make the forcing optional, and default to on...
>
> Cheers,
> Ulf

Went with a more general solution:
https://github.com/ulfalizer/Kconfiglib/commit/da40c014398f329b324a2eb9de062344e773dc74

You can now specify any encoding (or None, to use the encoding
specified in the environment), with "utf-8" as the default. That
default probably saves a bunch of pain in practice.

Cheers,
Ulf