[OE-core] [PATCH] dropbear: drop obsolete patch 0004-fix-2kb-keys.patch

[Andre McCurdy]

On Tue, May 29, 2018 at 11:56 AM, Richard Purdie
<richard.purdie at linuxfoundation.org> wrote:
> On Tue, 2018-05-29 at 11:26 -0700, Andre McCurdy wrote:
>> On Tue, May 29, 2018 at 10:29 AM, Khem Raj <raj.khem at gmail.com>
>> wrote:
>> >
>> > On Thu, May 24, 2018 at 7:27 PM, Andre McCurdy <armccurdy at gmail.com
>> > > wrote:
>> > >
>> > > This patch has been obsolete since 2005:
>> > >
>> > >   https://github.com/mkj/dropbear/commit/736f370dce614b717193f45d
>> > > 084e9e009de723ce
>> > Seems fine to drop but I would like to understand why we carried it
>> > thus far.
>> > What is the history of it in OE ?
>> It's older than the git history in oe-core (ie it was added in August
>> 2005 as part of the first commit to oe-core).
>>
>> The patch looks like it was initially created in March 2005 against
>> dropbear 0.45.
>>
>> In 2011 someone (incorrectly) added a tag "Upstream-Status:
>> Inappropriate [configuration]".
>>
>> That's about as much history as we have.
>
> Looks like it was added for 0.44:
>
> http://git.openembedded.org/openembedded/commit/packages/dropbear?id=ed6e40d6aff6cffabeff6f7fd6a9fee062566b2c
>
> and there was some associated discussion:
>
> http://familiar.handhelds.narkive.com/b1VGg2bI/problem-w-dropbear-ssh

That discussion ends with Matt Johnston (the dropbear author) saying
he will make a change in the next release to set the key size at
run-time rather than relying on a hardcoded maximum... which is
exactly what he did.

> so remote systems with 2048 length keys were somehow causing it
> problems?
>
> Cheers,
>
> Richard