05.10.2017 01:20, Sergey Ryazanov:
On Mon, Oct 2, 2017 at 2:33 AM, Sergey Ryazanov <ryazanov....@gmail.com> wrote:
Some boards have identical hardware id and differ only in hardware
revision (e.g. Acher C20 and Archer C20i). Such case confuse image
inspection code and it selects wrong board info structure.
Rework the board detection code to make it consider the hw revision
field. Now it returns best match board info:
* if possible then return board info with matched hw_id & hw_rev
* otherwise return first board info with matched hw_id (fallback to old
behaviour)
This patch by some cause is not picked up by patchwork. Should I resend it?
Hey Sergey,
please hold on with resending the patch.
At the moment I'm reviewing
https://github.com/lede-project/source/pull/1399 which will make your
patch obsolete by removing the code you're patching.
The PR drops any hardcoded board params from mktplinkfw2.c in favour of
commandline arguments, to provide only one way of creating tp-link
images. It has the nice benefit that this file most likely doesn't need
to be touched any more if new tp-link boards gets added.
Would be nice if you can give the PR a try to test for regressions.
Mathias
_______________________________________________
Lede-dev mailing list
Lede-dev@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/lede-dev
