Dear Takahiro,
In message 20190719083010.GT21948@linaro.org you wrote:
On Fri, Jul 19, 2019 at 10:14:06AM +0200, Wolfgang Denk wrote:
Dear AKASHI Takahiro,
In message 20190717082525.891-5-takahiro.akashi@linaro.org you wrote:
This patch shows how environment storage drivers should be modified at the minimum to support contexts.
This commit message is misleading. No part of this patch is related to contexts. What it acually does is adding support for variable sized environment blocks. The commit message should be fixed.
Please note that this patch (or even all the patches in this set) are not intended to be merged as they are. This is an RFC in order for me to determine if you agree with the approach I take here or not. This is the primary goal. The code itself is nothing but a help for your understandings of my basic ideas.
Agreed. Please consider my comment as help for restructuring this patrch set and improving commit messages.
Also this commit makes me wonder if you have tested your patches for bisectability.
Again, not.
I already mentioned that I think the whole patch set should be restructured. For example, adding support for volatile vaariables is one logical thing; adding support for auto-save another one; extending the env (and eventually called driver) functions by a dummy context parameter (which is initially 0 everywhere, and unused in the code) is a single step before you actually introduce contexts, etc.
For each step, code size changes (especially for SPL) and bisectability shpuld be checked. But these are actually general rules for all patches ;-)
Best regards,
Wolfgang Denk