Oooops, I think people are talking past each other here.
Subrata, nobody suggested that we shouldn't create a tigerlake/
directory, we should!
Arthur, nobody suggested to review this patch. I assume Subrata
deliberately ignores our development guidelines and wants this
rubber-stamped.
Our problem is not with the introduction of 8k LOC for Tigerlake.
Sure, more common code would be nice, but we understand that it's
not the best time. The problem is that you want to skip review.
Which multiplies future maintenance effort, just to save somebody
some days to write proper commits. That's very harmful for the
project as a whole. Without a reasonable commit history (from
which one can tell if a piece of code was reviewed for a particular
piece of silicon) maintenance cost increase so heavily that it
delays future work like more common code and eventually also
the next SoC generation. It also increases the regression rate,
it wastes resources on bug hunting, ... A broken Git history is a
technical debt that can only be repaid by starting from scratch.
I doubt that is the plan here.
So, instead of adding 8k LOC at once and unreviewed, we want you
to add the code topic by topic, and let it be reviewed commit by
commit. That shouldn't be hard. Just take the current state you
have achieved, throw everything away that isn't validated yet,
and put what is left into smaller, reviewable commits. Please!
To view, visit change 36087. To unsubscribe, or for help writing mail filters, visit settings.