When we say a linear regression model, it means linear in parameters. The independent variables are known values.
For example,
y=a+b*exp(x1) + c * log (x2)  + d * x3**(4) + error

is a linear regression.

There are many of other equations that can be translate into a linear regression problem. Please refer it to Williams Greene's Economitric Analysis.

我问的不是模型是否线形，我是问dummy是否线形

进来学习，大大们快来帮帮我们这些下菜鸟吧

Why you care about dummy是否线形? It is totally irrelevant here.  A dummy variable can even be coded as A, B, C, so on