LLchisq chokes when faced with non-likelihood based model

```
library(piecewiseSEM)
library(MASS)

df <- data.frame(
  a = rnorm(100),
  b = rnorm(100),
  c = rnorm(100),
  random = letters[1:4])

model <- psem(
  glmmPQL(a ~ b, random = ~ 1 | random, family = "gaussian", data = df),
  glmmPQL(b ~ c, random = ~ 1 | random, family = "gaussian", data = df)
)

summary(model)

LLchisq(model)
```