Adding a general abline in log-log ggplot2

I am trying to add a line to separate part of data in ggplot2. Following this thread: Adding linear model abline to log-log plot in ggplot

I tried

d = data.frame(x = 100*rlnorm(100), y = 100*rlnorm(100))
ggplot(d, aes(x, y)) + geom_point() + 
  geom_abline(intercept = 100, slope = -1, col='red') +
  scale_y_log10() + scale_x_log10()

but it did not plot the line. Note that the old plot approach got the line alright:

plot(d$x, d$y, log='xy')
abline(a = 100, b=-1, col='red', untf=TRUE)

This may not be the most elegant solution, but I usually define a separate data frame for predictions when I'm adding them to plots. I know that it's quicker in a lot of ways to add the model specification as part of the plot, but I really like the flexibility of having this as a separate object. Here's what I've got in mind in this case:

d = data.frame(x = 100*rlnorm(100), y = 100*rlnorm(100))

p = ggplot(d, aes(x,y)) + geom_point() + scale_x_log10() + scale_y_log10()

pred.func = function(x){
  100 - x
}

new.dat = data.frame(x = seq(from = 5, to = 90))
new.dat$pred = pred.func(new.dat$x)

p + geom_line(aes(x = x, y = pred), data = new.dat, col = "red")