I'm stuck on the final stages of a homework problem. Given $\mu(E) < \infty$ and $f$ measurable, we've been asked to show that
$\lim_{p \to p_0^-} \lVert f \rVert_p = \lVert f \rVert_{p_0}$.
Initially, I just approached this problem as a direct proof using Hölder's inequality, but that obviously doesn't work when $p_0 < 1$, so now I've changed tactics a bit.
I've handled the case where $f \notin L^{p_0}(E)$, I've shown that $f \in L^p$ if $f \in L^{p_0}$, and I've used DCT to arrive at the conclusion that $\lim_{p \to p_0^-} \int_E \lvert f \rvert^p = \int_E \lvert f \rvert^{p_0}$.
I'm running into trouble concluding from here that my limit is preserved by taking the $1/p$ and $1/p_0$ exponents on each side when $p_0 < 1$.
In the other questions I've looked at, either I couldn't figure out from the answers how the question of convergence was addressed in the case $p < 1$, or it wasn't a consideration to begin with.
Any help would be appreciated -- thank you!