In my calculus book the author writes: $$ U(cf, P) =cL(f, P) \quad (1) $$ whereby $\mathbb{R}\ni c < 0$ and $U(cf, P) = \displaystyle\sum_{k=1}^{n} \sup\left\{cf(x)\mid x \in [x_{k-1}, x_k]\right\} \Delta x$ denotes the upper sum, similarly $L(f, P) = \displaystyle\sum_{k=1}^{n} \inf\left\{cf(x)\mid x \in [x_{k-1}, x_k]\right\} \Delta x$ denotes the lower sum.
Shouldn't $(1)$ rather be $$ U(cf, P) =|c|L(f, P) ? $$
Since $c < 0$, we have $c = -|c|$ and
$$\begin{align}\sup\left\{cf(x)\mid x \in [x_{k-1}, x_k]\right\}&= \sup\left\{-|c|f(x)\mid x \in [x_{k-1}, x_k]\right\} \\&\underbrace{=}_{\text{by Property A}} -\inf\left\{|c|f(x)\mid x \in [x_{k-1}, x_k]\right\}\\ &\underbrace{=}_{\text{by Property B}} -|c|\inf\left\{|f(x)\mid x \in [x_{k-1}, x_k]\right\} \end{align}$$
Hence,
$$\begin{align} U(cf,P) &= \sum_{k=1}^n\sup\left\{cf(x)\mid x \in [x_{k-1}, x_k]\right\} \Delta x_k \\&= -\sum_{k=1}^n|c| \inf\left\{f(x)\mid x \in [x_{k-1}, x_k]\right\} \Delta x_k \\ &= -|c| L(f,P) = cL(f,P)\end{align}$$
Here we have used the following properties of supremum and infimum.
Property A: $\sup(S) = -\inf(-S)$ where $-S = \{ a \mid -a \in S\}$
Property B: If $\alpha > 0$, then $\inf(\alpha S) = \alpha \inf(S) $ where $\alpha S = \{\alpha a \mid a \in S\}$.
Property A is cited in most real analysis texts.
For a proof of Property B, note that for all $a \in S$, $\alpha \inf(S) \leqslant \alpha a $ which implies that $\alpha \inf(S) \leqslant \inf(\alpha S)$. We also have $\inf(\alpha S) \leqslant \alpha a$ which implies that $\alpha^{-1} \inf(\alpha S) \leqslant a \leqslant \inf(S)$ and, thus, $ \inf(\alpha S) \leqslant \alpha \inf(S)$.
$$$$