drivers/mtd: Add check for integer overflow

[Teufelchen1]

Contribution description

Hei 🦫

Fixes a minor potential for integer overflow.

Testing procedure

CI / good review
Maybe make -C tests/unittests tests-mtd test

[Teufelchen1]

Unrelated change, sneaking it in to make the file happy in uncrustify.

[bergzand]

Changes look good, thanks for the cleanup

Untested ack from my side

[riot-ci]

Murdock results

✔️ PASSED

3bd047a drivers/mtd: Add check for interger overflow

Success	Failures	Total	Runtime
10066	0	10066	13m:59s

Artifacts

• Documentation preview

[Teufelchen1]

Switched from the unsigned buildins to the "you will figure the type out" buildins.

[maribu]

/opt/riot-toolchain/msp430-elf/10.1.0-18/bin/../lib/gcc/msp430-elf/10.1.0/../../../../msp430-elf/bin/ld: section .bss LMA wraps around address space
collect2: error: ld returned 1 exit status

This means it got larger than what would fit into 16 bit address space, so it needs to be added to Makefile.ci.

Btw.: I get this a lot with the BUILD_IN_DOCKER=1 toolchain, and rarely with Alpine's toolchain. My guess is that the newer linker will perform the check if data exceeds the 16 bit address space after the garbage collection phase, and the older one did this before. This would explain why the same app build on my toolchain not only ends up fitting into 16 bit address space, but also into the (smaller) RAM and ROM (and behave as expected), while it fails in the CI.

But anyway, bumping Makefile.ci it is for now and if the toolchain gets updated, we can re-enable linking in the CI of that app.

[Teufelchen1]

Thanks for looking into it Maribu. I add (via script) the olimex-msp430-h1611 and telosb to the Makefile.ci

ah maybe I should have put it in an different commit.