[patch] Correct range checking in mallopt/mxfast/tcache [BZ #25194]

DJ Delorie dj@redhat.com
Wed Dec 4 04:47:00 GMT 2019


"Carlos O'Donell" <codonell@redhat.com> writes:
> Does this allow mallopt to accept an invalid large negative 
> value but after wrapping be treated as less than MAX_TCACHE_SIZE
> and thus be accepted instead of rejected?

Only if sizeof(int) > sizeof(size_t), which for our targets never
happens.  Negative values passed to mallopt will now result in failure,
because they'll map to positive size_t larger than MAX_TCACHE_SIZE.

Plus MAX_TCACHE_SIZE is "very small"



More information about the Libc-alpha mailing list