Re: using explicit_bzero

Peter Eisentraut <peter.eisentraut@2ndquadrant.com> writes:
> On 2019-07-18 00:45, Tom Lane wrote:
>> +1 for using the C11-standard name, even if that's not anywhere
>> in the real world yet.

> ISTM that a problem is that you cannot implement a replacement
> memset_s() as a wrapper around explicit_bzero(), unless you also want to
> implement the bound checking stuff.  (The "s"/safe in this family of
> functions refers to the bound checking, not the cannot-be-optimized-away
> property.)  The other way around it is easier.

Oh, hm.

> Also, the "s" family of functions appears to be a quagmire of
> controversy and incompatibility, so it's perhaps better to stay away
> from it for the time being.

Fair enough.

            regards, tom lane