Re: thread-safety: strerror_r()

Peter Eisentraut <peter@eisentraut.org> writes:
> I think we can apply these patches now to check this off the list of 
> not-thread-safe functions to check.

+1 for the first patch.  I'm less happy with

-    static char errbuf[36];
+    static char errbuf[128];

As a minor point, shouldn't this be

+    static char errbuf[PG_STRERROR_R_BUFLEN];

But the bigger issue is that the use of a static buffer makes
this not thread-safe, so having it use strerror_r to fill that
buffer is just putting lipstick on a pig.  If we really want
to make this thread-ready, we need to adopt the approach used
in libpq's fe-secure-openssl.c, where callers have to free the
buffer later.  Or maybe we could just palloc the result, and
trust that it's not in a long-lived context?

            regards, tom lane