FindDefinedSymbol() is subtly broken

From: John Naylor <jcnaylor(at)gmail(dot)com>
To: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: FindDefinedSymbol() is subtly broken
Date: 2018-05-19 07:23:43
Message-ID: CAJVSVGXM_n32hTTkircW4_K1LQFsJNb6xjs0pAP4QC0ZpyJfPQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Long story short: It expects an array of paths, but will die if it
can't find the symbol in the first path. Arguably it's a bug, but
since the original use case for multiple paths is long gone anyway, it
might never occur in practice. The attached patch changes it to expect
a single path.

Also attached: a few minor build script cleanups I've noticed along the way:
-more accurate error message
-s/print sprintf/printf/
-a recent perltidy change made a couple multi-line strings look odd,
so use heredocs (which other scripts use in this case anyway)
-make generated file footers match project style
-remove a duplicate comment

-John Naylor

Attachment Content-Type Size
0001-Update-FindDefinedSymbol-to-match-current-practice.patch text/x-patch 4.3 KB
last-minute-script-cleanups.patch text/x-patch 3.3 KB

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message John Naylor 2018-05-19 09:00:55 generating bootstrap entries for array types
Previous Message John Naylor 2018-05-19 06:41:10 Re: inconsistency and inefficiency in setup_conversion()