On 2024-04-09 Tu 09:46, Tom Lane wrote:
Michael Paquier <mich...@paquier.xyz> writes:
By the way, are you planning to do something like [1]? I've not
looked in details at the callers of IPC::Run::timeout, still the extra
debug output would be nice.
It needs more review I think - I didn't check every call site to see
if anything would be broken. I believe Andrew has undertaken a
survey of all the timeout/timer calls, but if he doesn't produce
anything I might have a go at it after awhile.
What I looked at so far was the use of is_expired, but when you look
into that you see that you need to delve further, to where timeout/timer
objects are created and passed around. I'll take a closer look when I
have done some incremental json housekeeping.
cheers
andrew
--
Andrew Dunstan
EDB: https://www.enterprisedb.com
