Re: margay fails assertion in stats/dsa/dsm code

On Thu, Jun 30, 2022 at 10:34 PM Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote:
> So... I think we should select a different default
> dynamic_shared_memory_type in initdb.c if defined(__sun__). Which is
> the least terrible? For sysv, it looks like all the relevant sysctls
> that used to be required to use sysv memory became obsolete/automatic
> in Sol 10 (note: Sol 9 is long EOL'd), so it should just work AFAICT,
> whereas for mmap mode your shared memory data is likely to cause file
> I/O because we put the temporary files in your data directory. I'm
> thinking perhaps we should default to dynamic_shared_memory_type=sysv
> for 15+. I don't really want to change it in the back branches, since
> nobody has actually complained about "posix" performance and it might
> upset someone if we change it for newly initdb'd DBs in a major
> release series. But I'm not an expert or even user of this OS, I'm
> just trying to fix the build farm; better ideas welcome.

Boy, relying on DSM for critical stuff sure is a lot of fun! This is
exactly why I hate adding new facilities that have to be implemented
in OS-dependent ways.

Changing the default on certain platforms to 'posix' or 'sysv'
according to what works best on that platform seems reasonable to me.
I agree that defaulting to 'mmap' doesn't seem like a lot of fun,
although I think it could be a reasonable choice on a platform where
everything else is broken. You could alternatively try to fix 'posix'
by adding some kind of code to work around that platform's
deficiencies. Insert handwaving here.

--
Robert Haas
EDB: http://www.enterprisedb.com