Lee Kindness wrote:
> Shridhar, Once the patches I've put forward are applied there's still
> a further change I've got planned which will remove the mutex locking
> in the common case - a NULL/DEFAULT connection parameter (I'll post a
> patch soon). This leaves the threaded case with comparable performance
> to the non-threaded case (both have a function call to get the
> connection, plus a getspecific call in the threaded case). As such is
> there much benefit in adding support for the connection being supplied
> by a struct pointer? You'd also have to add in something like "EXEC
> SQL GET DESCRIPTION xxx" to get the pointer too. How would it improve
> things over how they are in the test_thread_implicit test program?
>
> I still think it's worthwhile investigating the use of GCC's __thread
> storage class specifier to remove the use of pthread_*specific in this
> case. This would also be a help to the WIN32 port since this specifier
> maps well to similar constructs in Microsoft's and Borland's compilers
> (see "thread" item in the TODO at developer.postgresql.org).
I would like to avoid compiler-specific thread stuff unless tests can
show a performance benefit.
-- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610)
359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square,
Pennsylvania19073