Re: [HACKERS] Threads vs Processes (was: NuSphere and PostgreSQL for

From: Cyrille Chepelov <cyrille(at)chepelov(dot)org>
To: pgsql-hackers-win32(at)postgresql(dot)org
Subject: Re: [HACKERS] Threads vs Processes (was: NuSphere and PostgreSQL for
Date: 2003-09-25 18:22:18
Message-ID: 20030925182218.GB21627@chepelov.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers pgsql-hackers-win32

Le Thu, Sep 25, 2003, à 12:57:09PM -0400, Bruce Momjian a écrit:
> Keith Bottner wrote:
> > Typically variables that you want to be per-thread are stored in what
> > Microsoft calls Thread Local Storage (TLS). Variables that you want shared
> > you can just treat as globals and statics with the appropriate threading
> > synchronization primitives. With Windows 2000 and later you have up to 1088
> > TLS locations that you can use, of course these can be pointers to memory
> > which can store whatever you want.
> Goes GCC on Windows support TLS, or only Microsoft compilers?

Well, for sure Borland C++ Builder has support for TLS. Gcc too
apparently:
http://www-es.fernuni-hagen.de/cgi-bin/info2html?(gcc)Thread-Local

Doesn't seem specially hard to use, I guess the primary factor will be
the amount of static data currently used and how much of it should be
sent to the shared and non-shared bins.

-- Cyrille

--

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message markw 2003-09-25 18:44:19 Re: Is this a commit problem?
Previous Message Bruno Wolff III 2003-09-25 17:48:56 Re: pgsql procedures??

Browse pgsql-hackers-win32 by date

  From Date Subject
Next Message Keith Bottner 2003-09-25 19:16:24 Re: [HACKERS] Threads vs Processes (was: NuSphere and PostgreSQL for windows)
Previous Message Tom Lane 2003-09-25 17:03:02 Re: [HACKERS] Threads vs Processes