Thread: Globals Storage and Thread-safety
Hi All, This patch is to start up a globals storage environment and, in the future, thread-safety. It includes three new files: env/env.h -- Header for globals storage env/env.c -- Source for globals storage and creation env/thread.h -- Header for possible future thread implementation I created a new subsystem directory but they can go where appropriate. Implementation is very much like shared memory storage. The globals environment is a hashtable that stores stuctures for the various subsystems. Developers would be able to create structures to store and retrieve items that would normally be held in global variables. The hashtable is keyed by an int that identifies the specific global. I had to change dynahash.c to use MemoryContexts only if HASH_ALLOC and HASH_CONTEXT are not set in the control flags. I have included an example implementation for the MemoryContext globals. It uses #defines to redefine the globals to the stored structure if compiled with USE_GLOBAL_ENVIRONMENT. This is probably should be optimized to reduce the number of calls to retrieve storage but I will worry about that when and if the patch is accepted. Thanks, Myron Scott mkscott@sacadia.com
Attachment
This has been saved for the 7.3 release:
http://candle.pha.pa.us/cgi-bin/pgpatches2
---------------------------------------------------------------------------
mkscott@sacadia.com wrote:
>
> Hi All,
>
> This patch is to start up a globals storage environment
> and, in the future, thread-safety. It includes three new files:
>
> env/env.h -- Header for globals storage
> env/env.c -- Source for globals storage and creation
> env/thread.h -- Header for possible future thread implementation
>
> I created a new subsystem directory but they can go where appropriate.
>
> Implementation is very much like shared memory storage. The globals
> environment is a hashtable that stores stuctures for the various
> subsystems. Developers would be able to create structures to store
> and retrieve items
> that would normally be held in global variables. The hashtable is keyed
> by an int that identifies the specific global.
>
> I had to change dynahash.c to use MemoryContexts only if HASH_ALLOC and
> HASH_CONTEXT are not set in the control flags.
>
> I have included an example implementation for the MemoryContext globals.
> It uses #defines to redefine the globals to the stored
> structure if compiled with USE_GLOBAL_ENVIRONMENT. This is probably
> should be optimized to reduce the number of calls to retrieve storage but
> I will worry about that when and if the patch is accepted.
>
> Thanks,
>
> Myron Scott
> mkscott@sacadia.com
>
Content-Description:
[ Attachment, skipping... ]
>
> ---------------------------(end of broadcast)---------------------------
> TIP 2: you can get off all lists at once with the unregister command
> (send "unregister YourEmailAddressHere" to majordomo@postgresql.org)
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026
Your patch has been added to the PostgreSQL unapplied patches list at:
http://candle.pha.pa.us/cgi-bin/pgpatches
I will try to apply it within the next 48 hours.
---------------------------------------------------------------------------
mkscott@sacadia.com wrote:
>
> Hi All,
>
> This patch is to start up a globals storage environment
> and, in the future, thread-safety. It includes three new files:
>
> env/env.h -- Header for globals storage
> env/env.c -- Source for globals storage and creation
> env/thread.h -- Header for possible future thread implementation
>
> I created a new subsystem directory but they can go where appropriate.
>
> Implementation is very much like shared memory storage. The globals
> environment is a hashtable that stores stuctures for the various
> subsystems. Developers would be able to create structures to store
> and retrieve items
> that would normally be held in global variables. The hashtable is keyed
> by an int that identifies the specific global.
>
> I had to change dynahash.c to use MemoryContexts only if HASH_ALLOC and
> HASH_CONTEXT are not set in the control flags.
>
> I have included an example implementation for the MemoryContext globals.
> It uses #defines to redefine the globals to the stored
> structure if compiled with USE_GLOBAL_ENVIRONMENT. This is probably
> should be optimized to reduce the number of calls to retrieve storage but
> I will worry about that when and if the patch is accepted.
>
> Thanks,
>
> Myron Scott
> mkscott@sacadia.com
>
Content-Description:
[ Attachment, skipping... ]
>
> ---------------------------(end of broadcast)---------------------------
> TIP 2: you can get off all lists at once with the unregister command
> (send "unregister YourEmailAddressHere" to majordomo@postgresql.org)
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026
I have attached an updated patch for globals storage. Objects created in memory are not freed properly in the original patch. This patch properly uses the TopMemoryContext to create the globals hashtable. Thanks, Myron mkscott@sacadia.com On Fri, 22 Feb 2002, Bruce Momjian wrote: > > This has been saved for the 7.3 release: > > http://candle.pha.pa.us/cgi-bin/pgpatches2 > > --------------------------------------------------------------------------- > > mkscott@sacadia.com wrote: > > > > Hi All, > > > > This patch is to start up a globals storage environment > > and, in the future, thread-safety. It includes three new files: > > > > env/env.h -- Header for globals storage > > env/env.c -- Source for globals storage and creation > > env/thread.h -- Header for possible future thread implementation > > > > I created a new subsystem directory but they can go where appropriate. > > > > Implementation is very much like shared memory storage. The globals > > environment is a hashtable that stores stuctures for the various > > subsystems. Developers would be able to create structures to store > > and retrieve items > > that would normally be held in global variables. The hashtable is keyed > > by an int that identifies the specific global. > > > > I had to change dynahash.c to use MemoryContexts only if HASH_ALLOC and > > HASH_CONTEXT are not set in the control flags. > > > > I have included an example implementation for the MemoryContext globals. > > It uses #defines to redefine the globals to the stored > > structure if compiled with USE_GLOBAL_ENVIRONMENT. This is probably > > should be optimized to reduce the number of calls to retrieve storage but > > I will worry about that when and if the patch is accepted. > > > > Thanks, > > > > Myron Scott > > mkscott@sacadia.com > > > > Content-Description: > > [ Attachment, skipping... ] > > > > > ---------------------------(end of broadcast)--------------------------- > > TIP 2: you can get off all lists at once with the unregister command > > (send "unregister YourEmailAddressHere" to majordomo@postgresql.org) > > -- > Bruce Momjian | http://candle.pha.pa.us > pgman@candle.pha.pa.us | (610) 853-3000 > + If your life is a hard drive, | 830 Blythe Avenue > + Christ can be your backup. | Drexel Hill, Pennsylvania 19026 >