[Asterisk-code-review] res_pjsip: Always set async_operations to 1. (asterisk[master])

Joshua Colp asteriskteam at digium.com
Tue Apr 26 05:00:04 CDT 2022


Joshua Colp has uploaded this change for review. ( https://gerrit.asterisk.org/c/asterisk/+/18425 )


Change subject: res_pjsip: Always set async_operations to 1.
......................................................................

res_pjsip: Always set async_operations to 1.

The async_operations setting on a transport configures how
many simultaneous incoming packets the transport can handle
when multiple threads are polling and waiting on the transport.
As we only use a single thread this was needlessly creating
incoming packets when set to a non-default value, wasting memory.

ASTERISK-30006

Change-Id: I1915973ef352862dc2852a6ba4cfce2ed536e68f
---
M configs/samples/pjsip.conf.sample
A doc/UPGRADE-staging/res_pjsip_async_operations.txt
M res/res_pjsip/config_transport.c
M res/res_pjsip/pjsip_config.xml
4 files changed, 14 insertions(+), 3 deletions(-)



  git pull ssh://gerrit.asterisk.org:29418/asterisk refs/changes/25/18425/1

diff --git a/configs/samples/pjsip.conf.sample b/configs/samples/pjsip.conf.sample
index b1c1657..2d93bd1 100644
--- a/configs/samples/pjsip.conf.sample
+++ b/configs/samples/pjsip.conf.sample
@@ -1025,8 +1025,6 @@
 ;[transport]
 ;  SYNOPSIS: SIP Transport
 ;
-;async_operations=1     ; Number of simultaneous Asynchronous Operations
-                        ; (default: "1")
 ;bind=  ; IP Address and optional port to bind to for this transport (default:
         ; "")
 ; Note that for the Websocket transport the TLS configuration is configured
diff --git a/doc/UPGRADE-staging/res_pjsip_async_operations.txt b/doc/UPGRADE-staging/res_pjsip_async_operations.txt
new file mode 100644
index 0000000..cf9f942
--- /dev/null
+++ b/doc/UPGRADE-staging/res_pjsip_async_operations.txt
@@ -0,0 +1,7 @@
+Subject: res_pjsip
+
+The 'async_operations' setting on transports is no longer
+obeyed and instead is always set to 1. This is due to the
+functionality not being applicable to Asterisk and causing
+excess unnecessary memory usage. This setting will now be
+ignored but can also be removed from the configuration file.
diff --git a/res/res_pjsip/config_transport.c b/res/res_pjsip/config_transport.c
index d0a28cd..6319ecf 100644
--- a/res/res_pjsip/config_transport.c
+++ b/res/res_pjsip/config_transport.c
@@ -659,6 +659,12 @@
 		return -1;
 	}
 
+	if (transport->async_operations != 1) {
+		ast_log(LOG_WARNING, "The async_operations setting on transport '%s' has been set to '%d'. The setting can no longer be set and is always 1.\n",
+			transport_id, transport->async_operations);
+		transport->async_operations = 1;
+	}
+
 	perm_state = find_internal_state_by_transport(transport);
 	if (perm_state) {
 		ast_sorcery_diff(sorcery, perm_state->transport, transport, &changes);
diff --git a/res/res_pjsip/pjsip_config.xml b/res/res_pjsip/pjsip_config.xml
index ca5a266..9a9ef48 100644
--- a/res/res_pjsip/pjsip_config.xml
+++ b/res/res_pjsip/pjsip_config.xml
@@ -1622,7 +1622,7 @@
 						will not suffice.</para></note>
 				</description>
 				<configOption name="async_operations" default="1">
-					<synopsis>Number of simultaneous Asynchronous Operations</synopsis>
+					<synopsis>Number of simultaneous Asynchronous Operations, can no longer be set, always set to 1</synopsis>
 				</configOption>
 				<configOption name="bind">
 					<synopsis>IP Address and optional port to bind to for this transport</synopsis>

-- 
To view, visit https://gerrit.asterisk.org/c/asterisk/+/18425
To unsubscribe, or for help writing mail filters, visit https://gerrit.asterisk.org/settings

Gerrit-Project: asterisk
Gerrit-Branch: master
Gerrit-Change-Id: I1915973ef352862dc2852a6ba4cfce2ed536e68f
Gerrit-Change-Number: 18425
Gerrit-PatchSet: 1
Gerrit-Owner: Joshua Colp <jcolp at sangoma.com>
Gerrit-MessageType: newchange
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-code-review/attachments/20220426/b4c2cee3/attachment.html>


More information about the asterisk-code-review mailing list