Send Linux-ha-cvs mailing list submissions to
        [email protected]

To subscribe or unsubscribe via the World Wide Web, visit
        http://lists.community.tummy.com/mailman/listinfo/linux-ha-cvs
or, via email, send a message with subject or body 'help' to
        [EMAIL PROTECTED]

You can reach the person managing the list at
        [EMAIL PROTECTED]

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Linux-ha-cvs digest..."


Today's Topics:

   1. Linux-HA CVS: include by alan from 
      ([email protected])
   2. Linux-HA CVS: lib by alan from  ([email protected])
   3. Linux-HA CVS: heartbeat by alan from 
      ([email protected])
   4. Linux-HA CVS: debian by horms from 
      ([email protected])


----------------------------------------------------------------------

Message: 1
Date: Thu,  2 Feb 2006 13:58:56 -0700 (MST)
From: [email protected]
Subject: [Linux-ha-cvs] Linux-HA CVS: include by alan from 
To: [EMAIL PROTECTED]
Message-ID: <[EMAIL PROTECTED]>

linux-ha CVS committal

Author  : alan
Host    : 
Project : linux-ha
Module  : include

Dir     : linux-ha/include


Modified Files:
        hb_api_core.h 


Log Message:
Changed the priorities of a few of the realtime events in heartbeat.

===================================================================
RCS file: /home/cvs/linux-ha/linux-ha/include/hb_api_core.h,v
retrieving revision 1.15
retrieving revision 1.16
diff -u -3 -r1.15 -r1.16
--- hb_api_core.h       21 Dec 2005 02:34:32 -0000      1.15
+++ hb_api_core.h       2 Feb 2006 20:58:56 -0000       1.16
@@ -1,4 +1,4 @@
-/* $Id: hb_api_core.h,v 1.15 2005/12/21 02:34:32 gshi Exp $ */
+/* $Id: hb_api_core.h,v 1.16 2006/02/02 20:58:56 alan Exp $ */
 /*
  * hb_api_core_h: Internal definitions and functions for the heartbeat API
  *
@@ -37,14 +37,18 @@
 #include <ha_msg.h>
 
 /* Dispatch priorities for various kinds of events */
-#define        PRI_SENDSTATUS          G_PRIORITY_HIGH
-#define        PRI_DUMPSTATS           G_PRIORITY_LOW
-#define        PRI_AUDITCLIENT         G_PRIORITY_LOW
-#define        PRI_APIREGISTER         (G_PRIORITY_LOW-1)
-#define        PRI_CLUSTERMSG          G_PRIORITY_DEFAULT
-#define        PRI_FIFOMSG             PRI_CLUSTERMSG-1
-#define PRI_FREEMSG            G_PRIORITY_DEFAULT
-#define        PRI_CLIENTMSG           PRI_FIFOMSG
+#define        PRI_SENDSTATUS          (G_PRIORITY_HIGH)
+#define        PRI_SENDPKT             (PRI_SENDSTATUS+1)
+#define        PRI_READPKT             (PRI_SENDPKT+1)
+#define        PRI_FIFOMSG             (PRI_READPKT+1)
+
+#define PRI_CHECKSIGS          (G_PRIORITY_DEFAULT)
+#define PRI_FREEMSG            (PRI_CHECKSIGS+1)
+#define        PRI_CLIENTMSG           (PRI_FREEMSG+1)
+
+#define        PRI_APIREGISTER         (G_PRIORITY_LOW)
+#define        PRI_AUDITCLIENT         (PRI_APIREGISTER+1)
+#define        PRI_DUMPSTATS           (G_PRIORITY_LOW+20)
 
 void process_registerevent(IPC_Channel* chan,  gpointer user_data);
 




------------------------------

Message: 2
Date: Thu,  2 Feb 2006 14:01:37 -0700 (MST)
From: [email protected]
Subject: [Linux-ha-cvs] Linux-HA CVS: lib by alan from 
To: [EMAIL PROTECTED]
Message-ID: <[EMAIL PROTECTED]>

linux-ha CVS committal

Author  : alan
Host    : 
Project : linux-ha
Module  : lib

Dir     : linux-ha/lib/clplumbing


Modified Files:
        cl_malloc.c 


Log Message:
Turned off a cpu-expensive verification option that hasn't caught anything in 
many
months...

===================================================================
RCS file: /home/cvs/linux-ha/linux-ha/lib/clplumbing/cl_malloc.c,v
retrieving revision 1.23
retrieving revision 1.24
diff -u -3 -r1.23 -r1.24
--- cl_malloc.c 2 Feb 2006 05:15:22 -0000       1.23
+++ cl_malloc.c 2 Feb 2006 21:01:37 -0000       1.24
@@ -1,4 +1,4 @@
-/* $Id: cl_malloc.c,v 1.23 2006/02/02 05:15:22 alan Exp $ */
+/* $Id: cl_malloc.c,v 1.24 2006/02/02 21:01:37 alan Exp $ */
 /*
  * Copyright (C) 2000 Alan Robertson <[EMAIL PROTECTED]>
  *
@@ -51,6 +51,7 @@
  */
 
 #define        MARK_PRISTINE   1       /* Expensive in CPU time */
+#undef MARK_PRISTINE
 #define        MAKE_GUARD      1       /* Adds 'n' bytes memory - cheap in 
CPU*/
 #define        USE_ASSERTS     1
 #define        DUMPONERR       1




------------------------------

Message: 3
Date: Thu,  2 Feb 2006 15:30:30 -0700 (MST)
From: [email protected]
Subject: [Linux-ha-cvs] Linux-HA CVS: heartbeat by alan from 
To: [EMAIL PROTECTED]
Message-ID: <[EMAIL PROTECTED]>

linux-ha CVS committal

Author  : alan
Host    : 
Project : linux-ha
Module  : heartbeat

Dir     : linux-ha/heartbeat


Modified Files:
        hb_rexmit.c 


Log Message:
Fixed a bug where the random number generator was reseeded for every
random number we needed.  We use them on every retransmit.
This was very expensive - both to reseed the generator, and to go grab
a new number from /dev/urandom...

Probably ought to think about some kind of background consumer producer model
for generating random numbers for the future...
This would be really cool...


===================================================================
RCS file: /home/cvs/linux-ha/linux-ha/heartbeat/hb_rexmit.c,v
retrieving revision 1.4
retrieving revision 1.5
diff -u -3 -r1.4 -r1.5
--- hb_rexmit.c 21 Dec 2005 00:01:51 -0000      1.4
+++ hb_rexmit.c 2 Feb 2006 22:30:30 -0000       1.5
@@ -37,8 +37,8 @@
 
 static void    schedule_rexmit_request(struct node_info* node, seqno_t seq, 
int delay);
 
-static int max_rexmit_delay = 250;
-static GHashTable*             rexmit_hash_table = NULL;
+static int             max_rexmit_delay = 250;
+static GHashTable*     rexmit_hash_table = NULL;
 void hb_set_max_rexmit_delay(int);
 
 
@@ -47,6 +47,8 @@
        struct node_info* node;
 };
 
+static gboolean rand_seed_set = FALSE;
+
 void
 hb_set_max_rexmit_delay(int value)
 {
@@ -60,6 +62,8 @@
                       value);
        }
        max_rexmit_delay =value;
+       srand(cl_random());
+       rand_seed_set = TRUE;
        
        return;
 }
@@ -203,14 +207,20 @@
        return FALSE;
 }
 
+#define        RANDROUND       (RAND_MAX/2)
+
 static void
 schedule_rexmit_request(struct node_info* node, seqno_t seq, int delay)    
 {
-       unsigned long sourceid;
-       struct rexmit_info* ri;
+       unsigned long           sourceid;
+       struct rexmit_info*     ri;
+
        if (delay == 0){
-               srand(cl_random());
-               delay = (1.0* rand()/RAND_MAX)*max_rexmit_delay;
+               if (!rand_seed_set) {
+                       srand(cl_random());
+                       rand_seed_set = TRUE;
+               }
+               delay = ((rand()*max_rexmit_delay)+RANDROUND)/RAND_MAX;
        }
        
        ri = ha_malloc(sizeof(struct rexmit_info));




------------------------------

Message: 4
Date: Thu,  2 Feb 2006 18:15:17 -0700 (MST)
From: [email protected]
Subject: [Linux-ha-cvs] Linux-HA CVS: debian by horms from 
To: [EMAIL PROTECTED]
Message-ID: <[EMAIL PROTECTED]>

linux-ha CVS committal

Author  : horms
Host    : 
Project : linux-ha
Module  : debian

Dir     : linux-ha/debian/patches/series


Removed Files:
        2.0.0-1 


Log Message:
Debian packaging updates for 2.0.3

 * Add build dependancy on swig
 * As PYTHON_PATH has been added to configure.in it seems that autoconf1.4
   can no longer build this tree.  Currently autoconf is provided by
   autoconf1.4, and autoconfn is a virtual package provided by autoconf1.7,
   1.8 and 1.9.  Accordingly, replace the autoconf build dependancy with
   autoconf1.9 | autoconfn.  And add a build conflict on autoconf1.4.  This
   way the autoconf command should be autoconf1.7, 1.8 or 1.9, but not 1.4 
   as was previously the case.
 * Don't apply 00-ssh_scp_path.patch as it should no longer be needed,
   instead add a build dependancy on openssh-client.
   Also removed call to relevant configure options.
 * Removed configure option that no longer exists
   --enable-checkpointd
 * Removed configure options that are now on by default
   --enable-crm --enable-lrm
 * Add build dependancy on libgnutls11-dev, as 1.0 probably doesn't
   compile as _t types don't exist.
   http://lists.gnu.org/archive/html/help-gnutls/2004-08/msg00001.html






------------------------------

_______________________________________________
Linux-ha-cvs mailing list
[email protected]
http://lists.community.tummy.com/mailman/listinfo/linux-ha-cvs


End of Linux-ha-cvs Digest, Vol 27, Issue 14
********************************************

Reply via email to