Summary: | ASTERISK-04797: ToS can't be set properly | ||
Reporter: | Roy Sigurd Karlsbakk (rkarlsba) | Labels: | |
Date Opened: | 2005-08-09 06:30:26 | Date Closed: | 2008-01-15 15:44:11.000-0600 |
Priority: | Trivial | Regression? | No |
Status: | Closed/Complete | Components: | Core/General |
Versions: | Frequency of Occurrence | ||
Related Issues: | |||
Environment: | Attachments: | ( 0) siptosfix.patch ( 1) sip-tos-new.patch | |
Description: | ToS can't be set with hex in sip.conf..... with this patch it can be set with octal, if you want to :) | ||
Comments: | By: Roy Sigurd Karlsbakk (rkarlsba) 2005-08-09 07:05:55 new patch allows setting tos = 0 or 0x0 By: Roy Sigurd Karlsbakk (rkarlsba) 2005-08-09 07:11:38 and for one that noted 'why only octal', you can, of course set it using hex as well :) By: Vahan Yerkanian (vahan) 2005-08-09 08:06:01 just patched my cvs-head with your patch, here is the relevant output from asterisk -cvvvvvvvvv (with tos=lowdelay in sip.conf) == Parsing '/raid/etc/asterisk/sip.conf': Found == SIP Listening on 195.250.77.70:5060 == Using TOS bits 16 and with tos=0x18 == Parsing '/raid/etc/asterisk/sip.conf': Found == SIP Listening on 195.250.77.70:5060 == Using TOS bits 0 and with tos=184 == Parsing '/raid/etc/asterisk/sip.conf': Found == SIP Listening on 195.250.77.70:5060 == Using TOS bits 184 By: Roy Sigurd Karlsbakk (rkarlsba) 2005-08-09 08:09:56 you can check it with tcpdump -v udp as well :) works well By: Vahan Yerkanian (vahan) 2005-08-09 08:15:02 my sip.conf ----------- tos=0x18 my tcpdump ---------- 18:17:13.217069 IP (tos 0x0, ttl 64, id 63656, offset 0, flags [none], length: 60) sip.arminco.com.18896 > host-195-250-77-71.customer.arminco.com.24794: [bad udp cksum d2f6!] UDP, length: 32 By: Roy Sigurd Karlsbakk (rkarlsba) 2005-08-09 08:20:13 did you remember to reload? if i change tos in sip.conf and reload, it works for me. By: Vahan Yerkanian (vahan) 2005-08-09 08:23:25 I stand corrected, reload didn't work for me, had to restart asterisk itself. Now I see proper tos on tcpdump, however, all packets seem to have wrong udp checksum - not sure how major is this tho. Thanks. 18:23:47.103865 IP (tos 0x18, ttl 64, id 7400, offset 0, flags [none], length: 60) sip.arminco.com.13964 > phoenix.arminco.com.6092: [bad udp cksum fd98!] UDP, length: 32 18:23:47.123394 IP (tos 0x0, ttl 126, id 42003, offset 0, flags [none], length: 60) phoenix.arminco.com.6092 > sip.arminco.com.13964: [udp sum ok] UDP, length: 32 18:23:47.123756 IP (tos 0x18, ttl 64, id 7402, offset 0, flags [none], length: 60) sip.arminco.com.13964 > phoenix.arminco.com.6092: [bad udp cksum ce55!] UDP, length: 32 18:23:47.144011 IP (tos 0x0, ttl 126, id 42004, offset 0, flags [none], length: 60) phoenix.arminco.com.6092 > sip.arminco.com.13964: [udp sum ok] UDP, length: 32 18:23:47.144383 IP (tos 0x18, ttl 64, id 7404, offset 0, flags [none], length: 60) sip.arminco.com.13964 > phoenix.arminco.com.6092: [bad udp cksum 75ed!] UDP, length: 32 By: Roy Sigurd Karlsbakk (rkarlsba) 2005-08-09 08:29:08 wierd no checksum errors here...... By: Vahan Yerkanian (vahan) 2005-08-09 10:07:53 confirming that this patch fixed the problem, but I still see udp checksum errors on my side. By: Russell Bryant (russell) 2005-08-09 10:31:22 The format string to sscanf was improperly changed to %d from %i. All we have to do to fix this is to change it back. Patch uploaded. Disclaimer on file. By: Vahan Yerkanian (vahan) 2005-08-09 10:42:01 siptosfix.patch does the job too, confirmed :) By: Roy Sigurd Karlsbakk (rkarlsba) 2005-08-09 11:01:09 I still find strtoul() a better way to do this, but then, that's my opinion.... By: Mark Spencer (markster) 2005-08-09 12:16:55 Fixed in CVS head. Thanks! By: Digium Subversion (svnbot) 2008-01-15 15:44:10.000-0600 Repository: asterisk Revision: 6314 U trunk/channels/chan_sip.c ------------------------------------------------------------------------ r6314 | markster | 2008-01-15 15:44:10 -0600 (Tue, 15 Jan 2008) | 2 lines Allow ToS to be set in hex (bug ASTERISK-4797) ------------------------------------------------------------------------ http://svn.digium.com/view/asterisk?view=rev&revision=6314 |