Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Wed, May 12, 2010 at 4:31 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 11 May 2010 18:51, Bart Van Assche bvanass...@acm.org wrote: After having applied patch r18692 the Perl modules do now build with MSVC 6 and MSVC 2008. The MSVC 6 build went fine. Several crashes were reported however during the perl tests run as part of the MSVC 2008 build. These crashes have been fixed on the trunk and V5.5 branches via revisions r18375, r18369 and probably a few more commits. I've just tried building the 54x and 55x branches using MSVC 2008 (build.bat, including both Platform SDK and Perl support. In both cases, everything compiles OK, but the perl testing crashes out when it gets to the SNMP module. (Your vendor has not defined SNMP macro SNMP_DEFAULT_RETRIES) Right, I had already finished the default_store, ASN, OID and agent Perl modules but not yet the SNMP Perl module. I have added the details of the Win32 Perl regression test output on the bug tracker -- see also https://sourceforge.net/tracker/index.php?func=detailaid=3001079group_id=12694atid=112694 . Given that the errors are the same for both 54x and 55x lines, and in view of the length of time that the 54x release freeze has been in place, I'm inclined to the view that this probably isn't sufficient to delay the release any longer. We're never going to get all of the bugs out, but things feel good enough to run with. I understand this, but this also raises the question whether it will be possible to generate working Windows binaries for the upcoming 5.4 release. Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 11 May 2010 18:51, Bart Van Assche bvanass...@acm.org wrote: After having applied patch r18692 the Perl modules do now build with MSVC 6 and MSVC 2008. The MSVC 6 build went fine. Several crashes were reported however during the perl tests run as part of the MSVC 2008 build. These crashes have been fixed on the trunk and V5.5 branches via revisions r18375, r18369 and probably a few more commits. I've just tried building the 54x and 55x branches using MSVC 2008 (build.bat, including both Platform SDK and Perl support. In both cases, everything compiles OK, but the perl testing crashes out when it gets to the SNMP module. (Your vendor has not defined SNMP macro SNMP_DEFAULT_RETRIES) Given that the errors are the same for both 54x and 55x lines, and in view of the length of time that the 54x release freeze has been in place, I'm inclined to the view that this probably isn't sufficient to delay the release any longer. We're never going to get all of the bugs out, but things feel good enough to run with. Tomorrow is going to be busy (honours project deadline day), so nothing is going to happen then. I'm planning on cutting the final RC3 releases on Friday, with a view to making the full releases next week. Dave -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 8 May 2010 12:40, Bart Van Assche bvanass...@acm.org wrote: The Perl Counter64 code is buggy anyway, with or without patch #2890931. According to the strtoul() and strtoull() man pages, it is not guaranteed that errno is set when the conversion of a non-zero argument succeeded. See also http://www.opengroup.org/onlinepubs/95399/functions/strtoul.html. As far as I can tell - that's an issue with both strtoull and strtoul. Reverting to the 32-bit version wouldn't change anything as far as this is concerned. I've applied a simple patch to revert to strtoul whem compiling under Windows This allows the build.bat build to succeed when perl is enabled, while still allowing 64-bit support on other architectures. I'm not sure whether this also affect Cygwin and/or MinGW builds. (Or even whether the perl modules build at all in these environments). If you can let me know the status of the perl support with these builds, we can extend the patch if necessary. Otherwise I'm looking to cut the final final release candidates later this week. Dave -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Tue, May 11, 2010 at 5:51 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 8 May 2010 12:40, Bart Van Assche bvanass...@acm.org wrote: The Perl Counter64 code is buggy anyway, with or without patch #2890931. According to the strtoul() and strtoull() man pages, it is not guaranteed that errno is set when the conversion of a non-zero argument succeeded. See also http://www.opengroup.org/onlinepubs/95399/functions/strtoul.html. As far as I can tell - that's an issue with both strtoull and strtoul. Reverting to the 32-bit version wouldn't change anything as far as this is concerned. I've applied a simple patch to revert to strtoul whem compiling under Windows This allows the build.bat build to succeed when perl is enabled, while still allowing 64-bit support on other architectures. I'm not sure whether this also affect Cygwin and/or MinGW builds. (Or even whether the perl modules build at all in these environments). If you can let me know the status of the perl support with these builds, we can extend the patch if necessary. Otherwise I'm looking to cut the final final release candidates later this week. Although I consider r18689 as an ugly change, it should be sufficient to address the strtoull() linker errors that were reported by the MSVC batch build. Please note that this change must not be applied to the trunk or to the V5.5 branches since the strtoull.c source file is being compiled and linked by the Win32 batch build for these branches. After having applied patch r18692 the Perl modules do now build with MSVC 6 and MSVC 2008. The MSVC 6 build went fine. Several crashes were reported however during the perl tests run as part of the MSVC 2008 build. These crashes have been fixed on the trunk and V5.5 branches via revisions r18375, r18369 and probably a few more commits. Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Fri, May 7, 2010 at 3:55 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 6 May 2010 17:12, Bart Van Assche bvanass...@acm.org wrote: There is one open issue for the V5.4 branch I'm absolutely not happy with: http://www.mail-archive.com/net-snmp-users@lists.sourceforge.net/msg25049.html . That seems to be a consequence of patch #2890931 (to add support of Counter64 values within the perl module). I've had a quick attempt to add 'snmplib/strtoull.c' to the Win32 build environment, but without any success. I'm inclined to reverse the above patch (and re-open the tracker entry), so that we can get the releases out with no further delay. Unless anyone can provide a fix for the strtoull compilation problem Note that the only other occurance of this call are as follows: mibgroup/util_funcs sctp-mib/sctp{Scalars,Tables}_linux.c udp-mib/data_access/udp_endpoint_linux.c (all Linux-only code) if-mib/data_access_interface.c (any architecture) The Perl Counter64 code is buggy anyway, with or without patch #2890931. According to the strtoul() and strtoull() man pages, it is not guaranteed that errno is set when the conversion of a non-zero argument succeeded. See also http://www.opengroup.org/onlinepubs/95399/functions/strtoul.html. Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Thu, May 6, 2010 at 10:50 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 6 May 2010 17:12, Bart Van Assche bvanass...@acm.org wrote: There is one open issue for the V5.4 branch I'm absolutely not happy with: http://www.mail-archive.com/net-snmp-users@lists.sourceforge.net/msg25049.html . Which style(s) of build does this refer to? The steps I followed are: 1. Removed C:\usr\bin recursively. 2. Verified that there were no copies left of netsnmp.dll anywhere in PATH. 3. Opened a command prompt and selected MSVC 6 by running C:\Program Files\Microsoft Visual Studio\vc98\bin\VCVARS32.BAT. 4. Started win32\build.bat, enabled option 5 (Perl modules) and started the batch build. 5. Results: a dialog complaining about netsnmp_gethostbyname_v4 not found and build.bat finished with the following message: Build error (see perlmake.out) at build.pl line 249, line 2. The dialog that appeared: [image: dialog.png] Bart. dialog.png-- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 6 May 2010 17:12, Bart Van Assche bvanass...@acm.org wrote: There is one open issue for the V5.4 branch I'm absolutely not happy with: http://www.mail-archive.com/net-snmp-users@lists.sourceforge.net/msg25049.html. That seems to be a consequence of patch #2890931 (to add support of Counter64 values within the perl module). I've had a quick attempt to add 'snmplib/strtoull.c' to the Win32 build environment, but without any success. I'm inclined to reverse the above patch (and re-open the tracker entry), so that we can get the releases out with no further delay. Unless anyone can provide a fix for the strtoull compilation problem Note that the only other occurance of this call are as follows: mibgroup/util_funcs sctp-mib/sctp{Scalars,Tables}_linux.c udp-mib/data_access/udp_endpoint_linux.c (all Linux-only code) if-mib/data_access_interface.c (any architecture) Dave -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Wed, May 5, 2010 at 5:14 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 30 April 2010 11:07, Bart Van Assche bvanass...@acm.org wrote: The 5.4.3rc2 Cygwin IPv6 build was broken but has been fixed via r18607. Regression tests do not yet pass though -- see also https://sourceforge.net/tracker/?func=detailaid=2993522group_id=12694atid=112694 . I've now had a chance to look at your logs, as well as running my own tests. The results I'm seeing aren't quite as bad as yours - there are only five tests that fail (#s 19, 35, 45, 46 58) Checking the output, there seem to be two separate problems: a) Test 19 (T030) works fine until the agent is signalled with HUP to reload the configuration. The query that follows then fails. This is the first test that signals such a reload. This also appears to underly my failure for test 35 (but the logs from the tracker entry show a different problem) b) All tests attempt to use the utility function PROBE_FOR_PORT to identify suitable, unused port numbers for the various servers. This consistently fails. (Or rather, identifying a port number works, OK but assigning this to the variables SNMP_SNMPD_PORT, SNMP_SNMPTRAPD_PORT, etc fails) All tests therefore try to run both agent and trap receiver using the listening socket address udp:localhost: rather than udp:localhost:$PORT In the case of tests 45 46 (T120/121), this means that the second proxy agent attempts to listen on the same port as the master agent is already using. It cannot, and hence the proxy delegation fails. This problem also explains the failure of test 58 (T160) - the client command walks the UDP listening table, and looks for an entry with the port number SNMP_SNMPD_PORT. But since this variable is empty, there is no such matching entry in the table. I believe that the same problem underlies the rest of the failures that you are seeing. In each case, snmptrapd.log complains couldn't open udp:localhost: -- errno 1 (Operation not permitted) so the trap receiver isn't listening for incoming traps. I strongly suspect that hardwiring a suitable port number would allow these tests to succeed. (That's probably the next thing to check) I'm also seeing complaints from our anti-virus software whenever backquotes are used to invoke a sub-command (both here, and when running config.status). I don't know if this is causing the second problem, or whether this is a red herring. How do you want to proceed? Assuming that hardcoding port numbers will allow the tests to succeed, is this problem serious enough to warrent delaying the releases until it's fixed? The free port number detection issue has been fixed in r18651. With this fix the only test that still fails is test 58 because the Cygwin port of snmpd.exe reports incorrect information in the udpTable. See also https://sourceforge.net/tracker/?func=detailaid=2997492group_id=12694atid=112694. Another issue that I discovered and that is not covered by any regression test is that the information reported in tcpConnTable does not match the output of netstat. See also https://sourceforge.net/tracker/?func=detailaid=2997495group_id=12694atid=112694 . Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 6 May 2010 11:07, Bart Van Assche bvanass...@acm.org wrote: The free port number detection issue has been fixed in r18651. Good work. I've ported this (and other related patches) to the other relevant branches. The 5.2.x branch passes Cygwin regression tests OK (though not with IPv6 support). Something is odd wrt 53x, but that's less urgent, since it's not in release freeze mode. With this fix the only test that still fails is test 58 because the Cygwin port of snmpd.exe reports incorrect information in the udpTable. Another issue that I discovered and that is not covered by any regression test is that the information reported in tcpConnTable does not match the output of netstat. I'm inclined to suggest that these two problems shouldn't hold up the 5.4.3 (and 5.2.6) releases any longer - this process has dragged on long enough already! I'm proposing to package up the current code as rc3 tarballs tomorrow, with a view to these being the final releases. Yell now if you strongly disagree :-) Dave -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Thu, May 6, 2010 at 5:54 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 6 May 2010 11:07, Bart Van Assche bvanass...@acm.org wrote: The free port number detection issue has been fixed in r18651. Good work. I've ported this (and other related patches) to the other relevant branches. The 5.2.x branch passes Cygwin regression tests OK (though not with IPv6 support). Something is odd wrt 53x, but that's less urgent, since it's not in release freeze mode. With this fix the only test that still fails is test 58 because the Cygwin port of snmpd.exe reports incorrect information in the udpTable. Another issue that I discovered and that is not covered by any regression test is that the information reported in tcpConnTable does not match the output of netstat. I'm inclined to suggest that these two problems shouldn't hold up the 5.4.3 (and 5.2.6) releases any longer - this process has dragged on long enough already! I'm proposing to package up the current code as rc3 tarballs tomorrow, with a view to these being the final releases. Yell now if you strongly disagree :-) There is one open issue for the V5.4 branch I'm absolutely not happy with: http://www.mail-archive.com/net-snmp-users@lists.sourceforge.net/msg25049.html . Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 6 May 2010 17:12, Bart Van Assche bvanass...@acm.org wrote: There is one open issue for the V5.4 branch I'm absolutely not happy with: http://www.mail-archive.com/net-snmp-users@lists.sourceforge.net/msg25049.html. Which style(s) of build does this refer to? Dave -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 30 April 2010 11:07, Bart Van Assche bvanass...@acm.org wrote: The 5.4.3rc2 Cygwin IPv6 build was broken but has been fixed via r18607. Regression tests do not yet pass though -- see also https://sourceforge.net/tracker/?func=detailaid=2993522group_id=12694atid=112694. I've now had a chance to look at your logs, as well as running my own tests. The results I'm seeing aren't quite as bad as yours - there are only five tests that fail (#s 19, 35, 45, 46 58) Checking the output, there seem to be two separate problems: a) Test 19 (T030) works fine until the agent is signalled with HUP to reload the configuration. The query that follows then fails. This is the first test that signals such a reload. This also appears to underly my failure for test 35 (but the logs from the tracker entry show a different problem) b) All tests attempt to use the utility function PROBE_FOR_PORT to identify suitable, unused port numbers for the various servers. This consistently fails. (Or rather, identifying a port number works, OK but assigning this to the variables SNMP_SNMPD_PORT, SNMP_SNMPTRAPD_PORT, etc fails) All tests therefore try to run both agent and trap receiver using the listening socket address udp:localhost: rather than udp:localhost:$PORT In the case of tests 45 46 (T120/121), this means that the second proxy agent attempts to listen on the same port as the master agent is already using. It cannot, and hence the proxy delegation fails. This problem also explains the failure of test 58 (T160) - the client command walks the UDP listening table, and looks for an entry with the port number SNMP_SNMPD_PORT. But since this variable is empty, there is no such matching entry in the table. I believe that the same problem underlies the rest of the failures that you are seeing. In each case, snmptrapd.log complains couldn't open udp:localhost: -- errno 1 (Operation not permitted) so the trap receiver isn't listening for incoming traps. I strongly suspect that hardwiring a suitable port number would allow these tests to succeed. (That's probably the next thing to check) I'm also seeing complaints from our anti-virus software whenever backquotes are used to invoke a sub-command (both here, and when running config.status). I don't know if this is causing the second problem, or whether this is a red herring. How do you want to proceed? Assuming that hardcoding port numbers will allow the tests to succeed, is this problem serious enough to warrent delaying the releases until it's fixed? Dave -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Fri, Apr 30, 2010 at 9:22 PM, Bart Van Assche bart.vanass...@gmail.comwrote: On Sat, Apr 10, 2010 at 12:50 AM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: New release candidate versions of the 5.4.x and 5.2.x packages are now available from the SourceForge download servers. The main changes since the rc1 versions are: - Proper handling of multiple matching VACM entries (using the best match rather than the first one seen) This could potentially affect the behaviour of existing access control configurations. - Usage statistics for large disks are latched at 2Tb (2^31) rather than wrapping - The default AgentX target for Windows systems will listen on the localhost interface only rather than on all network interfaces. (5.4.3 only) This does not affect the default on Unix-based systems, which still use a named socket. The 5.2.6 release is envisaged as being the final version on the 5.2.x line. No further development is planned for this branch. It is hoped that these candidates can be released as the full versions as soon as possible, assuming no major issues come to light. Please test them out and report any problems that you might encounter. Download URLs: http://sourceforge.net/projects/net-snmp/files/net-snmp/5.2.6_pre-releases/ http://sourceforge.net/projects/net-snmp/files/net-snmp/5.4.3_pre-releases/ The 5.4.3 pre-release didn't build on MinGW. This has been fixed in r18631. Note: the 5.4 configure script fails to detect that recent MinGW versions supports IPv6. This issue has been fixed on the 5.5 branch and on the trunk. Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Thu, Apr 29, 2010 at 11:27 AM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 28 April 2010 12:02, Bart Van Assche bvanass...@acm.org wrote: The 5.4.3rc2 Cygwin IPv6 build was broken but has been fixed via r18607. Regression tests do not yet pass though -- see also https://sourceforge.net/tracker/?func=detailaid=2993522group_id=12694atid=112694 . Can you attach the logs from the tests that fail to the tracker entry, please. Under Unix-style systems, I'd expect these to be found under /tmp (one directory per failed test). I'm not sure whether this is the same for Cygwin. Do the regression tests succeed for a non-IPv6 build? If so, then I can probably investigate these problems myself. Unfortunately, our version of Cygwin does not include IPv6 support. The regression test output has been added to the Net-SNMP bug tracker. Regarding the non-IPv6 build: the results of the regression tests are the same as for the IPv6-enabled build (Summary: 41 / 59 succeeded.). Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Sat, Apr 10, 2010 at 12:50 AM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: New release candidate versions of the 5.4.x and 5.2.x packages are now available from the SourceForge download servers. The main changes since the rc1 versions are: - Proper handling of multiple matching VACM entries (using the best match rather than the first one seen) This could potentially affect the behaviour of existing access control configurations. - Usage statistics for large disks are latched at 2Tb (2^31) rather than wrapping - The default AgentX target for Windows systems will listen on the localhost interface only rather than on all network interfaces. (5.4.3 only) This does not affect the default on Unix-based systems, which still use a named socket. The 5.2.6 release is envisaged as being the final version on the 5.2.x line. No further development is planned for this branch. It is hoped that these candidates can be released as the full versions as soon as possible, assuming no major issues come to light. Please test them out and report any problems that you might encounter. Download URLs: http://sourceforge.net/projects/net-snmp/files/net-snmp/5.2.6_pre-releases/ http://sourceforge.net/projects/net-snmp/files/net-snmp/5.4.3_pre-releases/ The 5.4.3 pre-release didn't build on MinGW. This has been fixed in r18631. Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 28 April 2010 12:02, Bart Van Assche bvanass...@acm.org wrote: The 5.4.3rc2 Cygwin IPv6 build was broken but has been fixed via r18607. Regression tests do not yet pass though -- see also https://sourceforge.net/tracker/?func=detailaid=2993522group_id=12694atid=112694. Can you attach the logs from the tests that fail to the tracker entry, please. Under Unix-style systems, I'd expect these to be found under /tmp (one directory per failed test). I'm not sure whether this is the same for Cygwin. Do the regression tests succeed for a non-IPv6 build? If so, then I can probably investigate these problems myself. Unfortunately, our version of Cygwin does not include IPv6 support. Dave -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Sat, Apr 10, 2010 at 12:50 AM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: New release candidate versions of the 5.4.x and 5.2.x packages are now available from the SourceForge download servers. The main changes since the rc1 versions are: - Proper handling of multiple matching VACM entries (using the best match rather than the first one seen) This could potentially affect the behaviour of existing access control configurations. - Usage statistics for large disks are latched at 2Tb (2^31) rather than wrapping - The default AgentX target for Windows systems will listen on the localhost interface only rather than on all network interfaces. (5.4.3 only) This does not affect the default on Unix-based systems, which still use a named socket. The 5.2.6 release is envisaged as being the final version on the 5.2.x line. No further development is planned for this branch. It is hoped that these candidates can be released as the full versions as soon as possible, assuming no major issues come to light. Please test them out and report any problems that you might encounter. Download URLs: http://sourceforge.net/projects/net-snmp/files/net-snmp/5.2.6_pre-releases/ http://sourceforge.net/projects/net-snmp/files/net-snmp/5.4.3_pre-releases/ The 5.4.3rc2 Cygwin IPv6 build was broken but has been fixed via r18607. Regression tests do not yet pass though -- see also https://sourceforge.net/tracker/?func=detailaid=2993522group_id=12694atid=112694 . Bart. -- ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Sat, Apr 10, 2010 at 12:50 AM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: New release candidate versions of the 5.4.x and 5.2.x packages are now available from the SourceForge download servers. The main changes since the rc1 versions are: - Proper handling of multiple matching VACM entries (using the best match rather than the first one seen) This could potentially affect the behaviour of existing access control configurations. - Usage statistics for large disks are latched at 2Tb (2^31) rather than wrapping - The default AgentX target for Windows systems will listen on the localhost interface only rather than on all network interfaces. (5.4.3 only) This does not affect the default on Unix-based systems, which still use a named socket. The 5.2.6 release is envisaged as being the final version on the 5.2.x line. No further development is planned for this branch. It is hoped that these candidates can be released as the full versions as soon as possible, assuming no major issues come to light. Please test them out and report any problems that you might encounter. Download URLs: http://sourceforge.net/projects/net-snmp/files/net-snmp/5.2.6_pre-releases/ http://sourceforge.net/projects/net-snmp/files/net-snmp/5.4.3_pre-releases/ I just noticed that the perl modules are broken on the V5.4 branch when building on Win32. From win32/perlmake.out: [ ... ] agent.obj : error LNK2019: unresolved external symbol _strtoull referenced in function _XS_NetSNMP__agent__netsnmp_request_infoPtr_setValue [ ... ] Bart. -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 10 April 2010 08:53, Bart Van Assche bvanass...@acm.org wrote: Building 5.4.3.rc2 on Windows via build.bat fails: ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'socklen_t' : undeclared identifier ..\..\snmplib\snmpTCPDomain.c(187) : error C2146: syntax error : missing ';' before identifier 'farendlen' ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'farendlen' : undeclared identifier [ ... ] Eh? That code has been unchanged for nearly eight years! sigh Do you happen to know whether the earlier pre-releases compiled OK? What about 5.2.6.rc2 ? Is that broken as well? Dave -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On 10 April 2010 08:53, Bart Van Assche bvanass...@acm.org wrote: Building 5.4.3.rc2 on Windows via build.bat fails: [ ... ] cl.exe /nologo /MD /W3 /GX /O2 /I . /I .. /I ..\..\snmplib /I ..\.. /I ..\..\include /D WIN32 /D NDEBUG /D _CONSOLE /D _MBCS /D NETSNMP_DLL /Fp.\release\libsnmp_dll.pch /YX /Fo.\release\\ /Fd.\release\\ /FD /c ..\..\snmplib\snmpTCPDomain.c ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'socklen_t' : undeclared identifier ..\..\snmplib\snmpTCPDomain.c(187) : error C2146: syntax error : missing ';' before identifier 'farendlen' ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'farendlen' : undeclared identifier Can I check the exact settings you are working with, please. What version of Visual Studio are you using? What changes (if any) did you make to the build settings? I've just tried with a fresh tarball, using Visual Studio 2008 Professional, and accepting the default build settings. This ran into a couple of problems - but not anything like the errors shown above. Firstly, I found that 'build.pl' complained about needing to run the VCVARS32.bat script - despite the fact that I'd already run this. Then there were errors about inconsistent definitions within VC header files, which turned out to be due to the entry #define vsnprintf _vsnprintf within win32/net-snmp/net-snmp-config.h{,.in} Deleting this allowed the software to build OK. I don't know whether these are common problems, or related to our setup here. But I haven't been able to reproduce the error you list above. Dave -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
RE: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
From: Dave Shield [mailto:d.t.shi...@liverpool.ac.uk] Sent: Monday, April 12, 2010 3:37 AM On 10 April 2010 08:53, Bart Van Assche bvanass...@acm.org wrote: Building 5.4.3.rc2 on Windows via build.bat fails: ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'socklen_t' : undeclared identifier ..\..\snmplib\snmpTCPDomain.c(187) : error C2146: syntax error : missing ';' before identifier 'farendlen' ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'farendlen' : undeclared identifier [ ... ] Eh? That code has been unchanged for nearly eight years! sigh Do you happen to know whether the earlier pre-releases compiled OK? What about 5.2.6.rc2 ? Is that broken as well? Also, which version of VC are you using? Code has been known to break on VC upgrades. HTH, Mike -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev ___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Mon, Apr 12, 2010 at 4:20 PM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: On 10 April 2010 08:53, Bart Van Assche bvanass...@acm.org wrote: Building 5.4.3.rc2 on Windows via build.bat fails: [ ... ] cl.exe /nologo /MD /W3 /GX /O2 /I . /I .. /I ..\..\snmplib /I ..\.. /I ..\..\include /D WIN32 /D NDEBUG /D _CONSOLE /D _MBCS /D NETSNMP_DLL /Fp.\release\libsnmp_dll.pch /YX /Fo.\release\\ /Fd.\release\\ /FD /c ..\..\snmplib\snmpTCPDomain.c ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'socklen_t' : undeclared identifier ..\..\snmplib\snmpTCPDomain.c(187) : error C2146: syntax error : missing ';' before identifier 'farendlen' ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'farendlen' : undeclared identifier Can I check the exact settings you are working with, please. What version of Visual Studio are you using? What changes (if any) did you make to the build settings? I've just tried with a fresh tarball, using Visual Studio 2008 Professional, and accepting the default build settings. This ran into a couple of problems - but not anything like the errors shown above. Firstly, I found that 'build.pl' complained about needing to run the VCVARS32.bat script - despite the fact that I'd already run this. Then there were errors about inconsistent definitions within VC header files, which turned out to be due to the entry #define vsnprintf _vsnprintf within win32/net-snmp/net-snmp-config.h{,.in} Deleting this allowed the software to build OK. I don't know whether these are common problems, or related to our setup here. But I haven't been able to reproduce the error you list above. As far as I know win32/build.bat on the V5.4 branch only works with Visual Studio 6. This is what I get with default build settings for win32/build.bat: [ ... ] netsnmpmibs.lib(snmpNotifyTable.obj) : error LNK2001: unresolved external symbol _log_notification ..\bin\.\release/snmpd.exe : fatal error LNK1120: 1 unresolved externals NMAKE : fatal error U1077: 'link.exe' : return code '0x460' Stop. And this is what I get with Platform SDK Support, IPv6 support enabled and link type set to DLL, as shown above: [ ... ] cl.exe /nologo /MD /W3 /GX /O2 /I . /I .. /I ..\..\snmplib /I ..\.. /I ..\..\include /D WIN32 /D NDEBUG /D _CONSOLE /D _MBCS /D NETSNMP_DLL /Fp.\release\libsnmp_dll.pch /YX /Fo.\release\\ /Fd.\release\\ /FD /c ..\..\snmplib\snmpTCPDomain.c snmpTCPDomain.c ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'socklen_t' : undeclared identifier [ ... ] These results have been produced with the source code obtained from https://net-snmp.svn.sourceforge.net/svnroot/net-snmp/branches/V5-4-patches/net-snmp, r18491. No Subversion-administered files were modified except those overwritten by the build process (win32/libsnmp_dll/libsnmp.def and win32/net-snmp/net-snmp-config.h). Note: win32/win32.dsw builds fine with VS6, at least on my setup. Bart. -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users
Re: Net-SNMP 5.4.3.rc2 and 5.2.6.rc2 available for testing
On Sat, Apr 10, 2010 at 12:50 AM, Dave Shield d.t.shi...@liverpool.ac.ukwrote: New release candidate versions of the 5.4.x and 5.2.x packages are now available from the SourceForge download servers. The main changes since the rc1 versions are: - Proper handling of multiple matching VACM entries (using the best match rather than the first one seen) This could potentially affect the behaviour of existing access control configurations. - Usage statistics for large disks are latched at 2Tb (2^31) rather than wrapping - The default AgentX target for Windows systems will listen on the localhost interface only rather than on all network interfaces. (5.4.3 only) This does not affect the default on Unix-based systems, which still use a named socket. The 5.2.6 release is envisaged as being the final version on the 5.2.x line. No further development is planned for this branch. It is hoped that these candidates can be released as the full versions as soon as possible, assuming no major issues come to light. Please test them out and report any problems that you might encounter. Download URLs: http://sourceforge.net/projects/net-snmp/files/net-snmp/5.2.6_pre-releases/ http://sourceforge.net/projects/net-snmp/files/net-snmp/5.4.3_pre-releases/ Building 5.4.3.rc2 on Windows via build.bat fails: [ ... ] cl.exe /nologo /MD /W3 /GX /O2 /I . /I .. /I ..\..\snmplib /I ..\.. /I ..\..\include /D WIN32 /D NDEBUG /D _CONSOLE /D _MBCS /D NETSNMP_DLL /Fp.\release\libsnmp_dll.pch /YX /Fo.\release\\ /Fd.\release\\ /FD /c ..\..\snmplib\snmpTCPDomain.c ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'socklen_t' : undeclared identifier ..\..\snmplib\snmpTCPDomain.c(187) : error C2146: syntax error : missing ';' before identifier 'farendlen' ..\..\snmplib\snmpTCPDomain.c(187) : error C2065: 'farendlen' : undeclared identifier [ ... ] Bart. -- Download Intel#174; Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev___ Net-snmp-users mailing list Net-snmp-users@lists.sourceforge.net Please see the following page to unsubscribe or change other options: https://lists.sourceforge.net/lists/listinfo/net-snmp-users