Snort + SG-3100 = exited on signal 10
-
Ah man, I knew it! I knew the problem while running on the SG-1000 was more complex than the device doesn't have the processing power for it… I mean, while that may be true, it wouldn't exit with a Signal 10 error on start up. Signal 10 (to me) indicates the ARM chip is running malformed instructions meant for an x86 target.
Good luck @bmeeks ! May the programming gods bless you on this one!
-
Ah man, I knew it! I knew the problem while running on the SG-1000 was more complex than the device doesn't have the processing power for it… I mean, while that may be true, it wouldn't exit with a Signal 10 error on start up. Signal 10 (to me) indicates the ARM chip is running malformed instructions meant for an x86 target.
Good luck @bmeeks ! May the programming gods bless you on this one!
Thanks! This is a tough nut to crack. It's not an illegal instruction that's causing the Signal 10 in this case. Instead, it's a problem with something called unaligned memory access. You can Google that term for details about what it is. It's down all the way to the register level inside the CPU and how hardware memory access has to work. The root cause is what many consider poor or bad form C language programming practice when using pointers to reference data in memory. Intel x86 CPUs swallow these kinds of programmer issues and auto-correct them. In the old days, before tons of CPU on-die cache memory and all the fancy instruction execution pipelines of modern CPUs, there was a peformance penalty each time the CPU "fixed up" a C programmer's mistake. Not so much anymore, though. Modern Intel CPUs just basically instantly fix-up the unaligned memory access and there is no perceivable performance penalty. Thus there has not been a push to fix these problems in legacy C programming code. However, other CPUs such as the armv7 used in the SG-3100 don't perform these auto-fixups by default. So you get the errors. The preferred fix is to find all the poor programming practices in the C code and fix them at the source. That is easier to say that it is to actually do… :(. We're still working on it. The problems are within sections of the Snort binary and have nothing to do with the GUI package.
Bill
-
The preferred fix is to find all the poor programming practices in the C code and fix them at the source. That is easier to say that it is to actually do… :(.
Wow, I feel that pain. :-\
Sending you good vibes! :)
Steve
-
Just jumping into this thread to say that I have the same issue, and await a fix hopefully soon.
I originally thought I had misconfigured something.
-
OP Here.
Thank you guys for looking into this. I look forward to a fix.
-Ross
-
Hey guys, until we wait for the fix, you can use Suricata instead, not a complete replacement due to the lack of appid in my opnion, but it works.
The categories mentioned below worked fine.The only category I got errors inside Suricata.log was from malware-cnc, so I've disabled it.
I'm currently using the VRT paid rules, with the following categories:
snort_blacklist.rules
snort_browser-chrome.rules
snort_browser-firefox.rules
snort_browser-plugins.rules
snort_file-multimedia.rules
snort_file-office.rules
snort_file-pdf.rules
snort_malware-backdoor.rules
snort_os-windows.rulesDisabled the stream-events rules due to huge false positives.
And a few of the decoder-events due to same reason.If you guys have any doubts, just ask.
-
Hmm, interesting. Is Suricata still running OK for you?
I initially thought this also but found Suricata crashed out after some time. However I'm re-testing it now and it's still running….so far.
Steve
-
Yes, it
s working perfectly fine. Didn
t have a single crash so far, running only on my LAN, IPS mode (blocking mode enabled), not inline, didn`t test this yet. -
Hmm, interesting. Is Suricata still running OK for you?
I initially thought this also but found Suricata crashed out after some time. However I'm re-testing it now and it's still running….so far.
Steve
Are you running in inline mode or legacy mode? From what I can tell, inline mode isn't ready yet for the 3100 due to lack of driver support, which the team is working on.
-
Running in non-blocking mode currently. One step at a time ;)
Previously it wasn't running at all from what I could see but now seems good at 24hrs+.
Steve
-
Just checking back in. Any movement getting snort fully functional on SG-3100/ARM? I'm really interested in the new app detection stuff, so running Suricata doesn't scratch the itch. Really happy with my SG-3100 so far (but for this). I'm happy to help test/troubleshoot if my rig can be of assistance.
From the thread, it looks non-trivial based on some old bad programming habits. Not sure how hard that is to track down and fix :(
Thanks for any and all help!
Sean
-
Just checking back in. Any movement getting snort fully functional on SG-3100/ARM? I'm really interested in the new app detection stuff, so running Suricata doesn't scratch the itch. Really happy with my SG-3100 so far (but for this). I'm happy to help test/troubleshoot if my rig can be of assistance.
From the thread, it looks non-trivial based on some old bad programming habits. Not sure how hard that is to track down and fix :(
Thanks for any and all help!
Sean
No firm progress yet. I did manage to find where generally in the code it is failing (at least one point). It appears to be in the loading of the Stream5 preprocessor. Debugging this has proven challenging because when I build Snort with debugging enabled it does not crash! It only crashes with debugging disabled. Without the debugging symbols being enabled, troubleshooting the crash is very difficult.
I've not had much time to troubleshoot over the Christmas holidays. Since those are winding down, I should have more time to devote to the troubleshooting task. I have an SG-3100 appliance I am testing with. It was generously provided by the pfSense team.
Bill
-
Ouch, I literally just bought one of these today because I wanted to get introduced to pfSense and things like Snort. Saw mention elsewhere it didn't work on the SG1000 but missed this about the SG3100. I'm subscribed and best of luck but I think I'm going to put in more research on Qotom.
-
Ouch, I literally just bought one of these today because I wanted to get introduced to pfSense and things like Snort. Saw mention elsewhere it didn't work on the SG1000 but missed this about the SG3100. I'm subscribed and best of luck but I think I'm going to put in more research on Qotom.
Reports from other SG-3100 users indicate Suricata works fine on the hardware. Just use Suricata for now. There is no meaningful security difference between it and Snort. The only functional difference is Snort currently offers OpenAppID while Suricata does not, but then Suricata is multi-threaded and has Inline IPS Mode while Snort does not.
Bill
-
Yes Suricata seems to run fine on the SG-3100.
The only issue with it I have seen is that the package does not survive a firmware update for some reason I've yet to determine. It requires un-installing and then re-installing (not just hitting the reinstall button) after updating.
Not a huge issue unless you're following development snapshots and updating everyday. Like me. ;)
Steve
-
Finally some progress on getting Snort to run on the SG-3100 and other Netgate hardware with the armv6/armv7 processors!
I have a Snort package in testing that runs successfully on an SG-3100. The fix involves turning off compiler optimizations for the armv7 CPU used in the SG-3100. That produces a less efficient machine code binary, but at least it runs. I am discussing options with the pfSense team to get their input on how they would like to proceed with this (meaning use the non-optimized binary package that will run, but a little bit slower than an optimized binary would; or investing time and energy to figure out exactly how to fix the optimization routines in the compiler used for armv6/armv7 packages or else alter all the places in the Snort source code where unaligned access can happen). The root cause is the compiler optimizations appear to replace some byte-oriented instructions with multi-byte equivalents to gain a little speed, but the multi-byte equivalents do not support unaligned memory access and thus cause the crash. The details get a bit geeky from there, but you can read all about unaligned memory access and how the different CPUs handle it using a Google search for "unaligned access".
Bill
-
Great news, thank you bmeeks!
-
The fix for Snort on the SG-3100 will be ready soon. It is part of these two Snort updates:
(1) Snort binary update – https://github.com/pfsense/FreeBSD-ports/pull/497
(2) Snort GUI update – https://github.com/pfsense/FreeBSD-ports/pull/496Look for these to appear in Package Manager sometime next week.
Bill
-
I upgraded to the new snort package released in pfSense package manager, 3.2.9.6. I am still seeing issues with crashing. I did a complete removal and reinstall.
Jan 29 16:59:31 kernel mvneta0: promiscuous mode disabled
Jan 29 16:59:31 kernel pid 84107 (snort), uid 0: exited on signal 10
Jan 29 16:58:44 snort 84107 Commencing packet processing (pid=84107) -
@Missionary:
I upgraded to the new snort package released in pfSense package manager, 3.2.9.6. I am still seeing issues with crashing. I did a complete removal and reinstall.
Jan 29 16:59:31 kernel mvneta0: promiscuous mode disabled
Jan 29 16:59:31 kernel pid 84107 (snort), uid 0: exited on signal 10
Jan 29 16:58:44 snort 84107 Commencing packet processing (pid=84107)Could you try going to the interface settings in Snort, then <interface>Preprocs, then under "Stream5 Target-Based Stream Reassembly", uncheck "Track and reassemble TCP sessions. Default is Checked.". Save this and see if the crashes stop… after some troubleshooting with mine this is the only way I could get it to start.</interface>