[Bug 268246] crash and panic using pfsync on 13.1-RELEASE

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=268246

--- Comment #45 from jja...@gmail.com ---
(In reply to Kristof Provost from comment #44)

I ran memory tests, disk tests, and even completely rebuilt the machine before
opening a bug report. 

I'll drop this patch in today, and get back to you.

-- 
You are receiving this mail because:
You are the assignee for the bug.


[Bug 268246] crash and panic using pfsync on 13.1-RELEASE

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=268246

--- Comment #46 from Nick Hilliard  ---
am seeing this panic verbatim on 13.1-RELEASE-p3 (backup) with 12.3-RELEASE-p1
primary. Both instances running as xen VMs, xn driver.

-- 
You are receiving this mail because:
You are the assignee for the bug.


[Bug 268246] crash and panic using pfsync on 13.1-RELEASE

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=268246

--- Comment #47 from Nick Hilliard  ---
primary upgraded to 13.1-RELEASE-p3 and now it's crashing too, within seconds
of enabling pfsync (i.e. panic with kernel uptime 25s)

rc config variables:

--
ifconfig_xn3="mtu 1504 up"
ifconfig_vlan140="inet 10.x.y.1/24 vlan 140 vlandev xn3 up"

pf_enable="YES"
pflog_enable="YES"

pfsync_enable="YES"
pfsync_syncdev="vlan140"
pfsync_syncpeer="10.x.y.2"
pfsync_ifconfig="defer"
--

the vlan140 mtu is 1500 on both devices.

vlan140: flags=8843 metric 0 mtu 1500

-- 
You are receiving this mail because:
You are the assignee for the bug.


[Bug 266442] kernel page fault on packet with broken lengths if ipfilter is loaded

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=266442

--- Comment #3 from Cy Schubert  ---
Created attachment 239852
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=239852&action=edit
Notify pfil of dropped mangled packet

Can you try this patch, please. Though it fixes the problem the final patch may
return a different return code or may increment a counter. For now it avoids
the panic.

-- 
You are receiving this mail because:
You are on the CC list for the bug.


[Bug 266442] kernel page fault on packet with broken lengths if ipfilter is loaded

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=266442

Cy Schubert  changed:

   What|Removed |Added

 Attachment #239852|0   |1
is obsolete||

--- Comment #4 from Cy Schubert  ---
Created attachment 239853
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=239853&action=edit
Updated patch with DTrace probe

This patch adds a DTrace SDT.

-- 
You are receiving this mail because:
You are on the CC list for the bug.


[Bug 266442] kernel page fault on packet with broken lengths if ipfilter is loaded

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=266442

Cy Schubert  changed:

   What|Removed |Added

 Attachment #239853|0   |1
is obsolete||

--- Comment #5 from Cy Schubert  ---
Created attachment 239857
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=239857&action=edit
Also add to the fr_bad packet count.

In addition to a DTrace probe, add to the bad packet count. These are the
proposed two commits.

-- 
You are receiving this mail because:
You are on the CC list for the bug.


[Bug 266442] kernel page fault on packet with broken lengths if ipfilter is loaded

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=266442

Cy Schubert  changed:

   What|Removed |Added

 Status|Open|In Progress

-- 
You are receiving this mail because:
You are on the CC list for the bug.


[Bug 268246] crash and panic using pfsync on 13.1-RELEASE

2023-02-01 Thread bugzilla-noreply
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=268246

--- Comment #48 from Kristof Provost  ---
(In reply to Nick Hilliard from comment #47)
Do you have other non-1500 byte MTU interfaces on that machine?

I'm still unable to reproduce this panic. We do know it's related to defer mode
(which you have, so that much matches at least), but I wonder if there's an MTU
factor or not.

-- 
You are receiving this mail because:
You are the assignee for the bug.