Re: Armv7 (rpi2) getting stuck in buildworld for -current

On May 14, 2023, at 12:31, Mark Millard <[email protected]> wrote:

> On May 14, 2023, at 08:36, bob prohaska <[email protected]> wrote:
> 
>> Lately a Pi2 running -current has gotten stuck while buildworld is running.
>> There's no escape to debugger, no obvious errors on the console and only
>> modest swap use (tens of MB). So far the stoppages have been when building
>> clang.
>> 
>> One possible culprit is /boot/loader.conf, which has accumulated some baggage
>> over the years:
>> 
>> bob@www:/usr/src % more /boot/loader.conf
>> # Configure USB OTG; see usb_template(4).
>> hw.usb.template=3
>> umodem_load="YES"
>> # Disable the beastie menu and color
>> beastie_disable="YES"
>> loader_color="NO"
>> vm.pageout_oom_seq="4096"
>> #vm.pfault_oom_attempts="3"
>> vm.pfault_oom_attempts="120"
>> vm.pfault_oom_wait="20"
> 
> (I oroginally had a note here but I think it
> would just confuse things and not be tied to
> your problem.)  . . .
> 
> However, I expect that your user process had
> its kernel stack swapped out. See below.
> 
>> kern.cam.boot_delay="20000"
>> vfs.ffs.dotrimcons="1"
>> vfs.root_mount_always_wait="1"
>> filemon_load="YES"
>> net.inet.tcp.tolerate_missing_ts="1"
>> vm.swap_enabled=0
>> vm.swap_idle_enabled=0
> 
> Those last two lines are for avoiding having
> interactive sessions (sshd, serial console)
> processes end up with their kernel stacks swapped
> out. (But it does so by preventing such for all
> kernel stacks, not just the ones of interest.)
> 
> When a kernel stack is swapped out for a
> process/thread, the process/thread can not run
> at all until the kernel stack is is read back
> into the kernel.
> 
> Those last two lines you have in a file for
> tunables --but are not tunables:
> 
> # sysctl -T vm.swap_enabled
> # sysctl -T vm.swap_idle_enabled
> # 
> 
> Compared that to the check for the writable
> category:
> 
> # sysctl -W vm.swap_enabled
> vm.swap_enabled: 0
> # sysctl -W vm.swap_idle_enabled
> vm.swap_idle_enabled: 0
> # 
> 
> In my environment, I use /etc/sysctl.conf , which
> is a place appropriate for non-tunable but writable
> sysctl values:
> 
> # grep vm.swap_ /etc/sysctl.conf 
> vm.swap_enabled=0
> vm.swap_idle_enabled=0
> 
> I suggest moving the assignments to /etc/sysctl.conf .
> I expect that this will get rid of your problem once
> you reboot with them in a right place. (You can also
> interactively set them via sysctl use.)
> 
> I suggest avoiding confusions by not having copies of
> those 2 lines in /boot/loader.conf (where they will
> not work).
> 
>> --More--(END)
>> 
>> However, the problem emerged well after the changes above were made.
>> 
>> 
>> A running diary of experiments is at
>> http://www.zefox.net/~fbsd/rpi2/crashes/20230514/armv7hang
> 
> There you report reducing the swap space partition size.
> Were you getting the message about the swap possibly being
> mistuned prior to that?
> 
> For 1 GiByte of RAM 3647M looks to me to likely be a little
> below where that message about mistuning shows up. If you
> were not getting the message, the size should have been
> fine.
> 
> In other words, I expect it is appropriate to put back
> the original size (or some approximation of it that
> avoids the message about possibly being mistuned).
> 
> 
> Everything that you reported looks to me to be consistent
> with some kernel stacks having been swapped out for some
> processes/threads that would otherwise be involved in
> interactive I/O activity.
> 

Just adding an FYI about:

load-time-Tunable and Writable
vs.
load-time-Tunable (spanning both writable and read-only ones)
vs.
Writable (spanning both tunable and not)

and loader.conf vs. sysctl.conf appropriateness:

# sysctl -aTW | wc # ( all: either loader.conf or sysctl.conf )
     431     860   12410

# sysctl -aT | wc  # ( 617-431: just loader.conf )
     617    1231   17343

# sysctl -aW | wc  # ( 1415-431: just sysctl.conf )
    1415    2821   42158

===
Mark Millard
marklmi at yahoo.com