You are not logged in.
OK, here's where I'm at now.
First, I edited /boot/config.txt and attempted to change the kernel line to the version that I had installed via apt as follows: kernel=vmlinuz-5.10.0-19-arm64
The "file" command reported both files (the previous kernel8.img and the vmlinuz-5.10.0-19-arm64) as "Linux kernel ARM64 boot executable Image, little-endian, 4K pages" -- so I thought it just might work.
When I rebooted, the system was stuck at the initial color "rainbow" screen.
I resized my boot partition (it was 128MB on the ascii image, but 256MB on the chimaera image). I downloaded the file rpi-3-devuan-chimaera-5.10.110-arm64-ext4-2022-11-25.zip from the page https://arm-files.devuan.org/RaspberryPi%20Latest%20Builds/ ... then using dd I overwrote my boot partition with the one from that image. At first it would not complete the boot due to missing partition (I forgot to edit cmdline.txt). After I got that straightened out, it did boot into Linux devuan 5.10.110 #1 SMP PREEMPT Fri Nov 25 00:43:14 CET 2022 aarch64 GNU/Linux
So I guess that part is solved, sort of. I was hoping there would be a simpler way to just upgrade the kernel (like it works in amd64).
Also, upgrading the kernel didn't even fix the iptables problem.
# iptables -A INPUT -s 1.2.3.4 -j DROP
iptables/1.8.7 Failed to initialize nft: Protocol not supported
# iptables -V
iptables/1.8.7 Failed to initialize nft: Protocol not supported
I think I've got the solution for this, though.
I found on https://superuser.com/questions/1480986/iptables-1-8-2-failed-to-initialize-nft-protocol-not-supported:
update-alternatives --set iptables /usr/sbin/iptables-legacy
update-alternatives --set ip6tables /usr/sbin/ip6tables-legacy
I had to copy the /lib/modules/5.10.110 from the latest image file to my live system, too (which of course I forgot to do previously).
Now, iptables seems to be functioning correctly.
The only question remaining is, whether or not there's a simpler kernel upgrade process. At least things are working again, and it is possible to upgrade kernel, just a bit more time-consuming than I'm used to :-)
I've been reading a few topics on this forum, but haven't yet come across a satisfactory answer.
Here's my situation:
I've got a raspberry pi 3b+. I originally used the devuan ascii image on sd card, later moved to external usb drive. After running this setup for a few years, I finally (today) decided to dist-upgrade to beowulf, and then to chimaera. It all went pretty smoothly. The one issue that I've noticed during startup now is that iptables refuses to load the rules I have stored in a script. An example rule is: iptables -A INPUT -s 1.2.3.4 -j DROP and the message that was returned on beowulf is: iptables: Invalid argument. Run `dmesg' for more information.. dmesg contains nothing useful or related to the iptables command. On chimaera the error is: iptables v1.8.7 (nf_tables): Could not fetch rule set generation id: Invalid argument.
From some internet searching, I believe that the problem is due to an outdated kernel. uname -a result is: Linux devuan 4.16.14-v8+ #1 SMP PREEMPT Tue Jun 5 18:50:10 CEST 2018 aarch64 GNU/Linux.
I then tried: apt install linux-image-arm64 which proceeded to download and install linux-image-5.10.0-19-arm64_5.10.149-2_arm64.deb
This did create the files in /boot, but they were not loaded upon a system reboot, it's still on 4.16.14-v8+.
I *believe* (but may be mistaken) that the newer rpi images contain a newer kernel. Could I download the chimaera image, and copy the contents of its /boot to my /boot in order to load a newer kernel? Or is there another way to perform a kernel upgrade on a running system? Any help is greatly appreciated.
FYI, after adding repository, apt install loksh says...
Preparing to unpack .../archives/loksh_7.1-1_amd64.deb ...
Unpacking loksh (7.1-1) ...
Setting up loksh (7.1-1) ...
update-alternatives: error: alternative path /bin/loksh doesn't exist
dpkg: error processing package loksh (--configure):
installed loksh package post-installation script subprocess returned error exit status 2
loksh is located in /usr/bin, not in /bin
This occurred in devuan chimaera
Installing inetutils-ping (which removed iputils-ping) did remove that warning, so that part is solved. Thanks :-)
The change with tpm2-abrmd (adding the '-u tss' to line 17 of /etc/init.d/tpm2-abrmd) didn't help.
# service tpm2-abrmd start
Starting TPM2 Access Broker and Resource Management Daemon: tpm2-abrmdIllegal option -u
/sbin/start-stop-daemon: unable to stat //tss (No such file or directory)
failed!
I *do* have a user "tss" listed in /etc/passwd. I guess I'm wondering, do I need this service?
After a little bit of internet searching, I think I may not even have tpm enabled on my motherboard. Maybe that's part of the problem. I'll reboot and check the ol' bios.
OK, under the "Trusted Computing" section of bios setup, my motherboard informed me that no device was found. Guess it's not built-in and I never added it. So I'll remove the tpm2-abrmd package, not sure why it got pulled in. Sorry for the noise
This weekend I dist-upgraded from Beowulf to Chimaera.
The upgrade itself went well, with one package problem I am stuck at (I've uninstalled for now)
The package that fails is: "tpm2-abrmd" :
Setting up tpm2-abrmd (2.3.3-1+b2) ...
Starting TPM2 Access Broker and Resource Management Daemon: tpm2-abrmd
** (process:6742): WARNING **: 13:44:06.774: tcti_conf before: "device:/dev/tpm0"
** (tpm2-abrmd:6742): WARNING **: 13:44:06.774: tcti_conf after: "device:/dev/tpm0"
Refusing to run as root. Pass --allow-root if you know what you are doing.
failed!
invoke-rc.d: initscript tpm2-abrmd, action "start" failed.
dpkg: error processing package tpm2-abrmd (--configure):
installed tpm2-abrmd package post-installation script subprocess returned error exit status 1
I'm not going to "--allow-root", since I obviously don't know what I'm doing here. I've uninstalled the package (not sure what it is, or why I had it to begin with). Is this anything important? I could note, this is on my desktop computer, not an internet server, if that makes any difference.
The second question has to do with ping.
$ ping -c 1 127.0.0.1
ping: socket: Address family not supported by protocol
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
64 bytes from 127.0.0.1: icmp_seq=1 ttl=64 time=0.025 ms
Any fix for this warning? I keep getting the warning (Address family not supported by protocol), but the ping itself does succeed. Passing the "-4" switch will take the warning away; I might just make an alias to do that if there's not another workaround. Possibly related: https://github.com/iputils/iputils/issues/293.
These log entries look very similar to an issue I had at the beginning of the year.
Aug 24 02:19:07 devuan-ceres mysqld: 2022-08-24 2:19:07 0 [Note] InnoDB: Starting crash recovery from checkpoint LSN=1651026,1651026
Aug 24 02:19:07 devuan-ceres mysqld: 2022-08-24 2:19:07 0 [ERROR] InnoDB: Missing FILE_CHECKPOINT at 1651026 between the checkpoint 1651026 and the end 1651038.
Aug 24 02:19:07 devuan-ceres mysqld: 2022-08-24 2:19:07 0 [ERROR] Plugin 'InnoDB' init function returned error.
Aug 24 02:19:07 devuan-ceres mysqld: 2022-08-24 2:19:07 0 [ERROR] Unknown/unsupported storage engine: InnoDB
What fixed it for me was removing (actually, renaming, in case it was needed later; it wasn't) the file /var/lib/mysql/ib_logfile0
cd /var/lib/mysql
mv ib_logfile0 ib_logfile0.bad-renamed
service mariadb start
[root@devuan ~]# ss -tnlp | grep 4711 [root@devuan ~]#
Nothing.
Phooey. I was hoping another process (the old pihole?) was holding on to that port. You can't connect, because the service is not running. The service won't start (or rather, won't stay running) because of... something yet unknown.
Earlier you had posted this from your log:
[2022-08-19 18:05:28.696 1140M] Error opening IPv6 telnet socket: Address family not supported by protocol (97)
Would you mind trying this, add this line into the file /etc/pihole/setupVars.conf
IPV6_ADDRESS=
Then run
# service pihole-FTL start
If it doesn't start properly, check /var/log/pihole/FTL.log for other clues.
Also, in the web interface, there is a section "Pi-hole diagnosis" under the "Tools" menu. I'm not sure whether or not that is useful.
If all else fails, nuke it (including configs) and re-install. Sometimes that is quicker (but less satisfying) than finding the actual problem.
root@chimaera:~# service lighttpd status lighttpd is running.
can you open the pihole web admin page? ( try http://192.168.1.3/admin ... or whatever the machine's ip address is)
root@chimaera:~# cat /etc/pihole/setupVars.conf WEBPASSWORD=xxxx BLOCKING_ENABLED=true DNSSEC=false REV_SERVER=false PIHOLE_INTERFACE=eth0 PIHOLE_DNS_1=127.0.0.1#5335 PIHOLE_DNS_2= QUERY_LOGGING=true INSTALL_WEB_SERVER=true INSTALL_WEB_INTERFACE=true LIGHTTPD_ENABLED=true CACHE_SIZE=10000 DNS_FQDN_REQUIRED=true DNS_BOGUS_PRIV=true DNSMASQ_LISTENING=local root@chimaera:~#
Basically the same as mine, except for PIHOLE_DNS_1=127.0.0.1#5335
This might be valid, I'm not sure, but it's not how mine is set up at all. My pihole goes out to other DNS servers for its non-piholed queries. I also don't see IPV4_ADDRESS or IPV6_ADDRESS in your configuration. Mine is:
PIHOLE_DNS_1=208.67.222.222
PIHOLE_DNS_2=208.67.220.220
PIHOLE_DNS_3=198.6.100.194
PIHOLE_DNS_4=9.9.9.9
IPV4_ADDRESS=192.168.1.4/24
IPV6_ADDRESS=
[2022-08-19 18:05:28.675 1140M] Listening on port 4711 for incoming IPv4 telnet connections [2022-08-19 18:05:28.696 1140M] Error opening IPv6 telnet socket: Address family not supported by protocol (97)
As you can see, port 4711 is mentioned.
If fsockopen gets refused trying to connect to 127.0.0.1:4711, then there is a permissions problem with port 4711.
How do I troubleshoot that, specifically?
what happens when you run these:
root@devuan:/etc/pihole# ss -tnlp | grep 4711
LISTEN 0 5 127.0.0.1:4711 *:* users:(("pihole-FTL",pid=2819,fd=10))
LISTEN 0 5 ::1:4711 :::* users:(("pihole-FTL",pid=2819,fd=13))
root@devuan:/etc/pihole# telnet 127.0.0.1 4711
Trying 127.0.0.1...
Connected to 127.0.0.1.
Escape character is '^]'.
^]
telnet> quit
Connection closed.
root@devuan:/etc/pihole#
root@chimaera:~# service pihole-FTL status [ ] pihole-FTL is not running root@chimaera:~#
What is the output of:
service lighttpd status
FYI, my (very-old) FTL.php contains this:
function connectFTL($address, $port=4711)
{
if($address == "127.0.0.1")
{
// Read port
$portfile = file_get_contents("/var/run/pihole-FTL.port");
if(is_numeric($portfile))
$port = intval($portfile);
}
// Open Internet socket connection
$socket = @fsockopen($address, $port, $errno, $errstr, 1.0);
return $socket;
}
The next place to look would be /etc/pihole/setupVars.conf... check if anything looks off in there.
Oh, and don't forget the log files: /var/log/pihole/FTL.log and /var/log/pihole/pihole.log
The Pale Moon project does not currently create 32-bit binaries themselves, but they recommend the 32-bit builds by Steve Pusser. Links to his repos and deb packages for GTK2 and GTK3 versions are at the top of Pale Moon's 3rd party builds page here: http://www.palemoon.org/contributed-builds.shtml
Thanks for this, I knew about Steve's repository, but wasn't aware that 32-bit was included there (I rarely use my 32-bit machine).
This is strange:
$ aptitude versions firefox-esr
Package firefox-esr:
p 78.14.0esr-1~deb10u1 oldstable 500
i 91.12.0esr-1~deb10u1 oldstable-sec 500
Package firefox-esr:i386:
p 78.14.0esr-1~deb10u1 oldstable 500
p 78.15.0esr-1~deb10u1 oldstable-sec 500
It appears firefox-esr:i386 is indeed at 78.15.0 with beowulf (above). However, using ascii sources gives this:
$ aptitude versions firefox-esr
Package firefox-esr:
p 68.10.0esr-1~deb9u1 oldoldstable 500
p 91.11.0esr-1~deb9u1 oldoldstable- 500
i 91.12.0esr-1~deb10u1 100
Package firefox-esr:i386:
p 68.10.0esr-1~deb9u1 oldoldstable 500
p 91.11.0esr-1~deb9u1 oldoldstable- 500
I suppose you could (temporarily) change sources to ascii, apt-update, install firefox-esr (91.11.0), then restore sources to beowulf ...
I don't have the *exact* same setup as you (32-bit on beowulf)... I do have amd64 running beowulf, and i686 running chimaera. Both of them show firefox at 91.12.0esr. my arm64 running ascii shows firefox at 91.11.0esr. Have you run apt update recently? Do you have anything in /etc/apt/preferences.d/ that would prevent firefox upgrade (apt pinning)?
Apologies for the late response. It was, apparently, a hardware problem after all, although I'm not sure exactly what. Something with the CPU or motherboard (I didn't have a spare of either that I could interchange and narrow down the problem).
An alternate option for this particular case (apparmor) would be to create a file, e.g. /etc/apt/preferences.d/noapparmor
Contents:
Package: apparmor*
Pin: release n=beowulf
Pin-Priority: -1
Change the release name if you're not using beowulf.
I have 1 desktop that won't properly boot chimaera. This is a machine I used once a week.
It ran beowulf fine (I may just go back to that).
I updated the apt sources to chimaera, did the upgrade and dist-upgrade successfully. Rebooted into chimaera okay.
Some time later (the next week?) it wouldn't complete the boot -- it would go so far and then reboot itself (stuck in a reboot-loop).
I was able to boot into an older kernel, and it started up okay.
The following week, same problem. Only the older kernel wouldn't complete its boot either. I went to recovery mode, and was able to log in, exited and GUI started up fine.
The following week, nothing I did was successful. Even "recovery mode" does not complete far enough to get me to login -- it reboots and reboots and reboots! I let it continue this for over a half hour before shutting it down.
I've replaced this desktop with another one, so a fix is not something that is urgent or even necessary. I have noticed, in attempts to get it working, that it will successfully boot jessie desktop-live USB; but will not even boot up from chimaera desktop-live USB (it does the reboot thing).
Perhaps there is some incompatible hardware ... but if that's the case I am confused why it initially worked after the dist-upgrade. Should I just go back to beowulf?
I don't even know where to start looking for the cause ... syslog ? kern.log ? record a video of the boot to get the last message before it reboots ?
The chimaera desktop-live USB does work on other hardware so I don't think that it is a problem, either.
An old inxi output (showing the correct hardware at least; other info may be outdated)
System:
Host: fcc-audio Kernel: 5.8.0-0.bpo.2-amd64 x86_64 bits: 64 compiler: N/A
Desktop: Xfce 4.12.4 tk: Gtk 2.24.32 info: xfce4-panel wm: xfwm4
dm: SLiM 1.3.6 Distro: Devuan GNU/Linux 3 (beowulf)
Machine:
Type: Desktop Mobo: MSI model: FM2-A55M-E33 (MS-7721) v: 2.0 serial: N/A
BIOS: American Megatrends v: 11.6 date: 02/12/2015
CPU:
Topology: Quad Core model: AMD A10-6800K APU with Radeon HD Graphics
bits: 64 type: MCP arch: Piledriver rev: 1 L1 cache: 192 KiB
L2 cache: 2048 KiB
flags: lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm
bogomips: 32736
Speed: 1993 MHz min/max: 2000/4100 MHz boost: enabled Core speeds (MHz):
1: 1996 2: 1996 3: 2092 4: 2044
Graphics:
Device-1: AMD Richland [Radeon HD 8670D] vendor: Micro-Star MSI
driver: radeon v: kernel bus ID: 00:01.0 chip ID: 1002:990c
Display: server: X.Org 1.20.4 driver: ati,radeon
unloaded: fbdev,modesetting,vesa resolution: 1024x768~60Hz
OpenGL: renderer: AMD ARUBA (DRM 2.50.0 / 5.8.0-0.bpo.2-amd64 LLVM 7.0.1)
v: 4.3 Mesa 18.3.6 compat-v: 3.1 direct render: Yes
Audio:
Device-1: AMD Trinity HDMI Audio vendor: Micro-Star MSI
driver: snd_hda_intel v: kernel bus ID: 00:01.1 chip ID: 1002:9902
Device-2: AMD FCH Azalia vendor: Micro-Star MSI driver: snd_hda_intel
v: kernel bus ID: 00:14.2 chip ID: 1022:780d
Sound Server: ALSA v: k5.8.0-0.bpo.2-amd64
Network:
Device-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet
vendor: Micro-Star MSI driver: r8169 v: kernel port: e000 bus ID: 01:00.0
chip ID: 10ec:8168
IF: eth0 state: down mac: xx:xx:xx:xx:xx:xx
Drives:
Local Storage: total: 88.98 GiB used: 22.56 GiB (25.4%)
ID-1: /dev/sda vendor: Western Digital model: WD800JD-60LSA5
size: 74.53 GiB speed: 3.0 Gb/s serial: WD-XXXXXXXXXXXX rev: 1E03
temp: 35 C scheme: MBR
ID-2: /dev/sdf type: USB vendor: Toshiba model: TransMemory
size: 14.45 GiB serial: XXXXXXXXXXXXXXXXXXXXXXXX rev: 1.00 scheme: MBR
Partition:
ID-1: / size: 21.88 GiB used: 8.81 GiB (40.3%) fs: ext4 dev: /dev/sda2
ID-2: /boot size: 938.0 MiB used: 177.7 MiB (18.9%) fs: ext2
dev: /dev/sda1
ID-3: /home size: 41.92 GiB used: 6.25 GiB (14.9%) fs: ext4 dev: /dev/sda3
ID-4: swap-1 size: 8.41 GiB used: 0 KiB (0.0%) fs: swap dev: /dev/sda4
Sensors:
System Temperatures: cpu: 10.8 C mobo: N/A gpu: radeon temp: 0 C
Fan Speeds (RPM): N/A
Info:
Processes: 162 Uptime: 1h 48m Memory: 7.06 GiB used: 725.2 MiB (10.0%)
Init: SysVinit v: 2.93 runlevel: 2 default: 2 Compilers: gcc: 8.3.0
alt: 6/8 Shell: bash (sudo) v: 5.0.3 running in: xfce4-terminal
inxi: 3.0.32
Adding to this thread, in case someone finds this useful, or needs the info at some later point. To get the wifi (b43) working on this imac, I followed the steps at http://linuxwireless.sipsolutions.net/e … ivers/b43/ under "Other distributions" to build the driver manually.
In case that page goes down, basically it came down to this:
wget http://bues.ch/b43/fwcutter/b43-fwcutter-018.tar.bz2
wget http://www.lwfinger.com/b43-firmware/broadcom-wl-5.100.138.tar.bz2
tar xjf b43-fwcutter-018.tar.bz2
cd b43-fwcutter-018
make
sudo make install
cd ..
tar xjf broadcom-wl-5.100.138.tar.bz2
sudo b43-fwcutter -w /lib/firmware broadcom-wl-5.100.138/linux/wl_apsta.o
a reboot or modprobe b43 should get the device driver loaded, and wifi available. I also had to set the wireless interface to "wlan0" in wicd preferences., but this might be done automatically if you reboot rather than use the modprobe command.
It actually *did* boot the amd64 live desktop dvd. :-) Installation went pretty smoothly from there.
This machine doesn't look like the one pictured on your link, snork. The EFI situation could be similar though, I don't really know. It does seem to be from the same era, though. This is an all-in-one thing (basically the hardware is all mounted behind/under the screen. no tower, more difficult to replace individual parts).
inxi reports this as:
System: Host: imac Kernel: 4.19.0-14-amd64 x86_64 bits: 64 compiler: gcc v: 8.3.0 Desktop: Xfce 4.12.4 tk: Gtk 2.24.32
info: xfce4-panel wm: xfwm4 dm: SLiM 1.3.6 Distro: Devuan GNU/Linux 3 (beowulf)
Machine: Type: Desktop System: Apple product: iMac6,1 v: 1.0 serial: <> Chassis: type: 13 v: Mac-F4218FC8
serial: <>
Mobo: Apple model: Mac-F4218FC8 v: DVT serial: <> UEFI: Apple v: IM61.88Z.0093.B07.0706281250
date: 06/28/07
CPU: Topology: Dual Core model: Intel Core2 T7400 bits: 64 type: MCP arch: Core Merom rev: 6 L2 cache: 4096 KiB
flags: lm nx pae sse sse2 sse3 ssse3 vmx bogomips: 8645
Speed: 1589 MHz min/max: 1000/2167 MHz Core speeds (MHz): 1: 1101 2: 1049
Thank you both for the replies.
Unfortunately, I am not able to boot the amd64 debian buster netinstall cd, either. The prompt that comes up trying to boot any amd64 disc (so far) is:
1.
2.
Select CD-ROM Boot Type :
(and all keyboard input is ignored at this point)
Also, thanks for the hint on using the desktop *live* installer disc. I'll try that next, I think I should be able to get that to work. Even if I can't boot it directly, I should be able to mount it and run the installer from it.
I was recently given an old imac computer. It was running a very old mac os 10.6 and would not install a newer version. So I've decided to put devuan on it.
The CPU (an Intel Core 2 Duo) is indeed 64-bit capable. However, the computer refuses to boot devuan amd64 dvds. I've tried both ascii and beowulf, full desktop and netinstall; tried from both the internal drive and an external drive. It will not boot from these discs. What I have been able to do so far, is install from the i686 disc, then install an amd64 kernel and boot from it. I've been unable, at this point, to change all of the system packages over to amd64. I can mount and read the amd64 desktop installer dvd, but can't seem to start the installer from there without booting from it (?) I mainly want to use amd64 since it's required for most modern web browsers (most seemed to stop updating 32-bit versions several years ago). Any advice? I've read a few different "cross-grading" articles, but so far I've only managed to break apt/dpkg.
I use devuan ascii daily on my desktop, 3rd gen core i5 with 32gb ram, nvidia gtx650, dual monitor, 2tb + 250gb hdd
also on a raspberry pi 3b+ 16gb microsd
devuan jessie I use twice a week on laptop, it's got some amd cpu from around 2010-2012, 8gb ram and a 250gb hdd
also devuan jessie once a week on a very old desktop (actually planning to replace it with a newer desktop soon), pentium 4 with 512mb ram and a 40gb hdd
also considering to replace debian 8 with devuan ascii on a remote server (or two) that I manage for work
I think this is missing a part.
While parted does show the new size:
(parted) print
Model: SD SC16G (sd/mmc)
Disk /dev/mmcblk0: 15.9GB
Sector size (logical/physical): 512B/512B
Partition Table: msdos
Disk Flags:
Number Start End Size Type File system Flags
1 1049kB 135MB 134MB primary fat16 lba
2 135MB 15.9GB 15.8GB primary ext4
# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
mmcblk0 179:0 0 14.9G 0 disk
|-mmcblk0p1 179:1 0 128M 0 part /boot
`-mmcblk0p2 179:2 0 14.7G 0 part /
however ... the filesystem doesn't seem quite right:
# df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/root 1743136 777240 859300 48% /
devtmpfs 438328 0 438328 0% /dev
tmpfs 88764 1200 87564 2% /run
tmpfs 5120 4 5116 1% /run/lock
tmpfs 177520 0 177520 0% /run/shm
/dev/mmcblk0p1 130798 35754 95044 28% /boot
/dev/root size should be much larger, it is only reporting 1,743,136 KB (~1.7 GB).
To reply to my self (or is it preferred to edit the original post):
Apparently this didn't begin with the Ascii upgrade (which I did 2 weeks ago) -- /var/log/pm-suspend.log has a date of Fri Jun 29 (morning wakeup time). There's no record of Fri Jun 29 (night sleep time) or Sat Jun 30, Sun Jul 1 in /var/log/pm-suspend.log at all. This is likely when the suspend/resume hooks stopped running. This seems strange to me. I have rebooted in the meantime, which did not correct the issue. On Jun 29 I did remove mono and a bunch of things along with it... I'll see if any of that is related. So at this point it seems less likely a bug and more likely something I've done to my own system.
[Mon Jul 2 AM edit]
So last night I directly ran /usr/sbin/pm-suspend instead of /usr/bin/xfce4-session-logout --suspend
The result was the hooks were called on suspend/resume, /var/log/pm-suspend.log was updated, screen was not locked on resume (of course).
I suppose I could change my suspend routine to call /usr/bin/xflock4 then run /usr/sbin/pm-suspend, after granting my user account the ability to run /usr/sbin/pm-suspend . Still unclear as to what changed with xfce4-session-logout .
In Jessie this worked:
/usr/bin/xfce4-session-logout --suspend
before suspending, the system would call
/etc/pm.d/sleep/50_unmount
which I have unmount certain things...
Now in Ascii, the /etc/pm.d/sleep/50_unmount is no longer called when suspending.
In fact, none of the files in /etc/pm.d/sleep are called when resuming from suspend, either.
Is this a bug or intentional change, or do I have something else misconfigured?
I can work around this, but just wondering what has changed or if /etc/pm.d/sleep/* has been abandoned for some other way of calling scripts on suspend/hibernate/thaw/resume.
First ran on a (backed-up) virtual machine. The only "major" problem was with wicd, a missing file /etc/dbus-1/system.conf -- I copied from my host system and resumed the upgrade... it finished, rebooted, all seems to be okay.
Then backed up my "baremetal" system. Some other errors (depencencies?) partway through, had to run a "apt-get --fix-broken install" then resume the dist-upgrade. Same issue with wicd, copied system.conf from most recent backup and resumed upgrade. Rebooted, no gui. Reinstalled nvidia driver, rebooted, all seems okay. Remmina was uninstalled. Opened synaptic to some error, had to run "dpkg --configure -a", now synaptic works. Had to force specific remmina version (something about backports?), now it is okay. Had to redo some other minor tweaks, but overall a rather painless experience. Just kind of curious about that wicd thing, it seems the system.conf is deleted too early in the upgrade process?
I know this is supposed to work, in theory, but is it recommended to try debian 8 -> devuan ascii upgrade on a remote (ssh) machine? Will I lose ssh connectivity during the upgrade process, what other things should I be aware of before attempting it?
Again, thanks to the developers for their work in this project, and the people who provide forum support / other chat.