grobber's questions

I am on Ubuntu 20.04.3 LTS x86_64. The program w3m fails to launch as follows:

$ w3m -h

Wrong __data_start/_end pair
Aborted (core dumped)

I have tried

• installing and reinstalling;
• setting a stack-size bound with ulimit, as suggested here.

No luck.

The issue also came up in another thread, but that is specific to the Windows Subsystem for Linux, which I am not on.

Edit

Getting closer: this is a kernel issue.

The machine is running a custom-compiled kernel, currently at v5.17-rc3. If I go back to version 5.16 of the kernel I do not see the issue.

I'm not sure how things have changed for the 5.17 release, but it seems some modification to the kernel config is needed.

Edit 2

Found something promising on the kernel mailing list.

I have yet to test that patch, but it fits with what I'm seeing: the problem appearing after v5.17-rc1.

I am on Ubuntu 18.04. Every apt update && apt upgrade results in

W: Download is performed unsandboxed as root as file '/var/lib/apt/lists/partial/archive.ubuntu.com_ubuntu_dists_bionic_InRelease' couldn't be accessed by user '_apt'. - pkgAcquire::Run (13: Permission denied)

I am aware that this has come up. My problem is that as per this answer, I already have

/var/lib/apt/lists

owned by _apt:root, recursively: I've already run

sudo chown -Rv _apt:root /var/lib/apt/lists

And /var/lib/apt/lists/partial is chmodded 700, so _apt should have write access fine.

Additional info:

This all started when I symlinked /var/lib elsewhere, to a directory under /root/.

I had to do this because although I have root access, the machine was set up by someone else at work with /var as a separate partition. It was too small and got full, so the move+symlinking was the solution at the time.

Edit

Following the accepted answer, I moved the symlink target from under /root/; the warning is gone.

I know there are other posts here asking this, but those suggestions do not help, the configuration keeps changing and I simply cannot get my own changes to persist.

I have the Google 8.8.8.8 and 8.8.4.4 servers in my Network Manager configuration in the GUI, and

systemd-resolv --status returns

Link 2 (enp38s0f1)
      Current Scopes: DNS    
DefaultRoute setting: yes    
       LLMNR setting: yes    
MulticastDNS setting: no     
  DNSOverTLS setting: no     
      DNSSEC setting: no     
    DNSSEC supported: no     
  Current DNS Server: 8.8.8.8
         DNS Servers: 8.8.8.8
                      8.8.4.4
          DNS Domain: ~.   

Nevertheless, dig google.com shows

;; Query time: 0 msec
;; SERVER: 192.168.1.1#53(192.168.1.1)
;; WHEN: Mon Oct 05 11:08:26 EDT 2020
;; MSG SIZE  rcvd: 83

so my default route 192.168.1.1 as DNS. Furthermore, /etc/resolv.conf is a symlink to /run/systemd/resolve/resolv.conf, which in turn reads

# Generated by dhcpcd from enp38s0f1.dhcp, wlp0s20f3.dhcp
# /etc/resolv.conf.head can replace this line
domain fios-router.home
nameserver 192.168.1.1
# /etc/resolv.conf.tail can replace this line

The only enp38s0f1.dhcp file on this system (I've searched /) is /run/dhcpcd/resolv.conf/enp38s0f1.dhcp, which reads

# Generated by dhcpcd from enp38s0f1.dhcp
domain fios-router.home
search fios-router.home
nameserver 192.168.1.1

I have tried to write in other nameservers there, but they do not persist. In the past I've made the changes persistent by making the file immutable with chattr +i, but

lsattr /run/dhcpcd/resolv.conf/enp38s0f1.dhcp

returns

lsattr: Inappropriate ioctl for device While reading flags on /run/dhcpcd/resolv.conf/enp38s0f1.dhcp

so that attribute is not available here. And besides, I shouldn't have to make the file immutable anyway: there should presumably be some way I can control the DNS server..

Edit:

In response to comments below:

$ dpkg -l *dnsmasq*
---
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name             Version         Architecture Description
+++-================-===============-============-============================================
un  dnsmasq          <none>          <none>       (no description available)
ii  dnsmasq-base     2.80-1.1ubuntu1 amd64        Small caching DNS proxy and DHCP/TFTP server
un  dnsmasq-base-lua <none>          <none>       (no description available)

I am running TeXLive 2015 on 16.04 (32 bit), installed manually, according to the instructions at

https://www.tug.org/texlive/quickinstall.html

I have been trying to upgrade to TeXLive 2016. I understand I can proceed as before, resulting in the directory /usr/local/texlive/2016, etc., but I would like to install via apt instead.

The issue is: when I finish apt-get installing texlive-full (using the jonathonf ppa repository) I do not know how to get Ubuntu to switch to running pdflatex through 2016 instead of the old 2015.

The command tex -v still returns Tex Live 2015.