I’m setting up a new system, and I always create a new key when I bulid a new desktop… Having not done it in a few years, I wanted to see what the recomended ssh key looks like these days.

watching the founder of github talk about git.

https://www.youtube.com/watch?v=aolI_Rz0ZqY

here is what I learned:

I started from scratch and built up a new Neovim config.

Based on this video series:

• video by typecraft (thank you)

Pushed to github here: https://github.com/bartman/nvim-config

So today I learned about find -exec ... +

I put my camera SD/HC card into my laptop (running Debian/testing) and it didn’t mount. Usually I would just run the mount command to get it going:

    dmesg | tail
    mount /dev/mmcblk0p1 /mnt

That unfortunately has some annoyances and I decided to solve this finally.

I’m doing some IPv6 codig for a client and needed to setup a bunch of 6in4 tunnels.

Thre are many ways to do this through distribution init scripts (Debian, Fedora), but I wanted something less permanent and more dynamic for testing.

The procedure can be summarized in these steps:

• create a tunnel mytun between local 1.1.1.1 and remote 2.2.2.2

    ip tunnel add mytun mode sit local 1.1.1.1 \
                    remote 2.2.2.2 ttl 64 dev eth0
  

• give the local end an address

    ip addr add dev mytun f8c0::1.1.1.1/64
  

• bring up the tunnel

    ip link set dev mytun up
  

I have an LVM volume, with xfs on it, that is almost full:

$ df /scratch -h
Filesystem                Size  Used Avail Use% Mounted on
/dev/mapper/vg-scratch    180G  175G  5.4G  98% /scratch

$ sudo lvdisplay /dev/mapper/vg-scratch
...
  LV Size                180.00 GB
...

But I have some more space in the physical volume. Let’s grow the logical volume.

Just a quick note so I don’t forget now to enable console logging on systems running grub2 (like Ubuntu 10.04, Lucid).

• edit /etc/default/grub
  • set GRUB_CMDLINE_LINUX to "console=ttyS0,115200n8 console=tty0"
• run update-grub
• reboot

( more info can be found here )

I am working on a large kernel module which had just come out of a large (and fruitful) internal API refactoring exercise. I now want to go through and cull the unused functions.

It turns out, all that is needed is the readelf utility (part of binutils package).

If you’re trying to do post-mortem analysis on a crashed river, or trying to find kernel-level bottlenecks with oprofile, you need the decompressed kernel w/ debug symbols. This comes in a form of a vmlinux file. Some distributions ship debuginfo packages, namely RHEL. On Ubuntu this seems lacking.