Ubuntu on Random notes from mg

Booting ISO images, 2020 edition

marius@gedmin.as (Marius Gedminas) — Fri, 22 May 2020 18:57:21 +0300

Now that Ubuntu 20.04 LTS is out, it was time to update my bootable USB drive with a bunch of ISO images. Except I got tired of editing grub.cfg by hand. So I wrote a script.

Now all I have to do is plug in my USB drive, download a new ISO into it, run make in the /boot/grub subdirectory, and presto! It works.

What the script does:

finds all ISO files in ../../ubuntu/*.iso
extracts /boot/grub/grub.cfg from inside the ISO image (yes, I wrote an ISO 9660 filesystem parser in Python)
extracts the kernel command line
adds the iso-scan/filename=… argument to make it boot from an ISO image
generates a new grub.cfg with a menu listing all Ubuntu versions and all image variants (desktop vs server)

This works for all Ubuntu ISO images based on casper, which is a tool Ubuntu uses to produce their LiveCD images.

Notably it doesn’t work for older Ubuntu server images based on debian-installer.

I don’t know about other distributions and Ubuntu variants. Ubuntu variants ought to work, but my script might generate wrong titles for them.

New laptop: ThinkPad X390

marius@gedmin.as (Marius Gedminas) — Thu, 27 Jun 2019 01:09:48 +0300

The time has come to replace my trusty ThinkPad X220 with something more modern (that has enough horsepower to run Subnautica). I picked a ThinkPad X390 for the usual reasons (it’s a ThinkPad, it’s small and light and portable, the keyboard is reasonably sized, it has a trackpoint).

I installed Ubuntu 19.04 from a USB stick. This time I did not wipe Windows (because Subnautica), which gave me an unpleasant surprise: the Ubuntu installer does not allow you to enable full disk encryption if you’re using dual-boot.

Everything works great on Linux, except for the fingerprint scanner (a Synaptics device with ID 06cb:00bd in lsusb). Good news: Huang Vincent from Synaptics is working on a driver, so stay tuned!

The 13” 1920x1080 screen at ~160 dpi is a bit uncomfortable to use with my poor eyesight, so first I tried to use GNOME Tweaks to scale fonts to 120%. This worked okay-ish (a shame that Firefox ignores this and I had to tweak it separately) until I plugged in an external monitor (~80 dpi) where the large fonts were cartoonishly too large. Next, I enabled GNOME’s experimental fractional scaling support (I use a Ubuntu on Wayland session instead of the default one) and (after a reboot) set the zoom level on the internal screen to 125% (after resetting font scaling back to 100%, of course). Wayland apps look nice and crisp, X11 apps (Firefox) look fuzzy, but shrug at least I can read the text without squinting.

Surprise: I don’t hate the new keyboard! It took a while to get used to (and I caught myself using vim’s 0/$/^F/^B instead of Home/End/PageDown/PageUp way more than I usually do). I’m not sure why the backlit keyboard has two different light levels instead of just one. The mute LEDs on the F1 and F4 keys are too bright, especially at night, so I turn down the volume to the lowest level instead of muting. (The power LED is also a bit too bright – why did Lenovo make it white instead of the traditional nice green?)

I like the nice little physical camera switch that moves a little builtin plastic cover over the lens so I can be sure nobody’s spying on me :)

Things were less smooth on Windows 10: first it didn’t want to reboot (I waited like 20 minutes at the blue “Do not turn off your computer” screen before YOLOing the Power button for 10 seconds). Next, Bluetooth didn’t work (scanning for visible devices to add failed after a fraction of a second) until I updated the driver (Device Manager -> Bluetooth something something -> right-click -> Update driver) and/or installed Windows updates and rebooted. Other than that Windows 10 was a pleasant surprise – not something I could love, but at least something I don’t hate.

Subnautica runs … at 20 fps, using the lowest graphics settings. That’s integrated Intel graphics for you. (This was a deliberate choice: I want Linux compatibility more than I want to run games, so I avoid NVidia.)

Overall verdict: a nice little machine, would buy again.

Update: In August of 2020 GNOME Software offered a firmware update for the Synaptics Prometheus fingerprint scanner. The fingerprint scanner now works!

ImportError: No module named _md5

marius@gedmin.as (Marius Gedminas) — Wed, 07 Jul 2010 01:20:04 +0300

If you're using virtualenv, and after a system upgrade you get errors like

...
  File "...", line ...
    from hashlib import md5
  File "/usr/lib/python2.6/hashlib.py", line 63, in __get_builtin_constructor
     import _md5
ImportError: No module named _md5

this means that the copy of the python executable in your virtualenv/bin directory is outdated and you should update it:

$ cp /usr/bin/python2.6 /path/to/venv/bin/python

or, better yet, recreate the virtualenv.

Booting ISO images from a USB drive

marius@gedmin.as (Marius Gedminas) — Mon, 03 May 2010 20:24:31 +0300

Dear lazyweb, I would like to download an arbitrary ISO image (say, a Ubuntu 10.04 Desktop CD) into a directory of a USB flash drive, and then make that USB flash drive boot that ISO image. I do not want to

re-partition or re-format the flash drive (this eliminates usb-creator, AFAIU)
extract the contents of the ISO image into the root of the USB drive (this eliminates unetbootin)
skip the ISO's bootloader and directly boot the kernel+initramfs from the ISO (eliminates this recipe, and this recipe)

I just want a bootloader on the USB to read the VFAT filesystem, mount the ISO image as a loop device, then chain-load the bootloader from that ISO. Bonus points for having a menu letting me choose one of several ISO images. Running a script to edit a text file (say, grub's config) to get that menu is fine.

Is this even possible? If not, can I at least have two out of three (no partition/extraction, but skipping intrinsic bootloader is fine)?

Solution that I finally chose (from Ubuntu forums):

Plug in USB key. Find out the mount point (/media/disk) and the device name (/dev/sdx) of the USB key with

$ mount|grep /media

Install GRUB 2 into the USB key with

$ sudo grub-install --root-directory=/media/usbdisk /dev/sdx

If it says something about embedding being impossible and falling back to UNRELIABLE blocklist-based setup, run

$ sudo grub-install --root-directory=/media/usbdisk /dev/sdx --force

Download a CD image, let's say ubuntu-10.10-desktop-i386.iso. Put it into /media/usbdisk/ubuntu/.

Create a text file /media/usbdisk/boot/grub/grub.cfg with

menuentry "Ubuntu 10.10 (x86 desktop livecd)" {
    set isofile="/ubuntu/ubuntu-10.10-desktop-i386.iso"
    loopback loop $isofile
    linux (loop)/casper/vmlinuz boot=casper iso-scan/filename=$isofile quiet splash noprompt --
    initrd (loop)/casper/initrd.lz
}

You can have as many ISO images as you want, just make sure to add a menuentry for each. There's no need to run grub-install again after adding or removing a .iso file. Oh, and if you want to use an ISO file for a different distribution, you'll have to figure out the correct linux and initrd lines somehow.

Update: Ubuntu 13.04 changed the name of the kernel -- use (loop)/casper/vmlinuz.efi instead of (loop)/casper/vmlinuz.

I tested it with the following images

ubuntu-10.04-desktop-i386.iso
ubuntu-10.04-server-i386.iso
ubuntu-10.04-server-amd64.iso
ubuntu-10.10-desktop-i386.iso
ubuntu-11.04-desktop-i386.iso
ubuntu-11.10-desktop-i386.iso
ubuntu-12.04-desktop-i386.iso
ubuntu-12.04-desktop-amd64.iso
ubuntu-12.04-server-i386.iso
ubuntu-12.04-server-amd64.iso
ubuntu-12.10-desktop-i386.iso
ubuntu-12.10-desktop-amd64.iso
ubuntu-13.04-desktop-amd64.iso

This solution skips the CD-ROM's boot menu. I haven't found a better one.

Update: I wrote a script to generate a grub.cfg for me so I don't have to do it by hand any more.

df

marius@gedmin.as (Marius Gedminas) — Wed, 15 Jul 2009 01:33:02 +0300

Modern Linux system have all sorts of fake filesystems cluttering the output of df and mount: tmpfs, bind mounts, fuse for ~/.gvfs, etc. I have only one real partition on my laptop, yet mount returns 22 lines of output.

Question: are there any df-like utilities that filter out all the crap and show only interesting bits? The standard df as well as pydf both display 8 lines instead of 1. Discus is worse: it shows 20. GUI utilities like Baobab also suffer from this confusion, especially bind mounts.

Ironically, Ubuntu's update-motd gets confused by Ubuntu's private user directories and displays disk stats for ~/Private as if it were a real partition.

Buildbot issues on Ubuntu Hardy

marius@gedmin.as (Marius Gedminas) — Fri, 15 May 2009 15:33:26 +0300

Update: The story continues, but solution is not in sight yet.

I upgraded a buildbot slave to Ubuntu 8.04 (Hardy) recently and now I'm getting a strange intermittent failure: sometimes cp -r /local/dir /nfs/mounted/dir fails ("process killed by signal 1", i.e. SIGHUP).

I wonder if NFS is relevant or incidental to the issue?

Google finds an old thread from 2005, with a workaround (usepty=False), but I'd like to understand the problem before applying random fixes.

So far three different build steps doing cp -r have failed during 10 days. I've now changed them all to cp -rv, so I can at least see if the failure is in the middle of the copy or at the end, if it fails again.

Update: so far 4 build steps have failed on 6 separate occasions:

May  5 02:31: cp -r local-dir1 nfs-mounted-dir1  
May  6 02:31: cp -r local-dir1 nfs-mounted-dir1  
May  6 04:33: cp -r local-dir2 nfs-mounted-dir2  
May 15 02:00: cp -r local-dir3 nfs-mounted-dir3  
May 17 04:32: rm -rf nfs-mounted-dir4            
May 20 04:31: rm -rf nfs-mounted-dir4

I see no particular correlation between step duration and results, e.g. the rm -rf step usually takes between 2.2 and 4.6 seconds. The two SIGHUPs happened after 2.4 seconds.

They all make no output. When I changed the cp steps and added a -v, they stopped failing, but that could be just a coincidence.

We're having an email conversation with Jean-Paul Calderone ("exarkun") about the possibility of this being PTY-related, with no clear resolution so far.

And, hey, now this blog supports comments ;)

Filing bugs on GUI apps

marius@gedmin.as (Marius Gedminas) — Thu, 09 Apr 2009 22:29:21 +0300

Ubuntu people want users to use apport to report bugs. There's a command-line tool called 'ubuntu-bug' that you can use if you know the name of the package or at least the name of executable. There's a "Report a problem" menu item in many, but not all GUI apps.

Here's what you can do if the GUI app in question doesn't have that menu item, and you don't remember what it's called, and you're the same sort of a crazy command-line person that I am:

$ xprop|grep PID

Click on the app's window. Watch that shell command return a line that looks like

_NET_WM_PID(CARDINAL) = 807

Now run

$ ps 807  # substitute the real number

You'll see the command name, e.g.

  PID TTY      STAT   TIME COMMAND
  807 ?        S      0:02 /usr/lib/indicator-applet/indicator-applet --oaf-acti

Now you can run

$ ubuntu-bug /usr/lib/indicator-applet/indicator-applet

If you're not running Jaunty, you'll need to do one more step to find the name of the Ubuntu package:

$ dpkg -S /usr/lib/indicator-applet/indicator-applet
indicator-applet: /usr/lib/indicator-applet/indicator-applet

You can use that with apport-cli or on the Launchpad online bug reporting form.

For some (many?) programs you can short-circuit this trail by looking at the WM_CLASS property instead of the _NET_WM_PID property:

$ xprop|grep CLASS
WM_CLASS(STRING) = "indicator-applet", "Indicator-applet"

While there is no requirement for the window class name to match the name of the Ubuntu package or the name of the executable, it may give you a reasonable guess.

Upgrade to Ubuntu Jaunty

marius@gedmin.as (Marius Gedminas) — Mon, 06 Apr 2009 02:28:38 +0300

Ubuntu 9.04 is going to be released in around three weeks. As usual I couldn't wait (and saw that some bugs that were irritating me every day were fixed in Jaunty), so I upgraded to the current beta.

After a little hiccough at the beginning, the upgrade was the smoothest Ubuntu upgrade I've ever had: I spent those two and a half hours browsing the web, watching screencasts and chatting on IRC, while update-manager worked in the background. Firefox was mostly very responsive, only stuttering when update-manager got around to unpacking openoffice.org-common. There were few debconf or conffile questions (one from sysstat 2 minutes into the upgrade, then a conffile question after 1 hour and 20 minutes, then two more after 15 minutes, and one more 5 minutes later. And the last one 10 minutes later). There were no ugly theme changes or failing GNOME applet error messages during the upgrade. Near the end X-Chat automatically started showing new-style notifications (beautiful!) and Firefoxes nicely asked to be restarted with a fold-down notification bar.

Nice. Now, after a reboot things were not so nice: I couldn't login. After typing in my password and a couple of mode changes I was kicked back to the GDM prompt. I panicked and started logging into the text consoles and trying to run startx, quite in vain, since when I just tried gdm again it worked fine.

The intel video driver feels slower, as promised by the release notes, but it's acceptable as long as you don't try to rotate the external screen. Then it's horrible and unusable—a regression since Intrepid. I'll have to retry with UXA.

Compiz failed to enable a plugin (GNOME Compatibility), so a couple of key bindings didn't work (Alt+F1 to get the menu, Alt+F2 to get the run dialog, my custom keybinding to open a terminal) until someone on FreeNode told me what to enable.

The Flash plugin is now swfdec, and it is unable to cope with Youtube music videos—the sound is all choppy. I'm wondering if this is swfdec's fault, pulseaudio's fault (it's common knowledge that all audio problems stem from pulseaudio, right? ;) or X.org's fault (top shows it's X that's eating 90% CPU when swfdec is trying to play a video).

A lot of very irritating bugs are gone. I don't need to restart Compiz after playing with xrandr. X doesn't crash after I play with xrandr. Two-finger scrolling with the Synaptics touchpad doesn't produce phantom scroll-down-17-pages events when I take my fingers off the touchpad. The GNOME panels don't migrate to the external screen when I play with xrandr (but one of them jumps from the bottom to the top when I play some more). The new splash screen has a pretty gradient for its progress bar (but is displayed off-center, maybe because I added vga=872 to my GRUB kernel options list to avoid ugly stretching of text consoles). X.org no longer distorts the aspect ratio of 1024x768 when stretching it to fit the 1280x800 screen—now I get sensible black bars on the sides. The new notification bubbles are beautiful!. I could stare at them all day. (But the new indicator applet is ugly.)

Overall I'm happy. A bunch of very irritating bugs were replaced with a smaller bunch of somewhat less irritating bugs. The intel video slowdown scares me a bit, though, but the prettiness of the notification bubbles outweigh everything else. What can I say, I like pretty things—Ooh, shiny!

Submitting patches the Launchpad way

marius@gedmin.as (Marius Gedminas) — Mon, 06 Apr 2009 00:45:20 +0300

Today I happened to read about lazr.enum in a mailing list. I went to the PyPI page and saw raw ReStructuredText markup instead of a nicely formatted page. Now I know from prior experience that this happens when the package's description has an error in the markup. I thought I'd report a bug and provide a patch.

Leap of knowledge: since I know lazr.enum was created by the Launchpad.net team I could safely assume they were keeping the sources in Launchpad. Therefore I was pretty sure I could get them with

$ bzr branch lp:lazr.enum

so I ran that command and it worked.

Next I looked at setup.py to see how it produces the long_description field. It was reading the contents of a couple of text files, one of them being src/lazr/enum/README.txt. I looked at that and saw a .. toc-tree: directive that does not exists in plain docutils (it's a Sphinx extension).

I added up a couple of lines to setup.py to strip that out, tested it (with setup.py --long-description > test.rst; restview test.rst) committed to my local branch, and created a bug report in Launchpad. Then I was a bit lost, since I didn't know how to make my fix available. Attach a patch? Maybe, but I wanted to see if this distributed version control thing is good for anything else.

I thought that first I'd make that branch public, and then see if there was a way to link it to the bug report. I ran

$ bzr push lp:~mgedmin/lazr.enum/pypi-fix

which took a few seconds to create a new public branch on Launchpad with my fix in it (it would be nice if I didn't have to explicitly specify my Launchpad username and the project name—both of which bzr already knows—and just specify the name of the branch). Then I went back to my bug report and saw an option to link it to a branch. There was a search field in the popup that found my "~mgedmin/lazr.enum/pypi-fix" easily enough when I pasted it into the search box.

After clicking on the branch, I saw a "propose a merge" option. I did that and Launchpad sent an email to the developers asking them to merge my fix.

I made one mistake, I think: I should've created the bug report first, and then mentioned the bug number in my commit message (with bzr commit --fixes=NNN, although here I'm suddenly not sure if the bug number should be left bare, or prefixed with something like "lp" to indicate it was a Launchpad bug number?).

Other than that it was a pretty smooth experience. When will I be able to do that for Ubuntu packages?

Bug of the day

marius@gedmin.as (Marius Gedminas) — Fri, 03 Apr 2009 00:31:41 +0300

I think I'm going to blog about Ubuntu bugs I encounter during my day. Don't get me wrong—I love Ubuntu and haven't seen a better OS yet. But it has bugs.

Why blog and not report them to Launchpad? Many reasons: producing good bug reports is hard work; bugs rarely show up alone, and if I start filing one, I'll forget details about the others; sometimes it's unclear whether something is or is not a bug until you've written it down and looked at it; Launchpad is slow while previewing a blog post on the local machine is fast.

This is one of those long and rambling posts that you can skip without feeling guilty. ;-)

Things I did today:

I enabled my dual-head screen and had to restart compiz to avoid bug 317431.
After work I tried to switch to single-screen mode again and X promptly crashed—bug 298226.
When X restarted it picked a fugly stretched 1024x768 mode that looks very bad on my 1280x800 widescreen panel—because this was the only common mode between the internal panel and the external 1280x1024 LCD.
I used Ctrl+Alt+Backspace after pulling the VGA cable as a quick way to restart GDM with a saner resolution. I wonder what I'll do in Jaunty where Ctrl+Alt+Backspace is disabled by default. (BTW I agree with that decision, having accidentally pressed Ctrl+Alt+Backspace on several painful occasions.)
I suspended my laptop and put it in my backpack, then went home.
At home I tried to resume. The laptop suspended again immediately after waking up.
When I woke it up again, it did not accept my fingerprint (this often happens, I don't particularly mind—it's difficult to swipe the finger precisely).
I then typed my password, pressed Enter—and then gnome-screensaver spent an unreasonably long time verifying my password (this I do mind).
While it was verifying the password suddenly the GNOME panel showed up on top of the gnome-screensaver, with two new notification bubbles: one was the usual post-upgrade reboot required, and the other was a typical post-upgrade informational notice (one about refreshing ALSA configuration presets—I haven't seen it before). The interesting thing is that I don't remember running apt-get upgrade today, and /var/log/dpkg.log confirms I last upgraded things yesterday at noon.
I could not take a screenshot of the bizarreness described in step 3, but that's reasonable—the desktop was still locked. So I took a (crappy) picture with my cell phone (and now I'm too lazy to upload it—but I have proof! proof!).
After 30 seconds or so gnome-screensaver finally timed out, the GNOME panel disappeared, and I got a new fresh screensaver unlock dialog. This time it worked.
Just after unlocking the screen I saw a GNOME keyring prompting for my keyring password to enable Network Manager to get the WPA passphrase needed to log in.
It's the same as my login password. Shouldn't some PAM magic I recently read about in Ubuntu mailing lists unlock that keyring for me automatically on login? Oh, right, I log in using my fingerprint, not my password. Darn.

Now I feel like I should report item #3 (and #13) as a wishlist, item #6 as a bug (there's probably one open, with a bunch of duplicates, already), #8 as a bug (but that would be useless—I've no recipe for replicating that), #9 as a bug (again, how can I replicate?)—or, rather two bugs (why did the informational notices got delayed by a day?). Maybe three bugs: the untitled (four bugs!) window with the update information says

Refresh Advanced Linux Sound Architecture (ALSA) configuration presets

New Advanced Linux Sound Architecture (ALSA) configuration presets have been added. Please execute the asoundconf(1) set-default-card macro in a Terminal now to refresh your user's configuration presets. You may accomplish this task by executing the following command in a Terminal: asoundconf set-default-card

which reads like gibberish to any nontechnical person (and many technical persons too, I think).

Let's talk more about this notification: the first thought that comes into the mind of a technical person is: if you know what command I should run in a terminal, why don't you run it yourself, you stupid machine? The answer is “because that's not the complete command that you need to run”:

$ asoundconf set-default-card
You have omitted a necessary parameter.  Please see the
output from `asoundconf list`, and use one of those sound
card(s) as the parameter.

Okay, I'll bite

$ asoundconf list
Names of available sound cards:
Intel

There's only one card. Why isn't it selected as default automatically, without forcing me to jump through hoops in a terminal? Think of the children ~~(or, better, your mother)~~!

But now I remember that Pulseaudio is supposed to be the default ALSA card. What will happen if I select Intel here? Will it break Pulseaudio? Oh dear, I've no idea. I'm afraid to do anything now.

I think I will put on my regular-user-glasses and look at that dialog again:

Technobabble technobabble

gibberish gibberish gibberish

Okay, I hit the only button available—close. I'm done here. Sound appears to continue to work fine without me doing anything. Why was that dialog necessary? Just get rid of it.

I'm sorry. I did not intend this to become a rant. I wanted to make a list of reportable bugs that I could review later and file properly (or have helpful people look up and send me via email—this has happened before, and I was surprised and grateful).

X ate my keyboard, again

marius@gedmin.as (Marius Gedminas) — Fri, 06 Mar 2009 08:11:36 +0100

Last night I thought it would be fun to try to suspend my laptop with the 3G USB modem still plugged in and connected. This morning when I woke it up I had no keyboard.

Ctrl+Alt+F1 worked to switch to a text console (when I pressed it twice). The text console wasn't garbled, which was nice. killall gnome-settings-daemon or gnome-screensaver didn't fix anything. Suspending and resuming again didn't fix things either (not that there's any reason why it should). When I noticed that Ctrl+Alt+Backspace didn't work in X, I realized this wasn't an X grab issue.

/var/log/Xorg.0.log has a series of errors like this:

(EE) AT Translated Set 2 keyboard: Device has changed - disabling.
(WW) AT Translated Set 2 keyboard: Release failed (Invalid argument)

Sadly, there are no timestamps. I think this was printed once on resume, and then once on every VT switch.

xinput still lists "AT Translated Set 2 Keyboard" among its devices.

I remember now -- before suspending I had used sudo showkeys -s in an xterm (trying to see the scan code of the ThinkVantage button). Could've mucked up the keyboard mode?

Alt+SysRq+R (raw mode) didn't help, even when followed by more VT switches. I think I'm gonna restart the X server. Or reboot the whole PC, Ubuntu notifier is showing its reboot icon again...

Meta: I wrote this post on March 6, and then decided it wasn't interesting enough to post. And then posted it accidentally on March 20.

Update: looks like bug 327175.

X.org and stuck Ctrl

marius@gedmin.as (Marius Gedminas) — Fri, 27 Feb 2009 00:07:43 +0100

Warning: this is going to be a long story with a semi-happy ending. Skip it unless you enjoy tales of woe and debugging.

So, I open up my laptop, plug in a USB keyboard and mouse, start up Inkscape and start fooling around. Suddenly I notice strange behaviour:

I cannot select two objects in Inkscape by holding down ctrl or shift and clicking -- only the last object I clicked on becomes selected.
Menus don't pull down, although the highlighted bar follows my mouse
I cannot type any text into text boxes
Moving the mouse into a screen corner doesn't trigger the Expose-like effect
I cannot drag windows around by grabbing the title-bar
I cannot drag windows around by alt+dragging
Dragging the mouse inside an xterm creates a vertical selection

That last hint seems to indicate that the Ctrl key could be stuck. I try pressing it and releasing, then the other one, then both Ctrl keys on the USB keyboard. Surely, if X sees a press and a release event for each Ctrl it will realize none of them are still down? No such luck. I try to unplug the USB keyboard and mouse next. No results.

This is not the first time this has happened to me. Previously I found no exit out of this state other than killing X.org with (great pleasure and) Ctrl+Alt+Backspace. Surely there must be a better way?

I ssh in and start xev. MotionNotify events have state 0x4, which is control, I think. By the way, I only see mouse events in xev, keyboard events don't make it. The keyboard itself is alive at some level, as Caps Lock turns on its LED, and Ctrl+Alt+F1 gives me a (garbled and unusable) text console. Holding down Alt or Shift doesn't change the state of events seen by xev, though.

Did my keyboard map get lost? Is some X client grabbing all the keys? how do I recover?

I use x2x to connect to the laptop from my desktop. I now can use my desktop's mouse and keyboard to control the laptop. x2x works by injecting X events via XTest. I see mouse events x2x injects, but xev again shows nothing on the keyboard front.

I try to guess which X client might have the keyboard grab (if there is one). I killall compiz. I'm surprised that gnome-session doesn't restart it (or spawn a different window manager in its place). I start metacity manually. The problem is not gone. I kill metacity and start Compiz again.

I have a VNC server running (vino), but I've forgotten its password.

I notice a weird message in my xev log:

KeymapNotify event, serial 40, synthetic NO, window 0x0,
    keys:  4294967195 0   0   0   32  0   0   0   0   0   0   0   0   0   0   0   
           0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0

Huh? 4294967195 keys? That looks like an unsigned 32-bit int underflow. I scroll back and find the first KeymapNotify event seen by xev:

KeymapNotify event, serial 25, synthetic NO, window 0x0,
    keys:  0   0   0   0   32  0   0   0   0   0   0   0   0   0   0   0   
           0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0

Looks normal. Then I notice this at the end of the xev log:

FocusOut event, serial 40, synthetic NO, window 0x3e00001,
    mode NotifyWhileGrabbed, detail NotifyNonlinear

What's "NotifyWhileGrabbed" mean? How do I find the rogue app and kill it? xrestop shows me 35 X clients, do I just kill them one-by-one until the problem disappears? Some of those clients are <unknown> and show no PID.

I suspend the laptop (which is a very stupid idea when your other machine's mouse and keyboard are redirected to it via x2x) and resume it, hoping that gnome-screensaver will somehow overpower the existing application's lock with its own. Gnome-screensaver is nowhere in sight. At least my x2x connection becomes alive again and I have my keyboard & mouse back on the desktop.

I notice that I can no longer make any kinds of selections in gnome-terminal. Why? Window focus no longer follows mouse. xev no longer sees mouse motion events. It sees a couple of MappingNotify events when I plug in a (different) USB keyboard, though.

I killall gnome-screensaver (which was invisible, remember?) and can now again see motion events in xev and select windows with the mouse.

I start randomly killing applications. gnome-power-manager. nautilus. inkscape. firefox. gtk-window-decorator. gnome-panel. notification-daemon. gnome-terminal. pulseaudio (no reason). vino-server. update-notifier. seahorse-agent. gnome-keyring-daemon. bluetooth-applet. fast-user-switch-applet. multiload-applet-2. mixer_applet2. system-config-printer. gnome-settings-daemon.

And the system becomes ugly (no GNOME theme) but alive. I get a second xev window from an earlier attempt to type 'xev' in a terminal that was unable to receive key events.

Of course, since I killed all the actual applications and half of the necessary support programs my session is now useless, so I'll have to log out and log back in again. But at least in the future I'll know: when something like this goes wrong, killall gnome-settings-daemon.

Now I'd like to report a bug (the thought of hurling a brick through the responsible developer's window never crossed my mind, honest!), but without a reliable way of reproducing the problem will it be of any use?

I restart gnome-settings-daemon and it promptly invokes xrandr to set up a dual-head mode that confuses compiz. By "confuses" I mean displays a rotating cube in the top-left 1280x700 area of my 2560x1024 extended desktop, filling the rest with whatever was in the video memory last time I had a dual-head mode.

I try to start up Firefox on my desktop and put a link to the relevant bug, but Firefox quietly refuses to start up. Well, 'Segmentation fault' at the end of ~/.xsession-errors is quiet, isn't it? Thankfully, 'firefox http://someurl' for some reason works and opens a window. I cannot find the Compiz bug I remember (could it have been #135418?), but this looks like a better fit anyway: #317431 (and #206998 might be a duplicate). And here's my gnome-settings-daemon bug: #335201.

This is all on Ubuntu 8.10 (Intrepid Ibex).

Some days I just hate Linux. Then I remember that it's worse on other systems...

Largest apps on my laptop #2

marius@gedmin.as (Marius Gedminas) — Sun, 22 Feb 2009 12:28:41 +0200

I recently posted some data about applications taking up the most RAM on my laptop. That was after 9 days of uptime, while this is after 12 hours:

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
16033 root      20   0  527m 109m  11m S    3  5.5   6:52.82 Xorg
17834 mg        20   0  244m 105m  24m S    6  5.3   3:26.49 firefox
26425 mg        20   0 96872  58m  12m S    0  2.9   0:01.71 evince
16747 mg        20   0 90704  54m  19m S    0  2.8   0:10.70 tomboy
27169 mg        20   0  166m  53m  23m S    0  2.7   0:07.05 banshee-1
27167 mg        20   0 77392  31m  17m S    0  1.6   0:01.16 pidgin
16706 mg        20   0 86508  27m  18m S    0  1.4   0:35.35 gnome-panel
16708 mg        20   0 75196  20m  14m S    0  1.0   0:02.80 nautilus
20092 mg        20   0 61880  19m  11m S    1  1.0   0:07.08 gnome-terminal
16614 mg        20   0 58456  15m 9836 S    0  0.8   0:09.82 gnome-settings-

I don't have GNOME Do any more, and I've only one of the two PDFs open in Evince. I don't see multiload-applet on the first page of top output, which seems to indicate a slow leak. Evince has the same two documents. That concept doesn't quite apply to Banshee or Pidgin, but Pidgin's numbers are quite striking anyway (from 70 megs VIRT to 1.6 gigs VIRT in 9 days; thankfully RES only grows 2x during that time).

OS: Ubuntu 8.10, up-to-date with all the updates from -security, -updates, -proposed-updates and -backports.

Incidentally, I have 12 hours of uptime because my battery died while the laptops was suspended during my flight back home (either that, or it work up in the backpack, which is a scary thought). Apparently Ubuntu tried to hibernate when the battery was very low, which was a nice gesture. This didn't work out so well when resuming, since the kernels didn't match -- I had installed a kernel update, but hadn't rebooted. I don't think I ever used hibernation successfully in Linux.

Largest apps on my laptop

marius@gedmin.as (Marius Gedminas) — Wed, 18 Feb 2009 21:31:39 +0100

After reading Alexander Larsson's post on de-bloating nautilus I thought it would be interesting/useful to see what apps are eating my RAM, as a statistical data point if nothing else.

OS: Ubuntu 8.10. Uptime: 9 days, 20:45 (desktop session also started 9 days ago; I suspend and never log out). Top ten apps, according to top:

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
29594 mg        20   0  472m 269m  32m S    3 13.6  23:14.38 firefox
 7383 root      20   0  591m 182m  14m S    4  9.2 557:37.73 Xorg
 5199 mg        20   0  114m  67m  13m S    0  3.4   0:35.71 evince
17256 mg        20   0  230m  64m  24m S    0  3.3   4:10.93 banshee-1
29306 mg        20   0 1652m  62m  19m S    0  3.1   7:38.97 pidgin
 8060 mg        20   0  105m  59m  17m S    0  3.0   1:23.20 tomboy
 8015 mg        20   0  137m  42m  20m S    0  2.2  41:55.95 gnome-panel
 8017 mg        20   0  138m  41m  16m S    0  2.1   2:26.75 nautilus
 8063 mg        20   0 45876  30m 8788 S    0  1.6  33:22.14 multiload-apple
 8087 mg        20   0 80708  28m  15m S    0  1.4   4:45.49 gnome-do

Update: compare these numbers with what I get just 12 hours uptime.

Firefox started leaking memory quite rapidly lately, possibly after I upgraded to 3.0.6+nobinonly-0ubuntu0.8.10.1 exactly a week ago. I have to restart it once a day if I don't want my 2 gigs of RAM to fill up completely. I hadn't needed to do that before, memory usage stayed pretty constant.

I cannot explain the X.org numbers. pmap doesn't show me RSS numbers, but 150 megs of VIRT are attributed to the heap (an anonymous read-write mapping), while 256 megs look like the frame buffer ("resource2"). xrestop sees a total of 21027K of resources (pixmaps etc.) attributed to all the clients. I have a vague suspicion that this number doesn't include OpenGL textures used by Compiz, but I'm pretty clueless about those things. compiz --replace reduces Xorg's RSS down by 9 megabytes and increases VIRT by 5 megabytes. The increase is mirrored by xrestop, which now shows 25652K in total.

I have two PDFs open in the background since I intend to read them within the next couple of days, this explains the evince data.

I'm not happy about the Banshee memory usage. I wouldn't mind that much if it didn't insist on minimizing to the system tray.

I'm even less happy about Pidgin. Banshee at least has the excuse that it is built on top of Mono, which adds a whole new runtime & virtual machine. And why on Earth does a chat program need 1.6 gigs of virtual memory?

Tomboy: Mono again. But Tomboy is a killer application and I need it.

GNOME Panel: self-explanatory. The multitude of applets that I think that I need fill up both panels and slow down login times as well.

Nautilus: pretty consistent with Alexander's numbers, looks like I can expect improvements in Ubuntu 9.04 or at least 9.10.

Multiload applet: looks like I shouldn't have blamed GNOME Panel's memory usage on applets. 30 megs RSS for four ticking graphs seems a bit biggish, but maybe memory fragmentation is at fault.

GNOME do: people keep blogging about its coolness, then I install it, try it out twice, and forget it. The default GNOME Run dialog does what I need/am used to better.

Huawei E220 USB 3G modem

marius@gedmin.as (Marius Gedminas) — Thu, 05 Feb 2009 10:05:48 +0100

Almost works out of the box on Ubuntu. Will work out of the box in the forthcoming 9.04 release.

One curious little detail: according to the manual, a blinking green light means it's trying to find the GSM network (if it's blinking twice every 2.7 seconds) or that it's successfully found a GSM network (if it's blinking twice every 2.9 seconds). I'd like to have been on the meeting when this was decided. "I know! Let's make it blink 0.2 seconds faster to indicate it hasn't found a network yet! Brilliant!"

Update: given its shape and position next to my right-hand USB ports, it should double as a USB mouse.

On an unrelated note, Sweden is a very nice country.