Laptop arrived Sep 18, 2007. Took a good week from ordering to arrival, not bad, since IBM said 2 weeks. Got it with 2.0GHz core2duo, 160GB SATA, 2GB memory, 15.4" WSXGA+ (1680x1050) w/ NVIDIA Quadro FX 570M (256MB), Intel PRO/Wireless 3945ABG, 9 cell battery. Model number is 6459-CTO. It is confusing though this is found under the T61_14_Wide models on Lenovo's website, whereas this is a 15" laptop.

Fedora Core 7

First impressions:

   - first put the SATA (via BIOS) in compatibility mode .
     Disk is /dev/sda, DVD is /dev/scd0, a dual layer writer
     which appears to work fine.

   - this NVIDIA card is not detected in FC7 (FC9 ok, see below), 
     so had to do the slightly more tedious non-graphical install.
     After that was done, manually installed the NVIDIA driver, which then
     worked out of the box (with a small caveat, see below)

   - sounds doesn't work (known issue, can be solved by doing the ALSA
     drivers yourself
	** (17-feb-2008) finally got this to work with kernel  2.6.23.15-80.fc7 
  	   and the 1.0.16 also drivers from source (the kernel alsa didn't seem to
	   work for me)

   - wireless sometimes works, not sure when and when not. Sometimes 10b/sec,
     and i've seen 2MB/sec....
     I've tried the latest 3945 drivers from scratch. seems to be same.
     There is some suggestion that fixing the channel works better.
     New kernel (2.6.22.9-91.fc7) gave a very well working wireles after the
     first boot that then crawled to a halt and went back to the old mode.

	installed ipw3945 via FreshRPMS. Now loads this by default. Speed
	is only about 350 KB/s, but works (as eth1). Also notable is that
    	the frontpanel wireless light is now green, which iwlwifi wasn't doing.

	** with newer kernels it pretty much works out of the box, though 
	occasionally a reboot would be needed to reset the drivers as they
	sometimes get confused - ipw3945 is what i'm using with wlan0 

   - suspend to memory worked out of the box.

   - hibernation ("suspending to disk", uses swsusp) doesn't seem to work
     out of the box
	** not yet tested with the new kernel and nvidia driver ***
	** for a comparable model X61 and Ubuntu8 this worked out of the box **

   - somewhat SERIOUS: screen updates in certain programs are very slow.
     e.g. pgplot circles, OpenOffice programs, emacs scrolling

	SOLVED:  upgraded to NVIDIA-Linux-x86-100.14.19-pkg1.run, which
		 has explicit support for the FX 570M card.

   - oddities:
	suspend seems to work mostly, but sometimes it takes several
	minutes (sometimes aided by some Ctr-Alt-Fn  switching?); 
	patience seems to be the only way...

   - some applications (notably firefox, thunderbird, pidgin) crash and burn
     at random times. Not sure if that's me, or a relic of my . files

   - at random times - gimp seems to trigger it more than others - the whole
     desktop goes nuts as if doing a full refresh. Takes 10 secs or so to get
     back to live. This is VERY annoying. Also worried if that's not just
     me and my . files.

   - this is the pits: a few application (secondlife, ifrit) actually
     crash X. Nothing can be worse than that, short of locking up the
     laptop.

Ubuntu 7.10 (gutsy)

   - booted in safe mode, and came up utilizing full (wide) screen: nice
   - partition program scrashed.... could not continue. it filed a report
	seems to be that my /home is a symlink to a separare /home
	partition, and the "Migration assistant" couldn't quite
	figure it out. Solution: do /home later after the install.



Ubuntu 8.04.1

Came out 8.04 April 2008, but didn't try 8.04.1 (July version) until September.

After the install there was a failure from grub, luckily I had backed up /boot otherwise hard to get back to the other boots. Basically it was writing files in /boot, but no grub.conf. A manual fix (after putting back the old backup) helped. Next issue was that the gnome setup did not work well with the gnome files I had from the fedora's (between Fedora 7 and 9 also issues). However, a new user account worked fine. wow, gnome changing too fast.

Fedora 9

Came out May 13, 2008. No more Core in the name. Luckily the Nvidia is now detected in the boot. No more text install and manually installing the Nvidia driver. It seem suspend doesn't quite work anymore, so I'm still sticking to FC7 and waiting for a better kernel. Maybe should try the native Nvidia driver [though usually it is the other way around].

As of mid september 2008 I'm now using Fedora 9. Things seem to work with kernel 2.6.25.14-108.fc9.i686, though a newer one (2.6.26.3-29.fc9.i686) is available but no nvidia driver. Suspend seems to work, the recovery of wireless works most of the time.


cute mystery:

# df
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/sda5              9920624   7348976   2059580  79% /
tmpfs                  1020732         0   1020732   0% /dev/shm
/dev/sda7            201704368 188000916   3292132  99% /home
/dev/sda1              1019208     67700    898900   8% /a1
/dev/sda2              9920624   7979356   1429200  85% /a2
/dev/sda3              9920624   7348976   2059580  79% /a3

Looks like sda5 and sda3 are the SAME!!!!???

# mount
/dev/sda5 on / type ext3 (rw)					<----
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
tmpfs on /dev/shm type tmpfs (rw)
/dev/sda7 on /home type ext3 (rw)
none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)
sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw)
fusectl on /sys/fs/fuse/connections type fusectl (rw)
/dev/sda1 on /a1 type ext3 (rw)
/dev/sda2 on /a2 type ext3 (rw)
/dev/sda3 on /a3 type ext3 (rw)					<---


]# fdisk -l /dev/sda

Disk /dev/sda: 250.0 GB, 250059350016 bytes
255 heads, 63 sectors/track, 30401 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1         131     1052226   83  Linux
/dev/sda2             132        1406    10241437+  83  Linux
/dev/sda3            1407        2681    10241437+  83  Linux			<----
/dev/sda4            2682       30401   222660900    5  Extended
/dev/sda5            2682        3956    10241406   83  Linux			<----
/dev/sda6            3957        4478     4192933+  82  Linux swap / Solaris
/dev/sda7            4479       30401   208226466   83  Linux

did I get a RAID1 installed by default???

# more /etc/fstab
LABEL=/                 /                       ext3    defaults        1 1
tmpfs                   /dev/shm                tmpfs   defaults        0 0
devpts                  /dev/pts                devpts  gid=5,mode=620  0 0
sysfs                   /sys                    sysfs   defaults        0 0
proc                    /proc                   proc    defaults        0 0
LABEL=SWAP-sda6         swap                    swap    defaults        0 0
LABEL=/home             /home                   ext3    defaults        1 2
#
/dev/sda1               /a1                     ext3    noauto        0 0
/dev/sda2               /a2                     ext3    noauto        0 0
/dev/sda3               /a3                     ext3    noauto        0 0


In the end i figured out this was the result of installing FC7 and FC8
and /boot.  Manually fixed.

Fedora 10

Installed March 15, came out a little while earlier. Cause was that my ethernet seemed to have died, and it would only work on wireless. Popped in the old 160GB drive, and booted from the recovery & rescue partition using grub (since full windows was wiped long time ago). Miraculously all tests passed ok, the card was working fine. Booting back into linux then showed it still worked. Very odd.

After selecting the usual additional development and web server options, 1521 packages were installed.


- suspend didn't seem to work out of the box, so back to fedora 9 :-(

Fedora 11

Installed June 10, 2009. Hibernate/Suspend didn't initially work, but that's perhaps because no nvidia driver yet, since on F9 this works fine. This was indeed confirmed later. So, as of Aug 1, 2009, I'm running F11. After installing the Nvidia via RPMFusion (the merger of Dribble, Freshrpms, and Livna). Sadly, after the next kernel update and nvidia update, that driver is now misbehaving, and back into the originally not-suspending "nouveau" nvidia driver, this one now un-suspends fine.
Ah, after undoing the commented out baseurl='s in the rpmfusion-nonfree repos for yum, updating the kernel etc. I got the nvidia's to work again, seemingly both in 586 as well as in 686.PAE. Suspend seemst to work ok, so this is all quite a bit of a mess. Sound now also appears to be more reliable. This is Sep 7, 2009, kernel 2.6.30.5-43.fc11.i686.PAE and hopefully newer ones.

Fedora 13

Installed May 30, 2010. Still using ext4 for /, but ext3 for /home. The experience with running Ubuntu 9.10 and using the same /home/teuben was not a good experience. Whatever GNOME was doing from the Fedora side, Ubuntu would not display essential menus and render it useless. In Fedora 13 you select to be a desktop, development, web or minimal user. I choose development, but then wind up with no latex, openoffice or music for example. So some more work is needed. The usual additions via RPM Fusion to get mp3, video codecs, nvidia, flash, acroread, skype, picasa (just to name a few) is still to be done. This time I promised myself to keep a good log how I did this.

Issues

Ethernet Card

Twice now my ethernet "died", it would not start up anymore. It was clearly hardware related. You can see it because the green+orange lights would be steady, instead of the normal blinking mode (actually, green is steady, and orange would be blinking). As the system booted, it would blink, but somewhere during boot become steady. This happened both in Fedora and Ubuntu. The solution was erratic, not sure how to reproduce it, but a normal cold boot would not fix it, but cold from a battery removed state, maybe go in the BIOS, flip some options and save the BIOS and reboot. That seems to have fixed it. I call it the cosmic ray effect. Seen that in a Dell as well. No clue what's going on here.

Battery

The original IBM (9 cell) battery died mysteriously in the middle of using it. But as you can see from the plot(s) below, it had a pretty more-or-less linear steady loss of capacity, which a surprising revival (of one cell?) in mid 2008 (red dot in plot below). The new (6 cell) battery however keeps a very steady constant capacity, but then looses (a cell?) after about 2-3 months? (dark blue dots below)

Links


Reality check: how does my battery loose power:

or a new matplotlib style plot since the new battery (blue dots), since the old one (red dots) catastrophically died just like that, while sitting in Atlanta Bread:

This page was last modified on 30-Jun-2011 by teuben@astro.umd.edu.