Slow performance with Lucid on laptop/battery..

[azredwing]

I'm trying out Lucid (x86_64). I've got a few issues that I'm not sure are due to the beta or due to my laptop.

Lucid boots really quickly, but after getting to the main desktop, everything is just... slow and unresponsive. Opening any program (like, for instance, a terminal window) takes forever and many times the window greys out as if it's locked. (Running sudo apt-get update && sudo apt-get upgrade locks the terminal window up and I get asked if I want to force-close it.) Clicking shut-down (after clicking the menu button and waiting for the menu to appear) seems to not do anything until about 20 seconds later, and by then 4 of them pop up because I keep trying to click shut-down.

I can't really explain in further detail what is going on, other than it all seems very sluggish and unresponsive. Oddly, this behavior only seems to exist when the laptop isn't plugged in. I'm not sure what to make of this.

I'm running a Thinkpad T61 - I also just installed a 64GB SSD HDD (Patriot PS-100). I'm unsure if the new HDD is the culprit or this performance issue is in Lucid itself.

EDIT: Solution!

It turns out that there's an issue with libata and the SSD (dmesg logs are in the thread). Changing my BIOS SATA controller setting from "AHCI" to "Compatibility" resolved the issue. This did not occur with my SSD in Karmic, i.e., I ran the SSD fine with AHCI mode. Go figure.

[azredwing]

I can now definitively say that this problem exists only when the laptop is not plugged in upon boot. If I plug it in after it boots, the sluggishness remains. Unplugging the laptop results in the same slow performance as before. So as it stands, I have to keep the laptop plugged in in order for the OS to be responsive.

[hugmenot]

Are you using any naive power saving methods?

In particular make sure that you aren’t using the ›powersave‹ cpu scaling governor. You should be using ›ondemand‹

What does this show:
cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor

[azredwing]

Are you using any naive power saving methods?

In particular make sure that you aren’t using the ›powersave‹ cpu scaling governor. You should be using ›ondemand‹

What does this show:
cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor

I'm running a dual-core processor. Running the above on both CPUs returns ondemand and not powersave.

[azredwing]

sorry for the bump - but I feel like this could be a big deal if it's not just me but other Lucid laptop users. Does nobody else have this issue?

[miegiel]

sorry for the bump - but I feel like this could be a big deal if it's not just me but other Lucid laptop users. Does nobody else have this issue?

No problems here (32bit lucid 2.6.32-18-generic-pae). Not exactly the same hardware, but it's close.
user@machine:~$ sudo lshw -short
H/W path Device Class Description
================================================== =======
system Aspire 3810T
/0 bus Aspire 3810T
/0/0 memory 1MiB BIOS
/0/f memory 4GiB System Memory
/0/f/0 memory 2GiB SODIMM Synchronous 800 MHz (1.2 ns)
/0/f/1 memory 2GiB SODIMM Synchronous 800 MHz (1.2 ns)
/0/17 processor Intel(R) Core(TM)2 Duo CPU U9400 @ 1.40GHz
/0/17/18 memory 3MiB L2 cache
/0/17/1a memory 32KiB L1 cache
/0/17/1.1 processor Logical CPU
/0/17/1.2 processor Logical CPU
/0/19 memory 32KiB L1 cache
/0/100 bridge Mobile 4 Series Chipset Memory Controller Hub
/0/100/2 display Mobile 4 Series Chipset Integrated Graphics Controller
/0/100/2.1 display Mobile 4 Series Chipset Integrated Graphics Controller
/0/100/1a bus 82801I (ICH9 Family) USB UHCI Controller #4
/0/100/1a.1 bus 82801I (ICH9 Family) USB UHCI Controller #5
/0/100/1a.7 bus 82801I (ICH9 Family) USB2 EHCI Controller #2
/0/100/1b multimedia 82801I (ICH9 Family) HD Audio Controller
/0/100/1c bridge 82801I (ICH9 Family) PCI Express Port 1
/0/100/1c/0 wlan0 network Wireless WiFi Link 5100
/0/100/1c.2 bridge 82801I (ICH9 Family) PCI Express Port 3
/0/100/1c.2/0 eth0 network AR8131 Gigabit Ethernet
/0/100/1d bus 82801I (ICH9 Family) USB UHCI Controller #1
/0/100/1d.1 bus 82801I (ICH9 Family) USB UHCI Controller #2
/0/100/1d.2 bus 82801I (ICH9 Family) USB UHCI Controller #3
/0/100/1d.3 bus 82801I (ICH9 Family) USB UHCI Controller #6
/0/100/1d.7 bus 82801I (ICH9 Family) USB2 EHCI Controller #1
/0/100/1e bridge 82801 Mobile PCI Bridge
/0/100/1f bridge ICH9M-E LPC Interface Controller
/0/100/1f.2 scsi1 storage ICH9M/M-E SATA AHCI Controller
/0/100/1f.2/0.0.0 /dev/sda disk 320GB WDC WD3200BJKT-0
/0/100/1f.2/0.0.0/1 /dev/sda1 volume 64GiB Windows NTFS volume
/0/100/1f.2/0.0.0/2 /dev/sda2 volume 234GiB Extended partition
/0/100/1f.2/0.0.0/2/5 /dev/sda5 volume 4094MiB Linux swap / Solaris partition
/0/100/1f.2/0.0.0/2/6 /dev/sda6 volume 12GiB Linux filesystem partition
/0/100/1f.2/0.0.0/2/7 /dev/sda7 volume 12GiB Linux filesystem partition
/0/100/1f.2/0.0.0/2/8 /dev/sda8 volume 206GiB Linux filesystem partition
/0/100/1f.3 bus 82801I (ICH9 Family) SMBus Controller
/0/100/1f.6 generic 82801I (ICH9 Family) Thermal Subsystem

[azredwing]

Thanks for the input, miegiel. I see you're not using an SSD - perhaps this narrows down my issue to an SSD. I'm weary of replacing my good Karmic install on my non-SSD drive for sake of testing this, however. I'd like to get others with SSD laptops to corroborate what I suspect.


$ sudo lshw -short
H/W path Device Class Description
================================================== =====
system 7658CTO
/0 bus 7658CTO
/0/0 memory 128KiB BIOS
/0/6 processor Intel(R) Core(TM)2 Duo CPU T7700
/0/6/a memory 64KiB L1 cache
/0/6/c memory 4MiB L2 cache
/0/b memory 64KiB L1 cache
/0/2b memory 4GiB System Memory
/0/2b/0 memory 2GiB SODIMM DDR2 Synchronous 667 MHz
/0/2b/1 memory 2GiB SODIMM DDR2 Synchronous 667 MHz
/0/100 bridge Mobile PM965/GM965/GL960 Memory Cont
/0/100/2 display Mobile GM965/GL960 Integrated Graphi
/0/100/2.1 display Mobile GM965/GL960 Integrated Graphi
/0/100/19 eth0 network 82566MM Gigabit Network Connection
/0/100/1a bus 82801H (ICH8 Family) USB UHCI Contro
/0/100/1a.1 bus 82801H (ICH8 Family) USB UHCI Contro
/0/100/1a.7 bus 82801H (ICH8 Family) USB2 EHCI Contr
/0/100/1b multimedia 82801H (ICH8 Family) HD Audio Contro
/0/100/1c bridge 82801H (ICH8 Family) PCI Express Por
/0/100/1c.1 bridge 82801H (ICH8 Family) PCI Express Por
/0/100/1c.1/0 wlan0 network PRO/Wireless 3945ABG [Golan] Network
/0/100/1c.2 bridge 82801H (ICH8 Family) PCI Express Por
/0/100/1c.3 bridge 82801H (ICH8 Family) PCI Express Por
/0/100/1c.4 bridge 82801H (ICH8 Family) PCI Express Por
/0/100/1d bus 82801H (ICH8 Family) USB UHCI Contro
/0/100/1d.1 bus 82801H (ICH8 Family) USB UHCI Contro
/0/100/1d.2 bus 82801H (ICH8 Family) USB UHCI Contro
/0/100/1d.7 bus 82801H (ICH8 Family) USB2 EHCI Contr
/0/100/1e bridge 82801 Mobile PCI Bridge
/0/100/1e/0 bridge RL5c476 II
/0/100/1e/0.1 bus R5C832 IEEE 1394 Controller
/0/100/1f bridge 82801HBM (ICH8M-E) LPC Interface Con
/0/100/1f.1 scsi0 storage 82801HBM/HEM (ICH8M/ICH8M-E) IDE Con
/0/100/1f.1/0.0.0 /dev/sda disk 160GB ST9160823ASG
/0/100/1f.1/0.0.0/1 /dev/sda1 volume 149GiB EXT4 volume
/0/100/1f.2 scsi2 storage 82801HBM/HEM (ICH8M/ICH8M-E) SATA AH
/0/100/1f.2/0.0.0 /dev/sdb disk 63GB Patriot PS-100 6
/0/100/1f.2/0.0.0/1 /dev/sdb1 volume 3906MiB Linux swap volume
/0/100/1f.2/0.0.0/2 /dev/sdb2 volume 55GiB EXT4 volume
/0/100/1f.3 bus 82801H (ICH8 Family) SMBus Controlle
/1 power 42T4530


I will note that running this command, copy/pasting, and typing this response took a good 8 minutes. I also note that as I write this, the cursor constantly locks up and Chrome is freezing over and over again (same with Firefox) - maybe every 2-10 seconds. The computer isn't necessarily frozen, as the clock is still ticking and I can hover my mouse over icons and get tooltips. To prove the behavior isn't due to my browser, I tried to click Applications->Accessories->Terminal. The system hung up on clicking 'Accessories'. The menu pops up the same time my browser starts working again. This behavior exists both in a fresh install of Lucid as well as a Karmic install->Lucid upgrade.

[azredwing]

I suppose I'm the only person with this issue, then?

I'll probably revert to my non-SSD drive with Karmic on it. Lucid is totally unusable for me unless it's plugged in, and that pretty much defeats the purpose of having a laptop in the first place. :(

[miegiel]

I suppose I'm the only person with this issue, then?

I'll probably revert to my non-SSD drive with Karmic on it. Lucid is totally unusable for me unless it's plugged in, and that pretty much defeats the purpose of having a laptop in the first place. :(

That might help, or a clean install. :-k It could be a power management configuration hiccup, but I wouldn't really know how to troubleshoot that.

[azredwing]

That might help, or a clean install. :-k It could be a power management configuration hiccup, but I wouldn't really know how to troubleshoot that.

It's interesting because I originally started as a clean install, and that didn't work, which is why I started from Karmic and upgraded.

[hugmenot]

I have a Lenovo Thinkpad with an SSD in it. I have no problems at all. Neither on AC nor on battery.

So I think something odd is going on on your box. And from your description of the latencies you get, something rather grave goes wrong there.

No idea what it is however.

[miegiel]

It's interesting because I originally started as a clean install, and that didn't work, which is why I started from Karmic and upgraded.

What went wrong with the clean install? Did you try the alternate iso instead of the desktop iso? The alternate iso is more simple and less picky than the desktop iso, but it lacks the live-CD boot option.

[azredwing]

What went wrong with the clean install? Did you try the alternate iso instead of the desktop iso? The alternate iso is more simple and less picky than the desktop iso, but it lacks the live-CD boot option.

I'll give that a try. Nothing went wrong with the clean install, I just had the same latency issues and I wanted to see if the clean install was the issue. (Clearly it isn't.) I'll give the alternate iso a shot sometime this week.

[miegiel]

I'll give that a try. Nothing went wrong with the clean install, I just had the same latency issues and I wanted to see if the clean install was the issue. (Clearly it isn't.) I'll give the alternate iso a shot sometime this week.

Ok, I misunderstood you there. In that case the alternate iso might not help, but you never know ;)

There should be a Thinkpad T61 thread in the forums somewhere, maybe someone in the thread has already installed lucid too.

[azredwing]

Ok, I misunderstood you there. In that case the alternate iso might not help, but you never know ;)

There should be a Thinkpad T61 thread in the forums somewhere, maybe someone in the thread has already installed lucid too.

As expected, a clean install using the alternate iso did not work. Additionally, I tried a clean install with no swap partition and with my second HDD disconnected. Still having issues.

At this point I'm going to revert to Karmic. I'll see if I'm having issues with Karmic on the SSD as well.

[azredwing]

Running Karmic with SSD, 2GB swap and my 160GB secondary HDD (SATA 7200rpm) yields no problems - so the issue must be in Lucid.

I'll file a bug report... (but against what package?...) maybe someone else has some idea of what the hell is going on. I will also try an upgrade from Karmic to Lucid one more time to confirm the issue.

[hugmenot]

Have you looked around dmesg and in syslog?

[azredwing]

Running dmsg when my lappy is plugged in looks fine. Immediately after unplugging it, however, this occurs over and over again:


[ 78.263100] ata3.00: exception Emask 0x10 SAct 0x0 SErr 0x4040000 action 0xe frozen
[ 78.263104] ata3.00: irq_stat 0x00000040, connection status changed
[ 78.263107] ata3: SError: { CommWake DevExch }
[ 78.263110] ata3.00: failed command: READ DMA
[ 78.263114] ata3.00: cmd c8/00:08:67:63:42/00:00:00:00:00/e0 tag 0 dma 4096 in
[ 78.263115] res 50/00:08:5f:63:42/00:00:00:00:00/e0 Emask 0x10 (ATA bus error)
[ 78.263117] ata3.00: status: { DRDY }
[ 78.263121] ata3: hard resetting link
[ 79.010376] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
[ 84.010355] ata3.00: qc timeout (cmd 0xec)
[ 84.010376] ata3.00: failed to IDENTIFY (I/O error, err_mask=0x4)
[ 84.010383] ata3.00: revalidation failed (errno=-5)
[ 84.010396] ata3: hard resetting link
[ 84.360215] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
[ 84.363544] ata3.00: ACPI cmd ef/02:00:00:00:00:a0 (SET FEATURES) succeeded
[ 84.363555] ata3.00: ACPI cmd f5/00:00:00:00:00:a0 (SECURITY FREEZE LOCK) filtered out
[ 84.370494] ata3.00: ACPI cmd ef/02:00:00:00:00:a0 (SET FEATURES) succeeded
[ 84.370505] ata3.00: ACPI cmd f5/00:00:00:00:00:a0 (SECURITY FREEZE LOCK) filtered out
[ 84.373425] ata3.00: configured for UDMA/133
[ 84.380390] ata3.00: configured for UDMA/133
[ 84.380421] ata3: EH complete


So it looks to be a HDD issue for sure - but why only on battery mode?

I was thinking of upgrading the firmware on the SSD. (However, I don't have a Windows machine to do it. Ha.)

[azredwing]

It looks like some guys ran into the same things:

http://swiss.ubuntuforums.org/showthread.php?t=1034762

Not sure if this applies to me, since I'm not getting { DRDY ERR }'s. Somehow I hit a kernel panic and can't boot up, so I'm in the process of reinstalling (again!) now.

I will try again and set


options libata noacpi=1


in /etc/modprobe.d/options .

All signs seem to point to a HDD failure, but (1) this HDD is brand new and (2) it works fine when running Karmic.

(man, this CANNOT be good for my SSD.)

[azredwing]

Setting that option does nothing. I tried adding


noapic acpi=off


to /boot/grub/grub.cfg, but doing so actually causes the ata error to appear before getting to the login screen, even with the computer plugged in.

I'll keep investigating...

Does anyone know if there's a way to flash new firmware for the SSD HDD in Linux?

[azredwing]

$ sudo hdparm -i /dev/sda

/dev/sda:

Model=Patriot, FwRev=VER2.006, SerialNo=7B420701140800003044
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
RawCHS=16383/16/63, TrkSize=32256, SectSize=512, ECCbytes=4
BuffType=DualPort, BuffSize=1kB, MaxMultSect=1, MultSect=off
CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=124780320
IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2
UDMA modes: udma0 udma1 udma2 udma3 udma4 udma5 *udma6
AdvancedPM=yes: unknown setting WriteCache=enabled
Drive conforms to: unknown: ATA/ATAPI-3,4,5,6,7

* signifies the current active mode

[miegiel]

I had to use libata.force=noncq as a kernel parameter in grub to be able to boot linux on this laptop for a while. That bug is solved for me now, but there's a whole list of libata.... kernel parameters to give the kernel as it boots. Check this (http://www.kernel.org/doc/Documentation/kernel-parameters.txt), libata stuff is just above the middle of the long page. libata.noacpi is in there too, just use it like that, no =1 at the end (libata.acpi forces it on).

Hope this helps a bit, good luck.

[azredwing]

I had to use libata.force=noncq as a kernel parameter in grub to be able to boot linux on this laptop for a while. That bug is solved for me now, but there's a whole list of libata.... kernel parameters to give the kernel as it boots. Check this (http://www.kernel.org/doc/Documentation/kernel-parameters.txt), libata stuff is just above the middle of the long page. libata.noacpi is in there too, just use it like that, no =1 at the end (libata.acpi forces it on).

Hope this helps a bit, good luck.

libata.noacpi and libataforce=noncq did not work for me. Neither did upgrading my BIOS.

HOWEVER

I changed the SATA controller type from AHCI to "Compatible" in the BIOS and it seems to be running perfectly now! I will investigate whether it was JUST this that did the trick, or if it was a combination of that + the kernel boot options.

I'd like to point out again that the SSD was working just fine in Karmic with AHCI and no other kernel options. Should this issue be considered a bug and written up?

Thanks a ton for your help!

[miegiel]

libata.noacpi and libataforce=noncq did not work for me. Neither did upgrading my BIOS.

HOWEVER

I changed the SATA controller type from AHCI to "Compatible" in the BIOS and it seems to be running perfectly now! I will investigate whether it was JUST this that did the trick, or if it was a combination of that + the kernel boot options.

I'd like to point out again that the SSD was working just fine in Karmic with AHCI and no other kernel options. Should this issue be considered a bug and written up?

Thanks a ton for your help!

It might be a good idea to stick your laptop's SSD into a friends windows machine to update the firmware before reporting this as a bug. Laptop sata disks have the same size/type connectors as the larger sata disks in desktop PCs.

Your "Compatible" mode is the old ATA mode and a fall back in case of incompatibilities. It turns of some functions you might want to keep: NCQ (improves performance if your disk supports it too), hot-swapping disks (useful for servers, but not for laptops), improved security (on my laptop I can't secure the disk with a BIOS password unless AHCI is enabled) and some other little things.