I can confirm this problem in Hardy. It is an easily visible regression on my laptop somewhere between kernels 2.6.24-7-generic and 2.6.24-11-generic. Fortunately I kept the old -7 kernel so I can compare the results. Today I upgraded to 2.6.24-12-generic and the problem still persists.
I don't know if there's any archive of kernel deb binaries for all the versions between -7 and -11 -- if I had them I could exactly pinpoint the version which caused the regression. The problem is clearly related to disk I/O. With a large I/O operation, like copying an ISO file, the system becomes very unresponsive, mouse is jerky, I have trouble switching between terminal windows or from X to the virtual terminal (ctrl-alt-f1). If I start a large enough disk operation, it can starve the system resources so that only a hard reboot can help. The laptop is HP Compaq nx5000 with Pentium M 1.7 GHz, 1GB RAM, 80GB SATA disk, latest Hardy Heron. Here's a comparison of: $ time cp -p valhalla-i386-disc1.iso z (copying a 650M file on the same filesystem -- XFS) Linux fafik 2.6.24-7-generic #1 SMP Thu Feb 7 01:29:58 UTC 2008 i686 GNU/Linux real 0m35.307s user 0m0.060s sys 0m2.556s Linux fafik 2.6.24-12-generic #1 SMP Wed Mar 12 23:01:54 UTC 2008 i686 GNU/Linux real 5m31.690s user 0m0.340s sys 1m11.932s $ cat /proc/version_signature Ubuntu 2.6.24-12.22-generic During the slow I/O operation, top seems to indicate that some kernel daemons like kondemand, kblockd take a significant amount of CPU. Obviously load average skyrockets. top - 21:57:33 up 6 min, 3 users, load average: 8.87, 6.21, 2.68 Tasks: 117 total, 3 running, 112 sleeping, 0 stopped, 2 zombie Cpu(s): 3.7%us, 27.4%sy, 0.0%ni, 0.0%id, 2.0%wa, 5.0%hi, 61.9%si, 0.0%st Mem: 1027128k total, 1012132k used, 14996k free, 100k buffers Swap: 0k total, 0k used, 0k free, 802832k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 6389 wanted 20 0 3140 772 644 R 42.9 0.1 0:52.41 cp 1812 root 15 -5 0 0 0 S 8.6 0.0 0:02.76 scsi_eh_1 5347 uml-net 20 0 1708 428 360 S 5.3 0.0 0:03.76 uml_switch 43 root 15 -5 0 0 0 S 4.6 0.0 0:05.78 kblockd/0 4223 root 15 -5 0 0 0 S 4.6 0.0 0:00.84 kcryptd 5585 root 20 0 3408 1128 976 S 4.6 0.1 0:03.12 hald-addon-stor 4970 root 15 -5 0 0 0 S 4.0 0.0 0:06.80 kondemand/0 1479 root 15 -5 0 0 0 S 3.3 0.0 0:01.50 ata/0 2994 root 15 -5 0 0 0 S 3.3 0.0 0:09.86 ipw2200/0 5678 root 20 0 210m 10m 6720 S 3.3 1.0 0:08.70 Xorg ** Attachment added: "lspci-vvnn.log" http://launchpadlibrarian.net/12676983/lspci-vvnn.log -- Occasionally, disk goes crazy, system freezes https://bugs.launchpad.net/bugs/174185 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
