Ubuntu
linux package

  1. Bug #610055
  2. Comment #30

Comment 30 for bug 610055

Revision history for this message
Roman Yepishev (rye) wrote : Re: [pm-utils][maverick] system stutters, freezes, hangs, gratuitous 'Commit=0' errors

Hi, I don't really think the hangs in my case are caused by partition remounting. I have pretty frequent system lock-ups (and then releases) while being on the battery in Natty. I don't remember this behavior in Maverick and no hardware change was introduced. My dmesg is filled by the following entries:
[43352.278509] EXT4-fs (sdb6): re-mounted. Opts: errors=remount-ro,commit=600
[43352.406605] EXT4-fs (sdb1): re-mounted. Opts: commit=600
[43352.711492] EXT4-fs (sdb7): re-mounted. Opts: user_xattr,commit=600
[43400.040135] ata3.00: exception Emask 0x0 SAct 0x1 SErr 0x150000 action 0x6 frozen
[43400.040158] ata3: SError: { PHYRdyChg CommWake Dispar }
[43400.040169] ata3.00: failed command: READ FPDMA QUEUED
[43400.040190] ata3.00: cmd 60/08:00:f8:a9:1d/00:00:02:00:00/40 tag 0 ncq 4096 in
[43400.040193] res 40/00:ff:00:00:00/00:00:00:00:00/40 Emask 0x4 (timeout)
[43400.040202] ata3.00: status: { DRDY }
[43400.040216] ata3: hard resetting link
[43400.390116] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
[43400.672898] ata3.00: configured for UDMA/133
[43400.673024] ata3.00: device reported invalid CHS sector 0
[43400.673057] ata3: EH complete
[43441.040135] ata3.00: exception Emask 0x0 SAct 0x1 SErr 0x150000 action 0x6 frozen
[43441.040151] ata3: SError: { PHYRdyChg CommWake Dispar }
[43441.040163] ata3.00: failed command: READ FPDMA QUEUED
[43441.040183] ata3.00: cmd 60/08:00:00:c5:1d/00:00:02:00:00/40 tag 0 ncq 4096 in
[43441.040186] res 40/00:ff:00:00:00/00:00:00:00:00/40 Emask 0x4 (timeout)
[43441.040195] ata3.00: status: { DRDY }
[43441.040208] ata3: hard resetting link
[43441.390119] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
[43441.675629] ata3.00: configured for UDMA/133
[43441.675744] ata3.00: device reported invalid CHS sector 0
[43441.675765] ata3: EH complete
[43462.040160] ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x150000 action 0x6 frozen
[43462.040177] ata3: SError: { PHYRdyChg CommWake Dispar }
[43462.040186] ata3.00: failed command: SET FEATURES
[43462.040204] ata3.00: cmd ef/05:fe:00:00:00/00:00:00:00:00/40 tag 0
[43462.040207] res 40/00:ff:00:00:00/00:00:00:00:00/40 Emask 0x4 (timeout)
[43462.040216] ata3.00: status: { DRDY }
[43462.040229] ata3: hard resetting link
[43462.390122] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 300)

So it looks like my hard drive is being put to sleep to conserve the energy in a way that completely degrades its performance. The HDD led stays on for quite a long periods of time.

I tried doing hdparm -S 0, -B254 or -B255 (apm disabled on hdd) - the latter was OK in all previous versions, but it looks like that it is gets broken even before the hard drive. I've found the following in my pm-powersave.log:

/usr/lib/pm-utils/power.d/intel-audio-powersave true: success.
Running hook /usr/lib/pm-utils/power.d/journal-commit true:
Setting journal commit time for / to 600...Done.
Setting journal commit time for /boot to 600...Done.
Setting journal commit time for /home to 600...Done.

/usr/lib/pm-utils/power.d/journal-commit true: success.
Running hook /usr/lib/pm-utils/power.d/laptop-mode true:
Laptop mode enabled.

/usr/lib/pm-utils/power.d/laptop-mode true: success.
Running hook /usr/lib/pm-utils/power.d/pcie_aspm true:

/usr/lib/pm-utils/power.d/pcie_aspm true: success.
Running hook /usr/lib/pm-utils/power.d/readahead true:
Setting readahead for /dev/sdb6 to 3072...Done.
Setting readahead for /dev/sdb1 to 3072...Done.
Setting readahead for /dev/sdb7 to 3072...Done.

/usr/lib/pm-utils/power.d/readahead true: success.
Running hook /usr/lib/pm-utils/power.d/sata_alpm true:
Setting SATA APLM on host3 to min_power...Done.
Setting SATA APLM on host4 to min_power...Done.
Setting SATA APLM on host5 to min_power...Done.
Setting SATA APLM on host6 to min_power...Done.

/usr/lib/pm-utils/power.d/sata_alpm true: success.
Running hook /usr/lib/pm-utils/power.d/sched-powersave true:
**sched policy powersave ON

So i went forward and did...
$ sudo env PM_FUNCTIONS=/usr/lib/pm-utils/functions /usr/lib/pm-utils/power.d/sata_alpm false
Setting SATA APLM on host3 to max_performance...Done.
Setting SATA APLM on host4 to max_performance...Done.
Setting SATA APLM on host5 to max_performance...Done.
Setting SATA APLM on host6 to max_performance...Done.

And my machine unfroze. It looks like there's something interesting going on with SATA APML on my machine.
My machine info can be found in LP#693598, that's Acer Aspire 5520.