Ben Hutchings (2018-07-30): > Or perhaps by these changes listed in the changelog: > > * Improve battery life on laptops (Closes: #898629) > Thanks to Hans de Goede <hdegoede@redhat.com> > - ATA: A new SATA link-powermanagement-policy will be the default on all > Intel mobile chipsets. This can be reverted by passing > "ahci.mobile_lpm_policy=0" on the kernel commandline. > - sound: Enable Intel HDA codec power-saving by default with a 1 second > timeout. This can be overridden by passing "snd_hda_intel.power_save=0" > on the kernel commandline (0 is previously default). > > The hints there are meant as workarounds. Let us know if they work for > you. We can disable power-saving by default on specific hardware where > it doesn't work properly. I had not thought the issues were direct consequences of intentional changes. I should have, I am sorry for wasting your time that way. The audio issue is indeed fixed with: options snd-hda-intel power_save=0 in /etc/modprobe.d/snd_hda_intel.conf. Also, the 0.16 seconds delay is probably only a problem with my very specific use case. It can be forgotten. The non-responsiveness of the hard drive is NOT fixed by: options ahci mobile_lpm_policy=0 in /etc/modprobe.d/ahci.conf nor by by ahci.mobile_lpm_policy=0 on the kernel command line (I checked, for both cases, /sys/module/ahci/parameters/mobile_lpm_policy to see they were taken into account). I still get: [26825.694342] INFO: task md0_raid1:201 blocked for more than 120 seconds. [26825.694351] Tainted: G I 4.17.0-1-amd64 #1 Debian 4.17.8-1 [26825.694353] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [26825.694356] md0_raid1 D 0 201 2 0x80000000 [26825.694362] Call Trace: [26825.694374] ? __schedule+0x291/0x870 [26825.694381] schedule+0x28/0x80 [26825.694393] md_super_wait+0x6e/0xa0 [md_mod] [26825.694402] ? finish_wait+0x80/0x80 [26825.694415] write_page+0x249/0x330 [md_mod] [26825.694420] ? __schedule+0x299/0x870 [26825.694429] md_update_sb.part.62+0x367/0x840 [md_mod] [26825.694441] md_check_recovery+0x4ab/0x570 [md_mod] [26825.694451] raid1d+0x5c/0x880 [raid1] [26825.694461] ? lock_timer_base+0x67/0x80 [26825.694466] ? try_to_del_timer_sync+0x4d/0x80 [26825.694471] ? del_timer_sync+0x35/0x40 [26825.694476] ? schedule_timeout+0x181/0x380 [26825.694484] ? find_pers+0x70/0x70 [md_mod] [26825.694492] ? md_thread+0x122/0x160 [md_mod] [26825.694497] ? handle_read_error+0x4c0/0x4c0 [raid1] [26825.694504] md_thread+0x122/0x160 [md_mod] [26825.694510] ? finish_wait+0x80/0x80 [26825.694518] kthread+0x113/0x130 [26825.694523] ? kthread_create_worker_on_cpu+0x70/0x70 [26825.694529] ret_from_fork+0x35/0x40 And this is a rather serious bug. Do you think of anything I should try to get more diagnostics? Regards, -- Nicolas George
Attachment:
signature.asc
Description: Digital signature