Suspend-to-RAM on CentOS Server (Dell T20)

General support questions
Post Reply
jol
Posts: 2
Joined: 2016/04/23 10:44:08

Suspend-to-RAM on CentOS Server (Dell T20)

Post by jol » 2016/04/23 11:56:11

Hello,
I am new to CentOS, but do have some experience with other linux variants like OpenWRT or Bananian and I also do have a Windows server that I want to replace with a Linux based system. Thus I installed a CentOS (minimal) server (with encrypted partitions if that matters, need to fiddle with how to provide the keys reasonably) on a Dell T20. As the server is not used 24*7 I would like to install a simple monitor and when no activity is detected suspend-to-ram, from where I can then WOL the system again. WOL is already working, however the suspend-to-ram does not work as expected: the disk spins down (saving ca. 5W), network is inactive (WOL works, but a ssh session does not work), but the power led does not flash as other systems do and the power meter shows 45W - not exactly what I expected.
Dmesg shows something like this:

Code: Select all

[ 1682.283665] PM: Syncing filesystems ... done.
[ 1682.288218] Freezing user space processes ... (elapsed 0.001 seconds) done.
[ 1682.289376] Freezing remaining freezable tasks ... (elapsed 0.000 seconds) done.
[ 1682.290045] Suspending console(s) (use no_console_suspend to debug)
[ 1682.290301] sd 1:0:0:0: [sdb] Synchronizing SCSI cache
[ 1682.290349] sd 0:0:0:0: [sda] Synchronizing SCSI cache
[ 1682.290380] sd 1:0:0:0: [sdb] Stopping disk
[ 1682.290566] sd 0:0:0:0: [sda] Stopping disk
[ 1682.499311] e1000e: EEE TX LPI TIMER: 00000011
[ 1682.850513] PM: suspend of devices complete after 559.857 msecs
[ 1682.873589] PM: late suspend of devices complete after 23.052 msecs
[ 1682.873897] ehci-pci 0000:00:1d.0: System wakeup enabled by ACPI
[ 1682.875481] ehci-pci 0000:00:1a.0: System wakeup enabled by ACPI
[ 1682.876480] e1000e 0000:00:19.0: System wakeup enabled by ACPI
[ 1682.888488] xhci_hcd 0000:00:14.0: System wakeup enabled by ACPI
[ 1682.892537] PM: noirq suspend of devices complete after 18.930 msecs

[ 1706.955963] xhci_hcd 0000:00:14.0: System wakeup disabled by ACPI
[ 1706.969950] ehci-pci 0000:00:1a.0: System wakeup disabled by ACPI
[ 1706.971952] ehci-pci 0000:00:1d.0: System wakeup disabled by ACPI
[ 1706.973174] PM: noirq resume of devices complete after 32.248 msecs
[ 1706.977822] PM: early resume of devices complete after 4.611 msecs
[ 1706.977949] mei_me 0000:00:16.0: irq 29 for MSI/MSI-X
[ 1706.977985] e1000e 0000:00:19.0: System wakeup disabled by ACPI
[ 1706.978058] snd_hda_intel 0000:00:1b.0: irq 32 for MSI/MSI-X
[ 1706.979329] sd 0:0:0:0: [sda] Starting disk
[ 1706.979356] sd 1:0:0:0: [sdb] Starting disk
[ 1707.007303] e1000e 0000:00:19.0 em1: PHY Wakeup cause - Magic Packet
[ 1707.082162] e1000e 0000:00:19.0: irq 33 for MSI/MSI-X
[ 1707.286166] ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 1707.288175] ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
[ 1707.311703] ata2.00: configured for UDMA/133
[ 1709.881038] e1000e: em1 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
[ 1710.154326] ata1.00: configured for UDMA/133
[ 1710.181758] PM: resume of devices complete after 3201.348 msecs
[ 1710.181881] Restarting tasks ... done.
[ 1710.182177] video LNXVIDEO:00: Restoring backlight state
[root@dell02 systemd]#
I already removed any USB devices (after startup as I need to enter keys during boot). Any suggestion what to analyze or try in order to get suspend to work? Please don´t argue a server should run 24h, but I understand that it might not be typical and some package installations or configurations needed..
Thanks for any hint.
Joachim
Dell T20, CentOS Linux release 7.2.1511 (Core).

jol
Posts: 2
Joined: 2016/04/23 10:44:08

Re: Suspend-to-RAM on CentOS Server (Dell T20)

Post by jol » 2016/04/24 07:38:37

Did some additional research on my own. Following https://01.org/blogs/rzhang/2015/best-p ... ate-issues I added initcall_debug to the kernel command line. dmesg now reports during system boot:

Code: Select all

[    0.078502] calling  acpi_init+0x0/0x289 @ 1
[    0.078513] ACPI: Added _OSI(Module Device)
[    0.078514] ACPI: Added _OSI(Processor Device)
[    0.078515] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.078516] ACPI: Added _OSI(Processor Aggregator Device)
[    0.079662] ACPI: EC: Look up EC in DSDT
[    0.080885] ACPI: Executed 1 blocks of module-level executable AML code
[    0.082501] [Firmware Bug]: ACPI: BIOS _OSI(Linux) query ignored
[    0.082790] \_SB_:_OSC invalid UUID
[    0.082791] _OSC request data:1 1f
[    0.084480] ACPI: SSDT 00000000dbff5c18 003D3 (v01  PmRef  Cpu0Cst 00003001 INTL 20120711)
[    0.084770] ACPI: Dynamic OEM Table Load:
[    0.084772] ACPI: SSDT           (null) 003D3 (v01  PmRef  Cpu0Cst 00003001 INTL 20120711)
[    0.086569] ACPI: SSDT 00000000dbff5618 005AA (v01  PmRef    ApIst 00003000 INTL 20120711)
[    0.086910] ACPI: Dynamic OEM Table Load:
[    0.086911] ACPI: SSDT           (null) 005AA (v01  PmRef    ApIst 00003000 INTL 20120711)
[    0.088480] ACPI: SSDT 00000000dbff4d98 00119 (v01  PmRef    ApCst 00003000 INTL 20120711)
[    0.088766] ACPI: Dynamic OEM Table Load:
[    0.088767] ACPI: SSDT           (null) 00119 (v01  PmRef    ApCst 00003000 INTL 20120711)
[    0.090788] ACPI: Interpreter enabled
[    0.090793] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S1_] (20130517/hwxface-571)
[    0.090796] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S2_] (20130517/hwxface-571)
[    0.090799] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S3_] (20130517/hwxface-571)
[    0.090807] ACPI: (supports S0 S4 S5)
[    0.090808] ACPI: Using IOAPIC for interrupt routing
[    0.090834] HEST: Table parsing has been initialized.
[    0.090836] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
[    0.106848] ACPI: \_PR_.CPU4: failed to get CPU physical ID.
[    0.106851] ACPI: \_PR_.CPU5: failed to get CPU physical ID.
[    0.106853] ACPI: \_PR_.CPU6: failed to get CPU physical ID.
[    0.106855] ACPI: \_PR_.CPU7: failed to get CPU physical ID.
[    0.106995] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-3e])
[    0.106998] acpi PNP0A08:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI]
[    0.107362] acpi PNP0A08:00: _OSC: OS now controls [PCIeHotplug PME AER PCIeCapability]
According to http://askubuntu.com/questions/175793/w ... tput-means"[Firmware Bug]: ACPI: BIOS _OSI(Linux) query ignored" is not relevant. However when googling for "ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State" I found https://bugs.launchpad.net/ubuntu/+sour ... ug/1485223. How can I confirm this is the same with CentOS? How can I get that fix into my own system?
Dell T20, CentOS Linux release 7.2.1511 (Core).

User avatar
TrevorH
Site Admin
Posts: 33216
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: Suspend-to-RAM on CentOS Server (Dell T20)

Post by TrevorH » 2016/04/24 10:17:11

Since it's an upstream kernel bug that was not fixed until 2015-12-29 it's unlikely to have made its way into the RHEL kernel yet unless someone has reported it on bugzilla.redhat.com or via a support contract. Your first step would be to search b.r.c for existing bug reports about it and create one if not found.
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

Post Reply