Created attachment 88351 [details] full lspci When trying to suspend to ram S3 the machine, one of these things will happen, apparently randomly: 1) The screen goes black, everything freezes for about 15-20 seconds (but fans stay on), then the LED blinks and the systems returns to the desktop 2) The screen goes black, everything freezes for quite a long time (30-60 seconds), then the system finally goes to S3 3) The system goes immediately to S3 1) is by far the most common, followed by 2). I have seen 3) (the expected behaviour) only a couple of times. Usually when 1) happens, trying again will produce 2). However, a few times 1) happened 2-3 times in a row before 2). When 1) happens, in dmesg there is: [11615.245918] PM: Syncing filesystems ... done. [11615.295922] Freezing user space processes ... (elapsed 0.01 seconds) done. [11615.306991] Freezing remaining freezable tasks ... (elapsed 0.01 seconds) done. [11615.317981] Suspending console(s) (use no_console_suspend to debug) [11615.318331] sd 5:0:0:0: [sda] Synchronizing SCSI cache [11615.319217] sd 5:0:0:0: [sda] Stopping disk [11615.321015] i8042 kbd 00:08: wake-up capability enabled by ACPI [11645.363404] ata6.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen [11645.363407] ata6.00: cmd e0/00:00:00:00:00/00:00:00:00:00/a0 tag 0 [11645.363407] res 40/00:00:00:00:00/00:00:00:00:00/40 Emask 0x4 (timeout) [11645.363411] ata6: hard resetting link [11645.363412] ata6: nv: skipping hardreset on occupied port [11645.817242] ata6: SATA link up 3.0 Gbps (SStatus 123 SControl 300) [11645.823417] ata6.00: configured for UDMA/100 [11645.823419] ata6.00: device reported invalid CHS sector 0 [11645.823425] ata6: EH complete [11645.823427] sd 5:0:0:0: [sda] START_STOP FAILED [11645.823428] sd 5:0:0:0: [sda] [11645.823429] Result: hostbyte=0x00 driverbyte=0x08 [11645.823430] sd 5:0:0:0: [sda] [11645.823431] Sense Key : 0xb [current] [descriptor] [11645.823432] sd 5:0:0:0: [sda] [11645.823433] ASC=0x0 ASCQ=0x0 [11645.823435] dpm_run_callback(): 0xffffffff81305120 returns 134217730 [11645.823438] PM: Device 5:0:0:0 failed to suspend async: error 134217730 [11645.823468] PM: Some devices failed to suspend [11645.823520] ohci_hcd 0000:00:0a.0: setting latency timer to 64 [11645.823529] ehci_hcd 0000:00:0a.1: setting latency timer to 64 [11645.823620] pci 0000:00:0e.0: setting latency timer to 64 [11645.823877] i8042 kbd 00:08: wake-up capability disabled by ACPI [11645.834250] pata_amd 0000:00:0c.0: setting latency timer to 64 [11645.834261] sata_nv 0000:00:0d.0: setting latency timer to 64 [11645.835216] sata_nv 0000:00:0d.1: setting latency timer to 64 [11645.835495] snd_hda_intel 0000:00:0e.1: setting latency timer to 64 [11645.837251] ata8: port disabled--ignoring [11645.990823] ata7.01: ACPI cmd ef/03:42:00:00:00:b0 (unknown) filtered out [11645.990825] ata7.01: ACPI cmd ef/03:0c:00:00:00:b0 (unknown) filtered out [11645.994898] ata7.00: ACPI cmd ef/03:42:00:00:00:a0 (unknown) filtered out [11645.994899] ata7.00: ACPI cmd ef/03:0c:00:00:00:a0 (unknown) filtered out [11645.994905] ata7: nv_mode_filter: 0x739f&0x739f->0x739f, BIOS=0x7000 (0xc0c00000) ACPI=0x701f (60:60:0x1f) [11645.994907] ata7: nv_mode_filter: 0x739f&0x739f->0x739f, BIOS=0x7000 (0xc0c00000) ACPI=0x701f (60:60:0x1f) [11646.000653] ata7.00: configured for UDMA/33 [11646.006734] ata7.01: configured for UDMA/33 [11646.545940] ata5: SATA link down (SStatus 0 SControl 300) [11646.558933] ata2: SATA link down (SStatus 0 SControl 300) [11646.558941] ata3: SATA link down (SStatus 0 SControl 300) [11646.558966] ata1: SATA link down (SStatus 0 SControl 300) [11646.558971] ata4: SATA link down (SStatus 0 SControl 300) [11646.784547] PM: resume of devices complete after 961.445 msecs [11646.784739] Restarting tasks ... done. [11647.064163] forcedeth 0000:00:10.0: irq 42 for MSI/MSI-X [11647.064203] forcedeth 0000:00:10.0: eth0: MSI enabled [11647.064426] forcedeth 0000:00:10.0: eth0: no link during initialization [11647.064864] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready [11648.885199] forcedeth 0000:00:10.0: eth0: link up [11648.885608] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready When 2) happens, in dmesg there is: [11801.240224] PM: Syncing filesystems ... done. [11801.290312] Freezing user space processes ... (elapsed 0.01 seconds) done. [11801.301706] Freezing remaining freezable tasks ... (elapsed 0.01 seconds) done. [11801.312676] Suspending console(s) (use no_console_suspend to debug) [11801.313046] sd 5:0:0:0: [sda] Synchronizing SCSI cache [11801.315940] i8042 kbd 00:08: wake-up capability enabled by ACPI [11861.477568] ata6.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen [11861.477571] ata6.00: cmd ea/00:00:00:00:00/00:00:00:00:00/a0 tag 0 [11861.477571] res 40/00:00:00:00:00/00:00:00:00:00/40 Emask 0x4 (timeout) [11861.477575] ata6: hard resetting link [11861.477575] ata6: nv: skipping hardreset on occupied port [11861.931333] ata6: SATA link up 3.0 Gbps (SStatus 123 SControl 300) [11861.937504] ata6.00: configured for UDMA/100 [11861.937505] ata6.00: retrying FLUSH 0xea Emask 0x4 [11861.940305] ata6.00: device reported invalid CHS sector 0 [11861.940311] ata6: EH complete [11861.942599] sd 5:0:0:0: [sda] Stopping disk [11861.955310] PM: suspend of devices complete after 60665.827 msecs [11861.955428] PM: late suspend of devices complete after 0.116 msecs [11861.966334] ehci_hcd 0000:00:0a.1: wake-up capability enabled by ACPI [11861.977301] ohci_hcd 0000:00:0a.0: wake-up capability enabled by ACPI [11861.988420] PM: noirq suspend of devices complete after 33.004 msecs [11861.989052] ACPI: Preparing to enter system sleep state S3 [11861.989449] PM: Saving platform NVS memory [11861.989486] Disabling non-boot CPUs ... [11861.991182] smpboot: CPU 1 is now offline [11861.992744] smpboot: CPU 2 is now offline [11861.994234] smpboot: CPU 3 is now offline [11861.994496] Extended CMOS year: 2000 So it seems the disk (a Crucial C300 256GB) is failing to suspend. The motherboard is an M2N32-SLI deluxe, lspci attached. With version 2.6.37.2 it worked without problems 100% of the times.
Actually this looks like it is being suspended too early
Similar bug as Bug 48951, from 2.6.38 some hard disks fail to process the STANDBY IMMEDIATE command on suspend.
Full dmesgs for both v2.6.38(the bad one) and v2.6.37(the good one) are helpful, please attach them, thanks.
I'd love to do that but after a CMOS clear S3 mysteriously started working (it has now suspended correctly about ten times in a row). Note that I have put exactly the same values as before in the BIOS. If dmesg would be useful anyway, tell me and I'll gather them.
Yes, please attach the dmesg. I want to know which host controller this failing port belongs to, thanks.
Created attachment 89371 [details] dmesg from kernel 2.6.37.2
Created attachment 89381 [details] dmesg from kernel 3.6.2
Here it is, sorry for the delay
NVIDIA MCP55 SATA controller, using sata_nv driver. It seems this controller has problems dealing with standby command, I've seen some reports on this: http://marc.info/?l=linux-ide&m=133534061316338&w=2 Bug 48951 I'll close this bug and if the problem happened again, feel free to re-open it, thanks.