[PATCH 0/1] UBUNTU: SAUCE: Increase ATA_TMOUT_FF_WAIT to 5 seconds.

[Stefan Bader]

TJ wrote:
> Some users are experiencing kernel oops upon resume from suspend due to
> slow ATA devices:
> 
> [10692.148085] sd 0:0:0:0: [sda] Starting disk
> [10697.032030] ata1: link is slow to respond, please be patient (ready=0)
> [10698.992081] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
> [10699.338461] WARNING: at /build/buildd/linux-2.6.27/kernel/power/main.c:176 suspend_test_finish+0x74/0x80()
> 
> https://bugs.edge.launchpad.net/buga/286672
> 
> There was a related bug #318978 "Hard drive in Studio XPS 13 and 16
> cause a 17-18s resume time" which applied a patch (commit b65db6fd5d) to
> increase ATA_TMOUT_PMP_SRST_WAIT to 5 seconds. In that case it was
> increasing the time allowed for the soft reset.
> 
> https://bugs.edge.launchpad.net/bugs/331415
> 
> I suspect this issue could be solved by increasing the maximum
> ATA_TMOUT_FF_WAIT to 5 seconds in ata_wait_ready().
> 
> Please review the issue and the proposed patch which follows.
> 
> TJ.
> 
> 
I am not really convinced this will help. I might still miss something but this 
warning gets printed after 5s when there is still more than 3s until the 
deadline (which gets passed in by the caller). The only place where 
ATA_TMOUT_FF_WAIT gets used is to set the nodev_deadline. And this in turn 
seems to be used only if ata_link_online() is not true.
So my interpretation of this would be, that the 800ms timeout is not in effect 
here (otherwise the warning should not show up). And even if this would be 
increased, it gets limited by the deadline passed into the function.

Stefan

-- 

When all other means of communication fail, try words!