Bug 1084811

Summary: E-sata hotplugging stopped working since kernel 4.15
Product: [openSUSE] openSUSE Tumbleweed Reporter: Attila Schler <attila.schler>
Component: KernelAssignee: Hannes Reinecke <hare>
Status: RESOLVED WORKSFORME QA Contact: E-mail List <qa-bugs>
Severity: Major    
Priority: P5 - None CC: attila.schler, hare, mrmazda, tiwai
Version: CurrentFlags: tiwai: needinfo? (attila.schler)
Target Milestone: ---   
Hardware: x86-64   
OS: All   
Whiteboard:
Found By: --- Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Attachments: k4.12.14-hwinfo-hotp18.29
k4.12.14-kernelmsg-hotp18.29-off18.33
k4.15.8-hwinfo-hotp17.59
k4.15.8-kernelmsg-hotp17.59

Description Attila Schler 2018-03-11 19:20:50 UTC
I have noticed that on my Dell E6520 notebook the e-sata hotplugging stopped working since kernel 4.15.

Currently I can observe this problem with kernel:
4.15.7-1.6-x86_64
4.16.0-rc4-4.g7a5941a-x86-64

The e-sata cable is OK, and an e-sata device can be used with kernel 4.15 if it is turned on and plugged in before powering on my notebook.

(The e-sata hotplugging works well with the Leap kernel 4.12.14-lp150.5.3-x86_64.)

___
(I’m not an advanced user, but I’m ready to follow detailed instructions.)
Comment 1 Takashi Iwai 2018-03-16 11:26:30 UTC
Could you give the kernel messages on both working and non-working cases at e-SATA hotplug?

Also, give hwinfo output, too.

It's basically an upstream kernel regression and we should report to upstream.  But we need the basic hardware information for that.
Comment 2 Attila Schler 2018-03-17 17:57:59 UTC
Both hwinfo executed after hotplugging.

With kernel 4.15.8 the hotplugging was at ~17:59.

With kernel 4.12.14 the hotplugging was at ~18:29. 
~18:31 – disk mounted in KDE trough Dolphin
~18:32 – safely remove was clicked in Dolphin
~18:33 – the disk turned off
Comment 3 Attila Schler 2018-03-17 17:59:13 UTC
Created attachment 764051 [details]
k4.12.14-hwinfo-hotp18.29
Comment 4 Attila Schler 2018-03-17 18:00:11 UTC
Created attachment 764052 [details]
k4.12.14-kernelmsg-hotp18.29-off18.33
Comment 5 Attila Schler 2018-03-17 18:00:37 UTC
Created attachment 764053 [details]
k4.15.8-hwinfo-hotp17.59
Comment 6 Attila Schler 2018-03-17 18:01:00 UTC
Created attachment 764054 [details]
k4.15.8-kernelmsg-hotp17.59
Comment 7 Jiri Slaby 2018-06-15 13:28:26 UTC
Hmm, there is absolutely nothing about the actual plug in the new kernel. Does it still happen with 4.17?
Comment 8 Attila Schler 2018-06-15 23:00:51 UTC
(In reply to Jiri Slaby from comment #7)
> Hmm, there is absolutely nothing about the actual plug in the new kernel.
> Does it still happen with 4.17?

Yes, currently with 4.17.1-4.gdf028bb-default.
Comment 9 Attila Schler 2018-06-23 14:30:36 UTC
The same with kernel 4.18.0-rc1-2.gbf046be-default.
Comment 10 Attila Schler 2018-07-09 12:59:51 UTC
The drive can be detected with:

# echo 0 0 0 | tee /sys/class/scsi_host/host*/scan

And then can be mounted in Dolhpin as beforehand.

Currently I'm on kernel 4.18.0-rc3-2.gdd5896f-default, and the reported problem is still there.

I did not tried the mentioned command with other kernel versions.
Comment 11 Attila Schler 2018-07-09 13:16:41 UTC
Plus the mentioned command need to be used to "undetect" the disk after doing a "safely remove" in Dolphin. In this way will go away the sdb after its drive turned off.
Comment 12 Attila Schler 2022-07-03 16:15:42 UTC
tlp !!!

It seems that this issue is the side-effect of tlp. I just uninstalled it and e-sata hotplugging started working again.

(This tlp and the like stuff should never be automatically installed on user systems without the most explicit and detailed info on what it will do by default, and without explicit consent from the user.)