[Bug 216706] New: USBC resume callback takes far too long, between 650ms and 1200ms

[Bug 216706] New: USBC resume callback takes far too long, between 650ms and 1200ms

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=216706

            Bug ID: 216706
           Summary: USBC resume callback takes far too long, between 650ms
                    and 1200ms
           Product: Drivers
           Version: 2.5
    Kernel Version: 6.1.0-rc2
          Hardware: All
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: USB
          Assignee: drivers_usb@kernel-bugs.kernel.org
          Reporter: todd.e.brandt@intel.com
            Blocks: 178231
        Regression: No

Created attachment 303218
  --> https://bugzilla.kernel.org/attachment.cgi?id=303218&action=edit
lenb-Dell-XPS-13-9310_freeze.html

The new USBC resume callback is taking far too long in the resume phase. I
bisected the behavior to this commit:

[10402] 4e3a50293c2b21961f02e1afa2f17d3a1a90c7c8 is the first bad commit
[10402] commit 4e3a50293c2b21961f02e1afa2f17d3a1a90c7c8
[10402] Author: Heikki Krogerus <heikki.krogerus@linux.intel.com>
[10402] Date:   Fri Oct 7 13:09:51 2022 +0300
[10402]
[10402] usb: typec: ucsi: acpi: Implement resume callback
[10402]
[10402] The ACPI driver needs to resume the interface by calling
[10402] ucsi_resume(). Otherwise we may fail to detect connections
[10402] and disconnections that happen while the system is
[10402] suspended.
[10402]
[10402] Link: https://bugzilla.kernel.org/show_bug.cgi?id=210425
[10402] Fixes: a94ecde41f7e ("usb: typec: ucsi: ccg: enable runtime pm
support")
[10402] Cc: <stable@vger.kernel.org>
[10402] Signed-off-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
[10402] Link:
https://lore.kernel.org/r/20221007100951.43798-3-heikki.krogerus@linux.intel.com
[10402] Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

I have timelines from 7 separate machines that are now affected but this
behavior. We have a guidline that neither suspend or resume should take longer
than a second, and this is causing us problems.


Referenced Bugs:

https://bugzilla.kernel.org/show_bug.cgi?id=178231
[Bug 178231] Meta-bug: Linux suspend-to-mem and freeze performance optimization
-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 216706] USBC resume callback takes far too long, between 650ms and 1200ms

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=216706

--- Comment #1 from Todd Brandt (todd.e.brandt@intel.com) ---
Created attachment 303219
  --> https://bugzilla.kernel.org/attachment.cgi?id=303219&action=edit
lenb-Dell-XPS-13-9315_freeze.html

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 216706] USBC resume callback takes far too long, between 650ms and 1200ms

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=216706

--- Comment #2 from Todd Brandt (todd.e.brandt@intel.com) ---
Created attachment 303220
  --> https://bugzilla.kernel.org/attachment.cgi?id=303220&action=edit
otcpl-adl-m-1_freeze.html

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 216706] USBC resume callback takes far too long, between 650ms and 1200ms

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=216706

--- Comment #3 from Todd Brandt (todd.e.brandt@intel.com) ---
Created attachment 303221
  --> https://bugzilla.kernel.org/attachment.cgi?id=303221&action=edit
otcpl-adl-n-1_freeze.html

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 216706] USBC resume callback takes far too long, between 650ms and 1200ms

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=216706

--- Comment #4 from Todd Brandt (todd.e.brandt@intel.com) ---
Created attachment 303222
  --> https://bugzilla.kernel.org/attachment.cgi?id=303222&action=edit
otcpl-dell-9370-kbl_mem.html

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 216706] USBC resume callback takes far too long, between 650ms and 1200ms

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=216706

--- Comment #5 from Todd Brandt (todd.e.brandt@intel.com) ---
Created attachment 303223
  --> https://bugzilla.kernel.org/attachment.cgi?id=303223&action=edit
otcpl-hp-spectre-tgl_freeze.html

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 216706] USBC resume callback takes far too long, between 650ms and 1200ms

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=216706

--- Comment #6 from Todd Brandt (todd.e.brandt@intel.com) ---
Created attachment 303224
  --> https://bugzilla.kernel.org/attachment.cgi?id=303224&action=edit
otcpl-lenovo-tix1-tgl_freeze.html

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

Re: [Bug 216706] New: USBC resume callback takes far too long, between 650ms and 1200ms

[Greg KH]

On Fri, Nov 18, 2022 at 08:24:14PM +0000, bugzilla-daemon@kernel.org wrote:
> I have timelines from 7 separate machines that are now affected but this
> behavior. We have a guidline that neither suspend or resume should take longer
> than a second, and this is causing us problems.

What is making this arbritrary requirement?  If it is an issue, please
provide a working change to resolve it as your group can help implement
and test this.

thanks,

greg k-h