[Bug 230290] [patch] acpi: call sleep event handler when sleeping via command line

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=230290

            Bug ID: 230290
           Summary: [patch] acpi: call sleep event handler when sleeping
                    via command line
           Product: Base System
           Version: CURRENT
          Hardware: Any
                OS: Any
            Status: New
          Severity: Affects Many People
          Priority: ---
         Component: kern
          Assignee: bugs at FreeBSD.org
          Reporter: johalun0 at gmail.com
                CC: hps at FreeBSD.org

Created attachment 195759
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=195759&action=edit
Patch for acpi.c

When adding listeners to acpi_sleep_event I realized that suspend via
lid/button does not behave the same way as suspend via zzz or acpiconf. 

The problem is that acpi's ioctl requests sleep state directly, ignoring to
invoke registered handlers. This patch will fix that. 

The drawback of this patch is that we have no chance to return an error from
the ioctl call. To do that we need to do something like

if (ACPI_FAILURE(AcpiOsExecute(OSL_NOTIFY_HANDLER, 
        acpi_invoke_sleep_eventhandler, &state)))
    return ERROR;
else 
    return 0;

However, 'state' in this case does not live long enough to be passed to the
event handler so we need to add something like:

static int acpi_states[6] = {0,1,2,3,4,5}

And pass &acpi_states[state]

The purpose is to inform linuxkpi that we're suspending/resuming. For now the
power_suspend_early event is used but that does not include what state we're
suspending to so we're forced to assume we're always suspending to S3.

Which would be more preferable?

-- 
You are receiving this mail because:
You are the assignee for the bug.