The impact of this bug is that the suspend_test script might return false positives on some systems because of the way wakealarm might be supported. This fix provides a clean way to support wakealarm in all situations, assuming wakealarm is supported at all.
As requested on ubuntu-devel-announce  I've performed the "suspend_test --full" test. Unfortunately, the first test fails. In this test, the laptop is suspended and a wakeup-timer is set up. The wakeup-timer doesn't do what it's supposed to do and my laptop will only come back alive when I press the power button.
* My laptop is a HP Compaq 6710b
* In #307090 it's suggested to boot with "hpet=disable", which makes no difference in my case.
* It looks like http://bugzilla.kernel.org/show_bug.cgi?id=12013#c33 is related.
* See https://wiki.ubuntu.com/KernelTeam/SuspendResumeTesting/Feedback "| grep 6710b" as well.
pet@workmate:~$ uname -a
Linux workmate 2.6.28-9-generic #31-Ubuntu SMP Wed Mar 11 15:43:49 UTC 2009 x86_64 GNU/Linux
pet@workmate:~$ sudo sh -c 'echo 0 > /sys/class/rtc/rtc0/wakealarm'
pet@workmate:~$ sudo sh -c 'echo `date "+%s" -d "+ 5 minutes"` > /sys/class/rtc/rtc0/wakealarm'
pet@workmate:~$ cat /proc/driver/rtc
rtc_time : 18:12:16
rtc_date : 2009-03-27
alrm_time : 18:17:06
alrm_date : ****-**-27
alarm_IRQ : no
alrm_pending : no
24hr : yes
periodic_IRQ : no
update_IRQ : no
HPET_emulated : no
DST_enable : no
periodic_freq : 1024
batt_status : okay