clock_settime03

[rasibley]

While we updated to a recent commit for CKI, newly onboarded clock_settime03 test is failing with BROK: timer_create() failed: EOPNOTSUPP (95) error for all non x86_64 arches, see below:

>>> 80:clock_settime03 BROK <<<
###############################################################################
# Test Num    : 80                                                            #
# Test Case   : clock_settime03                                               #
# Test Result : BROK                                                          #
###############################################################################

     1	<<<test_start>>>
     2	tag=clock_settime03 stime=1596809785
     3	cmdline="clock_settime03"
     4	contacts=""
     5	analysis=exit
     6	<<<test_output>>>
     7	tst_test.c:1245: INFO: Timeout per run is 0h 05m 00s
     8	clock_settime03.c:41: INFO: Testing variant: syscall with old kernel spec
     9	clock_settime03.c:70: BROK: timer_create() failed: EOPNOTSUPP (95)
    10	
    11	Summary:
    12	passed   0
    13	failed   0
    14	skipped  0
    15	warnings 0
    16	<<<execution_status>>>
    17	initiation_status="ok"
    18	duration=0 termination_type=exited termination_id=2 corefile=no
    19	cutime=0 cstime=0
    20	<<<test_end>>>

[metan-ucw]

That is a bit strange, the syscall that fails is timer_create() with CLOCK_REALTIME_ALARM. I guess that there may be missing support for RTC wakeup either in kernel or hardware.

Does the test work if you change the CLOCK_REALTIME_ALARM into just CLOCK_REALTIME?

[jstancek]

http://lists.linux.it/pipermail/ltp/2020-August/018560.html

[jstancek]

From what I recall, the report was from ppc kvm guest.

[metan-ucw]

Well yes, but I do not see anything that would actually require the _ALARM clock, we do not suspend the OS in the middle, so it may be better to just change to plain CLOCK_REALTIME.

[Cypresslin]

For the record, we're seeing this failures on the following kernels as well (not fully tested on various releases yet):
Ubuntu Bionic 4.15 KVM kernel on KVM nodes (4.15.0-1072.73)
Ubuntu Xenial 4.4 kernel on P8 / s390x (4.4.0-189.219)

[wangli5665]

this has been solved in 21721a3