sysrq-w and sysrq-t useless with cgroups and systemd enabled?

[Andy Lutomirski]

For several kernel and Fedora versions now, I've gotten only garbage
out of sysrq-w and sysrq-t.  The output looks like this:

[432904.642680] rt_rq[0]:/system/home.mount
[432904.642681]   .rt_nr_running                 : 0
[432904.642682]   .rt_throttled                  : 0
[432904.642682]   .rt_time                       : 0.000000
[432904.642683]   .rt_runtime                    : 0.000000
[432904.642685]
[432904.642685] rt_rq[0]:/system/boot-efi.mount
[432904.642686]   .rt_nr_running                 : 0
[432904.642687]   .rt_throttled                  : 0
[432904.642688]   .rt_time                       : 0.000000
[432904.642689]   .rt_runtime                    : 0.000000
[432904.642690]
[432904.642690] rt_rq[0]:/system/boot.mount
[432904.642691]   .rt_nr_running                 : 0
[432904.642692]   .rt_throttled                  : 0
[432904.642693]   .rt_time                       : 0.000000
[432904.642694]   .rt_runtime                    : 0.000000
[432904.642695]
[432904.642695] rt_rq[0]:/system/fsck@.service
[432904.642696]   .rt_nr_running                 : 0
[432904.642697]   .rt_throttled                  : 0
[432904.642698]   .rt_time                       : 0.000000
[432904.642699]   .rt_runtime                    : 0.000000
[432904.642700]
[432904.642700] rt_rq[0]:/system/cryptsetup@.service
[432904.642701]   .rt_nr_running                 : 0
[432904.642702]   .rt_throttled                  : 0
[432904.642703]   .rt_time                       : 0.000000
[432904.642704]   .rt_runtime                    : 0.000000
[432904.642705]
[432904.642706] rt_rq[0]:/system/dev-mapper-vg_midnight_ssd\x2dswap.swap
[432904.642707]   .rt_nr_running                 : 0
[432904.642708]   .rt_throttled                  : 0
[432904.642708]   .rt_time                       : 0.000000
[432904.642709]   .rt_runtime                    : 0.000000

except that it goes on for pages and pages.  I can't find any
information about actual hung tasks in the mess (even by searching the
log -- I think the interesting information is either not there or is
so swamped by log spew that it's actually getting lost by my syslog
daemon).

Is this some kind of bad interaction between scheduling groups and the
task dumper that's getting exacerbated by the huge number of cgroups
that systemd creates?

--Andy
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/