This is weird. A few days ago I found that my desktop system (F23 fully updated) would no longer hibernate. A glance at "journal -xe" showed a bunch of lines like:
Jun 14 00:50:51 bree audit[12591]: SECCOMP auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12591 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c000003e syscall=273 compat=0 ip=0 Jun 14 00:50:51 bree kernel: audit: type=1326 audit(1465861851.266:7800): auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12591 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c00 Jun 14 00:50:59 bree audit[12599]: SECCOMP auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12599 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c000003e syscall=273 compat=0 ip=0 Jun 14 00:50:59 bree kernel: audit: type=1326 audit(1465861859.424:7801): auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12599 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c00 Jun 14 00:51:07 bree audit[12603]: SECCOMP auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12603 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c000003e syscall=273 compat=0 ip=0 Jun 14 00:51:07 bree kernel: audit: type=1326 audit(1465861867.128:7802): auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12603 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c00 Jun 14 00:51:20 bree audit[12608]: SECCOMP auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12608 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c000003e syscall=273 compat=0 ip=0 Jun 14 00:51:20 bree kernel: audit: type=1326 audit(1465861880.130:7803): auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12608 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c00 Jun 14 00:52:09 bree audit[12646]: SECCOMP auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12646 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c000003e syscall=273 compat=0 ip=0 Jun 14 00:52:09 bree kernel: audit: type=1326 audit(1465861929.397:7804): auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12646 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c00 Jun 14 00:52:25 bree audit[12662]: SECCOMP auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12662 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c000003e syscall=273 compat=0 ip=0 Jun 14 00:52:25 bree kernel: audit: type=1326 audit(1465861945.701:7805): auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12662 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c00 Jun 14 00:52:25 bree audit[12667]: SECCOMP auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12667 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c000003e syscall=273 compat=0 ip=0 Jun 14 00:52:25 bree kernel: audit: type=1326 audit(1465861945.768:7806): auid=1000 uid=1000 gid=1000 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12667 comm="chrome" exe="/opt/google/chrome/chrome" sig=0 arch=c00
and when I killed Chrome then hibernation worked again.
So one question is: what is the magic incantation to stop these alarms from Chrome?
However another question is how is this even possible? A process with normal user privileges can prevent the system from hibernating? Is this right?
poc
On Tue, 2016-06-14 at 10:34 +0100, Patrick O'Callaghan wrote:
This is weird. A few days ago I found that my desktop system (F23 fully updated) would no longer hibernate. A glance at "journal -xe" showed a bunch of lines like:
[...]
This is still happening to me, so as no-one has responded I reported it to BZ (NB: the SElinux violations I quoted are probably not related to this, as they've been going on for months):
https://bugzilla.redhat.com/show_bug.cgi?id=1346811
poc
On Wed, 2016-06-15 at 12:47 +0100, Patrick O'Callaghan wrote:
On Tue, 2016-06-14 at 10:34 +0100, Patrick O'Callaghan wrote:
This is weird. A few days ago I found that my desktop system (F23 fully updated) would no longer hibernate. A glance at "journal -xe" showed a bunch of lines like:
[...]
This is still happening to me, so as no-one has responded I reported it to BZ (NB: the SElinux violations I quoted are probably not related to this, as they've been going on for months):
Just for the record, it turned out to be insufficient swap. My system has 16GB of RAM and Chrome was eating about 5GB.
Now if only there was some way the system could have told me that, say an error log where useful messages would appear rather than a lot of cryptic text that boils down to "something failed" ...
poc
On 06/17/2016 08:47 AM, Patrick O'Callaghan wrote:
On Wed, 2016-06-15 at 12:47 +0100, Patrick O'Callaghan wrote:
On Tue, 2016-06-14 at 10:34 +0100, Patrick O'Callaghan wrote:
This is weird. A few days ago I found that my desktop system (F23 fully updated) would no longer hibernate. A glance at "journal -xe" showed a bunch of lines like:
[...]
This is still happening to me, so as no-one has responded I reported it to BZ (NB: the SElinux violations I quoted are probably not related to this, as they've been going on for months):
Just for the record, it turned out to be insufficient swap. My system has 16GB of RAM and Chrome was eating about 5GB.
Now if only there was some way the system could have told me that, say an error log where useful messages would appear rather than a lot of cryptic text that boils down to "something failed" ...
Geeze! We should have noticed that it was "hibernation". Yeah, that writes the entire RAM out to swap. Insufficient swap will sure cause hibernation to not work.
I guess the old saw about "make swap twice your RAM size" still holds true. I still set my systems to do it, but I'm an old, stick-in-the-mud, crusty and cranky curmudgeon. ---------------------------------------------------------------------- - Rick Stevens, Systems Engineer, AllDigital ricks@alldigital.com - - AIM/Skype: therps2 ICQ: 226437340 Yahoo: origrps2 - - - - ...Had this been an actual emergency, we would have fled in terror - - and you'd be on your own, pal! - ----------------------------------------------------------------------
On Fri, 2016-06-17 at 10:25 -0700, Rick Stevens wrote:
Now if only there was some way the system could have told me that,
say
an error log where useful messages would appear rather than a lot
of
cryptic text that boils down to "something failed" ...
Geeze! We should have noticed that it was "hibernation". Yeah, that writes the entire RAM out to swap. Insufficient swap will sure cause hibernation to not work.
It doesn't write the entire RAM, just what's in use, but I get it.
However I added a large swap file to bring the total up to 16GB and it *still* fails in the same way. And no, I'm not using more than 16GB of virtual memory:
$ free total used free shared buff/cache available Mem: 16361232 9910132 645876 497680 5805224 5503832 Swap: 16588792 3653100 12935692
The plot thickens.
poc
On Fri, 2016-06-17 at 18:37 +0100, Patrick O'Callaghan wrote:
On Fri, 2016-06-17 at 10:25 -0700, Rick Stevens wrote:
Now if only there was some way the system could have told me that,
say
an error log where useful messages would appear rather than a lot
of
cryptic text that boils down to "something failed" ...
Geeze! We should have noticed that it was "hibernation". Yeah, that writes the entire RAM out to swap. Insufficient swap will sure cause hibernation to not work.
It doesn't write the entire RAM, just what's in use, but I get it.
However I added a large swap file to bring the total up to 16GB and it *still* fails in the same way. And no, I'm not using more than 16GB of virtual memory:
$ free total used free shared buff/cache available Mem: 16361232 9910132 645876 497680 5805224 5503832 Swap: 16588792 3653100 12935692
The plot thickens.
It turns out that hibernation can't used multiple swap areas, so the only solution is to have a large enough swap partition to start with.
poc
On 06/19/2016 11:30 PM, Patrick O'Callaghan wrote:
On Fri, 2016-06-17 at 18:37 +0100, Patrick O'Callaghan wrote:
On Fri, 2016-06-17 at 10:25 -0700, Rick Stevens wrote:
Now if only there was some way the system could have told me that,
say
an error log where useful messages would appear rather than a lot
of
cryptic text that boils down to "something failed" ...
Geeze! We should have noticed that it was "hibernation". Yeah, that writes the entire RAM out to swap. Insufficient swap will sure cause hibernation to not work.
It doesn't write the entire RAM, just what's in use, but I get it.
However I added a large swap file to bring the total up to 16GB and it *still* fails in the same way. And no, I'm not using more than 16GB of virtual memory:
$ free total used free shared buff/cache available Mem: 16361232 9910132 645876 497680 5805224 5503832 Swap: 16588792 3653100 12935692
The plot thickens.
It turns out that hibernation can't used multiple swap areas, so the only solution is to have a large enough swap partition to start with.
poc
users mailing list users@lists.fedoraproject.org To unsubscribe or change subscription options: https://lists.fedoraproject.org/admin/lists/users@lists.fedoraproject.org Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines Have a question? Ask away: http://ask.fedoraproject.org
Might consider running a watchdog process on either chrome OR your swap file, which could throw an error if mem_in_use < mem_needed_for_hibernate
On Thu, 2016-07-07 at 22:58 +0000, Corey 'linuxmodder' Sheldon wrote:
Might consider running a watchdog process on either chrome OR your swap file, which could throw an error if mem_in_use < mem_needed_for_hibernate
I suspect that if it were a simple calculation then a hibernation attempt would fail immediately and not after around 30 seconds of disk activity, which is what happens.
poc
On Thu, 2016-07-07 at 22:58 +0000, Corey 'linuxmodder' Sheldon wrote:
Might consider running a watchdog process on either chrome OR your swap file, which could throw an error if mem_in_use < mem_needed_for_hibernate
I suspect that if it were a simple calculation then a hibernation attempt would fail immediately and not after around 30 seconds of disk activity, which is what happens.
poc
-
Corey Sheldon -
Patrick O'Callaghan -
Rick Stevens