close 866306
thanks
Hi Vagrant,
given that upstream adressed the report solely by merging the patch and
no further inquery about it, that haveged is not needed anymore unless
special setups, and given the bug is more than 7 years old with the last
message from 6 years ago.. I'm closing the bug.
On 2018-09-19, Natanael Copa wrote:
> On Mon, 03 Sep 2018 13:57:17 -0700
> Vagrant Cascadian wrote:
>> On 2018-09-03, Nicolas Braud-Santoni wrote:
>> > Could you confirm whether the patch solves the bug on your hardware?
>>
>> I'd be happy to try it out! Unfortunately, the patch sent to the deb
On Mon, 03 Sep 2018 13:57:17 -0700
Vagrant Cascadian wrote:
> On 2018-09-03, Nicolas Braud-Santoni wrote:
> > Hi Vagrant !
> >
> > Could you confirm whether the patch solves the bug on your hardware?
>
> I'd be happy to try it out! Unfortunately, the patch sent to the debian
> BTS is some xml
On Mon, 3 Sep 2018 21:54:08 +0200
Nicolas Braud-Santoni wrote:
> Control: tag -1 + patch upstream moreinfo
> Control: severity -1 important
>
> Hi Natanael !
>
> Thanks for providing the patch.
>
> Did you forward this patch upstream, and do you know whether it was
> included in the latest ups
On 2018-09-03, Nicolas Braud-Santoni wrote:
> Hi Natanael !
>
> Thanks for providing the patch.
>
> Did you forward this patch upstream, and do you know whether it was
> included in the latest upstream release?
>
>
> Hi Vagrant !
>
> Could you confirm whether the patch solves the bug on your hardwa
Control: tag -1 + patch upstream moreinfo
Control: severity -1 important
Hi Natanael !
Thanks for providing the patch.
Did you forward this patch upstream, and do you know whether it was
included in the latest upstream release?
Hi Vagrant !
Could you confirm whether the patch solves the bug o
Tag: Patch
Hi,
I bumped into this issue on Alpine Linux. It appears that the sysfs not
always report the cpu cache size on arm (both 32bit and 64 bit). It
reports size to be -1 and haveged does not handle that case.
I have seen the case on xgene and thunderx machines.
I have attached a patch th
On 2017-09-27, Guy Lunardi wrote:
> This issue still appears to exist against Linux 4.13.3.
And 4.13.4-2... (also tried 4.14-rc5, but that had other issues which
prevented me from testing it).
> Have not looked at haveged at all but noticed that passing "--data=16"
> instead on relying on the fa
This issue still appears to exist against Linux 4.13.3.
Have not looked at haveged at all but noticed that passing "--data=16"
instead on relying on the fallback seems to allow haveged to start properly.
Tested with armbian on a LibreComputer Le Potato.
HTH (I might try to look into haveged if I
Vagrant,
Sorry for the tardy reply - health issues the last couple of months.
Your strace text seems to indicate a memory allocation failure. The
first thing haveged does after printing "Starting up" is allocate the
collection buffer.
I haven't done much with haveged in a long time, don't th
Control: forwarded 866306 g...@issiweb.com
Hello, forwarding an issue from the Debian bug tracking system regarding
haveged on a particular arm board (odroid-c2) that fails on linux 4.12,
but works on linux 4.11... haven't yet tested with linux 4.13-rc*.
The full bug history is at:
https://bug
Mattia Rizzolo:
> Well, I wanted to try this, but when I went to just try a simple restart
> of hanvengd, it just started. Now all the amd64 nodes are running
> havengd seemingly without any issue
Interesting. So I suspect haveged starts before *something* it needs
at boot time. Sorry, no tim
On Thu, Jun 29, 2017 at 05:53:32PM +, Holger Levsen wrote:
> On Thu, Jun 29, 2017 at 07:49:31PM +0200, intrigeri wrote:
> > For the "disabling all such hardening" part:
> >
> > # systemctl edit haveged.service
> >
> > [Service]
> > SecureBits=
> > CapabilityBoundingSet=
> > NoNewPrivileges=no
control: clone -1 -2
control: retitle -2 haveged problems on some amd64 nodes
On Thu, Jun 29, 2017 at 07:49:31PM +0200, intrigeri wrote:
> For the "disabling all such hardening" part:
>
> # systemctl edit haveged.service
>
> [Service]
> SecureBits=
> CapabilityBoundingSet=
> NoNewPrivileges=no
>
Hi,
Holger Levsen:
> On Thu, Jun 29, 2017 at 02:05:08PM +0200, intrigeri wrote:
>> Holger Levsen:
>> > Process: 375 ExecStart=/usr/sbin/haveged --Foreground --verbose=1
>> > $DAEMON_ARGS (code=exited, status=226/NAMESPACE)
>>
>> 226/NAMESPACE suggests it's a problem with the hardening we do in
On 2017-06-29, intrigeri wrote:
> Vagrant Cascadian:
>> I've been trying to set up an Odroid-C2 for the Debian reproducible
>> builds test infrastructure, but haveged fails to start on this
>> particular machine with linux kernel 4.12. I've tried running the same
>> 4.12 kernels on other boards (pi
On Thu, Jun 29, 2017 at 02:05:08PM +0200, intrigeri wrote:
> Holger Levsen:
> > Process: 375 ExecStart=/usr/sbin/haveged --Foreground --verbose=1
> > $DAEMON_ARGS (code=exited, status=226/NAMESPACE)
>
> 226/NAMESPACE suggests it's a problem with the hardening we do in the
> systemd unit file. P
Holger Levsen:
> Process: 375 ExecStart=/usr/sbin/haveged --Foreground --verbose=1
> $DAEMON_ARGS (code=exited, status=226/NAMESPACE)
226/NAMESPACE suggests it's a problem with the hardening we do in the
systemd unit file. Please try disabling all such hardening, and if it
fixes stuff, re-enabl
Hi,
FWIW, we're seeing this now on our amd64 systems running stretch too:
eg
https://jenkins.debian.net/view/reproducible/view/Node_maintenance/job/reproducible_node_health_check_amd64_profitbricks1/
shows
● haveged.service - Entropy daemon using the HAVEGE algorithm
Loaded: loaded (/lib/sy
Hi Vagrant,
Vagrant Cascadian:
> I've been trying to set up an Odroid-C2 for the Debian reproducible
> builds test infrastructure, but haveged fails to start on this
> particular machine with linux kernel 4.12. I've tried running the same
> 4.12 kernels on other boards (pine64+) and haveged works
Package: haveged
Version: 1.9.1-5
Severity: normal
X-Debbugs-Cc: debian-...@lists.debian.org
I've been trying to set up an Odroid-C2 for the Debian reproducible
builds test infrastructure, but haveged fails to start on this
particular machine with linux kernel 4.12. I've tried running the same
4.1
21 matches
Mail list logo
