<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Tue, Mar 15, 2016 at 1:58 AM, Pavel Emelyanov <span dir="ltr"><<a href="mailto:xemul@virtuozzo.com" target="_blank">xemul@virtuozzo.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On 03/15/2016 02:33 AM, Filipe Brandenburger wrote:<br>
> Hi,<br>
><br>
> On Mon, Mar 14, 2016 at 4:14 PM, Saied Kazemi <<a href="mailto:saied@google.com">saied@google.com</a>> wrote:<br>
>> Yes, there's definitely a bug. Filipe was suggesting that perhaps the<br>
>> prctl() calls are failing not due to lack of support in the kernel but due<br>
>> to some other error (e.g., passing wrong parameters).<br>
><br>
> Yes. In fact Pavel included this change of brk -> sbrk which fixes<br>
> this issue on that specific line that was failing for us...<br>
><br>
>> This causes "criu check" to think the old API is not supported but in reality<br>
>> it is and that's why c/r works.<br>
>><br>
>> Below is a complete log showing all your patches applied. Hope this helps!<br>
><br>
>> $ sudo ./criu/criu check --ms<br>
>> Warn (cr-check.c:195): Skipping unssuported PR_SET_MM_MAP<br>
<br>
</span>Here it is -- the prctl checks all pass :) The warning here should not be a<br>
warning, I would say, it just says, that we don't care that SET_MM_MAP doesn't<br>
work and go ahead to check for older prctls.<br>
<span class=""><br>
>> Error (cr-check.c:629): Kernel doesn't support PTRACE_O_SUSPEND_SECCOMP<br>
<br>
</span>This is seccomp suspend API that should go to category 2 (optional)<br>
<span class=""><br>
>> Error (cr-check.c:673): Dumping seccomp filters not supported: Input/output<br>
>> error<br>
<br>
</span>Same here<br>
<span class=""><br>
>> Error (timerfd.c:55): timerfd: No timerfd support for c/r: Inappropriate<br>
>> ioctl for device<br>
<br>
</span>And here<br>
<span class=""><br>
>> Error (cr-check.c:307): fdinfo doesn't contain the mnt_id field<br>
<br>
</span>And actually here, though todays containers _typicaly_ create more than one mntns :(<br>
However, this doesn't block C/R of simple ones, so I'd still put this into category 2.<br>
<span class=""><br>
>> Warn (cr-check.c:773): Skipping unsupported AIO remap<br>
<br>
</span>Same here<br>
<span class=""><br>
>> Warn (cr-check.c:789): fdinfo doesn't contain the lock field<br>
<br>
</span>Ditto<br>
<span class=""><br>
>> Warn (cr-check.c:827): Skipping cgroup namespaces check<br>
<br>
</span>And that's it.<br>
<br>
So, from what I see your previous patch that splits all the checks into two cats,<br>
plus Filipe's fix (merged) plus my fix for brk/sbrk plus keeping all the prctls<br>
checks in category 1 should make us good criu check action.<br></blockquote><div><br></div><div>Yes, it's finally working the way we'd like it on my two systems. Will rebase on top of criu-dev and send a patch soon.</div><div><br></div><div>--Saied</div><div><br></div></div></div></div>