c/r failure

Oren Laadan orenl at cs.columbia.edu
Sun Apr 24 02:33:30 PDT 2011


Serge,

Do you still get this problem with rc2 ?

Oren.

On 03/23/2011 10:11 PM, Matt Helsley wrote:
> On Wed, Mar 23, 2011 at 08:34:52PM -0500, Serge E. Hallyn wrote:
>> Quoting Matt Helsley (matthltc at us.ibm.com):
>>> 	Unfortunately ckptinfo is broken. My most recent series of 18 patches
>>> to user-cr fixes it. I've attached the output using an earlier version
>>
>> Egads.
> 
> Incidentally, those patches don't change checkpoint or restart -- just
> the build and ckptinfo. So unless you want to patch your own ckptinfo
> or have some quirky old "boxes" like me, you don't need them.
> 
>>
>>> beginning. This was masked by some troubles with ckpt_err() which I still
>>> haven't been able to figure out -- the "error" reported in dmesg above
>>> is -512 but really should be -32 (EPIPE). You rewrote that code, didn't you? ;)
>>
>> Eeeeyeee know nothink.  Nothink!
>>
>>> Anyhow, today I managed to bisect my user-cr troubles down to:
>>>
>>> 5b97422c4c1342a128df508cda7c4639ecb24a36
>>>
>>> Revert was not clean and the resolved conflicts on top of the revert did
>>> not seem to fix the problem, soo... :/ all I can recommend at the moment is
>>> sticking to the earlier branches.
>>
>>
>>> Hope that helps. Again, sorry for the delay.
>>
>> Thanks, Matt.  It'll be at least a week before I have a chance to play with
>> it again, but I'll give earlier versions a shot.  Is there any particular
>> matching version for kernel+user-dr which you'd recomment at this point?
> 
> So far the closest thing I have to a working combination is this commit
> found via bisect:
> 
> v23-rc1 + user-cr @ f1d427d4ce1df44432fccd2e08c60992c8169b8c
> "Ghost tasks must be detached"
> 
> (which just follows: f0fdd38cb0a53594eec9d4e19671d0c33036517c
> "headers: regenerate for ckpt-v23-rc1 [x86_32]" -- nice to know it's
> about the right spot)
> 
> Watch out for non-matching syscall numbers. The clone_*.* source files
> will use old "default" values _very_ _quietly_ if they aren't found via
> the updated headers in ./include/linux . Syscall numbers in those files
> tend to get updated much less frequently.
> 
> I am testing on x86-32 in KVM at the moment and haven't run everything
> so YMMV.
> 
> Cheers,
> 	-Matt Helsley
> 


More information about the Containers mailing list