Resuming from a crashdump

João Carlos Mendes Luís jonny at jonny.eng.br
Tue Jan 25 13:07:18 PST 2005


Matthew Dillon wrote:
>     Well, I don't want do disuade you from trying, but I think you are
>     seriously underestimating the effort required to restore device state.
> 
>     You basically would either have to make all device drivers support a new 
>     hibernation/restore API (because it is not really possible to restore

     Shouldn't it be very similar to the suspend API?

>     a device driver based on a dump), or you would need to implement some
>     higher level utility code (e.g. scripts and such) to try to record and
>     restore the state at a higher level, such as for network interfaces,
>     and not allow any restored processes to run until that's done.  Either
>     way it would get messy very quickly.
> 
>     Also, if the machine has a lot of memory it could take longer to save
>     and restore then to reboot from scratch.  A typical laptop HD is 

     Indeed true, but when I use windows hibernate I don't intend to 
simply bypass boot procedures, I want to keep the workspace active.  If 
the workspace is not important, I simply shutdown, as it is also a 
profilatic memory cleaning.   ;-)

>     I think it would probably be more realistic to persue a process 
>     save/restore rather then a kernel save/restore.  The overhead is going
>     to be the disk I/O anyway and that seems to be about the same either
>     way (maybe less for a process restore), plus you can at least demand-load
>     the process restore.

     Let's suppose I want to keep my X desktop intact.  Should I 
save/restore processes in which order?  X server or X clients?  They 
need to interact with each other, and if one try to interact while the 
other has already "died", it will simply close and die too.  That's why 
I think a kernel save/restore is better for a full hibernation 
situation.  Of course single process hibernation is also useful!  I am a 
very old user, from the times where LaTeX had to core dump after 
initializing variables to startup faster later.   ;-)

                                         Jonny

-- 
João Carlos Mendes Luís - Networking Engineer - jonny at jonny.eng.br


More information about the freebsd-hackers mailing list