On Thu, Jun 24, 2010 at 5:23 PM, Bart Smaalders <bart.smaald...@oracle.com>wrote:
> On 06/24/10 16:57, Anup pemmaiah wrote: > >> >> We are planning to use first 1GB for solaris and the remaining memory is >> managed by a >> kernel module to support some code that we are porting from another OS. >> > > I don't think this won't work the way you want. > > Do you need the remaining memory to be physically contiguous? > Yes. > > When you talk about a kernel module, I assume you mean a Solaris kernel > module.... Yes, Solaris kernel module > if you set physmem, you're not going to be able to use any > Solaris calls to manage memory above that limit. Yes, will not be able to use Solaris calls to directly manage the physical mem above the limit. One of the way that we are doing is, the kernel module that manages the reserved mem above 1GB will map the physical addr to the kernel virtual address using the hat_kpm* calls. This module will be the interface between Solaris calls and the phys mem. > Why not do this > using Solaris facilities instead of trying to work around them? > We are trying to leverage Solaris facilities as much as possible. In this particular case, since there is lot of dependency on the module that manages the reserved memory, decided for the first phase to go the above route. Thanks Anup > > - Bart > > -- > Bart Smaalders Solaris Kernel Performance > bart.smaald...@oracle.com http://blogs.sun.com/barts > "You will contribute more with mercurial than with thunderbird." > _______________________________________________ > opensolaris-code mailing list > opensolaris-code@opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/opensolaris-code >
_______________________________________________ opensolaris-code mailing list opensolaris-code@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
