RFC: backporting GEOM to the 4.x branch

Matthew Dillon dillon at apollo.backplane.com
Thu Feb 3 19:08:19 PST 2005


:Actually, the main reason I want to have GEOM backported is GBDE. Perhaps I
:should just deGEOMify GEOM Based Disk Encryption. I just checked and GBDE
:consists of about 2000 lines of code. Would you be interested in integrating
:such deGEOMified Disk Encryption into DragonFly BSD?
:
:As far as BUF/BIO model changes are concerned, I understand that the I/O subsystem
:was supposed to be moved from KVA mapped buffers to page lists (using XIO to
:manage the page lists). Have any other significant changes been made to the FreeBSD
:BUF/BIO model that I should be aware of to be able to integrate deGEOMified Disk
:Encryption (dGDE) into DragonFly BSD?
:
:ALeine

    Sure, but I have a wish list a mile long and probably a thousand man-hours
    worth of work in my queue.  I'm not going to be porting that any time 
    soon.  You are certainly welcome to port it as a char device.  Our 
    disk abstraction is already fully recursive so it should be possible to
    stack any number of block devices on top of each other.  It would be nice
    to have a common utility to glue block devices together for this class
    of 'virtual' block device as well.

						-Matt





More information about the Kernel mailing list