On 21 April 2012 06:04, Stas Sergeev wrote: > It is better to have open_mapping() to open /dev/mem, > alloc_mapping_kmem() to map it somewhere, and mmap_mapping_kmem() > to do the flip - that's what I did in my patchset. Why don't > you like that approach?
Hmm, I don't see open_mapping changes in your patches. Maybe I missed it. In any case, I think I found a simpler approach (attached) for kmem, by first observing that alloc_mapping(MAPPING_KMEM...) was always immediately followed by a call to mmap(). The only place where they are not paired (just mmap_mapping there) is map_video_ram() in vc.c. So they can just be combined; and the call in map_video_ram() can still do the right thing (skipping the actual mmap to /dev/mem, and just doing mremap) by looking at kmem_map. In the end that changes mmap_mapping for KMEM to do what it should do conceptually, map an area of /dev/mem. What goes on behind the scenes (the remapping) is just to avoid reopening /dev/mem because root was dropped but the caller doesn't need to know these details. Bart
0001-Fold-alloc_mapping-into-mmap_mapping-for-MAPPING_KME.patch
Description: Binary data
------------------------------------------------------------------------------ For Developers, A Lot Can Happen In A Second. Boundary is the first to Know...and Tell You. Monitor Your Applications in Ultra-Fine Resolution. Try it FREE! http://p.sf.net/sfu/Boundary-d2dvs2
_______________________________________________ Dosemu-devel mailing list Dosemu-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dosemu-devel
