dcn@xxxxxxx (Dean Nelson) wrote: > > > The following patch modifies the gen_pool allocator (lib/genalloc.c) to > utilize a bitmap scheme instead of the buddy scheme. The purpose of this > change is to eliminate the touching of the actual memory being allocated. > > Since the change modifies the interface, a change to the uncached > allocator (arch/ia64/kernel/uncached.c) is also required. > > Signed-off-by: Dean Nelson <dcn@xxxxxxx> > > --- > > Andrew, > > Both Andrey Volkov and Jes Sorenson have expressed a desire that the > gen_pool allocator not write to the memory being managed. See the > following: > > http://marc.theaimsgroup.com/?l=linux-kernel&m=113518602713125&w=2 > http://marc.theaimsgroup.com/?l=linux-kernel&m=113533568827916&w=2 hm, fair enough. The patch is fairly large+intrusive. I trust it's been broadly tested? > -unsigned long gen_pool_alloc(struct gen_pool *poolp, int size) > +int gen_pool_add(struct gen_pool *pool, unsigned long addr, size_t size, > + int nid) > { > - int j, i, s, max_chunk_size; > - unsigned long a, flags; > - struct gen_pool_link *h = poolp->h; > + struct gen_pool_chunk *chunk; > + int nbits = size >> pool->min_alloc_order; > + int nbytes = sizeof(struct gen_pool_chunk) + nbits / 8; We can use BITS_PER_BYTE rather than "8". I have a suspicion that `nbytes' here needs to be rounded up. > + if (nbytes > PAGE_SIZE) { > + if (nid == GENALLOC_NID_NONE) > + chunk = vmalloc(nbytes); > + else > + chunk = vmalloc_node(nbytes, nid); > + } else { > + if (nid == GENALLOC_NID_NONE) > + chunk = kmalloc(nbytes, GFP_KERNEL); > + else > + chunk = kmalloc_node(nbytes, GFP_KERNEL, nid); > + } I don't think GENALLOC_NID_NONE needs to exist. If the caller passes in `nid=-1', kmalloc_node() will do what you want. Which is an apparently-undocumented feature... - : send the line "unsubscribe linux-ia64" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
![]() |