On 04/01/11 17:27, Christoph Egger wrote: > On 04/01/11 08:35, Lars Heidieker wrote: >> Hi, >> >> this is a part of the changes to the kernel memory management. >> It's a changing the subr_extent to use kmem(9) instead of >> malloc(9) essentially removing the MALLOC_TYPE from it. >> >> The next steps will provide the changes to the kmem(9)/pool(9) >> and the pool/uvm_km interface. The uvm_km will gain caches for >> multiplies of the page-size based on pool extending the current >> 1-page-sized cache. >> >> Making a patch for the pool(9) uvm_km interface changes will be >> next followed by the kmem(9)/malloc(9) merger. After that some >> changes that remove malloc(9) use from sys/uvm/* and large parts >> of sys/kern/* will follow. >> >> After that a big round of mechanical changes will follow that >> change the (then based on kmem) mallocs function signature by >> removing the MALLOC_TYPE completely or migrating the allocation >> to kmem(9). >> >> I made some changes to the initialization order in uvm_init, the >> idea behind this is to have the pool(9) allocators initialization >> at explicit stages, this will make initialization of a standard >> pool-allocators of different pool-page-sizes possible and will >> catch use before ready cases. (This part needs further testing as >> some changes in pmaps where necessary, I've compile tested all >> the pmaps involved and boot-tested i386/amd64/sparc64) > > Hi, > > this all sounds good to me. The Xen port requires kmem(9) to return > page aligned addresses to work correctly. kmem(9) currently doesn't > guarantee this. I just want to bring this to your awareness. > > Christoph > Hi,
the kmem implementation I'm working on has page aligned allocations for page size and larger allocations. Lars -- ------------------------------------ Mystische Erklärungen: Die mystischen Erklärungen gelten für tief; die Wahrheit ist, dass sie noch nicht einmal oberflächlich sind. -- Friedrich Nietzsche [ Die Fröhliche Wissenschaft Buch 3, 126 ]
