On Tue, Nov 12, 2024 at 10:15:44AM -0500, Liam R. Howlett wrote: > * Lorenzo Stoakes <lorenzo.stoakes@xxxxxxxxxx> [241108 08:57]: > > +VMA locks are special in that you must obtain an mmap **write** lock **first** > > +in order to obtain a VMA **write** lock. A VMA **read** lock however can be > > +obtained without any other lock (:c:func:`!lock_vma_under_rcu` will acquire then > > +release an RCU lock to lookup the VMA for you). > > This reduces the impact of a writer on readers by only impacting > conflicting areas of the vma tree. Sorry I totally missed this last time round, will add a note. > > + ================================= ===================== ======================================== =============== > > + Field Configuration option Description Write lock > > + ================================= ===================== ======================================== =============== > > + :c:member:`!anon_name` CONFIG_ANON_VMA_NAME A field for storing a mmap write, > > + :c:struct:`!struct anon_vma_name` VMA write. > > + object providing a name for anonymous > > + mappings, or :c:macro:`!NULL` if none > > + is set or the VMA is file-backed. > > These are ref counted and can be shared by more than one vma for > scalability. > > > + :c:member:`!swap_readahead_info` CONFIG_SWAP Metadata used by the swap mechanism mmap read, > > + to perform readahead. This field is swap-specific > > + accessed atomically. lock. > > + :c:member:`!vm_policy` CONFIG_NUMA :c:type:`!mempolicy` object which mmap write, > > + describes the NUMA behaviour of the VMA write. > > + VMA. > > These are also ref counted for scalability. I think actually on second thoughts I will add notes about this in v3, as it at least gives a bit more context.
