Re: [PATCH v2 2/2] overcommit: introduce mem-lock=on-fault

2024-12-11 Thread Daniil Tatianin
On 12/12/24 12:52 AM, Peter Xu wrote: On Wed, Dec 11, 2024 at 03:04:47AM +0300, Daniil Tatianin wrote: Locking the memory without MCL_ONFAULT instantly prefaults any mmaped anonymous memory with a write-fault, which introduces a lot of extra overhead in terms of memory usage when all you want t

Re: [PATCH v2 2/2] overcommit: introduce mem-lock=on-fault

2024-12-11 Thread Peter Xu
On Wed, Dec 11, 2024 at 03:04:47AM +0300, Daniil Tatianin wrote: > Locking the memory without MCL_ONFAULT instantly prefaults any mmaped > anonymous memory with a write-fault, which introduces a lot of extra > overhead in terms of memory usage when all you want to do is to prevent > kcompactd from

[PATCH v2 2/2] overcommit: introduce mem-lock=on-fault

2024-12-10 Thread Daniil Tatianin
Locking the memory without MCL_ONFAULT instantly prefaults any mmaped anonymous memory with a write-fault, which introduces a lot of extra overhead in terms of memory usage when all you want to do is to prevent kcompactd from migrating and compacting QEMU pages. Add an option to only lock pages laz