On Thu, Jun 22, 2023 at 03:10:47PM +0200, David Hildenbrand wrote:
> Maybe ... depends also on the uffd semantics as in 3).
UFFDIO_COPY|ZEROPAGE bypasses page cache for private file mappings, afaict.
We'll still got a limit on the inode size (so we can't COPY|ZEROPAGE over
that offset of vma) but
On 21.06.23 18:55, Peter Xu wrote:
On Wed, Jun 21, 2023 at 06:17:37PM +0200, David Hildenbrand wrote:
As documented, ram_block_discard_range() guarantees two things
a) Read 0 after discarding succeeded
b) Make postcopy work by triggering a fault on next access
And if we'd simply want to drop t
On Wed, Jun 21, 2023 at 06:17:37PM +0200, David Hildenbrand wrote:
> As documented, ram_block_discard_range() guarantees two things
>
> a) Read 0 after discarding succeeded
> b) Make postcopy work by triggering a fault on next access
>
> And if we'd simply want to drop the FALLOC_FL_PUNCH_HOLE:
>
On 21.06.23 17:55, Peter Xu wrote:
On Tue, Jun 20, 2023 at 03:03:51PM +0200, David Hildenbrand wrote:
ram_block_discard_range() cannot possibly do the right thing in
MAP_PRIVATE file mappings in the general case.
To achieve the documented semantics, we also have to punch a hole into
the file, p
On Tue, Jun 20, 2023 at 03:03:51PM +0200, David Hildenbrand wrote:
> ram_block_discard_range() cannot possibly do the right thing in
> MAP_PRIVATE file mappings in the general case.
>
> To achieve the documented semantics, we also have to punch a hole into
> the file, possibly messing with other M
ram_block_discard_range() cannot possibly do the right thing in
MAP_PRIVATE file mappings in the general case.
To achieve the documented semantics, we also have to punch a hole into
the file, possibly messing with other MAP_PRIVATE/MAP_SHARED mappings
of such a file.
For example, using VM templat