On Wed, 2 Jul 2025 15:15:01 +0100 Usama Arif [off-list ref] wrote:

[...]

In terms of the approach of doing this, IMHO, I dont think the way to do this
is controversial. After the great feedback from Lorenzo on the prctl series, the
approach would be for userpsace to make a call that just does for_each_vma of the process,
madvises the VMAs,

One dirty hack that I can think off the top of my head for doing this without
new kernel changes is, unsurprisingly, using DAMOS.  Using DAMOS, users can do
madvise(MADV_HUGEPAGE) to virtual address ranges of specific access patterns.
It is aimed to be used for hot regions, while using similar one of
MADV_NOHUGEPAGE for cold regions.  An experiment with a prototype[1] showed it
eliminates about 80% of internal fragmentation caused memory overhead while
keeping 46% of performance improvement under a constrained situation.

If you set the access pattern as any pattern, hence, you can do
madvise(MADV_HUGEPAGE) for effectively entire virtual address space of the
process.  DAMON user-space tool supports periodically tracking childs and
applying same DAMOS scheme to those.  So, for example, below hack could be
tried.

    # damo start $(pidof XXX) --damos_action hugepage --include_child_tasks

I'm working with Usama at Meta but not very closely involved in THP works, so
I'm not sure if this works for Usama's case and others.  I even not tried this
at all on any test environment.  So I'm not recommending this but just sharing
a thought for more brainsorming, and that's why I call this a dirty hack.

[1] https://assets.amazon.science/b7/2b/ce53222247739b174f2b54498d1a/daos-data-access-aware-operating-system.pdf


Thanks,
SJ

[...]