linux.git/tools/testing/vma, branch master

tools/testing/vma: add VMA userland tests for VMA flag functions

2026-02-12T23:42:59+00:00

Now we have the capability to test the new helpers for the bitmap VMA
flags in userland, do so.

We also update the Makefile such that both VMA (and while we're here)
mm_struct flag sizes can be customised on build.  We default to 128-bit to
enable testing of flags above word size even on 64-bit systems.

We add userland tests to ensure that we do not regress VMA flag behaviour
with the introduction when using bitmap VMA flags, nor accidentally
introduce unexpected results due to for instance higher bit values not
being correctly cleared/set.

As part of this change, make __mk_vma_flags() a custom function so we can
handle specifying invalid VMA bits.  This is purposeful so we can have the
VMA tests work at lower and higher number of VMA flags without having to
duplicate code too much.

Link: https://lkml.kernel.org/r/7fe6afe9c8c61e4d3cfc9a2d50a5d24da8528e68.1769097829.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes 
Reviewed-by: Liam R. Howlett 
Cc: Baolin Wang 
Cc: Barry Song 
Cc: David Hildenbrand 
Cc: Dev Jain 
Cc: Jason Gunthorpe 
Cc: Suren Baghdasaryan 
Cc: Vlastimil Babka 
Cc: Zi Yan 
Cc: Damien Le Moal 
Cc: "Darrick J. Wong" 
Cc: Jarkko Sakkinen 
Cc: Yury Norov 
Cc: Chris Mason 
Cc: Pedro Falcato 
Signed-off-by: Andrew Morton

tools/testing/vma: separate out vma_internal.h into logical headers

2026-02-12T23:42:59+00:00

The vma_internal.h file is becoming entirely unmanageable.  It combines
duplicated kernel implementation logic that needs to be kept in-sync with
the kernel, stubbed out declarations that we simply ignore for testing
purposes and custom logic added to aid testing.

If we separate each of the three things into separate headers it makes
things far more manageable, so do so:

* include/stubs.h  contains the stubbed declarations,
* include/dup.h    contains the duplicated kernel declarations, and
* include/custom.h contains declarations customised for testing.

[lorenzo.stoakes@oracle.com: avoid a duplicate struct define]
  Link: https://lkml.kernel.org/r/1e032732-61c3-485c-9aa7-6a09016fefc1@lucifer.local
Link: https://lkml.kernel.org/r/dd57baf5b5986cb96a167150ac712cbe804b63ee.1769097829.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes 
Reviewed-by: Liam R. Howlett 
Cc: Baolin Wang 
Cc: Barry Song 
Cc: David Hildenbrand 
Cc: Dev Jain 
Cc: Jason Gunthorpe 
Cc: Suren Baghdasaryan 
Cc: Vlastimil Babka 
Cc: Zi Yan 
Cc: Damien Le Moal 
Cc: "Darrick J. Wong" 
Cc: Jarkko Sakkinen 
Cc: Yury Norov 
Cc: Chris Mason 
Cc: Pedro Falcato 
Signed-off-by: Andrew Morton

tools/testing/vma: separate VMA userland tests into separate files

2026-02-12T23:42:59+00:00

So far the userland VMA tests have been established as a rough expression
of what's been possible.

Adapt it into a more usable form by separating out tests and shared
helper functions.

Since we test functions that are declared statically in mm/vma.c, we make
use of the trick of #include'ing kernel C files directly.

In order for the tests to continue to function, we must therefore also
this way into the tests/ directory.

We try to keep as much shared logic actually modularised into a separate
compilation unit in shared.c, however the merge_existing() and
attach_vma() helpers rely on statically declared mm/vma.c functions so
these must be declared in main.c.

Link: https://lkml.kernel.org/r/a0455ccfe4fdcd1c962c64f76304f612e5662a4e.1769097829.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes 
Reviewed-by: Liam R. Howlett 
Cc: Baolin Wang 
Cc: Barry Song 
Cc: David Hildenbrand 
Cc: Dev Jain 
Cc: Jason Gunthorpe 
Cc: Suren Baghdasaryan 
Cc: Vlastimil Babka 
Cc: Zi Yan 
Cc: Damien Le Moal 
Cc: "Darrick J. Wong" 
Cc: Jarkko Sakkinen 
Cc: Yury Norov 
Cc: Chris Mason 
Cc: Pedro Falcato 
Signed-off-by: Andrew Morton

mm: make vm_area_desc utilise vma_flags_t only

2026-02-12T23:42:59+00:00

Now we have eliminated all uses of vm_area_desc->vm_flags, eliminate this
field, and have mmap_prepare users utilise the vma_flags_t
vm_area_desc->vma_flags field only.

As part of this change we alter is_shared_maywrite() to accept a
vma_flags_t parameter, and introduce is_shared_maywrite_vm_flags() for use
with legacy vm_flags_t flags.

We also update struct mmap_state to add a union between vma_flags and
vm_flags temporarily until the mmap logic is also converted to using
vma_flags_t.

Also update the VMA userland tests to reflect this change.

Link: https://lkml.kernel.org/r/fd2a2938b246b4505321954062b1caba7acfc77a.1769097829.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes 
Reviewed-by: Pedro Falcato 
Reviewed-by: Liam R. Howlett 
Cc: Baolin Wang 
Cc: Barry Song 
Cc: David Hildenbrand 
Cc: Dev Jain 
Cc: Jason Gunthorpe 
Cc: Suren Baghdasaryan 
Cc: Vlastimil Babka 
Cc: Zi Yan 
Cc: Damien Le Moal 
Cc: "Darrick J. Wong" 
Cc: Jarkko Sakkinen 
Cc: Yury Norov 
Cc: Chris Mason 
Signed-off-by: Andrew Morton

mm: add basic VMA flag operation helper functions

2026-02-12T23:42:57+00:00

Now we have the mk_vma_flags() macro helper which permits easy
specification of any number of VMA flags, add helper functions which
operate with vma_flags_t parameters.

This patch provides vma_flags_test[_mask](), vma_flags_set[_mask]() and
vma_flags_clear[_mask]() respectively testing, setting and clearing flags
with the _mask variants accepting vma_flag_t parameters, and the non-mask
variants implemented as macros which accept a list of flags.

This allows us to trivially test/set/clear aggregate VMA flag values as
necessary, for instance:

	if (vma_flags_test(&flags, VMA_READ_BIT, VMA_WRITE_BIT))
		goto readwrite;

	vma_flags_set(&flags, VMA_READ_BIT, VMA_WRITE_BIT);

	vma_flags_clear(&flags, VMA_READ_BIT, VMA_WRITE_BIT);

We also add a function for testing that ALL flags are set for convenience,
e.g.:

	if (vma_flags_test_all(&flags, VMA_READ_BIT, VMA_MAYREAD_BIT)) {
		/* Both READ and MAYREAD flags set */
		...
	}

The compiler generates optimal assembly for each such that they behave as
if the caller were setting the bitmap flags manually.

This is important for e.g.  drivers which manipulate flag values rather
than a VMA's specific flag values.

We also add helpers for testing, setting and clearing flags for VMA's and
VMA descriptors to reduce boilerplate.

Also add the EMPTY_VMA_FLAGS define to aid initialisation of empty flags.

Finally, update the userland VMA tests to add the helpers there so they
can be utilised as part of userland testing.

Link: https://lkml.kernel.org/r/885d4897d67a6a57c0b07fa182a7055ad752df11.1769097829.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes 
Reviewed-by: Pedro Falcato 
Reviewed-by: Liam R. Howlett 
Cc: Baolin Wang 
Cc: Barry Song 
Cc: David Hildenbrand 
Cc: Dev Jain 
Cc: Jason Gunthorpe 
Cc: Suren Baghdasaryan 
Cc: Vlastimil Babka 
Cc: Zi Yan 
Cc: Damien Le Moal 
Cc: "Darrick J. Wong" 
Cc: Jarkko Sakkinen 
Cc: Yury Norov 
Cc: Chris Mason 
Signed-off-by: Andrew Morton

mm: use unmap_desc struct for freeing page tables

2026-02-12T23:42:56+00:00

Pass through the unmap_desc to free_pgtables() because it almost has
everything necessary and is already on the stack.

Updates testing code as necessary.

No functional changes intended.

[Liam.Howlett@oracle.com: fix up unmap desc use on exit_mmap()]
  Link: https://lkml.kernel.org/r/20260210214214.364856-1-Liam.Howlett@oracle.com
Link: https://lkml.kernel.org/r/20260121164946.2093480-12-Liam.Howlett@oracle.com
Signed-off-by: Liam R. Howlett 
Reviewed-by: Lorenzo Stoakes 
Cc: Suren Baghdasaryan 
Cc: Baoquan He 
Cc: Barry Song 
Cc: Chris Li 
Cc: David Hildenbrand 
Cc: David Hildenbrand 
Cc: Jann Horn 
Cc: Kairui Song 
Cc: Kemeng Shi 
Cc: Matthew Wilcox (Oracle) 
Cc: Michal Hocko 
Cc: Mike Rapoport 
Cc: Nhat Pham 
Cc: Pedro Falcato 
Cc: SeongJae Park 
Cc: Vlastimil Babka 
Signed-off-by: Andrew Morton

mm/vma: use unmap_desc in exit_mmap() and vms_clear_ptes()

2026-02-12T23:42:55+00:00

Convert vms_clear_ptes() to use unmap_desc to call unmap_vmas() instead of
the large argument list.  The UNMAP_STATE() cannot be used because the vma
iterator in the vms does not point to the correct maple state
(mas_detach), and the tree_end will be set incorrectly.  Setting up the
arguments manually avoids setting the struct up incorrectly and doing
extra work to get the correct pagetable range.

exit_mmap() also calls unmap_vmas() with many arguments.  Using the
unmap_all_init() function to set the unmap descriptor for all vmas makes
this a bit easier to read.

Update to the vma test code is necessary to ensure testing continues to
function.

No functional changes intended.

Link: https://lkml.kernel.org/r/20260121164946.2093480-10-Liam.Howlett@oracle.com
Signed-off-by: Liam R. Howlett 
Cc: Baoquan He 
Cc: Barry Song 
Cc: Chris Li 
Cc: David Hildenbrand 
Cc: David Hildenbrand 
Cc: Jann Horn 
Cc: Kairui Song 
Cc: Kemeng Shi 
Cc: Lorenzo Stoakes 
Cc: Matthew Wilcox (Oracle) 
Cc: Michal Hocko 
Cc: Mike Rapoport 
Cc: Nhat Pham 
Cc: Pedro Falcato 
Cc: SeongJae Park 
Cc: Suren Baghdasaryan 
Cc: Vlastimil Babka 
Signed-off-by: Andrew Morton

mm/memory: add tree limit to free_pgtables()

2026-02-12T23:42:54+00:00

The ceiling and tree search limit need to be different arguments for the
future change in the failed fork attempt.  The ceiling and floor variables
are not very descriptive, so change them to pg_start/pg_end.

Adding a new variable for the vma_end to the function as it will differ
from the pg_end in the later patches in the series.

Add a kernel doc about the free_pgtables() function.

Test code also updated.

No functional changes intended.

Link: https://lkml.kernel.org/r/20260121164946.2093480-6-Liam.Howlett@oracle.com
Signed-off-by: Liam R. Howlett 
Reviewed-by: Lorenzo Stoakes 
Reviewed-by: Pedro Falcato 
Cc: Baoquan He 
Cc: Barry Song 
Cc: Chris Li 
Cc: David Hildenbrand 
Cc: David Hildenbrand 
Cc: Jann Horn 
Cc: Kairui Song 
Cc: Kemeng Shi 
Cc: Matthew Wilcox (Oracle) 
Cc: Michal Hocko 
Cc: Mike Rapoport 
Cc: Nhat Pham 
Cc: SeongJae Park 
Cc: Suren Baghdasaryan 
Cc: Vlastimil Babka 
Signed-off-by: Andrew Morton

mm/rmap: separate out fork-only logic on anon_vma_clone()

2026-01-27T04:02:22+00:00

Specify which operation is being performed to anon_vma_clone(), which
allows us to do checks specific to each operation type, as well as to
separate out and make clear that the anon_vma reuse logic is absolutely
specific to fork only.

This opens the door to further refactorings and refinements later as we
have more information to work with.

Link: https://lkml.kernel.org/r/cf7da7a2d973cdc72a1b80dd9a73260519e8fa9f.1768746221.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes 
Reviewed-by: Liam R. Howlett 
Reviewed-by: Suren Baghdasaryan 
Cc: Barry Song 
Cc: Chris Li 
Cc: David Hildenbrand 
Cc: Harry Yoo 
Cc: Jann Horn 
Cc: Michal Hocko 
Cc: Mike Rapoport 
Cc: Pedro Falcato 
Cc: Rik van Riel 
Cc: Shakeel Butt 
Cc: Vlastimil Babka 
Signed-off-by: Andrew Morton

mm/rmap: remove anon_vma_merge() function

2026-01-27T04:02:21+00:00

This function is confusing, we already have the concept of anon_vma merge
to adjacent VMA's anon_vma's to increase probability of anon_vma
compatibility and therefore VMA merge (see is_mergeable_anon_vma() etc.),
as well as anon_vma reuse, along side the usual VMA merge logic.

We can remove the anon_vma check as it is redundant - a merge would not
have been permitted with removal if the anon_vma's were not the same (and
in the case of an unfaulted/faulted merge, we would have already set the
unfaulted VMA's anon_vma to vp->remove->anon_vma in dup_anon_vma()).

Avoid overloading this term when we're very simply unlinking anon_vma
state from a removed VMA upon merge.

Link: https://lkml.kernel.org/r/56bbe45e309f7af197b1c4f94a9a0c8931ff2d29.1768746221.git.lorenzo.stoakes@oracle.com
Signed-off-by: Lorenzo Stoakes 
Reviewed-by: Suren Baghdasaryan 
Reviewed-by: Liam R. Howlett 
Cc: Barry Song 
Cc: Chris Li 
Cc: David Hildenbrand 
Cc: Harry Yoo 
Cc: Jann Horn 
Cc: Michal Hocko 
Cc: Mike Rapoport 
Cc: Pedro Falcato 
Cc: Rik van Riel 
Cc: Shakeel Butt 
Cc: Vlastimil Babka 
Signed-off-by: Andrew Morton