Re: [PATCH v2 4/5] mm: make vma cache SLAB_TYPESAFE_BY_RCU

[Vlastimil Babka <vbabka@suse.cz>]

On 11/14/24 17:18, Suren Baghdasaryan wrote:
> 
> Ok, my vma->detached deprecation effort got to the point that my QEMU
> boots. The change is not pretty and I'm quite sure I did not cover all
> cases yet (like hugepages):
> 
>  arch/arm/kernel/process.c             |   2 +-
>  arch/arm64/kernel/vdso.c              |   4 +-
>  arch/loongarch/kernel/vdso.c          |   2 +-
>  arch/powerpc/kernel/vdso.c            |   2 +-
>  arch/powerpc/platforms/pseries/vas.c  |   2 +-
>  arch/riscv/kernel/vdso.c              |   4 +-
>  arch/s390/kernel/vdso.c               |   2 +-
>  arch/s390/mm/gmap.c                   |   2 +-
>  arch/x86/entry/vdso/vma.c             |   2 +-
>  arch/x86/kernel/cpu/sgx/encl.c        |   2 +-
>  arch/x86/um/mem_32.c                  |   2 +-
>  drivers/android/binder_alloc.c        |   2 +-
>  drivers/gpu/drm/i915/i915_mm.c        |   4 +-
>  drivers/infiniband/core/uverbs_main.c |   4 +-
>  drivers/misc/sgi-gru/grumain.c        |   2 +-
>  fs/exec.c                             |   2 +-
>  fs/hugetlbfs/inode.c                  |   3 +-
>  include/linux/mm.h                    | 111 +++++++++++++++++---------
>  include/linux/mm_types.h              |   6 --
>  kernel/bpf/arena.c                    |   2 +-
>  kernel/fork.c                         |   6 +-
>  mm/debug_vm_pgtable.c                 |   2 +-
>  mm/internal.h                         |   2 +-
>  mm/madvise.c                          |   4 +-
>  mm/memory.c                           |  39 ++++-----
>  mm/mmap.c                             |   9 +--
>  mm/nommu.c                            |   6 +-
>  mm/oom_kill.c                         |   2 +-
>  mm/vma.c                              |  62 +++++++-------
>  mm/vma.h                              |   2 +-
>  net/ipv4/tcp.c                        |   4 +-
>  31 files changed, 164 insertions(+), 136 deletions(-)
> 
> Many of the unmap_* and zap_* functions should get an `mm` parameter
> to make this work.

Ouch, thanks for the attempt!

> So, if we take this route, it should definitely be a separate patch,
> which will likely cause some instability issues for some time until
> all the edge cases are ironed out. I would like to proceed with this
> patch series first before attempting to deprecate vma->detached. Let
> me know if you have objections to this plan.

No objections!

>>
>> >
>> > >
>> > > >
>> > > > >
>> > > > > Either way, we should make it a function so it's easier to reuse for
>> > > > > whatever we need in the future, wdyt?
>> > > > >
>> > > > > To unsubscribe from this group and stop receiving emails from it, send an email to kernel-team+unsubscribe@android.com.
>> > > > >