Re: [PATCH v2 2/7] mm: introduce local state for lazy_mmu sections

[Jürgen Groß <jgross@xxxxxxxx>]

On 08.09.25 09:39, Kevin Brodsky wrote:
> arch_{enter,leave}_lazy_mmu_mode() currently have a stateless API
> (taking and returning no value). This is proving problematic in
> situations where leave() needs to restore some context back to its
> original state (before enter() was called). In particular, this
> makes it difficult to support the nesting of lazy_mmu sections -
> leave() does not know whether the matching enter() call occurred
> while lazy_mmu was already enabled, and whether to disable it or
> not.
>
> This patch gives all architectures the chance to store local state
> while inside a lazy_mmu section by making enter() return some value,
> storing it in a local variable, and having leave() take that value.
> That value is typed lazy_mmu_state_t - each architecture defining
> __HAVE_ARCH_ENTER_LAZY_MMU_MODE is free to define it as it sees fit.
> For now we define it as int everywhere, which is sufficient to
> support nesting.
>
> The diff is unfortunately rather large as all the API changes need
> to be done atomically. Main parts:
>
> * Changing the prototypes of arch_{enter,leave}_lazy_mmu_mode()
>    in generic and arch code, and introducing lazy_mmu_state_t.
>
> * Introducing LAZY_MMU_{DEFAULT,NESTED} for future support of
>    nesting. enter() always returns LAZY_MMU_DEFAULT for now.
>    (linux/mm_types.h is not the most natural location for defining
>    those constants, but there is no other obvious header that is
>    accessible where arch's implement the helpers.)
>
> * Changing all lazy_mmu sections to introduce a lazy_mmu_state
>    local variable, having enter() set it and leave() take it. Most of
>    these changes were generated using the following Coccinelle script:
>
> @@
> @@
> {
> + lazy_mmu_state_t lazy_mmu_state;
> ...
> - arch_enter_lazy_mmu_mode();
> + lazy_mmu_state = arch_enter_lazy_mmu_mode();
> ...
> - arch_leave_lazy_mmu_mode();
> + arch_leave_lazy_mmu_mode(lazy_mmu_state);
> ...
> }
>
> * In a few cases (e.g. xen_flush_lazy_mmu()), a function knows that
>    lazy_mmu is already enabled, and it temporarily disables it by
>    calling leave() and then enter() again. Here we want to ensure
>    that any operation between the leave() and enter() calls is
>    completed immediately; for that reason we pass LAZY_MMU_DEFAULT to
>    leave() to fully disable lazy_mmu. enter() will then re-enable it
>    - this achieves the expected behaviour, whether nesting occurred
>    before that function was called or not.
>
> Note: it is difficult to provide a default definition of
> lazy_mmu_state_t for architectures implementing lazy_mmu, because
> that definition would need to be available in
> arch/x86/include/asm/paravirt_types.h and adding a new generic
>   #include there is very tricky due to the existing header soup.
>
> Acked-by: Mike Rapoport (Microsoft) <rppt@xxxxxxxxxx>
> Signed-off-by: Kevin Brodsky <kevin.brodsky@xxxxxxx>
Reviewed-by: Juergen Gross <jgross@xxxxxxxx> # arch/x86


Juergen

Attachment: OpenPGP_0xB0DE9DD628BF132F.asc
Description: OpenPGP public key

Attachment: OpenPGP_signature.asc
Description: OpenPGP digital signature