* [PATCH] rust_binder: use lock_vma_under_rcu() in use_page_slow()
@ 2026-02-18 15:13 Alice Ryhl
2026-02-18 16:16 ` Liam R. Howlett
2026-02-18 16:21 ` Jann Horn
0 siblings, 2 replies; 3+ messages in thread
From: Alice Ryhl @ 2026-02-18 15:13 UTC (permalink / raw)
To: Greg Kroah-Hartman, Carlos Llamas
Cc: Miguel Ojeda, Boqun Feng, Gary Guo, Björn Roy Baron,
Benno Lossin, Andreas Hindborg, Trevor Gross, Danilo Krummrich,
Lorenzo Stoakes, Liam R. Howlett, Jann Horn, linux-kernel,
rust-for-linux, linux-mm, Alice Ryhl
There's no reason to lock the whole mm when we are doing operations on
the vma if we can help it, so to reduce contention, use the
lock_vma_under_rcu() abstraction.
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
---
Depends on:
https://lore.kernel.org/all/20260218-binder-vma-check-v2-0-60f9d695a990@google.com/
---
drivers/android/binder/page_range.rs | 37 ++++++++++++++++++------------------
1 file changed, 19 insertions(+), 18 deletions(-)
diff --git a/drivers/android/binder/page_range.rs b/drivers/android/binder/page_range.rs
index 67aae783e8b8b7cf60ecf7e711d5f6f6f5d1dbe3..9dfc154e5dd4e889c4f3aa89e5edb89434113e1a 100644
--- a/drivers/android/binder/page_range.rs
+++ b/drivers/android/binder/page_range.rs
@@ -435,24 +435,25 @@ unsafe fn use_page_slow(&self, i: usize) -> Result<()> {
//
// Using `mmput_async` avoids this, because then the `mm` cleanup is instead queued to a
// workqueue.
- check_vma(
- MmWithUser::into_mmput_async(self.mm.mmget_not_zero().ok_or(ESRCH)?)
- .mmap_read_lock()
- .vma_lookup(vma_addr)
- .ok_or(ESRCH)?,
- self,
- )
- .ok_or(ESRCH)?
- .vm_insert_page(user_page_addr, &new_page)
- .inspect_err(|err| {
- pr_warn!(
- "Failed to vm_insert_page({}): vma_addr:{} i:{} err:{:?}",
- user_page_addr,
- vma_addr,
- i,
- err
- )
- })?;
+ let mm = MmWithUser::into_mmput_async(self.mm.mmget_not_zero().ok_or(ESRCH)?);
+ {
+ let vma_read;
+ let mmap_read;
+ let vma = if let Some(ret) = mm.lock_vma_under_rcu(vma_addr) {
+ vma_read = ret;
+ check_vma(&vma_read, self)
+ } else {
+ mmap_read = mm.mmap_read_lock();
+ mmap_read
+ .vma_lookup(vma_addr)
+ .and_then(|vma| check_vma(vma, self))
+ };
+
+ match vma {
+ Some(vma) => vma.vm_insert_page(user_page_addr, &new_page)?,
+ None => return Err(ESRCH),
+ }
+ }
let inner = self.lock.lock();
---
base-commit: 2961f841b025fb234860bac26dfb7fa7cb0fb122
change-id: 20260217-binder-vma-rcu-e699d5752ad3
prerequisite-change-id: 20260217-binder-vma-check-b6fca42e986c:v2
prerequisite-patch-id: 4ca86894150aa7ee26c04440100cb71ce599ce80
prerequisite-patch-id: beb72c0aa2ce3d495dd69817507ae6885e5ae1e6
Best regards,
--
Alice Ryhl <aliceryhl@google.com>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] rust_binder: use lock_vma_under_rcu() in use_page_slow()
2026-02-18 15:13 [PATCH] rust_binder: use lock_vma_under_rcu() in use_page_slow() Alice Ryhl
@ 2026-02-18 16:16 ` Liam R. Howlett
2026-02-18 16:21 ` Jann Horn
1 sibling, 0 replies; 3+ messages in thread
From: Liam R. Howlett @ 2026-02-18 16:16 UTC (permalink / raw)
To: Alice Ryhl
Cc: Greg Kroah-Hartman, Carlos Llamas, Miguel Ojeda, Boqun Feng,
Gary Guo, Björn Roy Baron, Benno Lossin, Andreas Hindborg,
Trevor Gross, Danilo Krummrich, Lorenzo Stoakes, Jann Horn,
linux-kernel, rust-for-linux, linux-mm
* Alice Ryhl <aliceryhl@google.com> [260218 10:13]:
> There's no reason to lock the whole mm when we are doing operations on
> the vma if we can help it, so to reduce contention, use the
> lock_vma_under_rcu() abstraction.
>
> Signed-off-by: Alice Ryhl <aliceryhl@google.com>
FWIW for rust code..
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
> ---
> Depends on:
> https://lore.kernel.org/all/20260218-binder-vma-check-v2-0-60f9d695a990@google.com/
> ---
> drivers/android/binder/page_range.rs | 37 ++++++++++++++++++------------------
> 1 file changed, 19 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/android/binder/page_range.rs b/drivers/android/binder/page_range.rs
> index 67aae783e8b8b7cf60ecf7e711d5f6f6f5d1dbe3..9dfc154e5dd4e889c4f3aa89e5edb89434113e1a 100644
> --- a/drivers/android/binder/page_range.rs
> +++ b/drivers/android/binder/page_range.rs
> @@ -435,24 +435,25 @@ unsafe fn use_page_slow(&self, i: usize) -> Result<()> {
> //
> // Using `mmput_async` avoids this, because then the `mm` cleanup is instead queued to a
> // workqueue.
> - check_vma(
> - MmWithUser::into_mmput_async(self.mm.mmget_not_zero().ok_or(ESRCH)?)
> - .mmap_read_lock()
> - .vma_lookup(vma_addr)
> - .ok_or(ESRCH)?,
> - self,
> - )
> - .ok_or(ESRCH)?
> - .vm_insert_page(user_page_addr, &new_page)
> - .inspect_err(|err| {
> - pr_warn!(
> - "Failed to vm_insert_page({}): vma_addr:{} i:{} err:{:?}",
> - user_page_addr,
> - vma_addr,
> - i,
> - err
> - )
> - })?;
> + let mm = MmWithUser::into_mmput_async(self.mm.mmget_not_zero().ok_or(ESRCH)?);
> + {
> + let vma_read;
> + let mmap_read;
> + let vma = if let Some(ret) = mm.lock_vma_under_rcu(vma_addr) {
> + vma_read = ret;
> + check_vma(&vma_read, self)
> + } else {
> + mmap_read = mm.mmap_read_lock();
> + mmap_read
> + .vma_lookup(vma_addr)
> + .and_then(|vma| check_vma(vma, self))
> + };
> +
> + match vma {
> + Some(vma) => vma.vm_insert_page(user_page_addr, &new_page)?,
> + None => return Err(ESRCH),
> + }
> + }
>
> let inner = self.lock.lock();
>
>
> ---
> base-commit: 2961f841b025fb234860bac26dfb7fa7cb0fb122
> change-id: 20260217-binder-vma-rcu-e699d5752ad3
> prerequisite-change-id: 20260217-binder-vma-check-b6fca42e986c:v2
> prerequisite-patch-id: 4ca86894150aa7ee26c04440100cb71ce599ce80
> prerequisite-patch-id: beb72c0aa2ce3d495dd69817507ae6885e5ae1e6
>
> Best regards,
> --
> Alice Ryhl <aliceryhl@google.com>
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] rust_binder: use lock_vma_under_rcu() in use_page_slow()
2026-02-18 15:13 [PATCH] rust_binder: use lock_vma_under_rcu() in use_page_slow() Alice Ryhl
2026-02-18 16:16 ` Liam R. Howlett
@ 2026-02-18 16:21 ` Jann Horn
1 sibling, 0 replies; 3+ messages in thread
From: Jann Horn @ 2026-02-18 16:21 UTC (permalink / raw)
To: Alice Ryhl
Cc: Greg Kroah-Hartman, Carlos Llamas, Miguel Ojeda, Boqun Feng,
Gary Guo, Björn Roy Baron, Benno Lossin, Andreas Hindborg,
Trevor Gross, Danilo Krummrich, Lorenzo Stoakes, Liam R. Howlett,
linux-kernel, rust-for-linux, linux-mm
On Wed, Feb 18, 2026 at 4:13 PM Alice Ryhl <aliceryhl@google.com> wrote:
> There's no reason to lock the whole mm when we are doing operations on
> the vma if we can help it, so to reduce contention, use the
> lock_vma_under_rcu() abstraction.
>
> Signed-off-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Jann Horn <jannh@google.com>
> ---
> Depends on:
> https://lore.kernel.org/all/20260218-binder-vma-check-v2-0-60f9d695a990@google.com/
> ---
> drivers/android/binder/page_range.rs | 37 ++++++++++++++++++------------------
> 1 file changed, 19 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/android/binder/page_range.rs b/drivers/android/binder/page_range.rs
> index 67aae783e8b8b7cf60ecf7e711d5f6f6f5d1dbe3..9dfc154e5dd4e889c4f3aa89e5edb89434113e1a 100644
> --- a/drivers/android/binder/page_range.rs
> +++ b/drivers/android/binder/page_range.rs
> @@ -435,24 +435,25 @@ unsafe fn use_page_slow(&self, i: usize) -> Result<()> {
> //
> // Using `mmput_async` avoids this, because then the `mm` cleanup is instead queued to a
> // workqueue.
> - check_vma(
> - MmWithUser::into_mmput_async(self.mm.mmget_not_zero().ok_or(ESRCH)?)
> - .mmap_read_lock()
> - .vma_lookup(vma_addr)
> - .ok_or(ESRCH)?,
> - self,
> - )
> - .ok_or(ESRCH)?
> - .vm_insert_page(user_page_addr, &new_page)
> - .inspect_err(|err| {
> - pr_warn!(
> - "Failed to vm_insert_page({}): vma_addr:{} i:{} err:{:?}",
> - user_page_addr,
> - vma_addr,
> - i,
> - err
> - )
> - })?;
> + let mm = MmWithUser::into_mmput_async(self.mm.mmget_not_zero().ok_or(ESRCH)?);
> + {
> + let vma_read;
> + let mmap_read;
> + let vma = if let Some(ret) = mm.lock_vma_under_rcu(vma_addr) {
> + vma_read = ret;
> + check_vma(&vma_read, self)
nit: this might look nicer if the check_vma() calls were moved out of
the two separate branches
> + } else {
> + mmap_read = mm.mmap_read_lock();
> + mmap_read
> + .vma_lookup(vma_addr)
> + .and_then(|vma| check_vma(vma, self))
> + };
> +
> + match vma {
> + Some(vma) => vma.vm_insert_page(user_page_addr, &new_page)?,
> + None => return Err(ESRCH),
> + }
> + }
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-02-18 16:21 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-02-18 15:13 [PATCH] rust_binder: use lock_vma_under_rcu() in use_page_slow() Alice Ryhl
2026-02-18 16:16 ` Liam R. Howlett
2026-02-18 16:21 ` Jann Horn
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox