[PATCH v2] KFENCE: Clarify that sample allocations are not following NUMA or memory policies

Christoph Lameter via B4 Relay posted 1 patch 6 days, 2 hours ago
Documentation/dev-tools/kfence.rst | 4 +++-
lib/Kconfig.kfence                 | 8 +++++---
2 files changed, 8 insertions(+), 4 deletions(-)
[PATCH v2] KFENCE: Clarify that sample allocations are not following NUMA or memory policies
Posted by Christoph Lameter via B4 Relay 6 days, 2 hours ago
From: Christoph Lameter <cl@linux.com>

KFENCE manages its own pools and redirects regular memory allocations
to those pools in a sporadic way. The usual memory allocator features
like NUMA, memory policies and pfmemalloc are not supported.
This means that one gets surprising object placement with KFENCE that
may impact performance on some NUMA systems.

Update the description and make KFENCE depend on VM debugging
having been enabled.

Signed-off-by: Christoph Lameter <cl@linux.com>
---
Reviewed-by: Yang Shi <shy828301@gmail.com>
---
Changes in v2:
- Remove dependency on CONFIG_DEBUG_VM.
- Spelling fixes.
- Link to v1: https://lore.kernel.org/r/20250123-kfence_doc_update-v1-1-9aa8e94b3d0b@gentwo.org
---
 Documentation/dev-tools/kfence.rst | 4 +++-
 lib/Kconfig.kfence                 | 8 +++++---
 2 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/Documentation/dev-tools/kfence.rst b/Documentation/dev-tools/kfence.rst
index 541899353865..03062d0941dc 100644
--- a/Documentation/dev-tools/kfence.rst
+++ b/Documentation/dev-tools/kfence.rst
@@ -8,7 +8,9 @@ Kernel Electric-Fence (KFENCE) is a low-overhead sampling-based memory safety
 error detector. KFENCE detects heap out-of-bounds access, use-after-free, and
 invalid-free errors.
 
-KFENCE is designed to be enabled in production kernels, and has near zero
+KFENCE is designed to be low overhead but does not implement the typical
+memory allocation features for its samples like memory policies, NUMA and
+management of emergency memory pools. It has near zero
 performance overhead. Compared to KASAN, KFENCE trades performance for
 precision. The main motivation behind KFENCE's design, is that with enough
 total uptime KFENCE will detect bugs in code paths not typically exercised by
diff --git a/lib/Kconfig.kfence b/lib/Kconfig.kfence
index 6fbbebec683a..1f9f79df2d0a 100644
--- a/lib/Kconfig.kfence
+++ b/lib/Kconfig.kfence
@@ -11,8 +11,8 @@ menuconfig KFENCE
 	help
 	  KFENCE is a low-overhead sampling-based detector of heap out-of-bounds
 	  access, use-after-free, and invalid-free errors. KFENCE is designed
-	  to have negligible cost to permit enabling it in production
-	  environments.
+	  to have negligible cost. KFENCE does not support NUMA features
+	  and other memory allocator features for it sample allocations.
 
 	  See <file:Documentation/dev-tools/kfence.rst> for more details.
 
@@ -21,7 +21,9 @@ menuconfig KFENCE
 	  detect, albeit at very different performance profiles. If you can
 	  afford to use KASAN, continue using KASAN, for example in test
 	  environments. If your kernel targets production use, and cannot
-	  enable KASAN due to its cost, consider using KFENCE.
+	  enable KASAN due to its cost and you are not using NUMA and have
+	  no use of the memory reserve logic of the memory allocators,
+	  consider using KFENCE.
 
 if KFENCE
 

---
base-commit: d0d106a2bd21499901299160744e5fe9f4c83ddb
change-id: 20250123-kfence_doc_update-93b4576c25bb

Best regards,
-- 
Christoph Lameter <cl@gentwo.org>
Re: [PATCH v2] KFENCE: Clarify that sample allocations are not following NUMA or memory policies
Posted by Marco Elver 5 days, 19 hours ago
On Fri, 24 Jan 2025 at 03:06, Christoph Lameter via B4 Relay
<devnull+cl.gentwo.org@kernel.org> wrote:
>
> From: Christoph Lameter <cl@linux.com>
>
> KFENCE manages its own pools and redirects regular memory allocations
> to those pools in a sporadic way. The usual memory allocator features
> like NUMA, memory policies and pfmemalloc are not supported.
> This means that one gets surprising object placement with KFENCE that
> may impact performance on some NUMA systems.
>
> Update the description and make KFENCE depend on VM debugging
> having been enabled.

The commit message still incorrectly says "depend on VM debugging".

> Signed-off-by: Christoph Lameter <cl@linux.com>
> ---
> Reviewed-by: Yang Shi <shy828301@gmail.com>

If it's after '---', it will be ignored.

> ---
> Changes in v2:
> - Remove dependency on CONFIG_DEBUG_VM.
> - Spelling fixes.
> - Link to v1: https://lore.kernel.org/r/20250123-kfence_doc_update-v1-1-9aa8e94b3d0b@gentwo.org
> ---
>  Documentation/dev-tools/kfence.rst | 4 +++-
>  lib/Kconfig.kfence                 | 8 +++++---
>  2 files changed, 8 insertions(+), 4 deletions(-)
>
> diff --git a/Documentation/dev-tools/kfence.rst b/Documentation/dev-tools/kfence.rst
> index 541899353865..03062d0941dc 100644
> --- a/Documentation/dev-tools/kfence.rst
> +++ b/Documentation/dev-tools/kfence.rst
> @@ -8,7 +8,9 @@ Kernel Electric-Fence (KFENCE) is a low-overhead sampling-based memory safety
>  error detector. KFENCE detects heap out-of-bounds access, use-after-free, and
>  invalid-free errors.
>
> -KFENCE is designed to be enabled in production kernels, and has near zero
> +KFENCE is designed to be low overhead but does not implement the typical
> +memory allocation features for its samples like memory policies, NUMA and
> +management of emergency memory pools. It has near zero
>  performance overhead. Compared to KASAN, KFENCE trades performance for
>  precision. The main motivation behind KFENCE's design, is that with enough
>  total uptime KFENCE will detect bugs in code paths not typically exercised by
> diff --git a/lib/Kconfig.kfence b/lib/Kconfig.kfence
> index 6fbbebec683a..1f9f79df2d0a 100644
> --- a/lib/Kconfig.kfence
> +++ b/lib/Kconfig.kfence
> @@ -11,8 +11,8 @@ menuconfig KFENCE
>         help
>           KFENCE is a low-overhead sampling-based detector of heap out-of-bounds
>           access, use-after-free, and invalid-free errors. KFENCE is designed
> -         to have negligible cost to permit enabling it in production
> -         environments.
> +         to have negligible cost. KFENCE does not support NUMA features
> +         and other memory allocator features for it sample allocations.

This still doesn't parse: "for its sample allocations" ?

>           See <file:Documentation/dev-tools/kfence.rst> for more details.
>
> @@ -21,7 +21,9 @@ menuconfig KFENCE
>           detect, albeit at very different performance profiles. If you can
>           afford to use KASAN, continue using KASAN, for example in test
>           environments. If your kernel targets production use, and cannot
> -         enable KASAN due to its cost, consider using KFENCE.
> +         enable KASAN due to its cost and you are not using NUMA and have
> +         no use of the memory reserve logic of the memory allocators,
> +         consider using KFENCE.

This doesn't read well. As I said in the other mail, this is repeating
what you say above, but is mostly irrelevant in this context. The main
point of this paragraph is that KFENCE is no substitute for KASAN, but
can be a reasonable alternative if KASAN can't be used at all (in the
beginning, we had some folks thinking KFENCE is like KASAN, but faster
- but that's very wrong).

If you don't like the last sentence because of the "production"
reference, I'd suggest to just remove the whole sentence starting with
"If your kernel targets production use, ...".