2020-03-12 20:04:12

by Daniel Xu

[permalink] [raw]

Subject: [PATCH v3 1/4] kernfs: kvmalloc xattr value instead of kmalloc

---

xattr values have a 64k maximum size. This can result in an order 4
kmalloc request which can be difficult to fulfill. Since xattrs do not
need physically contiguous memory, we can switch to kvmalloc and not
have to worry about higher order allocations failing.

Signed-off-by: Daniel Xu <[email protected]>
---
fs/xattr.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/xattr.c b/fs/xattr.c
index 90dd78f0eb27..0d3c9b4d1914 100644
--- a/fs/xattr.c
+++ b/fs/xattr.c
@@ -817,7 +817,7 @@ struct simple_xattr *simple_xattr_alloc(const void *value, size_t size)
if (len < sizeof(*new_xattr))
return NULL;

- new_xattr = kmalloc(len, GFP_KERNEL);
+ new_xattr = kvmalloc(len, GFP_KERNEL);
if (!new_xattr)
return NULL;

@@ -882,7 +882,7 @@ int simple_xattr_set(struct simple_xattrs *xattrs, const char *name,

new_xattr->name = kstrdup(name, GFP_KERNEL);
if (!new_xattr->name) {
- kfree(new_xattr);
+ kvfree(new_xattr);
return -ENOMEM;
}
}
@@ -912,7 +912,7 @@ int simple_xattr_set(struct simple_xattrs *xattrs, const char *name,
spin_unlock(&xattrs->lock);
if (xattr) {
kfree(xattr->name);
- kfree(xattr);
+ kvfree(xattr);
}
return err;

--
2.21.1

2020-03-12 21:06:21

by Shakeel Butt

[permalink] [raw]

Subject: Re: [PATCH v3 1/4] kernfs: kvmalloc xattr value instead of kmalloc

---

On Thu, Mar 12, 2020 at 1:03 PM Daniel Xu <[email protected]> wrote:
>
> xattr values have a 64k maximum size. This can result in an order 4
> kmalloc request which can be difficult to fulfill. Since xattrs do not
> need physically contiguous memory, we can switch to kvmalloc and not
> have to worry about higher order allocations failing.
>
> Signed-off-by: Daniel Xu <[email protected]>

Reviewed-by: Shakeel Butt <[email protected]>

2020-03-12 21:10:25

by Andreas Dilger

[permalink] [raw]

Subject: Re: [PATCH v3 1/4] kernfs: kvmalloc xattr value instead of kmalloc

---

On Mar 12, 2020, at 2:03 PM, Daniel Xu <[email protected]> wrote:
>
> xattr values have a 64k maximum size. This can result in an order 4
> kmalloc request which can be difficult to fulfill. Since xattrs do not
> need physically contiguous memory, we can switch to kvmalloc and not
> have to worry about higher order allocations failing.
>
> Signed-off-by: Daniel Xu <[email protected]>

Reviewed-by: Andreas Dilger <[email protected]>

> ---
> fs/xattr.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/fs/xattr.c b/fs/xattr.c
> index 90dd78f0eb27..0d3c9b4d1914 100644
> --- a/fs/xattr.c
> +++ b/fs/xattr.c
> @@ -817,7 +817,7 @@ struct simple_xattr *simple_xattr_alloc(const void *value, size_t size)
> if (len < sizeof(*new_xattr))
> return NULL;
>
> - new_xattr = kmalloc(len, GFP_KERNEL);
> + new_xattr = kvmalloc(len, GFP_KERNEL);
> if (!new_xattr)
> return NULL;
>
> @@ -882,7 +882,7 @@ int simple_xattr_set(struct simple_xattrs *xattrs, const char *name,
>
> new_xattr->name = kstrdup(name, GFP_KERNEL);
> if (!new_xattr->name) {
> - kfree(new_xattr);
> + kvfree(new_xattr);
> return -ENOMEM;
> }
> }
> @@ -912,7 +912,7 @@ int simple_xattr_set(struct simple_xattrs *xattrs, const char *name,
> spin_unlock(&xattrs->lock);
> if (xattr) {
> kfree(xattr->name);
> - kfree(xattr);
> + kvfree(xattr);
> }
> return err;
>
> --
> 2.21.1
>


Cheers, Andreas

---

Attachments:

signature.asc (890.00 B)
Message signed with OpenPGP