* [PATCH] KEYS: user: Align the payload buffer
@ 2019-01-15 3:37 ` Eric Biggers
0 siblings, 0 replies; 17+ messages in thread
From: Eric Biggers @ 2019-01-15 3:37 UTC (permalink / raw)
To: keyrings, David Howells; +Cc: Aaro Koskinen, linux-fscrypt, linux-kernel
From: Eric Biggers <ebiggers@google.com>
Align the payload of "user" and "logon" keys so that users of the
keyrings service can access it as a struct that requires more than
2-byte alignment. fscrypt currently does this which results in the read
of fscrypt_key::size being misaligned as it needs 4-byte alignment.
Align to __alignof__(u64) rather than __alignof__(long) since in the
future it's conceivable that people would use structs beginning with
u64, which on some platforms would require more than 'long' alignment.
Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
---
include/keys/user-type.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/keys/user-type.h b/include/keys/user-type.h
index e098cbe27db54..12babe9915944 100644
--- a/include/keys/user-type.h
+++ b/include/keys/user-type.h
@@ -31,7 +31,7 @@
struct user_key_payload {
struct rcu_head rcu; /* RCU destructor */
unsigned short datalen; /* length of this data */
- char data[0]; /* actual data */
+ char data[0] __aligned(__alignof__(u64)); /* actual data */
};
extern struct key_type key_type_user;
--
2.20.1
^ permalink raw reply related [flat|nested] 17+ messages in thread* Re: [PATCH] KEYS: user: Align the payload buffer
2019-01-15 3:37 ` Eric Biggers
@ 2019-01-15 19:19 ` Aaro Koskinen
-1 siblings, 0 replies; 17+ messages in thread
From: Aaro Koskinen @ 2019-01-15 19:19 UTC (permalink / raw)
To: Eric Biggers; +Cc: keyrings, David Howells, linux-fscrypt, linux-kernel
Hi,
On Mon, Jan 14, 2019 at 07:37:16PM -0800, Eric Biggers wrote:
> From: Eric Biggers <ebiggers@google.com>
>
> Align the payload of "user" and "logon" keys so that users of the
> keyrings service can access it as a struct that requires more than
> 2-byte alignment. fscrypt currently does this which results in the read
> of fscrypt_key::size being misaligned as it needs 4-byte alignment.
>
> Align to __alignof__(u64) rather than __alignof__(long) since in the
> future it's conceivable that people would use structs beginning with
> u64, which on some platforms would require more than 'long' alignment.
>
> Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
> Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
> Cc: stable@vger.kernel.org
> Signed-off-by: Eric Biggers <ebiggers@google.com>
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
This fixes the "unaligned access" dmesg spam on SPARC when using ext4
encryption.
Thanks,
A.
> ---
> include/keys/user-type.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/keys/user-type.h b/include/keys/user-type.h
> index e098cbe27db54..12babe9915944 100644
> --- a/include/keys/user-type.h
> +++ b/include/keys/user-type.h
> @@ -31,7 +31,7 @@
> struct user_key_payload {
> struct rcu_head rcu; /* RCU destructor */
> unsigned short datalen; /* length of this data */
> - char data[0]; /* actual data */
> + char data[0] __aligned(__alignof__(u64)); /* actual data */
> };
>
> extern struct key_type key_type_user;
> --
> 2.20.1
>
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH] KEYS: user: Align the payload buffer
@ 2019-01-15 19:19 ` Aaro Koskinen
0 siblings, 0 replies; 17+ messages in thread
From: Aaro Koskinen @ 2019-01-15 19:19 UTC (permalink / raw)
To: Eric Biggers; +Cc: keyrings, David Howells, linux-fscrypt, linux-kernel
Hi,
On Mon, Jan 14, 2019 at 07:37:16PM -0800, Eric Biggers wrote:
> From: Eric Biggers <ebiggers@google.com>
>
> Align the payload of "user" and "logon" keys so that users of the
> keyrings service can access it as a struct that requires more than
> 2-byte alignment. fscrypt currently does this which results in the read
> of fscrypt_key::size being misaligned as it needs 4-byte alignment.
>
> Align to __alignof__(u64) rather than __alignof__(long) since in the
> future it's conceivable that people would use structs beginning with
> u64, which on some platforms would require more than 'long' alignment.
>
> Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
> Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
> Cc: stable@vger.kernel.org
> Signed-off-by: Eric Biggers <ebiggers@google.com>
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
This fixes the "unaligned access" dmesg spam on SPARC when using ext4
encryption.
Thanks,
A.
> ---
> include/keys/user-type.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/keys/user-type.h b/include/keys/user-type.h
> index e098cbe27db54..12babe9915944 100644
> --- a/include/keys/user-type.h
> +++ b/include/keys/user-type.h
> @@ -31,7 +31,7 @@
> struct user_key_payload {
> struct rcu_head rcu; /* RCU destructor */
> unsigned short datalen; /* length of this data */
> - char data[0]; /* actual data */
> + char data[0] __aligned(__alignof__(u64)); /* actual data */
> };
>
> extern struct key_type key_type_user;
> --
> 2.20.1
>
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [PATCH] KEYS: user: Align the payload buffer
2019-01-15 3:37 ` Eric Biggers
@ 2019-02-07 23:35 ` Eric Biggers
-1 siblings, 0 replies; 17+ messages in thread
From: Eric Biggers @ 2019-02-07 23:35 UTC (permalink / raw)
To: keyrings, David Howells; +Cc: Aaro Koskinen, linux-fscrypt, linux-kernel
On Mon, Jan 14, 2019 at 07:37:16PM -0800, Eric Biggers wrote:
> From: Eric Biggers <ebiggers@google.com>
>
> Align the payload of "user" and "logon" keys so that users of the
> keyrings service can access it as a struct that requires more than
> 2-byte alignment. fscrypt currently does this which results in the read
> of fscrypt_key::size being misaligned as it needs 4-byte alignment.
>
> Align to __alignof__(u64) rather than __alignof__(long) since in the
> future it's conceivable that people would use structs beginning with
> u64, which on some platforms would require more than 'long' alignment.
>
> Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
> Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
> Cc: stable@vger.kernel.org
> Signed-off-by: Eric Biggers <ebiggers@google.com>
> ---
> include/keys/user-type.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/keys/user-type.h b/include/keys/user-type.h
> index e098cbe27db54..12babe9915944 100644
> --- a/include/keys/user-type.h
> +++ b/include/keys/user-type.h
> @@ -31,7 +31,7 @@
> struct user_key_payload {
> struct rcu_head rcu; /* RCU destructor */
> unsigned short datalen; /* length of this data */
> - char data[0]; /* actual data */
> + char data[0] __aligned(__alignof__(u64)); /* actual data */
> };
>
> extern struct key_type key_type_user;
> --
> 2.20.1
>
Ping. David, are you planning to apply this?
- Eric
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH] KEYS: user: Align the payload buffer
@ 2019-02-07 23:35 ` Eric Biggers
0 siblings, 0 replies; 17+ messages in thread
From: Eric Biggers @ 2019-02-07 23:35 UTC (permalink / raw)
To: keyrings, David Howells; +Cc: Aaro Koskinen, linux-fscrypt, linux-kernel
On Mon, Jan 14, 2019 at 07:37:16PM -0800, Eric Biggers wrote:
> From: Eric Biggers <ebiggers@google.com>
>
> Align the payload of "user" and "logon" keys so that users of the
> keyrings service can access it as a struct that requires more than
> 2-byte alignment. fscrypt currently does this which results in the read
> of fscrypt_key::size being misaligned as it needs 4-byte alignment.
>
> Align to __alignof__(u64) rather than __alignof__(long) since in the
> future it's conceivable that people would use structs beginning with
> u64, which on some platforms would require more than 'long' alignment.
>
> Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
> Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
> Cc: stable@vger.kernel.org
> Signed-off-by: Eric Biggers <ebiggers@google.com>
> ---
> include/keys/user-type.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/keys/user-type.h b/include/keys/user-type.h
> index e098cbe27db54..12babe9915944 100644
> --- a/include/keys/user-type.h
> +++ b/include/keys/user-type.h
> @@ -31,7 +31,7 @@
> struct user_key_payload {
> struct rcu_head rcu; /* RCU destructor */
> unsigned short datalen; /* length of this data */
> - char data[0]; /* actual data */
> + char data[0] __aligned(__alignof__(u64)); /* actual data */
> };
>
> extern struct key_type key_type_user;
> --
> 2.20.1
>
Ping. David, are you planning to apply this?
- Eric
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH] KEYS: user: Align the payload buffer
2019-02-07 23:35 ` Eric Biggers
@ 2019-02-19 23:08 ` Eric Biggers
-1 siblings, 0 replies; 17+ messages in thread
From: Eric Biggers @ 2019-02-19 23:08 UTC (permalink / raw)
To: keyrings, David Howells; +Cc: Aaro Koskinen, linux-fscrypt, linux-kernel
On Thu, Feb 07, 2019 at 03:35:09PM -0800, Eric Biggers wrote:
> On Mon, Jan 14, 2019 at 07:37:16PM -0800, Eric Biggers wrote:
> > From: Eric Biggers <ebiggers@google.com>
> >
> > Align the payload of "user" and "logon" keys so that users of the
> > keyrings service can access it as a struct that requires more than
> > 2-byte alignment. fscrypt currently does this which results in the read
> > of fscrypt_key::size being misaligned as it needs 4-byte alignment.
> >
> > Align to __alignof__(u64) rather than __alignof__(long) since in the
> > future it's conceivable that people would use structs beginning with
> > u64, which on some platforms would require more than 'long' alignment.
> >
> > Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> > Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
> > Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
> > Cc: stable@vger.kernel.org
> > Signed-off-by: Eric Biggers <ebiggers@google.com>
> > ---
> > include/keys/user-type.h | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/include/keys/user-type.h b/include/keys/user-type.h
> > index e098cbe27db54..12babe9915944 100644
> > --- a/include/keys/user-type.h
> > +++ b/include/keys/user-type.h
> > @@ -31,7 +31,7 @@
> > struct user_key_payload {
> > struct rcu_head rcu; /* RCU destructor */
> > unsigned short datalen; /* length of this data */
> > - char data[0]; /* actual data */
> > + char data[0] __aligned(__alignof__(u64)); /* actual data */
> > };
> >
> > extern struct key_type key_type_user;
> > --
> > 2.20.1
> >
>
> Ping. David, are you planning to apply this?
>
> - Eric
Ping.
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH] KEYS: user: Align the payload buffer
@ 2019-02-19 23:08 ` Eric Biggers
0 siblings, 0 replies; 17+ messages in thread
From: Eric Biggers @ 2019-02-19 23:08 UTC (permalink / raw)
To: keyrings, David Howells; +Cc: Aaro Koskinen, linux-fscrypt, linux-kernel
On Thu, Feb 07, 2019 at 03:35:09PM -0800, Eric Biggers wrote:
> On Mon, Jan 14, 2019 at 07:37:16PM -0800, Eric Biggers wrote:
> > From: Eric Biggers <ebiggers@google.com>
> >
> > Align the payload of "user" and "logon" keys so that users of the
> > keyrings service can access it as a struct that requires more than
> > 2-byte alignment. fscrypt currently does this which results in the read
> > of fscrypt_key::size being misaligned as it needs 4-byte alignment.
> >
> > Align to __alignof__(u64) rather than __alignof__(long) since in the
> > future it's conceivable that people would use structs beginning with
> > u64, which on some platforms would require more than 'long' alignment.
> >
> > Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
> > Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
> > Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
> > Cc: stable@vger.kernel.org
> > Signed-off-by: Eric Biggers <ebiggers@google.com>
> > ---
> > include/keys/user-type.h | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/include/keys/user-type.h b/include/keys/user-type.h
> > index e098cbe27db54..12babe9915944 100644
> > --- a/include/keys/user-type.h
> > +++ b/include/keys/user-type.h
> > @@ -31,7 +31,7 @@
> > struct user_key_payload {
> > struct rcu_head rcu; /* RCU destructor */
> > unsigned short datalen; /* length of this data */
> > - char data[0]; /* actual data */
> > + char data[0] __aligned(__alignof__(u64)); /* actual data */
> > };
> >
> > extern struct key_type key_type_user;
> > --
> > 2.20.1
> >
>
> Ping. David, are you planning to apply this?
>
> - Eric
Ping.
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH] KEYS: user: Align the payload buffer
2019-02-19 23:08 ` Eric Biggers
(?)
@ 2019-02-20 13:32 ` David Howells
-1 siblings, 0 replies; 17+ messages in thread
From: David Howells @ 2019-02-20 13:32 UTC (permalink / raw)
To: Eric Biggers
Cc: dhowells, keyrings, Aaro Koskinen, linux-fscrypt, linux-kernel
Sent on to James.
David
^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH] KEYS: user: Align the payload buffer
@ 2019-02-20 13:32 ` David Howells
0 siblings, 0 replies; 17+ messages in thread
From: David Howells @ 2019-02-20 13:32 UTC (permalink / raw)
To: jmorris; +Cc: Aaro Koskinen, stable, Eric Biggers
From: Eric Biggers <ebiggers@google.com>
Align the payload of "user" and "logon" keys so that users of the
keyrings service can access it as a struct that requires more than
2-byte alignment. fscrypt currently does this which results in the read
of fscrypt_key::size being misaligned as it needs 4-byte alignment.
Align to __alignof__(u64) rather than __alignof__(long) since in the
future it's conceivable that people would use structs beginning with
u64, which on some platforms would require more than 'long' alignment.
Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: David Howells <dhowells@redhat.com>
---
include/keys/user-type.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/keys/user-type.h b/include/keys/user-type.h
index e098cbe27db5..12babe991594 100644
--- a/include/keys/user-type.h
+++ b/include/keys/user-type.h
@@ -31,7 +31,7 @@
struct user_key_payload {
struct rcu_head rcu; /* RCU destructor */
unsigned short datalen; /* length of this data */
- char data[0]; /* actual data */
+ char data[0] __aligned(__alignof__(u64)); /* actual data */
};
extern struct key_type key_type_user;
^ permalink raw reply related [flat|nested] 17+ messages in thread* [PATCH] KEYS: user: Align the payload buffer
@ 2019-02-20 13:32 ` David Howells
0 siblings, 0 replies; 17+ messages in thread
From: David Howells @ 2019-02-20 13:32 UTC (permalink / raw)
To: jmorris
Cc: Aaro Koskinen, stable, Eric Biggers, Aaro Koskinen, keyrings,
linux-security-module, linux-kernel
From: Eric Biggers <ebiggers@google.com>
Align the payload of "user" and "logon" keys so that users of the
keyrings service can access it as a struct that requires more than
2-byte alignment. fscrypt currently does this which results in the read
of fscrypt_key::size being misaligned as it needs 4-byte alignment.
Align to __alignof__(u64) rather than __alignof__(long) since in the
future it's conceivable that people would use structs beginning with
u64, which on some platforms would require more than 'long' alignment.
Reported-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Fixes: 2aa349f6e37c ("[PATCH] Keys: Export user-defined keyring operations")
Fixes: 88bd6ccdcdd6 ("ext4 crypto: add encryption key management facilities")
Cc: stable@vger.kernel.org
Signed-off-by: Eric Biggers <ebiggers@google.com>
Tested-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: David Howells <dhowells@redhat.com>
---
include/keys/user-type.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/keys/user-type.h b/include/keys/user-type.h
index e098cbe27db5..12babe991594 100644
--- a/include/keys/user-type.h
+++ b/include/keys/user-type.h
@@ -31,7 +31,7 @@
struct user_key_payload {
struct rcu_head rcu; /* RCU destructor */
unsigned short datalen; /* length of this data */
- char data[0]; /* actual data */
+ char data[0] __aligned(__alignof__(u64)); /* actual data */
};
extern struct key_type key_type_user;
^ permalink raw reply related [flat|nested] 17+ messages in thread* RE: [PATCH] KEYS: user: Align the payload buffer
2019-02-20 13:32 ` David Howells
@ 2019-02-20 13:52 ` David Laight
-1 siblings, 0 replies; 17+ messages in thread
From: David Laight @ 2019-02-20 13:52 UTC (permalink / raw)
To: 'David Howells', jmorris@namei.org
Cc: Aaro Koskinen, stable@vger.kernel.org, Eric Biggers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^ permalink raw reply [flat|nested] 17+ messages in thread
* RE: [PATCH] KEYS: user: Align the payload buffer
@ 2019-02-20 13:52 ` David Laight
0 siblings, 0 replies; 17+ messages in thread
From: David Laight @ 2019-02-20 13:52 UTC (permalink / raw)
To: 'David Howells', jmorris@namei.org
Cc: Aaro Koskinen, stable@vger.kernel.org, Eric Biggers,
Aaro Koskinen, keyrings@vger.kernel.org,
linux-security-module@vger.kernel.org,
linux-kernel@vger.kernel.org
From: David Howells
> Sent: 20 February 2019 13:32
>
> From: Eric Biggers <ebiggers@google.com>
>
> Align the payload of "user" and "logon" keys so that users of the
> keyrings service can access it as a struct that requires more than
> 2-byte alignment. fscrypt currently does this which results in the read
> of fscrypt_key::size being misaligned as it needs 4-byte alignment.
>
> Align to __alignof__(u64) rather than __alignof__(long) since in the
> future it's conceivable that people would use structs beginning with
> u64, which on some platforms would require more than 'long' alignment.
...
> include/keys/user-type.h | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/include/keys/user-type.h b/include/keys/user-type.h
> index e098cbe27db5..12babe991594 100644
> --- a/include/keys/user-type.h
> +++ b/include/keys/user-type.h
> @@ -31,7 +31,7 @@
> struct user_key_payload {
> struct rcu_head rcu; /* RCU destructor */
> unsigned short datalen; /* length of this data */
> - char data[0]; /* actual data */
> + char data[0] __aligned(__alignof__(u64)); /* actual data */
> };
I'd make the 'datalen' field 'unsigned int' at the same time.
It will use some of the hole you've made and generate better
code on most arches.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
^ permalink raw reply [flat|nested] 17+ messages in thread* Re: [PATCH] KEYS: user: Align the payload buffer
2019-02-20 13:52 ` David Laight
(?)
@ 2019-02-20 14:09 ` David Howells
2019-02-20 20:50 ` James Morris
-1 siblings, 1 reply; 17+ messages in thread
From: David Howells @ 2019-02-20 14:09 UTC (permalink / raw)
To: David Laight
Cc: dhowells, jmorris@namei.org, Aaro Koskinen,
stable@vger.kernel.org, Eric Biggers, keyrings@vger.kernel.org,
linux-security-module@vger.kernel.org,
linux-kernel@vger.kernel.org
David Laight <David.Laight@ACULAB.COM> wrote:
> I'd make the 'datalen' field 'unsigned int' at the same time.
> It will use some of the hole you've made and generate better
> code on most arches.
Most arches? I though most, if not all, arches had a load-word instruction.
Do you want to send me a patch for that? I'd rather not alter this patch at
this point. I can pass the additional patch to James for the next merge
window.
David
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [PATCH] KEYS: user: Align the payload buffer
2019-02-20 14:09 ` David Howells
@ 2019-02-20 20:50 ` James Morris
0 siblings, 0 replies; 17+ messages in thread
From: James Morris @ 2019-02-20 20:50 UTC (permalink / raw)
To: David Howells
Cc: David Laight, Aaro Koskinen, stable@vger.kernel.org, Eric Biggers,
keyrings@vger.kernel.org, linux-security-module@vger.kernel.org,
linux-kernel@vger.kernel.org
On Wed, 20 Feb 2019, David Howells wrote:
> David Laight <David.Laight@ACULAB.COM> wrote:
>
> > I'd make the 'datalen' field 'unsigned int' at the same time.
> > It will use some of the hole you've made and generate better
> > code on most arches.
>
> Most arches? I though most, if not all, arches had a load-word instruction.
>
> Do you want to send me a patch for that? I'd rather not alter this patch at
> this point. I can pass the additional patch to James for the next merge
> window.
Should this first one go into -rc?
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [PATCH] KEYS: user: Align the payload buffer
@ 2019-02-20 20:50 ` James Morris
0 siblings, 0 replies; 17+ messages in thread
From: James Morris @ 2019-02-20 20:50 UTC (permalink / raw)
To: David Howells
Cc: David Laight, Aaro Koskinen, stable@vger.kernel.org, Eric Biggers,
keyrings@vger.kernel.org, linux-security-module@vger.kernel.org,
linux-kernel@vger.kernel.org
On Wed, 20 Feb 2019, David Howells wrote:
> David Laight <David.Laight@ACULAB.COM> wrote:
>
> > I'd make the 'datalen' field 'unsigned int' at the same time.
> > It will use some of the hole you've made and generate better
> > code on most arches.
>
> Most arches? I though most, if not all, arches had a load-word instruction.
>
> Do you want to send me a patch for that? I'd rather not alter this patch at
> this point. I can pass the additional patch to James for the next merge
> window.
Should this first one go into -rc?
--
James Morris
<jmorris@namei.org>
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [PATCH] KEYS: user: Align the payload buffer
2019-02-20 20:50 ` James Morris
(?)
@ 2019-02-20 21:37 ` David Howells
-1 siblings, 0 replies; 17+ messages in thread
From: David Howells @ 2019-02-20 21:37 UTC (permalink / raw)
To: James Morris
Cc: dhowells, David Laight, Aaro Koskinen, stable@vger.kernel.org,
Eric Biggers, keyrings@vger.kernel.org,
linux-security-module@vger.kernel.org,
linux-kernel@vger.kernel.org
James Morris <jmorris@namei.org> wrote:
> Should this first one go into -rc?
Yes please.
David
^ permalink raw reply [flat|nested] 17+ messages in thread