This set of patches fixes some warning messages generated by gcc 4.6.0.
Please have a look at these if/when you have the time, and let me know
what might be changed etc..
cheers,
Justin P. Mattock
building with gcc 4.6 I'm getting a warning message:
CC security/keys/keyctl.o
security/keys/keyctl.c: In function 'keyctl_describe_key':
security/keys/keyctl.c:472:14: warning: variable 'key' set but not used
After reading key.h I noticed it says this:
NOTE! key_ref_t is a typedef'd pointer to a type that is not actually
defined. This is because we abuse the bottom bit of the reference to carry a
flag to indicate whether the calling process possesses that key in one of
its keyrings.
In this case the safest approach(in my mind) would be to just
mark the integer __unused. Keep in mind though Im not certain
if this is the right place for this value i.e. will this effect
*instkey or not(please check).
Signed-off-by: Justin P. Mattock <[email protected]>
---
security/keys/keyctl.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/security/keys/keyctl.c b/security/keys/keyctl.c
index 13074b4..d7bb74f 100644
--- a/security/keys/keyctl.c
+++ b/security/keys/keyctl.c
@@ -469,7 +469,7 @@ long keyctl_describe_key(key_serial_t keyid,
char __user *buffer,
size_t buflen)
{
- struct key *key, *instkey;
+ struct key *key __attribute__((unused)), *instkey;
key_ref_t key_ref;
char *tmpbuf;
long ret;
--
1.7.1.rc1.21.gf3bd6
Im getting this when compiling on gcc 4.6.0
CC [M] drivers/block/cryptoloop.o
drivers/block/cryptoloop.c: In function 'cryptoloop_init':
drivers/block/cryptoloop.c:46:8: warning: variable 'cipher' set but not used
The below fixes it for me. Please have a look and let me know.
Signed-off-by: Justin P. Mattock <[email protected]>
---
drivers/block/cryptoloop.c | 2 --
1 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/drivers/block/cryptoloop.c b/drivers/block/cryptoloop.c
index 8b6bb76..fb8a6fd 100644
--- a/drivers/block/cryptoloop.c
+++ b/drivers/block/cryptoloop.c
@@ -43,7 +43,6 @@ cryptoloop_init(struct loop_device *lo, const struct loop_info64 *info)
int cipher_len;
int mode_len;
char cms[LO_NAME_SIZE]; /* cipher-mode string */
- char *cipher;
char *mode;
char *cmsp = cms; /* c-m string pointer */
struct crypto_blkcipher *tfm;
@@ -56,7 +55,6 @@ cryptoloop_init(struct loop_device *lo, const struct loop_info64 *info)
strncpy(cms, info->lo_crypt_name, LO_NAME_SIZE);
cms[LO_NAME_SIZE - 1] = 0;
- cipher = cmsp;
cipher_len = strcspn(cmsp, "-");
mode = cmsp + cipher_len;
--
1.7.1.rc1.21.gf3bd6
Im seeing this building the kernel with gcc 4.6.0
CC [M] drivers/bluetooth/hci_bcsp.o
drivers/bluetooth/hci_bcsp.c: In function 'bcsp_prepare_pkt':
drivers/bluetooth/hci_bcsp.c:247:20: warning: operation on 'bcsp->msgq_txseq' may be undefined
Hopefully the below is a fix for this. Please let me know.
Signed-off-by: Justin P. Mattock <[email protected]>
---
drivers/bluetooth/hci_bcsp.c | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/bluetooth/hci_bcsp.c b/drivers/bluetooth/hci_bcsp.c
index 40aec0f..0f892e7 100644
--- a/drivers/bluetooth/hci_bcsp.c
+++ b/drivers/bluetooth/hci_bcsp.c
@@ -182,7 +182,7 @@ static struct sk_buff *bcsp_prepare_pkt(struct bcsp_struct *bcsp, u8 *data,
struct sk_buff *nskb;
u8 hdr[4], chan;
u16 BCSP_CRC_INIT(bcsp_txmsg_crc);
- int rel, i;
+ int rel, i, ret;
switch (pkt_type) {
case HCI_ACLDATA_PKT:
@@ -243,8 +243,8 @@ static struct sk_buff *bcsp_prepare_pkt(struct bcsp_struct *bcsp, u8 *data,
if (rel) {
hdr[0] |= 0x80 + bcsp->msgq_txseq;
- BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
- bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
+ BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq | ret);
+ ret = ++(bcsp->msgq_txseq) & 0x07;
}
if (bcsp->use_crc)
--
1.7.1.rc1.21.gf3bd6
Im getting a warning message when building with gcc 4.6.0
CC drivers/acpi/glue.o
drivers/acpi/glue.c: In function 'acpi_bind_one':
drivers/acpi/glue.c:163:7: warning: variable 'ret' set but not used
The below fixes it for me. Please have a look when you have
time and let me know(bit confused with the backwardness of
&acpi_dev->dev.kobj, &dev->kobji with these two ret's)
Signed-off-by: Justin P. Mattock <[email protected]>
---
drivers/acpi/glue.c | 3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c
index 4af6301..f146165 100644
--- a/drivers/acpi/glue.c
+++ b/drivers/acpi/glue.c
@@ -166,6 +166,9 @@ static int acpi_bind_one(struct device *dev, acpi_handle handle)
"firmware_node");
ret = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
"physical_node");
+ if (ret) {
+ printk(KERN_WARNING "dev%d: Failed to get exception\n", ret);
+ }
if (acpi_dev->wakeup.flags.valid) {
device_set_wakeup_capable(dev, true);
device_set_wakeup_enable(dev,
--
1.7.1.rc1.21.gf3bd6
gcc is giving me this during compiling:
CC security/selinux/ss/ebitmap.o
security/selinux/ss/ebitmap.c: In function 'ebitmap_netlbl_import':
security/selinux/ss/ebitmap.c:159:27: warning: variable 'e_sft' set but not used
The below fixes this warning for me.
(please check this whenever you have time).
Signed-off-by: Justin P. Mattock <[email protected]>
---
security/selinux/ss/ebitmap.c | 3 +--
1 files changed, 1 insertions(+), 2 deletions(-)
diff --git a/security/selinux/ss/ebitmap.c b/security/selinux/ss/ebitmap.c
index 04b6145..b7980ee 100644
--- a/security/selinux/ss/ebitmap.c
+++ b/security/selinux/ss/ebitmap.c
@@ -156,7 +156,7 @@ int ebitmap_netlbl_import(struct ebitmap *ebmap,
struct ebitmap_node *e_iter = NULL;
struct ebitmap_node *emap_prev = NULL;
struct netlbl_lsm_secattr_catmap *c_iter = catmap;
- u32 c_idx, c_pos, e_idx, e_sft;
+ u32 c_idx, c_pos, e_idx;
/* NetLabel's NETLBL_CATMAP_MAPTYPE is defined as an array of u64,
* however, it is not always compatible with an array of unsigned long
@@ -190,7 +190,6 @@ int ebitmap_netlbl_import(struct ebitmap *ebmap,
}
delta = c_pos - e_iter->startbit;
e_idx = delta / EBITMAP_UNIT_SIZE;
- e_sft = delta % EBITMAP_UNIT_SIZE;
while (map) {
e_iter->maps[e_idx++] |= map & (-1UL);
map = EBITMAP_SHIFT_UNIT_SIZE(map);
--
1.7.1.rc1.21.gf3bd6
Hi Justin,
* Justin P. Mattock <[email protected]> [2010-06-26 23:47:26 -0700]:
> Im seeing this building the kernel with gcc 4.6.0
> CC [M] drivers/bluetooth/hci_bcsp.o
> drivers/bluetooth/hci_bcsp.c: In function 'bcsp_prepare_pkt':
> drivers/bluetooth/hci_bcsp.c:247:20: warning: operation on 'bcsp->msgq_txseq' may be undefined
>
> Hopefully the below is a fix for this. Please let me know.
> Signed-off-by: Justin P. Mattock <[email protected]>
>
> ---
> drivers/bluetooth/hci_bcsp.c | 6 +++---
> 1 files changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/bluetooth/hci_bcsp.c b/drivers/bluetooth/hci_bcsp.c
> index 40aec0f..0f892e7 100644
> --- a/drivers/bluetooth/hci_bcsp.c
> +++ b/drivers/bluetooth/hci_bcsp.c
> @@ -182,7 +182,7 @@ static struct sk_buff *bcsp_prepare_pkt(struct bcsp_struct *bcsp, u8 *data,
> struct sk_buff *nskb;
> u8 hdr[4], chan;
> u16 BCSP_CRC_INIT(bcsp_txmsg_crc);
> - int rel, i;
> + int rel, i, ret;
>
> switch (pkt_type) {
> case HCI_ACLDATA_PKT:
> @@ -243,8 +243,8 @@ static struct sk_buff *bcsp_prepare_pkt(struct bcsp_struct *bcsp, u8 *data,
>
> if (rel) {
> hdr[0] |= 0x80 + bcsp->msgq_txseq;
> - BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
> - bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
> + BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq | ret);
> + ret = ++(bcsp->msgq_txseq) & 0x07;
> }
What are trying to do here? That is completely wrong, you are losting
the next txseq to be sent.
And please do not bother the linux-bluetooth mailing list with patches
to other subsystems. Send them in a way that only Bluetooth patches will
come to linux-bluetooth.
Regards,
--
Gustavo F. Padovan
http://padovan.org
Justin P. Mattock <[email protected]> wrote:
> In this case the safest approach(in my mind) would be to just
> mark the integer __unused. Keep in mind though Im not certain
> if this is the right place for this value i.e. will this effect
> *instkey or not(please check).
This is the wrong approach. Either the variable should be got rid of, or it
should be used to replace all the other calls to key_ref_to_ptr() in
keyctl_describe_key().
David
Justin P. Mattock <[email protected]> wrote:
> gcc is giving me this during compiling:
> CC security/selinux/ss/ebitmap.o
> security/selinux/ss/ebitmap.c: In function 'ebitmap_netlbl_import':
> security/selinux/ss/ebitmap.c:159:27: warning: variable 'e_sft' set but not used
>
> The below fixes this warning for me.
> (please check this whenever you have time).
> Signed-off-by: Justin P. Mattock <[email protected]>
Acked-by: David Howells <[email protected]>
(though I wonder if e_sft should be used for something...)
Justin P. Mattock <[email protected]> wrote:
> + if (ret) {
> + printk(KERN_WARNING "dev%d: Failed to get exception\n", ret);
> + }
That's not a good warning because it's a meaningless string and you're passing
the error number to the dev%d.
Better would perhaps be:
"dev%d: Failed to create physical_node sysfs link: %d\n"
Note also that you're only checking the result of one sysfs_create_link().
You should probably check both.
Also you're introducing a pair of unnecessary braces as there's only one
statement in the if-body.
David
Justin P. Mattock <[email protected]> wrote:
> Im getting this when compiling on gcc 4.6.0
> CC [M] drivers/block/cryptoloop.o
> drivers/block/cryptoloop.c: In function 'cryptoloop_init':
> drivers/block/cryptoloop.c:46:8: warning: variable 'cipher' set but not used
> The below fixes it for me. Please have a look and let me know.
>
> Signed-off-by: Justin P. Mattock <[email protected]>
Acked-by: David Howells <[email protected]>
Justin P. Mattock <[email protected]> wrote:
> - BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
> - bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
> + BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq | ret);
> + ret = ++(bcsp->msgq_txseq) & 0x07;
I don't know what you're trying to do here, but you seem to be trying to send
the computed value back in time.
The problem is that the compiler is confused about why a '++' operator makes
any sense here. It doesn't. It should be a '+ 1' instead. I think what you
want is:
- bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
+ bcsp->msgq_txseq = (bcsp->msgq_txseq + 1) & 0x07;
David
Fix abuse of the preincrement operator as detected when building with gcc
4.6.0:
CC [M] drivers/bluetooth/hci_bcsp.o
drivers/bluetooth/hci_bcsp.c: In function 'bcsp_prepare_pkt':
drivers/bluetooth/hci_bcsp.c:247:20: warning: operation on 'bcsp->msgq_txseq' may be undefined
Reported-by: Justin P. Mattock <[email protected]>
Signed-off-by: David Howells <[email protected]>
---
drivers/bluetooth/hci_bcsp.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/bluetooth/hci_bcsp.c b/drivers/bluetooth/hci_bcsp.c
index 40aec0f..42d69d4 100644
--- a/drivers/bluetooth/hci_bcsp.c
+++ b/drivers/bluetooth/hci_bcsp.c
@@ -244,7 +244,7 @@ static struct sk_buff *bcsp_prepare_pkt(struct bcsp_struct *bcsp, u8 *data,
if (rel) {
hdr[0] |= 0x80 + bcsp->msgq_txseq;
BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
- bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
+ bcsp->msgq_txseq = (bcsp->msgq_txseq + 1) & 0x07;
}
if (bcsp->use_crc)
keyctl_describe_key() turns the key reference it gets into a usable key pointer
and assigns that to a variable called 'key', which it then ignores in favour of
recomputing the key pointer each time it needs it. Make it use the precomputed
pointer instead.
Without this patch, gcc 4.6 reports that the variable key is set but not used:
building with gcc 4.6 I'm getting a warning message:
CC security/keys/keyctl.o
security/keys/keyctl.c: In function 'keyctl_describe_key':
security/keys/keyctl.c:472:14: warning: variable 'key' set but not used
Reported-by: Justin P. Mattock <[email protected]>
Signed-off-by: David Howells <[email protected]>
---
security/keys/keyctl.c | 12 +++++-------
1 files changed, 5 insertions(+), 7 deletions(-)
diff --git a/security/keys/keyctl.c b/security/keys/keyctl.c
index 639226a..b2b0998 100644
--- a/security/keys/keyctl.c
+++ b/security/keys/keyctl.c
@@ -505,13 +505,11 @@ okay:
ret = snprintf(tmpbuf, PAGE_SIZE - 1,
"%s;%d;%d;%08x;%s",
- key_ref_to_ptr(key_ref)->type->name,
- key_ref_to_ptr(key_ref)->uid,
- key_ref_to_ptr(key_ref)->gid,
- key_ref_to_ptr(key_ref)->perm,
- key_ref_to_ptr(key_ref)->description ?
- key_ref_to_ptr(key_ref)->description : ""
- );
+ key->type->name,
+ key->uid,
+ key->gid,
+ key->perm,
+ key->description ?: "");
/* include a NUL char at the end of the data */
if (ret > PAGE_SIZE - 1)
On Mon, 2010-06-28 at 13:52 +0100, David Howells wrote:
> Justin P. Mattock <[email protected]> wrote:
>
> > - BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
> > - bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
> > + BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq | ret);
> > + ret = ++(bcsp->msgq_txseq) & 0x07;
>
> I don't know what you're trying to do here, but you seem to be trying to send
> the computed value back in time.
>
> The problem is that the compiler is confused about why a '++' operator makes
It's even worse as that expression is explicitly undefined (and should
be fixed anyways and unconditionally).
> any sense here. It doesn't. It should be a '+ 1' instead. I think what you
> want is:
>
> - bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
> + bcsp->msgq_txseq = (bcsp->msgq_txseq + 1) & 0x07;
Yes, that's looks like the most probable intention of it - at least for
one who doesn't know the bluetooth code.
Bernd
--
Bernd Petrovitsch Email : [email protected]
LUGA : http://www.luga.at
Hi David,
* David Howells <[email protected]> [2010-06-28 13:57:52 +0100]:
> Fix abuse of the preincrement operator as detected when building with gcc
> 4.6.0:
>
> CC [M] drivers/bluetooth/hci_bcsp.o
> drivers/bluetooth/hci_bcsp.c: In function 'bcsp_prepare_pkt':
> drivers/bluetooth/hci_bcsp.c:247:20: warning: operation on 'bcsp->msgq_txseq' may be undefined
>
> Reported-by: Justin P. Mattock <[email protected]>
> Signed-off-by: David Howells <[email protected]>
> ---
>
> drivers/bluetooth/hci_bcsp.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/bluetooth/hci_bcsp.c b/drivers/bluetooth/hci_bcsp.c
> index 40aec0f..42d69d4 100644
> --- a/drivers/bluetooth/hci_bcsp.c
> +++ b/drivers/bluetooth/hci_bcsp.c
> @@ -244,7 +244,7 @@ static struct sk_buff *bcsp_prepare_pkt(struct bcsp_struct *bcsp, u8 *data,
> if (rel) {
> hdr[0] |= 0x80 + bcsp->msgq_txseq;
> BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
> - bcsp->msgq_txseq = ++(bcsp->msgq_txseq) & 0x07;
> + bcsp->msgq_txseq = (bcsp->msgq_txseq + 1) & 0x07;
> }
>
> if (bcsp->use_crc)
>
Acked-by: Gustavo F. Padovan <[email protected]>
--
Gustavo F. Padovan
http://padovan.org
On 06/28/2010 05:38 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> In this case the safest approach(in my mind) would be to just
>> mark the integer __unused. Keep in mind though Im not certain
>> if this is the right place for this value i.e. will this effect
>> *instkey or not(please check).
>
> This is the wrong approach. Either the variable should be got rid of, or it
> should be used to replace all the other calls to key_ref_to_ptr() in
> keyctl_describe_key().
>
> David
>
I see your patch you sent for this.. vary nice!
Thanks!
Justin P. Mattock
On 06/28/2010 05:44 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> gcc is giving me this during compiling:
>> CC security/selinux/ss/ebitmap.o
>> security/selinux/ss/ebitmap.c: In function 'ebitmap_netlbl_import':
>> security/selinux/ss/ebitmap.c:159:27: warning: variable 'e_sft' set but not used
>>
>> The below fixes this warning for me.
>> (please check this whenever you have time).
>> Signed-off-by: Justin P. Mattock<[email protected]>
>
> Acked-by: David Howells<[email protected]>
>
> (though I wonder if e_sft should be used for something...)
>
alright..
Justin P. Mattock
On 06/28/2010 05:48 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> + if (ret) {
>> + printk(KERN_WARNING "dev%d: Failed to get exception\n", ret);
>> + }
>
> That's not a good warning because it's a meaningless string and you're passing
> the error number to the dev%d.
>
> Better would perhaps be:
>
> "dev%d: Failed to create physical_node sysfs link: %d\n"
>
I'll throw that in.
> Note also that you're only checking the result of one sysfs_create_link().
> You should probably check both.
>
this is where I get confused with: &acpi_dev->dev.kobj, &dev->kobji
it's like a criss cross of code
> Also you're introducing a pair of unnecessary braces as there's only one
> statement in the if-body.
>
> David
>
would just removing ret be good or will things go out of whack because
of no ret
Justin P. Mattock
On 06/28/2010 05:52 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> - BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
>> - bcsp->msgq_txseq = ++(bcsp->msgq_txseq)& 0x07;
>> + BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq | ret);
>> + ret = ++(bcsp->msgq_txseq)& 0x07;
>
> I don't know what you're trying to do here, but you seem to be trying to send
> the computed value back in time.
>
I was under the impression that hdr[0] |= 0x80 + bcsp->msgq_txseq;
is computing a value for BT_DBG then ret = ++(bcsp->msgq_txseq)& 0x07
computes a value as well i.e.
BT_DBG("Sending packet with seqno %u", hdr[0] | ret);
> The problem is that the compiler is confused about why a '++' operator makes
> any sense here. It doesn't. It should be a '+ 1' instead. I think what you
> want is:
>
> - bcsp->msgq_txseq = ++(bcsp->msgq_txseq)& 0x07;
> + bcsp->msgq_txseq = (bcsp->msgq_txseq + 1)& 0x07;
>
> David
>
yeah I did play around with the ++ and noticed the compiler would be
satisfied if the ++ was not there, but didn't think to just add a + 1
Justin P. Mattock
On 06/28/2010 05:57 AM, David Howells wrote:
> Fix abuse of the preincrement operator as detected when building with gcc
> 4.6.0:
>
> CC [M] drivers/bluetooth/hci_bcsp.o
> drivers/bluetooth/hci_bcsp.c: In function 'bcsp_prepare_pkt':
> drivers/bluetooth/hci_bcsp.c:247:20: warning: operation on 'bcsp->msgq_txseq' may be undefined
>
> Reported-by: Justin P. Mattock<[email protected]>
> Signed-off-by: David Howells<[email protected]>
> ---
>
> drivers/bluetooth/hci_bcsp.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/bluetooth/hci_bcsp.c b/drivers/bluetooth/hci_bcsp.c
> index 40aec0f..42d69d4 100644
> --- a/drivers/bluetooth/hci_bcsp.c
> +++ b/drivers/bluetooth/hci_bcsp.c
> @@ -244,7 +244,7 @@ static struct sk_buff *bcsp_prepare_pkt(struct bcsp_struct *bcsp, u8 *data,
> if (rel) {
> hdr[0] |= 0x80 + bcsp->msgq_txseq;
> BT_DBG("Sending packet with seqno %u", bcsp->msgq_txseq);
> - bcsp->msgq_txseq = ++(bcsp->msgq_txseq)& 0x07;
> + bcsp->msgq_txseq = (bcsp->msgq_txseq + 1)& 0x07;
> }
>
> if (bcsp->use_crc)
>
>
ahh.. so it's o.k. to add the value after bcsp->msgq_txseq instead of
before. Anyways build clean over here..
Thanks!
Justin P. Mattock
Justin P. Mattock <[email protected]> wrote:
> would just removing ret be good or will things go out of whack because of no
> ret
Don't you then get warnings for not capturing the result?
David
On 06/28/2010 11:47 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> would just removing ret be good or will things go out of whack because of no
>> ret
>
> Don't you then get warnings for not capturing the result?
>
> David
>
not sure if it would give warnings for not capturing the result.
then doing this would be the way to go.
if (ret) {
printk(KERN_WARNING "dev%d: Failed to create physical_node sysfs link:
%d\n");
}
but like you had said unnecessary braces, and only one statement
I'll look at this some more too see if I come up with anything.
Justin P. Mattock
no need to have extra cc's on this, so I took some people
off since they dont pertain to this area.
Justin P. Mattock
On Mon, 28 Jun 2010, David Howells wrote:
> keyctl_describe_key() turns the key reference it gets into a usable key pointer
> and assigns that to a variable called 'key', which it then ignores in favour of
> recomputing the key pointer each time it needs it. Make it use the precomputed
> pointer instead.
>
> Without this patch, gcc 4.6 reports that the variable key is set but not used:
>
> building with gcc 4.6 I'm getting a warning message:
> CC security/keys/keyctl.o
> security/keys/keyctl.c: In function 'keyctl_describe_key':
> security/keys/keyctl.c:472:14: warning: variable 'key' set but not used
>
> Reported-by: Justin P. Mattock <[email protected]>
> Signed-off-by: David Howells <[email protected]>
Applied to
git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6#next
--
James Morris
<[email protected]>
o.k. after stepping out for a while.. I'm finally sitting down and
looking at this. below is what I came up with.. hopefully it's in the
area/vecinity of what might be a good idea for this(if not then let me know)
From da5cfa463f29ff3fe4af3874649db0809e50ab96 Mon Sep 17 00:00:00 2001
From: Justin P. Mattock <[email protected]>
Date: Mon, 28 Jun 2010 20:14:50 -0700
Subject: [PATCH] glue.c
Signed-off-by: Justin P. Mattock <[email protected]>
---
drivers/acpi/glue.c | 12 +++++++++---
1 files changed, 9 insertions(+), 3 deletions(-)
diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c
index 4af6301..970d7f3 100644
--- a/drivers/acpi/glue.c
+++ b/drivers/acpi/glue.c
@@ -145,6 +145,7 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
{
struct acpi_device *acpi_dev;
acpi_status status;
+ int fn, pn;
if (dev->archdata.acpi_handle) {
dev_warn(dev, "Drivers changed 'acpi_handle'\n");
@@ -160,12 +161,17 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
status = acpi_bus_get_device(handle, &acpi_dev);
if (!ACPI_FAILURE(status)) {
- int ret;
- ret = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
+ fn = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
"firmware_node");
- ret = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
+ pn = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
"physical_node");
+ if (fn) {
+ printk(KERN_WARNING "dev%d: Failed to create firmware_node: %d\n",
status, fn);
+ }else if (pn) {
+ printk(KERN_WARNING "dev%d: Failed to create physical_node: %d\n",
status, pn);
+ return 0;
+ }
if (acpi_dev->wakeup.flags.valid) {
device_set_wakeup_capable(dev, true);
device_set_wakeup_enable(dev,
--
1.7.1.rc1.21.gf3bd6
keep in mind Im not exactly sure what should go into the printk
as for words to say, and functions, so I just used status, fn/pn
for the two because I was getting a not enough function error.
Justin P. Mattock
Justin P. Mattock <[email protected]> wrote:
> + if (fn) {
> + printk(KERN_WARNING "dev%d: Failed to create
> firmware_node: %d\n", status, fn);
> + }else if (pn) {
> + printk(KERN_WARNING "dev%d: Failed to create
> physical_node: %d\n", status, pn);
> + return 0;
> + }
The if-statement should be correctly indented (it's inside another
if-body, so needs to be one more tab over) and there needs to be a space
before the else.
You should probably split your printks up so they don't exceed 80 chars too,
for example:
printk(KERN_WARNING
"dev%d: Failed to create physical_node: %d\n",
status, pn);
Also 'status' is probably the wrong thing to print as the number in "dev%d".
If it worked, that should be unconditionally AE_OK, I think. Can you not use
dev_warn() or similar instead or printk?
David
On 06/29/2010 08:47 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> + if (fn) {
>> + printk(KERN_WARNING "dev%d: Failed to create
>> firmware_node: %d\n", status, fn);
>> + }else if (pn) {
>> + printk(KERN_WARNING "dev%d: Failed to create
>> physical_node: %d\n", status, pn);
>> + return 0;
>> + }
>
> The if-statement should be correctly indented (it's inside another
> if-body, so needs to be one more tab over) and there needs to be a space
> before the else.
>
o.k., so it should look something like this:
if (fn) {
code...
} else if (pn) {
> You should probably split your printks up so they don't exceed 80 chars too,
> for example:
>
> printk(KERN_WARNING
> "dev%d: Failed to create physical_node: %d\n",
> status, pn);
>
> Also 'status' is probably the wrong thing to print as the number in "dev%d".
> If it worked, that should be unconditionally AE_OK, I think. Can you not use
> dev_warn() or similar instead or printk?
>
> David
>
alright, I'll look at this today. I'm not the best at making printks in
fact I'm more intimidated by them..(so with this in mind, I'm going to
sit and make myself learn this, so I atleast have a better idea of doing
these than I have now.)
Justin P. Mattock
On 06/29/2010 08:47 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> + if (fn) {
>> + printk(KERN_WARNING "dev%d: Failed to create
>> firmware_node: %d\n", status, fn);
>> + }else if (pn) {
>> + printk(KERN_WARNING "dev%d: Failed to create
>> physical_node: %d\n", status, pn);
>> + return 0;
>> + }
>
> The if-statement should be correctly indented (it's inside another
> if-body, so needs to be one more tab over) and there needs to be a space
> before the else.
>
> You should probably split your printks up so they don't exceed 80 chars too,
> for example:
>
> printk(KERN_WARNING
> "dev%d: Failed to create physical_node: %d\n",
> status, pn);
>
> Also 'status' is probably the wrong thing to print as the number in "dev%d".
> If it worked, that should be unconditionally AE_OK, I think. Can you not use
> dev_warn() or similar instead or printk?
>
> David
>
o.k. heres another go at this.. this goes through, using dev_warn
but am uncertain about using these three for the right info dev, %p,
dev_acpi
but give it a look whenever you have the time, and let me know
then I'll go from there..
From 34485b5709dc9ad18c57be8c672236580300e05c Mon Sep 17 00:00:00 2001
From: Justin P. Mattock <[email protected]>
Date: Tue, 29 Jun 2010 14:47:42 -0700
Subject: [PATCH ]ACPI:glue.c
Signed-off-by: Justin P. Mattock <[email protected]>
---
drivers/acpi/glue.c | 15 ++++++++++++---
1 files changed, 12 insertions(+), 3 deletions(-)
diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c
index 4af6301..69ca24d 100644
--- a/drivers/acpi/glue.c
+++ b/drivers/acpi/glue.c
@@ -145,6 +145,7 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
{
struct acpi_device *acpi_dev;
acpi_status status;
+ int fn, pn;
if (dev->archdata.acpi_handle) {
dev_warn(dev, "Drivers changed 'acpi_handle'\n");
@@ -160,12 +161,20 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
status = acpi_bus_get_device(handle, &acpi_dev);
if (!ACPI_FAILURE(status)) {
- int ret;
- ret = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
+ fn = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
"firmware_node");
- ret = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
+ pn = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
"physical_node");
+ if (fn) {
+ dev_warn(dev, "dev:%p Failed to create firmware_node: %d\n",
+ acpi_dev, fn);
+
+ } else if (pn) {
+ dev_warn(dev, "dev:%p Failed to create physical_node: %d\n",
+ acpi_dev, pn);
+ return AE_ERROR;
+ }
if (acpi_dev->wakeup.flags.valid) {
device_set_wakeup_capable(dev, true);
device_set_wakeup_enable(dev,
--
1.7.1.rc1.21.gf3bd6
Justin P. Mattock
Justin P. Mattock <[email protected]> wrote:
> if (!ACPI_FAILURE(status)) {
> - int ret;
>
> - ret = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
> + fn = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
> "firmware_node");
> - ret = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
> + pn = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
> "physical_node");
> + if (fn) {
That new if-statement still needs indenting one more tab stop. It's indented
the same as the previous if-statement, but is actually in the body of that
previous if-statement.
The body of the second if-statement should be indented one tab beyond the if,
and else/else-if statements and the final closing brace should be indented
level with the if:
if (...) {
body;
} else if (...) {
body;
} else {
body;
}
so that they line up vertically.
> + dev_warn(dev, "dev:%p Failed to create firmware_node: %d\n",
> + acpi_dev, fn);
The "dev:%p " seems like it ought to be superfluous if you're using
dev_warn(), and certainly, returning the pointer isn't really useful, I
suspect.
However, at this point you have two device struct pointers: dev and
&acip_dev->dev, so printing them both is may be good. Perhaps something like:
+ dev_warn(&acpi_dev->dev,
+ "Failed to create firmware_node link to %s %s: %d\n",
+ dev_driver_string(dev), dev_name(dev), fn);
David
On 06/30/2010 02:13 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> if (!ACPI_FAILURE(status)) {
>> - int ret;
>>
>> - ret = sysfs_create_link(&dev->kobj,&acpi_dev->dev.kobj,
>> + fn = sysfs_create_link(&dev->kobj,&acpi_dev->dev.kobj,
>> "firmware_node");
>> - ret = sysfs_create_link(&acpi_dev->dev.kobj,&dev->kobj,
>> + pn = sysfs_create_link(&acpi_dev->dev.kobj,&dev->kobj,
>> "physical_node");
>> + if (fn) {
>
> That new if-statement still needs indenting one more tab stop. It's indented
> the same as the previous if-statement, but is actually in the body of that
> previous if-statement.
>
> The body of the second if-statement should be indented one tab beyond the if,
> and else/else-if statements and the final closing brace should be indented
> level with the if:
>
> if (...) {
> body;
> } else if (...) {
> body;
> } else {
> body;
> }
>
> so that they line up vertically.
Thanks for the info on this, I really appreciate it. I'll look at this
today, and resend.
>
>> + dev_warn(dev, "dev:%p Failed to create firmware_node: %d\n",
>> + acpi_dev, fn);
>
> The "dev:%p " seems like it ought to be superfluous if you're using
> dev_warn(), and certainly, returning the pointer isn't really useful, I
> suspect.
>
I kept receiving an new warning for using acpi_dev the %p was the only
option I saw in the Documentation that worked
> However, at this point you have two device struct pointers: dev and
> &acip_dev->dev, so printing them both is may be good. Perhaps something like:
>
> + dev_warn(&acpi_dev->dev,
> + "Failed to create firmware_node link to %s %s: %d\n",
> + dev_driver_string(dev), dev_name(dev), fn);
>
> David
>
o.k. I'll look at this today, and see if I can find/locate the device
name and string for these.
Justin P. Mattock
o.k. hopefully this is getting close to being right.. I ended up
spending the later half of the morning looking for
dev_driver_string(dev) until I read device.h and realized you actually
use that it's self to gather the debug info etc..(I admit it, I'm a newbie!)
anyways here is the updated patch, let me know if something need a
changing..
From 16df81551d1899e650ae28ea8d41931f7c391c7a Mon Sep 17 00:00:00 2001
From: Justin P. Mattock <[email protected]>
Date: Wed, 30 Jun 2010 12:34:32 -0700
Subject: [PATCH]acpi:glue.c Fix Fix warning: variable 'ret' set but not used
Fix a warning message generated by gcc:
Im getting a warning message when building with gcc 4.6.0
CC drivers/acpi/glue.o
drivers/acpi/glue.c: In function 'acpi_bind_one':
drivers/acpi/glue.c:163:7: warning: variable 'ret' set but not used
Signed-off-by: Justin P. Mattock <[email protected]>
Signed-off-by: David Howells <[email protected]>
---
drivers/acpi/glue.c | 16 +++++++++++++---
1 files changed, 13 insertions(+), 3 deletions(-)
diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c
index 4af6301..11ad510 100644
--- a/drivers/acpi/glue.c
+++ b/drivers/acpi/glue.c
@@ -145,6 +145,7 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
{
struct acpi_device *acpi_dev;
acpi_status status;
+ int fn, pn;
if (dev->archdata.acpi_handle) {
dev_warn(dev, "Drivers changed 'acpi_handle'\n");
@@ -160,12 +161,21 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
status = acpi_bus_get_device(handle, &acpi_dev);
if (!ACPI_FAILURE(status)) {
- int ret;
- ret = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
+ fn = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
"firmware_node");
- ret = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
+ pn = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
"physical_node");
+ if (fn) {
+ dev_warn(&acpi_dev->dev,
+ "Failed to create firmware_node link to %s %s: %d\n",
+ dev_driver_string(dev), dev_name(dev), fn);
+ } else if (pn) {
+ dev_warn(&acpi_dev->dev,
+ "Failed to create physical_node link to %s %s: %d\n",
+ dev_driver_string(dev), dev_name(dev), pn);
+ return AE_ERROR;
+ }
if (acpi_dev->wakeup.flags.valid) {
device_set_wakeup_capable(dev, true);
device_set_wakeup_enable(dev,
--
1.7.1.rc1.21.gf3bd6
cheers,
Justin P. Mattock
From: "Gustavo F. Padovan" <[email protected]>
Date: Mon, 28 Jun 2010 10:12:30 -0300
> Hi David,
>
> * David Howells <[email protected]> [2010-06-28 13:57:52 +0100]:
>
>> Fix abuse of the preincrement operator as detected when building with gcc
>> 4.6.0:
>>
>> CC [M] drivers/bluetooth/hci_bcsp.o
>> drivers/bluetooth/hci_bcsp.c: In function 'bcsp_prepare_pkt':
>> drivers/bluetooth/hci_bcsp.c:247:20: warning: operation on 'bcsp->msgq_txseq' may be undefined
>>
>> Reported-by: Justin P. Mattock <[email protected]>
>> Signed-off-by: David Howells <[email protected]>
...
> Acked-by: Gustavo F. Padovan <[email protected]>
Applied, thanks everyone.
Justin P. Mattock <[email protected]> wrote:
> + if (fn) {
> + dev_warn(&acpi_dev->dev,
> + "Failed to create firmware_node link to %s %s: %d\n",
> + dev_driver_string(dev), dev_name(dev), fn);
> + } else if (pn) {
> + dev_warn(&acpi_dev->dev,
> + "Failed to create physical_node link to %s %s: %d\n",
> + dev_driver_string(dev), dev_name(dev), pn);
> + return AE_ERROR;
> + }
There's one more question to ask yourself: do you really need two dev_warn()
statements? You could have just one that prints both error values:
if (fn || pn)
dev_warn(&acpi_dev->dev,
"Failed to create link(s) to %s %s:"
" fn=%d pn=%d\n",
dev_driver_string(dev), dev_name(dev),
fn, pn);
Not sure it's worth going that far. You could reduce it still further:
if (fn || pn)
dev_warn(&acpi_dev->dev,
"Failed to create link(s) to %s %s:"
" %d\n",
dev_driver_string(dev), dev_name(dev),
fn ?: pn);
Is it that important to know which failed to be created, or that both failed
to be created?
David
On 07/01/2010 02:31 AM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> + if (fn) {
>> + dev_warn(&acpi_dev->dev,
>> + "Failed to create firmware_node link to %s %s: %d\n",
>> + dev_driver_string(dev), dev_name(dev), fn);
>> + } else if (pn) {
>> + dev_warn(&acpi_dev->dev,
>> + "Failed to create physical_node link to %s %s: %d\n",
>> + dev_driver_string(dev), dev_name(dev), pn);
>> + return AE_ERROR;
>> + }
>
> There's one more question to ask yourself: do you really need two dev_warn()
> statements? You could have just one that prints both error values:
>
> if (fn || pn)
> dev_warn(&acpi_dev->dev,
> "Failed to create link(s) to %s %s:"
> " fn=%d pn=%d\n",
> dev_driver_string(dev), dev_name(dev),
> fn, pn);
ah... I did think about that a few days ago, but had no idea how to
really follow through with this.. and from looking at what you did, it's
as simple as a || b
>
> Not sure it's worth going that far. You could reduce it still further:
>
> if (fn || pn)
> dev_warn(&acpi_dev->dev,
> "Failed to create link(s) to %s %s:"
> " %d\n",
> dev_driver_string(dev), dev_name(dev),
> fn ?: pn);
I don't mind resending with your change to this.
>
> Is it that important to know which failed to be created, or that both failed
> to be created?
>
> David
>
maybe a simple test(prog) case can be created to simulate what this is
doing, just to make sure.
Justin P. Mattock
> Not sure it's worth going that far. You could reduce it still further:
>
> if (fn || pn)
> dev_warn(&acpi_dev->dev,
> "Failed to create link(s) to %s %s:"
> " %d\n",
> dev_driver_string(dev), dev_name(dev),
> fn ?: pn);
>
> Is it that important to know which failed to be created, or that both failed
> to be created?
>
> David
>
I like this..
fn ?: pn (will this give us the results from the above question?
_both failed to be created_)
a bit confused with the whole: "?:" though
*condition ? value if true : value if false* (what if both are true
what if both are false or does it matter?)
here is the patch itself with the change:
Fix a warning message generated by gcc:
Im getting a warning message when building with gcc 4.6.0
CC drivers/acpi/glue.o
drivers/acpi/glue.c: In function 'acpi_bind_one':
drivers/acpi/glue.c:163:7: warning: variable 'ret' set but not used
Signed-off-by: Justin P. Mattock <[email protected]>
Signed-off-by: David Howells <[email protected]>
---
drivers/acpi/glue.c | 14 +++++++++++---
1 files changed, 11 insertions(+), 3 deletions(-)
diff --git a/drivers/acpi/glue.c b/drivers/acpi/glue.c
index 4af6301..23b16e6 100644
--- a/drivers/acpi/glue.c
+++ b/drivers/acpi/glue.c
@@ -145,6 +145,7 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
{
struct acpi_device *acpi_dev;
acpi_status status;
+ int fn, pn;
if (dev->archdata.acpi_handle) {
dev_warn(dev, "Drivers changed 'acpi_handle'\n");
@@ -160,12 +161,19 @@ static int acpi_bind_one(struct device *dev,
acpi_handle handle)
status = acpi_bus_get_device(handle, &acpi_dev);
if (!ACPI_FAILURE(status)) {
- int ret;
- ret = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
+ fn = sysfs_create_link(&dev->kobj, &acpi_dev->dev.kobj,
"firmware_node");
- ret = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
+ pn = sysfs_create_link(&acpi_dev->dev.kobj, &dev->kobj,
"physical_node");
+ if (fn || pn) {
+ dev_warn(&acpi_dev->dev,
+ "Failed to create link(s) to %s %s:"
+ " %d\n",
+ dev_driver_string(dev), dev_name(dev),
+ fn ?: pn);
+ return AE_ERROR;
+ }
if (acpi_dev->wakeup.flags.valid) {
device_set_wakeup_capable(dev, true);
device_set_wakeup_enable(dev,
-- 1.7.1.rc1.21.gf3bd6
Justin P. Mattock
Justin P. Mattock <[email protected]> wrote:
> a bit confused with the whole: "?:" though
> *condition ? value if true : value if false* (what if both are true
> what if both are false or does it matter?)
If you say:
cond ?: false_value
then you'll get cond if cond is non-zero and false_value if it isn't.
I suspect it's a gccism.
David
On 07/01/2010 05:10 PM, David Howells wrote:
> Justin P. Mattock<[email protected]> wrote:
>
>> a bit confused with the whole: "?:" though
>> *condition ? value if true : value if false* (what if both are true
>> what if both are false or does it matter?)
>
> If you say:
>
> cond ?: false_value
>
> then you'll get cond if cond is non-zero and false_value if it isn't.
>
> I suspect it's a gccism.
>
> David
>
I'll have to experiment with this one..
?:
Justin P. Mattock