2015-09-28 06:29:07

by Mutharaju, Prasanna (P.)

Subject: [PATCH] Bluetooth:hci_qca: Changed unsigned long to bool

From: Prasanna Karthik <[email protected]>

'retransmit' being set in HCI_IBS_TX_WAKING case,
using bool would be efficient.

Signed-off-by: Prasanna Karthik <[email protected]>
---
drivers/bluetooth/hci_qca.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index 21f4ea4..3fdb17d 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -347,7 +347,7 @@ static void hci_ibs_wake_retrans_timeout(unsigned long =
arg)
struct hci_uart *hu =3D (struct hci_uart *)arg;
struct qca_data *qca =3D hu->priv;
unsigned long flags, retrans_delay;
- unsigned long retransmit =3D 0;
+ bool retransmit;
=20
BT_DBG("hu %p wake retransmit timeout in %d state",
hu, qca->tx_ibs_state);
@@ -358,7 +358,7 @@ static void hci_ibs_wake_retrans_timeout(unsigned long =
arg)
switch (qca->tx_ibs_state) {
case HCI_IBS_TX_WAKING:
/* No WAKE_ACK, retransmit WAKE */
- retransmit =3D 1;
+ retransmit =3D true;
if (send_hci_ibs_cmd(HCI_IBS_WAKE_IND, hu) < 0) {
BT_ERR("Failed to acknowledge device wake up");
break;
--=20
1.9.1

2015-09-28 07:58:52

by Mutharaju, Prasanna (P.)

Subject: Re: [PATCH] Bluetooth:hci_qca: Changed unsigned long to bool

On Mon, Sep 28, 2015 at 09:15:47AM +0200, Szymon Janc wrote:
> Hi,
>=20
> On Monday 28 of September 2015 06:29:07 Mutharaju, Prasanna wrote:
> > From: Prasanna Karthik <[email protected]>
> >=20
> > 'retransmit' being set in HCI_IBS_TX_WAKING case,
> > using bool would be efficient.
> >=20
> > Signed-off-by: Prasanna Karthik <[email protected]>
> > ---
> > drivers/bluetooth/hci_qca.c | 4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >=20
> > diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
> > index 21f4ea4..3fdb17d 100644
> > --- a/drivers/bluetooth/hci_qca.c
> > +++ b/drivers/bluetooth/hci_qca.c
> > @@ -347,7 +347,7 @@ static void hci_ibs_wake_retrans_timeout(unsigned l=
ong
> > arg) struct hci_uart *hu =3D (struct hci_uart *)arg;
> > struct qca_data *qca =3D hu->priv;
> > unsigned long flags, retrans_delay;
> > - unsigned long retransmit =3D 0;
> > + bool retransmit;
> >=20
> > BT_DBG("hu %p wake retransmit timeout in %d state",
> > hu, qca->tx_ibs_state);
> > @@ -358,7 +358,7 @@ static void hci_ibs_wake_retrans_timeout(unsigned l=
ong
> > arg) switch (qca->tx_ibs_state) {
> > case HCI_IBS_TX_WAKING:
> > /* No WAKE_ACK, retransmit WAKE */
> > - retransmit =3D 1;
> > + retransmit =3D true;
> > if (send_hci_ibs_cmd(HCI_IBS_WAKE_IND, hu) < 0) {
> > BT_ERR("Failed to acknowledge device wake up");
> > break;
>=20
> Isn't this patch giving you a warning about retransmit being used=20
> uninitialized? (it is never set to false)
>=20
> --=20
> BR
> Szymon Janc

Nope, It doesn't show compiler warning, bool will be initialized to false
by default by the compiler. I will send a V2 of the patch with retransmit
set to false if needed. =

2015-09-28 07:15:47

by Szymon Janc

Subject: Re: [PATCH] Bluetooth:hci_qca: Changed unsigned long to bool

Hi,

On Monday 28 of September 2015 06:29:07 Mutharaju, Prasanna wrote:
> From: Prasanna Karthik <[email protected]>
>
> 'retransmit' being set in HCI_IBS_TX_WAKING case,
> using bool would be efficient.
>
> Signed-off-by: Prasanna Karthik <[email protected]>
> ---
> drivers/bluetooth/hci_qca.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
> index 21f4ea4..3fdb17d 100644
> --- a/drivers/bluetooth/hci_qca.c
> +++ b/drivers/bluetooth/hci_qca.c
> @@ -347,7 +347,7 @@ static void hci_ibs_wake_retrans_timeout(unsigned long
> arg) struct hci_uart *hu = (struct hci_uart *)arg;
> struct qca_data *qca = hu->priv;
> unsigned long flags, retrans_delay;
> - unsigned long retransmit = 0;
> + bool retransmit;
>
> BT_DBG("hu %p wake retransmit timeout in %d state",
> hu, qca->tx_ibs_state);
> @@ -358,7 +358,7 @@ static void hci_ibs_wake_retrans_timeout(unsigned long
> arg) switch (qca->tx_ibs_state) {
> case HCI_IBS_TX_WAKING:
> /* No WAKE_ACK, retransmit WAKE */
> - retransmit = 1;
> + retransmit = true;
> if (send_hci_ibs_cmd(HCI_IBS_WAKE_IND, hu) < 0) {
> BT_ERR("Failed to acknowledge device wake up");
> break;

Isn't this patch giving you a warning about retransmit being used
uninitialized? (it is never set to false)

--
BR
Szymon Janc