2008-12-10 16:15:00

by Hannes Eder

Subject: [PATCH 4/9] mISDN: fix sparse warnings

Impact: make an exported symbol non static

Fix this warning:

drivers/isdn/hardware/mISDN/hfcmulti.c:176:1: warning: symbol 'plx_lock' shadows an earlier one
drivers/isdn/hardware/mISDN/hfcmulti.c:175:19: originally declared here

Signed-off-by: Hannes Eder <[email protected]>
---

Note that patch introduces this new warning:

drivers/isdn/hardware/mISDN/hfcmulti.c:175:12: warning: symbol 'plx_lock' was not declared. Should it be static?

Should 'plx_lock' be declared in a header file? Or is it really
neccessart to export it?


drivers/isdn/hardware/mISDN/hfcmulti.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/isdn/hardware/mISDN/hfcmulti.c b/drivers/isdn/hardware/mISDN/hfcmulti.c
index 599d832..b4595f7 100644
--- a/drivers/isdn/hardware/mISDN/hfcmulti.c
+++ b/drivers/isdn/hardware/mISDN/hfcmulti.c
@@ -172,7 +172,7 @@ static int interrupt_registered;

static struct hfc_multi *syncmaster;
static int plxsd_master; /* if we have a master card (yet) */
-static spinlock_t plx_lock; /* may not acquire other lock inside */
+spinlock_t plx_lock; /* may not acquire other lock inside */
EXPORT_SYMBOL(plx_lock);

#define TYP_E1 1
--
1.5.6.3

2008-12-10 22:01:28

by Jiri Slaby

Subject: Re: [PATCH 4/9] mISDN: fix sparse warnings

On 12/10/2008 04:13 PM, Hannes Eder wrote:
> Should 'plx_lock' be declared in a header file? Or is it really
> neccessart to export it?

As there is no other user, I vote for unexporting it.

2008-12-11 20:05:55

by Hannes Eder

Subject: Re: [PATCHv2 4/9] mISDN: fix sparse warning

Impact: un-export symbol 'plx_lock'

Fix this warning:

drivers/isdn/hardware/mISDN/hfcmulti.c:176:1: warning: symbol 'plx_lock' shadows an earlier one
drivers/isdn/hardware/mISDN/hfcmulti.c:175:19: originally declared here

Signed-off-by: Hannes Eder <[email protected]>
---

On Wed, Dec 10, 2008 at 11:01 PM, Jiri Slaby <[email protected]> wrote:
> On 12/10/2008 04:13 PM, Hannes Eder wrote:
>> Should 'plx_lock' be declared in a header file? Or is it really
>> neccessart to export it?
>
> As there is no other user, I vote for unexporting it.

In that case patch number 4 looks like this:

drivers/isdn/hardware/mISDN/hfcmulti.c | 1 -
1 files changed, 0 insertions(+), 1 deletions(-)

diff --git a/drivers/isdn/hardware/mISDN/hfcmulti.c b/drivers/isdn/hardware/mISDN/hfcmulti.c
index 599d832..970201d 100644
--- a/drivers/isdn/hardware/mISDN/hfcmulti.c
+++ b/drivers/isdn/hardware/mISDN/hfcmulti.c
@@ -173,7 +173,6 @@ static int interrupt_registered;
static struct hfc_multi *syncmaster;
static int plxsd_master; /* if we have a master card (yet) */
static spinlock_t plx_lock; /* may not acquire other lock inside */
-EXPORT_SYMBOL(plx_lock);

#define TYP_E1 1
#define TYP_4S 4
--
1.5.6.3