| Message ID | 20190404084204.GA20193@kadam (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | remoteproc: silence an uninitialized variable warning | expand |
On 4/4/2019 2:12 PM, Dan Carpenter wrote: > Smatch complains that "ret" might be uninitialized. I can see why it > generates the warning, but I don't know if it's actually possible. > Anyway initializing "ret" here is harmless. > > Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> > --- > drivers/soc/qcom/mdt_loader.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c > index 1c488024c698..fc58d660692f 100644 > --- a/drivers/soc/qcom/mdt_loader.c > +++ b/drivers/soc/qcom/mdt_loader.c > @@ -188,7 +188,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw, > > if (reloc_base) > *reloc_base = mem_reloc; > - > + ret = 0; You are overriding the value here, better keep it at the start. Thanks, Mukesh > out: > kfree(fw_name); >
On Thu, Apr 04, 2019 at 02:47:39PM +0530, Mukesh Ojha wrote: > > On 4/4/2019 2:12 PM, Dan Carpenter wrote: > > Smatch complains that "ret" might be uninitialized. I can see why it > > generates the warning, but I don't know if it's actually possible. > > Anyway initializing "ret" here is harmless. > > > > Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> > > --- > > drivers/soc/qcom/mdt_loader.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c > > index 1c488024c698..fc58d660692f 100644 > > --- a/drivers/soc/qcom/mdt_loader.c > > +++ b/drivers/soc/qcom/mdt_loader.c > > @@ -188,7 +188,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw, > > if (reloc_base) > > *reloc_base = mem_reloc; > > - > > + ret = 0; > > > You are overriding the value here, better keep it at the start. I like how I wrote it. It makes it clear that this is the success path. regards, dan carpenter
On 4/4/2019 3:06 PM, Dan Carpenter wrote: > On Thu, Apr 04, 2019 at 02:47:39PM +0530, Mukesh Ojha wrote: >> On 4/4/2019 2:12 PM, Dan Carpenter wrote: >>> Smatch complains that "ret" might be uninitialized. I can see why it >>> generates the warning, but I don't know if it's actually possible. >>> Anyway initializing "ret" here is harmless. >>> >>> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> >>> --- >>> drivers/soc/qcom/mdt_loader.c | 2 +- >>> 1 file changed, 1 insertion(+), 1 deletion(-) >>> >>> diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c >>> index 1c488024c698..fc58d660692f 100644 >>> --- a/drivers/soc/qcom/mdt_loader.c >>> +++ b/drivers/soc/qcom/mdt_loader.c >>> @@ -188,7 +188,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw, >>> if (reloc_base) >>> *reloc_base = mem_reloc; >>> - >>> + ret = 0; >> >> You are overriding the value here, better keep it at the start. > I like how I wrote it. It makes it clear that this is the success path. But think about a case when this request_firmware_into_buf fails or the snippet above that can set the ret with error value right.. is not that possible?..and ret = 0 stops propagating the error properly. Thanks Mukesh > > regards, > dan carpenter >
On Fri, Apr 05, 2019 at 12:38:19PM +0530, Mukesh Ojha wrote: > > On 4/4/2019 3:06 PM, Dan Carpenter wrote: > > On Thu, Apr 04, 2019 at 02:47:39PM +0530, Mukesh Ojha wrote: > > > On 4/4/2019 2:12 PM, Dan Carpenter wrote: > > > > Smatch complains that "ret" might be uninitialized. I can see why it > > > > generates the warning, but I don't know if it's actually possible. > > > > Anyway initializing "ret" here is harmless. > > > > > > > > Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> > > > > --- > > > > drivers/soc/qcom/mdt_loader.c | 2 +- > > > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > > > > > diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c > > > > index 1c488024c698..fc58d660692f 100644 > > > > --- a/drivers/soc/qcom/mdt_loader.c > > > > +++ b/drivers/soc/qcom/mdt_loader.c > > > > @@ -188,7 +188,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw, > > > > if (reloc_base) > > > > *reloc_base = mem_reloc; > > > > - > > > > + ret = 0; > > > > > > You are overriding the value here, better keep it at the start. > > I like how I wrote it. It makes it clear that this is the success path. > > > But think about a case when this request_firmware_into_buf fails or the > snippet above that can set the ret with error value > > right.. is not that possible?..and ret = 0 stops propagating the error > properly. > Gar. I read "goto out;" instead of "break;". It probably should be goto out though... Anyway, I will resend. Thanks! regards, dan carpenter
diff --git a/drivers/soc/qcom/mdt_loader.c b/drivers/soc/qcom/mdt_loader.c index 1c488024c698..fc58d660692f 100644 --- a/drivers/soc/qcom/mdt_loader.c +++ b/drivers/soc/qcom/mdt_loader.c @@ -188,7 +188,7 @@ static int __qcom_mdt_load(struct device *dev, const struct firmware *fw, if (reloc_base) *reloc_base = mem_reloc; - + ret = 0; out: kfree(fw_name);
Smatch complains that "ret" might be uninitialized. I can see why it generates the warning, but I don't know if it's actually possible. Anyway initializing "ret" here is harmless. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> --- drivers/soc/qcom/mdt_loader.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)