Re: [PATCH] PCI: rcar: Check for OF device match early

[Simon Horman]

On Tue, Jan 31, 2017 at 10:43:39PM +0100, Geert Uytterhoeven wrote:
> Hi Bjorn,
> 
> On Tue, Jan 31, 2017 at 7:09 PM, Bjorn Helgaas  wrote:
> > On Tue, Jan 31, 2017 at 04:33:15PM +0100, Geert Uytterhoeven wrote:
> >> On Tue, Jan 31, 2017 at 4:10 PM, Bjorn Helgaas  wrote:
> >> > A match in the rcar_pcie_of_match[] table is required, so check that 
> >> > first,
> >> > before we start setting up things that need to be undone if it fails.  No
> >> > functional change intended.
> >> >
> >> > Signed-off-by: Bjorn Helgaas 
> >> > ---
> >> >  drivers/pci/host/pcie-rcar.c |   10 +-
> >> >  1 file changed, 5 insertions(+), 5 deletions(-)
> >> >
> >> > diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
> >> > index 0d9b96c3c49d..c91ff0b91be8 100644
> >> > --- a/drivers/pci/host/pcie-rcar.c
> >> > +++ b/drivers/pci/host/pcie-rcar.c
> >> > @@ -1129,6 +1129,10 @@ static int rcar_pcie_probe(struct platform_device 
> >> > *pdev)
> >> > int err;
> >> > int (*hw_init_fn)(struct rcar_pcie *);
> >> >
> >> > +   of_id = of_match_device(rcar_pcie_of_match, dev);
> >> > +   if (!of_id || !of_id->data)
> >> > +   return -EINVAL;
> >> > +
> >>
> >> As this driver is DT-only, none of the above can fail, and you could just 
> >> do
> >>
> >>hw_init_fn = of_device_get_match_data(dev);
> >>
> >> instead, getting rid of of_id completely.
> >
> > Oh, I really like that, thanks for pointing that out!
> >
> > I was about to say that I personally would not check of_id->data for NULL,
> > because it is only NULL if somebody adds an entry to rcar_pcie_of_match
> > without a .data member.  In that case, I'd rather take the NULL pointer
> > dereference than return -EINVAL because it's too easy to ignore the
> > -EINVAL.
> >
> > What do you think about the following?
> 
> Thanks, looks fine!
> 
> > commit 25bd3aa972ee32f04590aa68b2b785dce36b036a
> > Author: Bjorn Helgaas 
> > Date:   Tue Jan 31 08:45:49 2017 -0600
> >
> > PCI: rcar: Use of_device_get_match_data() to simplify probe
> >
> > This is a DT-only driver, so the only way to call rcar_pcie_probe() is 
> > to
> > match an entry in rcar_pcie_of_match[], so of_id cannot be NULL.
> >
> > Furthermore, of_id->data can only be NULL if an rcar_pcie_of_match[] 
> > entry
> > has a NULL .data member.  That's a driver defect, and we don't want to
> > return -EINVAL, which is easy to ignore.  We'd rather take the NULL 
> > pointer
> > dereference so we notice the problem and fix it.
> >
> > Use of_device_get_match_data() to retrieve the hw_init_fn pointer.  No
> > functional change intended.
> >
> > Suggested-by: Geert Uytterhoeven 
> > Signed-off-by: Bjorn Helgaas 
> 
> Reviewed-by: Geert Uytterhoeven 

Acked-by: Simon Horman 

Re: [PATCH] PCI: rcar: Check for OF device match early

[Bjorn Helgaas]

On Tue, Jan 31, 2017 at 04:33:15PM +0100, Geert Uytterhoeven wrote:
> Hi Bjorn,
> 
> On Tue, Jan 31, 2017 at 4:10 PM, Bjorn Helgaas  wrote:
> > A match in the rcar_pcie_of_match[] table is required, so check that first,
> > before we start setting up things that need to be undone if it fails.  No
> > functional change intended.
> >
> > Signed-off-by: Bjorn Helgaas 
> > ---
> >  drivers/pci/host/pcie-rcar.c |   10 +-
> >  1 file changed, 5 insertions(+), 5 deletions(-)
> >
> > diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
> > index 0d9b96c3c49d..c91ff0b91be8 100644
> > --- a/drivers/pci/host/pcie-rcar.c
> > +++ b/drivers/pci/host/pcie-rcar.c
> > @@ -1129,6 +1129,10 @@ static int rcar_pcie_probe(struct platform_device 
> > *pdev)
> > int err;
> > int (*hw_init_fn)(struct rcar_pcie *);
> >
> > +   of_id = of_match_device(rcar_pcie_of_match, dev);
> > +   if (!of_id || !of_id->data)
> > +   return -EINVAL;
> > +
> 
> As this driver is DT-only, none of the above can fail, and you could just do
> 
>hw_init_fn = of_device_get_match_data(dev);
> 
> instead, getting rid of of_id completely.

Oh, I really like that, thanks for pointing that out!

I was about to say that I personally would not check of_id->data for NULL,
because it is only NULL if somebody adds an entry to rcar_pcie_of_match
without a .data member.  In that case, I'd rather take the NULL pointer
dereference than return -EINVAL because it's too easy to ignore the
-EINVAL.

What do you think about the following?


commit 25bd3aa972ee32f04590aa68b2b785dce36b036a
Author: Bjorn Helgaas 
Date:   Tue Jan 31 08:45:49 2017 -0600

PCI: rcar: Use of_device_get_match_data() to simplify probe

This is a DT-only driver, so the only way to call rcar_pcie_probe() is to
match an entry in rcar_pcie_of_match[], so of_id cannot be NULL.

Furthermore, of_id->data can only be NULL if an rcar_pcie_of_match[] entry
has a NULL .data member.  That's a driver defect, and we don't want to
return -EINVAL, which is easy to ignore.  We'd rather take the NULL pointer
dereference so we notice the problem and fix it.

Use of_device_get_match_data() to retrieve the hw_init_fn pointer.  No
functional change intended.

Suggested-by: Geert Uytterhoeven 
Signed-off-by: Bjorn Helgaas 

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index 0d9b96c3c49d..cb07c45c1858 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -1125,7 +1125,6 @@ static int rcar_pcie_probe(struct platform_device *pdev)
struct device *dev = >dev;
struct rcar_pcie *pcie;
unsigned int data;
-   const struct of_device_id *of_id;
int err;
int (*hw_init_fn)(struct rcar_pcie *);
 
@@ -1149,11 +1148,6 @@ static int rcar_pcie_probe(struct platform_device *pdev)
if (err)
return err;
 
-   of_id = of_match_device(rcar_pcie_of_match, dev);
-   if (!of_id || !of_id->data)
-   return -EINVAL;
-   hw_init_fn = of_id->data;
-
pm_runtime_enable(dev);
err = pm_runtime_get_sync(dev);
if (err < 0) {
@@ -1162,6 +1156,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
}
 
/* Failure to get a link might just be that no cards are inserted */
+   hw_init_fn = of_device_get_match_data(dev);
err = hw_init_fn(pcie);
if (err) {
dev_info(dev, "PCIe link down\n");

[PATCH] PCI: rcar: Check for OF device match early

[Bjorn Helgaas]

A match in the rcar_pcie_of_match[] table is required, so check that first,
before we start setting up things that need to be undone if it fails.  No
functional change intended.

Signed-off-by: Bjorn Helgaas 
---
 drivers/pci/host/pcie-rcar.c |   10 +-
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index 0d9b96c3c49d..c91ff0b91be8 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -1129,6 +1129,10 @@ static int rcar_pcie_probe(struct platform_device *pdev)
int err;
int (*hw_init_fn)(struct rcar_pcie *);
 
+   of_id = of_match_device(rcar_pcie_of_match, dev);
+   if (!of_id || !of_id->data)
+   return -EINVAL;
+
pcie = devm_kzalloc(dev, sizeof(*pcie), GFP_KERNEL);
if (!pcie)
return -ENOMEM;
@@ -1149,11 +1153,6 @@ static int rcar_pcie_probe(struct platform_device *pdev)
if (err)
return err;
 
-   of_id = of_match_device(rcar_pcie_of_match, dev);
-   if (!of_id || !of_id->data)
-   return -EINVAL;
-   hw_init_fn = of_id->data;
-
pm_runtime_enable(dev);
err = pm_runtime_get_sync(dev);
if (err < 0) {
@@ -1162,6 +1161,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
}
 
/* Failure to get a link might just be that no cards are inserted */
+   hw_init_fn = of_id->data;
err = hw_init_fn(pcie);
if (err) {
dev_info(dev, "PCIe link down\n");