| Message ID | 20211224192626.15843-3-prabhakar.mahadev-lad.rj@bp.renesas.com (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | net: Use platform_get_irq*() variants to fetch IRQ's | expand |
Hi Andy, Thank you for the review. On Sat, Dec 25, 2021 at 11:24 AM Andy Shevchenko <andy.shevchenko@gmail.com> wrote: > > > > On Friday, December 24, 2021, Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> wrote: >> >> platform_get_resource(pdev, IORESOURCE_IRQ, ..) relies on static >> allocation of IRQ resources in DT core code, this causes an issue >> when using hierarchical interrupt domains using "interrupts" property >> in the node as this bypasses the hierarchical setup and messes up the >> irq chaining. >> >> In preparation for removal of static setup of IRQ resource from DT core >> code use platform_get_irq(). >> >> Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> >> --- >> drivers/net/ethernet/marvell/pxa168_eth.c | 9 +++++---- >> 1 file changed, 5 insertions(+), 4 deletions(-) >> >> diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c >> index 1d607bc6b59e..52bef50f5a0d 100644 >> --- a/drivers/net/ethernet/marvell/pxa168_eth.c >> +++ b/drivers/net/ethernet/marvell/pxa168_eth.c >> @@ -1388,7 +1388,6 @@ static int pxa168_eth_probe(struct platform_device *pdev) >> { >> struct pxa168_eth_private *pep = NULL; >> struct net_device *dev = NULL; >> - struct resource *res; >> struct clk *clk; >> struct device_node *np; >> int err; >> @@ -1419,9 +1418,11 @@ static int pxa168_eth_probe(struct platform_device *pdev) >> goto err_netdev; >> } >> >> - res = platform_get_resource(pdev, IORESOURCE_IRQ, 0); >> - BUG_ON(!res); >> - dev->irq = res->start; >> + err = platform_get_irq(pdev, 0); >> + if (err == -EPROBE_DEFER) > > > What about other errors? > Ouch I missed it... > >> >> + goto err_netdev; >> + BUG_ON(dev->irq < 0); > > > ??? What is this and how it supposed to work? > .. should have been BUG_ON(dev->irq < 0); Cheers, Prabhakar >> >> + dev->irq = err; >> dev->netdev_ops = &pxa168_eth_netdev_ops; >> dev->watchdog_timeo = 2 * HZ; >> dev->base_addr = 0; >> -- >> 2.17.1 >> > > > -- > With Best Regards, > Andy Shevchenko > >
> >> + goto err_netdev; > >> + BUG_ON(dev->irq < 0); > > > > > > ??? What is this and how it supposed to work? > > > .. should have been BUG_ON(dev->irq < 0); Is this fatal to the machine as a whole, now way to recover, all that can be done is to limit the damage while it explodes? If not, please use WARN_ON(), not BUG_ON(). There is an email from Linus about this, not using BUG_ON() in general. Andrew
On 25.12.2021 13:19, Lad, Prabhakar wrote: > Hi Andy, > > Thank you for the review. > > On Sat, Dec 25, 2021 at 11:24 AM Andy Shevchenko > <andy.shevchenko@gmail.com> wrote: >> >> >> >> On Friday, December 24, 2021, Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> wrote: >>> >>> platform_get_resource(pdev, IORESOURCE_IRQ, ..) relies on static >>> allocation of IRQ resources in DT core code, this causes an issue >>> when using hierarchical interrupt domains using "interrupts" property >>> in the node as this bypasses the hierarchical setup and messes up the >>> irq chaining. >>> >>> In preparation for removal of static setup of IRQ resource from DT core >>> code use platform_get_irq(). >>> >>> Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> >>> --- >>> drivers/net/ethernet/marvell/pxa168_eth.c | 9 +++++---- >>> 1 file changed, 5 insertions(+), 4 deletions(-) >>> >>> diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c >>> index 1d607bc6b59e..52bef50f5a0d 100644 >>> --- a/drivers/net/ethernet/marvell/pxa168_eth.c >>> +++ b/drivers/net/ethernet/marvell/pxa168_eth.c >>> @@ -1388,7 +1388,6 @@ static int pxa168_eth_probe(struct platform_device *pdev) >>> { >>> struct pxa168_eth_private *pep = NULL; >>> struct net_device *dev = NULL; >>> - struct resource *res; >>> struct clk *clk; >>> struct device_node *np; >>> int err; >>> @@ -1419,9 +1418,11 @@ static int pxa168_eth_probe(struct platform_device *pdev) >>> goto err_netdev; >>> } >>> >>> - res = platform_get_resource(pdev, IORESOURCE_IRQ, 0); >>> - BUG_ON(!res); >>> - dev->irq = res->start; >>> + err = platform_get_irq(pdev, 0); >>> + if (err == -EPROBE_DEFER) >> >> >> What about other errors? >> > Ouch I missed it... >> >>> >>> + goto err_netdev; >>> + BUG_ON(dev->irq < 0); >> >> >> ??? What is this and how it supposed to work? >> > .. should have been BUG_ON(dev->irq < 0); Usage of BUG_ON() is discouraged. Better handle the error w/o stopping the whole system. > > Cheers, > Prabhakar >>> >>> + dev->irq = err; >>> dev->netdev_ops = &pxa168_eth_netdev_ops; >>> dev->watchdog_timeo = 2 * HZ; >>> dev->base_addr = 0; >>> -- >>> 2.17.1 >>> >> >> >> -- >> With Best Regards, >> Andy Shevchenko >> >>
Hi Heiner, Thank you for the review. On Sun, Dec 26, 2021 at 11:15 AM Heiner Kallweit <hkallweit1@gmail.com> wrote: > > On 25.12.2021 13:19, Lad, Prabhakar wrote: > > Hi Andy, > > > > Thank you for the review. > > > > On Sat, Dec 25, 2021 at 11:24 AM Andy Shevchenko > > <andy.shevchenko@gmail.com> wrote: > >> > >> > >> > >> On Friday, December 24, 2021, Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> wrote: > >>> > >>> platform_get_resource(pdev, IORESOURCE_IRQ, ..) relies on static > >>> allocation of IRQ resources in DT core code, this causes an issue > >>> when using hierarchical interrupt domains using "interrupts" property > >>> in the node as this bypasses the hierarchical setup and messes up the > >>> irq chaining. > >>> > >>> In preparation for removal of static setup of IRQ resource from DT core > >>> code use platform_get_irq(). > >>> > >>> Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> > >>> --- > >>> drivers/net/ethernet/marvell/pxa168_eth.c | 9 +++++---- > >>> 1 file changed, 5 insertions(+), 4 deletions(-) > >>> > >>> diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c > >>> index 1d607bc6b59e..52bef50f5a0d 100644 > >>> --- a/drivers/net/ethernet/marvell/pxa168_eth.c > >>> +++ b/drivers/net/ethernet/marvell/pxa168_eth.c > >>> @@ -1388,7 +1388,6 @@ static int pxa168_eth_probe(struct platform_device *pdev) > >>> { > >>> struct pxa168_eth_private *pep = NULL; > >>> struct net_device *dev = NULL; > >>> - struct resource *res; > >>> struct clk *clk; > >>> struct device_node *np; > >>> int err; > >>> @@ -1419,9 +1418,11 @@ static int pxa168_eth_probe(struct platform_device *pdev) > >>> goto err_netdev; > >>> } > >>> > >>> - res = platform_get_resource(pdev, IORESOURCE_IRQ, 0); > >>> - BUG_ON(!res); > >>> - dev->irq = res->start; > >>> + err = platform_get_irq(pdev, 0); > >>> + if (err == -EPROBE_DEFER) > >> > >> > >> What about other errors? > >> > > Ouch I missed it... > >> > >>> > >>> + goto err_netdev; > >>> + BUG_ON(dev->irq < 0); > >> > >> > >> ??? What is this and how it supposed to work? > >> > > > .. should have been BUG_ON(dev->irq < 0); > > Usage of BUG_ON() is discouraged. Better handle the error w/o stopping > the whole system. > Agreed, if everyone is OK i'll create a separate patch for removal of BUG_ON() and the later patch for using platform_get_irq(). Cheers, Prabhakar
diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c index 1d607bc6b59e..52bef50f5a0d 100644 --- a/drivers/net/ethernet/marvell/pxa168_eth.c +++ b/drivers/net/ethernet/marvell/pxa168_eth.c @@ -1388,7 +1388,6 @@ static int pxa168_eth_probe(struct platform_device *pdev) { struct pxa168_eth_private *pep = NULL; struct net_device *dev = NULL; - struct resource *res; struct clk *clk; struct device_node *np; int err; @@ -1419,9 +1418,11 @@ static int pxa168_eth_probe(struct platform_device *pdev) goto err_netdev; } - res = platform_get_resource(pdev, IORESOURCE_IRQ, 0); - BUG_ON(!res); - dev->irq = res->start; + err = platform_get_irq(pdev, 0); + if (err == -EPROBE_DEFER) + goto err_netdev; + BUG_ON(dev->irq < 0); + dev->irq = err; dev->netdev_ops = &pxa168_eth_netdev_ops; dev->watchdog_timeo = 2 * HZ; dev->base_addr = 0;
platform_get_resource(pdev, IORESOURCE_IRQ, ..) relies on static allocation of IRQ resources in DT core code, this causes an issue when using hierarchical interrupt domains using "interrupts" property in the node as this bypasses the hierarchical setup and messes up the irq chaining. In preparation for removal of static setup of IRQ resource from DT core code use platform_get_irq(). Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> --- drivers/net/ethernet/marvell/pxa168_eth.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)