Re: [PATCH] PCI: Add dummy pci_irqd_intx_xlate() for CONFIG_PCI=n build

[Bjorn Helgaas]

On Fri, Jan 19, 2018 at 11:18:59AM +0000, Lorenzo Pieralisi wrote:
> On Fri, Jan 19, 2018 at 10:39:06AM +0100, Niklas Cassel wrote:
> > If CONFIG_PCI=n and CONFIG_PCI_DRA7XX_EP=y, the build fails with:
> > 
> >   drivers/pci/dwc/pci-dra7xx.c:229:11: error: 'pci_irqd_intx_xlate'
> >   undeclared here (not in a function)
> > 
> > When building drivers/pci/dwc/pci-dra7xx.c without CONFIG_PCI,
> > gcc is usually smart enough to realize that RC specific code
> > is unreachable and can thus be eliminated.
> > 
> > However, gcc cannot do this if there is a function that isn't
> > even declared.
> > 
> > Hence fix the issue by introducing a dummy for pci_irqd_intx_xlate().
> > 
> > Acked-by: Arnd Bergmann <arnd@xxxxxxxx>
> > Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@xxxxxxx>
> > Signed-off-by: Niklas Cassel <niklas.cassel@xxxxxxxx>
> 
> Thanks for the patch.
> 
> I think the commit log should be rewritten (when you write it think at
> someone reading it in some time - what you wrote won't probably make
> any sense) but first, Bjorn are you picking this up ? It does not make
> much sense to squash it in one of my pci/dwc branch commits - therefore
> I am asking, just let me know.

I did apply this, as follows.  I don't know if the commit log makes
any more sense; a lot of the sense is in the eye of the author :)

But anyway, I put this on pci/misc and re-merged my "next" branch so
this is applied before the dra7xx changes, so there shouldn't be any
bisection issue.


commit 80db6f08b7af93eddc9487535e6150b220262637
Author: Niklas Cassel <niklas.cassel@xxxxxxxx>
Date:   Fri Jan 19 10:39:06 2018 +0100

    PCI: Add dummy pci_irqd_intx_xlate() for CONFIG_PCI=n build
    
    Some hardware can operate in either "host" or "endpoint" mode, which means
    there can be both a host bridge driver and an endpoint driver for the same
    device.  Those drivers share a lot of code, so sometimes they live in the
    same source file.
    
    The host bridge driver requires CONFIG_PCI=y because it enumerates PCI
    devices below the bridge using the PCI core.  The endpoint driver does not
    require CONFIG_PCI=y because it runs in an embedded kernel on the other
    side of the device, e.g., on an adapter card.
    
    pci-dra7xx.c contains both host and endpoint drivers.  If we select only
    the endpoint driver (CONFIG_PCI=n and CONFIG_PCI_DRA7XX_EP=y), the unneeded
    host driver is still compiled.  It references pci_irqd_intx_xlate(), which
    is not present when CONFIG_PCI=n, which causes this error:
    
      drivers/pci/dwc/pci-dra7xx.c:229:11: error: 'pci_irqd_intx_xlate' undeclared here (not in a function)
    
    Add a dummy pci_irqd_intx_xlate() for the CONFIG_PCI=n case.
    
    [bhelgaas: changelog]
    Signed-off-by: Niklas Cassel <niklas.cassel@xxxxxxxx>
    Signed-off-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
    Acked-by: Arnd Bergmann <arnd@xxxxxxxx>
    Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@xxxxxxx>

diff --git a/include/linux/pci.h b/include/linux/pci.h
index 6079ab46191f..febe7f653689 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -1686,6 +1686,13 @@ static inline int pci_get_new_domain_nr(void) { return -ENOSYS; }
 #define dev_is_pf(d) (false)
 static inline bool pci_acs_enabled(struct pci_dev *pdev, u16 acs_flags)
 { return false; }
+static inline int pci_irqd_intx_xlate(struct irq_domain *d,
+				      struct device_node *node,
+				      const u32 *intspec,
+				      unsigned int intsize,
+				      unsigned long *out_hwirq,
+				      unsigned int *out_type)
+{ return -EINVAL; }
 #endif /* CONFIG_PCI */
 
 /* Include architecture-dependent settings and functions */