| Message ID | 20191115135842.119621-1-wei.liu@kernel.org |
|---|---|
| State | New |
| Delegated to: | Bjorn Helgaas |
| Headers | show |
| Series | PCI: build Broadcom PAXC quirks unconditionally | expand |
On Fri, 15 Nov 2019 at 13:58, Wei Liu <wei.liu@kernel.org> wrote: > > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > in. Removing the ifdef will allow us to build the driver as a module. > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > --- > Alternatively, we can change the condition to: > > #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE > . > > I chose to remove the ifdef because that's what other quirks looked like > in this file. Bjorn and Ray, any comment on this trivial patch? Wei.
Hi Wei, On 11/15/19 5:58 AM, Wei Liu wrote: > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > in. Removing the ifdef will allow us to build the driver as a module. > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > --- > Alternatively, we can change the condition to: > > #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE > . > > I chose to remove the ifdef because that's what other quirks looked like > in this file. > --- > drivers/pci/quirks.c | 2 -- > 1 file changed, 2 deletions(-) > > diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c > index 320255e5e8f8..cd0e7c18e717 100644 > --- a/drivers/pci/quirks.c > +++ b/drivers/pci/quirks.c > @@ -2381,7 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM, > PCI_DEVICE_ID_TIGON3_5719, > quirk_brcm_5719_limit_mrrs); > > -#ifdef CONFIG_PCIE_IPROC_PLATFORM > static void quirk_paxc_bridge(struct pci_dev *pdev) > { > /* > @@ -2405,7 +2404,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge); > -#endif > > /* > * Originally in EDAC sources for i82875P: Intel tells BIOS developers to > The change looks good to me. Thanks! Reviewed-by: Ray Jui <ray.jui@broadcom.com>
On 11/25/19 6:55 AM, Wei Liu wrote: > On Fri, 15 Nov 2019 at 13:58, Wei Liu <wei.liu@kernel.org> wrote: >> >> CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built >> in. Removing the ifdef will allow us to build the driver as a module. >> >> Signed-off-by: Wei Liu <wei.liu@kernel.org> >> --- >> Alternatively, we can change the condition to: >> >> #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE >> . >> >> I chose to remove the ifdef because that's what other quirks looked like >> in this file. > > Bjorn and Ray, any comment on this trivial patch? > > Wei. > I just reviewed it and it looks good to me. Thanks! Ray
On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > in. Removing the ifdef will allow us to build the driver as a module. > > Signed-off-by: Wei Liu <wei.liu@kernel.org> Sorry, I missed this thinking it would be under drivers/pci/controller and hence handled by Lorenzo. So I guess this doesn't fix a build problem, but without this patch, we just don't run the quirk if the driver is a module, right? > --- > Alternatively, we can change the condition to: > > #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE > > I chose to remove the ifdef because that's what other quirks looked like > in this file. > --- > drivers/pci/quirks.c | 2 -- > 1 file changed, 2 deletions(-) > > diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c > index 320255e5e8f8..cd0e7c18e717 100644 > --- a/drivers/pci/quirks.c > +++ b/drivers/pci/quirks.c > @@ -2381,7 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM, > PCI_DEVICE_ID_TIGON3_5719, > quirk_brcm_5719_limit_mrrs); > > -#ifdef CONFIG_PCIE_IPROC_PLATFORM > static void quirk_paxc_bridge(struct pci_dev *pdev) > { > /* > @@ -2405,7 +2404,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge); > -#endif > > /* > * Originally in EDAC sources for i82875P: Intel tells BIOS developers to > -- > 2.24.0 >
On Tue, 26 Nov 2019 at 23:05, Bjorn Helgaas <helgaas@kernel.org> wrote: > > On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: > > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > > in. Removing the ifdef will allow us to build the driver as a module. > > > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > > Sorry, I missed this thinking it would be under drivers/pci/controller > and hence handled by Lorenzo. > > So I guess this doesn't fix a build problem, but without this patch, > we just don't run the quirk if the driver is a module, right? > Yes, this is correct. Without this patch, the quirk doesn't get to run if the driver is a module. Wei.
On Wed, 27 Nov 2019 at 10:59, Wei Liu <wei.liu@kernel.org> wrote: > > On Tue, 26 Nov 2019 at 23:05, Bjorn Helgaas <helgaas@kernel.org> wrote: > > > > On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: > > > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > > > in. Removing the ifdef will allow us to build the driver as a module. > > > > > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > > > > Sorry, I missed this thinking it would be under drivers/pci/controller > > and hence handled by Lorenzo. > > > > So I guess this doesn't fix a build problem, but without this patch, > > we just don't run the quirk if the driver is a module, right? > > > > Yes, this is correct. > > Without this patch, the quirk doesn't get to run if the driver is a module. > Hi Bjorn Are you satisfied with the patch? What do I need to do to get it merged? Thanks, Wei.
On Tue, Dec 03, 2019 at 02:06:57PM +0000, Wei Liu wrote: > On Wed, 27 Nov 2019 at 10:59, Wei Liu <wei.liu@kernel.org> wrote: > > > > On Tue, 26 Nov 2019 at 23:05, Bjorn Helgaas <helgaas@kernel.org> wrote: > > > > > > On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: > > > > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > > > > in. Removing the ifdef will allow us to build the driver as a module. > > > > > > > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > > > > > > Sorry, I missed this thinking it would be under drivers/pci/controller > > > and hence handled by Lorenzo. > > > > > > So I guess this doesn't fix a build problem, but without this patch, > > > we just don't run the quirk if the driver is a module, right? > > > > Yes, this is correct. > > > > Without this patch, the quirk doesn't get to run if the driver is a module. > > Are you satisfied with the patch? What do I need to do to get it merged? You needn't do anything. I'll clarify the changelog (the patch doesn't actually *enable* building the driver as a module; it merely ensures that we include the quirk in that case). This is too late for v5.5, so it will get merged for v5.6 unless the modular driver itself was enabled for v5.5. Bjorn
On Tue, 3 Dec 2019 at 14:42, Bjorn Helgaas <helgaas@kernel.org> wrote: > > On Tue, Dec 03, 2019 at 02:06:57PM +0000, Wei Liu wrote: > > On Wed, 27 Nov 2019 at 10:59, Wei Liu <wei.liu@kernel.org> wrote: > > > > > > On Tue, 26 Nov 2019 at 23:05, Bjorn Helgaas <helgaas@kernel.org> wrote: > > > > > > > > On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: > > > > > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > > > > > in. Removing the ifdef will allow us to build the driver as a module. > > > > > > > > > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > > > > > > > > Sorry, I missed this thinking it would be under drivers/pci/controller > > > > and hence handled by Lorenzo. > > > > > > > > So I guess this doesn't fix a build problem, but without this patch, > > > > we just don't run the quirk if the driver is a module, right? > > > > > > Yes, this is correct. > > > > > > Without this patch, the quirk doesn't get to run if the driver is a module. > > > > Are you satisfied with the patch? What do I need to do to get it merged? > > You needn't do anything. I'll clarify the changelog (the patch > doesn't actually *enable* building the driver as a module; it merely > ensures that we include the quirk in that case). > > This is too late for v5.5, so it will get merged for v5.6 unless the > modular driver itself was enabled for v5.5. > OK. Thanks for the clarification. Wei. > Bjorn
On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > in. Removing the ifdef will allow us to build the driver as a module. > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > --- > Alternatively, we can change the condition to: > > #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE > . > > I chose to remove the ifdef because that's what other quirks looked like > in this file. > --- > drivers/pci/quirks.c | 2 -- > 1 file changed, 2 deletions(-) > > diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c > index 320255e5e8f8..cd0e7c18e717 100644 > --- a/drivers/pci/quirks.c > +++ b/drivers/pci/quirks.c > @@ -2381,7 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM, > PCI_DEVICE_ID_TIGON3_5719, > quirk_brcm_5719_limit_mrrs); > > -#ifdef CONFIG_PCIE_IPROC_PLATFORM > static void quirk_paxc_bridge(struct pci_dev *pdev) > { > /* > @@ -2405,7 +2404,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge); > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge); > -#endif Is there a reason this quirk can't be moved to drivers/pci/controller/pcie-iproc-platform.c? That would make it much less subtle because it would be compiled if and only if the driver itself is compiled. If it needs to be here in quirks.c, please include a note about the reason. > /* > * Originally in EDAC sources for i82875P: Intel tells BIOS developers to > -- > 2.24.0 >
On 12/10/19 4:09 PM, Bjorn Helgaas wrote: > On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: >> CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built >> in. Removing the ifdef will allow us to build the driver as a module. >> >> Signed-off-by: Wei Liu <wei.liu@kernel.org> >> --- >> Alternatively, we can change the condition to: >> >> #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE >> . >> >> I chose to remove the ifdef because that's what other quirks looked like >> in this file. >> --- >> drivers/pci/quirks.c | 2 -- >> 1 file changed, 2 deletions(-) >> >> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c >> index 320255e5e8f8..cd0e7c18e717 100644 >> --- a/drivers/pci/quirks.c >> +++ b/drivers/pci/quirks.c >> @@ -2381,7 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM, >> PCI_DEVICE_ID_TIGON3_5719, >> quirk_brcm_5719_limit_mrrs); >> >> -#ifdef CONFIG_PCIE_IPROC_PLATFORM >> static void quirk_paxc_bridge(struct pci_dev *pdev) >> { >> /* >> @@ -2405,7 +2404,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge); >> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge); >> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge); >> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge); >> -#endif > > Is there a reason this quirk can't be moved to > drivers/pci/controller/pcie-iproc-platform.c? That would make it much > less subtle because it would be compiled if and only if the driver > itself is compiled. > > If it needs to be here in quirks.c, please include a note about the > reason. > There's no particular reason and yes it could be moved to pcie-iproc.c. If that's preferred (and it sounds like it is) then we can do that. Thanks, Ray >> /* >> * Originally in EDAC sources for i82875P: Intel tells BIOS developers to >> -- >> 2.24.0 >>
On Tue, Dec 10, 2019 at 04:28:47PM -0800, Ray Jui wrote: > > > On 12/10/19 4:09 PM, Bjorn Helgaas wrote: > > On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: > > > CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built > > > in. Removing the ifdef will allow us to build the driver as a module. > > > > > > Signed-off-by: Wei Liu <wei.liu@kernel.org> > > > --- > > > Alternatively, we can change the condition to: > > > > > > #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE > > > . > > > > > > I chose to remove the ifdef because that's what other quirks looked like > > > in this file. > > > --- > > > drivers/pci/quirks.c | 2 -- > > > 1 file changed, 2 deletions(-) > > > > > > diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c > > > index 320255e5e8f8..cd0e7c18e717 100644 > > > --- a/drivers/pci/quirks.c > > > +++ b/drivers/pci/quirks.c > > > @@ -2381,7 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM, > > > PCI_DEVICE_ID_TIGON3_5719, > > > quirk_brcm_5719_limit_mrrs); > > > -#ifdef CONFIG_PCIE_IPROC_PLATFORM > > > static void quirk_paxc_bridge(struct pci_dev *pdev) > > > { > > > /* > > > @@ -2405,7 +2404,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge); > > > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge); > > > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge); > > > DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge); > > > -#endif > > > > Is there a reason this quirk can't be moved to > > drivers/pci/controller/pcie-iproc-platform.c? That would make it much > > less subtle because it would be compiled if and only if the driver > > itself is compiled. > > > > If it needs to be here in quirks.c, please include a note about the > > reason. > > There's no particular reason and yes it could be moved to pcie-iproc.c. > > If that's preferred (and it sounds like it is) then we can do that. Yes, please, that would be great! No #ifdefs, plus the code won't be compiled into x86 and other arches that never use that driver. Bjorn
On 12/10/19 4:34 PM, Bjorn Helgaas wrote: > On Tue, Dec 10, 2019 at 04:28:47PM -0800, Ray Jui wrote: >> >> >> On 12/10/19 4:09 PM, Bjorn Helgaas wrote: >>> On Fri, Nov 15, 2019 at 01:58:42PM +0000, Wei Liu wrote: >>>> CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built >>>> in. Removing the ifdef will allow us to build the driver as a module. >>>> >>>> Signed-off-by: Wei Liu <wei.liu@kernel.org> >>>> --- >>>> Alternatively, we can change the condition to: >>>> >>>> #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE >>>> . >>>> >>>> I chose to remove the ifdef because that's what other quirks looked like >>>> in this file. >>>> --- >>>> drivers/pci/quirks.c | 2 -- >>>> 1 file changed, 2 deletions(-) >>>> >>>> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c >>>> index 320255e5e8f8..cd0e7c18e717 100644 >>>> --- a/drivers/pci/quirks.c >>>> +++ b/drivers/pci/quirks.c >>>> @@ -2381,7 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM, >>>> PCI_DEVICE_ID_TIGON3_5719, >>>> quirk_brcm_5719_limit_mrrs); >>>> -#ifdef CONFIG_PCIE_IPROC_PLATFORM >>>> static void quirk_paxc_bridge(struct pci_dev *pdev) >>>> { >>>> /* >>>> @@ -2405,7 +2404,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge); >>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge); >>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge); >>>> DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge); >>>> -#endif >>> >>> Is there a reason this quirk can't be moved to >>> drivers/pci/controller/pcie-iproc-platform.c? That would make it much >>> less subtle because it would be compiled if and only if the driver >>> itself is compiled. >>> >>> If it needs to be here in quirks.c, please include a note about the >>> reason. >> >> There's no particular reason and yes it could be moved to pcie-iproc.c. >> >> If that's preferred (and it sounds like it is) then we can do that. > > Yes, please, that would be great! No #ifdefs, plus the code won't be > compiled into x86 and other arches that never use that driver. > > Bjorn > Okay, Wei, let me know if you'd like to help with that? If not, I can make the change. Thanks, Ray
diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index 320255e5e8f8..cd0e7c18e717 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -2381,7 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5719, quirk_brcm_5719_limit_mrrs); -#ifdef CONFIG_PCIE_IPROC_PLATFORM static void quirk_paxc_bridge(struct pci_dev *pdev) { /* @@ -2405,7 +2404,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge); DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge); DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge); DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge); -#endif /* * Originally in EDAC sources for i82875P: Intel tells BIOS developers to
CONFIG_PCIE_IPROC_PLATFORM only gets defined when the driver is built in. Removing the ifdef will allow us to build the driver as a module. Signed-off-by: Wei Liu <wei.liu@kernel.org> --- Alternatively, we can change the condition to: #ifdef CONFIG_PCIE_IPROC_PLATFORM || CONFIG_PCIE_IPROC_PLATFORM_MODULE . I chose to remove the ifdef because that's what other quirks looked like in this file. --- drivers/pci/quirks.c | 2 -- 1 file changed, 2 deletions(-)