linux-next: libata/avr32 build failure

linux-next: libata/avr32 build failure

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 555 bytes --]

Hi Jeff,

For a while the linux-next build for avr32 defconfig (and others) has
been failing like this:

drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' undeclared here (not in a function)

It seems to be because the declaration of ata_pci_default_filter is
protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
only.

Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
March 29).

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: linux-next: libata/avr32 build failure

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 802 bytes --]

Hi Jeff,

On Fri, 4 Apr 2008 17:10:40 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> For a while the linux-next build for avr32 defconfig (and others) has
> been failing like this:
> 
> drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' undeclared here (not in a function)
> 
> It seems to be because the declaration of ata_pci_default_filter is
> protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
> only.
> 
> Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
> March 29).

The bug looks like it was introduced by commit
dc14c0c5012855bfc9a5f76056b92f198d52834c ("libata: implement and use ops
inheritance").

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: linux-next: libata/avr32 build failure

[Jeff Garzik]

Stephen Rothwell wrote:
> Hi Jeff,
> 
> For a while the linux-next build for avr32 defconfig (and others) has
> been failing like this:
> 
> drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' undeclared here (not in a function)
> 
> It seems to be because the declaration of ata_pci_default_filter is
> protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
> only.
> 
> Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
> March 29).

fixed being pushed right now (patch from Tejun fixed it)

Re: linux-next: libata/avr32 build failure

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 771 bytes --]

On Fri, 04 Apr 2008 04:06:56 -0400 Jeff Garzik <jeff@garzik.org> wrote:
>
> Stephen Rothwell wrote:
> > 
> > For a while the linux-next build for avr32 defconfig (and others) has
> > been failing like this:
> > 
> > drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' undeclared here (not in a function)
> > 
> > It seems to be because the declaration of ata_pci_default_filter is
> > protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
> > only.
> > 
> > Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
> > March 29).
> 
> fixed being pushed right now (patch from Tejun fixed it)

Thanks.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: linux-next: libata/avr32 build failure

[Randy Dunlap]

On Fri, 04 Apr 2008 04:06:56 -0400 Jeff Garzik wrote:

> Stephen Rothwell wrote:
> > Hi Jeff,
> > 
> > For a while the linux-next build for avr32 defconfig (and others) has
> > been failing like this:
> > 
> > drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' undeclared here (not in a function)
> > 
> > It seems to be because the declaration of ata_pci_default_filter is
> > protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
> > only.
> > 
> > Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
> > March 29).
> 
> fixed being pushed right now (patch from Tejun fixed it)

Some fixes other than what was in today's (April 04)
[git patches] libata fixes?

I hope there is more.  I don't see how that fixes this error.
Nor does gcc see how, after I apply that patch.

---
~Randy

Re: linux-next: libata/avr32 build failure

[Jeff Garzik]

Randy Dunlap wrote:
> On Fri, 04 Apr 2008 04:06:56 -0400 Jeff Garzik wrote:
> 
>> Stephen Rothwell wrote:
>>> Hi Jeff,
>>>
>>> For a while the linux-next build for avr32 defconfig (and others) has
>>> been failing like this:
>>>
>>> drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' undeclared here (not in a function)
>>>
>>> It seems to be because the declaration of ata_pci_default_filter is
>>> protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
>>> only.
>>>
>>> Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
>>> March 29).
>> fixed being pushed right now (patch from Tejun fixed it)
> 
> Some fixes other than what was in today's (April 04)
> [git patches] libata fixes?
> 
> I hope there is more.  I don't see how that fixes this error.
> Nor does gcc see how, after I apply that patch.

It wasn't sent in #upstream-fixes, it was applied to #upstream...

Is that where you are looking?

	Jeff

Re: linux-next: libata/avr32 build failure

[Randy Dunlap]

Jeff Garzik wrote:
> Randy Dunlap wrote:
>> On Fri, 04 Apr 2008 04:06:56 -0400 Jeff Garzik wrote:
>>
>>> Stephen Rothwell wrote:
>>>> Hi Jeff,
>>>>
>>>> For a while the linux-next build for avr32 defconfig (and others) has
>>>> been failing like this:
>>>>
>>>> drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' 
>>>> undeclared here (not in a function)
>>>>
>>>> It seems to be because the declaration of ata_pci_default_filter is
>>>> protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
>>>> only.
>>>>
>>>> Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
>>>> March 29).
>>> fixed being pushed right now (patch from Tejun fixed it)
>>
>> Some fixes other than what was in today's (April 04)
>> [git patches] libata fixes?
>>
>> I hope there is more.  I don't see how that fixes this error.
>> Nor does gcc see how, after I apply that patch.
> 
> It wasn't sent in #upstream-fixes, it was applied to #upstream...
> 
> Is that where you are looking?

I'm just trying to build linux-next with or without today's "libata fixes".
ATA with CONFIG_PCI=n still has build errors, as reported above.

-- 
~Randy

Re: linux-next: libata/avr32 build failure

[Jeff Garzik]

Randy Dunlap wrote:
> Jeff Garzik wrote:
>> Randy Dunlap wrote:
>>> On Fri, 04 Apr 2008 04:06:56 -0400 Jeff Garzik wrote:
>>>
>>>> Stephen Rothwell wrote:
>>>>> Hi Jeff,
>>>>>
>>>>> For a while the linux-next build for avr32 defconfig (and others) has
>>>>> been failing like this:
>>>>>
>>>>> drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' 
>>>>> undeclared here (not in a function)
>>>>>
>>>>> It seems to be because the declaration of ata_pci_default_filter is
>>>>> protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
>>>>> only.
>>>>>
>>>>> Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
>>>>> March 29).
>>>> fixed being pushed right now (patch from Tejun fixed it)
>>>
>>> Some fixes other than what was in today's (April 04)
>>> [git patches] libata fixes?
>>>
>>> I hope there is more.  I don't see how that fixes this error.
>>> Nor does gcc see how, after I apply that patch.
>>
>> It wasn't sent in #upstream-fixes, it was applied to #upstream...
>>
>> Is that where you are looking?
> 
> I'm just trying to build linux-next with or without today's "libata fixes".
> ATA with CONFIG_PCI=n still has build errors, as reported above.

OK, that's expected.

	Jeff

Re: linux-next: libata/avr32 build failure

[Randy Dunlap]

On Fri, 04 Apr 2008 16:23:53 -0400 Jeff Garzik wrote:

> Randy Dunlap wrote:
> > Jeff Garzik wrote:
> >> Randy Dunlap wrote:
> >>> On Fri, 04 Apr 2008 04:06:56 -0400 Jeff Garzik wrote:
> >>>
> >>>> Stephen Rothwell wrote:
> >>>>> Hi Jeff,
> >>>>>
> >>>>> For a while the linux-next build for avr32 defconfig (and others) has
> >>>>> been failing like this:
> >>>>>
> >>>>> drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' 
> >>>>> undeclared here (not in a function)
> >>>>>
> >>>>> It seems to be because the declaration of ata_pci_default_filter is
> >>>>> protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
> >>>>> only.
> >>>>>
> >>>>> Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
> >>>>> March 29).
> >>>> fixed being pushed right now (patch from Tejun fixed it)
> >>>
> >>> Some fixes other than what was in today's (April 04)
> >>> [git patches] libata fixes?
> >>>
> >>> I hope there is more.  I don't see how that fixes this error.
> >>> Nor does gcc see how, after I apply that patch.
> >>
> >> It wasn't sent in #upstream-fixes, it was applied to #upstream...
> >>
> >> Is that where you are looking?
> > 
> > I'm just trying to build linux-next with or without today's "libata fixes".
> > ATA with CONFIG_PCI=n still has build errors, as reported above.
> 
> OK, that's expected.

Uh, that's not good.  What does it mean?

(a) linux-next is not pulling the correct branch of your git tree
or
(b) linux-next just won't work with libata
or
(c) we have too many trees in the forest
or
something else?  (if so, what?)


---
~Randy

Re: linux-next: libata/avr32 build failure

[Jeff Garzik]

Randy Dunlap wrote:
> On Fri, 04 Apr 2008 16:23:53 -0400 Jeff Garzik wrote:
> 
>> Randy Dunlap wrote:
>>> Jeff Garzik wrote:
>>>> Randy Dunlap wrote:
>>>>> On Fri, 04 Apr 2008 04:06:56 -0400 Jeff Garzik wrote:
>>>>>
>>>>>> Stephen Rothwell wrote:
>>>>>>> Hi Jeff,
>>>>>>>
>>>>>>> For a while the linux-next build for avr32 defconfig (and others) has
>>>>>>> been failing like this:
>>>>>>>
>>>>>>> drivers/ata/libata-core.c:127: error: 'ata_pci_default_filter' 
>>>>>>> undeclared here (not in a function)
>>>>>>>
>>>>>>> It seems to be because the declaration of ata_pci_default_filter is
>>>>>>> protected by CONFIG_PCI but the usage above is protected by CONFIG_ATA
>>>>>>> only.
>>>>>>>
>>>>>>> Also reported by Randy Dunlap (on March 26) and Haavard Skinnemoen (on
>>>>>>> March 29).
>>>>>> fixed being pushed right now (patch from Tejun fixed it)
>>>>> Some fixes other than what was in today's (April 04)
>>>>> [git patches] libata fixes?
>>>>>
>>>>> I hope there is more.  I don't see how that fixes this error.
>>>>> Nor does gcc see how, after I apply that patch.
>>>> It wasn't sent in #upstream-fixes, it was applied to #upstream...
>>>>
>>>> Is that where you are looking?
>>> I'm just trying to build linux-next with or without today's "libata fixes".
>>> ATA with CONFIG_PCI=n still has build errors, as reported above.
>> OK, that's expected.
> 
> Uh, that's not good.  What does it mean?
> 
> (a) linux-next is not pulling the correct branch of your git tree
> or
> (b) linux-next just won't work with libata
> or
> (c) we have too many trees in the forest
> or
> something else?  (if so, what?)

It means you are inside the window between unfixed linux-next and 
tomorrow's fixed linux-next, and that you attempted to seek the proper 
fix in the wrong location.

So the next time linux-next pulls from libata, it should be fixed.

Or, if you don't want to wait, you should get the fix from the location 
noted (#upstream branch), because the fix was not among those sent to 
Linus yesterday.

	Jeff