Re: [Qemu-devel] [PULL 00/45] Include reorganization + PCI patch queue

[Alexander Graf]

On 19.12.2012, at 03:00, Anthony Liguori wrote:

> Anthony Liguori <address@hidden> writes:
> 
>> Paolo Bonzini <address@hidden> writes:
>> 
>>> And how does it conflicts with Alex's ppc pull request? It is a fast
>>> forward from origin/master as of now (commit a8a826a, exec: refactor
>>> cpu_restore_state, 2012-12-04), and that includes Alex's commits...
>> 
>> It's not a conflict.  You both have problems with your pull requests
>> because I assume neither of you have libfdt installed which means a
>> bunch of ppc is not being built.
>> 
>> We really ought to pull in fdt as a submodule to prevent this type of
>> thing...

Does making libfdt a submodule really change things? Developers would still 
have to init the submodule. They could just as well install the libfdt-devel 
packet from their distro of choice instead :).


Alex

> 
> We the patches below, the changes pass all my tests so as soon as you
> make these changes, it'll go in.
> 
> Regards,
> 
> Anthony Liguori
> 
>> 
>> Paolo, you need to incorporate:
>> 
>> diff --git a/hw/ppc/e500.c b/hw/ppc/e500.c
>> index 5b16096..aa54fd8 100644
>> --- a/hw/ppc/e500.c
>> +++ b/hw/ppc/e500.c
>> @@ -18,7 +18,8 @@
>> #include "qemu-common.h"
>> #include "e500.h"
>> #include "e500-ccsr.h"
>> -#include "net.h"
>> +#include "net/net.h"
>> +#include "qemu/config-file.h"
>> #include "hw/hw.h"
>> #include "hw/serial.h"
>> #include "hw/pci/pci.h"
>> 
>> diff --git a/hw/spapr.c b/hw/spapr.c
>> index 071cf41..fdd1eb6 100644
>> --- a/hw/spapr.c
>> +++ b/hw/spapr.c
>> @@ -49,6 +49,7 @@
>> 
>> #include "exec/address-spaces.h"
>> #include "hw/usb.h"
>> +#include "qemu/config-file.h"
>> 
>> #include <libfdt.h>
>> 
>> Michael, you need to incorporate:
>> 
>> diff --git a/hw/ppc/e500plat.c b/hw/ppc/e500plat.c
>> index ea4134c..4deb02a 100644
>> --- a/hw/ppc/e500plat.c
>> +++ b/hw/ppc/e500plat.c
>> @@ -14,7 +14,7 @@
>> #include "e500.h"
>> #include "../boards.h"
>> #include "sysemu/device_tree.h"
>> -#include "hw/pci.h"
>> +#include "hw/pci/pci.h"
>> 
>> static void e500plat_fixup_devtree(PPCE500Params *params, void *fdt)
>> {
>> 
>> I can't do this through a merge commit because the branches being merged
>> aren't bisectable so please update and resend.  It's probably easiest to
>> do it all through Paolo's branch to make sure there are no conflicts
>> between the two branches.
>> 
>> Regards,
>> 
>> Anthony Liguori
>> 
>>> 
>>> Paolo
>>> 
>>> 
>>> 
>>> On Tue, Dec 18, 2012 at 10:11 PM, Paolo Bonzini <address@hidden> wrote:
>>> 
>>>> Il 18/12/2012 21:59, Anthony Liguori ha scritto:
>>>>> But this series breaks the build of make check.  It's not just one test
>>>>> case but almost every single one.  I think you must have eliminated an
>>>>> implicit include of qemu-common.h which makes PRId64 et al all
>>>>> undeclared.
>>>>> 
>>>>> I started fixing this too but it became too much for a merge commit
>>>>> since it affects almost all tests.
>>>>> 
>>>>> Can you fixup make check and send this series out?  I'll make sure to
>>>>> check tomorrow and merge your pull request if I see it tomorrow.
>>>>> Otherwise, let me know when you plan on sending it and I'll make sure to
>>>>> be available to merge it.
>>>> 
>>>> Hmm, I must have pushed the wrong branch because I remember this failure.
>>>> 
>>>> Paolo
>>>> 
>