Re: [Qemu-devel] [PATCH v2 16/29] Include exec/memory.h slightly less

[Philippe Mathieu-Daudé]

On 8/8/19 2:16 PM, Markus Armbruster wrote:
> Philippe Mathieu-Daudé <address@hidden> writes:
> 
>> On 8/6/19 5:14 PM, Markus Armbruster wrote:
>>> Drop unnecessary inclusions from headers.  Downgrade a few more to
>>> exec/hwaddr.h.
>>>
>>> Signed-off-by: Markus Armbruster <address@hidden>
>>> ---
>>>  hw/audio/lm4549.h                   | 1 +
>>>  hw/display/edid-region.c            | 1 +
>>>  hw/display/tc6393xb.c               | 1 +
>>>  hw/net/can/can_sja1000.h            | 1 +
>>>  hw/net/ne2000.c                     | 1 +
>>>  hw/xtensa/xtensa_memory.h           | 1 -
>>>  include/hw/arm/boot.h               | 1 -
>>>  include/hw/arm/fsl-imx7.h           | 1 -
>>>  include/hw/arm/soc_dma.h            | 2 +-
>>>  include/hw/block/flash.h            | 2 +-
>>>  include/hw/boards.h                 | 1 +
>>>  include/hw/char/parallel.h          | 1 -
>>>  include/hw/display/milkymist_tmu2.h | 1 +
>>>  include/hw/display/tc6393xb.h       | 2 --
>>>  include/hw/display/vga.h            | 2 +-
>>>  include/hw/hw.h                     | 1 -
>>>  include/hw/i2c/pm_smbus.h           | 1 +
>>>  include/hw/i2c/smbus_eeprom.h       | 1 +
>>>  include/hw/misc/auxbus.h            | 1 +
>>>  include/hw/ppc/xics.h               | 1 +
>>>  include/hw/usb.h                    | 1 +
>>>  include/hw/virtio/virtio.h          | 1 +
>>>  migration/colo.c                    | 1 +
>>>  migration/migration.h               | 1 +
>>>  migration/postcopy-ram.c            | 1 +
>>>  migration/rdma.c                    | 1 +
>>>  26 files changed, 20 insertions(+), 10 deletions(-)
>>>
>> [...]> diff --git a/migration/colo.c b/migration/colo.c
>>> index 9f84b1fa3c..0b5b620391 100644
>>> --- a/migration/colo.c
>>> +++ b/migration/colo.c
>>> @@ -23,6 +23,7 @@
>>>  #include "io/channel-buffer.h"
>>>  #include "trace.h"
>>>  #include "qemu/error-report.h"
>>> +#include "qemu/rcu.h"
>>>  #include "migration/failover.h"
>>>  #ifdef CONFIG_REPLICATION
>>>  #include "replication.h"
>>> diff --git a/migration/migration.h b/migration/migration.h
>>> index 26f01d00f6..a7882af67a 100644
>>> --- a/migration/migration.h
>>> +++ b/migration/migration.h
>>> @@ -14,6 +14,7 @@
>>>  #ifndef QEMU_MIGRATION_H
>>>  #define QEMU_MIGRATION_H
>>>  
>>> +#include "exec/cpu-common.h"
>>>  #include "qapi/qapi-types-migration.h"
>>>  #include "qemu/thread.h"
>>>  #include "qemu/coroutine_int.h"
>>> diff --git a/migration/postcopy-ram.c b/migration/postcopy-ram.c
>>> index 9faacacc9e..56054d0a73 100644
>>> --- a/migration/postcopy-ram.c
>>> +++ b/migration/postcopy-ram.c
>>> @@ -25,6 +25,7 @@
>>>  #include "ram.h"
>>>  #include "qapi/error.h"
>>>  #include "qemu/notify.h"
>>> +#include "qemu/rcu.h"
>>>  #include "sysemu/sysemu.h"
>>>  #include "sysemu/balloon.h"
>>>  #include "qemu/error-report.h"
>>> diff --git a/migration/rdma.c b/migration/rdma.c
>>> index 3036221ee8..b0e27b6174 100644
>>> --- a/migration/rdma.c
>>> +++ b/migration/rdma.c
>>> @@ -25,6 +25,7 @@
>>>  #include "qemu/error-report.h"
>>>  #include "qemu/main-loop.h"
>>>  #include "qemu/module.h"
>>> +#include "qemu/rcu.h"
>>>  #include "qemu/sockets.h"
>>>  #include "qemu/bitmap.h"
>>>  #include "qemu/coroutine.h"
>>>
>>
>> To keep this patch cleaner, can you use a preparatory patch adding the
>> various "qemu/rcu.h"?
> 
> How would I explain that new patch?
> 
> Other patches that drop includes from headers also fix up their users by
> adding the includes that are actually needed there, e.g.:
> 
> * PATCH 03: hw/qdev-core.h dropped, fixups add only qemu/bitops.h or
>   hw/irq.h.
> 
> * PATCH 10: hw/ide/internal.h dropped, some fixups add only
>   exec/address-spaces.h or sysemu/sysemu.h.
> 
> * PATCH 17: qom/object.h dropped, fixup adds only hw/hw.h.
> 
> * PATCH 20: qemu/main-loop.h dropped, some fixups add only block/aio.h
>   or qemu/thread.h.  Touches five times as many files as this one...
> 
> * More.

I See.

> Does each of them need preparatory patches?

If those are only 4, why not... However this series is already big
enough, and what matters more: bisectable. Preparatory 3-lines patch
makes no difference, I was just aiming at perfection ;)

I am happy with the current organization.

> Would it help if their commit messages explained the fixups more
> clearly?

No, not that much, each commit is already well explained.

Sorry for the noise!

Phil,