Re: [Qemu-devel] [PATCH 1/4] savevm: introduce little endian variants of savevm routines

[Anthony Liguori <aliguori@us.ibm.com>]

Peter Maydell <peter.maydell@linaro.org> writes:

> On 10 December 2012 14:29, Anthony Liguori <aliguori@us.ibm.com> wrote:
>> Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
>> ---
>>  qemu-file.h |  7 +++++++
>>  savevm.c    | 45 +++++++++++++++++++++++++++++++++++++++++++++
>>  2 files changed, 52 insertions(+)
>>
>> diff --git a/qemu-file.h b/qemu-file.h
>> index d64bdbb..ac5286c 100644
>> --- a/qemu-file.h
>> +++ b/qemu-file.h
>> @@ -94,6 +94,9 @@ static inline void qemu_put_ubyte(QEMUFile *f, unsigned int v)
>>  void qemu_put_be16(QEMUFile *f, unsigned int v);
>>  void qemu_put_be32(QEMUFile *f, unsigned int v);
>>  void qemu_put_be64(QEMUFile *f, uint64_t v);
>> +void qemu_put_le16(QEMUFile *f, unsigned int v);
>> +void qemu_put_le32(QEMUFile *f, unsigned int v);
>> +void qemu_put_le64(QEMUFile *f, uint64_t v);
>
> Do we want to add a comment here somewhere that says the le versions
> are for backcompat with a specific thing and the be ones are the
> ones to use in new code (well, new code not using vmstate)?

Yeah, that's a good idea.  I've unfortunately found a couple more cases
of this (writing native endian to the wire).

Regards,

Anthony Liguori

>
> -- PMM