Avi Kivity wrote:
On 12/07/2009 05:50 PM, Anthony Liguori wrote:
While it's always fun to argue about standards interpretation, I
wanted to capture some action items from the discussion that I think
there is agreement about. Since I want to make changes for 0.12, I
think it would be best to try and settle these now so we can do this
before -rc2.
For 0.12.0-rc2:
I will send out a patch tonight or tomorrow changing qemu_malloc() to
return malloc(1) when size=0 only for production builds (via
--enable-zero-mallocs). Development trees will maintain their
current behavior.
Since active development is ceasing on 0.12, I'd suggest not having
separate behaviour for devel and production. Do we want patches for
n==0 array allocations at this time?
Covering every qemu_malloc instance this close to the GA is too risky.
I agree that having separate behavior is less than ideal but I think
it's the only sane way forward.
I'd really like to see Markus' patch applied.
For 0.12, that doesn't seem like a possibility.
For 0.13:
Someone (Marcus?) will introduce four new allocation functions.
type *qemu_new(type, n_types);
type *qemu_new0(type, n_types);
type *qemu_renew(type, mem, n_types);
type *qemu_renew0(type, mem, n_types);
I'd like to see separate functions for arrays and single objects, to
avoid ", 1)" everywhere.
qemu_new()
qemu_new0()
qemu_new_array()
qemu_new_array0()
qemu_renew_array()
qemu_renew_array0()
Like I said, I'm not tied to naming. I'll defer this to whoever
contributes the patch and signs up for the conversion work.
In addition, Markus' patch should be applied to master to avoid
regressions while the code is converted.
Let's separate that discussion as it's an independent consideration.
Regards,
Anthony Liguori
