Re: [Libvirt-cim] [PATCH 1 of 6] Turns out that the sdc_rasd_prop functions might need to report status or see the passed in reference. This updates the typedef and the memory functions, since those are already in the tree
Dan Smith wrote:
JG> # HG changeset patch
JG> # User Jay Gagnon <grendel(a)linux.vnet.ibm.com>
JG> # Date 1194557266 18000
JG> # Node ID fd6deb234772b44fe549a3513a47115a01e20f7d
JG> # Parent bd1b1067d106ecc1546c3b2436a43f59f8eaba15
JG> Turns out that the sdc_rasd_prop functions might need to report status or see the
passed in reference. This updates the typedef and the memory functions, since those are
already in the tree.
This one looks pretty straightforward. It looks like we could apply
this one independently if we wanted to, right? A rather
self-contained change, considering the size of the rest of the set.
This sets up the new function header for all the callbacks. I think it
can be applied separately in that it can be applied before the other
ones, but if we are going to put those in then this one will have to be
there. Did that answer the question?
JG> + if (!ret) {
JG> + cu_statusf(_BROKER, s, CMPI_RC_ERR_FAILED,
JG> + "Could not copy RASD.");
JG> + }
This is a total nit, so forgive my OCD, but I think that everywhere
else, the style of a cu_statusf() call is:
cu_statusf(broker, status, <newline>
CMPI_RC_ERR_FOO, <newline>
"Error message", [opt if short, else newline],
[opt]);
Could I convince you to change these? :)
Yea, no prob.
--
-Jay