Re: [libvirt] [PATCH 06/11] conf: Get rid of virDomainDeviceDefPostParseOne

On Mon, Oct 01, 2018 at 12:41 PM +0200, Peter Krempa <pkrempa(a)redhat.com&gt; wrote: > On Mon, Oct 01, 2018 at 12:27:41 +0200, Marc Hartmayer wrote: >> On Sat, Sep 29, 2018 at 04:09 AM +0200, John Ferlan <jferlan(a)redhat.com&gt; wrote: >> > On 9/20/18 1:44 PM, Marc Hartmayer wrote: >> >> Move the domainPostParseDataAlloc/Free calls to >> >> virDomainDeviceDefParse. As an consequence >> >> virDomainDeviceDefPostParseOne is superfluous and can therefore be >> >> removed. >> >> >> >> Signed-off-by: Marc Hartmayer <mhartmay(a)linux.ibm.com&gt; >> >> Reviewed-by: Boris Fiuczynski <fiuczy(a)linux.ibm.com&gt; >> >> --- >> >> src/conf/domain_conf.c | 37 +++++++++++-------------------------- >> >> 1 file changed, 11 insertions(+), 26 deletions(-) >> >> >> > >> > I'm not against this per se; however, I should not that the code was >> > specifically extracted in commit e168bc8a. >> >> There are the following three functions: >> >> virDomainDeviceDefParse >> virDomainDeviceDefPostParse >> virDomainDeviceDefPostParseOne >> >> Peter introduced the function “virDomainDeviceDefPostParseOne” to avoid >> the allocation of private data across the callbacks. This is absolutely >> fine. >> >> What I’ve done is, I moved the domainPostParseDataAlloc/Free calls to >> virDomainDeviceDefParse instead of having an extra wrapper function >> (virDomainDeviceDefPostParse/One) for that. With this change I can reuse >> the QEMU caps for the virDomainDeviceDefValidate call in >> virDomainDeviceDefParse as well. > > For the above it's not necessary to open-code what virDomainDeviceDefPostParseOne > in a rather massive function. You can pass the opaque in if you want. I don’t get it. Where can I pass the opaque?