On 25/07/16 10:45, Wei Liu wrote: > On Thu, Jul 21, 2016 at 06:17:36PM +0100, Andrew Cooper wrote: >> It was never intended for records such as these to be sent with zero content. >> > Wouldn't it be better to modify write_split_record to ignore zero > content instead of patching up different places?
That would catch the HVM_PARAMS, but not the other 3, which have an extra 8 byte header. However, having write_split_record() conditionally drop records is a violation of the principle of least surprise. Choosing to skip a record is an important decision and needs to be easy to spot in the code. ~Andrew _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
