Eric Blake <[email protected]> wrote: > On 3/7/19 6:28 AM, Juan Quintela wrote: >> "Dr. David Alan Gilbert" <[email protected]> wrote: >>> * Juan Quintela ([email protected]) wrote: >>>> We need to send this field when we add compression support. As we are >>>> still on x- stage, we can do this kind of changes. >>>> >>>> Signed-off-by: Juan Quintela <[email protected]> >>> >>> OK, so that kind of makes sense - in that what you want to do is read >>> the whole compressed chunk in one go. >>> >>> Consider it might be a good idea to add some spare entries here; >>> say a uint64_t dummy[4] so that if you need any more fields in the >>> future it'll be easy to add them without breaking compatibility. >> >> We have a version field, so ... we can handle it relatively gratiously, >> but I see your point. >> >> How many? > > What is the current struct size? Is there a nice power-of-2 or > cache-line-size multiple to round it out to?
MultiFDInit_t padded to 64 bytes. (7 uint8_t + 4 uint64_t padding) Relatively easy to get it to a multiple of 2 and cacheline. MultiFDPacket_t padded to 32 bytes (5 * 64). (4 uint64_t padding) This one is too far away from a multiple of two (512bytes). Would have required 28 uint64_t of padding. I am using this ones, ok? Later, Juan.
