Blue Swirl wrote: > On 5/13/10, Jan Kiszka <jan.kis...@web.de> wrote: >> From: Jan Kiszka <jan.kis...@siemens.com> >> >> Some legacy users (mostly PC devices) of vmstate_register manage >> instance IDs on their own, and that unfortunately in a way that is >> incompatible with automatically generated ones. This so far prevents >> switching those users to vmstates that are registered by qdev. >> >> To establish a migration path, this patch introduces the concept of >> alias IDs. They can be passed to an extended vmstate registration >> service, and qdev provides a set service to be used during device init. >> find_se will consider the alias in addition to the default ID. We can >> then start generating the default ID automatically and writing it on >> vmsave, thus converting that format without breaking support for upward >> migration. > > If this is only for compatibility, I think the name should show it, > like vmstate_set_compat_instance_id(), or > vmstate_set_legacy_instance_id(). That way, if there happens to be an > incompatible version bump, the function name suggests that it can be > removed.
Hmm, makes some sense, not for the vmstate interface (no new code outside the core should touch it anymore), but for clarifying the qdev part. > > The function should also take a last_legacy_version_id parameter. > Consider for example that a vmstate format with the legacy ID is > currently used with version_id of 2. We also start using this > compatibility system. A new, compatible version 3 arrives but we only > want to support legacy ID for version 2, as indicated by > last_legacy_version_id=2. Then with a new version, let's say 5, which > is no longer compatible with 2 or 3, the legacy ID stuff can finally > be thrown away. qdev.c should check if last_legacy_id >= > minimum_version_id and complain otherwise. OK, will look into this. Jan
signature.asc
Description: OpenPGP digital signature
