Re: [Qemu-devel] [RFC PATCH v6 00/20] replay additions

[Pavel Dovgalyuk]

> From: Ciro Santilli [mailto:address@hidden 
> Awesome! This worked with my test procedure described at: 
> https://stackoverflow.com/questions/46970215/how-to-use-qemus-deterministic-record-and-replay-feature-for-a-linux-kernel-boo
> While we are a it, can we clarify the following points the docs?
> 1) Usage of the record/replay: mentions that `-net none` is required, unless 
> a network patch is applied.
> But the network section that follows explains how to record network requests, 
> and I've tested it, so I think the `-net none` / patch requirement is just 
> outdated and should be removed.
> 2) There are mentions of "reverse execution" and "reverse debugging" on the 
> "Record/replay" section.
> Are those supported? If yes, explain there how to use them. Otherwise, remove 
> the mentions.
> I would love to see that feature!

Not yet, but they are the next to apply after this series.
I'll update the docs.

> 3) Is the snapshot feature mentioned in the "Snapshotting" section something 
> specific to record and replay?
> If yes, can we detail further how to use it besides passing 
> `rrsnapshot=snapshot_name`? How to create and restore the snapshots?
> If not, why does the `rrsnapshot=snapshot_name` option exist?
> 4) Maybe give the required disk and network commands under "Usage of the 
> record/replay:"
> -drive file=disk.qcow,if=none,id=img-direct
> -drive driver=blkreplay,if=none,image=img-direct,id=img-blkreplay
> -device ide-hd,drive=img-blkreplay
> -netdev user,id=net1 -device rtl8139,netdev=net1
> -object filter-replay,id=replay,netdev=net1
> People likely always want those, but it currently requires reading through 
> the whole doc file to put everything together.

Thanks for reviewing, I'll add the documentation update to the series.


Pavel Dovgalyuk