> On 21 May 2025, at 11:20, Michael Paquier <mich...@paquier.xyz> wrote:
>
> Rebased, to fix a conflict I've introduced with a different commit.
I've looked into the patch set and have some more questions:
1. What if we flush many times? Is it supposed to overwrite
injection_points.data?
2. What if we restart many times? first startup will remove the file... maybe
remove it explicitly?
3. InjectionPoint private data is not handled, is it OK?
4. How session-local points are expected to be flushed? into which session they
will be loaded? Maybe let's document that they are not saved?
Besides these, cool new abilities and a test for a bug, looks good to me.
Best regards, Andrey Borodin.
