mem(4): allow configuring livedump via ioctl
AbandonedPublic
Actions

Authored by mhorne on Feb 8 2022, 8:14 PM.

Details

Reviewers

kib
markj

Summary

This should implement the functionality covered by D34069.

Test Plan

Tested that dumpon -z|Z /dev/mem succeeds and results in a compressed dump.

Need to test the encryption paths.

Diff Detail

Repository

rS FreeBSD src repository - subversion

Lint

Lint Passed

Unit

No Test Coverage

Build Status

Buildable 44343
Build 41231: arc lint + arc unit

Event Timeline

mhorne created this revision.Feb 8 2022, 8:14 PM

Herald added subscribers: Contributor Reviews (src), imp. · View Herald TranscriptFeb 8 2022, 8:14 PM

mhorne requested review of this revision.Feb 8 2022, 8:14 PM

Harbormaster completed remote builds in B44343: Diff 102511.Feb 8 2022, 8:14 PM

mhorne added a parent revision: D33813: Add new vnode dumper to support live minidumps.Feb 8 2022, 8:14 PM

mhorne edited the summary of this revision. (Show Details)Feb 8 2022, 8:19 PM

mhorne edited the test plan for this revision. (Show Details)

mhorne added reviewers: kib, markj.

mhorne mentioned this in D34069: livedump: add an option for dump compression.Feb 8 2022, 8:33 PM

I'm sorry for the delay.

sys/dev/mem/memdev.c
160	Why exactly do we have separate ioctls for configuring and invoking a livedump? Is it only so that we can have dumpon perform configuration and have savecore actually retrieve the dump? If so I'd really prefer to have savecore do everything, but I see it's a bit inconvenient to have to shoehorn everything there... OTOH I still can't see a reason we'd want the kernel to perform encryption and compression rather than doing it in userspace. Did we discuss this somewhere already?

Is this review obsolete now?

In D34209#785515, @markj wrote:

Is this review obsolete now?

Yep, along with D34069.

Revision Contents
Changeset List

Path

Size

sys/

dev/

mem/

memdev.c

33 lines

Diff 102511

View Options

mem(4): allow configuring livedump via ioctlAbandonedPublicActions