Page MenuHomeFreeBSD
Differential D49560

pci: Only re-route IRQs based on firmware on x86
ClosedPublic
Actions

Authored by cperciva on Mar 28 2025, 6:12 PM.
Tags
None
Referenced Files
Unknown Object (File)
Sun, Feb 8, 2:27 PM
Unknown Object (File)
Sat, Feb 7, 7:47 PM
Unknown Object (File)
Sat, Feb 7, 3:25 AM
Unknown Object (File)
Tue, Feb 3, 9:34 PM
Unknown Object (File)
Tue, Feb 3, 8:39 AM
Unknown Object (File)
Sat, Jan 31, 2:45 AM
Unknown Object (File)
Sat, Jan 31, 12:08 AM
Unknown Object (File)
Tue, Jan 27, 8:31 PM
View All 81 Files
Subscribers
zarychtam_plan-b.pwste.edu.pl

Details

Reviewers
jhb
imp
Commits
rG0e33c2e6df7a: pci: Only re-route IRQs based on firmware on x86
Summary

There is a (very historical) call to pci_assign_interrupt for the
purpose of routing IRQs which may have been set up wrong by x86 BIOS
or firmware. On non-x86 systems, this is unnecessary; and on INTRNG
systems it results in a (synthetic) IRQ leak and ultimately a kernel
panic after many hotplug/unplug cycles.

Suggested by: jhb
MFC after: 2 weeks
Sponsored by: Amazon

Diff Detail

Repository
rG FreeBSD src repository
Lint
Lint Skipped
Unit
Tests Skipped
Build Status
Buildable 63186
Build 60070: arc lint + arc unit

Event Timeline

cperciva requested review of this revision.Mar 28 2025, 6:12 PM
cperciva created this revision.
Harbormaster completed remote builds in B63186: Diff 152809.Mar 28 2025, 6:13 PM
cperciva mentioned this in D49543: Correctly track index of first free irq map entry.Mar 28 2025, 6:32 PM
jhb accepted this revision.Mar 28 2025, 6:43 PM
This revision is now accepted and ready to land.Mar 28 2025, 6:43 PM
Closed by commit rG0e33c2e6df7a: pci: Only re-route IRQs based on firmware on x86 (authored by cperciva). · Explain WhyMar 29 2025, 8:18 PM
This revision was automatically updated to reflect the committed changes.
cperciva added a commit: rG0e33c2e6df7a: pci: Only re-route IRQs based on firmware on x86.
zarychtam_plan-b.pwste.edu.pl added a subscriber: zarychtam_plan-b.pwste.edu.pl.EditedMar 30 2025, 7:27 PM

Could it bring a new flow of messages into the kernel message buffer? After today's upgrade of my desktop PC, I see plenty of "Can't enable IRQ/MSI because no handler is installed," which didn't appear in the past.

It was triggered by D49543 (likely) and seems to be harmless so far.

cperciva added a reverting change: rG680d34896c36: Revert "pci: Only re-route IRQs based on firmware on x86".Mar 31 2025, 9:01 PM

Revision Contents
Changeset List

PathSize
sys/
dev/
pci/
2 lines

Diff 152809

View Options

sys/dev/pci/pci.c

Loading...