Skip to content

aarch64: fix MPIDR encoding above 16 VPs #4373

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Henry Li (henryli001) wants to merge 2 commits into from
Closed

aarch64: fix MPIDR encoding above 16 VPs #4373

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6f94e0a
vm_topology: fix ARM64 MPIDR affinity packing
henryli001 Sep 3, 2026
985332a
ci: retry unrelated x64 VMM test failures
henryli001 Sep 3, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
40 changes: 30 additions & 10 deletions vm/vmcore/vm_topology/src/processor/aarch64.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,18 @@ use super::VpInfo;
use super::VpTopologyInfo;
use aarch64defs::MpidrEl1;

fn mpidr_for_vp_index(vp_index: u32, uni_proc: bool) -> MpidrEl1 {
// ICC_SGIxR directly addresses 16 CPUs at affinity level 0. Match KVM's
// vCPU ID mapping by carrying higher bits into affinity levels 1 and 2.
MpidrEl1::new()
.with_res1_31(true)
.with_u(uni_proc)
.with_aff0((vp_index & 0xf) as u8)
.with_aff1(((vp_index >> 4) & 0xff) as u8)
.with_aff2(((vp_index >> 12) & 0xff) as u8)
.with_aff3(0)
}
Comment on lines +21 to +25

/// ARM64-specific topology information.
#[cfg_attr(feature = "inspect", derive(inspect::Inspect))]
#[derive(Debug, Copy, Clone)]
Expand Down Expand Up @@ -190,17 +202,8 @@ impl TopologyBuilder<Aarch64Topology> {
return Err(InvalidTopology::InvalidGicNrIrqs(nr));
}
let mpidrs = (0..proc_count).map(|vp_index| {
// TODO: construct mpidr appropriately for the specified
// topology.
let uni_proc = proc_count == 1;
let mut aff = (0..4).map(|i| (vp_index >> (8 * i)) as u8);
MpidrEl1::new()
.with_res1_31(true)
.with_u(uni_proc)
.with_aff0(aff.next().unwrap())
.with_aff1(aff.next().unwrap())
.with_aff2(aff.next().unwrap())
.with_aff3(aff.next().unwrap())
mpidr_for_vp_index(vp_index, uni_proc)
});
let gic_version = self.arch.platform.gic_version;
self.build_with_vp_info(mpidrs.enumerate().map(move |(id, mpidr)| {
Expand Down Expand Up @@ -283,3 +286,20 @@ impl ProcessorTopology<Aarch64Topology> {
self.arch.platform.gic_nr_irqs
}
}

#[cfg(test)]
mod tests {
use super::*;

#[test]
fn mpidr_rolls_over_aff0_after_sixteen_vps() {
let vp15 = mpidr_for_vp_index(15, false);
assert_eq!((vp15.aff2(), vp15.aff1(), vp15.aff0()), (0, 0, 15));

let vp16 = mpidr_for_vp_index(16, false);
assert_eq!((vp16.aff2(), vp16.aff1(), vp16.aff0()), (0, 1, 0));

let vp111 = mpidr_for_vp_index(111, false);
assert_eq!((vp111.aff2(), vp111.aff1(), vp111.aff0()), (0, 6, 15));
}
}
Loading