[PATCH v2 1/3] KVM: arm64: Disable TRBE Trace Buffer Unit when running in guest context

[Leo Yan]

On Fri, Feb 27, 2026 at 09:21:33PM +0000, Will Deacon wrote:
> The nVHE world-switch code relies on zeroing TRFCR_EL1 to disable trace
> generation in guest context when self-hosted TRBE is in use by the host.
> 
> Per D3.2.1 ("Controls to prohibit trace at Exception levels"), clearing
> TRFCR_EL1 means that trace generation is prohibited at EL1 and EL0 but
> per R_YCHKJ the Trace Buffer Unit will still be enabled if
> TRBLIMITR_EL1.E is set. R_SJFRQ goes on to state that, when enabled, the
> Trace Buffer Unit can perform address translation for the "owning
> exception level" even when it is out of context.
> 
> Consequently, we can end up in a state where TRBE performs speculative
> page-table walks for a host VA/IPA in guest/hypervisor context depending
> on the value of MDCR_EL2.E2TB, which changes over world-switch. The
> potential result appears to be a heady mixture of SErrors, data
> corruption and hardware lockups.
> 
> Extend the TRBE world-switch code to clear TRBLIMITR_EL1.E after
> draining the buffer, restoring the register on return to the host. This
> unfortunately means we need to tackle CPU errata #2064142 and #2038923
> which add additional synchronisation requirements around manipulations
> of the limit register. Hopefully this doesn't need to be fast.
> 
> Cc: Marc Zyngier <maz at kernel.org>
> Cc: Oliver Upton <oupton at kernel.org>
> Cc: James Clark <james.clark at linaro.org>
> Cc: Leo Yan <leo.yan at arm.com>
> Cc: Suzuki K Poulose <suzuki.poulose at arm.com>
> Cc: Fuad Tabba <tabba at google.com>
> Cc: Alexandru Elisei <alexandru.elisei at arm.com>
> Fixes: a1319260bf62 ("arm64: KVM: Enable access to TRBE support for host")
> Signed-off-by: Will Deacon <will at kernel.org>

I tested this on my Orion6 board in nVHE mode (kvm-arm.mode=nvhe).

I launched a VM with several threads running sleep 0.1 in a loop inside
the VM shell.  Then, I collected TRBE trace data on the host side:

  $ perf record -e cs_etm// -a -- sleep 100
  [ perf record: Woken up 74 times to write data ]
  Warning:
  Processed 4798137 events and lost 4 chunks!

  Check IO/CPU overload!

  Warning:
  Processed 9608 samples and lost 100.00%!

  Failed to open /proc/schedstat
  [ perf record: Captured and wrote 42401.333 MB perf.data ]

Tested-by: Leo Yan <leo.yan at arm.com>