Errata
Specification Update
33
AZ29.
Split Locked Stores May Not Trigger the Monitoring Hardware
Problem:
Logical processors normally resume program execution following the MWAIT, when
another logical processor performs a write access to a WB cacheable address within the
address range used to perform the MONITOR operation. Due to this erratum, a logical
processor may not resume execution until the next targeted interrupt event or O/S timer
tick following a locked store that spans across cache lines within the monitored address
range.
Implication:
The logical processor that executed the MWAIT instruction may not resume execution
until the next targeted interrupt event or O/S timer tick in the case where the monitored
address is written by a locked store which is split across cache lines.
Workaround:
Do not use locked stores that span cache lines in the monitored address range.
Status:
For the steppings affected, see the Summary Tables of Changes.
AZ30.
Programming the Digital Thermal Sensor (DTS) Threshold May Cause
Unexpected Thermal Interrupts
Problem:
Software can enable DTS thermal interrupts by programming the thermal threshold and
setting the respective thermal interrupt enable bit. When programming DTS value, the
previous DTS threshold may be crossed. This will generate an unexpected thermal
interrupt.
Implication:
Software may observe an unexpected thermal interrupt occur after reprogramming the
thermal threshold.
Workaround:
In the ACPI/OS implement a workaround by temporarily disabling the DTS threshold
interrupt before updating the DTS threshold value.
Status:
For the steppings affected, see the Summary Tables of Changes.
AZ31.
Writing Shared Unaligned Data that Crosses a Cache Line without
Proper Semaphores or Barriers May Expose a Memory Ordering Issue
Problem:
Software which is written so that multiple agents can modify the same shared unaligned
memory location at the same time may experience a memory ordering issue if multiple
loads access this shared data shortly thereafter. Exposure to this problem requires the
use of a data write which spans a cache line boundary.
Implication:
This erratum may cause loads to be observed out of order. Intel has not observed this
erratum with any commercially available software or system.
Workaround:
Software should ensure at least one of the following is true when modifying shared data
by multiple agents:
The shared data is aligned
Proper semaphores or barriers are used in order to prevent concurrent data
accesses.
Status:
For the steppings affected, see the Summary Tables of Changes.