EasyManua.ls Logo

Intel T8300 - Core 2 Duo 2.4GHz 800MHz 3MB Socket P Mobile CPU - Page 39

Intel T8300 - Core 2 Duo 2.4GHz 800MHz 3MB Socket P Mobile CPU
56 pages
Print Icon
To Next Page IconTo Next Page
To Next Page IconTo Next Page
To Previous Page IconTo Previous Page
To Previous Page IconTo Previous Page
Loading...
Errata
Specification Update 39
AZ44. NMIs May Not Be Blocked by a VM-Entry Failure
Problem: The Intel® 64 and IA-32 Architectures Software Developer's Manual, Volume 3B: System
Programming Guide, Part 2 specifies that, following a VM-entry failure during or after
loading guest state, “the state of blocking by NMI is what it was before VM entry.” If
non-maskable interrupts (NMIs) are blocked and the “virtual NMIs” VM-execution control
set to 1, this erratum may result in NMIs not being blocked after a VM-entry failure
during or after loading guest state.
Implication: VM-entry failures that cause NMIs to become unblocked may cause the processor to
deliver an NMI to software that is not prepared for it.
Workaround: VMM software should configure the virtual-machine control structure (VMCS) so that VM-
entry failures do not occur.
Status: For the steppings affected, see the Summary Tables of Changes.
AZ45. Partial Streaming Load Instruction Sequence May Cause the Processor
to Hang
Problem: Under some rare conditions, when multiple streaming load instructions (MOVNTDQA) are
mixed with non-streaming loads that split across cache lines, the processor may hang.
Implication: Under the scenario described above, the processor may hang. Intel has not observed this
erratum with any commercially available software.
Workaround: It is possible for the BIOS to contain a workaround for this erratum. However, streaming
behavior may be re-enabled by setting bit 5 to 1 of the MSR at address 0x21 for software
development or testing purposes. If this bit is changed, then a read-modify-write should
be performed to preserve other bits of this MSR. When the streaming behavior is enabled
and using streaming load instructions, always consume a full cache line worth of data
and/or avoid mixing them with non-streaming memory references. If streaming loads are
used to read partial cache lines, and mixed with non-streaming memory references, use
fences to isolate the streaming load operations from non-streaming memory operations.
Status: For the steppings affected, see the Summary Tables of Changes.

Related product manuals