Specification Update 35
BV52. Instructions Retired Event May Over Count Execution of IRET
Instructions
Problem: Under certain conditions, the performance monitoring event Instructions Retired (Event
C0H, Unmask 00H) may over count the execution of IRET instruction.
Implication: Due to this erratum, performance monitoring event Instructions Retired may over
count.
Workaround: None identified.
Status: For the steppings affected, see the Summary Tables of Changes.
BV53. PCIe* Link May Unexpectedly Exit Loopback State
Problem: The PCIe Port is capable of functioning as 3 independent PCIe controllers. Due to this
erratum, if more than one of the controllers is in Loopback.Active state and configured
as a loopback slave and if any one of these controllers transition to Loopback.Exit, all
controllers in Loopback.Active will transition to Loopback.Exit.
Implication: Loopback.Active state on a given Link may unexpectedly exit. Software should avoid
configuring more than one of the PCIe Controllers as Loopback slave concurrently.
Workaround: PCIe endpoints should avoid configuring more than one of PCIe Controllers as Loopback
slave.
Status: For the steppings affected, see the Summary Tables of Changes.
BV54. The RDRAND Instruction Will Not Execute as Expected
Problem: On processors that support the RDRAND instruction, that capability should be reported
via the setting of CPUID.01H:ECX.RDRAND[bit 30]. Due to this erratum, that bit will
not be set, and the execution of the RDRAND instruction will result in a #UD exception.
Implication: Software will not be able to utilize the RDRAND instruction
Workaround: It is possible for the BIOS to contain a workaround for this erratum to report RDRAND
as present via CPUID and allow proper execution of RDRAND.
Status: For the steppings affected, see the Summary Tables of Changes.
BV55. A PCIe* Device That Initially Transmits Minimal Posted Data Credits
May Cause a System Hang
Problem: Under certain conditions, if a PCIe device that initially transmits posted data credits
less than Max_Payload_Size/16 + 4 (16B/4DW is unit of data flow control) and is the
target of a Peer-to-Peer write of Max_Payload_Size, the system may hang due to
Posted Data credit starvation.
Implication: Under certain conditions, the processor may encounter a Posted Data credit starvation
scenario and hang.
Workaround: A BIOS code change has been identified and may be implemented as a workaround for
this erratum.
Status: For the steppings affected, see the Summary Tables of Changes.
To Next Page
Loading...
Need help?
General
