UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

AR# 70435

Virtex UltraScale+ HBM Controller - AXI_RRESP Can Be Incorrect in Some Conditions

Description

Version Found: HBM v1.0

Version Resolved: See (Xilinx Answer 69267)


With ES HMB devices, the AXI read data slave response (RRESP) can be erroneous in either of the following conditions:

  • I enable ECC correction and enable ECC scrubbing
  • I enable ECC correction and a partial word write (read-modify-write) operation

In these combinations, the RRESP might not display an error when there is an uncorrectable ECC error detected, or it might display an error when there is none.  

See the figure below.




Solution

To work around this issue, do not use either ECC scrubbing or Partial Word Writes in combination with ECC Correction when using HBM ES devices.

These issues are fixed in production HBM devices.

Change log

05/15/2018Updated Solution
06/04/2018removed parity retry from the list of conditions that cause behavior
06/18/2018added figure
03/13/2019Updated to indicate this behavior only occurs with ES devices and is fixed in production devices
AR# 70435
Date 04/01/2019
Status Active
Type Known Issues
Devices
Tools
IP
Page Bookmarked