PASADENA, Calif. Engineers have found the root cause of a computer reset that occurred two months ago on NASA's Mars Science Laboratory and have determined how to correct it.
The fix involves changing how certain unused data-holding locations, called registers, are configured in the memory management of the type of computer chip used on the spacecraft. Billions of runs on a test computer with the modified register configuration yielded no repeat of the reset behavior. The mission team made this software change on the spacecraft's computer last week and confirmed this week that the update is successful.
The reset occurred Nov. 29, 2011, three days after launch, during use of the craft's star scanner. The cause has been identified as a previously unknown design idiosyncrasy in the memory management unit of the Mars Science Laboratory computer processor. In rare sets of circumstances unique to how this mission uses the processor, cache access errors could occur, resulting in instructions not being executed properly. This is what happened on the spacecraft on Nov. 29.
"Good detective work on understanding why the reset occurred has yielded a way to prevent it from occurring again," said Mars Science Laboratory Deputy Project Manager Richard Cook of NASA's Jet Propulsion Laboratory, Pasadena, Calif. "The successful resolution of this problem was the outcome of productive teamwork by engineers at the computer manufacturer and JPL."
The Mars-bound spacecraft performed a brief alignment activity using its star scanner and sun sensor on Jan. 26. During the alignment observations, the star scanner detected Mars.
The fix involves changing how certain unused data-holding locations, called registers, are configured in the memory management of the type of computer chip used on the spacecraft. Billions of runs on a test computer with the modified register configuration yielded no repeat of the reset behavior. The mission team made this software change on the spacecraft's computer last week and confirmed this week that the update is successful.
The reset occurred Nov. 29, 2011, three days after launch, during use of the craft's star scanner. The cause has been identified as a previously unknown design idiosyncrasy in the memory management unit of the Mars Science Laboratory computer processor. In rare sets of circumstances unique to how this mission uses the processor, cache access errors could occur, resulting in instructions not being executed properly. This is what happened on the spacecraft on Nov. 29.
"Good detective work on understanding why the reset occurred has yielded a way to prevent it from occurring again," said Mars Science Laboratory Deputy Project Manager Richard Cook of NASA's Jet Propulsion Laboratory, Pasadena, Calif. "The successful resolution of this problem was the outcome of productive teamwork by engineers at the computer manufacturer and JPL."
The Mars-bound spacecraft performed a brief alignment activity using its star scanner and sun sensor on Jan. 26. During the alignment observations, the star scanner detected Mars.
No comments:
Post a Comment