Advanced Micro Devices Corporation

Other

Ramkrishna Vepa

Sherman E Lee

Leonid Grossman

Robert Alan Williams

Din-I Tsai

David Gary Roberts

Jiu An

Glen William Gibson

Henry Yeh

A system and method for facilitates a fast transmission of packet information into the buffers without unnecessary delays, thereby increasing overall system performance. The method and system comprises sending a packet of information located in a local buffer to a media, generating an interrupt signal indicating a completed transfer of the packet to the media and availability of the local buffer for receiving a next packet, and sending the next packet of information from a queue to the local buffer in response to the interrupt signal, wherein the generation of the interrupt signal after each data packet is transmitted to the media does not affect the overall operation of the processing system.

Method and system for fast data transmissions in a processing system

A method for identifying an error condition due to a faulty cable connection in a network comprising a plurality of computer systems, at least one of the computer systems including a network adapter with the network adapter including a media access control (MAC) unit, includes initializing a plurality of mechanisms for tracking a plurality of conditions in the MAC of the network adapter. The method further includes receiving a transmit demand request in the network adapter and updating the plurality of mechanisms according to a current status of each of the plurality of conditions. In addition, the method includes determining whether a predetermined threshold has been reached in one or more of the plurality of mechanisms, wherein when one or more of the plurality of mechanisms has reached the predetermined threshold, a faulty cable connection is identified. In a system aspect, the system includes a plurality of counting mechanism for tracking each of a plurality of error conditions. The system further includes a media access control (MAC) unit for differentiating each of the plurality of error conditions. In addition, the system includes an error identification mechanism for determining when any of the plurality of counting mechanisms has reached a predetermined threshold.

Method and system for identifying an error condition due to a faulty

A data transfer method with peripheral precharge wherein a starting portion of an output data block targeted for a peripheral is transferred to an output buffer for the peripheral using programmed I/O or slave cycles and wherein a virtual address of the output data block is translated into a physical address while the peripheral begins transferring the starting portion from the output buffer and over an output path. A direct memory access operation by the peripheral transfers a remainder portion of the output data block into the output buffer.

Precharging an output peripheral for a direct memory access operation

A system and method for detecting and gracefully recovering from a peripheral device fault has been disclosed. The method detects whether a peripheral device has suffered from a peripheral device fault. Where the peripheral device fault has occurred, the method determines whether any of a plurality of processes executable by the peripheral device is currently being executed by the peripheral device. The plurality of processes comprises those process which could result in significant loss of data, loss of connection to a network or adversely affect the performance of the peripheral device if the peripheral device is reset during execution of any of the plurality of processes. If none of the plurality of processes is being executed by the peripheral device, the method automatically resets the peripheral device. According to the method and system disclosed, peripheral devices can be made to recover from faults without user intervention, without loss of connection to any networks, and with minimal loss of data. Reliability is improved, thereby increasing overall system performance.

Method and system for detection of and graceful recovery from a

A method and system for resetting a peripheral device which could use a variety of buses have been disclosed. The method and system determine what bus type the peripheral device has. The method and system then automatically execute a reset process capable of resetting the peripheral device having that bus type. A beneficial aspect of this invention is allowing a peripheral device to gracefully recover from a fault. In this aspect, the method detects whether a peripheral device fault has occurred. Where the fault has occurred, the method determines whether any of a plurality of processes executable by the peripheral device is being executed. The processes comprise those processes which could result in significant loss of data, loss of connection to a network, or adversely affect performance if the peripheral device is reset during execution. If none of the processes is being executed, the method automatically resets the peripheral device. The automatic reset process determines the bus type of the peripheral device. The automatic reset process then automatically executes a reset process capable of resetting the peripheral device having that bus type. According to the method and system, a peripheral device which may use a variety of buses can be reset or made to recover from faults without user intervention, without loss of connection to any networks, and with minimal loss of data.

Growing community of inventors

Fremont, CA, United States of America

Ramkrishna Vepa