www.ti.com

Architecture

2.16.1.2Transmit Packet Completion Interrupts

The transmit DMA engine has eight channels, with each channel having a corresponding interrupt (TXPENDn). The transmit interrupts are level interrupts that remain asserted until cleared by the CPU.

Each of the eight transmit channel interrupts may be individually enabled by setting the corresponding bit in the transmit interrupt mask set register (TXINTMASKSET) to 1. Each of the eight transmit channel interrupts may be individually disabled by clearing the corresponding bit in the transmit interrupt mask clear register (TXINTMASKCLEAR) to 0. The raw and masked transmit interrupt status may be read from the transmit interrupt status (unmasked) register (TXINTSTATRAW) and the transmit interrupt status (masked) register (TXINTSTATMASKED), respectively.

When the EMAC completes the transmission of a packet, the EMAC issues an interrupt to the CPU by writing the packet’s last buffer descriptor address to the appropriate channel queue’s transmit completion pointer located in the state RAM block. The interrupt is generated by the write when enabled by the interrupt mask, regardless of the value written.

Upon interrupt reception, the CPU processes one or more packets from the buffer chain and then acknowledges an interrupt by writing the address of the last buffer descriptor processed to the queue’s associated transmit completion pointer in the transmit DMA state RAM.

The data written by the host (buffer descriptor address of the last processed buffer) is compared to the data in the register written by the EMAC port (address of last buffer descriptor used by the EMAC). If the two values are not equal (which means that the EMAC has transmitted more packets than the CPU has processed interrupts for), the transmit packet completion interrupt signal remains asserted. If the two values are equal (which means that the host has processed all packets that the EMAC has transferred), the pending interrupt is cleared. The value that the EMAC is expecting is found by reading the transmit channel n completion pointer register (TXnCP).

The EMAC write to the completion pointer actually stores the value in the state RAM. The CPU written value does not actually change the register value. The host written value is compared to the register content (which was written by the EMAC) and if the two values are equal then the interrupt is removed; otherwise, the interrupt remains asserted. The host may process multiple packets prior to acknowledging an interrupt, or the host may acknowledge interrupts for every packet.

2.16.1.3Receive Packet Completion Interrupts

The receive DMA engine has eight channels, which each channel having a corresponding interrupt (RXPENDn). The receive interrupts are level interrupts that remain asserted until cleared by the CPU.

Each of the eight receive channel interrupts may be individually enabled by setting the corresponding bit in the receive interrupt mask set register (RXINTMASKSET) to 1. Each of the eight receive channel interrupts may be individually disabled by clearing the corresponding bit in the receive interrupt mask clear register (RXINTMASKCLEAR) to 0. The raw and masked receive interrupt status may be read from the receive interrupt status (unmasked) register (RXINTSTATRAW) and the receive interrupt status (masked) register (RXINTSTATMASKED), respectively.

When the EMAC completes a packet reception, the EMAC issues an interrupt to the CPU by writing the packet's last buffer descriptor address to the appropriate channel queue's receive completion pointer located in the state RAM block. The interrupt is generated by the write when enabled by the interrupt mask, regardless of the value written.

Upon interrupt reception, the CPU processes one or more packets from the buffer chain and then acknowledges one or more interrupt(s) by writing the address of the last buffer descriptor processed to the queue's associated receive completion pointer in the receive DMA state RAM.

The data written by the host (buffer descriptor address of the last processed buffer) is compared to the data in the register written by the EMAC (address of last buffer descriptor used by the EMAC). If the two values are not equal (which means that the EMAC has received more packets than the CPU has processed interrupts for), the receive packet completion interrupt signal remains asserted. If the two values are equal (which means that the host has processed all packets that the EMAC has received), the pending interrupt is de-asserted. The value that the EMAC is expecting is found by reading the receive channel n completion pointer register (RXnCP).

SPRUEQ6–December 2007

Ethernet Media Access Controller (EMAC)/Management Data Input/Output (MDIO)

57

Submit Documentation Feedback

Page 56 Page 58
Page 57
Image 57
Texas Instruments TMS320DM646x manual Transmit Packet Completion Interrupts, Receive Packet Completion Interrupts
Page 56 Page 58

TMS320DM646x specifications

The Texas Instruments TMS320DM646x series is a powerful family of digital media processors designed to handle high-performance applications in video, imaging, and audio processing. These devices leverage advanced technologies to deliver efficient processing capabilities for a variety of embedded systems, making them ideal for multimedia solutions.

At the core of the TMS320DM646x is the versatile DSP architecture, which optimizes performance for digital signal processing tasks. This architecture allows for real-time processing, enabling the devices to handle complex algorithms necessary for image and video compression, thereby meeting the rigorous demands of modern multimedia applications.

One of the standout features of the TMS320DM646x series is its dual-core architecture. This consists of a Digital Signal Processor (DSP) alongside an ARM-based application processor. The DSP is predominantly employed for critical processing tasks, allowing it to execute high-throughput data streams efficiently, while the ARM processor manages control tasks and user interfaces. This division of labor enhances overall system performance and responsiveness.

The series supports a wide range of video formats and technologies, including HD video encoding and decoding, which accommodates HD resolution content essential for today’s multimedia applications. Furthermore, the TMS320DM646x integrates hardware accelerators for video compression standards such as H.264 and MPEG-4, which significantly reduce the processing burden on the CPU, resulting in lower power consumption and higher efficiency.

Networking capabilities are another significant feature of the TMS320DM646x. With support for Ethernet, the device can handle streaming media applications and connectivity, facilitating the transmission of high-quality audio and video over the internet. This connectivity is crucial for developing robust IPTV and streaming solutions.

Power management is a primary focus in the design of the TMS320DM646x series. The processors are built to operate efficiently with minimal power consumption, making them suitable for portable and battery-operated devices. The low power characteristics do not compromise performance, enabling high computational capabilities while maintaining energy efficiency.

Overall, the Texas Instruments TMS320DM646x series offers an exceptional combination of processing power, advanced multimedia capabilities, and energy efficiency. It optimally supports a wide array of applications, from video processing and image analysis to audio encoding. This comprehensive feature set, along with its robust architecture, positions the TMS320DM646x as a leading choice for developers in the digital media space.
Top Page Image Contents