Percepio®, the leader in visual trace diagnostics for embedded systems and the Internet of Things (IoT), today released Tracealyzer 4.6 with official support for Zephyr RTOS and Microsoft Azure RTOS ThreadX. The new release also includes Percepio’s next-generation trace recorder library with improved support for snapshot trace. Snapshot recording opens the door to using Percepio DevAlert®, Percepio’s cloud service for monitoring application code executing in deployed IoT devices.
“The new trace recorder library is our next-generation platform for runtime monitoring that benefits both Tracealyzer and DevAlert. This enables us to accelerate development over the coming years and has been crucial for the new support for Azure RTOS ThreadX and Zephyr RTOS,” said Percepio CEO and founder Johan Kraft.
The new trace recorder in Tracealyzer 4.6 is designed to be far easier to port to other software platforms and to allow more efficient monitoring of deployed IoT systems and tracing of multi-core systems. The new recorder library features more efficient snapshot recording, where trace data is saved in a ring buffer on the target system, for both Zephyr and ThreadX, in addition to the previously supported streaming recording.
Johan Kraft adds: “Both streaming trace and snapshot trace are now supported in the same solution and use the same trace format, which makes it far easier to maintain and improve. Moreover, this will soon enable us to provide an official SDK for partners and customers that wish to integrate Tracealyzer and DevAlert with their RTOS of choice, or with bare-metal firmware.”
“Tracealyzer support for Zephyr is a very welcome addition to the embedded and IoT ecosystems. The ability to visualize code execution is a key tool for developers working with resource-constrained devices to be able to efficiently debug defects and release products,” says Kate Stewart, VP of Dependable Embedded Systems, The Linux Foundation.
Another innovation is the new dynamic legend feature, which improves the legibility of Tracealyzer’s timeline views, so users can quickly see which events belong to which process or thread. This is particularly valuable for operating systems such as Linux, where the number of threads can become unwieldy.