![Freescale Semiconductor PXR4030 Reference Manual Download Page 1246](http://html1.mh-extra.com/html/freescale-semiconductor/pxr4030/pxr4030_reference-manual_23306601246.webp)
Enhanced Time Processing Unit (eTPU2)
29-78
Freescale Semiconductor
PXR40 Microcontroller Reference Manual, Rev. 1
a particular function varies with each system. The PWM thread response is faster if there are no other
active channels than if other channels are also active. In addition, changing the priority scheme and
channel number assignments can change performance for a function even if the same set of functions are
still active.
Each function is divided into treads, as shown in
(see also the
eTPU Reference Manual
for
details). The eTPU Microengine executes one thread of a function at a time. For example, the Microengine
might execute thread 1 of PWM, then thread 3 of DIO, then thread 2 of PWM, then thread 2 of SM, and
so on. The amount of time the eTPU Microengine grants a function to execute a thread varies with the
number of microcode instructions in the thread.
Since there is only one eTPU Microengine (in each eTPU Engine), the eTPU cannot actually execute the
software for multiple functions simultaneously. However, the hardware for each of the channels is
independent. This means that, for example, all 32 channel signals can change thread at the same moment,
provided that the function software sets up the channel hardware to do so beforehand.
With Host CPU code, the system designer assigns functions to channels and initializes the functions. After
initialization, functions typically run without Host intervention, except for eTPU channel interrupts to the
Host to give or receive information. Most functions can run continuously with periodic servicing from the
eTPU Microengine. As required, the channels request service from the eTPU Microengine, and the eTPU
Scheduler determines the order in which the channels are serviced. Worst-case latency for a channel can
be derived from the details of the priority scheme that the scheduler uses (see
Figure 29-35. Function Threads
29.4.2.2
Using Worst-Case Latency Estimates to Evaluate Performance
Once WCL is found for a channel, the user must determine how to use this number to analyze performance.
To analyze the performance of a channel running the PWM function, for example, some information about
what happens in each thread is necessary.
DIO Function Threads
PWM Function Threads
SM Function Threads
S1
S2
S3
S4
S5
S6
S1
S2
S3
S4
S5
S6
S1
S2
S3
S4
Summary of Contents for PXR4030
Page 1: ...PXR40 Microcontroller Reference Manual Devices Supported PXR4030 PXR4040 PXR40RM Rev 1 06 2011...
Page 30: ...PXR40 Microcontroller Reference Manual Rev 1 Freescale Semiconductor xxx...
Page 40: ...PXR40 Microcontroller Reference Manual Rev 1 xl Freescale Semiconductor...
Page 66: ...Memory Map PXR40 Microcontroller Reference Manual Rev 1 2 4 Freescale Semiconductor...
Page 120: ...Signal Descriptions 3 54 Freescale Semiconductor PXR40 Microcontroller Reference Manual Rev 1...
Page 860: ...FlexCAN Module 24 50 Freescale Semiconductor PXR40 Microcontroller Reference Manual Rev 1...
Page 1167: ...Decimation Filter Freescale Semiconductor 28 53 PXR40 Microcontroller Reference Manual Rev 1...
Page 1168: ...Decimation Filter 28 54 Freescale Semiconductor PXR40 Microcontroller Reference Manual Rev 1...