JDSU 2111A-10SLHP Digital expansion module

Description

JDSU 2111A-10SLHP Digital expansion module
① fully shadowed. When an embedded system program is run, all code is copied from non-volatile memory (Flash, ROM, etc.) to RAM to run.

② demand paging. Copy only part of the code into RAM. This approach does import/export management of pages in RAM, and only maps code and data to RAM if access is in virtual memory but not in physical RAM, resulting in page errors.

③ eXecute In Place (XIP). At system startup, the code is not copied to RAM, but is executed directly in a non-volatile storage location. RAM holds only the part of the data that needs to be constantly changing, as shown in Figure 1. If the reading speed of non-volatile memory is similar to RAM, XIP can save time for copying and decompressing. NOR Flash and ROM read speed is relatively fast (about 100 ns), suitable for XIP; The read operation of NAND Flash is sector-based, and the speed is relatively slow (μs level), so it is not suitable to implement XIP.

Figure 1 Comparison of full mapping and XIP

XIP can be divided into the following two types:

① Kernel XIP. Running the kernel directly in Flash/ROM saves time on copying and image decompression. The Linux 2.6.10 kernel already includes XIP support.

② Application program XIP. Execute the application code directly from where it is stored, rather than loading it into RAM, so that the first execution of the application is faster. Use of application XIP should be based on the file system that supports it.