DocID024207 Rev 1

5/11

AN4246

Flash code protection

1.2

Write Protection

The Write protection, applied by a Flash area (sector), protects the content of the specified

sectors against code update or erase.

One option bit is used to activate the write protection for each Flash sector. When the Write

programmed.

Table 3 shows the sector Write protection depending on the RDP Level.

Note:

Under these conditions, the integrity of the embedded firmware written in these sectors is

guaranteed against any modification.

1.3

Proprietary Code Read Out Protection

The Proprietary Code Read Out Protection (PCROP) is an alternative protection which is

applied also by sector, allowing the protection of specific code (intellectual property) against

attacks.

The PCROP implements 2 main features on the microcontroller code protection and the

code management.

Table 4 compares both PCROP features to the RDP protection method.

This protection is based on an execute-only mechanism. The Flash code area can only be

reached by the STM32 CPU (as an instruction code), while all other accesses (DMA, debug

and CPU data read) are strictly prohibited.

While protecting the executable code against reading, a side effect generated by this

execute-only mechanism makes the protected code itself (executed from this area) unable

to access the associated data values stored in the same area (e.g. literal pool). In order to

avoid the need of data accesses in this area (specially for literal pool accesses), a specific

command line option must be chosen in the ARM/Keil compiler:

(armcc --no_literal_pools --max_string_in_code = 0).

Table 3. Write Protection

Levels

Description

Level 0 or 1

1. The sector Write protection is very important for safety functions. If they are programmed in the

write protected sectors, these functions are fully protected against accidental erase or update.

Level 2

2. A write protected sector cannot be erased or modified, either intentionally or not.

Table 4. Protection against attacks

Type of Protection

Comparison

External attacks

Similar to the protection offered by RDP (but which can be restricted to a

specific Flash area)

Internal attacks (such

as Trojan horse type)

Possible use of some “unsecured” third party code in an application, while

still preserving the privacy of some parts of the code