# MPU demo LPC1768 link script problem

**URL:** <https://forums.freertos.org/t/mpu-demo-lpc1768-link-script-problem/1332>\
**Category:** Kernel\
**Created:** [July 8, 2015, 4:51am UTC](https://forums.freertos.org/t/mpu-demo-lpc1768-link-script-problem/1332 "2015-07-08T04:51:24Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![system](https://d2v6vdsk2p900z.cloudfront.net/original/1X/7fc8fdb729c645aab5b20282bbe34e57cb407be6.jpeg) [@system](https://forums.freertos.org/u/system)\
**Post date:** [July 8, 2015, 4:51am UTC](https://forums.freertos.org/t/mpu-demo-lpc1768-link-script-problem/1332/1 "2015-07-08T04:51:24Z")

</div>

**[docsis](https://sourceforge.net/u/docsis/)** wrote on [Wednesday, July 08, 2015](https://sourceforge.net/p/freertos/discussion/382005/thread/3bfa786c/#f58d):

In file rtosdemo\_rdb1768\_Debug.ld, The privileged\_data section simply set as bss:

```
privileged_data :
{
	_bss = .;
	*(privileged_data)
	/* Non kernel data is kept out of the first 256 bytes of SRAM. */
} > SRAM	

```

But in task.c, not all variable is zero initial value:

```
PRIVILEGED_DATA static volatile TickType_t xNextTaskUnblockTime = portMAX_DELAY;

```

Not sure if this is an issue.

---

<div class="post-metadata">

**Author:** ![rtel](https://d3iykfbgfh8zhg.cloudfront.net/user_avatar/forums.freertos.org/rtel/32/31_2.png) [@rtel](https://forums.freertos.org/u/rtel)\
**Post date:** [July 9, 2015, 10:32am UTC](https://forums.freertos.org/t/mpu-demo-lpc1768-link-script-problem/1332/2 "2015-07-09T10:32:25Z")

</div>

**[rtel](https://sourceforge.net/u/rtel/)** wrote on [Thursday, July 09, 2015](https://sourceforge.net/p/freertos/discussion/382005/thread/3bfa786c/#ba9d):

Thanks for pointing this out - it is an interesting one and I’m afraid I don’t have an answer for how it worked _however_ the head revision code no longer initialises this variable where it is declared, and instead initialises it immediately before the scheduler is started. The change was made explicitly to ensure the kernel itself had nothing outside the bss section.

You can see the head revision version here:  
[https://sourceforge.net/p/freertos/code/HEAD/tree/trunk/FreeRTOS/Source/tasks.c#l250](https://sourceforge.net/p/freertos/code/HEAD/tree/trunk/FreeRTOS/Source/tasks.c#l250)

Regards.

---

<div class="post-metadata">

**Author:** ![system](https://d2v6vdsk2p900z.cloudfront.net/original/1X/7fc8fdb729c645aab5b20282bbe34e57cb407be6.jpeg) [@system](https://forums.freertos.org/u/system)\
**Post date:** [July 10, 2015, 12:26am UTC](https://forums.freertos.org/t/mpu-demo-lpc1768-link-script-problem/1332/3 "2015-07-10T00:26:10Z")

</div>

**[docsis](https://sourceforge.net/u/docsis/)** wrote on [Friday, July 10, 2015](https://sourceforge.net/p/freertos/discussion/382005/thread/3bfa786c/#ba9d/06fb):

Thank you for your reply. Yes, this is the only privileged variable with no-zero initial value.

There are some other minor issues in link script, maybe also not issues:

1. If enable timer, \_Privileged\_Data\_Region\_Size exceed 256.

2. If change flash start address to other value than zero, link fail. Have to change:

to

```
.text :
	{
		/* Non privileged code kept out of the first 16K or flash. */
		. = ORIGIN( FLASH ) + _Privileged_Functions_Region_Size;

```

In my environment, need change these to get right link result.  
3. Also have to change this to put .bss after \_Privileged\_Data\_Region\_Size.

```
. = ORIGIN( SRAM ) + _Privileged_Data_Region_Size;

.bss :
{
*(.bss*)
*(COMMON)
_ebss = .;
} > SRAM

```

to

```
.bss :
{
. = ORIGIN( SRAM ) + _Privileged_Data_Region_Size;
*(.bss*)
*(COMMON)
_ebss = .;
} > SRAM
```