|
Hi, I used STM8s105k6 to design our products with floating arithmetic
libraries (fast floating library is deactivated)
When I enter to debug mode using Rlink on RIDE7, target worked well.
And after starting the debug session, i can press CTRL+D to exit debug
mode, and can confirm the target is still working.
But when i reclosing the target power, it look suspiciously likefloating
calculation routine is skipped.
Could someone tell me why target is working during debug session and
not working during no-debug session ?
Raisonance Version : 7.30.10.0159
STM8 2.32.10.0307
Project code size: 17046, data : 1280
STACK SIZE: 0x200 (Default)
Fast Floating Library : Deactivated
Optimization : 0
Default Global Memory : Data
Here is the map file result....
TYPE BASE LENGTH RELOCATION SEGMENT NAME
---- ---- ------ ---------- ------------
* * * * * * * * D A T A M E M O R Y * * * * * * * *
DATA 000000H 000004H PAGE0 CSTDLIB_REGISTERS
DATA 000004H 00001CH INSECTION0 ?ZDT?...
DATA 000020H 000015H INSECTION0 ?ZDT?...
DATA 000035H 000012H INSECTION0 ?ZDT?...
DATA 000047H 000010H INSECTION0 ?ZDT?...
DATA 000057H 000010H INSECTION0 ?ZDT?...
DATA 000067H 000009H INSECTION0 ?ZDT?...
DATA 000070H 000003H INSECTION0 ?ZDT?HAL_LED
DATA 000073H 000002H INSECTION0 ?ZDT?HAL_UART
DATA 000075H 000001H INSECTION0 ?ZDT?HAL_OC
DATA 000076H 000001H INSECTION0 ?ZDT?HAL_SCHEDULER
DATA 000077H 000022H INSECTION0 ?DT?...
DATA 000099H 000014H INSECTION0 ?DT?...
DATA 0000ADH 00000CH INSECTION0 ?DT?...
DATA 0000B9H 000004H INSECTION0 ?DT?...
DATA 0000BDH 000092H INSECTION0 ?EDT?...
DATA 00014FH 00003AH INSECTION0 ?EDT?...
DATA 000189H 00000CH INSECTION0 ?EDT?...
DATA 000195H 000009H INSECTION0 ?EDT?...
DATA 00019EH 000008H INSECTION0 ?EDT?HAL_UART
DATA 0001A6H 00015AH INSECTION ?PR??HAL_EEPROM_PROGRAM_BLOCK_OPERATION?HAL_EEPROM
000300H 000300H *** GAP ***
DATA 000600H 000200H UNIT ?C_STACK?SEG
|
Hi,
Looks like an interesting problem...
There is not much difference between debug mode and normal mode, except that in debug the whole flash is writeable, which sometimes changes the algorithms (when they are not correctly written).
Can you try to reduce the code and show us a (short) piece of code that exhibits the issue, so that we can reproduce it in our Labs?
Thanks,
Hi, Bruno-san
Thanks for your reply and information about the difference between debug and release mode.
At first, I tried to move some variables from DATA space to PAGE0 space.
(When I checked the map file, it gave me an advise as "SUGGESTED OPTIMIZATION".)
After moving some variables to page0 (by adding page0 declaration in from of variable type),
floating point routine works well.
I'd like to confirm which part was the cause. so, I'm continuing the code inspection now.
Here is the cut of code (floating point routine), but it can't regenerate problem...and it's difficult to cut the code...
When i use the former code (Set #if directive), problem ocurred. but when i used the latter
code, problem didn't occur.
-------------------------------
#define NZEROS 2
#define NPOLES 2
#define GAIN 2.502908710e+02
static float xv[NZEROS+1], yv[NPOLES+1];
uint8_t sensor_get_data(int16_t * p_sensor)
{
int16_t temp;
/* Fetch Data */
if (hal_sensor_io_get_data() == NG)
{
/* Reset sensor */
hal_sensor_io_reset();
return NG;
}
else
{
/* Do Nothing */
};
/* Store Current Data */
temp = ((p_g_sensor_packet->data - g_sensor_offset_std) / cg_hal_sensor_io_scale_factor[0]);
xv[0] = xv[1];
xv[1] = xv[2];
xv[2] = (float)temp * 0.003621682151492864300;
yv[0] = yv[1];
yv[1] = yv[2];
yv[2] =
xv[0] +
2 * xv[1] +
xv[2] +
( -0.8371816513 * yv[0]) +
( 1.8226949252 * yv[1]);
*p_sensor = (int16_t)yv[2];
return OK;
}
#else
uint8_t sensor_get_data(int16_t * p_sensor)
{
/* Fetch Data */
if (hal_sensor_io_get_data() == NG)
{
/* Reset sensor */
hal_sensor_io_reset();
return NG;
}
else
{
/* Do Nothing */
};
/* Store Current Data */
*p_sensor = ((p_g_sensor_packet->data - g_sensor_offset_std) / cg_hal_sensor_io_scale_factor[0]);
return OK;
}
#endif