Ah, bit OPTCR_DB1M of the STM32_FLASH_OPTCR is used on stm32f2x.c:930 to
determine of the MCU has one or two flash banks... (42x/43x/469/479 1024 kiByte
devices).
And bit OPTCR_NDBANK of STM32_FLASH_OPTCR for F76x/77x devices right below.
~~~
/* Devices with > 1024 kiByte always are dual-banked */
if (flash_size_in_kb > 1024)
stm32x_info->has_large_mem = true;
/* F42x/43x/469/479 1024 kiByte devices have a dual bank option */
if ((device_id & 0xfff) == 0x419 || (device_id & 0xfff) == 0x434) {
uint32_t optiondata;
retval = target_read_u32(target, STM32_FLASH_OPTCR,
&optiondata);
if (retval != ERROR_OK) {
LOG_DEBUG("unable to read option bytes");
return retval;
}
if ((flash_size_in_kb > 1024) || (optiondata & OPTCR_DB1M)) {
stm32x_info->has_large_mem = true;
LOG_INFO("Dual Bank %d kiB STM32F42x/43x/469/479
found", flash_size_in_kb);
} else {
stm32x_info->has_large_mem = false;
LOG_INFO("Single Bank %d kiB STM32F42x/43x/469/479
found", flash_size_in_kb);
}
}
/* F76x/77x devices have a dual bank option */
if ((device_id & 0xfff) == 0x451) {
uint32_t optiondata;
retval = target_read_u32(target, STM32_FLASH_OPTCR,
&optiondata);
if (retval != ERROR_OK) {
LOG_DEBUG("unable to read option bytes");
return retval;
}
if (optiondata & OPTCR_NDBANK) {
stm32x_info->has_large_mem = false;
LOG_INFO("Single Bank %d kiB STM32F76x/77x found",
flash_size_in_kb);
} else {
stm32x_info->has_large_mem = true;
max_sector_size_in_kb >>= 1; /* sector size divided by
2 in dual-bank mode */
LOG_INFO("Dual Bank %d kiB STM32F76x/77x found",
flash_size_in_kb);
}
}
~~~
*Trying to figure out if that is possible for STM32Lx MCUs as well...*
---
** [tickets:#148] STM32L0x: flash size and dual bank support**
**Status:** new
**Milestone:** 0.9.0
**Created:** Sun Mar 19, 2017 01:47 AM UTC by Aurelio Lucchesi
**Last Updated:** Sun Mar 19, 2017 03:42 PM UTC
**Owner:** nobody
Flash programming large binaries to my STM32L082 board failed,
and I had to recompile OpenOCD in order to make it work.
Without modification, it looks like this:
~~~
$ openocd -f interface/stlink-v2.cfg -f target/stm32l0.cfg -c "program
build/prj.elf reset exit verify"
...
Info : Device: STM32L0xx (Cat.5)
Info : STM32L flash has dual banks. Bank (0) size is 128kb, base address is
0x8000000
...
Error: checksum mismatch - attempting binary compare
diff 0 address 0x08020000. Was 0x3e instead of 0x69
diff 1 address 0x08020001. Was 0x79 instead of 0x21
diff 2 address 0x08020002. Was 0x79 instead of 0x3d
diff 3 address 0x08020003. Was 0x74 instead of 0x59
diff 4 address 0x08020004. Was 0x6f instead of 0x59
diff 5 address 0x08020005. Was 0x73 instead of 0x5f
...
~~~
Then I've realized that there are two flash program memory banks:
STM32L0xx category 5 MCUs come with 192 KiB of Flash program memory:
* Bank 1: starting at 0x0800 0000, size 96 KiB
* Bank 2: starting at 0x0801 8000, size 96 KiB
So I've added a second one to a copy of the stm32l0.cfg script:
~~~
# flash size will be probed
set _FLASHNAME $_CHIPNAME.flash
flash bank $_FLASHNAME stm32lx 0x08000000 0 0 0 $_TARGETNAME
# add second flash bank
set _FLASHNAME $_CHIPNAME.flash1
flash bank $_FLASHNAME stm32lx 0x08018000 0x18000 0 0 $_TARGETNAME
~~~
But that revealed another problem:
~~~
$ openocd -f interface/stlink-v2.cfg -f ./stm32l0.cfg -c "program
build/prj.elf reset exit verify"
...
Info : Device: STM32L0xx (Cat.5)
Info : STM32L flash has dual banks. Bank (0) size is 128kb, base address is
0x8000000
Warn : couldn't use loader, falling back to page memory writes
Info : Device: STM32L0xx (Cat.5)
Warn : STM32L flash bank base address config is incorrect. 0x8018000 but should
rather be 0x8000000 or 0x8020000
Error: auto_probe failed
** Programming Failed **
shutdown command invoked
~~~
So I tried to fix things by rebuilding OpenOCD, and changed the size of
the first (and second) flash page in src/flash/nor/stm32lx.c:261 from
128 to 96 KiB:
~~~
{
.id = 0x447,
.revs = stm32_447_revs,
.num_revs = ARRAY_SIZE(stm32_447_revs),
.device_str = "STM32L0xx (Cat.5)",
.page_size = 128,
.pages_per_sector = 32,
.max_flash_size_kb = 192,
.first_bank_size_kb = 96, // 128,
.has_dual_banks = true,
.flash_base = 0x40022000,
.fsize_base = 0x1FF8007C,
},
~~~
That actually works for my hardware with the modified cfg script:
~~~
$ ~/git/openocd/src/openocd -f interface/stlink-v2.cfg -f ./stm32l0.cfg -c
"program build/prj.elf reset exit verify" --search
~/git/openocd/tcl
Open On-Chip Debugger 0.10.0+dev-00093-g6b2acc02 (2017-03-18-23:04)
...
Info : Device: STM32L0xx (Cat.5)
Info : STM32L flash has dual banks. Bank (0) size is 96kb, base address is
0x8000000
Warn : couldn't use loader, falling back to page memory writes
Info : Device: STM32L0xx (Cat.5)
Info : STM32L flash has dual banks. Bank (1) size is 96kb, base address is
0x8018000
Info : ignoring flash probed value, using configured bank size: 96kbytes
...
** Verified OK **
** Resetting Target **
~~~
So I've asked on IRC and tried to submit a really simple patch for that:
http://openocd.zylin.com/#/c/4073/
But like Cezary pointed out: This **does not solve the problem of all STM32L0**
Category 5 (or 3) devices, since they all have a different memory layout...
See also:
http://openocd.zylin.com/#/c/3554/
STM32L0x1 Ref Manual: http://st.com/resource/en/reference_manual/DM00108282.pdf
STM32L0x2 Ref Manual: http://st.com/resource/en/reference_manual/DM00108281.pdf
STM32L0x3 Ref Manual: http://st.com/resource/en/reference_manual/DM00095744.pdf
I don't have access to other hardware to test things out nor enough knowledge
about OpenOCD to really fix the issue for all cases...
-----
Possible solutions that came to mind:
1) Don't use hardcoded size values. - Like Karl Palsson pointed out:
Using the probed flash size (fsize_base/flash_size_in_kb, 0x1FF8004C on
STM32L0x2,
33.1.1 Flash size register) and the device ID (32.4.1 MCU device ID code,
stm32lx_read_id_code()), it should be possible to figure out the exact flash
memory layout (size, addresses, dual bank).
2) Hardcode the values for all possible combinations for all STM32L0 MCU
types/categories code the sizes.
3) Maybe the easiest: Make it possible to change first_bank_size_kb in a
config script? Or is this already possible somehow?
I'd like to help fixing this issue...
---
Sent from sourceforge.net because [email protected] is
subscribed to https://sourceforge.net/p/openocd/tickets/
To unsubscribe from further messages, a project admin can change settings at
https://sourceforge.net/p/openocd/admin/tickets/options. Or, if this is a
mailing list, you can unsubscribe from the mailing list.------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
OpenOCD-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openocd-devel
