Nucleo64 STM32F091RC "Device not found in database"

Question

Nucleo64 STM32F091RC "Device not found in database"

diond31 opened this issue 7 years ago · comments

I would like to develop on the STM32 M0 line, but unfortunately QSTLink2 responds with "Device not found in database!" when trying to connect.

Fabien Poussin commented 7 years ago

Yes

W. Tad Morgan · Answer 1 · Fri Jun 23 2017 07:13:02 GMT+0800 (China Standard Time)

Same here:

3714 - Info: ChipID: 0x442
3715 - Error: Did not find chipID!

W. Tad Morgan · Answer 2 · Fri Jun 23 2017 09:35:14 GMT+0800 (China Standard Time)

Is it as simple as putting this in the devices.xml file?

<device type="STM32F09x" coretype="CM0">
  <core_id>0x0bb11477</core_id>
  <chip_id>0x442</chip_id>
  <flash_size_reg>0x1FFFF7CC</flash_size_reg>
  <flash_int_reg>0x40022000</flash_int_reg>
  <buffer_size>0x1000</buffer_size>
  <loader>loader_f0.bin</loader>
</device>

diond31 · Answer 3 · Fri Jun 23 2017 23:57:25 GMT+0800 (China Standard Time)

Tad, it is that "simple"...Brilliant!

It used to say 32 devices and connect resulted in "Device not found in database! Disconnecting... Disconnected..." message that so many others have experienced.

For others that might find this thread useful: you will have to change permissions to devices.xml prior to editing.

Thanks again Tad

Adrian Fewster · Answer 4 · Sat Jun 24 2017 00:02:44 GMT+0800 (China Standard Time)

Are the missing parameters required? I guess not ; <device type="STM32F09x" coretype="CM0"> <core_id>0x0bb11477</core_id> <chip_id>0x442</chip_id> <flash_size_reg>0x1FFFF7CC</flash_size_reg> <flash_int_reg>0x40022000</flash_int_reg> <loader>loader_f4.bin</loader> <SR_BSY>0x10</SR_BSY> <CR_STRT>0x10</CR_STRT> <CR_LOCK>0x1F</CR_LOCK> <CR_SER>0x1</CR_SER> <CR_PGSIZE>0x8</CR_PGSIZE> <buffer_size>0x1000</buffer_size> <loader>loader_f0.bin</loader> </device>

…

On 23/06/17 11:35, W. Tad Morgan wrote: |<device type="STM32F09x" coretype="CM0"> <core_id>0x0bb11477</core_id> <chip_id>0x442</chip_id> <flash_size_reg>0x1FFFF7CC</flash_size_reg> <flash_int_reg>0x40022000</flash_int_reg> <buffer_size>0x1000</buffer_size> <loader>loader_f0.bin</loader> </device>| On 24/06/17 01:57, diond31 wrote: Tad, it is that "simple"...Brilliant! image <https://user-images.githubusercontent.com/22030039/27490049-b361e4e2-57f1-11e7-8b1e-a34ee010e906.png> It used to say 33 devices and connected resulted in "Device not found in database! Disconnecting... Disconnected..." message that so many others have experienced. Thanks again Tad — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#38 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIVGYCn4oy7_5P9LIB-G9unPzmfB_cJIks5sG-BmgaJpZM4Ntqil>.

Fabien Poussin · Answer 5 · Sat Jun 24 2017 00:03:44 GMT+0800 (China Standard Time)

Glad you could fix it by yourself!
Can you submit a pull request to merge the fix?

Thanks

Adrian Fewster · Answer 6 · Sat Jun 24 2017 00:08:10 GMT+0800 (China Standard Time)

I just asked a simple question if the missing parameters are required as for the format shown for the other types and I gave a guessed example. I had a similar situation with the STM32F446RE a couple of months back,. Tads example for this one lacks three lines seen in the others, is that important ?

…

On 24/06/17 02:03, Fabien Poussin wrote: Glad you could fix it by yourself! Can you submit a pull request to merge the fix? Thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#38 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIVGYBeHGcavWOdZtBKO8YMIlgkjVwzcks5sG-HigaJpZM4Ntqil>.

Fabien Poussin · Answer 7 · Sat Jun 24 2017 00:20:46 GMT+0800 (China Standard Time)

Yes that should be correct.

diond31 · Answer 8 · Sat Jun 24 2017 00:27:51 GMT+0800 (China Standard Time)

fpoussin...I'll pull the devices.xml and supply a merge request.

vk4tux...I suspect in some cases they are, but I don't have a complete understanding of the xml file parameters as they are related to Qstlink2, the ARM family of architectures, or the device specific data sheets.

This fix was a cut and paste of the "STM32F07x" XML and changing with tadmorgan's fix (which I verified by watch the terminal window after running "qstlink2 -v"). I know the Nucelo-STM32F07x only has 128k of flash, so I expected the flash size to be wrong as the Nucleo-STM32F09x has 256k, but the screen shot above shows that Qstlink2 found all 256k.

Fabien Poussin · Answer 9 · Sat Jun 24 2017 00:31:39 GMT+0800 (China Standard Time)

Flash size is detect at runtime by reading a register, so it should always be correct.

You just need to make sure the buffer size is not bigger than the ram (0X1000 here, so 8Kb)

diond31 · Answer 10 · Sat Jun 24 2017 00:32:18 GMT+0800 (China Standard Time)

reopen to pull request

diond31 · Answer 11 · Sat Jun 24 2017 01:56:18 GMT+0800 (China Standard Time)

@fpoussin the STM32F09x has 32Kb of SRAM. Would changing <buffer_size>0x8000</buffer_size> allow access to all 32Kb of SRAM?