cpu/stm32/usbdev_fs: allow inverted disconnect GPIO #21718
Conversation
maribu
added
Type: new feature
github-actions
bot
added
Platform: ARM
Did you also test it with a board where the pin is not inverted? One of the upstream boards? I think I have a board with USB peripheral somewhere to test this. |
|
No, I "tested" this by "staring hard at the diff". But I'm having a Bluepill somewhere in reach, I'll give it a spin in a sec. |
|
Tested just now: So even re-enumeration after reboot (the important part) does still work. |
There was a problem hiding this comment.
I wonder if the file should rather be in the cpu_stm32_usbdev subgroup [3], because currently the entries are not added to that group and the cpu_stm32 page also does not show the content explicitly [1]. You have to nagivate to the file [2], which is not easy to find IMO.
[1] https://ci.riot-os.org/results/76a1ab86808d47e8ba3d70c69f14fcec/doc-preview/group__cpu__stm32.html
[2] https://ci.riot-os.org/results/76a1ab86808d47e8ba3d70c69f14fcec/doc-preview/cpu__usbdev_8h.html
[3] https://ci.riot-os.org/results/76a1ab86808d47e8ba3d70c69f14fcec/doc-preview/group__cpu__stm32__usbdev.html
| #define USBDEV_CPU_DMA_REQUIREMENTS __attribute__((aligned(USBDEV_CPU_DMA_ALIGNMENT))) | ||
|
|
||
| /** | ||
| * @name Flags used in @ref stm32_usbdev_fs_config_t |
There was a problem hiding this comment.
This doesn't appear to work so great: https://ci.riot-os.org/results/76a1ab86808d47e8ba3d70c69f14fcec/doc-preview/cpu__usbdev_8h.html
maribu
force-pushed
the
cpu/stm32/usbdev_fs/inverted-disconnect-pin
branch
from
2e0f128 to
3b8fef7
Compare
|
Ah, we could do better having the different two flavors of the USB peripheral more prominently appear in the doc. I'll rework it here soonish |
The STM32F3 requires a dedicated digital signal to emulate a disconnect event by pulling D+ down via a 1.5 kOhm resistor. Some boards, such as the OLIMEXINO-STM32F3, do not directly connect a GPIO but place a transistor in between. Depending on the exact implementation, the logic level may end up being inverted compared to directly connecting a GPIO. This adds a flag member to the `stm32_usbdev_fs_config_t` and a new flag to indicate inverted logic. In addition the members in the struct are sorted by alignment, as this is a foolproof algorithm to prevent wasting memory on unneeded padding. Finally, the USB driver is adapted to honor the flag.
maribu
force-pushed
the
cpu/stm32/usbdev_fs/inverted-disconnect-pin
branch
from
3b8fef7 to
3e6e965
Compare
Contribution description
The STM32F3 requires a dedicated digital signal to emulate a disconnect event by pulling D+ down via a 1.5 kOhm resistor. Some boards, such as the OLIMEXINO-STM32F3, do not directly connect a GPIO but place a transistor in between. Depending on the exact implementation, the logic level may end up being inverted compared to directly connecting a GPIO.
This adds a flag member to the
stm32_usbdev_fs_config_tand a new flag to indicate inverted logic. In addition the members in the struct are sorted by alignment, as this is a foolproof algorithm to prevent wasting memory on unneeded padding.Finally, the USB driver is adapted to honor the flag.
Testing procedure
No regressions on STM32 boards with the
usbdev_fsUSB peripheral (e.g. Nucleo-F303Ze, Bluepill with STM32F1, p-Nucleo-WB55).We do not have a board upstream that has an inverted disconnect pin yet, but I pinky-promise I tested this with one downstream for which this fixed an issue of USB not re-enumerating on reboot.
Issues/PRs references
None