mirror of
https://github.com/felis/USB_Host_Shield_2.0.git
synced 2024-03-22 11:31:26 +01:00
Merge branch 'xxxajk' of github.com:felis/USB_Host_Shield_2.0 into xxxajk
This commit is contained in:
commit
336e5f37aa
1 changed files with 10 additions and 0 deletions
10
Usb.cpp
10
Usb.cpp
|
@ -567,6 +567,16 @@ uint8_t USB::DefaultAddressing(uint8_t parent, uint8_t port, bool lowspeed) {
|
||||||
* This is broken. We need to enumerate differently.
|
* This is broken. We need to enumerate differently.
|
||||||
* It causes major problems with several devices if detected in an unexpected order.
|
* It causes major problems with several devices if detected in an unexpected order.
|
||||||
*
|
*
|
||||||
|
*
|
||||||
|
* Oleg - I wouldn't do anything before the newly connected device is considered sane.
|
||||||
|
* i.e.(delays are not indicated for brevity):
|
||||||
|
* 1. reset
|
||||||
|
* 2. GetDevDescr();
|
||||||
|
* 3a. If ACK, continue with allocating address, addressing, etc.
|
||||||
|
* 3b. Else reset again, count resets, stop at some number (5?).
|
||||||
|
* 4. When max.number of resets is reached, toggle power/fail
|
||||||
|
* If desired, this could be modified by performing two resets with GetDevDescr() in the middle - however, from my experience, if a device answers to GDD()
|
||||||
|
* it doesn't need to be reset again
|
||||||
* New steps proposal:
|
* New steps proposal:
|
||||||
* 1: get address pool instance. exit on fail
|
* 1: get address pool instance. exit on fail
|
||||||
* 2: pUsb->getDevDescr(0, 0, constBufSize, (uint8_t*)buf). exit on fail.
|
* 2: pUsb->getDevDescr(0, 0, constBufSize, (uint8_t*)buf). exit on fail.
|
||||||
|
|
Loading…
Reference in a new issue