Skip to content

micropython/usb/usb-device/usb/device/core.py: Add alternate setting support plus a few fixes and improvements. #1031

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
HLammers wants to merge 3 commits into
base: master
Choose a base branch
from
Open

micropython/usb/usb-device/usb/device/core.py: Add alternate setting support plus a few fixes and improvements. #1031

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
942c80c
Update core.py
HLammers Jul 5, 2025
befd307
Update core.py
HLammers Jul 5, 2025
7424b12
Moved assignment out of assert statement
HLammers Jul 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
39 changes: 24 additions & 15 deletions micropython/usb/usb-device/usb/device/core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,10 +3,11 @@
# These contain the classes and utilities that are needed to
# implement a USB device, not any complete USB drivers.
#
# MIT license; Copyright (c) 2022-2024 Angus Gratton
# MIT license; Copyright (c) 2022-2024 Angus Gratton, 2025 Harm Lammers
from micropython import const
import machine
import struct
import time

@projectgus projectgus Jul 1, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Confused by this as PR description says:

Removed import time (not used)

... and it is indeed not used, from what I can see?


try:
from _thread import get_ident
Expand Down Expand Up @@ -108,7 +109,6 @@ def config( # noqa: PLR0913
device_class=0,
device_subclass=0,
device_protocol=0,
config_str=None,
max_power_ma=None,
remote_wakeup=False,
):
Expand Down Expand Up @@ -166,7 +166,11 @@ def maybe_set(value, idx):
# Keep track of the interface and endpoint indexes
itf_num = builtin_driver.itf_max
ep_num = max(builtin_driver.ep_max, 1) # Endpoint 0 always reserved for control
while len(strs) < builtin_driver.str_max:
while (
len(strs) < builtin_driver.str_max - 1
): # This is possibly unnecessary or wrong because
# https://docs.micropython.org/en/latest/library/machine.USBDevice.html
# states all string values except index 0 should be plain ASCII
Comment on lines +171 to +173

@projectgus projectgus Jul 1, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What's happening here is that we're expanding the strs list to cover the built-in driver indexes, because these str index values are actually handled by C code and never read by Python code. By putting something here we make sure that the indexes elsewhere in the driver match what the host sees.

This is why the documentation says the strings arguments should be strings - the values provided by the Python caller should be strings, these values are internal placeholders to pad the index out.

I assume the off-by-one is a real bug, though - I'm guessing you found that strings were reported wrong without this?

strs.append(None) # Reserve other string indexes used by builtin drivers
initial_cfg = builtin_driver.desc_cfg or (b"\x00" * _STD_DESC_CONFIG_LEN)

Expand Down Expand Up @@ -204,10 +208,11 @@ def maybe_set(value, idx):
)

# Configuration string is optional but supported
iConfiguration = 0
if configuration_str:
iConfiguration = len(strs)
strs.append(configuration_str)
else:
iConfiguration = 0

if max_power_ma is not None:
# Convert from mA to the units used in the descriptor
Expand Down Expand Up @@ -665,6 +670,7 @@ def interface(
bInterfaceSubClass=_INTERFACE_SUBCLASS_NONE,
bInterfaceProtocol=_PROTOCOL_NONE,
iInterface=0,
bAlternateSetting=0,
):
# Utility function to append a standard Interface descriptor, with
# the properties specified in the parameter list.
Expand All @@ -680,7 +686,7 @@ def interface(
_STD_DESC_INTERFACE_LEN, # bLength
_STD_DESC_INTERFACE_TYPE, # bDescriptorType
bInterfaceNumber,
0, # bAlternateSetting, not currently supported
bAlternateSetting,
bNumEndpoints,
bInterfaceClass,
bInterfaceSubClass,
Expand Down Expand Up @@ -791,17 +797,18 @@ class Buffer:
# approximate a Python-based single byte ringbuffer.
#
def __init__(self, length):
self._l = length
self._b = memoryview(bytearray(length))
# number of bytes in buffer read to read, starting at index 0. Updated
# number of bytes in buffer ready to read, starting at index 0. Updated
# by both producer & consumer.
self._n = 0
# start index of a pending write into the buffer, if any. equals
# start index of a pending write into the buffer, if any. Equals
# len(self._b) if no write is pending. Updated by producer only.
self._w = length

def writable(self):
# Number of writable bytes in the buffer. Assumes no pending write is outstanding.
return len(self._b) - self._n
return self._l - self._n

def readable(self):
# Number of readable bytes in the buffer. Assumes no pending read is outstanding.
Expand All @@ -815,15 +822,15 @@ def pend_write(self, wmax=None):
# this many bytes long.
#
# (No critical section needed as self._w is only updated by the producer.)
self._w = self._n
end = (self._w + wmax) if wmax else len(self._b)
return self._b[self._w : end]
self._w = (_w := self._n)
end = (_w + wmax) if wmax else self._l
return self._b[_w:end]

def finish_write(self, nbytes):
# Called by the producer to indicate it wrote nbytes into the buffer.
ist = machine.disable_irq()
try:
assert nbytes <= len(self._b) - self._w # can't say we wrote more than was pended
assert nbytes <= self._l - self._w # can't say we wrote more than was pended
if self._n == self._w:
# no data was read while the write was happening, so the buffer is already in place
# (this is the fast path)
Expand All @@ -834,13 +841,14 @@ def finish_write(self, nbytes):
#
# As this updates self._n we have to do it in the critical
# section, so do it byte by byte to avoid allocating.
_b = self._b
while nbytes > 0:
self._b[self._n] = self._b[self._w]
_b[self._n] = _b[self._w]
self._n += 1
self._w += 1
nbytes -= 1

self._w = len(self._b)
self._w = self._l
finally:
machine.enable_irq(ist)

Expand All @@ -866,10 +874,11 @@ def finish_read(self, nbytes):
assert nbytes <= self._n # can't say we read more than was available
i = 0
self._n -= nbytes
_b = self._b
while i < self._n:
# consumer only read part of the buffer, so shuffle remaining
# read data back towards index 0 to avoid fragmentation
self._b[i] = self._b[i + nbytes]
_b[i] = _b[i + nbytes]
i += 1
finally:
machine.enable_irq(ist)
Expand Down
Loading