[bluez/bluez] ae20c4: monitor: Add L2CAP channel details to analyze output

[bluez/bluez] ae20c4: monitor: Add L2CAP channel details to analyze output

[Luiz Augusto von Dentz]

  Branch: refs/heads/1081747
  Home:   https://github.com/bluez/bluez
  Commit: ae20c44933a860640ee6fda24b47c6137a082a42
      https://github.com/bluez/bluez/commit/ae20c44933a860640ee6fda24b47c6137a082a42
  Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
  Date:   2026-04-15 (Wed, 15 Apr 2026)

  Changed paths:
    M monitor/analyze.c

  Log Message:
  -----------
  monitor: Add L2CAP channel details to analyze output

Enhance btmon analyze mode to display richer per-channel information:

- Add MTU, MPS, and mode fields to struct l2cap_chan
- Parse Configure Request/Response options for BR/EDR channels to
  extract MTU (option 0x01) and mode (option 0x04)
- Add l2cap_le_sig() to parse LE signaling (CID 5) for LE Credit
  Based Connection Request/Response and Enhanced Credit Connection
  Request, extracting PSM, MTU, MPS, and mode
- Display fixed channel names (ATT, L2CAP Signaling, SMP) for CID <= 7
- Show PSM in both decimal and hex format
- Print mode, MTU, and MPS for dynamic channels


  Commit: 846c8a570c4c0ac12377e85c4b60379778b847f0
      https://github.com/bluez/bluez/commit/846c8a570c4c0ac12377e85c4b60379778b847f0
  Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
  Date:   2026-04-15 (Wed, 15 Apr 2026)

  Changed paths:
    M monitor/analyze.c

  Log Message:
  -----------
  monitor: Use wall-clock duration for throughput and add min/max speed

The previous speed calculation used the sum of per-packet latencies as
the denominator.  For TX, each latency measures the time from command
submission to completion event -- when multiple packets are in-flight
simultaneously these overlap, making the sum much larger than wall-clock
time and deflating the reported speed.  For RX the inter-arrival sum
roughly equals wall-clock time.  This asymmetry caused the same data
stream to report vastly different speeds depending on direction.

Fix by computing speed from wall-clock duration (last packet timestamp
minus first packet timestamp), which gives consistent results regardless
of direction.  Also add 1-second windowed throughput tracking to report
min and max speed alongside the average.


  Commit: 76d5b3aec064b8bba40ac04a8618f16ca5794f16
      https://github.com/bluez/bluez/commit/76d5b3aec064b8bba40ac04a8618f16ca5794f16
  Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
  Date:   2026-04-15 (Wed, 15 Apr 2026)

  Changed paths:
    M doc/btmon-l2cap.rst

  Log Message:
  -----------
  doc: Add throughput estimation section to btmon-l2cap.rst

Document the throughput statistics reported by btmon analyze mode,
including speed calculation methodology, channel details (PSM, mode,
MTU, MPS), and caveats about inter-packet latency vs wall-clock time.
Also document the windowed throughput approach used by btsnoop-analyzer.


Compare: https://github.com/bluez/bluez/compare/ae20c44933a8%5E...76d5b3aec064

To unsubscribe from these emails, change your notification settings at https://github.com/bluez/bluez/settings/notifications