* [bluez/bluez] b93baa: monitor: Check valid range of CE Length
@ 2026-09-10 17:27 Luiz Augusto von Dentz
0 siblings, 0 replies; only message in thread
From: Luiz Augusto von Dentz @ 2026-09-10 17:27 UTC (permalink / raw)
To: linux-bluetooth
Branch: refs/heads/master
Home: https://github.com/bluez/bluez
Commit: b93baad72d27ec3fc5c42f80a6b21e9f3edfe014
https://github.com/bluez/bluez/commit/b93baad72d27ec3fc5c42f80a6b21e9f3edfe014
Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M monitor/packet.c
Log Message:
-----------
monitor: Check valid range of CE Length
The connection event length recommended in requests by a Peripheral has
a valid range of 0x0001 to 0x7CFF (Time = N * 125 us, Time Range:
0.125 ms to 3.999875 s), so print values outside of it as reserved
instead of decoding them as a time.
Commit: 9d35770cb64777b3a44b9deccc3683231486af0c
https://github.com/bluez/bluez/commit/9d35770cb64777b3a44b9deccc3683231486af0c
Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M emulator/btdev.c
Log Message:
-----------
emulator/btdev: Check valid range of CE Length
The connection event length recommended in requests by a Peripheral has
a valid range of 0x0001 to 0x7CFF (Time = N * 125 us, Time Range:
0.125 ms to 3.999875 s), so reject LE Connection Rate Request and LE Set
Default Rate Parameters with values outside of it.
Commit: 51da643c8271468013804e57b24a55ea4207410b
https://github.com/bluez/bluez/commit/51da643c8271468013804e57b24a55ea4207410b
Author: Zijun Hu <zijun.hu@oss.qualcomm.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M doc/hci-protocol.rst
Log Message:
-----------
doc/hci-protocol: Fix missing & on mtu argument in BT_RCVMTU example
getsockopt() requires a pointer argument, but mtu was passed by value.
Fix by using &mtu instead of mtu.
Commit: 4be273b132a72c5eaeeac9b9b9cf19c241b321d0
https://github.com/bluez/bluez/commit/4be273b132a72c5eaeeac9b9b9cf19c241b321d0
Author: Zijun Hu <zijun.hu@oss.qualcomm.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M monitor/broadcom.c
M monitor/broadcom.h
M monitor/intel.c
M monitor/intel.h
M monitor/msft.c
M monitor/msft.h
M monitor/packet.c
M monitor/vendor.c
M monitor/vendor.h
Log Message:
-----------
monitor: s/vendor_(evt|event)/\1_vendor/
Kernel include/net/bluetooth/hci.h defines:
#define HCI_EVENT_PKT 0x04
#define HCI_VENDOR_PKT 0xff
#define HCI_EV_VENDOR 0xff
HCI_VENDOR_PKT and HCI_EV_VENDOR share the same value, but are very
different, as the table below shows:
+------------+-----------------------+---------------------------+
| | HCI_VENDOR_PKT | HCI_EV_VENDOR |
+------------+-----------------------+---------------------------+
| Defined by | Vendor | BT SIG |
+------------+-----------------------+---------------------------+
| Hierarchy | Parallel to | Nested inside |
| | HCI_EVENT_PKT | HCI_EVENT_PKT |
+------------+-----------------------+---------------------------+
| Direction | RX and TX | RX only |
+------------+-----------------------+---------------------------+
| kernel <-> | HCI_VENDOR_PKT, ... | 0x04, HCI_EV_VENDOR, ... |
| userspace | | |
+------------+-----------------------+---------------------------+
| kernel <-> | RX: 23, vendor packet | RX: 3, HCI_EV_VENDOR, ... |
| monitor | TX: 22, vendor packet | |
+------------+-----------------------+---------------------------+
monitor is about to gain HCI_VENDOR_PKT support. To avoid confusing the
two, rename what deals with HCI_EV_VENDOR to follow its pattern:
s/vendor_(evt|event)/\1_vendor/
Commit: 160395d93d14d6f0e2db30c217a5c5d2ea6e49a4
https://github.com/bluez/bluez/commit/160395d93d14d6f0e2db30c217a5c5d2ea6e49a4
Author: Zijun Hu <zijun.hu@oss.qualcomm.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M doc/btsnoop-protocol.rst
M monitor/analyze.c
M monitor/packet.c
M monitor/packet.h
M src/shared/btsnoop.h
Log Message:
-----------
monitor: Support vendor packets
Kernel include/net/bluetooth/hci.h defines its packet type as:
#define HCI_VENDOR_PKT 0xff
- Defined by vendor
- Bidirectional (RX and TX)
- kernel <-> userspace: HCI_VENDOR_PKT followed by vendor packet
- kernel <-> monitor: vendor packet has TX opcode (22) or RX
opcode (23)
Support them in tool btmon.
Commit: 06638a3983fd26f7ed141304b48f50ead2c6972e
https://github.com/bluez/bluez/commit/06638a3983fd26f7ed141304b48f50ead2c6972e
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M client/btpclient/bap.c
M client/btpclient/bap.h
M client/btpclient/btpclient.c
M client/btpclient/btpclient.h
M src/shared/btp.h
Log Message:
-----------
client/btpclient: Add BTP_EV_BAP_ASE_FOUND support
This is used at least for BAP/UCL/SCC/BV-004-C and BAP/UCL/SCC/BV-019-C
tests.
Commit: 6679d89c7c6d3dd4c6b25c808bd58e1cc69f10fb
https://github.com/bluez/bluez/commit/6679d89c7c6d3dd4c6b25c808bd58e1cc69f10fb
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M Makefile.tools
M client/btpclient/bap.c
M src/shared/btp.h
Log Message:
-----------
client/btpclient: Get Codec capabilities on ASE reply
Commit: 1f4dbf07531e95ee83ca24b3ece72f1551b8966b
https://github.com/bluez/bluez/commit/1f4dbf07531e95ee83ca24b3ece72f1551b8966b
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M Makefile.tools
A client/btpclient/ascs.c
A client/btpclient/ascs.h
M client/btpclient/btpclient.c
M client/btpclient/btpclient.h
M client/btpclient/core.c
M client/btpclient/gap.c
A client/btpclient/vendor.c
A client/btpclient/vendor.h
M src/shared/btp.h
Log Message:
-----------
client/btpclient: Add ASCS BTP support
The audio endpoints are registered by ASCS setup command of VENDOR
service at start of the test.
For SelectProperties() D-Bus method, the reply is postponed after
auto-pts has requested the ASE and PACS characteritic values.
This allows to pass the BAP/UCL/SCC/* tests.
Commit: 0aa5118f4c8782c4604937e26ec063a262e34010
https://github.com/bluez/bluez/commit/0aa5118f4c8782c4604937e26ec063a262e34010
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M Makefile.tools
M client/btpclient/btpclient.h
M client/btpclient/core.c
A client/btpclient/pacs.c
A client/btpclient/pacs.h
M src/shared/btp.h
Log Message:
-----------
client/btpclient: Add BTP_OP_PACS_SET_LOCATION support
This allows to store the expected location for a local endpoint.
Commit: a0d138b16027d8597d134dab4edb344588db0f0d
https://github.com/bluez/bluez/commit/a0d138b16027d8597d134dab4edb344588db0f0d
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M client/btpclient/bap.c
M client/btpclient/btpclient.h
M src/shared/btp.h
Log Message:
-----------
client/btpclient: Add BTP_OP_BAP_SEND support
Commit: 2a0cee7a5af1072bf75239bc8c1fc65bb43033d3
https://github.com/bluez/bluez/commit/2a0cee7a5af1072bf75239bc8c1fc65bb43033d3
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M client/btpclient/ascs.c
M client/btpclient/ascs.h
M client/btpclient/bap.c
M client/btpclient/btpclient.c
M client/btpclient/btpclient.h
M client/btpclient/vendor.c
M src/shared/btp.h
Log Message:
-----------
client/btpclient: Add ASCS support for BAP/UCL/STR/* tests
Commit: b2d1281907992e7d294f445fba0c03399f3ecc96
https://github.com/bluez/bluez/commit/b2d1281907992e7d294f445fba0c03399f3ecc96
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M client/btpclient/ascs.c
M client/btpclient/bap.c
Log Message:
-----------
client/btpclient: CIG/CIS assignment based on Client/Server role
Previously, CIG/CIS ids were eagerly assigned to each ASE as soon as
it was discovered (CIG 0, next available CIS), regardless of whether
the btpclient was acting as the Unicast Client or Server for that ASE.
This did not match how CIG/CIS ids are actually negotiated, nor what
auto-pts expects: unicast client tests expect CIG/CIS to start at 0/0,
while unicast server tests expect them to start at 1/1.
- As Unicast Client, the CIG/CIS ids are chosen locally when building
the QoS properties returned from SelectProperties(), starting at 0.
- As Unicast Server, the CIG/CIS ids are only known once the transport
reports them; if not yet available, all of the device's ASEs should
share a common CIG (1) with CIS ids starting at 1 as well.
Move get_next_cis() from bap.c to ascs.c and extend it to take a base
CIS value, so it can be reused both for the Client and Server code
paths (base 0 and base 1 respectively) and skips ASEs whose CIS id is
still unset. Leave ase->cig_id and ase->cis_id unset
(BT_ISO_QOS_CIG_UNSET/BT_ISO_QOS_CIS_UNSET) at ASE discovery time, and
resolve them lazily:
- In get_properties_reply(), assign CIG 0 and the next CIS starting at
0 when building the QoS reply for SelectProperties() (Client role).
- In ascs_proxy_added(), when the transport does not yet report a
CIG/CIS, assign CIG 1 to all of the device's ASEs and derive CIS ids
starting at 1 via the new set_cig_cis()/set_cig_cis_data helpers
(Server role).
Commit: 6cf877bf5a61bcb7cd851bc5a98e5a4361608970
https://github.com/bluez/bluez/commit/6cf877bf5a61bcb7cd851bc5a98e5a4361608970
Author: Frédéric Danis <frederic.danis@collabora.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M client/btpclient/ascs.c
M client/btpclient/bap.c
M client/btpclient/bap.h
M client/btpclient/btpclient.c
M client/btpclient/btpclient.h
Log Message:
-----------
client/btpclient: Add stream auto acquire for BAP/USR/STR/* tests
The sink stream should be auto acquired and data sent on it.
Commit: ed279b3516ee06f39c4a16b7c02a101de1ec61a6
https://github.com/bluez/bluez/commit/ed279b3516ee06f39c4a16b7c02a101de1ec61a6
Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M Makefile.am
M test/functional/requirements.txt
Log Message:
-----------
build: run the test suite in parallel
Both the unit tests and the functional tests were running serially,
which makes "make check" unnecessarily slow.
The automake test harness spawns the individual tests from a sub-make,
so adding -j to AM_MAKEFLAGS for the check-TESTS/recheck targets is
enough to parallelize the unit tests.
The functional tests are parallelized with pytest-xdist. The loadgroup
scheduler is required since pytest-bluezenv marks the tests sharing a
host/VM setup with an xdist_group, so that they are kept on the same
worker instead of each worker re-creating the same setup.
The number of concurrent tests defaults to nproc and can be overridden
with CHECK_JOBS, e.g. "make CHECK_JOBS=1 check" for a serial run.
Assisted-by: Claude:claude-opus-5
Commit: 51d1b7924e5580b724e8cbbfe31b5633d381ae65
https://github.com/bluez/bluez/commit/51d1b7924e5580b724e8cbbfe31b5633d381ae65
Author: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Date: 2026-09-10 (Thu, 10 Sep 2026)
Changed paths:
M .github/workflows/sync.yml
Log Message:
-----------
workflows: allow the sync workflow to be triggered manually
The scheduled trigger is heavily throttled by Github and the run can be
delayed for hours, which makes it hard to verify a change or to recover
from a failed run. Add workflow_dispatch so the workflow can be started
on demand.
Compare: https://github.com/bluez/bluez/compare/19ebd6a2124d...51d1b7924e55
To unsubscribe from these emails, change your notification settings at https://github.com/bluez/bluez/settings/notifications
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2026-09-10 17:27 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-09-10 17:27 [bluez/bluez] b93baa: monitor: Check valid range of CE Length Luiz Augusto von Dentz
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox