Align the Easymesh R1 smoke test with the DUT flashing procedure from FEAT-29.
This commit aligns the DUT deploy scripts from the prplMesh repository.
Reference: PPM-4139
Signed-off-by: Maarten De Decker <maarten.dedecker@mind.be>
During the course of the test, containers are created and tested to properly survive a simulated software upgrade.
However, pcm does not guarantee that indexes will be the same after SWU, especially if there were any gaps
in numbering prior to the upgrade.
This change to the test makes it index-agnostic by replacing all possibly affected indexes with placeholders.
References: PPW-1787
Closes: PPW-1787
Signed-off-by: a.kuznetsov <a.kuznetsov@inango-systems.com>
Currently the final DNS host check after configuration restore
accepts only one digit in the generated alias. The failing
testbed-02 job reaches instance 10, restores a valid object, and
reports this Cram difference:
- "Alias": "cpe-X_PRPLWARE-COM_Host-\d", (re)
+ "Alias": "cpe-X_PRPLWARE-COM_Host-10",
The comparable testbed-01 control job reaches instance 9 and passes
the same check. Both jobs successfully restore the DNS object, so
the failure depends on the allocator crossing the single-digit
boundary, not on invalid restored state.
The additional testbed-02 topology from PCF-2504 is the supported
trigger for the one-instance delta. It exposes the restrictive
expectation but does not make the restored alias invalid.
So lets fix the final alias expectation to accept one or more
digits, matching the equivalent check earlier in the test. This
keeps the test independent of the current multi-instance allocator
position.
Closes: PPW-1934
References: PCF-2504
Link: https://gitlab.com/prpl-foundation/prplos/prplos/-/jobs/15294595185
Link: https://gitlab.com/prpl-foundation/prplos/prplos/-/jobs/15300965299
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
Several parameters for Download/Activate RPC are defined but are not
implemented yet, so setting them will not take any action.
Update package documentation to explicitly state it
Implements: PCF-2340
References: FEAT-356, FEAT-357, FEAT-358
Signed-off-by: Mykhailo Polnyi <mykhailo.polnyi@globallogic.com>
Currently the BulkData MQTT cram test assumes that every DUT uses the
legacy SN[A-F0-9]+ serial-number format. SFL devices instead expose the
manufacturing serial through readmfg, including values such as
ospv2_A8C2463D425C and freedom_58E403D21127.
This makes a valid report fail the complete-payload grep even though the
reported SN is exactly DeviceInfo.SerialNumber. Both OSPv2 and Freedom
jobs show the following assertion failure:
$ R 'cat /tmp/mqtt_subscriber' | \
> grep -cP '^\{"Report":\[\{"SN":"SN[A-F0-9]+","CollectionTime":"[0-9]{13}"\}\]\}$'
-[1-9][0-9]* (re)
+0
+[1]
So lets fix the test by reading DeviceInfo.SerialNumber from the DUT and
passing that value to jq. Require exactly one report and compare its SN
with the live serial. Validate CollectionTime independently as a string
of exactly 13 digits. Print the payload when validation fails so future
cram diffs preserve the malformed message.
Closes: PPW-2102
References: PCF-2643
Assisted-by: codex:gpt-5.6-sol:high
Signed-off-by: Petr Štetiar <petr.stetiar@prplfoundation.org>
Problem:
network_security.t fails deterministically on latest-24.10. The
service-readiness block added by !2697 (adc82ee3, PPW-887) starts
with "$ R '" without the two-space indent cram requires for a
command line.
Root cause:
cram does not parse the malformed line as a command, so it is
absorbed as expected output of the preceding nmap check. Its
continuation lines then attach to that check and run in the local
testbed shell instead of on the DUT via the R alias:
/bin/sh: 6: pgrep: not found
/bin/sh: 14: Syntax error: Unterminated quoted string (no-eol)
adc82ee3 was reverted (72f0994e) once DUT jobs were finally
triggered and exposed the failure; no DUT cram job ran against any
revision of !2697 before it merged.
Solution:
Re-add the service-readiness wait, indenting "$ R '" by two spaces
so cram parses it as a command and dispatches the wait script to the
DUT through the R alias, matching every other "$ R '...'" usage in
this test suite.
Reference: PPW-2150, PPW-887, PPW-455
Signed-off-by: Alexandre Fiset <alexandre.fiset@softathome.com>
When finishing the pwhm-wpa3-cm.t test script, MLDUnits are currently
restored to the same values as in Freedom (0: Private MLD, 1: Guest
MLD).
However, MLO is not yet supported on OSPv2, so MLDUnits should instead
remain at their default value (-1, as set through MXL UPDK).
Closes: PPW-1713
References: PCF-2395
Signed-off-by: Houssem Dafdouf <houssem.dafdouf_ext@softathome.com>
After re-enabling the wifi-scheduler script, update the test to:
- Add backhaul interfaces
- Use the correct prplMesh AccessPoint creation method
- Use the generic wifi_dm helper
- Adapt AP status checks to keep the script generic
References: PPW-774
Signed-off-by: Houssem Dafdouf <houssem.dafdouf_ext@softathome.com>
The test was hardcoded to expect wan and br-lan as interface names,
but the actual kernel/netdev interface name that showed up in the mangle table was eth1.
Instead of resolving the interface name at test time (hardcoded),
we resolve it at test run time as a dynamic resolution by
walking through the data model to extract the real netdev name whatever the project (ospv2/freedom)
the test now emit OK if this check passes or FAIL with explicit message if the check fails.
Tracked-by: PPW-1998
Signed-off-by: Yousfi Neji <yousfi.neji_ext@softathome.com>
Integrates:
- d12d632a: ("fix reload-time crash in AMX bus and interface handling", 2026-06-05)
References: PPW-2041
Signed-off-by: ahmed sassi <ahmed.sassi_ext@softathome.com>
When removing the last BSS of an AP MLD, the associated MLD
control interface may be left unused, leaving its control
socket lingering.
Invoke hostapd_cleanup_unused_mlds() after BSS removal to
clean up unused MLD objects. This ensures that stale MLD
state and control sockets are properly release
Fixes: PCF-2605
Signed-off-by: Marvin Lu <marvin.lu@wnc.com.tw>
Remove the deprecated Device.DeviceInfo.Reboots.reboot() RPC and its input
parameters (Cause, Reason) from the obuspa.expected fixture.
This change aligns the test expectations with the current reboot-service
model where the reboot() RPC is no longer exposed.
References: PPW-1960
Signed-off-by: ahmed sassi <ahmed.sassi_ext@softathome.com>
Integrates following changes:
reboot-service:
- reboot-service: Reboots.reboot() should not be part of the GSDM (26b126f7)
tr181-device:
- tr181-device: Reboots.reboot() is now protected (29c3f3c9)
References: PPW-1960
Signed-off-by: prplbot <bot@prplfoundation.org>
uboot-tools (pylibfdt) failed to compile with:
fatal error: Python.h: No such file or directory
Add python3-dev to the apt-get install step to provide the
missing header.
Reference: PPW-2121
Signed-off-by: Mouna Aloui <mouna.aloui@softathome.com>
Integrates:
-Update the CRAM test expectations to use the correct parameter name Stats.NACK instead of the misspelled Stats.NAK
References: PPW-1952
Signed-off-by: FAOUZI YAHYA <faouzi.yahya_ext@softathome.com>
Integrates:
-Update the obuspa expected fixture to add:
Device.SoftwareModules.ExecutionUnit.{i}.EnvVariableNumberOfEntries
Device.SoftwareModules.ExecutionUnit.{i}.HostObjectNumberOfEntries
References: PPW-1954
Signed-off-by: Antoine Philippe <antoine.philippe@softathome.com>
Integrates:
-Update the obuspa expected fixture to include
Device.SoftwareModules.ExecEnv.{i}.ApplicationDataNumberOfEntries
for timingila.
References : PPW-1954
Signed-off-by: ahmed sassi <ahmed.sassi_ext@softathome.com>
Context:
========
A first attempt to reorganize the pwhm usp tests for PPW-1606 was
introduced in commit 7560c16e81.
However, CI was failing because some required alias definitions were
missing from the OSPv2 cram tests. The change was reverted in commit
c7ea8021e0.
Let's rework the change and include the missing alias definitions so
the test reorganization is handled properly.
Initial problem:
================
During the integration of PPM-3671, the 103-apmld.t script started
failing due to an incorrect security mode set by previous tests.
Cause:
======
The security mode was unexpectedly modified during the execution of
102-pwhm-usp.t while toggling AP1. This happens because prplMesh now
runs in controller mode as part of PPM-3671.
Solution:
=========
When testing pure PWHM features, prplMesh should not be running to
avoid conflicts. This was previously the case, except for the AP1
toggling step executed in wifi-usp.t.
Move the USP-related test cases to more appropriate locations:
- USP socket creation and PWHM USP event checks (AP1 toggling):
moved to 100-pwhm.t where prplMesh is kept stopped
- prplMesh connection to the PWHM direct socket:
moved to 110-prplmesh-dataelements.t
- 102-pwhm-usp.t is so removed from fast track test scripts
Closes: PPW-1606
References: PPM-3671
Signed-off-by: Houssem Dafdouf <houssem.dafdouf_ext@softathome.com>
Move AssociatedDeviceNumberOfEntries to match the current HomePlug data model output ordering in obuspa.expected.
Refrences: PPW-1953
Signed-off-by: ahmed sassi <ahmed.sassi_ext@softathome.com>
Updating the test 005-build-config during PCF-2624.
Update the test to add CONFIG_FWI_FLASH_PARTITIONS="u-boot,kernel,rootfs"
References: PCF-2624, PPW-2154
Signed-off-by: Hela Saadani <hela.saadani_ext@softathome.com>
New packages ivykis and liburing appear in the expected build config
because upstream openwrt/packages commit 8fcf45d05a changed syslog-ng
to depend on an external ivykis package (instead of the previously
embedded copy). ivykis in turn pulls in liburing when io_uring support
is enabled in the kernel.
References: PCF-1982
Signed-off-by: Hela Saadani <hela.saadani_ext@softathome.com>
