Approximately 20% of devices are experiencing intermittent boot failures with this kernel version. The issue appears to be related to auto login failures, where an incorrect password is being detected on the serial console during the login process.
This intermittent regression is noticed on stable-rc 6.15.5-rc2 and Linux mainline master v6.16-rc5. This regressions is only specific to the devices not on the qemu's.
Test environments: - dragonboard-410c - dragonboard-845c - e850-96 - juno-r2 - rk3399-rock-pi-4b - x86
Regression Analysis: - New regression? Yes - Reproducibility? 20% only
Test regression: 6.15.5-rc2 v6.16-rc5 auto login failed Login incorrect
Reported-by: Linux Kernel Functional Testing lkft@linaro.org
## log in problem
runner-ns46nmmj-project-40964107-concurrent-0 login: # Password: Login incorrect runner-ns46nmmj-project-40964107-concurrent-0 login:
## Investigation The following three patches were reverted and the system was re-tested. The previously reported issues are no longer observed after applying the reverts.
serial: imx: Restore original RXTL for console to fix data loss commit f23c52aafb1675ab1d1f46914556d8e29cbbf7b3 upstream.
serial: core: restore of_node information in sysfs commit d36f0e9a0002f04f4d6dd9be908d58fe5bd3a279 upstream.
tty: serial: uartlite: register uart driver in init [ Upstream commit 6bd697b5fc39fd24e2aa418c7b7d14469f550a93 ]
Reference bug report lore link, - https://lore.kernel.org/stable/CA+G9fYvidpyHTQ179dAJ4TSdhthC-Mtjuks5iQjMf+ov...
## Test links * log 1: https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-6.15.y/build/v6.15.... * log 2: https://qa-reports.linaro.org/api/testruns/29021720/log_file/ * log 3: https://qa-reports.linaro.org/lkft/linux-mainline-master/build/v6.16-rc5/tes... * Boot test: https://regressions.linaro.org/lkft/linux-stable-rc-linux-6.15.y/v6.15.4-264... * LAVA jobs 1: https://lkft.validation.linaro.org/scheduler/job/8344153#L1186 * LAVA jobs 2: https://lkft.validation.linaro.org/scheduler/job/8343870#L1266
## Build * kernel: 6.15.5-rc2 * git: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git * git commit: f6977c36decb0875e78bdb8599749bce1e84c753 * git describe: v6.15.4-264-gf6977c36decb * test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-6.15.y/build/v6.15....
## Test Regressions (compared to v6.15.3-590-gd93bc5feded1) * dragonboard-410c, boot - clang-20-lkftconfig - clang-nightly-lkftconfig-kselftest - gcc-13-lkftconfig-debug
* dragonboard-845c, boot - clang-20-lkftconfig - korg-clang-20-lkftconfig-lto-thing
* dragonboard-845c-compat, boot - gcc-13-lkftconfig-compat
* e850-96, boot - gcc-13-lkftconfig-no-kselftest-frag
* juno-r2, boot - clang-20-lkftconfig - gcc-13-lkftconfig-debug - gcc-13-lkftconfig-kselftest
* rk3399-rock-pi-4b, boot - clang-20-lkftconfig
* x86, boot - clang-20-lkftconfig - clang-20-lkftconfig-no-kselftest-frag - clang-nightly-lkftconfig-kselftest - clang-nightly-lkftconfig-lto-thing - gcc-13-defconfig-preempt_rt - gcc-13-lkftconfig-no-kselftest-frag
-- Linaro LKFT https://lkft.linaro.org
On Wed, Jul 09, 2025 at 04:26:53PM +0530, Naresh Kamboju wrote:
Approximately 20% of devices are experiencing intermittent boot failures with this kernel version. The issue appears to be related to auto login failures, where an incorrect password is being detected on the serial console during the login process.
This intermittent regression is noticed on stable-rc 6.15.5-rc2 and Linux mainline master v6.16-rc5. This regressions is only specific to the devices not on the qemu's.
Test environments:
- dragonboard-410c
- dragonboard-845c
- e850-96
- juno-r2
- rk3399-rock-pi-4b
- x86
Regression Analysis:
- New regression? Yes
- Reproducibility? 20% only
Test regression: 6.15.5-rc2 v6.16-rc5 auto login failed Login incorrect
Reported-by: Linux Kernel Functional Testing lkft@linaro.org
## log in problem
runner-ns46nmmj-project-40964107-concurrent-0 login: # Password: Login incorrect runner-ns46nmmj-project-40964107-concurrent-0 login:
## Investigation The following three patches were reverted and the system was re-tested. The previously reported issues are no longer observed after applying the reverts.
serial: imx: Restore original RXTL for console to fix data loss commit f23c52aafb1675ab1d1f46914556d8e29cbbf7b3 upstream.
serial: core: restore of_node information in sysfs commit d36f0e9a0002f04f4d6dd9be908d58fe5bd3a279 upstream.
tty: serial: uartlite: register uart driver in init [ Upstream commit 6bd697b5fc39fd24e2aa418c7b7d14469f550a93 ]
As stated before, those are 3 totally independent changes. Any chance you can nail this down to just one of the above?
thanks,
greg k-h
+
On Wed, 9 Jul 2025 at 16:31, Greg Kroah-Hartman gregkh@linuxfoundation.org wrote:
On Wed, Jul 09, 2025 at 04:26:53PM +0530, Naresh Kamboju wrote:
Approximately 20% of devices are experiencing intermittent boot failures with this kernel version. The issue appears to be related to auto login failures, where an incorrect password is being detected on the serial console during the login process.
This intermittent regression is noticed on stable-rc 6.15.5-rc2 and Linux mainline master v6.16-rc5. This regressions is only specific to the devices not on the qemu's.
Test environments:
- dragonboard-410c
- dragonboard-845c
- e850-96
- juno-r2
- rk3399-rock-pi-4b
- x86
Regression Analysis:
- New regression? Yes
- Reproducibility? 20% only
Test regression: 6.15.5-rc2 v6.16-rc5 auto login failed Login incorrect
Reported-by: Linux Kernel Functional Testing lkft@linaro.org
## log in problem
runner-ns46nmmj-project-40964107-concurrent-0 login: # Password: Login incorrect runner-ns46nmmj-project-40964107-concurrent-0 login:
## Investigation The following three patches were reverted and the system was re-tested. The previously reported issues are no longer observed after applying the reverts.
serial: imx: Restore original RXTL for console to fix data loss commit f23c52aafb1675ab1d1f46914556d8e29cbbf7b3 upstream.
serial: core: restore of_node information in sysfs commit d36f0e9a0002f04f4d6dd9be908d58fe5bd3a279 upstream.
tty: serial: uartlite: register uart driver in init [ Upstream commit 6bd697b5fc39fd24e2aa418c7b7d14469f550a93 ]
As stated before, those are 3 totally independent changes. Any chance you can nail this down to just one of the above?
You're right, since this issue is intermittent, it's challenging to reproduce consistently. Pinpointing the exact commit would be ideal, but it will require more time.
thanks,
greg k-h
Thanks, Naresh
linux-stable-mirror@lists.linaro.org