On Fri, May 8, 2026 at 8:53 AM Max Merchel <[email protected]> wrote:
>
> Adds support for TQMa6UL[L]x. It has 256M and 512M RAM variants which
> are detected by the SPL by initializing them in descending order.
> It can boot from SD, MMC, SPI and USB (SDP). SPI however requires an
> additional prepended NXP header image which is currently unsupported
> in u-boot.
>
> Pipeline: 
> https://dev.azure.com/u-boot/u-boot/_build/results?buildId=13172&view=results

Interesting that it passed Azure's pipeline. It failed on Gitlab's CI:

https://source.denx.de/u-boot/custodians/u-boot-imx/-/jobs/1451666

The trick is that the error is not explicit. Tom helped me to debug
this error offline.

>From Tom:

"And so in another terminal:
$ ps uxwww | grep O=
trini     834394  0.0  0.0   3068  2176 pts/1    S+   14:07   0:00
make O=/tmp/tmptj4mqomn KCONFIG_IGNORE_DUPLICATES=1
include/config/auto.conf
trini    1069113  0.0  0.0   3068  2176 pts/1    S+   14:08   0:00
make O=/tmp/tmpqg2bkydk KCONFIG_IGNORE_DUPLICATES=1
include/config/auto.conf
trini    1481999  0.0  0.0   3068  2176 pts/1    S+   14:10   0:00
make O=/tmp/tmpwgav3dki KCONFIG_IGNORE_DUPLICATES=1
include/config/auto.conf
trini    1879754  0.0  0.0   3068  2176 pts/1    S+   14:11   0:00
make O=/tmp/tmp7xt1epoi KCONFIG_IGNORE_DUPLICATES=1
include/config/auto.conf
trini    2190009  0.0  0.0   9148  2176 pts/3    S+   14:18   0:00
grep --color=auto O=
And so:
$ grep SYS_CONFIG_NAME /tmp/tmptj4mqomn/.config
/tmp/tmpqg2bkydk/.config /tmp/tmpwgav3dki/.config
/tmp/tmp7xt1epoi/.config
/tmp/tmptj4mqomn/.config:CONFIG_SYS_CONFIG_NAME="tqma6ul_mba6ul"
/tmp/tmpqg2bkydk/.config:CONFIG_SYS_CONFIG_NAME="tqma6ul_mba6ul"
/tmp/tmpwgav3dki/.config:CONFIG_SYS_CONFIG_NAME="tqma6ul_mba6ul"
/tmp/tmp7xt1epoi/.config:CONFIG_SYS_CONFIG_NAME="tqma6ul_mba6ul"
And that is set in board/tq/tqma6ul/Kconfig and so that leads me to:
configs/tqma6ulx_mba6ul_qspi_defconfig
configs/tqma6ulxl_mba6ul_qspi_defconfig
configs/tqma6ullxl_mba6ul_qspi_defconfig
configs/tqma6ullx_mba6ul_qspi_defconfig

And I see at least the first two don't define ENV_SECT_SIZE which is
probably a migration error on their part. Which also means these weren't
build tested by them before posting."

Reply via email to