On Fri, Oct 29, 2021 at 12:56 PM Adam Ford <aford...@gmail.com> wrote: > > On Fri, Oct 22, 2021 at 6:16 PM Marcel Ziswiler <mar...@ziswiler.com> wrote: > > > > From: Marcel Ziswiler <marcel.ziswi...@toradex.com> > > > > > > With the move to using binman to generate SPL aka u-boot-spl-ddr.bin and > > U-Boot proper aka u-boot.itb every board now covers such configuration > > in its own U-Boot specific device tree include. Move the comon part of > > that configuration to the common imx8mm-u-boot.dtsi include file. > > > > The initial patch fixes an issue with intermediate binary naming for the > > imx8mm-cl-iot-gate. And subsequent patches further clean up that dtsi. > > > > This series is based on my refresh of Adam's imx8mm_beacon switch to > > binman patch [1], my refresh of Michael's sl28 patches [2] and Tim's > > switching imx8mm_venice to using binman to pack images [3]. > > Thanks for doing the rebase. I wonder if we should drop the > imx8mm-beacon patch and just add the adjusted defconfig changes rather > than applying either my patch or the rebased patch, then undoing it > right away with this series. I like the idea of merging them all > together into one nice series using the common imx8mm-u-boot.dtsi > file, but I wonder if we should split this into another sub-file > called im8mm-lpddr4-u-boot.dtsi. If someone uses a different DDR, > these referenced files won't be appropriate for their DDR type. > > I tried applying this series on u-boot/master, but it didn't apply > correctly, so the rest of my message might not apply if there is a > patch somewhere that fixes it, and I just missed it. > > 1. The build fails because there is a file in a location different > than what's expected: > > make flash.bin CROSS_COMPILE=aarch64-linux-gnu- -j8 > <snip> > BINMAN flash.bin > MKIMAGE flash.bin > ./tools/mkimage: Can't open spl/u-boot-spl-ddr.bin: No such file or directory > make[1]: *** [arch/arm/mach-imx/Makefile:167: flash.bin] Error 1 > make[1]: *** Deleting file 'flash.bin' > make: *** [Makefile:1516: flash.bin] Error 2 > > If I copy u-boot-spl-ddr.bin from the main build directory to the spl > directory and run make again, it completes successfully. > > 2. I noticed that the offset doesn't appear to be jumping to the > right location. I am not sure if something has changed in binman, but > I intentionally loaded this over a much older flash.bin file because I > wanted to make sure the correct version of u-boot loaded, but it did > not. > > Notice the version mismatch between SPL and U-Boot. > > U-Boot SPL 2021.10-00983-geed0aab6ff-dirty (Oct 29 2021 - 14:24:10 -0500) > WDT: Started watchdog@30280000 with servicing (60s timeout) > Trying to boot from MMC1 > NOTICE: BL31: v2.2(release):imx_5.4.24_er3-0-gb0a00f22b-dirty > NOTICE: BL31: Built : 03:05:00, May 12 2021 > > > U-Boot 2019.04-4.19.35-1.1.0+g4d377539a1 (Jun 03 2021 - 19:41:40 +0000) > > CPU: Freescale i.MX8MMQ rev1.0 1600 MHz (running at 500 MHz) > CPU: Industrial temperature grade (-40C to 105C) at 56C > Reset cause: POR > Model: Beacon EmbeddedWorks i.MX8M Mini Development Kit > DRAM: 2 GiB > MMC: FSL_SDHC: 0, FSL_SDHC: 1, FSL_SDHC: 2 > Loading Environment from MMC... OK > In: serial > Out: serial > Err: serial > > I can try to determine the proper offset again, but there is another > binman patch that might make this easier if the elf properly has the > offset defined [4]. > > > > > > [1] https://marc.info/?l=u-boot&m=163493902915471 > > [2] https://marc.info/?l=u-boot&m=163494122816224 > > [3] https://marc.info/?l=u-boot&m=163355140309636 > > [4] - > https://patchwork.ozlabs.org/project/uboot/patch/20211026033058.430010-6-...@chromium.org/ > > > If there is a patch I missed somewhere, please let me know, and I will > go back and try again. I'll try to figure out what's going on with the > u-boot offset sometime this weekend. > > > > > Changes in v3: > > - Integrated imx8mm_beacon as well. > > - Re-based on top of imx/master. > > > > Changes in v2: > > - New patch preparing cl-iot-gate. > > - New patch preparing cl-iot-gate. > > - Morph common binman dtsi into regular common u-boot dtsi as discussed > > on mailing list. > > - Re-based. > > - Rather than doing that clean-up to the separate binman dtsi now do it > > to the common u-boot dtsi one. > > - Rather than doing that clean-up to the separate binman dtsi now do it > > to the common u-boot dtsi one. > > - Rather than doing that clean-up to the separate binman dtsi now do it > > to the common u-boot dtsi one. > > > > Changes in v1: > > - This was suggested by Simon on my earlier patch set upon which we > > decided to first break this up into a common binman dtsi and then > > subsequently do such further improvements. > > > > Marcel Ziswiler (7): > > imx8mm-cl-iot-gate: fix imximage intermediate binary naming > > arm64: dts: imx8mm-cl-iot-gate-u-boot.dtsi: alphabetically re-order > > arm64: dts: imx8mm-cl-iot-gate-u-boot.dtsi: use common > > imx8mm-u-boot.dtsi > > arm64: dts: imx8mm: use common binman configuration > > arm64: dts: imx8mm-u-boot.dtsi: alphabetically re-order properties > > arm64: dts: imx8mm-u-boot.dtsi: explicitly add spl filename > > arm64: dts: imx8mm-u-boot.dtsi: improve odd blob-ext naming > > > > arch/arm/dts/imx8mm-beacon-kit-u-boot.dtsi | 122 ---------- > > .../dts/imx8mm-cl-iot-gate-optee-u-boot.dtsi | 209 ++++-------------- > > arch/arm/dts/imx8mm-cl-iot-gate-u-boot.dtsi | 188 +++------------- > > arch/arm/dts/imx8mm-evk-u-boot.dtsi | 123 ----------- > > arch/arm/dts/imx8mm-u-boot.dtsi | 149 ++++++++++++- > > arch/arm/dts/imx8mm-venice-u-boot.dtsi | 119 ---------- > > arch/arm/dts/imx8mm-verdin-u-boot.dtsi | 122 +--------- > > .../imximage-8mm-lpddr4.cfg | 2 +- > > 8 files changed, 212 insertions(+), 822 deletions(-) > > > > -- > > 2.26.2 > >
I'm very unclear where we stand right now as well as far as imx goes. As far as I understand the merge window for 2022.01 is now closed yet several of us still have imx patches sitting out there that likely got stuck because of merge difficulties. I have not gotten a response from Stefano how he wants us to deal with these patches. I was hoping we would all be switched over to using binman for 2022.01. Best regards, Tim