Hi, IMHO PATCHv4 was better, but in the end I don't have strong feelings about this. Btw. I think this patchset is a good demonstration of frustrating upstream kernel development can be considering PATCHv5 is basically the same as PATCHv1. Thanks for making us go in circles :( Changes since PATCHv4: * https://lore.kernel.org/lkml/20210305115058.92284-1-sebastian.reichel@xxxxxxxxxxxxx/ * use DT property instead of sysfs config option, like the initial patch version did as requested by Greg. Changes since PATCHv3: * https://lore.kernel.org/lkml/1539249903-6316-1-git-send-email-fabien.lahoudere@xxxxxxxxxxxxx/ * rewrote commit message to provide a lot more details why this is needed * rebased to torvalds/master (5.12-rc1-dontuse), also applies on top of linux-next * use sysfs_emit() instead of sprintf -- Sebastian Fabien Lahoudere (2): serial: imx: Add DMA buffer configuration via DT ARM: dts: imx53-ppd: add dma-info nodes .../bindings/serial/fsl-imx-uart.yaml | 12 +++++++++ arch/arm/boot/dts/imx53-ppd.dts | 2 ++ drivers/tty/serial/imx.c | 25 +++++++++++++------ 3 files changed, 32 insertions(+), 7 deletions(-) -- 2.30.2
