Hi Greg, > Subject: [PATCH 2/3] tty: serial: lpuart: support UPIO_MEM32 for lpuart32 This patch was sent out wrongly. Not affect the other three patches which belongs the whole patch set described in cover-letter. Should I sent out v2? Or you are fine if the other 3 patches are ok? Thanks, Peng. > > From: Peng Fan <peng.fan@xxxxxxx> > > "earlycon" no need to specify the value string since it uses stdout-path > parameters. However when earlycon and normal console are not using the > same uart port, we need specify value string to earlycon, this is what we need > to do when support dual linux using jailhouse hypervisor. The 2nd linux will > use the uart of the 1st linux as earlycon. > > earlycon=lpuart32,mmio32,0x5a060010,115200 not work for i.MX8QXP. > It is because lpuart32_early_console_setup not support little endian. > > Since the original code is to support UPIO_MEM32BE, so if not UPIO_MEM32, > we still take it as UPIO_MEM32BE > > Acked-by: Fugang Duan <fugang.duan@xxxxxxx> > Signed-off-by: Peng Fan <peng.fan@xxxxxxx> > --- > drivers/tty/serial/fsl_lpuart.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/drivers/tty/serial/fsl_lpuart.c b/drivers/tty/serial/fsl_lpuart.c index > 54c7735406f8..6ad0af073e62 100644 > --- a/drivers/tty/serial/fsl_lpuart.c > +++ b/drivers/tty/serial/fsl_lpuart.c > @@ -2376,7 +2376,9 @@ static int __init > lpuart32_early_console_setup(struct earlycon_device *device, > if (!device->port.membase) > return -ENODEV; > > - device->port.iotype = UPIO_MEM32BE; > + if (device->port.iotype != UPIO_MEM32) > + device->port.iotype = UPIO_MEM32BE; > + > device->con->write = lpuart32_early_write; > return 0; > } > -- > 2.16.4
