Hi, (adding Hans in Cc) On 20/07/2015 at 15:05:39 -0400, Peter Hurley wrote : > Hi Alexandre, > > On 07/20/2015 09:56 AM, Alexandre Belloni wrote: > > On ARM, it is necessary to use the _relaxed versions of read[bl] and > > write[bl] to ensure we do the correct endianness conversion when running in > > big endian. > > Keep using avr32 __raw version on avr32 to avoid breakage. > > > > Signed-off-by: Alexandre Belloni <alexandre.belloni@xxxxxxxxxxxxxxxxxx> > > --- > > > > This patch applies after [PATCH v4 0/5] tty/serial: at91: add support to FIFOs > > http://lists.infradead.org/pipermail/linux-arm-kernel/2015-July/353704.html > > > > drivers/tty/serial/atmel_serial.c | 22 ++++++++++++++++++++++ > > 1 file changed, 22 insertions(+) > > > > diff --git a/drivers/tty/serial/atmel_serial.c b/drivers/tty/serial/atmel_serial.c > > index e91b3b2f0590..451acf054281 100644 > > --- a/drivers/tty/serial/atmel_serial.c > > +++ b/drivers/tty/serial/atmel_serial.c > > @@ -194,6 +194,7 @@ to_atmel_uart_port(struct uart_port *uart) > > return container_of(uart, struct atmel_uart_port, uart); > > } > > > > +#ifdef CONFIG_AVR32 > > The whole point of i/o accessors is to encapsulate arch-specific > differences, so one of these arches isn't doing it right, or the wrong > i/o accessors are being used. > I'm not an avr32 specialist but I think __raw accessors are doing big endian accesses but _relaxed accessors are doing litlle endian accesses. Maybe Hans can clarify what's needed here. -- Alexandre Belloni, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
