On Wed, Mar 11, 2015 at 8:31 AM, Peter Hurley <peter@xxxxxxxxxxxxxxxxxx> wrote: > On 03/11/2015 10:29 AM, Tim Kryger wrote: >> On Wed, Mar 11, 2015 at 6:19 AM, Peter Hurley <peter@xxxxxxxxxxxxxxxxxx> wrote: >> In cases where it prints, the uart has likely become non-functional. > > Why? > > The initial port line settings come from the console line settings > and the h/w was already initialized when the console was registered, > so ISTM the worst that could happen there is UART_TX is written with > the value intended for UART_DLL. Additionally, UART_IER would be written with the value intended for UART_DLH should the attempt to write UART_LCR with the DLAB set fail. That is bad but if the attempt to write UART_LCR with DLAB clear fails, all subsequent writes intended for UART_TX will reach UART_DLL instead. The same is true of reads of UART_RX. > Feel free to submit an alternative. I liked your earlier proposal where the logging was only skipped if the port that failed was also the console. -- 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
