29.10.2019 03:14, Dmitry Osipenko пишет: > 28.10.2019 17:27, Peter De Schrijver пишет: >> On Tue, Jul 23, 2019 at 05:52:44AM +0300, Dmitry Osipenko wrote: >>> Unset "enable" bit means that divider is in bypass mode, hence it doesn't >>> have any effect in that case. Please note that there are no known bugs >>> caused by the missing check. >>> >> >> Technically this is not quite true, but for the purposes of CCF you can >> treat it that way. This bits defines if the value in the lower 16 bits >> of the divider register is used to configure the divider or if the >> contents of the UART DLM/DLL registers is used. So the divider isn't >> actually bypassed, it's just configured differently. >> In practice this bit is only set when the divider is non-zero when doing >> set rate. So the extra test isn't strictly needed as long as the sw >> running before the kernel also ensures the bit is only set when the >> divider is non-zero. >> >> Acked-by: Peter De Schrijver <pdeschrijver@xxxxxxxxxx> > > Thank you for the clarification. I hope that bootloader doesn't enable > the divider because it looks like standard 8250 driver won't be ready > for that. But serial-tegra driver seems should be good. Actually, it should be good because I missed that UART clocks are per-initialized in the clk driver init table. I'll update commit's message and send a new version of this patch.
