Messages in this thread | | | From | "Hans-Peter Jansen" <> | Subject | Re: 2.6.25.3: serial problem (minicom) | Date | Wed, 18 Jun 2008 02:15:45 +0200 |
| |
Am Dienstag, 17. Juni 2008 schrieb Alan Cox: > On Mon, 16 Jun 2008 23:22:54 -0500 (CDT) > > "R.L. Horn" <lists@eastcheap.org> wrote: > > On Mon, 16 Jun 2008, Alan Cox wrote: > > > Whatever it implies the behaviour should not have changed between > > > 2.6.24 and 2.6.25. Nobody AFAIK sat down and decided to change it. > > > > And, besides, I've gotten reports that the usb-serial drivers still > > behave the same as with 2.6.24. > > > > It looks like the call to tty_termios_encode_baud_rate() in > > drivers/serial/8250.c is the culprit. If I comment it out, everything > > appears to go back to normal (seemingly with no undesired side > > effects). > > > > Why the call is there (it didn't replace anything else in the 2.6.24.7 > > version of 8250.c, though it did in serial_core.c) remains a mystery to > > me. > > Ah ok I know what the bug is - it was fixed in 2.6.26-rc as follows > > + /* Don't rewrite B0 */ > + if (tty_termios_baud_rate(termios)) > + tty_termios_encode_baud_rate(termios, baud, baud); > }
Alan, could this issue lead to dysfunctional serial dcf77 receivers, too? (and could you point my to the related git changeset?)
I'm using ntpd "127.127.8.0 mode 16" devices since a decade now (RAWDCF receiver: DTR=low/RTS=high) and after upgrading to openSUSE 11.0, which is using 2.6.25.5, those get no power anymore :-(.
Or do you have another idea? I currently prepare a test system, and would be ready for kernel patching tomorrow..
TIA, Pete
| |