On Tue, Jan 11, 2011 at 18:12:25, Michael Williamson wrote: > [...] > This was pretty much what the original patch I had proposed did. There are > actually two termios flags that cause MS interrupts to be enabled by the > driver, CLOCAL (disabled) or CRTSCTS (enabled). I observed that some > process after the kernel had completed booting (getty, I think) > was calling set_termios with CLOCAL disabled (not CRTSCTS enabled). I could > probably alter the getty args to work around it, but it's a bit of a hole to > not have something in the kernel to preclude sending the OS to "interrupt > hell". > > Is there another way to accomplish what you are suggesting that is different > from the original patch? > > Thanks for the comments. > > -Mike > > Background: The original patch (not accepted) is here > https://patchwork.kernel.org/patch/442671/ > > This kicked a second email thread: > http://marc.info/?l=linux-serial&m=129409970003827&w=4 > > From that thread, it was suggested to add a port flag to the driver > flags structure for disabling MS interrupts. > > Driver Tweak patch (not accepted): > https://patchwork.kernel.org/patch/450771/ > > From that thread, it was suggested to override the I/O methods and hide > the problem back in the platform code (this patch). > > All three proposed patches work for me. I'd gladly submit a fourth if > suggested -- this problem pretty much kills a couple of platform > configurations we are working on... > > > Thanks, > > Prakash > > Thanks for clarifying. Looks like this is what we can do as of now. Thanks, Prakash -- 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
