Skip to contentSkip to conversation

serial console problems under 2.4.4/5

DISCUSSION Narayan Desai 2 participants 4 messages May 29, 2001

lore
Share & tools
lore
#0 lore raw

Hi. I have started having serial console problems in the last bunch of kernel releases. I have tried various 2.4.4 and 2.4.5 ac kernels (up to and including 2.4.5-ac4) and the problem has persisted. The problem is basically that serial console doesn't recieve. I know that the hardware works properly, and lilo and other kernels even work fine from serial console. As far as I can tell, only one direction is broken, serial receive. (ie, an echo foo > /dev/ttyS0 shows up on the serial console, but an echo to the serial device on the other end of the console never shows up)

The serial hardware is detected:
Serial driver version 5.05b (2001-05-03) with MANY_PORTS SHARE_IRQ SERIAL_PCI enabled ttyS00 at 0x03f8 (irq = 4) is a 16550A
ttyS01 at 0x02f8 (irq = 3) is a 16550A

the pertinant parts of .config are:
CONFIG_SERIAL=y
CONFIG_SERIAL_CONSOLE=y
# CONFIG_SERIAL_EXTENDED is not set
CONFIG_SERIAL_NONSTANDARD=y
# CONFIG_COMPUTONE is not set
CONFIG_ROCKETPORT=y
# CONFIG_CYCLADES is not set

Has anyone else seen this, or have any ideas what is broken? thanks...
-nld

#1 lore raw

In article <[email protected]>,

Quoted 4 lines · Narayan Desai

Hi. I have started having serial console problems in the last bunch of kernel releases. I have tried various 2.4.4 and 2.4.5 ac kernels (up to and including 2.4.5-ac4) and the problem has persisted. The problem is basically that serial console doesn't recieve.

The serial driver now pays attention to the CREAD bit. Sysvinit clears it, so that's where it goes wrong.

I don't think this change should have gone into a 'stable' kernel version. 2.5.0 would have been fine, not 2.4.4

Mike.

#2 lore raw
In reply to Narayan Desai
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Mike" == Miquel van Smoorenburg <[email protected]> writes:

Mike> In article <[email protected]>,

Quoted 4 lines · Mike> Narayan Desai <[email protected]> wrote:
Quoted 4 lines

Hi. I have started having serial console problems in the last bunch of kernel releases. I have tried various 2.4.4 and 2.4.5 ac kernels (up to and including 2.4.5-ac4) and the problem has persisted. The problem is basically that serial console doesn't recieve.

Mike> The serial driver now pays attention to the CREAD bit. Sysvinit Mike> clears it, so that's where it goes wrong.

Mike> I don't think this change should have gone into a 'stable' Mike> kernel version. 2.5.0 would have been fine, not 2.4.4

How would I go about resetting this so that serial console worked again? thanks...
-nld

#3 lore raw
In reply to Narayan Desai

In article <[email protected]>,

Quoted 9 lines · Narayan Desai
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Mike" == Miquel van Smoorenburg <[email protected]> writes:

Mike> In article <[email protected]>,

Quoted 4 lines · Mike> Narayan Desai <[email protected]> wrote:
Quoted 4 lines

Hi. I have started having serial console problems in the last bunch of kernel releases. I have tried various 2.4.4 and 2.4.5 ac kernels (up to and including 2.4.5-ac4) and the problem has persisted. The problem is basically that serial console doesn't recieve.

Quoted 5 lines

Mike> The serial driver now pays attention to the CREAD bit. Sysvinit Mike> clears it, so that's where it goes wrong.

Mike> I don't think this change should have gone into a 'stable' Mike> kernel version. 2.5.0 would have been fine, not 2.4.4

Okay it was 2.4.3 when it went in

Quoted 2 lines

How would I go about resetting this so that serial console worked again? thanks...

Fix sysvinit. Oh and all getty programs etc - everything that mucks around with termios.

Alternatively revert the change to drivers/char/serial.c

See also http://www.mail-archive.com/[email protected]/msg42426.html

Oh *ahem* I just read the entire thread again and I remembered it wrong. It's not the CREAD handling perse, it's something else that has changed that caused this, but it appears that a solution hasn't been found yet. Except for fixing all user-space programs. Sysvinit, agetty, busybox etc etc

Okay so ignore my earlier comments. Sorry.

Mike.