What do I have to do to get my serial ports back?
devfs: v1.22 (20021013) Richard Gooch ([email protected])
devfs: boot_options: 0x1
pnp: the driver 'serial' has been registered
pnp: pnp: match found with the PnP device '00:0f' and the driver 'serial'
pnp: the device '00:0f' has been activated
devfs_register(tts/2): could not append to parent, err: -17
tts/2 at I/O 0x3e8 (irq = 4) is a 16550A
pnp: pnp: match found with the PnP device '00:10' and the driver 'serial'
pnp: the device '00:10' has been activated
devfs_register(tts/3): could not append to parent, err: -17
tts/3 at I/O 0x2e8 (irq = 3) is a 16550A
Effectively as my mouse is on the serial port and it
obviously does not work in gpm or with X directly.
ttyS0: LSR safety check engaged!
ttyS0: LSR safety check engaged!
ttyS0: LSR safety check engaged!
ttyS1: LSR safety check engaged!
ttyS1: LSR safety check engaged!
ttyS1: LSR safety check engaged!
That's coming from gpm startup, at least one of it, although
I have /dev/tts/0 in my gpm.conf... Ah, it's hardcoded in the
binary, it seems (doing a strings on it).
Anyway, does anyone have a clue as to why my serial modules
are not working any longer since 2.5.53 (and neither 2.5.55)?
I am not missing it much for X but at least gpm is a good
thing.
Thanks,
Alexander
On Sat, Jan 11, 2003 at 11:12:42AM +0100, Alexander Koch wrote:
> devfs: v1.22 (20021013) Richard Gooch ([email protected])
> devfs: boot_options: 0x1
> pnp: the driver 'serial' has been registered
> pnp: pnp: match found with the PnP device '00:0f' and the driver 'serial'
> pnp: the device '00:0f' has been activated
> devfs_register(tts/2): could not append to parent, err: -17
> tts/2 at I/O 0x3e8 (irq = 4) is a 16550A
> pnp: pnp: match found with the PnP device '00:10' and the driver 'serial'
> pnp: the device '00:10' has been activated
> devfs_register(tts/3): could not append to parent, err: -17
> tts/3 at I/O 0x2e8 (irq = 3) is a 16550A
Argh. What I suspect has happened here is:
- the serial driver probed the ports tts/0 and tts/1 at I/O 0x3f8 and
0x2f8, and found two UARTS there. So it claimed IO resources at
these addresses.
(please show earlier messages in the boot log to confirm this.)
- PNP came along, and noticed that 0x3f8 and 0x2f8 were used, so re-
located the ports to 0x3e8 and 0x2e8, and told the serial driver
about them.
Since the serial driver knew there were ports at tts/0 and tts/1, it
allocated tts/2 and tts/3 to the "PNP" ports.
However, the original ports found are no longer at 0x3f8 and 0x2f8,
and therefore no longer work.
> That's coming from gpm startup, at least one of it, although
> I have /dev/tts/0 in my gpm.conf... Ah, it's hardcoded in the
> binary, it seems (doing a strings on it).
A "get you working" fix should be to tell gpm.conf to use /dev/tts/2
As for the -17 (EEXIST) error with devfs, I'd need to see the other
serial messages earlier in the boot log to work out what's going on,
as well as the kernel version the messages came from.
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
Hi Russel,
On Sat, 11 January 2003 09:44:35 +0000, Russell King wrote:
> Argh. What I suspect has happened here is:
>
> - the serial driver probed the ports tts/0 and tts/1 at I/O 0x3f8 and
> 0x2f8, and found two UARTS there. So it claimed IO resources at
> these addresses.
>
> (please show earlier messages in the boot log to confirm this.)
I have attached the dmesg from the 2.5.55 boot.
> - PNP came along, and noticed that 0x3f8 and 0x2f8 were used, so re-
> located the ports to 0x3e8 and 0x2e8, and told the serial driver
> about them.
>
> Since the serial driver knew there were ports at tts/0 and tts/1, it
> allocated tts/2 and tts/3 to the "PNP" ports.
>From reading that log explicitely, yes, PNP finds the same
ports again.
> A "get you working" fix should be to tell gpm.conf to use /dev/tts/2
Ah, this works, thanks lots!
> As for the -17 (EEXIST) error with devfs, I'd need to see the other
> serial messages earlier in the boot log to work out what's going on,
> as well as the kernel version the messages came from.
2.5.55 and 2.5.53 both have this and their log diffs not
at anything related to serial/pnp, etc.
Hope this helps a little for debug. If you need more
information please let me know.
Thanks lots,
Alexander