| 1 |
Leonid Podolny wrote this: |
| 2 |
> Hi, all, |
| 3 |
> I lack some background to debug this issue. |
| 4 |
> I have two serial ports on my box. Naturally, I would expect them to be |
| 5 |
> ttyS0 and ttyS1. Indeed, dmesg shows the following: |
| 6 |
> |
| 7 |
> [snip] |
| 8 |
> Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled |
| 9 |
> pnp: Device 00:06 activated. |
| 10 |
> 00:06: ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A |
| 11 |
> pnp: Device 00:07 activated. |
| 12 |
> 00:07: ttyS1 at I/O 0x2f8 (irq = 3) is a 16550A |
| 13 |
> [snip] |
| 14 |
> |
| 15 |
> However, the actual nodes that udev (or hotplug?) creates are: |
| 16 |
> |
| 17 |
> [snip] |
| 18 |
> vyhuhol ~ # ls -l /dev/ttyS* |
| 19 |
> crw-rw---- 1 root uucp 4, 66 2007-06-26 09:04 /dev/ttyS2 |
| 20 |
> crw-rw---- 1 root uucp 4, 67 2007-06-26 09:04 /dev/ttyS3 |
| 21 |
> [snip] |
| 22 |
> |
| 23 |
> Needless to say, I'm unable to communicate with the port unless I |
| 24 |
> manually mknod ttyS0 with minor 64. |
| 25 |
> The question is: how do I make hotplug (or udev?) enumerate them |
| 26 |
> correctly, as ttyS0 and ttyS1? |
| 27 |
> |
| 28 |
|
| 29 |
I had this kind of problem some weeks ago. I resolved it rmmod'ing |
| 30 |
8250_pnp and 8250, sleeping 1 and modprobing them again in local.start. |
| 31 |
|
| 32 |
I know it wasn't a nice solution, but it worked for me. |
Archives