Re: [Xenomai-core] [patch] serial driver fixes/improvements

[Jan Kiszka <kiszka@domain.hid>]

[-- Attachment #1: Type: text/plain, Size: 1642 bytes --]

Jan Kiszka wrote:
> Hannes Mayer wrote:
> 
>>Hi Jan et al.!
>>
>>Jan Kiszka wrote:
>>[...]
>>
>>
>>>+modprobe xeno_16550A ioaddr=<io1>[,<io2>...] irq=<irq1>[,<irq2>...]
>>>+                     [tx_fifo=<len1>[,<len2>...]] [start_index=<index>]
>>>+
>>>+Arguments:
>>>+    ioaddr<N>   - I/O address of device <N> (e.g. 0x3f8 for ttyS0)
>>>+    irq<N>      - interrupt number of device <N> (e.g. 4 for ttyS0)
>>>+    tx_fifo<N>  - Transmitter FIFO size in bytes of device <N>,
>>>default is 16
>>>+    start_index - First device instance number to be used, default is 0
>>
>>
>>If I may throw in my 2 eurocents:
>>I think the Arguments-section is a bit misleading.
>>One doesn't specify ioaddr1, ioaddr2 etc. on the command line, as suggested
>>by ioaddr<N>, but ioaddr=[1.dev],[2.dev]...
>>Same for irq and tx_fifo.
>>
>>What about this:
>>
>>Arguments:
>>    ioaddr      - I/O addresses of devices (e.g. 0x3f8 for ttyS0),
>>separated by comma
>>    irq         - interrupt numbers of devices (e.g. 4 for ttyS0),
>>separated by comma
>>    tx_fifo     - Transmitter FIFO sizes in bytes of devices, default is
>>16, separated by comma
>>    start_index - First device instance number to be used, default is 0
>>=> rtser0
>>
>>Just trying to make it as clear as possible :-)
>>
> 
> 
> Indeed, my version is misleading, your version makes more sense. Just
> the parameter for ttyS0 are not necessarily the one you listed, only

Oops, actually I originally sorted ttyS0 to 0x3f8, not you. As I said:
don't trust anything I produce between 3 and 5 am...

> typically. Will post a patch.
> 

Patch posted, feel free to comment again!

Thanks,
Jan

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 256 bytes --]