Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 18 May 1998 10:28:11 +0930
From:      Greg Lehey <grog@lemis.com>
To:        Alexey Lukin <al@cn.ua>, hackers@FreeBSD.ORG
Subject:   Re: Missing sio0 and sio1 (was: HELP! Please, HELP!)
Message-ID:  <19980518102811.J427@freebie.lemis.com>
In-Reply-To: <355EC236.4626669B@cn.ua>; from Alexey Lukin on Sun, May 17, 1998 at 01:55:50PM %2B0300
References:  <6jleo9$ee0@elit.elit.chernigov.ua> <355EA834.A657FA38@cn.ua> <19980517195812.B427@freebie.lemis.com> <355EC236.4626669B@cn.ua>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 17 May 1998 at 13:55:50 +0300, Alexey Lukin wrote:
> Hi, Greg!
>
> Greg Lehey wrote:
> [...]
>> ?
>> ? I giot the same sort of problems with different new motherboards.
>> ? Well, the SIO detection code is TOO sophisticated and does not recognise
>> ? MOST of on-board 16550 sios.
>>
>> I haven't heard anybody claim that before.  Can you substantiate this
>> claim?  Which motherboards have you used?  What UARTs do they use?
>> Did the patch work with all of them?
>
> I tried motherboards of 2 wendors: "Micro Star" (MS-5156)
> and few lastI "Iwill" boards.

OK, we know about the IWill boards.  That's what the patch is for.
Have you tried it?

> All boards have 16550A chips. Linux 2.0.32,33 recognized sio ports and
> worked well,
> but 2.2.5-R, 2.2.6-R refused to DETECT sio ports.

(minor nit) could you please try to write approximately even-length
lines?  Something like this:

> All boards have 16550A chips. Linux 2.0.32,33 recognized sio ports
> and worked well, but 2.2.5-R, 2.2.6-R refused to DETECT sio ports.

It makes things a lot easier to read. (end nit)

No, modern boards use integrated multi-I/O chips.  The IWill boards
use an ALi chip which includes a UART which looks like a 16550, but
it's not a 16550.  In particular, it has additional register which
need to be initialized.

>> In any case, code which doesn't recognize all UARTS is not *too*
>> sophisticated.
>
> I mean only style of /sys/i386/isa/sio.c.

Sorry, I don't understand.

>> At the moment, we don't know that this affects more than a fraction of
>> all motherboards.  Yuri wasn't confident enough to commit this patch
>> (which, so far, has always worked), because he was afraid it might
>> break the majority for whom the current code works.
>
> Ok, Ok. But I spent a week changing motherboards for simplest worksation
> :-)

If it makes you any happier, it took me about that long, too.

> IMHO, sio detection procedure may have options "STRICT" to do what
> it tries to do. If I just want to have 9600 console or mouse on sio
> device, I do not need all the troubles with detection.  BTHW, SIO
> code itself works just fine on 115200.
>
> And please, exuse me if I said something in offesive manner. It's my
> pure English, not the intention.

Yes, we're thinking about it.  The problem is that so far very few
people have been affected, and until a number of people are prepared
to try this out with their hardware, we're concerned that the fix may
break a significant number of other hardware.

Greg
--
See complete headers for address and phone numbers
finger grog@lemis.com for PGP public key

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980518102811.J427>