Date: Sat, 2 Jan 1999 09:09:42 -0500 (EST) From: Peter Dufault <dufault@hda.com> To: current@FreeBSD.ORG Subject: if_ep probe failures related to reading back settings Message-ID: <199901021409.JAA18234@hda.hda.com>
next in thread | raw e-mail | index | archive | help
I have trouble properly reading the settings back from a 3C509 on
a 180MHz Cyrix MediaGX crash box. The result is that it won't
always probe in at boot.
I find that by increasing the DELAY by a factor of 10 and more
importantly adding a delay before reading each serial bit of data
(I assume we're reading back a serial EEPROM):
Before:
> outb(id_port, 0x80 + offset);
> DELAY(1000);
> for (i = 0; i < 16; i++)
> data = (data << 1) | (inw(id_port) & 1);
After:
>#define DELAY_MULTIPLE 10
> outb(id_port, 0x80 + offset);
> for (i = 0; i < 16; i++) {
> DELAY(DELAY_MULTIPLE * 1000);
> data = (data << 1) | (inw(id_port) & 1);
> }
that it always works. Without this I get signs of not reading
back the data properly - for example, sometimes it will boot
with a corrupt hardware address so that on another system I'll see
this:
> arp: 198.252.184.2 moved from 00:20:ff:ff:af:a0 to 00:20:eb:e8:96:3a
> arp: 198.252.184.2 moved from 00:20:eb:e8:96:3a to e0:00:af:a0:96:3a
> de0: receive: e0:00:af:a0:96:3a: bad crc
> arp: 198.252.184.2 moved from e0:00:af:a0:96:3a to 00:20:af:a0:e5:8e
> arp: 198.252.184.2 moved from 00:20:af:a0:e5:8e to 00:20:af:a0:96:3a
Can someone check if we may be violating timings on reading that
serial bus, or is it likely that this Cyrix has an inw that is too
fast and I have to monkey with some BIOS settings?
Peter
--
Peter Dufault (dufault@hda.com) Realtime development, Machine control,
HD Associates, Inc. Safety critical systems, Agency approval
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199901021409.JAA18234>