Date: Sun, 27 Mar 2005 04:07:39 +0200 From: Stefan Haglund <stefan.haglund@crystalnorth.com> To: freebsd-hardware@freebsd.org Subject: KT133A quirks present in FreeBSD [5.3]? Message-ID: <4246156B.3040909@crystalnorth.com>
next in thread | raw e-mail | index | archive | help
I have a private FreeBSD 5.3 file- & webserver sporting an IWill
KK266-R (KT133A/686B) motherboard, running a Athlon XP 1600+
(1400MHz). The board has an onboard CMD649 IDE RAID controller, that I
use as an ordinary IDE controller. This controller handles all four
disks in the system, other than the boot disk. The NIC is an Intel
PRO/100 (fxp).
I have a problem with files uploaded to the server is sometimes
getting corrupted. I've tried a lot of possible solutions (changing
PCI slots, BIOS settings, etc) to no avail.
My question is - is there a quirk fix in FreeBSD for the KT133A
northbridge, as (AFAIK) there is in Linux?
Snippet from Linux 2.6.11 - drivers/pci/quirks.c:
/*
* VIA Apollo KT133 needs PCI latency patch
* Made according to a windows driver based patch by George E.
Breese
* see PCI Latency Adjust on
[1]http://www.viahardware.com/download/viatweak.shtm
* Also see [2]http://www.au-ja.org/review-kt133a-1-en.phtml for
* the info on which Mr Breese based his work.
*
* Updated based on further information from the site and also on
* information provided by VIA
*/
static void __devinit quirk_vialatency(struct pci_dev *dev)
{
struct pci_dev *p;
u8 rev;
u8 busarb;
/* Ok we have a potential problem chipset here. Now see if we have
a buggy southbridge */
p = pci_get_device(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_82C686,
NULL);
if (p!=NULL) {
pci_read_config_byte(p, PCI_CLASS_REVISION, &rev);
/* 0x40 - 0x4f == 686B, 0x10 - 0x2f == 686A; thanks Dan Hollis
*/
/* Check for buggy part revisions */
if (rev < 0x40 || rev > 0x42)
goto exit;
} else {
p = pci_get_device(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8231,
NULL);
if (p==NULL) /* No problem parts */
goto exit;
pci_read_config_byte(p, PCI_CLASS_REVISION, &rev);
/* Check for buggy part revisions */
if (rev < 0x10 || rev > 0x12)
goto exit;
}
/*
* Ok we have the problem. Now set the PCI master grant to
* occur every master grant. The apparent bug is that under
high
* PCI load (quite common in Linux of course) you can get data
* loss when the CPU is held off the bus for 3 bus master
requests
* This happens to include the IDE controllers....
*
* VIA only apply this fix when an SB Live! is present but
under
* both Linux and Windows this isnt enough, and we have seen
* corruption without SB Live! but with things like 3 UDMA IDE
* controllers. So we ignore that bit of the VIA
recommendation..
*/
pci_read_config_byte(dev, 0x76, &busarb);
/* Set bit 4 and bi 5 of byte 76 to 0x01
"Master priority rotation on every PCI master grant */
busarb &= ~(1<<5);
busarb |= (1<<4);
pci_write_config_byte(dev, 0x76, busarb);
printk(KERN_INFO "Applying VIA southbridge workaround.\n");
exit:
pci_dev_put(p);
}
References
1. http://www.viahardware.com/download/viatweak.shtm
2. http://www.au-ja.org/review-kt133a-1-en.phtml
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4246156B.3040909>