Re: Chipset SiS735 / NIC SiS 900 - PR kern/30836
Dan Langille wrote:
Hi folks,
It appears I am not the first to buy a K7S5A motherboard (an SiS 735
chipset) with integrated NIC. I can't get my NIC to work either.
Bill: I'm ready to help with this. If you so wish, I can give you a
login on the box.
I received these two patches from J. Lemon. They sort of do something.
I can see the first two blocks of the MAC address but it doesn't move
to the next two pairs and that is as far as they go towards working. I
converted what JLemon sent to patches to stable. I have attached them.
I have two systems using the K7S5A motherboard and one that uses an
Amptron 830LM, which is also based on the SiS 735 chipset.
Kent
cheers.
--
Kent Stewart
Richland, WA
mailto:[EMAIL PROTECTED]
http://users.owt.com/kstewart/index.html
FreeBSD News http://daily.daemonnews.org/
--- if_sis.c.orig Sat Dec 29 03:45:11 2001
+++ if_sis.cSat Dec 29 04:40:18 2001
@@ -439,6 +439,31 @@
pci_write_config(bridge, 0x48, reg ~0x40, 1);
return;
}
+
+static void sis635_mac_addr(sc, dest)
+ struct sis_softc*sc;
+ caddr_t dest;
+{
+ u_int32_t filtsave;
+
+ filtsave = CSR_READ_4(sc, SIS_RXFILT_CTL);
+
+ SIS_SETBIT(sc, SIS_CSR, SIS_CSR_RELOAD);
+ CSR_WRITE_4(sc, SIS_CSR, 0);
+
+ /* Disable the the receive filter. */
+ SIS_CLRBIT(sc, SIS_RXFILT_CTL, SIS_RXFILTCTL_ENABLE);
+
+ /* Get MAC address */
+ CSR_WRITE_4(sc, SIS_RXFILT_CTL, NS_FILTADDR_PAR0);
+ ((u_int16_t *)dest)[0] = CSR_READ_2(sc, SIS_RXFILT_DATA);
+ CSR_WRITE_4(sc, SIS_RXFILT_CTL, NS_FILTADDR_PAR1);
+ ((u_int16_t *)dest)[1] = CSR_READ_2(sc, SIS_RXFILT_DATA);
+ CSR_WRITE_4(sc, SIS_RXFILT_CTL, NS_FILTADDR_PAR2);
+ ((u_int16_t *)dest)[2] = CSR_READ_2(sc, SIS_RXFILT_DATA);
+
+ CSR_WRITE_4(sc, SIS_RXFILT_CTL, filtsave);
+}
#endif
static int sis_miibus_readreg(dev, phy, reg)
@@ -891,6 +916,8 @@
command == SIS_REV_630E ||
command == SIS_REV_630EA1)
sis_read_cmos(sc, dev, (caddr_t)eaddr, 0x9, 6);
+ else if (command == SIS_REV_635A)
+ sis635_mac_addr(sc, (caddr_t)eaddr);
else
#endif
sis_read_eeprom(sc, (caddr_t)eaddr,
--- if_sisreg.h.origSat Dec 29 03:45:36 2001
+++ if_sisreg.h Sat Dec 29 03:58:00 2001
@@ -105,6 +105,7 @@
#define SIS_CSR_RX_RESET 0x0020
#define SIS_CSR_SOFTINTR 0x0080
#define SIS_CSR_RESET 0x0100
+#define SIS_CSR_RELOAD 0x0400
#define SIS_CFG_BIGENDIAN 0x0001
#define SIS_CFG_PERR_DETECT0x0008
@@ -367,6 +368,7 @@
#define SIS_REV_630E 0x0081
#define SIS_REV_630S 0x0082
#define SIS_REV_630EA1 0x0083
+#define SIS_REV_635A 0x0090
/*
* NatSemi vendor ID
@@ -410,6 +412,8 @@
#define CSR_WRITE_4(sc, reg, val) \
bus_space_write_4(sc-sis_btag, sc-sis_bhandle, reg, val)
+#define CSR_READ_2(sc, reg)\
+ bus_space_read_2(sc-sis_btag, sc-sis_bhandle, reg)
#define CSR_READ_4(sc, reg)\
bus_space_read_4(sc-sis_btag, sc-sis_bhandle, reg)
Re: 4.5-PRE panic in vfs_cache with softupdates, via chipset
On Thu, 10 Jan 2002, Nils Holland wrote: On Wed, Jan 09, 2002 at 10:03:09PM +, Gavin Atkinson stood up and spoke: I was running buildworld, with /usr/src and /usr/obj on seperate partitions, both with softupdates on. I have the problematic VIA southbridge. I was building world, then suspended it (ctrl-z) before suspending the laptop with 'apm -z'. Later, woke it up, and about 20 seconds later resumed the build. Two or three more files were compiled before the kernel panic occured. Welcome to the club ;-) Now, I'm no kernel expert either, but I wonder if you can reproduce your problem. If it's similar to mine, you should be able exactly what you did when the panic occured, and in about two out of three cases the problem should be there again. No - I think that it's irregular. It's only happened twice, once while running buildworld, and once whlie running grep on /usr/src/sys. I certanly cannot make it panic on command. If I had more money on my bank account right now, I'd probably simply replace my mainboard with an unaffected one (either no VIA chipset or one of the latest VIA things) and send my current one to Soren for further investigation. I guess that if he'd see how strange it performs, he'd throw it out of the window after a short period of time, and that's what I almost did for about 20 times so far ;-) heh. sadly this is a brand new laptop... Thanks, Gavin To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
New cdboot ISO available
A new experimental CD bootstrap was recently added to -stable for the i386 architecture that allows the full contents of the CD to be used when booting. The method we currently uses relies on the BIOS emulating a floppy drive from a floppy image file on the CD. This limits the size of the kernel that we can boot off of. The new CD bootstrap allows a full GENERIC kernel to be used when booting off of a CD, allowing more drivers to work out of the box. However, not all BIOS's that support the older method support this newer method of CD booting. The release engineering team would like feedback as to roughly how many machines do not work with the new bootstrap utility as well as how many machines do work. If most machines work with the new bootstrap, then we may switch to using the new bootstrap on the release ISO images. Note that this only affects booting directly from the CD. Booting from floppies and then installing off of the CD will work the same with both images. The cdboot image can be found at ftp://releng4.freebsd.org/pub/FreeBSD/releases/i386/ISO-IMAGES/4.5-RC1-cdboot.iso The MD5 checksum of the image is: MD5 (4.5-RC1-cdboot.iso) = 294c50cdfe77f51b8224ae9679afeabf This new bootstrap utility is used on Windows NT 4, Windows ME, and Windows 2000 CD's (and probably Windows XP CD's), so if your system has a 'Designed for' sticker featuring one of those operating systems, it will probably work with cdboot. Please test this ISO and provide feedback to the -stable or -qa lists thanks. RC1 x86 cdboot Errata : 1. The kernel configuration tool USERCONFIG does not currently work on the cdboot ISO. Thanks! - The FreeBSD 4.5 Release Engineering Team To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Cvsup 09/01/2002 /make buildworld problem
Hello! Problem is make buildworld for cvsup 09.01/2002 make buildword === usr.bin /usr/src/share/mk/bsd.subdir.mk , line 81: Inconsistent operator for ftp make: fatal errors encountered -- cannot continue *** Error code 1 Stop in /usr/src *** Error code 1 To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
RE: Bigger Harddisk
And there's Dan O'Connor's excellent info on http://www.mostgraveconcern.com/freebsd/ See the Moving to a Larger Hard Drive section. Lawrence Farr EPC Direct Limited -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Christopher Schulte Sent: 10 January 2002 07:46 To: Erich Zigler; [EMAIL PROTECTED] Subject: Re: Bigger Harddisk At 12:20 AM 1/10/2002 -0600, Erich Zigler wrote: I have recently freed up a couple partitions on my hard drive. I would like to expand the FreeBSD partition. Is this possible? If you're running FreeBSD 4.4 or later, growfs(8) may be able to help you. -- I like work. I can sit here and watch it for hours. --chris To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Re: Cvsup 09/01/2002 /make buildworld problem
Spider wrote: Hello! Problem is make buildworld for cvsup 09.01/2002 make buildword === usr.bin /usr/src/share/mk/bsd.subdir.mk , line 81: Inconsistent operator for ftp make: fatal errors encountered -- cannot continue *** Error code 1 Stop in /usr/src *** Error code 1 I had no problems cvsuping RELENG_4 an hour ago and doing a buildworld. What did you cvsup? Is the buildword a typing error for your email? Kent To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message . -- Kent Stewart Richland, WA mailto:[EMAIL PROTECTED] http://users.owt.com/kstewart/index.html FreeBSD News http://daily.daemonnews.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
Euro and XFree86-4.1.0, the canonical way?
how i got the euro: using Option XkbLayout de(euro) in /etc/X11/XF86Config did not work. The xkb compiler complained in many ways (clash with nodeadkeys variant and more), even when not using german keyboard. the following worked using LANG=de_DE.ISO8859-15 AND xmodmap -e 'keycode 26 = e E EuroSign cent' does the thing. I cannot believe that xmodmap is the only way ... -- Fritz Heinrichmeyer mailto:[EMAIL PROTECTED] FernUniversitaet Hagen, LG ES, 58084 Hagen (Germany) tel:+49 2331/987-1166 fax:987-355 http://www-es.fernuni-hagen.de/~jfh To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-stable in the body of the message
