Re: [Re: Why do both network interfaces have the same ethernetaddress ?]
By default, the qfe interfaces on our E4000 use the same MAC. This is easily changed, however... Tom On Fri, 18 Feb 2000, Joshua Uziel wrote: > If memory serves me (and I could be way off), it also depends on > which card you have. Some of the older ones use the system's > MAC address, while some of the newer ones have their own MAC > address per port (like fe/qfe's, etc...) > > But yeah... I could be way off, and would have to check with a > friend of mine at SUN (who are now on their weekend break). > > * Mr. James W. Laferriere <[EMAIL PROTECTED]> [000218 23:27]: > > > > Hello Eric, Yup I found that too . But this is an old sparc20 > > and the openprom version is way to old(I'm sure) . Tnx, JimL > > > > On 18 Feb 2000, Eric wrote: > > > According to the OpenPROM manual, the "local-mac-address?" > > > parameter (default "false") controls whether or not all > > > interfaces assume the MAC address of the system. > > > > > > I don't have any system with multiple interfaces with which > > > to test, so perhaps give that a try. I don't know if that > > > property is queried by the Solaris drivers, or if the PROM > > > takes care of the job independently. > > > > > > Feel free to jump down my throat if I am incorrect. > > -- > Joshua Uziel, Senior Linux Consultant, Linuxcare, Inc. > 415.354.4878 tel, 415.701.7457 fax > [EMAIL PROTECTED], http://www.linuxcare.com/ > Linuxcare. Support for the revolution. > - > To unsubscribe from this list: send the line "unsubscribe sparclinux" in > the body of the message to [EMAIL PROTECTED] > - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: [Re: Why do both network interfaces have the same ethernet address ?]
If memory serves me (and I could be way off), it also depends on which card you have. Some of the older ones use the system's MAC address, while some of the newer ones have their own MAC address per port (like fe/qfe's, etc...) But yeah... I could be way off, and would have to check with a friend of mine at SUN (who are now on their weekend break). * Mr. James W. Laferriere <[EMAIL PROTECTED]> [000218 23:27]: > > Hello Eric, Yup I found that too . But this is an old sparc20 > and the openprom version is way to old(I'm sure) . Tnx, JimL > > On 18 Feb 2000, Eric wrote: > > According to the OpenPROM manual, the "local-mac-address?" > > parameter (default "false") controls whether or not all > > interfaces assume the MAC address of the system. > > > > I don't have any system with multiple interfaces with which > > to test, so perhaps give that a try. I don't know if that > > property is queried by the Solaris drivers, or if the PROM > > takes care of the job independently. > > > > Feel free to jump down my throat if I am incorrect. -- Joshua Uziel, Senior Linux Consultant, Linuxcare, Inc. 415.354.4878 tel, 415.701.7457 fax [EMAIL PROTECTED], http://www.linuxcare.com/ Linuxcare. Support for the revolution. - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: [Re: Why do both network interfaces have the same ethernetaddress ?]
Hello Eric, Yup I found that too . But this is an old sparc20 and the openprom version is way to old(I'm sure) . Tnx, JimL On 18 Feb 2000, Eric wrote: > According to the OpenPROM manual, the "local-mac-address?" > parameter (default "false") controls whether or not all > interfaces assume the MAC address of the system. > > I don't have any system with multiple interfaces with which > to test, so perhaps give that a try. I don't know if that > property is queried by the Solaris drivers, or if the PROM > takes care of the job independently. > > Feel free to jump down my throat if I am incorrect. ++ | James W. Laferriere | System Techniques | Give me VMS | | NetworkEngineer | 25416 22nd So | Give me Linux | | [EMAIL PROTECTED] | DesMoines WA 98198 | only on AXP | ++ - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: Why do both network interfaces have the same ethernet address ?
someone sent mail offlist about this eeprom setting (from eeprom man page):
local-mac-address? If true, network drivers use their
own MAC address, not system's.
Defaults to false.
looks like a good place to start.
On Fri, Feb 18, 2000 at 06:39:39PM -0800, Mr. James W. Laferriere chortled:
>
> Hello Horst,
>
> On Fri, 18 Feb 2000, Horst von Brand wrote:
> > Mark Luntzel <[EMAIL PROTECTED]> said:
> > > On Thu, Feb 17, 2000 at 10:12:53PM -0300, Horst von Brand chortled:
> > > > Mark Luntzel <[EMAIL PROTECTED]> said:
> > > > > On Thu, Feb 17, 2000 at 12:43:24PM +0100,
> > > > >[EMAIL PROTECTED] chortled:
> > > > > yes! you run into major problems when your design calls for both NIC's to
> > > > > be on the same subnet, as I found out. I never did find a way around
> > > > > this, other tha n to split subnets. Does anyone find this as annoying as
> > > > > I did? Is there a way around it, other than what I did?
> >
> > > > That design makes absolutely no sense. You won't get more data {in,out}
> > > > that way, only less as both ethX's are fighting for the Ether.
> >
> > > it only makes sense if you have IP constraints - such that it cost my
> > > company m ore $ to have IP space on another subnet. The question isnt so
> > > much about movin g more or less data, its about keeping costs down. sorry
> > > if I wasnt more clear. in any case, it wasnt possible to accomplish my
> > > objective a the time with that particular design, so, water under the
> > > bridge.
> >
> > How does this relieve IP constraints? It gives you _less_ access to the
> > net than just one interface, and uses up an IP address. You win only if the
> > interfaces are on different physical nets, and then the fact that they have
> > the same MAC is inconsecuential. And if you have two interfaces, you can
> > set up a private net behind that machine, then most IP constranints become
> > moot.
> I am not interested in giving the second interface an IP .
> Quite the contrary all I want it to have is a differant MAC
> to allow its being used for sniffing & some other playing
> around I want to do . But this -does- (afaik) require a
> unique MAC address for this to work . I hope this clears
> up the reasons I am very interested in getting a differant
> MAC address assigned to a secondary ether card in my sparc20 .
> JimL
> > --
> > Horst von Brand [EMAIL PROTECTED]
> > Casilla 9G, Viña del Mar, Chile +56 32 672616
>++
>| James W. Laferriere | System Techniques | Give me VMS |
>| NetworkEngineer | 25416 22nd So | Give me Linux |
>| [EMAIL PROTECTED] | DesMoines WA 98198 | only on AXP |
>++
--
Now a message from Mr. T. to all of the server-fingerin', Aphex Twin-listenin',
Linux-usin', pot smokin' kids out there: "Don't be a foo', stay in schoo'!"
-
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of the message to [EMAIL PROTECTED]
Re: Why do both network interfaces have the same ethernet address?
Hello Horst,
On Fri, 18 Feb 2000, Horst von Brand wrote:
> Mark Luntzel <[EMAIL PROTECTED]> said:
> > On Thu, Feb 17, 2000 at 10:12:53PM -0300, Horst von Brand chortled:
> > > Mark Luntzel <[EMAIL PROTECTED]> said:
> > > > On Thu, Feb 17, 2000 at 12:43:24PM +0100,
> > > >[EMAIL PROTECTED] chortled:
> > > > yes! you run into major problems when your design calls for both NIC's to
> > > > be on the same subnet, as I found out. I never did find a way around
> > > > this, other tha n to split subnets. Does anyone find this as annoying as
> > > > I did? Is there a way around it, other than what I did?
>
> > > That design makes absolutely no sense. You won't get more data {in,out}
> > > that way, only less as both ethX's are fighting for the Ether.
>
> > it only makes sense if you have IP constraints - such that it cost my
> > company m ore $ to have IP space on another subnet. The question isnt so
> > much about movin g more or less data, its about keeping costs down. sorry
> > if I wasnt more clear. in any case, it wasnt possible to accomplish my
> > objective a the time with that particular design, so, water under the
> > bridge.
>
> How does this relieve IP constraints? It gives you _less_ access to the
> net than just one interface, and uses up an IP address. You win only if the
> interfaces are on different physical nets, and then the fact that they have
> the same MAC is inconsecuential. And if you have two interfaces, you can
> set up a private net behind that machine, then most IP constranints become
> moot.
I am not interested in giving the second interface an IP .
Quite the contrary all I want it to have is a differant MAC
to allow its being used for sniffing & some other playing
around I want to do . But this -does- (afaik) require a
unique MAC address for this to work . I hope this clears
up the reasons I am very interested in getting a differant
MAC address assigned to a secondary ether card in my sparc20 .
JimL
> --
> Horst von Brand [EMAIL PROTECTED]
> Casilla 9G, Viña del Mar, Chile +56 32 672616
++
| James W. Laferriere | System Techniques | Give me VMS |
| NetworkEngineer | 25416 22nd So | Give me Linux |
| [EMAIL PROTECTED] | DesMoines WA 98198 | only on AXP |
++
-
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of the message to [EMAIL PROTECTED]
Re: Why do both network interfaces have the same ethernet address ?
Tom Kyle <[EMAIL PROTECTED]> said: [...] > Personally, I think it was done this way to make software licenses easier > to track (ie, if your nic self-destructs or you upgrade, you don't have to > go change your license). Doesn't help, as I can happily use your MAC here on my net. AFAIK, the CPUs have a serial number that can't be changed, and you could (try to) use that for this. But that doesn't work either, as those schemes are usually easy to make inoperative by editing the binary a bit ;-) > I've had good experiences with using IP aliases on our 1000e when I need > to have two addresses on the same subnet. You should be able to do an > 'ifconfig hme0:1 123.45.67.89' yadda yadda. The same idea (I forget the syntax) works on Solaris, BTW. -- Horst von Brand [EMAIL PROTECTED] Casilla 9G, Viña del Mar, Chile +56 32 672616 - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: Why do both network interfaces have the same ethernet address ?
> > > it only makes sense if you have IP constraints - such that it cost my > > company m ore $ to have IP space on another subnet. The question isnt so > > much about movin g more or less data, its about keeping costs down. sorry > > if I wasnt more clear. in any case, it wasnt possible to accomplish my > > objective a the time with that particular design, so, water under the > > bridge. > > How does this relieve IP constraints? It gives you _less_ access to the > net than just one interface, and uses up an IP address. You win only if the > interfaces are on different physical nets, and then the fact that they have > the same MAC is inconsecuential. And if you have two interfaces, you can > set up a private net behind that machine, then most IP constranints become > moot. this is all academic now, FWIW. some ISP's offer packages that give you a certain block of IP addresses, to go outside of these costs you more $. thats all I am saying. I certainly dont agree that its a good or right way to have things, but business is business. Not sure what you mean by "_less_ access to the net". You dont seem to understand that I was constrained by *cost* issues. Also, you assume that I could have gotten away with having one of these interfaces on a private net. bzz. sorry. in any case, as I said before, this isnt an issue anymore. what fun venture capital can bring. > -- > Horst von Brand [EMAIL PROTECTED] > Casilla 9G, Viña del Mar, Chile +56 32 672616 -- Now a message from Mr. T. to all of the server-fingerin', Aphex Twin-listenin', Linux-usin', pot smokin' kids out there: "Don't be a foo', stay in schoo'!" - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: [Re: Why do both network interfaces have the same ethernet address ?]
According to the OpenPROM manual, the "local-mac-address?" parameter (default "false") controls whether or not all interfaces assume the MAC address of the system. I don't have any system with multiple interfaces with which to test, so perhaps give that a try. I don't know if that property is queried by the Solaris drivers, or if the PROM takes care of the job independently. Feel free to jump down my throat if I am incorrect. E > > Hello All, Having seen that Sloaris has gotten around this > > predicument I find myself wondering why linux has not ? > > Is it a security issue ?? I can sure see it being one. (but) > > Is it ... ? Tia, JimL Get free email and a permanent address at http://www.netaddress.com/?N=1 - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: Why do both network interfaces have the same ethernet address ?
Mark Luntzel <[EMAIL PROTECTED]> said:
> On Thu, Feb 17, 2000 at 10:12:53PM -0300, Horst von Brand chortled:
> > Mark Luntzel <[EMAIL PROTECTED]> said:
> > > On Thu, Feb 17, 2000 at 12:43:24PM +0100,
> > >[EMAIL PROTECTED] chortled:
> > > yes! you run into major problems when your design calls for both NIC's to
> > > be on the same subnet, as I found out. I never did find a way around
> > > this, other tha n to split subnets. Does anyone find this as annoying as
> > > I did? Is there a way around it, other than what I did?
> > That design makes absolutely no sense. You won't get more data {in,out}
> > that way, only less as both ethX's are fighting for the Ether.
> it only makes sense if you have IP constraints - such that it cost my
> company m ore $ to have IP space on another subnet. The question isnt so
> much about movin g more or less data, its about keeping costs down. sorry
> if I wasnt more clear. in any case, it wasnt possible to accomplish my
> objective a the time with that particular design, so, water under the
> bridge.
How does this relieve IP constraints? It gives you _less_ access to the
net than just one interface, and uses up an IP address. You win only if the
interfaces are on different physical nets, and then the fact that they have
the same MAC is inconsecuential. And if you have two interfaces, you can
set up a private net behind that machine, then most IP constranints become
moot.
--
Horst von Brand [EMAIL PROTECTED]
Casilla 9G, Viña del Mar, Chile +56 32 672616
-
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of the message to [EMAIL PROTECTED]
RE: Why do both network interfaces have the same ethernet address ?
You're right, it is tied to the host id, but I believe the newer machines have the host id as part of the hardware address. The host id seemes to be '80' + last six digits of the MAC. Madness. ;) Tom On Fri, 18 Feb 2000 [EMAIL PROTECTED] wrote: > > -Original Message- > > From: Tom Kyle [mailto:[EMAIL PROTECTED]] > > > > Nope, the PCI-based Sparcs have a single hardware address, > > just like the SBus machines. > > Ho, hum.. > > > Personally, I think it was done this way to make software > > licenses easier > > to track (ie, if your nic self-destructs or you upgrade, you > > don't have to > > go change your license). > > Yeah, but most Sun licensing that I've experienced uses > the machine's 'hostid' rather than the hardware address. > My experience is limited though YMMV ;-) > > -Andy > > -- > Andrew Walker Kvaerner Oil & Gas a.s. > [EMAIL PROTECTED]P.O. Box 222, > [EMAIL PROTECTED] N-1324 Lysaker, Norway > > . Will you speak out? . > - > To unsubscribe from this list: send the line "unsubscribe sparclinux" in > the body of the message to [EMAIL PROTECTED] > - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: FYI: Stock 2.3.46 doesn't link
Fixes for this are in 2.3.47-pre3 already. Later, David S. Miller [EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
FYI: Stock 2.3.46 doesn't link
(Dual HyperSparc RT620 SparcStation10 machine) ... kernel/kernel.o: In function `access_one_page': kernel/kernel.o(.text+0x13294): undefined reference to `flush_icache_page' kernel/kernel.o(.text+0x13294): relocation truncated to fit: R_SPARC_WDISP30 flush_icache_page mm/mm.o: In function `do_swap_page': mm/mm.o(.text+0x19b0): undefined reference to `flush_icache_page' mm/mm.o(.text+0x19b0): relocation truncated to fit: R_SPARC_WDISP30 flush_icache_page mm/mm.o: In function `do_no_page': mm/mm.o(.text+0x1d64): undefined reference to `flush_icache_page' mm/mm.o(.text+0x1d64): relocation truncated to fit: R_SPARC_WDISP30 flush_icache_page mm/mm.o: In function `free_pgtables': mm/mm.o(.text+0x2f58): undefined reference to `pgd_index' mm/mm.o(.text+0x2f58): relocation truncated to fit: R_SPARC_WDISP30 pgd_index mm/mm.o(.text+0x2f64): undefined reference to `pgd_index' mm/mm.o(.text+0x2f64): relocation truncated to fit: R_SPARC_WDISP30 pgd_index make: *** [vmlinux] Error 1 .. But The CVS version is nice 'n shiny ;) Cheers, Florian. - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: ll-rw-blk.c in 2.3.46 on dual HyperSparc RT620
On Fri, 18 Feb 2000, Florian-Daniel Otel wrote: >ll_rw_blk.c:256: structure has no member named `back_merges_fn' ftp://ftp.*.kernel.org/pub/linux/kernel/people/andrea/patches/v2.3/2.3.46pre5/elevator-merge-typo-1.gz Andrea - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: ll-rw-blk.c in 2.3.46 on dual HyperSparc RT620
sed 's/back_merges_fn/back_merge_fn/' TMP mv TMP ll_rw_blk.c Later, David S. Miller [EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
ll-rw-blk.c in 2.3.46 on dual HyperSparc RT620
Hi all, On a SparcStation10: % cat /proc/cpuinfo cpu : ROSS HyperSparc RT620 fpu : ROSS HyperSparc combined IU/FPU promlib : Version 3 Revision 2 prom: 2.25 type: sun4m ncpus probed: 2 ncpus active: 2 Cpu0Bogo: 66.56 Cpu1Bogo: 66.56 MMU type: ROSS HyperSparc invall : 0 invmm : 0 invrnge : 0 invpg : 0 contexts: 4096 CPU0: online CPU1: online ..2.3.46: make[3]: Entering directory `/usr/src/linux-2.3.46/drivers/block' gcc -D__KERNEL__ -I/usr/src/linux-2.3.46/include -D__SMP__ -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -fno-strict-aliasing -m32 -pipe -mno-fpu -fcall-used-g5 -fcall-used-g7 -DEXPORT_SYMTAB -c ll_rw_blk.c ll_rw_blk.c:237: warning: static declaration for `generic_plug_device' follows non-static ll_rw_blk.c: In function `blk_init_queue': ll_rw_blk.c:256: structure has no member named `back_merges_fn' ll_rw_blk.c: In function `generic_make_request': ll_rw_blk.c:682: warning: `__entry' might be used uninitialized in this function ll_rw_blk.c:950: warning: `entry' might be used uninitialized in this function ll_rw_blk.c:566: warning: `entry' might be used uninitialized in this function make[3]: *** [ll_rw_blk.o] Error 1 make[3]: Leaving directory `/usr/src/linux-2.3.46/drivers/block' make[2]: *** [first_rule] Error 2 make[2]: Leaving directory `/usr/src/linux-2.3.46/drivers/block' make[1]: *** [_subdir_block] Error 2 make[1]: Leaving directory `/usr/src/linux-2.3.46/drivers' make: *** [_dir_drivers] Error 2 [root@courvoisier linux]# (Caveat emptor: Didn't had the time for a detailed dig through the sources and to see who/how/why ) Cheers, Florian P.S. The only reason I try 2.3.46 is to use the new 2.3 PPP driver for a PPPOE server on a 100Mbit Eth (the 2.2 driver sucks and ppp-2.3.x, x<=11 goes bananas). Any alternate ideas/reality check welcome. - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
RE: Why do both network interfaces have the same ethernet address ?
> -Original Message- > From: Tom Kyle [mailto:[EMAIL PROTECTED]] > > Nope, the PCI-based Sparcs have a single hardware address, > just like the SBus machines. Ho, hum.. > Personally, I think it was done this way to make software > licenses easier > to track (ie, if your nic self-destructs or you upgrade, you > don't have to > go change your license). Yeah, but most Sun licensing that I've experienced uses the machine's 'hostid' rather than the hardware address. My experience is limited though YMMV ;-) -Andy -- Andrew Walker Kvaerner Oil & Gas a.s. [EMAIL PROTECTED] P.O. Box 222, [EMAIL PROTECTED]N-1324 Lysaker, Norway . Will you speak out? . - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
RE: Why do both network interfaces have the same ethernet address?
Nope, the PCI-based Sparcs have a single hardware address, just like the SBus machines. This really befuddled some Marimba reps who came in yesterday. ;) Personally, I think it was done this way to make software licenses easier to track (ie, if your nic self-destructs or you upgrade, you don't have to go change your license). I've had good experiences with using IP aliases on our 1000e when I need to have two addresses on the same subnet. You should be able to do an 'ifconfig hme0:1 123.45.67.89' yadda yadda. Tom On Thu, 17 Feb 2000 [EMAIL PROTECTED] wrote: > > -Original Message- > > From: Mr. James W. Laferriere [mailto:[EMAIL PROTECTED]] > > > > Hello All, Having seen that Sloaris has gotten around this > > predicument I find myself wondering why linux has not ? > > Is it a security issue ?? I can sure see it being one. (but) > > Is it ... ? Tia, JimL > > I wouldn't call it a security issue - more a Sun design decision. > On older Sun hardware (can't speak for Ultra's 'cos I live mostly > in sun4c/sun4m land) the hardware address is programmed into the > NVRAM chip for the workstation, and this address is automatically > assigned to all the network interfaces (e.g. the ie/le/qe/hme cards). > As long as the interfaces are connected to different network segments > then this never causes a problem. > > On both Solaris and Linux you can change the hardware address of > each individual interface through the ifconfig command, but you > should make sure that the address you are assigning really is free. > > I just confirmed that an SS10 running Solaris 2.6 has the same hw address > for both hme0 and le0 - in what way has Solaris "gotten around this" where > Linux hasn't? The only thing I can think of is that newer PCI based cards > probably have a hw-address assigned to the card from the manufacturer as > is usual in the PC world. Then the problem automagically disappears. > Maybe I'm missing something here.. > > -Andy > > -- > Andrew Walker Kvaerner Oil & Gas a.s. > [EMAIL PROTECTED]P.O. Box 222, > [EMAIL PROTECTED] N-1324 Lysaker, Norway > > . Will you speak out? . > - > To unsubscribe from this list: send the line "unsubscribe sparclinux" in > the body of the message to [EMAIL PROTECTED] > - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
Re: Why do both network interfaces have the same ethernet address ?
Mark Luntzel <[EMAIL PROTECTED]> said:
> On Thu, Feb 17, 2000 at 12:43:24PM +0100, [EMAIL PROTECTED] chortled:
> yes! you run into major problems when your design calls for both NIC's to
> be on the same subnet, as I found out. I never did find a way around
> this, other tha n to split subnets. Does anyone find this as annoying as
> I did? Is there a way around it, other than what I did?
That design makes absolutely no sense. You won't get more data {in,out}
that way, only less as both ethX's are fighting for the Ether.
> > On both Solaris and Linux you can change the hardware address of
> > each individual interface through the ifconfig command, but you
> > should make sure that the address you are assigning really is free.
And that has worked from the SunOS 4.0 times, to the very least.
--
Horst von Brand [EMAIL PROTECTED]
Casilla 9G, Viña del Mar, Chile +56 32 672616
-
To unsubscribe from this list: send the line "unsubscribe sparclinux" in
the body of the message to [EMAIL PROTECTED]
Re: RH6.1 on sun4?
On Thu, Feb 17, 2000 at 09:02:52PM +, Matthew Kirkwood wrote: > Hi, > > [ this may perhaps be more appropriate on a redhat list, but > I know that David and Jakub hang out here ] > > I'm attempting to revive my 4/330 again[0], and to this end I > have acquired a RH6.1 CD. > > The last time I installed it was with the last Ultrapenguin, > which was based around a 5.9 beta, as I recall, and it wasn't > too hard to get the thing going - I just had to extract all > of the installer stuff into an NFS root directory, drop in a > new ld.so (which doesn't seem to be necessary any more) and > boot with init=/bin/sh to make sure that everything was OK > before I fired off the installer. > > These days, things seem a lot harder, and I'm a little at a > loss to know what to do. I'm guessing that, with some hacking, > I could attach a the RH61 install initrd to a sun4 kernel and > boot that, but unfortunately piggyback doesn't seem to like > the headerless sun4 kernels and I can't find a detached initrd > to attach anyway. Red Hat Linux uses the TILO bootloader for TFTP images. See misc/src/trees/tilo-0.2/ on the CD (or FTP or wherever). It should have support for sun4, though it was never tested. You just should modify the /tilo shell script in there so that it passes sun4=, size4=, root4= arguments instead of the 4c or 4u, plus you should kill the a.out header at the end (say using dd to shift it 32 bytes down). If you'll have further problems, just let me know. Once (if ever) SILO support for sun4 is done, I think we'll put sun4 support into the distribution. Cheers, Jakub ___ Jakub Jelinek | [EMAIL PROTECTED] | http://sunsite.mff.cuni.cz/~jj Linux version 2.3.46 on a sparc64 machine (1343.49 BogoMips) ___ - To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of the message to [EMAIL PROTECTED]
