Re: Binding process to a fixed processor
Actually I am looking for some command or system call which I can execute from my user level program so that I can bind one of my process to a processor. like pbind command in Linux.. Or do I have to write a system call to do that ?? Dennis Julian Elischer [EMAIL PROTECTED] wrote: Dennis George wrote: Hi, I am working on freeBSD 5.2. Dennis Julian Elischer wrote: Dennis George wrote: Hi all, I am working on a intel based multi processor system. I like to know how can I bind one process permanently to one processor. and other one for general use. You can bind a thread to one processor, in the kernel, but I don't know offhand if there is a user interface for it however.. (I'd have to go look at the code again). (goes to look) There is code that can bind a thread to the current processor that it is on, but nothing uses it that I can see.. If you wrote a kernel module you could write your own syscall to use it.. This is of course different from binding a thread to a processor EXCLUSIVELY so that no other thread can use it. thanks in advance Dennis which version of the system are you using? - Do you Yahoo!? Win 1 of 4,000 free domain names from Yahoo! Enter now. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED] - Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED] - Do you Yahoo!? Y! Messenger - Communicate in real time. Download now. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Support for SMT in latest FreeBSD
Hi, Acutally I was wondering if there is no support for SMT / SMP then can freeBSD support dual processors. Or can I utlize dual-processor in its fullness ? Dennis Julian Elischer [EMAIL PROTECTED] wrote: Dennis George wrote: Hi all, I looking for SMT capability in freeBSD.. And found the following extract in a document... saying that 4.3 BSD has no support for SMT. does the current/latest version of freeBSCD (5.2 or 6.0) has the support for SMT ?? yes.. it's the current area of development. what do you want and I can tell you how well we support you.. It (4.3 BSD) has no support for processor affinity or binding. It also has no mechanism for distinguishing between CPUs of varying capability, which is important for SMT (Symmetric Multi-Threading). Thanks in advance... Dennis - Do you Yahoo!? Take Yahoo! Mail with you! Get it on your mobile phone. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED] __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Binding process to a fixed processor
Hi all, I am working on a intel based multi processor system. I like to know how can I bind one process permanently to one processor. and other one for general use. thanks in advance Dennis - Do you Yahoo!? Win 1 of 4,000 free domain names from Yahoo! Enter now. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Support for SMT in latest FreeBSD
Hi all, I looking for SMT capability in freeBSD.. And found the following extract in a document... saying that 4.3 BSD has no support for SMT. does the current/latest version of freeBSCD (5.2 or 6.0) has the support for SMT ?? It (4.3 BSD) has no support for processor affinity or binding. It also has no mechanism for distinguishing between CPUs of varying capability, which is important for SMT (Symmetric Multi-Threading). Thanks in advance... Dennis - Do you Yahoo!? Take Yahoo! Mail with you! Get it on your mobile phone. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Binding process to a fixed processor
Hi, I am working on freeBSD 5.2. Dennis Julian Elischer [EMAIL PROTECTED] wrote: Dennis George wrote: Hi all, I am working on a intel based multi processor system. I like to know how can I bind one process permanently to one processor. and other one for general use. thanks in advance Dennis which version of the system are you using? - Do you Yahoo!? Win 1 of 4,000 free domain names from Yahoo! Enter now. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED] - Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Finding MTU
Hi all, Can anybody tell me how to find the MTU (Maximum Transmitting Unit) in freeBSD programatically... Thanks in advance Dennis - Do you Yahoo!? Take Yahoo! Mail with you! Get it on your mobile phone. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Kernel Debugging
Hi, I am studying the kernel source of FreeBSD. I like to know the flow of packets from NIC to different modules of Kernel and then to the user-level. I studied the code and identified some of the functions through which the kernel handles network packets. But I want to check from where the control goes to that function... So I decided to debug the kernel since I have only one machine I am not checking for kgdb I decided to work with DDB so compiled the kernel with option DDB and with debugging enabled (-g option). After installing the new kernel I rebooted the system and in the boot prompt I gave -d option to enter the debugging module. My problem is that I can't provide any breakpoints with this method... It gives me error saying symbol not found. Can anybody tell me where is the problem thanks Joseph - Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage! ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Zero copy Socket
Hi all, Is there any limitations in using zero copy socket I mean... do the sender has to follow any specific protocol (like the packet payload should be page alligned or so.) ??? Thanks in advance Dennis - Do you Yahoo!? Yahoo! Mail Address AutoComplete - You start. We finish. ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Kernel Debugging
Hi, Thanks for quick response... My Vmware is running on Windows XP machine In this installed two FreeBSD5.2 vritual machines Now can you tell me how do I enable serial communication between them ? And can you tell me any funtion which is called after the call to sysinit (after which the symbols get loaded)... so that I can set a break point on that function by seeing its address by nm Thanks in advance Dennis Julian Elischer [EMAIL PROTECTED] wrote: Daniel O'Connor wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Wed, 18 Aug 2004 15:41, Dennis George wrote: I am studying the kernel source of FreeBSD. I like to know the flow of packets from NIC to different modules of Kernel and then to the user-level. I studied the code and identified some of the functions through which the kernel handles network packets. But I want to check from where the control goes to that function... So I decided to debug the kernel since I have only one machine I am not checking for kgdb I decided to work with DDB so compiled the kernel with option DDB and with debugging enabled (-g option). After installing the new kernel I rebooted the system and in the boot prompt I gave -d option to enter the debugging module. My problem is that I can't provide any breakpoints with this method... It gives me error saying symbol not found. Can anybody tell me where is the problem Maybe you should try vmware then you can pretend you have 2 machines :) Not sure how to answer your actual question though. In some releases of freebsd.. you can not see symbols at the moment that boot -d stops. for some reason they have not been enabled yet. so use nm to find the address of something that will be called and use the address to set a breakpoint there.. when it stops you will have symbols. I also would suggest vmware.. I've used it very successfully in the past.. you can use the nmdm device to set up s 'serial link' with the virtual machine and debug across that. this is from my web page: A screenshot of a FreeBSD system running under vmware, running X11, with its console output redirected out one serial port (with 'tip' on the other side of the nullmodem), and stopped in the kernel debugger, which is redirected to another serial port (actually the nullmodem driver again)with gdb (in the guise of xxgdb) showing the breakpoint and the current instruction pointer. You can also see that I was running ddb on the console port for a while too, all without disturbing the X11 display. Hopefully a more explicit description of how to do this will be forthcoming. The gif file is here. and the URL for the gif is: http://www.freebsd.org/~julian/VM-Debug_screenshot.gif - -- Daniel O'Connor software and network engineer for Genesis Software - http://www.gsoft.com.au The nice thing about standards is that there are so many of them to choose from. -- Andrew Tanenbaum GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.4 (FreeBSD) iD8DBQFBIvkh5ZPcIHs/zowRAhWKAJ90opCOwSjJVvWBTy7LhIufOzf/tgCfX+er n0sG45x5nGJez+nq1PxiZ+g= =wyB4 -END PGP SIGNATURE- ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED] - Do you Yahoo!? Yahoo! Mail is new and improved - Check it out! ___ [EMAIL PROTECTED] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
problem: FreeBSD iplog tun-device
For some reason iplog doesn't work for tun-devices I don't know why, I think it happens after forking a new process, then a resource is not available. note this happens only if I run iplog -i tun0, iplog -i xl0 for example works fine. Maybe it happens because tun-device is opened by user-ppp. But a few month ago iplog works fine with tun-device although the device was opened. I added two ifconfigs and one processtrace of iplog. --- Nipsi# ifconfig tun0 tun0: flags=8051UP,POINTOPOINT,RUNNING,MULTICAST mtu 1492 inet6 fe80::210:4bff:feb3:ecd5%tun0 prefixlen 64 scopeid 0xc inet 213.23.32.65 -- 145.253.1.164 netmask 0xff00 Opened by PID 204 xl0: flags=8843UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST mtu 1500 inet 172.16.1.1 netmask 0x broadcast 172.16.255.255 inet6 fe80::210:4bff:feb3:ecd5%xl0 prefixlen 64 scopeid 0x1 ether 00:10:4b:b3:ec:d5 media: Ethernet autoselect (100baseTX) status: active --- -- Nipsi# truss -S iplog -i tun0 __sysctl(0xbfbffad0,0x2,0x28068c08,0xbfbffacc,0x0,0x0) = 0 (0x0) mmap(0x0,32768,0x3,0x1002,-1,0x0)= 671522816 (0x2806a000) geteuid()= 0 (0x0) getuid() = 0 (0x0) getegid()= 0 (0x0) getgid() = 0 (0x0) open(/var/run/ld-elf.so.hints,0,00)= 3 (0x3) read(0x3,0xbfbffab0,0x80)= 128 (0x80) lseek(3,0x80,0) = 128 (0x80) read(0x3,0x2806e000,0x51)= 81 (0x51) close(3) = 0 (0x0) access(/usr/lib/libpcap.so.2,0)= 0 (0x0) open(/usr/lib/libpcap.so.2,0,027757775450) = 3 (0x3) fstat(3,0xbfbffaf8) = 0 (0x0) read(0x3,0xbfbfeac8,0x1000) = 4096 (0x1000) mmap(0x0,106496,0x5,0x2,3,0x0) = 67184 (0x28072000) mmap(0x2808a000,4096,0x3,0x12,3,0x17000) = 671653888 (0x2808a000) mmap(0x2808b000,4096,0x3,0x1012,-1,0x0) = 671657984 (0x2808b000) close(3) = 0 (0x0) access(/usr/lib/libc_r.so.4,0) = 0 (0x0) open(/usr/lib/libc_r.so.4,0,027757775450) = 3 (0x3) fstat(3,0xbfbffaf8) = 0 (0x0) read(0x3,0xbfbfeac8,0x1000) = 4096 (0x1000) mmap(0x0,745472,0x5,0x2,3,0x0) = 671662080 (0x2808c000) mmap(0x28122000,24576,0x3,0x12,3,0x95000)= 672276480 (0x28122000) mmap(0x28128000,106496,0x3,0x1012,-1,0x0)= 672301056 (0x28128000) close(3) = 0 (0x0) sigaction(SIGILL,0xbfbffb50,0xbfbffb38) = 0 (0x0) sigprocmask(0x1,0x0,0x28068b3c) = 0 (0x0) sigaction(SIGILL,0xbfbffb38,0x0) = 0 (0x0) getpid() = 16306 (0x3fb2) fcntl(0x0,0x3,0x0) = 2 (0x2) fcntl(0x1,0x3,0x0) = 2 (0x2) fcntl(0x2,0x3,0x0) = 2 (0x2) pipe() = 3 (0x3) fcntl(0x3,0x3,0x0) = 2 (0x2) fcntl(0x3,0x4,0x6) = 0 (0x0) fcntl(0x4,0x3,0x0) = 2 (0x2) fcntl(0x4,0x4,0x6) = 0 (0x0) readlink(/etc/malloc.conf,0xbfbff9a4,63) ERR#2 'No such file or directory' mmap(0x0,4096,0x3,0x1002,-1,0x0) = 672407552 (0x28142000) break(0x8059000) = 0 (0x0) break(0x805a000) = 0 (0x0) break(0x805b000) = 0 (0x0) break(0x805c000) = 0 (0x0) mmap(0xbfaff000,4096,0x0,0x1000,-1,0x0) = -1078988800 (0xbfaff000) gettimeofday(0x28124188,0x0) = 0 (0x0) break(0x8066000) = 0 (0x0) sigaltstack(0x2813fd40,0x0) = 0 (0x0) sigaction(SIGHUP,0x0,0x2813c5c0) = 0 (0x0) sigaction(SIGINT,0x0,0x2813c5d8) = 0 (0x0) sigaction(SIGQUIT,0x0,0x2813c5f0)= 0 (0x0) sigaction(SIGILL,0x0,0x2813c608) = 0 (0x0) sigaction(SIGTRAP,0x0,0x2813c620)= 0 (0x0) sigaction(SIGABRT,0x0,0x2813c638)= 0 (0x0) sigaction(SIGEMT,0x0,0x2813c650) = 0 (0x0) sigaction(SIGFPE,0x0,0x2813c668) = 0 (0x0) sigaction(SIGBUS,0x0,0x2813c698) = 0 (0x0) sigaction(SIGSEGV,0x0,0x2813c6b0)= 0 (0x0) sigaction(SIGSYS,0x0,0x2813c6c8) = 0 (0x0) sigaction(SIGPIPE,0x0,0x2813c6e0)= 0 (0x0) sigaction(SIGALRM,0x0,0x2813c6f8)
Re: keep-state rule for icmp, really stateful ???
Sorry I missed something ... forget - Original Message - From: Peter Pentchev [EMAIL PROTECTED] To: Dennis Berger [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Sent: Thursday, August 02, 2001 7:26 PM Subject: Re: keep-state rule for icmp, really stateful ??? On Thu, Aug 02, 2001 at 05:22:36PM +0200, Dennis Berger wrote: Hi I have the following rule allowing traceroute and ping to my server. 200 allow icmp from any to any keep-state in recv tun0 icmptype 8 Now I would assume that this rule generate two dynamic rules back. The fire one is a rule that initiates ping to work properly it's just a dynamic ICMP rule 00200 2623 220332 (T 30, # 43) ty 0 icmp, 134.100.58.115 0 - 213.23.32.88 0 and the second that the traceroute UDP taffic from port 33434-33960 can pass in. But what happans ... the rule 200 doesn't open a second dynamic rule to allow udp traffic to specific ports back in, the traceroute UDP traffic will be blocked. To keep the icmp packetfiltering stateful it would be nice to implement this clean. Or maybe it is already implemented in CURRENT tree. What's the current state ? E.. maybe it's just me, but I just can't see how a rule that says 'allow icmp' should allow UDP traffic to pass through.. Maybe you haven't shown us all the rules? (And I don't necessarily mean 'all the rules pertaining to icmp and traceroute'.. it might as well be that some other rule, which you do not consider relevant, is blocking your traceroute packets.) G'luck, Peter -- I am jealous of the first word in this sentence. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
keep-state rule for icmp, really stateful ???
Hi I have the following rule allowing traceroute and ping to my server. "200 allow icmp from any to any keep-state in recv tun0 icmptype 8" Now I wouldassume that this rule generate two dynamic rules back. The fire one is a rule that initiates ping to work properly it's just a dynamic ICMP rule 00200 2623 220332 (T 30, # 43) ty 0 icmp, 134.100.58.115 0 - 213.23.32.88 0 and the second that the traceroute UDP tafficfrom port33434-33960 can pass in. But what happans ... the rule 200 doesn't opena second dynamic rule to allow udp traffic to specific ports back in, the traceroute UDP traffic will be blocked. To keep the icmp packetfiltering stateful it would be nice to implement this clean. Or maybe it is already implemented in CURRENT tree. What's the current state ? greets Dennis
somebody implement ppp nat punch-fw ?
Hi, A function available from libalias, the PKT_ALIAS_PUNCH_FW one is currently not implemented in ppp. Maybe somebody could implement it in ppp. It's definitly not muchwork, so I request if somebody could do it. greets Dennis
Re: BSDI and Marketing 101
At 01:06 AM 04/27/2001, Mark Sergeant wrote: Then again you could also download an iso of 4.3 for free, go to BSD Mall and get a 4 CD set for 39.95 etc. The thing you pay for when you are buying these box sets is usually the manual, some also include a certain amount of telephone support etc. So while one may be 29.95 and the other another 100 dollars more you may not get a book / as good a book with the 29.95 version. Also you may or may not get phone support etc with the one you pay 100 for, I know many companies that would pay 129.95 for an OS that you get a book, support n cd's for, hell I know companies that buy Windows 98 / 2000 :P I guess you missed the POINT, which was that mandrake was selling the same LINUX functionality for $30.??? You dont get your foot in the door for $129. All you get is a return from the bookstore in 6 months. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: gcc -O bug
At 05:23 PM 04/27/2001, you wrote: In message [EMAIL PROTECTED], Dennis writes: At 05:51 PM 04/26/2001, Peter Seebach wrote: (Go ahead, dismiss me as being unfairly biased against C.) Done. Like I said, its not worthy of debate. Wow. I think that pretty much summarizes your knowledge of the situation. :) Retards often have silly grins... Shall we debate that the Ford Taurus is better than an S500 next? Some things you dont need to justify. If you dont know it, you are just a dope. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: gcc -O bug
At 09:11 AM 04/26/2001, you wrote: On Thu, Apr 26, 2001 at 11:00:15PM +1000, Greg Black wrote: Anton Berezin wrote: | Could you provide the Perl script as well? That would be pointless. The issue is with the C ... I know that. | I am quite sure it can be | made to run faster. In fact, it is almost always possible in Perl to | closely match the perfomance of a C program for this kind of | application. Nonsense (unless the C program is written by an idiot). Nope. The real nonsense is what you say. Perl core is written in a highly optimized C using very polished algorithms. As long as the Perl script is written in such a way as to minimize the number of OPCODEs executed and maximize the time spent inside the OPCODE executor engine, it is not exactly trivial to beat it in C, unless you are willing to spent a considerable time polishing your code (which is not worth it for your typical log analyzer). Don't try to argue this ridiculous point on this list. You are badly overmatched. You are so wrong that its not worthy of debate. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: gcc -O bug
At 05:51 PM 04/26/2001, Peter Seebach wrote: In message [EMAIL PROTECTED], Dennis writes: Don't try to argue this ridiculous point on this list. You are badly overmatched. You are so wrong that its not worthy of debate. Which is presumably why you offered no arguments. Actually, this is a fairly well-demonstrated result. Anything that depends mostly on the operation of, say, regexp code, and doesn't spend most of its time doing flow control will be fairly comparable in C and perl. Slower? Quite possibly. *much* slower? Not normally. I think the standing estimate is that competently-written perl will take no more than three times as long as carefully-written C for most perl-ish tasks. Matrix multiplies are an obvious exception. In practice, perl is likely to beat C substantially on most exrpession-matching code, because most C programmers write very inefficient matching code, and perl is good at it. (Go ahead, dismiss me as being unfairly biased against C.) Done. Like I said, its not worthy of debate. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
BSDI and Marketing 101
I saw BSDIs retail product of FreeBSD in a local bookstore the other day, and it became sadly clear why LINUX, although highly inferior, is so much more widely used. Right next to Freebsd (priced at $129.95) was Mandrake LINUX for $29.95. What they dont seem to realize is that people who know its worth more than linux also know they dont have to pay $129. for free software with fancy packaging and paid support. Of course they are building systems now.. db To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: BSDI and Marketing 101
At 08:12 PM 04/26/2001, David O'Brien wrote: On Thu, Apr 26, 2001 at 07:03:47PM -0400, Dennis wrote: I saw BSDIs retail product of FreeBSD in a local bookstore the other day, and it became sadly clear why LINUX, although highly inferior, is so much more widely used. Right next to Freebsd (priced at $129.95) was Mandrake LINUX for $29.95. I should use your quote from this morning... but I won't. You don't understand channel marketing. And I'll just leave it at that. Actually I do. Channel marketing requires a marketing base, which they dont have. You have to establish a base before you can gouge. I know that 0 X 129. anything X 29. do you? DB To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: BSDI and Marketing 101
At 08:24 PM 04/26/2001, George Reid wrote: On Thu, 26 Apr 2001, Dennis wrote: This is relevant to -hackers in what way? well you copied hackers, so you either think so also or you are just an ass. Learn to ignore things you dont care about, you'll live longer. I either think so w.r.t what? I didn't make a definite statement. Learn to read properly, you'll make yourself look like less of a dickhead. You asked if it was relevent, and I said you much think so also. Its not brain surgery Your need to continuously criticize me no matter how trivial the subject is very satisfying to me. Makes you look like even more of a loser than you are. db To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: SMP in 2.4 (fwd)
At 10:17 PM 04/18/2001, Rik van Riel wrote: On Wed, 18 Apr 2001, Dennis wrote: You think Intel isn't going to market dual/quad ia64 machines? Yes, but who'll need them? If nobody needed them, what would be the point in SELLING them ? I know you don't trust our technical instinct, but you might at least consider the business instinct of companies like Intel, IBM or Unisys (who all sell big SMP systems). I didnt say they shouldnt support SMP, only that complicating the OS with highly SMP-specific code to make it slightly more efficient when 99% of users dont need it is a questionable endeavor. And as for the "but you can wait 2 years until UP is faster than today's SMP" doesn't quite work for eg. investment banking and stock funds. More computing power means better calculations, which means more money. And for folks like them, computing power is not measured in FLOPS, but in ACRES. And when you're talking 3 acres of computing power, you'd better have some decend density (ie. SMP in 2U rackmounted boxes, or something similarly suitable). Your point is moot, as you already have SMP support. The question is whether squeezing a few extra cycles out (SMPng) is worth making the OS significantly more complex, particularly when more computing power is always on the way. I understand there is a language thing, but I went out of my way to say that i wasnt saying that SMP shouldnt be supported. It already is, and its been done very cleanly in a way that doesnt compromise the integrity of the OS internals. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: SMP in 2.4 (fwd)
At 01:34 PM 04/19/2001, Jeremiah Gowdy wrote: Your point is moot, as you already have SMP support. The question is whether squeezing a few extra cycles out (SMPng) is worth making the OS significantly more complex, particularly when more computing power is always on the way. Much of the code is being simplified and cleaned up. And it's not a "few extra cycles". I do admit im in a vacuum here, as I havent seen any 5.0 code. Im assuming it will be as ugly and problemattic as linux (which was unfortunately how this thread got started, but some linux moron crossposting)...and thats not fair as there are much better programmers in FBSD's camp than linux's. If its done relatively transparently, then its a big win. If it makes all of the drivers a new learning experience, then its not. db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: SMP in 2.4 (fwd)
At 01:12 PM 04/18/2001, Rik van Riel wrote: Hi, better back out SMPng real fast, otherwise you'll get into a flamewar with Dennis again ;) I just fear that "ng" will have the same negative connotations that "NT" did. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: SMP in 2.4 (fwd)
At 02:21 PM 04/18/2001, you wrote: Feel free to test it and contribute your bug reports to the developers. We make -current available for this reason, you know.. Kris Are there any sites or articles anybody knows of that desribe the differences between the 'SMPnow' and SMPng? In other words, what changes are there to ng? Is the source tree for 5.0 posted anywhere, or a spec on the driver requirements? I only see tarball chunks in the snapshots. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: SMP in 2.4 (fwd)
At 02:15 PM 04/18/2001, Kris Kennaway wrote: On Wed, Apr 18, 2001 at 01:17:03PM -0400, Dennis wrote: At 01:12 PM 04/18/2001, Rik van Riel wrote: Hi, better back out SMPng real fast, otherwise you'll get into a flamewar with Dennis again ;) I just fear that "ng" will have the same negative connotations that "NT" did. Feel free to test it and contribute your bug reports to the developers. We make -current available for this reason, you know.. Kris No thanks. I treasure these tranquil days without endless race conditions, lockups and undebuggable code. I see that the more stressful days approach. I'll stick with single processor and count on my buddies at intel to raise the bar by 75% every year without having to introduce the instability that SMPng will undoubted suffer with for long periods. A 1.5Ghz processor can outperform 2 fully saturated PCI buses, so its not going to help much in the networking world, which is where I live. Processing power is already exceeding the busses capabilities. Its nice to have a processor for user space and one for kernel/interrupt space, but going beyond that to seriously adulterate the OS to squeeze a few extra cycles in a world where processors are jumping 20% in speed every few months seems counterproductive. You dont put 2 engines in a car to make it faster, you get a faster engine. It seems that there is a lack of foresight here...you're losing a year or more of engineering time and before SMPng is stablilized the IA-64 will be out and most multiprocessor applications will be rushing to move over to that. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: SMP in 2.4 (fwd)
It seems that there is a lack of foresight here...you're losing a year or more of engineering time and before SMPng is stablilized the IA-64 will be out and most multiprocessor applications will be rushing to move over to that. I know that engaging you in conversation is a futile exercise but I'd like to point out that we've got our feet in both boats right now as well as some others that you don't mention. Ia64 is nearly working, we are taking the platform quite seriously. You think Intel isn't going to market dual/quad ia64 machines? Yes, but who'll need them? You say that "engaging me" is useless, yet you dont make a point that has anything to do with what i said. My question is, Is it worth it to tear apart the FreeBSD internals and to significantly adulterate the kernel proper with SMP-specifics to sqeeze some extra performance in the wake of the forthcoming performance boosts seems counterproductive. We dont program in assembler (much), because the extra performance isnt worth the effort. Im not saying that SMP should not be supported, only that making the OS and drivers more cumbersome for a small performance boost may be counterproductive in the long run. Db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
stupid pet tricks (using ifconfig)
Wierdness in 4.2. Scenario: interface fxp0 has address 100.1.1.1. Use it with this address for awhile. decide to change it to 100.1.1.5. do: ifconfig fxp0 delete 100.1.1.1 ifconfig fxp0 100.1.1.5 netmask 255.255.255.0 viewing ifconfig shows the new address. HOWEVER, pinging 100.1.1.99, the freebsd machine sends out 100.1.1.1, the OLD address. Is this cached/saved somewhere and not getting cleaned up? Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: [OT] Re: Intel driver doc's Take 2.
At 03:59 PM 04/04/2001, Eric Lee Green wrote: On Tue, 3 Apr 2001, Dennis wrote: * Dennis ([EMAIL PROTECTED]) [010403 18:17]: [stuff] Just a reminder: estinc.com is NOT the same domain name as etinc.com and the opinions of employees and management at Enhanced Software Technologies Inc. (estinc.com) have nothing to do with etinc.com . Just clearing up possible confusion there (due to 1 character difference in domain name). is this still going on? lol To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 06:17 PM 04/03/2001, T. William Wells wrote: Its not a "proprietary tree". I dont have time to clean it up and submit patches. But you do seem to have time to keep arguing with people??? I'm sure you'll have time to bitch again if 4.4 doesn't meet your needs because you didn't submit some patch you needed. Only because the same morons (like yourself) continue ad infinitum to post your useless comments publicly. I am opposed to supporting individuals or corporations whose principals cannot manage simple disagreements with civility. It makes it clear what the consequences will be _to me_ should I, in my capacity as consultant, ever have a dispute with such an individual or organization. I don't need that sort of stress nor would I be willing to expose my clients to such behavior. Nor do we, as a corporation, need the stress of dealing with customers with such attitudes. So it works out, doesnt it? When are you going to get it into your heads? You are not supporting me by buying what we sell. You are making a business decision, paying a price because you believe the product has at least the value of the price to you. This "consumer" attitude that you are doing a company a favor by buying something from them is completely misguided. Most companies are not some ISP or consultant struggling to pay its bills. WE are doing you a favor by making our technology available to you at a fair price. If you dont see it that way, then you have a serious problem. Because Cisco, and Intel, and 3Com and yes, Emerging Technologies will survive without your business and your attitude. We dont expect to make every sale. This is why, unless there is no reasonable alternative, your products will not be on my short-list of solutions. Ok, and unless we are totally desperate for cash (dont count on it) we wont sell anything to you. Deal? You've just made a world class business decision. Burning bridges with a vendor that you may someday need is absolutely brilliant. Now lets drop it and get back to work. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 07:37 PM 04/03/2001, you wrote: | Ok, and unless we are totally desperate for cash (dont count on it) we wont | sell anything to you. Deal? You've just made a world class business | decision. Burning bridges with a vendor that you may someday need is | absolutely brilliant. Cool, can I please go on the list of people you won't sell things to? You're already on it. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 08:24 PM 04/03/2001, T. William Wells wrote: The message referred to below was sent by me in *PRIVATE E-MAIL*. Dennis [EMAIL PROTECTED] posted that private e-mail to this public forum. He did not have my permission to do so. (But, yes, I stand by my sentiments -- I just didn't want them further cluttering this list.) sorry, everything in my hackers mailbox gets copied to the list. You didnt have permission to send me private email, so dont do it again. :-) To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: [OT] Re: Intel driver doc's Take 2.
At 07:30 PM 04/03/2001, Rick Bradley wrote: * Dennis ([EMAIL PROTECTED]) [010403 18:17]: I am opposed to supporting individuals or corporations whose principals cannot manage simple disagreements with civility. It makes it clear what the consequences will be _to me_ should I, in my capacity as consultant, ever have a dispute with such an individual or organization. I don't need that sort of stress nor would I be willing to expose my clients to such behavior. [...] This "consumer" attitude that you are doing a company a favor by buying something from them is completely misguided. Most companies are not some ISP or consultant struggling to pay its bills. WE are doing you a favor by making our technology available to you at a fair price. If you dont see it that way, then you have a serious problem. Because Cisco, and Intel, and 3Com and yes, Emerging Technologies will survive without your business and your attitude. We dont expect to make every sale. [...] Ok, and unless we are totally desperate for cash (dont count on it) we wont sell anything to you. Deal? You've just made a world class business decision. Burning bridges with a vendor that you may someday need is absolutely brilliant. Can you go ahead and put me on that list of bridge burners too? You guys are so pathetic. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 06:38 PM 03/30/2001, [EMAIL PROTECTED] wrote: Dennis writes: . My competitors probably sell twice as many boards as I do and I'll bet that I make more profit than they do. Selling more is not necessarily good. Selling more can be very bad. WHO you sell to and HOW MUCH they pay are more important. Its all about MARGIN. And you lose margin when everyone has the same information. Gold worlds! I think me as customer so I know, that my and my family life is as good as all margins are low. So I do not buy Intel's products many years and newer buy M$ - no matter how good (of bad) M$ products are. By the way, I manage a lot of purchasing (I am a consultant). And I am not alone. Several grains of sand dont make a beach. Im sure you dont drive a mercedes either, but Im sure they are doing ok without y ou. A vendor does not have to sell their product to every person in the world using that type of product to be successful. In fact M$ got in a bit of trouble trying to do so. Im not sure why you fellows dont get that. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 06:38 PM 03/30/2001, [EMAIL PROTECTED] wrote: Dennis writes: . My competitors probably sell twice as many boards as I do and I'll bet that I make more profit than they do. Selling more is not necessarily good. Selling more can be very bad. WHO you sell to and HOW MUCH they pay are more important. Its all about MARGIN. And you lose margin when everyone has the same information. Gold worlds! I think me as customer so I know, that my and my family life is as good as all margins are low. So I do not buy Intel's products many years and newer buy M$ - no matter how good (of bad) M$ products are. By the way, I manage a lot of purchasing (I am a consultant). And I am not alone. What do you Russians make? Like $90. a month? Now theres a market worth going after :-) Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 02:18 PM 03/31/2001, David O'Brien wrote: On Fri, Mar 30, 2001 at 08:49:55PM +0100, Koster, K.J. wrote: Its not a "proprietary tree". I dont have time to clean it up and submit patches. But you do seem to have time to keep arguing with people??? I'm sure you'll have time to bitch again if 4.4 doesn't meet your needs because you didn't submit some patch you needed. Only because the same morons (like yourself) continue ad infinitum to post your useless comments publicly. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 10:40 AM 03/30/2001, you wrote: On Thu, Mar 29, 2001 at 09:40:05PM -0500, Dennis wrote: Im sure that Intel is really sweating over your decision. I think they should. When people realise they are buying into something tainted and undisclosed only open to an 'elite' crowd (like dennis) then people will look elsewhere. Its only "tainted" in the minds of "source weenies". Your thinking is not mainstream. Your implication that a company with teams of marketing and legal gurus is just so naive about the freebsd market that they dont "get it" is comical. I know you dont want to hear this, but "hackers" are generally undesirable customers. They complain a lot, think they know everything, refuse to read documentation (mainly because they are used to not having any)...so "losing" their business isnt that unprofitable. do you think that the MSCE at some big company that buys our FreeBSD-based firewall/bandwidth manager complains daily about there being no GUI support for divert functions? Or that it doesnt have the latest version of ssh? Nope. They are just happy that they got what the paid for, a nice firewall that controls their bandwidth. And guess what? They also dont whine about not getting discounts because they are an "isp" or a "reseller" or a "good guy". They pay the same price and dont complain. Thats a model customer. Im 100% sure that there arent discussions in the board rooms at General Motors about using Servers with realtek ethernet controllers because intel requires an NDA for disclosure of their eepro100. Intel DOES release the info, they just dont want YOU to release it. You can write a driver, and you can sell it, you just cant give away the info. That serves the mainstream community. And thats where the money is. If you stop writing drivers for FreeBSD for intel products, guess what? Someone else will write one and make a lot of money off of it. So you are just going to create opportunities for the very people you hate, the capitalists. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 12:49 PM 03/30/2001, Peter Seebach wrote: In message [EMAIL PROTECTED], Dennis writes: [snip] Dennis, everything you're saying sounds exactly like the people who were saying, five or ten years ago, that Linux would *never* make *any* difference, because Microsoft had already won. Microsoft has won in the markets that existed at the time. Unix is more suitable for internet services, so it has a sizable chunk of the internet market. BSDi failed miserably against MS with an arguably better product in the server market. Source availability didnt help much. And lets face it. If MS had a good product, they would have a much larger market share and linux would be a non-issue. MS just makes shitty stuff. Its not about "open source". its about how well it works. I've said it 1000 times but none of the source weenies want to hear it. Linux started to make headway when it started to work well. The fact that its cheap helps too, with or without source. If there is a measurable population of people to whom open specs are important, open specs are a competitive advantage. Over time, they are likely to win if all else is equal... and in the long run, all else *is* equal. Open specs are a competitive disadvantage, because all players are equal. NetBSD will never be much better than FreeBSD (or vice-versa) because they keep stealing each others ideas and code. Those that sign an NDA have an advantage of those that dont, that the whole point. Is General Motors worried about using a card for which the drivers require an NDA? No. Is Home Depot, who are running a lot of boxes on Linux, more likely to standardize on a few thousand cards that their programmers assure them are "safer for us"? Yes. I'll bet you $.50 they use intel or 3com cards. The pressure need not be overwhelming to be real. Over time, yes, I expect to see more vendors release hardware specs, because failure to do so can cost them *at least some* sales. The number of sales seems to be steadily going up. It can be very small today and still be a big deal in five years. "some" sales dont matter. You dont understand the trade-offs, which often are negative. My competitors probably sell twice as many boards as I do and I'll bet that I make more profit than they do. Selling more is not necessarily good. Selling more can be very bad. WHO you sell to and HOW MUCH they pay are more important. Its all about MARGIN. And you lose margin when everyone has the same information. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 02:50 PM 03/30/2001, Drew Eckhardt wrote: In message [EMAIL PROTECTED], [EMAIL PROTECTED] m writes: And lets face it. If MS had a good product, they would have a much larger market share and linux would be a non-issue. MS just makes shitty stuff. Its not about "open source". Directly, it isn't. Indirectly, it is. its about how well it works. Although that comes from the software being open source. If it's open source and broken enough to affect me or my employer, This is a lot like trying to convince a jew that theres nothing wrong with cheeseburgers. Open source has its place, but its not going to take over the world. Deal with it. FreeBSD is not even better than Windows out of the box, because you need an experienced unix person to set up the box. It doesnt matter if its better, because you can't FIND experienced unix people. The economics of open source will not allow it to dominate, because there arent enough good programmers to make it work. Im arguing with a guy from "poohsticks.org". What am i thinking? lol db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 11:01 AM 03/29/2001, you wrote: On Wed, 28 Mar 2001, Dennis wrote: At 04:22 PM 03/28/2001, Chistopher S. Weimann wrote: On Wed, Mar 14, 2001 at 12:33:21PM -0500, Dennis wrote: Your logic is backwards. You think that rewarding mediocre companies will scare good companies into wanting a piece of the pie. The only thing that it will do is consume these companies so that the good companies can have a larger share of the more profitable sun/NT market, and convince them that they want no part of the "free" market if they have to compete with cut-rate hardware from hungry companies. Ok, let me get this Free Market thing straight. Not buying from a good company that provides a useful product and instead buying from a bad company that doesn't provide a useful product will make things better. That seems to be what you are saying dennis. No, I said just the opposite. This was in response to someone suggesting that we boycott companies like Intel for not providing full disclosure on their boards, and reward companies that do by touting their products. So I said that promoting lesser products because they are "cooperative" will make good hardware less available to the freebsd community, which might make some little people feel powerful but it wont serve the user base, which I assume is the goal. You seem to keep inferring that all vendors who disclose full programming information somehow have "lesser" hardware. [other trivial stuff snipped] I think that boycotting Intel and 3Com says enough to dispute your argument. The possible fact that some "good hardware" is disclosed doesnt make for good counterpoint. Generally, companies that "just crank out hardware" disclose their hardware specs., and releasing source is a last ditch effort when a company finds its software not good enough to sell. Value added vendors dont release such things, and value-added vendors tend to be the more dominant vendors. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 07:20 PM 03/29/2001, Soren Kristensen wrote: Sorry everybody, I have to express my opinion now. Dennis, it seems like that you keep repeating yourself here And you keep being wrong. As a hardware designer myself, I can assure you that there is no connection between hardware quality and level of documentation. And having fought with Intel getting documentation myself on other parts, I have long time ago decided to only use Intel parts if I really have to, just like more and more engineers decides. Im sure that Intel is really sweating over your decision. Using yourself as an example was brilliant. YOU are exactly the kind of engineer/company that I am talking about. YOU will release your design specs so the FreeBSD community will get to use your hardware instead of intels. Wooopee do. Most other vendors, incl the best ones, don't have any problem releasing full documentation. 3com never has, nor has intel,and they generally are considered top vendors. Still zero valid points. Dlink and realtek are clone vendors, as are kingston and most others. The 2 market leaders dont release their docs. Intel just don't get it, they had had that arrogant attitude for years, and now that they don't own the PC market anymore, they're going to pay for it. You, my friend, are a joke. Im sorry (other list members)...do you have any clue how much money intel makes, and how insignificant your (wrong) opinion is to them? and, while I dont have the actual numbers, I'll bet that intel and 3com have well over 50% of the linux/BSD free software market without releasing their docs. Why? because they make good, cheap boards that are available from many sources worldwide. do you want to take that away from BSD users by "boycotting" their hardware? why cant this thread just die? db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: # of bpf devices
At 01:32 PM 03/28/2001, Poul-Henning Kamp wrote: In message [EMAIL PROTECTED], Dennis writes: It seems that only 256 bpf devices are supported. How painful would it be to increase that number...I assume its an 8bit varable somewhere? Are there other caveats? It's pretty trivial. Send a patch when you are done. I was hoping to get some useful insight before I looked into it..or if there was a sound reason for not expanding them. Im hopeful someone brighter will answer. dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: # of bpf devices
At 03:17 PM 03/28/2001, David O'Brien wrote: On Wed, Mar 28, 2001 at 02:49:53PM -0500, Dennis wrote: It seems that only 256 bpf devices are supported. How painful would it be to increase that number...I assume its an 8bit varable somewhere? Are there other caveats? It's pretty trivial. Send a patch when you are done. I was hoping to get some useful insight before I looked into it..or if there was a sound reason for not expanding them. Im hopeful someone brighter will answer. You did. "pretty trival". I.E. will not take much work, and no good reason not to expand them. Since you've said so many times you can hack your own system, sounds like PHK told you what you needed to know. What a helpful bunch. thanks. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 04:22 PM 03/28/2001, Chistopher S. Weimann wrote: On Wed, Mar 14, 2001 at 12:33:21PM -0500, Dennis wrote: Your logic is backwards. You think that rewarding mediocre companies will scare good companies into wanting a piece of the pie. The only thing that it will do is consume these companies so that the good companies can have a larger share of the more profitable sun/NT market, and convince them that they want no part of the "free" market if they have to compete with cut-rate hardware from hungry companies. Ok, let me get this Free Market thing straight. Not buying from a good company that provides a useful product and instead buying from a bad company that doesn't provide a useful product will make things better. That seems to be what you are saying dennis. No, I said just the opposite. This was in response to someone suggesting that we boycott companies like Intel for not providing full disclosure on their boards, and reward companies that do by touting their products. So I said that promoting lesser products because they are "cooperative" will make good hardware less available to the freebsd community, which might make some little people feel powerful but it wont serve the user base, which I assume is the goal. Get it? Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: # of bpf devices
At 04:56 PM 03/28/2001, Mike Smith wrote: It seems that only 256 bpf devices are supported. How painful would it be to increase that number...I assume its an 8bit varable somewhere? Are there other caveats? It's pretty trivial. Send a patch when you are done. I was hoping to get some useful insight before I looked into it..or if there was a sound reason for not expanding them. Im hopeful someone brighter will answer. You did. "pretty trival". I.E. will not take much work, and no good reason not to expand them. Since you've said so many times you can hack your own system, sounds like PHK told you what you needed to know. What a helpful bunch. thanks. Just ignore them. I know it's hard, but once you get used to it, it'll hurt less. Trust me on this one. 8) it doesnt "hurt" at all. Dealing with bitter losers is part of the public experience :-) Thanks for the tip. i'll forward it to the customer who needs it and let him do the work. I've got some more flames to deflect :-) Dennis Anyway, I just had a quick look, and I think that your basic problem is that MAKEDEV uses the wrong encoding for devices above 255. This is fixed in -CURRENT, and if you bring back the unit2minor changes from there to -STABLE you should be in business. If this works, please file a PR so that it gets fixed. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
# of bpf devices
It seems that only 256 bpf devices are supported. How painful would it be to increase that number...I assume its an 8bit varable somewhere? Are there other caveats? VPNs and extensive frame relay setups with DHCP require more than 256 devices. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: Porting a Linux driver to FreeBSD with ioctl return values
At 04:05 AM 03/27/2001, Devin Butterfield wrote: On Monday 26 March 2001 11:24, Alfred Perlstein wrote: * Coleman Kane [EMAIL PROTECTED] [010326 22:40] wrote: Yeah, that's basically what I had to do in tdfx. You can take a look int src/sys/dev/tdfx/tdfx_pci.c under tdfx_ioctl(...) to get an idea of what needs to be done, if you need more info. Tdfx basically implements the API from device_3dfx in Linux. Is there anyone we can PLEAD with to explain that to the linnux people that that's a broken way to implement ioctl()? Everything in LINUX is implemented in a broken way. Whats interesting is that they like it that way, and even try to justify it. skbufs are really neat. no chaining, so if you are doing encapsulations you get to reallocate the entire buffer and copy overs tons of "attached info". Very high performance. the biggest caveat with linux ioctls() is that there are only 10 commands available. So you are best off only using 1 and using sub-functions within your application, or you may run out. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Porting a Linux driver to FreeBSD with ioctl return values
At 04:05 AM 03/27/2001, Devin Butterfield wrote: On Monday 26 March 2001 11:24, Alfred Perlstein wrote: * Coleman Kane [EMAIL PROTECTED] [010326 22:40] wrote: Yeah, that's basically what I had to do in tdfx. You can take a look int src/sys/dev/tdfx/tdfx_pci.c under tdfx_ioctl(...) to get an idea of what needs to be done, if you need more info. Tdfx basically implements the API from device_3dfx in Linux. Is there anyone we can PLEAD with to explain that to the linnux people that that's a broken way to implement ioctl()? oh, also, they MUST be negative. so you return (-ENODEV)...some bad upper layer code they decided to never fix. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: AW: Best Gigabit ethernet for 4.x
Personally I hate the attitude and policies expressed by the NDA on chips that are long out of development. I can understand it for 'new on the block' designs, as that is how competition works. But once a chip has moved from the prototype sample mode into full production, I think the chip manufactures should publish publically on the web (where it is almost 'cost less' to do so) for all implementers to have the information. As it is, it seems in terms of Intel and other chip manufacturers more profitable to make 'strategic' business partnerships with 'big software houses' (for example, buying a stake in LynxOS now LynxOS Works and Blue Cat Linux), than to let the world have a crack at the information. In order to "secure" those "big contracts" they have to give relative assurances that the resulting boards wont be cloned, allowing others to cash in on extensive software and marketing costs of these companies. Theres no sense spending big money to establish a market if anyone can come along and take away your margins by selling cut-rate hardware. Hopefully SOME of you understand this. Its not intel worried about their chips being cloned (they have a team of international lawyers to police that), its their partners that they are protecting. At some point some of you will get it. I think the best strategy is to get intel to have a freebsd driver (as they do for linux)...which would do more for freebsd than boycotting such products as some of you have suggested. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
RE: Good server motherboard?
At 12:42 PM 03/27/2001, you wrote: On Tue, 27 Mar 2001, Ed Henderson wrote: You mention that I should consider software RAID if I only need level 1. How much processing overhead is there? My system will be running with a single 1.2GHz Athlon CPU. Software RAID like Vinum has proven to be faster then HW RAID. Mainly because of HW Raid card's using poorly sized stripes. There is a benchmark test of Vinum vs some DPT Cards that brad knowles put on the web. If you want the URL I would be happy to dig it up for you. Vinum is FAST++, and you can enable softupdates on Vinum as well. So you get speed and reliability. And I use this combo on alot of Postfix mail servers for the mail queue. I couldn't be happier. Vinum is incredible. Now I just have to keep working on some decent doc's for it :-) Will the machine boot off of the second drive if the first one fails? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: old business (was Re: Intel driver doc's Take 2.)
At 12:50 PM 03/25/2001, Matthew Jacob wrote: On Sat, 24 Mar 2001, Dennis wrote: If the if_wx driver sucks, why not fix it rather than trying to coerce a mega-companies with a deep political structure to change is policies? But if youre not going to maintain it, dont do it at all. You cant stick it to users by deciding later that you dont want to support it anymore. I am going to fix it (but this has been low on my priority list- you got the bucks to pay for this, buddy, and make it more attractive then the other things I'm currently getting paid to work on? money talks...), and I *also* am trying to change Intel's policy. If I pay for it, I own it. Thats the caveat of "open-source" db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 07:47 PM 03/24/2001, you wrote: On 24 Mar 2001, at 19:59, Dennis wrote: the only thing more annoying the 2 people having a discussion is a third person telling them to stop. Feel free not to read any more messages in this thread. Feel free to read the list charter. You two are in a pissing contest unreleated to this list. Please take it elsewhere. We are discussing the issue. TOO BAD if you dont like it. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: ipchains ported to FreeBSD
At 04:33 AM 03/01/2001, mouss wrote: At 22:20 28/02/01 -0600, Dan Nelson wrote: In the last episode (Mar 01), jett tayer said: can ipchains / iptables be ported to FreeBSD... this is a suggestion if u dont mind. We've already got ipfw and ipfilter; why in the world would we need a third packet-filtering systam? :) add to this that ipchains will certainly be replaced by iptables! at least there is a GUI for ipchains (albeit a lousy one). Is there one for ipfilter anywhere? Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: AW: Best Gigabit ethernet for 4.x
At 05:04 PM 11/18/2000, Kenneth D. Merry wrote: On Sat, Nov 18, 2000 at 11:33:29 -0500, Dennis wrote: At 04:28 PM 11/17/2000, Schmalzbauer, Harald wrote: I just heard that Intel doesn't supply documentation on ther chipset and the FreeBSD and Linux support is quiet bad. The Netgear GA620 is said to be twice as fast. The same Chipset (Alteon Tigon/AceNIC) is on the 3com985. Are all of the cards supported that use this chipset? I read somewhere that the netgear card has a smallish buffer, and that the alteon was a better choice. How does the 3com card compare in that respect? The Netgear boards have 512K SRAM, the 3Com boards have 1MB SRAM. You can get Alteon-branded boards (with either 512K or 1MB SRAM), but generally only directly from Alteon, and you're going to pay more than you would for either the 3Com or Netgear boards. The 3Com and Netgear boards are identical to the Alteon boards. The only difference is they've got "Netgear" or "3Com" silk-screened on them, and the Alteon boards don't have any logos on them. FWIW, 3Com is buying Alteon's NIC group. Apparantly (according to an Alteon engineer who posted on the linux-acenic list) they're just buying the technology, not hiring the engineers: When you say "identical", that implies that they are the same...or do you just mean that they use the same parts? It seems unlikely that alteon would allow netgear to license its product and them sell it for 1/2 the price. Price is not an issue at this level as performance is tantamount, but Im not sure you'd need more than 512M with unix unless you have several NICs in a box. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 01:33 PM 03/24/2001, [EMAIL PROTECTED] wrote: On Sat, 24 Mar 2001, Luigi Rizzo wrote: I have read the thread for a while, and i wonder: why in the world someone should go through the effort and responsibility of SIGNING THE NDA _and_ negotiating with Intel for getting permissions to redistribute the code ? sleep deprived venting I made the effort to try and work things out for users like dennis. Who constantly have problems with Intel changing their PHY's, and our driver not getting updated. Because Intel wont give doc's out without an NDA. Now that should tell people like dennis, that our developers are *not interested* in writing binary only drivers, and/or signing NDA's. You use the term "our developers" as if you are some sort of closed cult. I have NEVER complained about Intel not releasing full information on their boards.That whining has come exclusively from hackers that didnt feel like doing the work. I complained about FreeBSD having a "maintainer" for a very important driver who didnt do any maintaining. I fixed the persistent PHY problem in an afternoon with info from the available linux driver supplied by intel. You dont need to get intel to change its policy to support the board. Its just an excuse to not do it. There are plenty of resources available. If the if_wx driver sucks, why not fix it rather than trying to coerce a mega-companies with a deep political structure to change is policies? But if youre not going to maintain it, dont do it at all. You cant stick it to users by deciding later that you dont want to support it anymore. I complained for days and then fixed the fxp driver in about 4 hours. Maybe its time to do work and complain less. dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
If the if_wx driver sucks, why not fix it rather than trying to coerce a mega-companies with a deep political structure to change is policies? But if youre not going to maintain it, dont do it at all. You cant stick it to users by deciding later that you dont want to support it anymore. You can't "fix" a device driver "correctly" when you don't have the right docs. Most drivers are written without full docs. Intel supplied drivers for linux are available for both eepro100 and gigabit cards. The info is out there. Cobbling together the info to produce a driver...THATS what open source is all about. I complained for days and then fixed the fxp driver in about 4 hours. Maybe its time to do work and complain less. Uh bro, if you have patches, submit them! That's what opensource is all about... of course you knew that too. Im not an "open-source" developer. but then again, you knew that. Plus I offered to help and J. Lemon bit off my head. Plus Im sure you wouldnt want Intels copyright in "your" driver. db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 02:45 PM 03/24/2001, [EMAIL PROTECTED] wrote: On Sat, 24 Mar 2001, Richard Hodges wrote: Thanks for the tip on the XL driver. Over the years, I have seen many questions about which driver/card was the best, and all the answers were pretty vague, or suggested that Intel had the edge. Yes. It's currently my understanding that the xl driver is indeed as reliable and as fast as the fxp. I know what your saying though. The fxp has been touted as the fastest. Years ago it was the de driver for DEC cards that was the screamer, then it became the fxp, but the xl driver is just as fast. thats not true at all. We have traced many of our heavily loaded customers problems to the XL driver. Replacing them with intels solved the problems. But for general use you are probably correct. Db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 03:12 PM 03/24/2001, Will Andrews wrote: On Sat, Mar 24, 2001 at 03:11:54PM -0500, Dennis wrote: Most drivers are written without full docs. Intel supplied drivers for linux are available for both eepro100 and gigabit cards. The info is out there. Cobbling together the info to produce a driver...THATS what open source is all about. [...] Im not an "open-source" developer. but then again, you knew that. Plus I If you're not an opensource developer, then you don't know what yer talking about. Why do you even bother subscribing to any freebsd lists if you aren't going to adopt any sort of opensource behavior? Is being an "open-source" developer now a necessary condition for being a "hacker" or "developer"? You academics crack me up. You have it completely backwards. Most open-source people dont know what they are talking about. This conversation about intel is like a bunch of women discussing what its like to be hit in the balls. its downright comical. They are in the business of making money. If you think they dont know the exact impact of releasing the information then you need to get out of your cubical a hell of a lot more. For your info, Bub, what makes the BSD license attractive is its usability by commercial vendors, so maybe you should go play in Linuxland because you are the one in the wrong camp, not me. the ability to take code, fix it and incorporate it into a product WITHOUT giving back source is the ENTIRE CONCEPT of the BSD license. And why does all of your email have that stupid attachment? Whats the matter, cant figure out how to use an open-source mailer? :-) DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 04:19 PM 03/24/2001, Will Andrews wrote: On Sat, Mar 24, 2001 at 04:12:34PM -0500, Dennis wrote: For your info, Bub, what makes the BSD license attractive is its usability by commercial vendors, so maybe you should go play in Linuxland because you are the one in the wrong camp, not me. the ability to take code, fix it and incorporate it into a product WITHOUT giving back source is the ENTIRE CONCEPT of the BSD license. Obviously so, but it serves no point to complain about an opensource driver's problem on an opensource list if you're going to fix it in your proprietary tree, say you did so, and not pass it on. Since nobody else gets to see your "fix", it won't solve a thing, for you or FreeBSD. Its not a "proprietary tree". I dont have time to clean it up and submit patches. "hackers" doesnt imply "open-souce". This is a "general technical discussion" list, according to freebsd.org. db To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Intel driver doc's Take 2.
At 04:07 PM 03/24/2001, Dan Langille wrote: On 24 Mar 2001, at 16:12, Dennis wrote: And why does all of your email have that stupid attachment? Whats the matter, cant figure out how to use an open-source mailer? :-) It's called a PGP signature. Could you two kids please take this pissing contest off -hackers? Thanks. the only thing more annoying the 2 people having a discussion is a third person telling them to stop. Feel free not to read any more messages in this thread. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Routing latency
At 02:43 AM 03/20/2001, you wrote: I'm using the de driver. Alas, the NICs seems quite old. They are 21140's. I've only got one 21143. I think there is a 3COM 3c905b in the lab too. Would it be better to use the 21143 + 3com than two 21140s? definitely : in my packet blaster, I get an order of magnitude less packet drops with a 3c905 than with a dc NIC (which is on a multi-port NIC : the PCI-PCI bridge may be a hindrance there) not my experience -- with the 21143 i can blast 140kpacket/s and receive them with no problems. For sure the "de" driver might have its own problems, but i think a lot of packet drops also depend on the card not being properly set for full duplex (which can cause collisions and lots of drops). You should initially test mono-directional in a controlled environment to avoid "collisions" to compare the true efficiency of the driver. dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
RE: Routing latency
At 02:04 AM 03/20/2001, Mrten Wikstrm wrote: [snip] triggers every second and steals too much cpu. So my question is, how can I decrease this routing delay? Were you loading the interface, or just passing nominal streams? What pps did you pass through the box? Most likely the "delays" are only seen when the machine is close to capacity (the slow CPU you are using doesnt help). I sent 2 packets/s, three UDP streams with 60, 200 and 1000 bytes sized packets respectively. I also tried just one stream with 60 bytes packets and the same behaviour occured. 20k pps is probably beyond the capacity of a 200Mhz PPRO machine to forward on an ongoing basis (ie if other processes are running at the machine). the way the machine behaves over capacity is not as important as its abiltiy to continue running. How it works during normal operations its what is important. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Routing latency
At 02:32 PM 03/19/2001, Thierry Herbelot wrote: Hello, the FreeBSD TCP/IP stack uses the "system tick timer" for some delay (maybe only for TCP). you may want to use a HZ=1000 option (see the LINT config file) in a recompiled kernel and see if things go better. (moreover, the dc(4) driver which is used for your NIC has some interesting performance improvements in the forthcoming 4.3-Release) TfH Cool. Is the 21143 now started in store-and-forward mode and has the mandatory watchdog timeout been fixed? Im getting tired of hacking it every release. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Routing latency
At 09:22 AM 03/19/2001, Mrten Wikstrm wrote: I've performed a routing test between a FreeBSD box and a Linux box. I measured the latency and the result was not what I had expected. Both systems had the peak at 100 us (microseconds), but whereas the Linux box had _no_ packet over 200 us, the FreeBSD box delayed some packets up to 2 ms! Looking at the time series, it seems that the packets are delayed at regular intervals, about every second. My guess is that some timer interrupt triggers every second and steals too much cpu. So my question is, how can I decrease this routing delay? Were you loading the interface, or just passing nominal streams? What pps did you pass through the box? Most likely the "delays" are only seen when the machine is close to capacity (the slow CPU you are using doesnt help). Latency under load and general latency are very different. Differing methods of handling backup conditions may have different goals; the proper goal is overall stability and NOT packet efficiency. It doesnt matter how fast a man runs if he doesnt finish the race. The problem with LINUX is that it works to a point and then chokes, while freebsd works up to higher thresholds. You cant evaluate a subsystem with one somewhat bogus test, without looking at the system as a whole. If you are using the dc driver, make certain it is operating in store-and-forward mode, the default configuration starts in a mode that only works on 10mb/s connections. dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Routing latency
At 07:20 PM 03/19/2001, Will Andrews wrote: On Mon, Mar 19, 2001 at 07:14:54PM -0500, Dennis wrote: Cool. Is the 21143 now started in store-and-forward mode and has the mandatory watchdog timeout been fixed? Im getting tired of hacking it every release. Submit a PR to fix the problem? I never got an answer (as usual) from bill paul when I made the suggestions, and noone seemed interested in getting it fixed. He seems to get insulted when I infer that he did something wrong. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 06:35 AM 03/14/2001, Jeroen Ruigrok/Asmodai wrote: -On [20010310 01:00], Lyndon Nerenberg ([EMAIL PROTECTED]) wrote: Perhaps a first step towards leaning on the vendors for documentation is to publically declare our support for those vendors who *do* release documentation under reasonable terms. One way to do this is to acknowledge those vendors in the hardware section of the handbook, and encourage people to support them by buying thier products. Typically companies that are quick to release docs are the weaker companies, because they need sales or dont have a clear target market. Plus the best technologies are usually proprietary at least in the beginning of their deployment. So your strategy will guarantee alignment with many mediocre products and few of the best, which doesnt seem to be in the best interests of anyone. Telling people that they cant use Intel or 3com cards will more likely drive them to other OSes then hurt either of the fore mentioned companies. A better strategy would be to welcome ALL vendors AND binary distributions (who may release source under NDA if they chose) so that the best products could be available for FreeBSD without the adversity of the "geek-revolution" that you propose. That's what Soeren and me did. HighPoint was very forthcoming with documentation and as part of that synergy they put the FreeBSD Hardware logo on their frontpage http://www.highpoint-tech.com and us mentioning it on our webpages as well as being a primary recommendation over other vendors back when we got little information out of other vendors. I'm sure that High-point's competitors are shivering in their boots over this powerful alliance. DB Emerging Technologies, Inc. - http://www.etinc.com ISA and PCI T1/T3/V35/HSSI Cards for FreeBSD and LINUX Multiport T1 and HSSI/T3 UNIX-based Routers Bandwidth Management Standalone Systems Bandwidth Management software for LINUX and FreeBSD To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 11:32 AM 03/14/2001, you wrote: Dennis wrote: At 06:35 AM 03/14/2001, Jeroen Ruigrok/Asmodai wrote: -On [20010310 01:00], Lyndon Nerenberg ([EMAIL PROTECTED]) wrote: Perhaps a first step towards leaning on the vendors for documentation is to publically declare our support for those vendors who *do* release documentation under reasonable terms. One way to do this is to acknowledge those vendors in the hardware section of the handbook, and encourage people to support them by buying thier products. Typically companies that are quick to release docs are the weaker companies, because they need sales or dont have a clear target market. Plus the best technologies are usually proprietary at least in the beginning of their deployment. So your strategy will guarantee alignment with many mediocre products and few of the best, which doesnt seem to be in the best interests of anyone. Telling people that they cant use Intel or 3com cards will more likely drive them to other OSes then hurt either of the fore mentioned companies. A better strategy would be to welcome ALL vendors AND binary distributions (who may release source under NDA if they chose) so that the best products could be available for FreeBSD without the adversity of the "geek-revolution" that you propose. I think you underestimate the number of faceless servers thoughout the world running FreeBSD or Linux. The jobs they do aren't glamourous, but they have to be done (cheap), so they don't get the front page accolades that Sun UE10ks get, but they do get used. My point is that it will have no impact, so you will only hurt the FreeBSD community. You cant strong-arm companies into making their intellectual properly rights publicly available. its a losing argument. That's what Soeren and me did. HighPoint was very forthcoming with documentation and as part of that synergy they put the FreeBSD Hardware logo on their frontpage http://www.highpoint-tech.com and us mentioning it on our webpages as well as being a primary recommendation over other vendors back when we got little information out of other vendors. I'm sure that High-point's competitors are shivering in their boots over this powerful alliance. Probably not, but they might wonder why High-points sales are stronger than they should be, and why there is so much positive material about High-point cards on the web (while their own cards are barely mentioned). People who want to get work done don't want to mess with a company that tries to stop them (by not releasing specs or drivers for the OS you're using) and will instead go with the open-minded competetor. selling to geeks is not most companies marketing strategy. What you fail to understand is the negative impact on sales when some taiwanese company clones the hardware and you effectively end up cannibalizing your own business with your efforts. Your also just as likely to get negative press because the guy that writes the driver for your hardware does a lousy job, and the resulting driver sucks and people then think your hardware sucks because most geeks can't separate the hardware from the driver. Your logic is backwards. You think that rewarding mediocre companies will scare good companies into wanting a piece of the pie. The only thing that it will do is consume these companies so that the good companies can have a larger share of the more profitable sun/NT market, and convince them that they want no part of the "free" market if they have to compete with cut-rate hardware from hungry companies. Driving away companies with good products because you dont like their policies is counterproductive. the only reason people use windows is because of their relationships with vendors who sell products that people want. its not about the OS, its about what you can do with it. dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 02:31 PM 03/14/2001, Jeroen Ruigrok/Asmodai wrote: -On [20010314 17:38], Dennis ([EMAIL PROTECTED]) wrote: At 06:35 AM 03/14/2001, Jeroen Ruigrok/Asmodai wrote: That's what Soeren and me did. HighPoint was very forthcoming with documentation and as part of that synergy they put the FreeBSD Hardware logo on their frontpage http://www.highpoint-tech.com and us mentioning it on our webpages as well as being a primary recommendation over other vendors back when we got little information out of other vendors. I'm sure that High-point's competitors are shivering in their boots over this powerful alliance. Well Dennis, I congratulate you. Be assured that with this attitude you just displayed you made me decide never to recommend ET Inc., for any of my present and future projects. Clearly you dont get it. Its like teaching a fish to fly. LOL To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 01:47 PM 03/14/2001, you wrote: On Wed, Mar 14, 2001 at 09:09:15AM -0800, Alfred Perlstein wrote: how many times does windows crash because of poorly written drivers rather than flaws in the core OS? (*) ALL the time. Microsoft has given the UC-Davis security and formal verification lab a multi-year grant to look at this problem. (the approach being researched is "model checking") Why would they need to do that? Every time you load a program it updates the libraries, breaking older programs. Its a philosophical problem. You dont need a grant to figure it out. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 02:31 PM 03/14/2001, Jeroen Ruigrok/Asmodai wrote: -On [20010314 17:38], Dennis ([EMAIL PROTECTED]) wrote: At 06:35 AM 03/14/2001, Jeroen Ruigrok/Asmodai wrote: That's what Soeren and me did. HighPoint was very forthcoming with documentation and as part of that synergy they put the FreeBSD Hardware logo on their frontpage http://www.highpoint-tech.com and us mentioning it on our webpages as well as being a primary recommendation over other vendors back when we got little information out of other vendors. I'm sure that High-point's competitors are shivering in their boots over this powerful alliance. Well Dennis, I congratulate you. Be assured that with this attitude you just displayed you made me decide never to recommend ET Inc., for any of my present and future projects. Im sure that we will survive quite nicely without your recommendations. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 10:37 PM 03/10/2001, Peter Seebach wrote: If anyone has a specific part number or model information about the new unsupported PHY, I'd be happy to look it up and tell you what, if anything, I can find out. I can't send out copies of the source without some kind of formal approval, but I could certainly at least answer questions like "do we have a BSD-flavored driver that works with this". As it turns out (as usual), its not an "unsupported PHY" but an error in the assumption that the correct PHY information is where DG's logic thinks it should be in the eeprom. Reading the PHY info from the part directly allows you to correctly identify and set up the phy, at least on the SuperMicro MB that I was having problems with. For those with boards that give the "unsupported PHY" message who want a "quick fix", you might try just forcing the address to 1 and the device_type to 7, as every eepro100 that I've tested uses the 82555 PHY. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: non-working fxp cards
At 02:45 PM 03/12/2001, Jonathan Lemon wrote: I would like anyone who has a fxp card which doesn't work with the current driver to contact me in order to test out an alternate driver. -- Jonathan in case you havent read my posts, I've fixed the problem with mine. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: non-working fxp cards
At 04:44 PM 03/12/2001, Jonathan Lemon wrote: On Mon, Mar 12, 2001 at 04:44:38PM -0500, Dennis wrote: At 02:45 PM 03/12/2001, Jonathan Lemon wrote: I would like anyone who has a fxp card which doesn't work with the current driver to contact me in order to test out an alternate driver. -- Jonathan in case you havent read my posts, I've fixed the problem with mine. Glad to hear it. In that case, I expect you _NOT_ to use my new driver. And why is that? I thought you just might like some guidance. Feel free to beat on it on your own. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 01:11 AM 03/10/2001, Bill Paul wrote: I think its been mentioned several times in this and other threads that intel has a driver for LINUX that is effective documentation on the board, and the code is public (although you may have to stick an intel copyright in the code also). Whoever mentioned this was not thinking clearly. A manual is effective documentation for a NIC. Sample driver code alone is not. It's handy, but it's not enough. When you write a driver, you make certain design decisions based on the information in the manual and the OS you're developing for. By forcing someone to rely soley on your driver to see how the board works, you're limiting their ability to make their own design decisions. What works well for Windows or Linux may be mediocre for BSD. Besides, Intel engineers have a knack for choosing really confusing register names. confusing or not, the logic to fix the driver is available. You can whine about there being "no full documentation", but guess what? FreeBSD doesnt have "full documentation" either. Do any of your drivers have "full documentation" for anyone that want to modify them? Fixing Greenman's driver is no party either as he hasnt documented any of the phy-related stuff he uses. And again, saying "but there's a Linux driver" just gives vendors an excuse to perpetuate their stupidity. I'm not keen to give them this opportunity. You "guys" regularly say to me "you have the source, fix it". Source that works IS documentation. As someone whose read a few controller specs in my time, I can tell you that "full documentation" is sometimes a lot less useful than code that works, because the docs dont always make it clear what needs to be done to achieve a certain goal. You guys continue not to understand why companies dont disclose board info freely. You end up competing with your own customers. They dont want people buying gray market parts and selling $9. boards. Its very easy to clone a board with 2 chips on it these days. I'm sorry, that doesn't wash. *I* am not trying to compete with anyone. Lord knows I can't afford to fabricate my own controller chips on my salary. its not about you, man, its about the clone manufacturers that can make cards that use your or intel's drivers without any engineering. You dont have to "fabricate chips", you buy them from Intel. Thats what I mean by "competing with your own customers". Intel sells chips for $8. and boards for $32. they odnt want to have to compete with boards that sell for $12. with their $8. chips. Your lame argument might be that "they sell the chips anyway" , but that doesnt work, becuase they make money on the boards at 32 and dont at 12. Notice that there arent really any Intel eepro100 clones? Because intel makes sure that the spec isnt public, so they can go after anyone that clones them. Western Digital in the 80s learned the hard way. You do all the marketing, get software written for your cards, and then the taiwanese cash in on it and you have to lower your prices to where you cant make enough money to get back your marketing outlay. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
The tulip cards can be quirky, if nothing else. I used to like the VIA Rhine cards, because they were cheap, and I had no problems with them... until suddenly they started crashing at 100Mbps. I don't know why; I ran some of them under very heavy loads at 100Mbps. I can't tell whether it was new cards or a driver change. Cards generally arent "quirky"; drivers are incomplete. Its all about the software. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 01:49 PM 03/10/2001, Romain Kang wrote: As a newcomer to this, I'm a little confused. There's a slew of datasheets at Intel's web site http://www.intel.com/design/network/datashts/index.htm that don't seem to require NDA. (Just this week, I used the 82559 docs to implement a polled version of if_fxp). Is anyone up on the latest legal stuff? There was a ruling that universities cant be held liable for releasing NDA informationuniversities and states I think. I know we couldnt sell source to universities or the government because it wasnt protected (ie you couldnt sue them if it leaked out). Maybe we can get an academic to sign something and leak out the info :-) Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 03:37 PM 03/10/2001, Tim wrote: Is there a point of contact at Intel that we could all send e-mail to or even send a formal letter? I am sure my buying 50 or so boards a year isn't going to make a dent at Intel's bottom line, but considering how their stock is doing lately and if we all contribute... Um, I dont think your going to get intel to change its policy The policy exists for a reason. Plus, they are already selling lots of boards with the existing driver, so how many more will they sell? the delta isnt enough for them to put someone on the case. What will really get them is some press on how bad their drivers are. Their eepro100 driver for linux is unusable under load. and from what I've hard their gigabit driver isnt much better. Letters dont do anything. its all about image. Dennis Thanks, Tim To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: call for testers: port aggregation netgraph module
At 03:42 PM 03/10/2001, Justin T. Gibbs wrote: Each link is checked once every second to see if the link is still up. An attempt to send a packet over a dead link will cause the packet to be shifted over to the next link in the bundle. Any chance this can be done through an async event rather than by polling? I've been meaning to ask about this...is there a reason that ethernet drivers dont call if_up and if_down like serial drivers on cable events? This is needed for load balancing so that the UP flag can be used instead of polling or an event. Of course a polling protocol is needed also. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
You've got a valid problem. Go away. "You've got a valid problem, go away." huh?? His points are very valid about maintenance of the `fxp' driver. His views on how to make something happen are what is a little out of touch. I tried sitting with my hands folded. It didnt work. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
Call BSDi (numbers on their website) and ask to speak to Gary Johnson (CEO) or Mark Garver (senior VP) and ask them yourself. If you get a useful answer, please tell the rest of us; especially me, since I burnt out trying to make it happen. Ive spoken with Mr Johnson several times. hes the biggest reason that BSDI is what it is. Amazing that hes still there. He has no concept of what the market wants. First he wanted to challenge microsoft (at similar prices) for the server market now he wants to be redhat. funny guy. If you're unhappy with the results you're getting from Gary, try calling Kirk McKusick. Keep kicking people. Wait. Isn't "kicking people" what I'm getting beat up for in this thread? I dont call BSDI (anymore). Its generally a waste of time. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
RE: if_fxp status?
At 04:58 AM 03/08/2001, Koster, K.J. wrote: Dear All, Harrruummm. (Gong!) Harrruummm. (Gong!) :-) its been discussed several times on this list. No need to go into it again. Then why the hell are you posting if there is no need to go into it again? It is extremely bad form to post asking for something without supplying details or at least a pointer to where they are (msgid's etc). If i regurgitate, then the flames will fly about "reposting". I chose the lower octane method to save gas. I am willing to ship an offending fxp card to whoever offers to fix this problem. We have about 9 useless cards left, IIRC. The 10th is currently on David Greenman's desk. If you are unwilling to do that, then stop bothering us! Indeed, I think Dennis could solve two problems by just dropping FreeBSD right away. maybe the 1000+ companies that use our stuff to run their businesses would disagree with that. Dennis seems to have a knack for rubbing everyone else the wrong way, David said he'd fix this and hasn't. We all know this, so let's please leave it at that. Let's separate the technical issues from the personal ones. I think maybe you need a nice vegetable cocktail. I recommend tomato/celery/carrot juice with a snip of parsley. Its very soothing. You'll live longer. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp status?
At 01:14 AM 03/08/2001, Poul-Henning Kamp wrote: In message [EMAIL PROTECTED], Peter Wemm writes: Dennis wrote: At 05:20 PM 03/07/2001, you wrote: : :The people involved know. : Harrruummm. (Gong!) Harrruummm. (Gong!) -Matt its been discussed several times on this list. No need to go into it again. Then why the hell are you posting if there is no need to go into it again? It is extremely bad form to post asking for something without supplying details or at least a pointer to where they are (msgid's etc). There are other people who can fix it, but they cannot do anything if you refuse to supply all the precise details in an easy to access form. If you are unwilling to do that, then stop bothering us! Indeed, I think Dennis could solve two problems by just dropping FreeBSD right away. Why do I get flamed for expressing my concerns about a very important driver and PHK, who hasn't said one intelligent thing on the list in years, regularly makes inflammatory statements with no content whatsoever such as the above and on-one ever "disciplines" him? reminds me of an old george orwell novell Jordan? DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
if_fxp - the real point
I dont have time for weenie flame wars with people who are more interested in ignoring problems than fixing them. The point here seems very simple. The intel NICs are on a large number of MBs and the eepro100 is the most popular card on the market. So why is there vitually no maintainer for arguably the most important driver in the freebsd tree when there are maintainers for scads of obscure, must less used cards? freebsd touts a "core team" which provides "direction"...does the "direction" include letting important drivers fall out of maintenance in favor of some crappy netgear card that chokes at 3,000pps? Keeping mainstream FreeBSD releases up to date is more important then working on next years release. Otherwise you just have another linux. DB PS: Whatever happened to all of the "support" that BSDI was going to provide? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 12:35 PM 03/08/2001, Mike Smith wrote: The point here seems very simple. The intel NICs are on a large number of MBs and the eepro100 is the most popular card on the market. So why is there vitually no maintainer for arguably the most important driver in the freebsd tree when there are maintainers for scads of obscure, must less used cards? You appear to have a very wrong idea about how these things work. There's a maintainer for the fxp driver, who currently happens to be out of circulation. Nobody else has stepped up to take it on because obviously nobody is motivated to do the work. As for the rest; there's basically one maintainer for these "scads" of (not actually very) obscure cards - Bill Paul. maybe commercial vendors would be willing to fund some freebsd projects if there was a positive relationship. And DG has been MIA for a long time, not just recently. The last time the fxp driver was broken i found the fix and pointed him at the code. I guess I'll have to fix it this time also, as if maintaining drivers for 6 serial cards and the bwmgr isnt enough for me to do. Maybe I'll sell it this time. :-) PS: Whatever happened to all of the "support" that BSDI was going to provide? Call BSDi (numbers on their website) and ask to speak to Gary Johnson (CEO) or Mark Garver (senior VP) and ask them yourself. If you get a useful answer, please tell the rest of us; especially me, since I burnt out trying to make it happen. Ive spoken with Mr Johnson several times. hes the biggest reason that BSDI is what it is. Amazing that hes still there. He has no concept of what the market wants. First he wanted to challenge microsoft (at similar prices) for the server market now he wants to be redhat. funny guy. So what is the "relationship" that was announced? If they dont provide support or funds, what do they do? DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
RE: if_fxp - the real point
At 12:50 PM 03/08/2001, Andy [TECC NOPS] wrote: There's a maintainer for the fxp driver, who currently happens to be out of circulation. Nobody else has stepped up to take it on because obviously nobody is motivated to do the work. Would love to step up and produce a patch, just too busy at the mo working on other things. However, if this thread is still raging when I get some spare time I'd be happy to contib code. In the meantime, I'll just continue reading this cute conversation. "cuteness" is in the eye of the beholder :-) DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp - the real point
At 12:20 PM 03/08/2001, Peter Seebach wrote: In message [EMAIL PROTECTED], Dennis writes: PS: Whatever happened to all of the "support" that BSDI was going to provide? I have no doubt that the BSDi sales office would be happy to sell you a contract. For that matter, I believe they are quite happy to do funded development. Want the fxp driver fixed in FreeBSD? Contact your local sales critter, describe what you want done, get an estimate, and if you like the price, pay it. That's how other people with "thousands of customers" get key hardware support that's a bigger priority for them than it is for other people, and it's not hard. No. Keeping supported drivers up to date is part of the business of distributing an OS. Thats what "supported" implies. The driver is out of date. Noone is looking for a feature here. We just want it to work. -s To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp status?
At 11:57 AM 03/08/2001, [EMAIL PROTECTED] wrote: I've hesitated to be a part of this, but I can't stand by and watch someone be denigrated that I've both learned from and that has contributed huge amounts of code to the project. too bad you dont consider contributions to users of freebsd that arent free. The fact that they are willing to pay for those contributions is de facto evidence that they are of great value. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
if_fxp status?
I really dont care to ask this on the list, but DG doesnt answer my private emails, so I have little choice. has any progress been made on making the if_fxp driver work with the latest intel NICs? Its been over 3 months since the problems creeped up, and its becoming a serious problem explaining to potential freebsd users why is doesnt work. thanks, DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp status?
At 12:18 PM 03/07/2001, Jonathan Lemon wrote: In article local.mail.freebsd-hackers/[EMAIL PROTECTED] you write: I really dont care to ask this on the list, but DG doesnt answer my private emails, so I have little choice. has any progress been made on making the if_fxp driver work with the latest intel NICs? Its been over 3 months since the problems creeped up, and its becoming a serious problem explaining to potential freebsd users why is doesnt work. You might receive more informed answers if you would detail exactly what problems that you are seeing, and what hardware you are using. The people involved know. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp status?
At 05:34 PM 03/07/2001, Thierry Herbelot wrote: Luigi Rizzo wrote: [SNIP] Now, the 21143 (which is a pretty nice chip and has available documentation and a decent driver, "dc") is discontinued, but there are clones which work reasonably well (and are even cheaper, around $30 or so at compusa, i think netgear or linksys does one of these cards). I'd go with them. No multiport card, at least as far as i know. What : no more dc multiport board ? are you sure of this ? if so, it may be wise to buy some D-link 570-TX boards in advance for my project. The 21143 is not discontinued. Its an afterthought perhaps, but still in use by Dlink for sure and certainly others. Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp status?
At 05:20 PM 03/07/2001, you wrote: : :The people involved know. : Harrruummm. (Gong!) Harrruummm. (Gong!) -Matt its been discussed several times on this list. No need to go into it again. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Hardware question - WAN port for FreeBSD router!
At 10:05 AM 03/01/2001, you wrote: Hi all. Please forgive me if this post is altogether in the wrong place, but I need help finding a certain piece of hardware. I have a number of firewalls running (very happily) on FreeBSD. Each one is connected on the outside NIC to a Cisco 1601 Router, which is connected in turn to an NTU etc I am sure I could save a lot by installing a WAN card directly into the BSD firewall, thereby doing without the Router altogether - FreeBSD can do all the routing I need after all! So - is there such a WAN serial card available? And (most important) one which is supported by the necessary BSD drivers? www.etinc.com V.35, EIA-530, X.21, RS-232, HSSI. Support for FreeBSD v3.4, v4.1 and v4.2. PPP, Frame Relay (1024 DLCIs per line), densities to 4 ports per line, 16 ports per system. All cards include integrated Bandwidth Management software. We also sell complete prebuilt !U routers with up to 4 ports with complete GUI management. Dennis Emerging Technologies, Inc. - http://www.etinc.com ISA and PCI T1/T3/V35/HSSI Cards for FreeBSD and LINUX Multiport T1 and HSSI/T3 UNIX-based Routers Bandwidth Management Standalone Systems Bandwidth Management software for LINUX and FreeBSD To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
FreeBSD cant boot from ZIP - Is it true?
Is it true that freebsd can't boot from a zip drive? Dennis To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: FreeBSD cant boot from ZIP - Is it true?
At 02:20 PM 02/15/2001, Mike Smith wrote: Is it true that freebsd can't boot from a zip drive? No. You can't boot from the parallel-port drives, but that's a feature of the PC not expecting to be able to boot from a printer rather than a FreeBSD issue. I've personally booted FreeBSD from the old, old ATA Zip drives, the ATAPI ones and of course the SCSI drives, in both whole-disk and sliced (as from the factory) modes. I believe you, but no-one seems to know how to do it. FreeBSD seems to get confused between the hard drive and the ZIP and it becomes a mess very quickly. This is with ATAPI IDE drives btw. Whats the "trick"? Dennis -- ... every activity meets with opposition, everyone who acts has his rivals and unfortunately opponents also. But not because people want to be opponents, rather because the tasks and relationships force people to take different points of view. [Dr. Fritz Todt] V I C T O R Y N O T V E N G E A N C E To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: ADSL and PPPoE question
At 10:02 PM 02/12/2001, Julian Elischer wrote: [EMAIL PROTECTED] wrote: On Mon, 12 Feb 2001 08:20:55 -0800, Julian Elischer wrote: [EMAIL PROTECTED] wrote: I'm very sorry if this is a stupid question. In our company, we want to set up a small network of about 20 PCs. ADSL seems like a good inexpensive solution, and I understand that FreeBSD with Netgraph can act like a gateway for our computers. are they in different places? No - the same place. Negraph/ppp can act as a gateway for pppoe connections but I am not sure how that helps you. How do you get the ADSL sessions to terminate on an ethernet in your office? (does your ISP provide that service?) The "way" do to this is by getting a frame T1 and running a DLCI to each of your customers, and do ethernet bridging over the frame. You have 1 line into your hub and you can service many customers on the frame. We have customers services 200+ customers on a single T1 and 800+ and a T3. A cisco 1600 will choke at about 150 customers (and perform like crap as you approach it). Dennis What I don't understand is whether we will have to use IP aliasing (NAT) or we can have our own routable IP range. That very much depends on what you think the topology looks like.? I thought something like this: [ISP] | | - Office [ADSL] | | [FreeBSD Box] | | | | | | | | [A][B][C][D] where A, B, C, D all have their own routable IPs. So according to what you said, FreeBSD would need to establish a separate PPPoE session for each of the computers A, B, C and D, provided the ISP supports multiple PPPoE sessions over the single ADSL line? no that would give 4 connections to the FreeBSD machine, and not 4 connections to the client machines. Does your ISP even use PPPoE? it is possible to have ADSL without it.. I need to know if this configuration is possible, so I will know what to ask the ISP sales and support people. So far they haven't been very helpful. Thank you very much for your reply what the ISP will try sell you will be: [ISP] | | - Office [ADSL] | | ethernet +--+--+--+ | | | | | | | | [A][B][C][D] which is quite possible with ethernet attached ADSL modems. this might even be an ok way for you to go, but the client machines will be 'vulnerable' on the network to hackers. A more common answer is: [ISP] | | - Office [ADSL] | | ethernet [FreeBSD Box] | | | | | | | | [A][B][C][D] where the freebsd box does NAT and the other machines have addresses assigned in the 10.x.x.x space or 192.168.x.x space. this gives you some protection of the client boxes. -- __--_|\ Julian Elischer / \ [EMAIL PROTECTED] ( OZ) World tour 2000-2001 --- X_.---._/ v To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: ADSL and PPPoE question
At 11:02 AM 02/12/2001, [EMAIL PROTECTED] wrote: I'm very sorry if this is a stupid question. In our company, we want to set up a small network of about 20 PCs. ADSL seems like a good inexpensive solution, and I understand that FreeBSD with Netgraph can act like a gateway for our computers. What I don't understand is whether we will have to use IP aliasing (NAT) or we can have our own routable IP range. Since ADSL is a Point-to-point link over Ethernet, does the protocol / one ADSL line support more than one IP, or will everything have to be aliased? FYI: We support direct IP over ethernet bridging on our cards without the pppoe nonsense on FreeBSD. Bridging over both frame relay and PTP lines are supported. www.etinc.com Dennis Thank you Takashi Kiguchi ___ Send a cool gift with your E-Card http://www.bluemountain.com/giftcenter/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: watchdog bugging us.
At 08:18 PM 01/31/2001, Reddy Crashalott wrote: [disclaimer: this is an invalid sender and return address; replies should be posted to the list or perhaps better not posted at all, finally i am rid of that cursed e-mail reachability, free at last] :: When i do tail -f /var/log/messages, it gives me this. :: :: Jan 31 15:03:49 beta /kernel: xl0: watchdog timeout :: Jan 31 15:04:24 beta last message repeated 6 times I've seen this on questionable HP NetSwerver hardware, that you may be able to solve by either trying your xl ethernet card in a different PCI slot, or in your BIOS doing something about the IRQs. By doing the latter, I got things to work without problems, while otherwise things were unacceptably bad or stopped working totally after a short time, depending on which of several flavours of NICs I was using. Its really questionable drivers...BP's if_dc driver will timeout by design if you dont patch the driver. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: hardware not raising interrupts
At 07:48 AM 01/30/2001, Suma H R wrote: Hi, We are writing a network device driver for an ethernet interface network card. The hardware IRQ status register shows an Interrupt to be Pending. But, vmstat -i doesn't show any interrupts from our device. Any idea why it could be happening? maybe you didnt register it or enable it on your MB? Thanks, Suma To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp driver info (which card then?)
At 08:51 AM 01/26/2001, Mike Wade wrote: On Fri, 26 Jan 2001, Greg Lehey wrote: Performance isn't even the main thing. As I said earlier, it's plain bloody unreliable. Linux people avoid the EtherExpress because they think something is wrong with the card. They were surprised when I reported that it works without any problems under FreeBSD. Do we really want to change that? Slightly off subject but with all the discussion about not Intel playing nicely with the FreeBSD developers... I've always had the best reliability, performance, and lower CPU usage with the Intel EtherExpress Pro 10/100B cards in FreeBSD (and Solaris x86 for that matter). Are there better cards out there that I should be looking at? Why dont some people get the point even when you hit them in the head with a hammer? The point is that the driver quality is more important than the "card" To get completely off base, this is which is why we SELL our software. Implementation technique is usually more decisive in determining functionality and performance than the hardware itself. its something that people in the know are willing to pay for (sometimes). Certainly some hardware is better than others, but a bad driver with good hardware is useless. DB To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp driver info (which card then?)
At 09:47 AM 01/26/2001, Jim Sander wrote: Linux people avoid the EtherExpress because they think something is wrong with the card. Intel EtherExpress Pro 10/100B cards in FreeBSD These cards work well in our many 3.x and 4.x systems. But I just built up a Redhat 6.2 box with one, and all seemed to be working fine, but after a while I started having various problems starting net services. The box would boot, but often would "hang" indefinitely when "Starting eth0" - requiring a hard reboot. I swapped to another EE-Pro NIC, new MB, different RAM, other cables, everything, but no change. the eepro100 driver is badly broken in linux (havent you been paying attention?). it took me a few hours to fix it. They dont reset the card properly on an overrun, which causes it to lock up. Clearly the driver as is is unusable in a heavy use environment. DB After I switched to a linksys NIC, voila- everything worked without a problem. (so far) Of course the Intel NICs still work perfectly when put into a spare BSD system. So it's *not* that the cards themselves are unreliable. Perhaps the drivers controlling them? Perhaps a weird MB/NIC conflict of some sort? -=Jim=- To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: if_fxp driver info (which card then?)
At 11:43 AM 01/26/2001, Aleksandr A.Babaylov wrote: Mike Wade writes: On Fri, 26 Jan 2001, Greg Lehey wrote: Performance isn't even the main thing. As I said earlier, it's plain bloody unreliable. Linux people avoid the EtherExpress because they think something is wrong with the card. They were surprised when I reported that it works without any problems under FreeBSD. Do we really want to change that? Slightly off subject but with all the discussion about not Intel playing nicely with the FreeBSD developers... I've always had the best reliability, performance, and lower CPU usage with the Intel EtherExpress Pro 10/100B cards in FreeBSD (and Solaris x86 for that matter). Are there better cards out there that I should be looking at? 3C905 I disagree. The if_fxp driver is far superior to the if_xl driver. In other OS's your mileage may vary. DB -- @BABOLO http://links.ru/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message