Re: Hooking 2 Networks
Hooking is illegal in the USA. Even over networks. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hooking 2 Networks
Thats why the call it 'escorting' :) Hahaha, I spat out some pepsi when I read that Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Problems with the rl driver...
Your card's resources are turned off. Use your computer's BIOS/CMOS/whatever configuration program to activate resources on the card. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hang after soft reboot.
If the extra hacking was done, what would be wrong with that occurring at the end of the reboot process, rather than the start of the boot process? It probably wouldn't *hurt* anything, though, it wouldn't *do* anything, either. When power cycles the ATA will initialize to the native state. So, any assertion of pins prior to power cycle would be moot. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hang after soft reboot.
Gah. Correction: that was either a fluke, the problem is intermittent, or I hit the reset button without thinking and through it had worked. Probabbly the last option. More caffeine please. Damn... I was hoping that worked.. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
[Fwd: Re: perplexing problem (non-ATA66 cable or device error...)]
Original Message From: - Thu Feb 20 14:16:57 2003 X-Mozilla-Status: 0001 X-Mozilla-Status2: Message-ID: [EMAIL PROTECTED] Date: Thu, 20 Feb 2003 14:16:52 -0500 From: northern snowfall [EMAIL PROTECTED] User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:0.9.4.1) Gecko/20020518 Netscape6/6.2.3 X-Accept-Language: en-us MIME-Version: 1.0 To: Ron Andreasen [EMAIL PROTECTED] Subject: Re: perplexing problem (non-ATA66 cable or device error...) References: [EMAIL PROTECTED] Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit acd0:read data overrun 34/0 acd0:MODE_SENSE_BIG command timeout - resetting ata0:resetting devices ..ad0:DMA limited to UDMA33, non-ATA66 cable or device These messages occur frequently when you've got a CDROM in the drive thats been burned improperly. The image is usually written too large for the media. Did you burn the FreeBSD iso yourself? Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
[Fwd: Re: different disk geometry]
Original Message From: - Thu Feb 20 14:28:55 2003 X-Mozilla-Status: 0001 X-Mozilla-Status2: Message-ID: [EMAIL PROTECTED] Date: Thu, 20 Feb 2003 14:28:50 -0500 From: northern snowfall [EMAIL PROTECTED] User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:0.9.4.1) Gecko/20020518 Netscape6/6.2.3 X-Accept-Language: en-us MIME-Version: 1.0 To: Michael Soboleff [EMAIL PROTECTED] Subject: Re: different disk geometry References: 000901c2d8f0$6e8fe570$[EMAIL PROTECTED] Content-Type: text/plain; charset=KOI8-R; format=flowed Content-Transfer-Encoding: 7bit I have 40Gb IBM IDE drive, while booting freebsd 4.7 shows me 79780/16/63 geometry, but! sysinstall gives me another numbers : 5005/255/63. The QUESTION is it OK? Usually, what sysinstall gives is ok. Just make sure the number of sectors that sysinstall perceives is the same as or under the actual sector number. You can check by doing: 79780 * 16 * 63 = 80,418,240 sectors (512eight-bit octets each) 5005 * 255 * 63 = 80,405,325 sectors Any array of proper C/H/S can be given. What actually happens here is that Sysinstall determines the total amount of sectors on the disk (most likely in LBA mode), then determines the C/H/S based on most-likely-candidate mapping. 63 is maximum value for sectors per head. 255 is the maximum value for heads per cylinder. Thus, (Total_Sectors / (nHeads * nSectors)) = nCyls. If you are unhappy with Sysinstall's chosen mapping due to loss of X number of sectors, consult your hard disk's documentation to determine the valid C/H/S value or total sector number. Edit as appropriate. As a side note, don't give a number that exceeds the total number of sectors on the disk. The driver may attempt to access these sectors expecting OK from the controller, but, receiving ERROR. This may cause the driver to tell you something more malicious is occuring than is not, causing trouble down the road. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hang after soft reboot.
Hi, Hello On a fresh, clean install of 4.7-RELEASE my (old) machine consistently hangs while booting after soft reboot. Powering down and back up again, or pressing the reset button will boot the machine, but if it is rebooted by the OS then it hangs after detecting the isa bus isa0: ISA bus on motherboard. This is usually a manifestation of old power interfaces (or lack there-of) mingling with devices that need to be told when to assert a RESET via power management. For example, I have a hurd of Compaq Deskpro that all use the old VIA 586 power controller. Because no OS (that I know of ... except for maybe win*?) has a driver for this chip, I get the same hang you talk about on a warm boot. Since the chipset isn't thunked properly, it doesn't let the ATA know to RESET. This can be a problem when probing the disk for information, but, normally doesn't affect loading the boot sector or the first couple of cylinders. I forget the restriction atm... Anyways, thats most likely your culprit. This is a normal boot up, with the hang point indicated: hangs here on soft reboot ata0 at port 0x1f0-0x1f7,0x3f6 irq 14 on isa0 ata1 at port 0x170-0x177,0x376 irq 15 on isa0 Solution? Never warm boot. Don P.S. if you keep warm booting you might corrupt the ATA To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hang after soft reboot.
Would there be any chance that the system is working with some generic driver and there is a specific driver for my hard drive controller (it's an ISA card), which would solve this? Well... all ATA in freebsd are pretty much conglomerated into the same driver... but, the problem isn't really so much the ATA as it is that the ATA expects a PM telling it what to do. Since this is something that must be done via the hardware at boot, FreeBSD doesn't really have a way to tell the ATA what to do. One solution might be hacking extra reset controls, etc, into the ATA driver so that this functionality is asserted on boot. Then, though, there is a possible chicken-egg issue: you're initializing the disk and snarfing data off the disk at the same time as attempting a hard reset, which, might cause a lock (or worse). I think your most painless solution is upgrading your mother board. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hard error??
I used to get this error on a FreeBSD while using a perfectly stable harddrive. That harddrive is managed via Solaris now, but, I determined the issue during its FreeBSD usage was DMA. If you are running two disks on the same ATA channel with different DMA capabilities, the capabilities may be causing scrambles in the negotiation of I/O on the line. The solution is to put ATA drives that use _only_ the same DMA caps on the same ATA channel. If you only have two drives, simply put ATA0.1 on ATA1.0. This stopped my falling back to PIO messages and probably saved the disk from hard failure caused by misuse. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hard error??
Is this a bug in the FreeBSD ATA driver then?
Its entirely possible, but, I, personally, wouldn't know for sure.
I'm just getting in to the depths of the ATA specs. It may not be
a bug so much as a lack of handling specific DMA issues.
Maybe someone should CC freebsd-{hardware,hackers}@
Don
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-questions in the body of the message
Re: Hard error??
Anyway, it seems like I have just got to get myself a new drive. On that note, has anybody got any idea what I should go for? Any vendors whose drives do NOT cave in after half a year? ;) I choose Maxtor for several reasons. First off, I've been using Maxtor disks the most since I started out in computers and haven't had one fail yet (running every OS i've tested). Now that I'm alittle more experienced, I use Maxtor because of its standing credibility with me, and, because the Chairman of the T13[1] (technical committee for ATA[-ATAPI] development) is from Maxtor Corporation. They are most likely to want to adhere to a published specification (along with other T13 members), rather than develop chipsets that are rushed to keep up with a $25 billion a year industry. Don [1] T13 technical committee http://www.t13.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: network issue
I'm guessing more than likely your DNS server is external to your internal LAN and you don't have an internal DNS to manage RFC1918 IPs. If this is the case, this is why pings will *seem* to fail. They are trying to look up your internal addresses (which will fail with an internet connection up fairly quickly) but hang because there is no connection via your ISP to a DNS server to respond no, there is no PTR for that A. Get yourself an internal DNS and you should be ok Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: network issue revisited
(which rules out DNS problems). Not unless he does ping -n. If not, the A will still attempt to be resolved. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: network issue
Brian Henning wrote: Ok, i am willing to try out the internal dns server but, i don't know which machine to run it on. Any suggestions? Whichever box doesn't act as your most-used-workstation, or, the router if its capable of running a server. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: xe0: discard oversize frame (ether type 800 flags 3 len 1518 max 1514)
This was fixed about a week ago. You need to upgrade to the latest -CURRENT (you could probably get away with just pulling the latest if_xe.c, though). Was the problem that the ether driver was not snipping off the 16bit checksum on full-length frames? Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: xe0: discard oversize frame (ether type 800 flags 3 len 1518 max 1514)
Yes. Specifically, it wasn't telling the higher layers of the stack that the checksum bytes were always there, but it does that now. Right on. That was my guess. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Why is there no JFS?
One with a license that will let it be distributed in the core. That lets out GPL'ed code, and I believe it lets out XFS as well, though I'm not positive on that. Just FYI, IBM's JFS is GPL'd, IIRC, according 2 the WWW site for JFS. Hah, yay for acronyms. http://www-124.ibm.com/developerworks/oss/jfs/index.html Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: PCI interpretation question AND Superprobe question
pci bus 0x0001 cardnum 0x00 function 0x00: vendor 0x1039 device 0x6326 SiS 6326 The SiS 6326 has 8MB or 4MB on-board RAM. You should be able to let X autodetect the video size by selecting 8MB in X86Config. If there is less X will figure it out with a probe. FYI, the BASE[0-2] values are either memory mapped I/O or classic port I/O for transferring data to/from the PCI. It is not RAM. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: printf...! BSD
Just try reading the FreeBSD kernel source. All the answers are right there. Why read a book or an article about how it works when you can see how it works for yourself =) Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: I'm running out of swapspace
One way to do it is to create a memory file system linked to a file on-disk. In FreeBSD 5.0-RELEASE, to create a 128MB additional swap space, I've tried: dd if=/dev/zero of=/swap0 count=128 bs=1m chmod 600 /swap0 mdconfig -a -t vnode -f /swap0 -u 3 swapon /dev/md3 This will link the 128 megabyte file into swap space recognized by the system. Execute swapinfo to confirm: sandstone.north_ % swapinfo Device 1K-blocks UsedAvail Capacity Type /dev/ad0s1b52428810136 514152 2%Interleaved /dev/md3 131072 1876 129196 1%Interleaved Total 65536012012 643348 2% sandstone.north_ % Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: freebsd installer is braindead
i just tried 4.7 and it doesn't work either. how do you people install freebsd? both 4.7 and 5.0 installers are braindead. does anyone have a work around for the problem of the installer extracting into the mfsroot instead of /mnt? Did you try using another ethernet card besides the 8139? I had the same problems when using this chip. Finally switching to the Kingston is what saved the install. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: freebsd installer is braindead
what about the problem of it extracting into the mfsroot? switching nics fixes that? Yea, that was actually the main problem I was referring to. The look up problem can be eluded by aborting then restarting the installation. I did this in Expert Mode so I could go straight to media configuration once, thunking the Ether device to manifest the look up bug. Then, during stall, interrupted the look-up to restart the entire sysinstall instance. The second media initialization is clean. Switching from the 8139 fixed the mfsroot bug as well. I have no idea what the link between the 8139 and the mfsroot is. I'm putting money on an interrupt layer mishandling. None the less, switching NICs fixed it. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: 5.0-release install problem
drivers.flp is your friend :) It happened to me, and I'm no fool! Well, thats kind of the point of the problem. Drivers.flp isn't needed. Sysinstall loads the 8139 driver from its base. The ATA drivers are in the base, as well (of course), so the question is: why does the bug in the 8139 driver manifest in a mishandling of the mounted file system? This leads me to believe that the bug isn't 8139-dependant, but, something rooted in the design of the interrupt code itself. Don i want to be.. as deep... as the ocean To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: 5.0-release install problem
Too deep for me. The cd9660 kld is on drivers.flp and would hopefully help with Shane's Error mounting /dev/ac0 on /dist: Operation not supported by device (19) problem. Oops. Thats what I get for trying to do 10 things at once. This was meant for freebsd installer is braindead thread. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: printf ... !
This is the definition of the write(2) system call. You should also check the implementation of printf(3) at the libc sources. Look at /usr/src/lib/libc/stdio/printf.c for more details about the way printf() works in userlevel programs. Right on. I think, however, that Auge is looking for a trace, not jsut the printf source. More or less we bust out of printf() in libc to: vfprintf/usr/src/lib/libc/stdio/vfprintf.c __sprint/usr/src/lib/libc/stdio/vfprintf.c __sfvwrite/usr/src/lib/libc/stdio/fvwrite.c _swrite/usr/src/lib/libc/stdio/stdio.c This is where you make a direct call to the file stream's write function. In printf's case this is __swrite. __swrite/usr/src/lib/libc/stdio/stdio.c Simply, __swrite() calls _write (which equates to write()). Write is a simple syscall trampoline that manifests int 0x80. This, of course, calls the kernel, which dumps the thread into SYS_write. Skipping the interrupt code we can drop into SYS_write: write/sys/kern/sys_generic.c You should be able to trace the code from here. Don nosotros tenemos mas influencia con sus hijos que tu tienes... pero los queremos.. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: installer: / write failed device full
well, i put my drive on the sym controller (since it's not seeing my aic7x chips) and now when i starts downloading and extracting the binary sets it stops and says there's no space. this is obviously wrong, since just before that it does a newfs on an almost 9gig slice. has anyone else had this problem? this is on my alpha. also, only certain sites work from the installer. they work fine using ncftp, but most others hang while looking up I've also had this problem. Instead of installing into the mounted target drive the installer attempts to fill up /, which is just mfs. I have newfs bug is, but, I've had the same occurance. However, it doesn't seem to *do* anything... As far as the look up hang: this occurs due to mishandled interrupts in the network controller. I don't think its driver specific. Once I installed a PCI device and configured the IRQ to use a non-shared interrupt, the installer was A-OK. I used a Kingston KNE111TX/100B. The base system is a 400mhz Xeon Intel CPU with 128MB RAM. ATA0.0 has a 6449 MB WD. ATA1.0 has a LITE-ON CD. The only other card on the motherboard is an Intel i740 AGP. The motherboard is an Intel 440BX AGPset GA-6BXC. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: installer: / write failed device full
I have newfs bug is, but, I've had the same occurance. However, it doesn't seem to *do* anything... Control+- flip err: .s/I have/I don't know what the cause of the/ Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
5.25 Floppy
Morning, all; I'm trying to get a Mitsumi D509V3 1.2MB 5.25 floppy drive to work on FreeBSD 4.2.6. The operating system reports the drive is available and definitely makes contact with the drive (visual confirmation: LED). The issue is during read/write from the drive. Error message: fd0c: hard error reading fsbn 0 (No status) I have the proper drive type set in the BIOS. FreeBSD seems to agree according to the dmesg: fd0: 1200-KB 5.25 drive on fdc0 drive 0 I've been doing simple read tests using: dd if=/dev/fd0 count=1 bs=512 | hexdump ; Any suggestions? Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: 5.25 Floppy
I thought that was the newer ones. Weren't the old style 5.25s 640K? It's been so long . . . Model independant Where the heck did you even find a working one? Around the house. We have three... Even the 3.5s are pretty much beyond usefulness for me now that the net is everywhere and the CDRWs are so easy and cheap, but those 5.25s would be pretty interesting for nostalgia. Nostalgia, sure... but, hacking an 8086 boot disk just to access an even more ancient hard disk with an original 70s FORTRAN compiler and libraries? Much cooler. Either way, I'll bet Daxbert's advice will at least set you in the right direction. Good luck. I appreciate the luck =) But, I tried that before posting to the lists. No such luck in that direction. The driver plumbs the fd?.* interface along with the fd? interface based on values probed from the CMOS, so, its basically a namespace bind on static size media. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: Hard Drive
It would be nice if this information could be removed from the disklabel. I guess if that's not feasible, a notation in the documentation would be nice (I was confused about this about a year ago). If it were up to me, I'd rather keep it in. I get sick of thunking my documentation database just to figure out what RPM drive X runs at for OS'es I use (and will still use in the future) that consider the value relevant. disklabel -r X is much easier than bothering NFS and a db thunk. If something needs to be done, make a note in the disklabel manpage. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: 5.25 Floppy
Here's the rundown: SSDD 8-sector: 160K This is all that was available in MS-DOS 1.0, IIRC. SSDD 9-sector: 180K DSDD 8-sector: 320K DSDD 9-sector: 360K This was the default starting with MS-DOS 2.1, IIRC. DSHD 15-sector: 1.2G (At least I think it was 15-sector...) Only the last two appear to be supported by FreeBSD 4.7, at least by default. Yeh, according to FreeBSD's /sys/isa/fd.c and /sys/sys/fdcio.h 360k 1.2M 720k are the only ones supported. This, in particular, is a Mitsumi D509V3 which is a 1.2M. Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: languages
what are the bbest three languages to learn? C, java and as many flavors of ASM as possible =) English, Chinese, and Latin. Of course, that does depend on what you're going to use them for. Sysadmin / International Diplomat for the Preservation and Perpetuation of universal Freedom and Peace. Unless... electrons suddenly consolidate into terrorist cells... Oh wait... is that what a DoS attack is? Don To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: languages
While C might be a good starting point UNIX is written primarily in C. Solaris, IRIX, AIX, Linux, etc, and, yes, even FreeBSD, are all developed in C. My own OS is written in C (*wink*). So is my favorite OS to hack in: plan9. People thought this trend might decrease as the years passed on, yet, they have not. Operating system design trends tend to be object oriented in design, yet, reside in a C language primary. With the movement towards an IA64 platform, I think the trend towards object oriented conceptualization will persist, while base code will stay in the C domain. UltraSPARC and other 64bit families may be designed with OOB in mind, that doesn't seem to be the trend in utilization throughout the research community, both public and private. and Java might teach you object orientation skills one might choose C++ over C or even over Java. Objects are simply an abstract of perception relative to one's environment. That abstraction changes with every individual to a degree, yet, stays founded on a generic concept of orientation. This foundation can be maintained in any language. There is no pure OOB language, nor is there a best OOB language. Instead of talking about portability/usability/etc I will simply summarize Java by saying: Sun rules. As for ASM, it gives you a good background over how a computer works but it's not suitable for every programer. Every programmer must learn underlying architecture to comprehend the design and intent of his application, no matter what level of the OSI (or another model) the app resides in. This relation is inherent in every abstraction of computer design. A great example is the recent security flaws in some Wayne County/Michigan web sites. The developers did not understand the underlying architecture of the internet, or, even more trite, the design of web site transaction state and HTTP. This failure to dig deeper caused users to reveal other users' credit card information simply by substituting names in the web site's normal functionality. Underlying architecture comprehension not suitable for every programmer? Don dead cats... dead rats... To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: languages
The Hurd project is beeing developed in C++. Right on While I do like to program in C, I think C++ code is much better organized and it's a lot easier to read. Well, that depends on who is doing the coding, of course. That is applicable to any language. I will admit that BSD code can be *very* messy; even for the advanced C hackers. If you want to see some very clean and well written C code, check out plan9. plan9.bell-labs.com Bell Labs gets props for a good reason, IMO Don blood in the streets its up to my knees... To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
Re: languages
I can see how C and asm fit into that picture. One's the portable assembler you right most of your code in, and the other is the non-portable assembler that you right the other bits in. How does Java fit? Java is just the next level of abstraction. C may be portable in source form, but, of course, its not in binary form. Java attempts to take that next step. There are several candidates for this scope of binary abstraction, of course; Java just happens to be my favorite. I'm beginning to think the original posters goal was to start a long, off-topic discussion. I agree. I suspected this at first; though, I kind of felt like playing the role of oxygen. Besides, if he wanted information he got it =) Don sucking on the soldier's brain... To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-questions in the body of the message
