Re: Stale modules (Re: panic in the morning)
On Wed, 19 Apr 2000, Kris Kennaway wrote: On Wed, 19 Apr 2000, Christoph Kukulies wrote: I cvsup'ed, built world and kernel. Hhmm, actually I see no reason why there should be a problem since everything should be done by make world. make world doesn't build a kernel. Making a kernel doesn't build modules. This bit me again the other day when updating, as well - panic at boot when loading a stale linux.ko. If making world _and_ kernel doesn't build modules, what _then_? Leif To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
Leif Neland wrote: make world doesn't build a kernel. Making a kernel doesn't build modules. This bit me again the other day when updating, as well - panic at boot when loading a stale linux.ko. If making world _and_ kernel doesn't build modules, what _then_? Making world builds kernel modules. If you followed this thread before, I stated that modules should be built with the kernel. After all, they ARE kernel modules, and are part of the kernel, and not the world. I'd like to discuss further the possibility of creating some sort of mechanism where the modules can be built with the kernel. Also, we can have some sort of option in LINT or GENERIC where a keyword, such as module, can be put somewhere in the kernel config file line to compile certain drivers as modules instead of statically linking them into the kernel. Which mailing list would be appropriate for discussing kernel modules, etc.? - Donn To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
attachable driver modules (Re: Stale modules (Re: panic in the morning))
On Sun, 23 Apr 2000, Donn Miller wrote: I'd like to discuss further the possibility of creating some sort of mechanism where the modules can be built with the kernel. Also, we can have some sort of option in LINT or GENERIC where a keyword, such as module, can be put somewhere in the kernel config file line to compile certain drivers as modules instead of statically linking them into the kernel. An item which would probably go a long way towards demystifying the kernel process would be to use loader.rc to add device modules at load time; for instance, all the Ethernet cards, the list of which is always in flux. Theoretically, it should be possible to go as far as the disk, tape, CD, etc. drives (no point in floppies), and even a choice between the console and a VT220 on a port. Initial load would bring in a /GENERIC from which a GUI with check the boxes would configure loader.rc and subsequent reboots would be with the standard /kernel. To modify an existing /kernel, reboot with /GENERIC... or make the editor accessible, maybe even including an X option. The Linux crowd, or at least Caldera and Corel, are trying to beautify the install... I have ranted for over 20 years, starting with Bell Labs, BSD 4.0, and numerous keynotes in Europe in the 80s (too inflammatory to the 100+ U.S. feel-good *ix vendors). I can remember being impressed with SUN's GUI install in the 1982 SUN 2 (I still have both a "macho" SUN 1 and a SUN 2). We, who as a whole live by RTFM (and my extension of RTFS-source), have not been willing to grant the unwashed an audience. But, keep in mind, Bill Gate$, with a snappy check the box setup, marginalized us --and still does despite an "almost usable" product with daily or hourly BSODs vs. my last uptime of 272 days! It's all well and good that us 'cognizenti' or 'intelligentsia' are perfectly happy with a command line interface. But X permits me to run at least 4 'xterm' windows, pine, Xemacs, dclock, xcdplayer, xshisen, and Netscape --and have as many as 40 open Netscape windows (great for picking e-News articles in a batch or diversions --no backstepping). I could care less about KDE's or Gnome's flash (in fact, KDE's visible primary folders plus the icons are a nuisance) --hence I use the old standby: 'TWM' which works fine for me. the bottom line is if we (and Linux and the rest of the crowd) wish to push Gate$' $lop off the desktop, we need to be user- friendly --that, and hope Judge Jackson forces 'Office' into Open Source which would sure light BadBillyG's fire! attila out... -- Caveat: No Microsoft programs or processes were used in the creation or distribution of this message. If you are using a Microsoft program to view this message, please be advised I am not responsible for any harm you may suffer as a result. ___ ___ ___ _ __ ___ | _ ) __| \ freeBSD: The Power to Serve! _ __ ___ | _ \__ \ |) | Release 5.0-CURRENT - The Fast Lane! __ _ |___/___/___/ It's like a wigwam -- no Gates, no Windows, and an Apache inside. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
On Sun, 23 Apr 2000, Leif Neland wrote: make world doesn't build a kernel. Making a kernel doesn't build modules. This bit me again the other day when updating, as well - panic at boot when loading a stale linux.ko. If making world _and_ kernel doesn't build modules, what _then_? Making them both together. Making world _does_ build modules as I said above, you'll just run into problems if you, say, make your kernel after cvsupping 3 days later. Kris In God we Trust -- all others must submit an X.509 certificate. -- Charles Forsythe [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
On Sun, 23 Apr 2000, Donn Miller wrote: Which mailing list would be appropriate for discussing kernel modules, etc.? freebsd-arch. Kris In God we Trust -- all others must submit an X.509 certificate. -- Charles Forsythe [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
In message [EMAIL PROTECTED] Alex Zepeda writes: : Perhaps it's time to implement some sort of versioning in the modules to : prevent them from being loaded into the incorrect kernel. In theory that sounds nice, but in -current the kernel ABI changes too quickly for that to be effecitve. The general rule in -current is that the modules must be compiled at the same time as the kernel, or from the same sources. Anything else might work, but is unsupported. I don't know of a good way to serialize (eg assign a serial number to) the sources that would work. For -stable, long periods of time pass between the ABI changes forcing a recompile, so it could be keyed off the FreeBSD version somehow. Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
On Wed, Apr 19, 2000 at 06:15:55PM -0700, Kris Kennaway wrote: On Wed, 19 Apr 2000, Christoph Kukulies wrote: I cvsup'ed, built world and kernel. Hhmm, actually I see no reason why there should be a problem since everything should be done by make world. make world doesn't build a kernel. Making a kernel doesn't build modules. This bit me again the other day when updating, as well - panic at boot when loading a stale linux.ko. \begin{newbie question} So why aren't the modules built with the kernel instead of with the world? \end{newbie question} --Stijn -- FreeBSD: | When I was in school, I cheated on my metaphysics exam: I The power to serve | looked into the soul of the boy sitting next to me. -- www.freebsd.org | Woody Allen To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
Warner Losh wrote: In message [EMAIL PROTECTED] Alex Zepeda writes: : Perhaps it's time to implement some sort of versioning in the modules to : prevent them from being loaded into the incorrect kernel. In theory that sounds nice, but in -current the kernel ABI changes too quickly for that to be effecitve. The general rule in -current is that the modules must be compiled at the same time as the kernel, or from the same sources. Anything else might work, but is unsupported. I don't know of a good way to serialize (eg assign a serial number to) the sources that would work. Why not just use a UTC timestamp? If your kernel is newer than the module then don't load the module. Maybe make it a sysctl as well so that rapid developer folks can disable it. Even if it was used just as a warning it would remind people to think about the modules. I've lost count of the number of times that I've chased a kernel bug before remembering to rebuild modules. I know a timestamp doesn't guarantess ABI compatibility but it solves the most common problem, which is rebuilding a kernel but forgetting to rebuild the modules. Paul. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
On Thu, Apr 20, 2000 at 11:56:08AM +0100, Paul Richards wrote: that rapid developer folks can disable it. ITYM "rabid". And I kinda resent that.. ;-) -- Will Andrews [EMAIL PROTECTED] GCS/E/S @d- s+:++:- a---+++ C++ UB P+ L- E--- W+++ !N !o ?K w--- ?O M+ V-- PS+ PE++ Y+ PGP t++ 5 X++ R+ tv+ b++ DI+++ D+ G+ e- h! r--+++ y? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
Will Andrews wrote: On Thu, Apr 20, 2000 at 11:56:08AM +0100, Paul Richards wrote: that rapid developer folks can disable it. ITYM "rabid". And I kinda resent that.. ;-) I really did mean rapid, as in those that are installing kernels every 10 mins to test changes. Though, these days there does seem to be a lot of foaming at the mouth :-) Paul. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
Stijn Hoop wrote: On Wed, Apr 19, 2000 at 06:15:55PM -0700, Kris Kennaway wrote: make world doesn't build a kernel. Making a kernel doesn't build modules. This bit me again the other day when updating, as well - panic at boot when loading a stale linux.ko. \begin{newbie question} So why aren't the modules built with the kernel instead of with the world? \end{newbie question} Good question. With Linux, you have the option of building the modules when you compile the kernel by doing a "make modules". I think maybe we could use a similar approach. For example, I think we should put an option into config(8) where we can choose the modules we want to build. Like, we add the keyword "module" somewhere to the driver we want to add as a modules. Then, we could add an option "make modules" and "make install_modules" so that they could be built/installed with the kernel. After all, modules ARE a part of the kernel... - Donn To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
Donn Miller wrote: Stijn Hoop wrote: On Wed, Apr 19, 2000 at 06:15:55PM -0700, Kris Kennaway wrote: make world doesn't build a kernel. Making a kernel doesn't build modules. This bit me again the other day when updating, as well - panic at boot when loading a stale linux.ko. \begin{newbie question} So why aren't the modules built with the kernel instead of with the world? \end{newbie question} Good question. With Linux, you have the option of building the modules when you compile the kernel by doing a "make modules". I think maybe we could use a similar approach. For example, I think we should put an option into config(8) where we can choose the modules we want to build. Like, we add the keyword "module" somewhere to the driver we want to add as a modules. Then, we could add an option "make modules" and "make install_modules" so that they could be built/installed with the kernel. After all, modules ARE a part of the kernel... Looks like *really* nice idea. This would allow to solve "stale modules" problem at minimal cost. -Maxim To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
On Thu, Apr 20, 2000 at 11:24:49AM +0200, Stijn Hoop wrote: \begin{newbie question} So why aren't the modules built with the kernel instead of with the world? \end{newbie question} I bitched about this on committers the other day and marcel told me it was being worked on. -- Bill Fumerola - Network Architect Computer Horizons Corp - CVM e-mail: [EMAIL PROTECTED] / [EMAIL PROTECTED] Office: 800-252-2421 x128 / Cell: 248-761-7272 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
On Thu, 20 Apr 2000, Maxim Sobolev wrote: Then, we could add an option "make modules" and "make install_modules" so that they could be built/installed with the kernel. After all, modules ARE a part of the kernel... Looks like *really* nice idea. This would allow to solve "stale modules" problem at minimal cost. First we need to address the problem of "multiple kernels". Linux does this by having the modules associated with a particular kernel in a directory whose name is kernel dependent. After that, I personally think that we should treat the "kernel" as if it is just another module in the set. Rather than building a kernel, you ALWAYS build/rebuild/install the entire set. As long as the proper Makefiles are controlling the build, it won't be "expensive" for make to examine each module to verify that it is up-to-date. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: Stale modules (Re: panic in the morning)
On Thu, 20 Apr 2000 21:58:49 +0300, Maxim Sobolev [EMAIL PROTECTED] said: Looks like *really* nice idea. This would allow to solve "stale modules" problem at minimal cost. Long ago, in a galaxy far, far away... (well, OK, seven years ago, in FreeBSD 1.0) I was working on developing a loadable-module system for FreeBSD (which I eventually threw out when Terry Lambert donated his LKM implementation). In my version, the config file included statements like `module foo', which config(8) interpreted as `link the files listed in the files file as ``optional foo'' into a separate module foo.o'. We can't do this for everything marked `optional' because not all of those combinations actually make sense. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same [EMAIL PROTECTED] | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
panic in the morning
With a kernel built Apr 18th (and also today, Apr 19th) I'm getting panics when starting up in multiuser mode but also strange things happen when trying to boot /kernel -s In the latter case I was dropped into ddb after being prompted with something like mountroot I'm running i4b (isdnd) and the old PCI shims (ncr), FWIW. Also I don't know if it is now a bug or a feature: I compiled DDB into the kernel this morning after getting the panics (which were in strcmp() or strcpy() in the kernel, process was 'sh' in the panic log. After booting this DDB kernel first time the kernel didn't panic apparently but it was 'shutdown' smoothly (bufdaemon was stopped and kernel was halted). (looked like a feature but I couldn't see any benefit since I don't know whether it really panick'ed and if, where it did). All in all strange things are happening at the moment. Also ftpd tells me some module could not be loaded (auth_pam). I was hardly able to get up . Only with heavily hitting ^C through the daemon starup phase I got the kernel running. Maybe some daemon is the culprit (Mysqld, squid, sshd)? Will supply dmesg and CONFIG later - in a hurry at the moment... -- Chris Christoph P. U. Kukulies [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: panic in the morning
On Wed, Apr 19, 2000 at 12:44:22PM +0200, Christoph Kukulies wrote: With a kernel built Apr 18th (and also today, Apr 19th) I'm getting panics when starting up in multiuser mode but also strange things happen when trying to boot /kernel -s In the latter case I was dropped into ddb after being prompted with something like mountroot I'm running i4b (isdnd) and the old PCI shims (ncr), FWIW. Also I don't know if it is now a bug or a feature: I compiled DDB into the kernel this morning after getting the panics (which were in strcmp() or strcpy() in the kernel, process was 'sh' in the panic log. After booting this DDB kernel first time the kernel didn't panic apparently but it was 'shutdown' smoothly (bufdaemon was stopped and kernel was halted). (looked like a feature but I couldn't see any benefit since I don't know whether it really panick'ed and if, where it did). All in all strange things are happening at the moment. Also ftpd tells me some module could not be loaded (auth_pam). I was hardly able to get up . Only with heavily hitting ^C through the daemon starup phase I got the kernel running. Maybe some daemon is the culprit (Mysqld, squid, sshd)? Will supply dmesg and CONFIG later - in a hurry at the moment... OK, here goes: Copyright (c) 1992-2000 The FreeBSD Project. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 5.0-CURRENT #1: Wed Apr 19 12:13:34 CEST 2000 [EMAIL PROTECTED]:/usr/src/sys/compile/MONKAVMIFB Timecounter "i8254" frequency 1193182 Hz Timecounter "TSC" frequency 499946900 Hz CPU: Pentium III/Pentium III Xeon (499.95-MHz 686-class CPU) Origin = "GenuineIntel" Id = 0x673 Stepping = 3 Features=0x387f9ffFPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,PN,MMX,FXSR,XMM real memory = 134152192 (131008K bytes) avail memory = 126861312 (123888K bytes) Preloaded elf kernel "kernel" at 0xc037f000. Pentium Pro MTRR support enabled npx0: math processor on motherboard npx0: INT 16 interface pcib0: Intel 82443BX (440 BX) host to PCI bridge on motherboard pci0: PCI bus on pcib0 pcib1: Intel 82443BX (440 BX) PCI-PCI (AGP) bridge at device 1.0 on pci0 pci1: PCI bus on pcib1 pci1: NVidia Riva Ultra Vanta TNT2 graphics accelerator at 0.0 irq 11 isab0: Intel 82371AB PCI to ISA bridge at device 7.0 on pci0 isa0: ISA bus on isab0 atapci0: Intel PIIX4 ATA33 controller port 0xf000-0xf00f at device 7.1 on pci0 ata0: at 0x1f0 irq 14 on atapci0 ata1: at 0x170 irq 15 on atapci0 pci0: Intel 82371AB/EB (PIIX4) USB controller at 7.2 irq 10 chip1: Intel 82371AB Power management controller port 0x5000-0x500f at device 7.3 on pci0 pcm0: AudioPCI ES1371 port 0xe400-0xe43f irq 12 at device 10.0 on pci0 xl0: 3Com 3c905B-TX Fast Etherlink XL port 0xe800-0xe87f mem 0xe900-0xe97f irq 10 at device 11.0 on pci0 xl0: Ethernet address: 00:50:04:52:4f:73 miibus0: MII bus on xl0 xlphy0: 3Com internal media interface on miibus0 xlphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto ncr0: ncr 53c810a fast10 scsi port 0xec00-0xecff mem 0xe9001000-0xe90010ff irq 11 at device 12.0 on pci0 ncr0: driver is using old-style compatability shims fdc0: NEC 72065B or clone at port 0x3f0-0x3f5,0x3f7 irq 6 drq 2 on isa0 fdc0: FIFO enabled, 8 bytes threshold fd0: 1440-KB 3.5" drive on fdc0 drive 0 atkbdc0: Keyboard controller (i8042) at port 0x60,0x64 on isa0 atkbd0: AT Keyboard irq 1 on atkbdc0 vga0: Generic ISA VGA at port 0x3c0-0x3df iomem 0xa-0xb on isa0 sc0: System console on isa0 sc0: VGA 16 virtual consoles, flags=0x200 sio0 at port 0x3f8-0x3ff irq 4 flags 0x10 on isa0 sio0: type 16550A sio1 at port 0x2f8-0x2ff irq 3 flags 0x10 on isa0 sio1: type 16550A ppc0: Parallel port at port 0x378-0x37f irq 7 on isa0 ppc0: Generic chipset (EPP/NIBBLE) in COMPATIBLE mode ppbus0: IEEE1284 device found Probing for PnP devices on ppbus0: ppi0: Parallel I/O on ppbus0 ppi0: can't allocate irq device_probe_and_attach: ppi0 attach returned 12 lpt0: Printer on ppbus0 lpt0: Interrupt-driven port plip0: PLIP network interface on ppbus0 isic0 at port 0x1b40-0x1b5f,0x1720-0x173f,0x720-0x73f,0xf20-0xf3f,0x1340-0x135f,0x340-0x35f,0xb40-0xb5f irq 5 flags 0x4 on isa0 isic0: AVM A1 or Fritz!Card Classic unknown: PNP can't assign resources unknown0: PNP0200 at port 0-0xf,0x81-0x83,0x87,0x89-0x8b,0x8f-0x91,0xc0-0xdf drq 4 on isa0 unknown1: PNP0100 at port 0x40-0x43 irq 0 on isa0 unknown2: PNP0b00 at port 0x70-0x71 irq 8 on isa0 unknown: PNP0303 can't assign resources pca0: AT-style speaker sound at port 0x61 on isa0 unknown3: PNP0c04 at port 0xf0-0xff irq 13 on isa0 unknown4: PNP0c01 at iomem 0-0x9,0xfffe-0x,0x10-0x7ff on isa0 unknown5: PNP0c02 at iomem 0xf-0xf3fff,0xf4000-0xf7fff,0xf8000-0xf,0xd-0xd3fff on isa0 unknown6: PNP0a03 at port 0x294-0x297,0x4d0-0x4d1,0xcf8-0xcff,0x480-0x48f,0x4000-0x403f,0x5000-0x501f on isa0 unknown: PNP0501 can't assign
Re: panic in the morning
On Wed, Apr 19, 2000 at 03:24:05PM +0200, Christoph Kukulies wrote: On Wed, Apr 19, 2000 at 12:44:22PM +0200, Christoph Kukulies wrote: With a kernel built Apr 18th (and also today, Apr 19th) I'm getting panics when starting up in multiuser mode but also strange things happen when trying to boot /kernel -s In the latter case I was dropped into ddb after being prompted with something like mountroot I'm running i4b (isdnd) and the old PCI shims (ncr), FWIW. Also I don't know if it is now a bug or a feature: I compiled DDB into the kernel this morning after getting the panics (which were in strcmp() or strcpy() in the kernel, process was 'sh' in the panic log. After booting this DDB kernel first time the kernel didn't panic apparently but it was 'shutdown' smoothly (bufdaemon was stopped and kernel was halted). (looked like a feature but I couldn't see any benefit since I don't know whether it really panick'ed and if, where it did). All in all strange things are happening at the moment. Also ftpd tells me some module could not be loaded (auth_pam). I was hardly able to get up . Only with heavily hitting ^C through the daemon starup phase I got the kernel running. Maybe some daemon is the culprit (Mysqld, squid, sshd)? The panic is gone. I disabled the loading of the linuxulator in /etc/rc.local and re-brandelf'ed it. Unfortunately I have no longer any proof that it really *was* the linux emulator (should have made a copy). So what. -- Chris Christoph P. U. Kukulies [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: panic in the morning
On Wed, Apr 19, 2000 at 10:18:38AM -0400, Bill Fumerola wrote: On Wed, Apr 19, 2000 at 03:49:28PM +0200, Christoph Kukulies wrote: The panic is gone. I disabled the loading of the linuxulator in /etc/rc.local and re-brandelf'ed it. Unfortunately I have no longer any proof that Yikes! Nonesense. Wipe out. rubbish. Make unsaid! brandelf is for linux binaries, not the module. Had a bad day. it really *was* the linux emulator (should have made a copy). So what. Module out of date with kernel? I cvsup'ed, built world and kernel. Hhmm, actually I see no reason why there should be a problem since everything should be done by make world. -- Bill Fumerola - Network Architect Computer Horizons Corp - CVM e-mail: [EMAIL PROTECTED] / [EMAIL PROTECTED] Office: 800-252-2421 x128 / Cell: 248-761-7272 -- Chris Christoph P. U. Kukulies [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message