Re: Yaboot Config Question
On Fri, 6 Apr 2001, Matthew Grant wrote: }I am trying to get Debian 2.2 on a 9600/200MP. }When I reboot I get the Disk with the X in it. Which means I }need to configure My boot loader. I can't get into OF I get a }black screen. I think tis is because when one boots into OF (at least on the initial PCI machines) the default input/output is a serial port. You need to set the OF input/output variables to the keyboard and screen if you'd like to work in OF. This was done using a program called BootVars but you would need to run that from within MacOS. }I boot with the floppies and go to a shell I mount my root }partion (/dev/sda3) to /target and run ybin -b /dev/sda3. }It tells me there's no config file. or it tells me it's not a }HFS file system. Yaboot needs an HFS system and installs a series of files with all the attributes, so that Open Firmware thinks it's a MacOS system folder on an HFS file system. But this "System Folder" will boot whatever linux kernel you point to. }I am used to Lilo and debian on i386. would some one give me a }clue on what steps I am suppose to do to configure yaboot. }If I am doing this ass backwards send me a Link to were I can }learn the PowerPC way. Take a look at: http://linuxppc.org/hardware/blueg3/articles/yaboot.php3 I used those instructions to install yaboot on my PB G3 and it worked pretty well. Nothing about the instructions are LinuxPPC specific except maybe for the disk partitioning. And you may need to do that to create the small HFS partition where yaboot goes. But I was able to use these instuctions to put Potato 2.2 on my PB. HTH cheers vinai
Re: upgrade to MacOS 9.1, now lilo dissapears
On Sat, 7 Apr 2001, Grant Miller wrote: }At some point after I erased the OS 9.0.4 partition, a disk called bootstrap }showed up on the MacOS desktop. This disk has never appeared before in }MacOS. The bootstrap disk is at /dev/hda9 is named Apple_Bootstrap and is }of type HFS. } }Once the bootstrap disk showed up in MacOS, lilo did not appear when booting. } }I was able to boot off of the Debian CD and get into Linux and run ybin to }get lilo back, but whenever I booted into MacOS, lilo would get nuked. } }.. } }Is there a way to get the Apple_Bootstrap partition to not be seen by MacOS }even though it is a HFS partition? Boot off your CD and launch "Drive Setup". Then under the "customize volumes" menu options, select the Apple_bootstrap partition and set the options so that it is not automatically mounted at startup. Then you will have to reboot into Linux one more time and run ybin. If this does not work, will probably have to use pdisk to manually re- create the Apple_bootstrap partition. cheers vinai
Re: potato on PB G3
On Wed, 11 Apr 2001, Philipp von Weitershausen wrote: } I learned from HOW-TOs that yaboot (lilo-like tool for powerpc?!?) } preferably wants a small ext2 partition for /boot like you do } sometimes on ix86 platforms as well. Actually, for booting, it is not a small ext2 partition. It is an HFS partition, and when you are using pdisk to create it, you have to create it with type "Apple_Bootstrap". } I did that. Then it wanted to copy the base system but after having } done that, it wanted to it again, and again, and again. Then I } selected the next step from the menu manually - Installing "QUIK". } What is that? Whatever it is - it failed with a message that says } something like "QUIK is not for PowerPCs". Quik is an OF bootloader for old world machines. I am not sure if it works on the new world machines. } Can somebody please point me at a reasonable good tutorial/HOW-TO. } http://www.powermaclinux.org is nice, but doesn't take you from the } beginning till the end. Take a look at: http://linuxppc.org/hardware/blueg3/articles/yaboot.php3 None of the information there is LinuxPPC specific, except booting and drive parititioning. But I was able to use those pages (and the yaboot FAQ linked to on that page) to get Potato installed on my Lombard just a few days ago. The other quirk I found with the debian installer was that it did not like you skipping a step or making an error in a given step. I was able to get around the step I wanted to skip, but it was in a rather strange and non-systematic way. I basically went through the options till it let me continue ... cheers vinai
Re: potato on PB G3
On Wed, 11 Apr 2001, Philipp von Weitershausen wrote: } Ok, here's what I've done: } } I rebooted from the binary-1 CD and got into fdisk. I deleted } everything I could delete, so basicly just the partitions I had } created during earlier tries to install Linux. I wasn't able to } delete the driver partitions or the map itself. If you want MacOS on your machine, you should definitely NOT delete the driver partitions or the partition map, as these are how MacOS "knows" about your HD. } After having installed other standard packages, I had a look at the } hfsutils and figured that I could use those to format the two hfs } partitions. But although I labeled them and read the man pages again } and again, the MacOS installation system that I booted off the CD } doesn't recognize them as volumes (the hfsutils do though). If I start } the MacOS hard drive utility that does the "initialization" it lists } the CD drive and the ATA hard drive, saying that the hard drive is } "unrecognized" or "uninitialized" or something (dunno what the } original word was - my MacOS is German). The problem is: I don't dare } press that "Initialize" button because I once did that and it wiped } the complete hard drive. How can I get MacOS on this volume? If you deleted any of Apple's driver or patches partition, it might explain why the disk is not being recognized by Apple's utilities. What I have done in the past (on pretty much all of my installations) is to first partition the HD with Apple's utility (because I usually want to have MacOS around), and leave space for Linux at the end of the drive. Then from within the linux install, I delete that HFS partition and divide it up as I want to ... This might mean a re-install, but I am not certain. cheers vinai
Re: potato on PB G3
On Wed, 11 Apr 2001, Ethan Benson wrote: } if you want macos, then you do need to first partition with the macos } partitioner but you must create the placeholder partition for linux at } the BEGINNING of the disk *NOT* the end. } } otherwise you can't put the bootstrap partition first and i get } craploads of mail about how to fiddle with OF. always always put the } bootstrap partition BEFORE any macos partitions, Actually, you _CAN_. Your utility seems to work better than you give it credit for. My partition table looks like: # type name length base (size ) system /dev/hda1 Apple_partition_map Apple 63 @ 1 ( 31.5k) Partition map /dev/hda2 Apple_Driver43 Macintosh 54 @ 64 ( 27.0k) Driver 4.3 /dev/hda3 Apple_Driver43 Macintosh 74 @ 118 ( 37.0k) Driver 4.3 /dev/hda4 Apple_Driver_ATA Macintosh 54 @ 192 ( 27.0k) Unknown /dev/hda5 Apple_Driver_ATA Macintosh 74 @ 246 ( 37.0k) Unknown /dev/hda6Apple_Patches Patch Partition 512 @ 320 (256.0k) Unknown /dev/hda7Apple_HFS untitled 512 @ 832 ( 2.4G) HFS /dev/hda8Apple_HFS untitled 2 1740800 @ 5120832 (850.0M) HFS /dev/hda9 Apple_Bootstrap bootstrap 10240 @ 6861632 ( 5.0M) Unknown /dev/hda10 Apple_UNIX_SVR2 /1024000 @ 6871872 (500.0M) Linux native /dev/hda11 Apple_UNIX_SVR2 swap 77824 @ 7895872 ( 38.0M) Linux swap /dev/hda12 Apple_UNIX_SVR2 /usr 1843200 @ 7973696 (900.0M) Linux native /dev/hda13 Apple_UNIX_SVR2 /home2868784 @ 9816896 ( 1.4G) Linux native As you can see, I have the bootstrap partition after all my MacOS stuff and it works perfectly, as I am writing this on that very machine under Potato, running my own 2.2.18 kernel. This is on a Lombard PB. I don't think you will be getting "craploads of e-mail" about this ... cheers vinai
Re: potato on PB G3
On Thu, 12 Apr 2001, Philipp von Weitershausen wrote: } In case I should consider a complete reinstallation and I would do it } like you do, I'd clean the partition table with 'i', create the } bootstrap as 2nd, create my Linux partitions and then create the HFS } volumes, right? If you are going to keep MacOS around, I would still recommend you do the initial formatting with Apple's Utility first. The initial small partitions will contain the requisite info for MacOS, then the first partition AFTER those should be the Apple_Bootstrap partition for your Linux install, then after that, you should be able to divide up space between MacOS and linux as you desire. The only reason I did NOT put Apple_Bootstrap before MacOS partitions was because I just wanted to keep all my MacOS partitions on the first part of the disk, and my linux "stuff" on the latter. After playing with OF commands a little, if I lose the settings that allow the PB to boot of yaboot automatically, I can manually specify the OF commands, and then reset them for auto-boot from within Linux. Bottom line is: if you are going to keep MacOS around, don't touch the driver and patch partitions. But you can partition your disk to your heart's content after that ;-) Just keep some notes around, "in case" ;-) cheers vinai
openmotif and ppc .debs source ?
Hey Folks, Has anyone out there put together a .deb of openmotif ? If so - where would I be able to download it ? And a more general question - where can we find PPC .debs ? I recently moved from LinuxPPC to Debian PPC, so I am still trying to find my way around ... thanks all vinai
Re: ADB mouse not wroking on 9600/200
On Mon, 16 Apr 2001, Mark Simos wrote: }do i need to recreate that? how? # cd dev # mknod adbmouse c 10 10 if you need to recreate it. cheers vinai
reqs for vlc ?
Hi Folks, I wanted to try out playing DVD's under Potato PPC. So I grabbed a vlc .tgz archive from www.videolan.org. Everything seemed to compile, but on launching vlc, I get a window, the DVD drive spins for a while, but nothing else seems to happen. I let it spin for a while in case it is just really taking its time, but no go. Do I need a 2.4.* kernel with UDF support to run vlc ? Currently, I am using a custom self-built kernel, compiled from the 2.2.19 sources from kernel.org. I run the 3.3.6 version of XFree from potato 2.2r2 and this is all on a Lombard/400/DVD. Any ideas on why I can't use vlc at all ? Thanks. vinai
Re: reqs for vlc ?
On Mon, 7 May 2001, Michel [ISO-8859-1] Dänzer wrote: } How is it supposed to read the data from disc without UDF support? ;) I was hoping that vlc would take care of file system translation without the need for kernel support, like hfsutils do ... } There are Lombards with DVD drives? Sure it's not a Pismo? No: the original Lombard 400's had DVD drives, as well as a hardware DVD decoder. And it does have the HDI SCSI port at the back, so it is not a Pismo ;-) } Playing DVDs will be the opposite of fun without the XVideo extension } which is only in XFree86 4.x and only officially in the r128 driver. } } Besides, sound is choppy for me on a Pismo/400 with vlc, xine is } better but video isn't smooth; see the other recent thread about this. I hear you. It is just that all of the machines I have use some flavour of the ATI Mach64 chip, and the 2.2.** kernels allow me to do colour and resolution switching on the command line and at boot time. The kernel drivers in 2.4.+ have not been as well behaved (at least for me) and it was a little hard to jusitfy moving up to 2.4.+ without this hardware working ... Thanks much. cheers vinai
kernel options (was: reqs for vlc ?)
On Tue, 8 May 2001, Andrew Sharp wrote: }Michel Dänzer wrote: } }> Maybe there are UDF patches for 2.2 (or do DVD players depend on }> other stuff from 2.4?) . } } 2.2 kernels have had support for DVDs since at least 2.2.17. How would I turn on that support? I've tried the "Experimental Drivers" and I did not the option to turn that on ... This is with the stable 2.2.19 sources from kernel.org ? Also - I have another issue with kernel compilation. Using Ben H.'s PPC specific tree from penguinppc.org, I am in the following quandry: - If I DO NOT use the INPUT_LAYER option, XFree 3.3.6 won't recognize my mouse device ( /dev/mouse soft-linked to /dev/input/mice). On trying to launch X, I get the "Fatal server error" - no such device /dev/mouse. - If I DO use the INPUT_LAYER option, my keyboard does not work at all (this is from a console as I boot into runlevel 3) and I have to reboot and switch kernels to recover function. What do I need to do to have Ben's kernel (2.4.4) recognize my Lombard's keyboard ? Thanx. vinai
kernel packago 2.6.8-9 sources ?
On Tue, 11 Jan 2005, Sven Luther wrote: > Have you tried the debian powerpc kernels 2.6.8-8 and -9, which include > support for G5 thermal management ? > > Actually everyone should upgrade to 2.6.8-9 because of the various security > issues fixed by it. Hi, I was wondering where one would obtain the sources for this package? The link: http://packages.debian.org/testing/devel/kernel-build-2.6.8-powerpc still points to the 2.6.8-6 sources. Thanks vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Xserve G5 disks
Hi, I encountered a similar problem when I was trying to install Debian on a SATA drive in my 8500. I wanted to partition the disk for MacOS classic and Debian, but Apple's Disk utility would not recognize the drive. I ended up using my old debian installer to partition the disk, and used cp -ax to do a direct copy of my old system, partition by partition, to the new disk. In your case, you'd probably have to do the opposite - have Apple's disk utility partition the disk as you wish, and use your current Debian disk to format and copy over your partitions. You will probably have to edit /etc/fstab and /etc/yaboot.conf on the system you copied over to reflect the new disk layout, but it should be a quick and dirty way to get you up and running on your Xserve. cheers vinai P.S. How noisy is your Xserve in light to moderate usage ? On Thu, 13 Jan 2005, John Koskie wrote: > Hi, > > Does anyone know where to get a driver for the Hitachi SATA drives that > come with the Xserve G5. > > I successfully installed Sarge on one of the machines by placing a > Maxotr SATA drive in a G5 tower, installing onto that drive, compiling > the latest 2.6.8 kernel, then placing the Maxtor disk into the > Xserveit works fine. > > However, when I try the same with the Hitachi hard drive that comes wit > hthe Xserve the disk cannot be accessed. Discover cannot correctly > identify it. Th disk is labeled: Hitachi Deskstar HDS722580VLSA80 SATA > and "Apple HDD Firmware". > > I haven't filed a bug report yet since it seems I just don't know the > right driver. > > Thanks for any help > > John -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Last ditch on PlanB driver
On Sun, 16 Jan 2005, Sean Jewett wrote: > Here's a dumb question... what is videodev ? It could be that I have it > compiled in monolithic, I tend to stay away from modules if I know it's > something I'm going to need loaded in one way or the other (for > troubleshooting I did do planb as a module for ease of hacking purposes). Here's a snippet from my current kernel .config: # # Multimedia devices # # CONFIG_VIDEO_DEV is not set video_dev seems to activate the v4l API in the kernel, which then gives you access to mutlimedia input devices. IIRC - I think you do need to specify this option to get access to the planb device, as well as other video input hardware... cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: PPC Firewire support on 2.4.18-newpmac/Debian-3.0 r1 install CD?
Wolfgang, In the interests of time, I would recommend your trying the ext2 VFS extension for Mac OS X, which can be found at: http://sourceforge.net/projects/ext2fsx/ I've been using a powerbook drive in a firewire case on my Pismo PB for the better part of the last 2 years. This software is a kernel extension, and allows you to access ext2 volumes from within Mac OS X. I think it also works for ext3, but it ignores the journal. The only problems I've had with this software is from trying to work with files on the ext2 volumes from the Finder - that does (or did) not work. However, working with a command line from Terminal.app, or from an xterm in X11 should be just fine. And at least this way, you don't have to worry about Firewire support. >From what I recall, firewire support started working stably (for me at least) at ~ 2.4.16 to 2.4.18, so the Debian installer might be at that edge of support. If your disk is in another format, like xfs or jfs or reiser, this won't work ... Good luck ! cheers vinai On Thu, 20 Jan 2005, Wolfgang Pfeiffer wrote: > Hi All > > Sorry for asking here instead of browsing Google: But I have not even 24 > hours to rescue some configs from a broken hard disk. Data, i.e. Linux > configs, that are lost forever if I don't find some way to access them in > the next few hours on the old disk. I can't keep the old disk, as > Apple, as it seems, wants it back after I got a new one via the > warranty I still have with an "Apple Care Protection Plan" ... > > The details: > The hard disk broke on a PowerBook G4 (TitaniumIV). The repair service > already installed a new disk to this machine. I need access to the old > disk: The repair service will give me the chance to access the data on > the old disk via a firewire connection from the Titanium to the broken > disk. This will happen tomorrow noon. > > Current software on the Titanium: > > A very rudimentary Debian/testing system is installed: Just enough > packages to get the machine booting from the new hard disk, with some > additional stuff like curl, lynx etc.. The kernel version on this > system is a 2.4.18-newpmac. > > My idea now was to either boot the Titanium from the Debian/3.0 r1 install > CD, to start the first few installer steps and then to copy the data > from the old, broken via firewire connected disk to the new disk > inside the Titanium. Or, alternatively, simply boot the Debian system > from the new disk and try to connect it to the old, via firewire > connected disk outside the PowerBook. > > The problem: I do not know, whether the Debian 3.0 r1 installer > system - that is, the 2.4.18 kernel - will *see* the old, via firewire to > the Titanium connected disk. > > And I also don't know anything about firewire technology until now; I > just had a look to the 2.4.18 config on the Titanium /boot dir, and I > see several instances of "CONFIG_IEEE1394*" modules. This actually > means this kernel is ready for firewire connections? Positive? > > Excerpt from the current 2.4.18 config: > -- > CONFIG_IEEE1394=m > CONFIG_IEEE1394_PCILYNX=m > CONFIG_IEEE1394_OHCI1394=m > CONFIG_IEEE1394_VIDEO1394=m > CONFIG_IEEE1394_SBP2=m > CONFIG_IEEE1394_RAWIO=m > -- > > I consider installing a newer 2.6 (2.4?) kernel for the fresher > firewire drivers: > > Does anyone know where to get a readily installable, pre-compiled > ppc kernel that does not boot via initrd: I don't want this initrd stuff > on my machine, if possible: It is complicating things unnecessarily, > AFAICT .. > > And last question - important because I need to find a way to mount > the old disk outside: > How does the kernel call a hard disk that is connected via firewire: > /dev/hd[?] ... Or something else? > > Best Regards > > And thanks in anticipation > > Wolfgang > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: powerbook startup problem - folder with a question mark ???
> I have a powerbook g4 550 running all the time Debian unstable with > kernel 2.6.8-powerpc and OSX sometimes, mainly for spike. > > For the last week or so, when I start the machine instead of having > the yaboot screen appears I get a little folder with a question mark > inside... This doesn't happen every time I start the machine. > > When it happened the first time, I booted with a debian bootable > install cd and the hard drive was not recognized. (???) I booted > with the Apple Hardware Test, run all the tests and it didn't find any > problems. I waited a couples of hours and then the machine booted > fine. ? > > I have been trying to identify a pattern but no chance. I use my > laptop everyday and the problem never occurred in the morning at > startup. So far, I had the problem almost systematically after I > leave work and then go to some place and boot the machine. Note that > I always do a clean shutdown. I then wait a couple of hours and then > it boot fine. > > What could it be, firmware? hard drive? Is there any software to > test the machine? This is just a guess, but you might want to check, from within OS X that your machine is not trying to boot off a network device (this is in the "Startup Disk" pane in "System Preferences". Just make sure the boot device is your yaboot partition ... vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: albook g4 + external hard drive
On Thu, 10 Feb 2005, Tamas K Papp wrote: If you have specific models which are working with your linux, I would appreciate if you shared that. I am looking for a cable-powered one. I've had really good luck with an external case with Oxford's 911 chips, and housing an IBM 20 G laptop drive. If you're in the US (which I am guessing you are, from your email, take a look at OWC's site for some options). If you want something powered by the USB or firewire port, you'll be restricted to enclosures that house 2.5" (laptop) hard drives. Some cases even have both ports, so you can hook the drive up to whatever is available on the computer in front of you. As to the type of cable, I don't think it really matters, as with the laptop hard drives, you probably aren't going to reach the performance difference between 480 Mb/s USB and firewire. There is a difference, but only close to the limits of the transfer rates. The only thing to be aware of is that if you plug your drive into a USB chain with even a single 11 Mb/s USB device, the entire chain drops to that speed. If I understand the specs correctly, in a firewire chain, each device will "negotiate" with the bus to run at its own speed, while not slowing down the rest. With USB, everything drops to the speed of the slowest device. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: where's tfm?
On Mon, 14 Feb 2005, Brad Boyer wrote: I doubt it gets the kind of testing that the newer stuff does. I think only the 7x00 and 8x00 models had the chaos bridge. Hey Folks, I just set up an 8600 at home with the sole purpose of playing around with multimedia stuff on it. I'm actually quite willing to spend some time to try to get the control video device working if it means I could recover one of my PCI slots (the one in question, currently occupied by an ATI Mach 64 card). I have no device driver programming experience, but I've programmed with C for the last few years, and am quite willing to learn here, if some of the more knowledgeable folks are willing to answer sticky questions when they come up ;-) Even it means going back to Mac OS classic and taking apart system extensions, etc, I'm willing to give it a go ... So here are my first couple of questions. I had "testing" on this mac but I'm quite willing to wipe the install and start from scratch - and I probably will, just to test out the new installer... But if I do kernel work and I have questions, what sources would be the best to use ? And how would I get these sources ? thanx vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: iBook G4: repeatingly connect/disconnect of usb storage device
On Tue, 15 Feb 2005, David Pye wrote: On Tuesday 15 February 2005 15:04, steinm wrote: Feb 15 12:31:29 localhost kernel: hub 4-0:1.0: over-current change on port 2 That suggests to me that the amount of current it's trying to pull exceeds what the USB port is able/willing to provide. Try an externally powered drive? Or go through a powered USB hub - these aren't too bad. I got one that seemed to be pretty well built, as was able to supply power to all 4 of its ports. Cost was ~ $40 US. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Linux-PPC, 802.11g, and B&W new world power mac, internal v.90 modems?
> I have a PC with an 802.11b wifi card running debian linux. Currently > it's supported by the madwifi project, using the atheros chipset. It > also requires a PCI 2.2 compliant motherboard. Does the B&W g3 have a > pci compliant motherboard? Where can I find information like this > about other mac models? Take a look at Apple's developer docs for your particular mac. I think you might be out of luck though - I just had a quick look at the Xserve developer PDF and it's slots are 2.1 (this is the G$ Xserve). > Also, to what degree are the internal v.90 modems supported? What > device do they usually show up as? I think only the newer software based modems from conexant are not well supported. In the old G3s, that should still be a good ol' hardware modem, and should show up as a serial port (/dev/ttyS[0-4]). cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
install "testing" on an 8500 ?
Hey Folks, I'm about to wipe the Debian install on my 8500. I'm keeping a system backup on a separate SCSI drive, but I wanted to see if I could do an install, using either BootX + appropriate RAM disk with installer, or floppy images (if testing of this software is desired) to see if I can created a fully (non Mac OS dependent) bootable system on a SATA disk with quik. I've been following the threads on 8600 and G3 Old World installation problems over the last few days. I saw where I could grab floppy disk images for testing, but where can I download a RAM disk with the d-i installer ? This would be for the "testing" version of Debian ... Thanks all. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Which Kernel for Airport Support?
On Thu, 24 Feb 2005, Ben Hill wrote: It is an Airport Extreme unfortunately - looks like I'm SOL... If you're willing to get an external USB device, you might want to look into a thread within the last couple of weeks about 802.11 devices with linux compatibility. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Which Kernel for Airport Support?
On Thu, 24 Feb 2005, Eric Gaumer wrote: The airport card is backward compatible correct? In other words one can find a Mac OSX user and proposition him/her to swap cards (airport for airport extreme). I think a friend of mine found an airport card on Ebay. So here's another related question - anyone know of 802.11g cards which will fit in the 1st generation "airport" slot (like in the Pismo, and most likely the first TiBooks)? From what I've read, this slot is a 16 bit PCMCIA slot, so 32-bit cardbus cards don't work in it. Does such a beast exist - a 16-bit PCMCIA 802.11g card (linux compatible, of course) ?? cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Problem installing on PowerMac 8600
Hey Folks, No luck booting from a floppy or making it through the installer. Here is what I did, and hopefully, folks can point me in the right direction. I went to http://www.nl.debian.org/devel/debian-installer/ and the first thing I tried was the "floppy" set of powerpc images. However, my 8500 would not boot of a floppy made from these images. I tried writing the images with both rawwrite on a PC and with dd on my current installation of debian. No joy with the "boot.img" disk written either way... And I don't think it's a problem with the floppy drive itself - I put in the floppies in MacOS 9, and they showed up on the desktop and I was able to see the kernel file on the disk. I also tried to just use the kernel and ramdisk files from the "netinst" powerpc CD image. I was able to boot the machine, but I couldn't get it to do the network install I poked around Sven's links, and saw new kernel images. But I didn't see anything that looked like an installer ramdisk, or floppy images. Are there newer floppy images I can test ? Or an ramdisk which can do a network install ? thanx vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Problem installing on PowerMac 8600
On Fri, 25 Feb 2005, Patrick Bachmann wrote: Don't be so sure about that. After I installed MacOS 9.1 on the box, I was able to mount those floppies and take a look at them. But when I wanted to write to them, within MacOS I just got an error message saying, that it won't work. Cool, eh? If you still have MacOS on your drive you can still go the BootX way, if you want to get to some results quickly. ... I think that Holger's links point at the latest floppies. But I bet, Holger can tell more about that ;-> That sucks... Maybe I'll have to invest in a floppy cleaning kit too... I swear, I read Holger's message about 5 times last night, and thought it a redundant link against the location on the devel/debian-installer page. I wasn't paying attention as closely as I should have been ... I'll grab those images sometime this weekend out of Sven's directory and try again. I'll be glad when we have the miBoot issues squared away, so we can have proper boot images. Thanks again Patrick. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Problem installing on PowerMac 8600
> Don't be so sure about that. After I installed MacOS 9.1 on the box, I > was able to mount those floppies and take a look at them. But when I > wanted to write to them, within MacOS I just got an error message > saying, that it won't work. Cool, eh? > If you still have MacOS on your drive you can still go the BootX way, > if you want to get to some results quickly. Well, I poked around Sven's directories (even for the images he put up yesterday) but no luck. I found a few sets of floppies. I downloaded a few and I was able to see the penguin sitting with the mac, but after the kernel loaded, my display went black, and the system rebooted about 3 minutes later. I think my problem is that for whatever reason, my Mach 64 card does not at ALL play nice with the kernel frame buffer drivers. So I have to use the "No video" setting in BootX. If the kernels on these images do force some video mode, that might be the point of failure in my case. Back to the drawing board ... ;-) vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: And d-i initrd images (vmlinux and initrd.gz only) available too. (was Re: New 2.4.27 powerpc kernel packages ready, please test (including nubus and apus support) ...)
Hi, I did try the new floppy images, but due to issues with my video card not liking any video modes that the kernel wants to seem to enforce, I could not get those to work. The floppy kernel images would boot my 8500, but as soon as the kernel would start to boot, the video would disappear, and the machine would reboot a few minutes later. I did try a couple of the initrd images, but I could not get any to work either, probably (my guess) because you need a matching kernel to work these ramdisk images ? Will any of these images work with a kernel compiled from the 2.4.27 kernel.org sources ? vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: And d-i initrd images (vmlinux and initrd.gz only) available too. (was Re: New 2.4.27 powerpc kernel packages ready, please test (including nubus and apus support) ...)
On Mon, 28 Feb 2005, Sven Luther wrote: On Sun, Feb 27, 2005 at 03:04:11PM -0600, vinai wrote: I did try the new floppy images, but due to issues with my video card not liking any video modes that the kernel wants to seem to enforce, I could not get those to work. The floppy kernel images would boot my 8500, but as soon as the kernel would start to boot, the video would disappear, and the machine would reboot a few minutes later. This is nothing new though, and was present already in the normal d-i floppy-2.4 images, right ? Did you try the floppy ones or the floppy-2.4 ones ? I believe I tried both (went back and forth between floppy, floppy-2.4 and then ramdisk images with my own kernel) I did try a couple of the initrd images, but I could not get any to work either, probably (my guess) because you need a matching kernel to work these ramdisk images ? Will any of these images work with a kernel compiled from the 2.4.27 kernel.org sources ? No chance. But if you have a running linux install on it, you could try installing the 2.4.27 packages, and use quik/bootx to boot it. Should work, and this is really the test i am after. Okay - I was leaning towards just grabbing one of the debian kernels and doing what I do with my kernels - forcing the video setting with BootX. I'm also looking to migrate from 2.4 to 2.6 - I'm interested in learning about device driver programming. Could you point me to the location of the 2.4 and 2.6 kernels + ramdisks you want tested? I can at least try these with BootX, then proceed from that point ... cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: And d-i initrd images (vmlinux and initrd.gz only) available too. (was Re: New 2.4.27 powerpc kernel packages ready, please test (including nubus and apus support) ...)
> Try these ones : > > http://people.debian.org/~luther/d-i/new-2.4-kernels > > I guess both the powerpc/netboot/2.4 or powerpc/netboot/2.6 should do, > and work with BootX. Finally - a breakthrough !! ;-) The 2.4 kernel, even with all of the settings in BootX which work with my own kernel, will not boot my 8500. After I select that kernel image, it doesn't matter what setting, or if I use the matching ramdisk or not, but all I get is a black sreen + no system activity. I can't say that I tried every single iteration of options (though it felt like I came close :), but nothing seemed to work for me. The 2.6 kernel and ramdisk though, seemed to work perfectly, at least up till the part where I had to partition the system disks. I did have a little trouble getting the udebs from the mirrors but I think my ISP network is more the cause of that. However, since I wasn't actually ready to install to this system yet, I did not proceed past that point. Thanks again Sven. The 2.6 based installer seemed to find all of my hardware, so I don't foresee any major issues. However, when I do a backup of what I currently need on my existing installation, I will do a complete re-install, and I can post back here the results. That will not be till this weekend, most likely. Thanx vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Getting to command line from GDM or KDM...
On Tue, 1 Mar 2005, Charles Read wrote: Simple one here... how can I get back to the command line once the computer automatically starts GDM? You have a couple of options. I'm not at my debian box right now, but I think you can type (from an xterm, as root): # init 2 or # init 3 Failing those, look in /etc and it's sub-directories (most likely, the /etc/init.d directory), find the script which starts kdm/gdm, and call it with the "stop" switch # /etc/init.d/gdm stop I apologize if these are not the exact file names or locations, but my debian box is not in front of my right now, and these commands do work for what you asked on both RedHat and Suse systems. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: initrd.img ?
> If you compile in "enough" drivers, instead of making them modules, > you won't need an initrd. "Enough" means at least the driver for > your hard disk, the filesystem for your root partition, and maybe > your keyboard, mouse, and video. I don't have any experience with the 2.6 kernels, but I've booted 2.4 kernels over the last few years withour a ramdisk. The only modules I've had are for USB and firewire (which are not needed for boot on my 8500). vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: backing up before debian install
> 1. is it OK for debian and OS X if I just shrink the OS X partition >using partman, and put an ext3 filesystem partition after it? Or >should I put it before, etc, or the whole thing is not relevant? I'm not sure. There was a discussion on this list sometime within the last couple of months on this topic. I THINK partman can resize a HFS+ partition, but not with journaling enabled. Check the thread to make sure. > 2. before using partman, I would like to make a backup of the whole >disk. What tools do you suggest for that? > >I can now boot the sarge netinstall cd (current, daily release) and >mount partitions on my firewire/usb2 OWC 250 gb drive. I thought >of making a backup into a file on one of the partitions, using the >tools on the sarge install disk. Is this feasible? What tools >should I use? I thought of dd, but I don't know how could I make >it compare the data it has written with the disk, and I don't know >which device name to use for the whole disk (not just a partition). >If things don't work out, will I be able to restore from this? If you want to make a fully functional backup of your mac os x system, I would recommend looking at Carbon Copy Cloner from Mike Bombich (I think his website is bombich.com). It is donation-ware, but I've used it for the last year to year and a half to backup my laptop drive to an external firewire drive. It can create bootable partitions, complete disk images of existing partitions that can be backed up to CD or DVD, but not bootable disk images. Last time I checked, I think the site said it was for OS X 10.2, but I've successfully used it on 10.3.x for the last few months, and it still seems to work. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
d-i "testing" installer kernels + Power Mac 8500
Hi, I had a chance to test the d-i installer and wanted to provide a little feedback. This was with the kernel and ramdisk found in: http://people.debian.org/~luther/d-i/new-2.4-kernels/powerpc/netboot/ I did NOT use the ones in the 2.4 sub-directory, just this directory itself. I thought these were 2.6 kernels (if they were not in the 2.4 sub-directory ?)... Installation is to a Power Mac 8500, 1 G RAM, 2 hard drives: a 4.5 GB SCSI IBM disk, with Mac OS 9 (with BootX) and a backup of my old debian testing installation, and a WD Raptor SATA HD, 36 G, running off a FirmTek SATA PCI card. CPU is a G3 / 375 MHz card, but I also have a dual-CPU 604e card I want to try to use. Everything went pretty smoothly. Having been working with linux for ~ 12 years now, the original debian installer never gave me any trouble to really complain about. However, d-i seemed to really improve on the old installer, and things went very smoothly. I guess the part of the installation in which I deviated from a person who might be newer to linux was that I manually selected the disk partitions, as I wanted to set it up with the possibility of migrating this disk to a new world machine at some point in the future, in the never-ending upgrade cycle ;-) quik installation DID fail. However, I have not applied the OF patches this generation of machines seems to need, so I wasn't surprised that it failed at this point. The only real troubling part was that after the first reboot, where system initialization and installation is completed, the powerpc-2.6.8-smp (forgive me if that's not the exact package name, but it was the 2.6 PowerPC kernel, with SMP support) choked on my video setup again (ATI Mach 64 card, connected to an LCD, booting into a text console). It did boot the machine - I heard disk activity, and when I rebooted using my old 2.4 kernel, fsck reported errors on / (which was ext2, as per quik requirements). So I booted into the new d-i installation, using my old 2.4 kernel, and copied over my old /lib/modules/2.4.27 directory, and finished the setup. So here's my question - is the installer kernel I used (location above) a 2.4 or 2.6 kernel ? And if it is a 2.6, why did it work, and the boot kernel not (for console video)? Does anyone have a 2.6 kernel .config file for an old world machine ? I was planning on giving the non-smp version of this kernel a try, but I had to call it a night by the time I was able to create a working system. Any comments, suggestions, answers are welcome. Again, I have to say I was pretty happy and impressed with how smoothly everything went, save for the glitches I came across. And even those, I was able to overcome easily enough. Thanks to everyone who worked on all these components for the tremendous effort !! cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
newer kernels and console video in 7300-8600 macs ?
Hey Folks, Does anyone have console video working with any recent debian kernel (2.4 or 2.6 line) on "Old World" Power Macs ? I've tried both on-board video (the "control" device), an ATI card, 2.4.27 pmac kernels, 2.6.8 pmac kernels, and nothing seems to work. I can use my own 2.4.27 (compiled from kernel.org sources), and the kernel with the "testing" version of the debian installer allowed me to go through that installer. But other than that, when using a new kernel, I have to log in "blind" and start X ... Any pointers? If I wanted to hack away at these drivers myself, how or where do I start? thanx vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gnome silliness
> does anyone know why gnome would be demanding 24 bit color. > i would rather 16bt 800x600 over 24bit 640x480!! > > i cant find anytihng in ~/.gnome* /usr/share/gnome/* etc > > Dean This shouldn't have anything to do with GNOME, but with your xfree settings. Take a look at the file /etc/X11/XF86Config-4. Once your video card and monitor can handle it, there should be no issue with switching down in colour and up in resolution. vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#298565: bug in kernel-patch-powerpc-2.4.27
Package: kernel-patch-powerpc-2.4.27 When trying to boot with kernel-image-2.4.27-powerpc-pmac, console video disappears after the "arch: exit" kernel boot message. This happens on a Power Macintosh 8500, with both the motherboard video, and a PCI video card. The attached config file used with the 2.4.27 sources from kernel.org do NOT result in disappearing video.# # Automatically generated make config: don't edit # # CONFIG_UID16 is not set # CONFIG_RWSEM_GENERIC_SPINLOCK is not set CONFIG_RWSEM_XCHGADD_ALGORITHM=y CONFIG_HAVE_DEC_LOCK=y # # Code maturity level options # CONFIG_EXPERIMENTAL=y # CONFIG_ADVANCED_OPTIONS is not set # # Loadable module support # CONFIG_MODULES=y # CONFIG_MODVERSIONS is not set CONFIG_KMOD=y # # Platform support # CONFIG_PPC=y CONFIG_PPC32=y CONFIG_6xx=y # CONFIG_40x is not set # CONFIG_44x is not set # CONFIG_POWER3 is not set # CONFIG_POWER4 is not set # CONFIG_8xx is not set # CONFIG_8260 is not set CONFIG_PPC_STD_MMU=y CONFIG_ALL_PPC=y # CONFIG_APUS is not set # CONFIG_SPRUCE is not set # CONFIG_LOPEC is not set # CONFIG_SANDPOINT is not set # CONFIG_PPLUS is not set # CONFIG_PRPMC750 is not set # CONFIG_PAL4 is not set # CONFIG_GEMINI is not set # CONFIG_SMP is not set # CONFIG_ALTIVEC is not set # CONFIG_TAU is not set CONFIG_PPC_ISATIMER=y # # General setup # CONFIG_HIGHMEM=y CONFIG_HIGHMEM_START=0xfe00 CONFIG_LOWMEM_SIZE=0x3000 CONFIG_KERNEL_START=0xc000 CONFIG_TASK_SIZE=0x8000 # CONFIG_ISA is not set # CONFIG_EISA is not set # CONFIG_SBUS is not set # CONFIG_MCA is not set CONFIG_PCI=y CONFIG_NET=y CONFIG_SYSCTL=y CONFIG_SYSVIPC=y # CONFIG_BSD_PROCESS_ACCT is not set CONFIG_KCORE_ELF=y CONFIG_BINFMT_ELF=y CONFIG_KERNEL_ELF=y # CONFIG_BINFMT_MISC is not set # CONFIG_OOM_KILLER is not set CONFIG_PCI_NAMES=y # CONFIG_HOTPLUG is not set # CONFIG_PCMCIA is not set # # Parallel port support # # CONFIG_PARPORT is not set # CONFIG_GEN_RTC is not set # CONFIG_PPC_RTC is not set # CONFIG_PPC601_SYNC_FIX is not set CONFIG_PROC_DEVICETREE=y # CONFIG_PREP_RESIDUAL is not set # CONFIG_PROC_PREPRESIDUAL is not set CONFIG_PPCBUG_NVRAM=y # CONFIG_CMDLINE_BOOL is not set # # Memory Technology Devices (MTD) # # CONFIG_MTD is not set # # Plug and Play configuration # # CONFIG_PNP is not set # CONFIG_ISAPNP is not set # # Block devices # # CONFIG_BLK_DEV_FD is not set # CONFIG_BLK_DEV_XD is not set # CONFIG_PARIDE is not set # CONFIG_BLK_CPQ_DA is not set # CONFIG_BLK_CPQ_CISS_DA is not set # CONFIG_CISS_SCSI_TAPE is not set # CONFIG_CISS_MONITOR_THREAD is not set # CONFIG_BLK_DEV_DAC960 is not set # CONFIG_BLK_DEV_UMEM is not set # CONFIG_BLK_DEV_SX8 is not set CONFIG_BLK_DEV_LOOP=y # CONFIG_BLK_DEV_NBD is not set CONFIG_BLK_DEV_RAM=y CONFIG_BLK_DEV_RAM_SIZE=8192 CONFIG_BLK_DEV_INITRD=y # CONFIG_BLK_STATS is not set # # Multi-device support (RAID and LVM) # # CONFIG_MD is not set # CONFIG_BLK_DEV_MD is not set # CONFIG_MD_LINEAR is not set # CONFIG_MD_RAID0 is not set # CONFIG_MD_RAID1 is not set # CONFIG_MD_RAID5 is not set # CONFIG_MD_MULTIPATH is not set # CONFIG_BLK_DEV_LVM is not set # # Networking options # CONFIG_PACKET=y # CONFIG_PACKET_MMAP is not set # CONFIG_NETLINK_DEV is not set CONFIG_NETFILTER=y # CONFIG_NETFILTER_DEBUG is not set # CONFIG_FILTER is not set CONFIG_UNIX=y CONFIG_INET=y # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set # CONFIG_IP_PNP is not set # CONFIG_NET_IPIP is not set # CONFIG_NET_IPGRE is not set # CONFIG_ARPD is not set # CONFIG_INET_ECN is not set CONFIG_SYN_COOKIES=y # # IP: Netfilter Configuration # CONFIG_IP_NF_CONNTRACK=y CONFIG_IP_NF_FTP=y # CONFIG_IP_NF_AMANDA is not set # CONFIG_IP_NF_TFTP is not set # CONFIG_IP_NF_IRC is not set # CONFIG_IP_NF_QUEUE is not set CONFIG_IP_NF_IPTABLES=y CONFIG_IP_NF_MATCH_LIMIT=y CONFIG_IP_NF_MATCH_MAC=y # CONFIG_IP_NF_MATCH_PKTTYPE is not set CONFIG_IP_NF_MATCH_MARK=y CONFIG_IP_NF_MATCH_MULTIPORT=y CONFIG_IP_NF_MATCH_TOS=y # CONFIG_IP_NF_MATCH_RECENT is not set # CONFIG_IP_NF_MATCH_ECN is not set # CONFIG_IP_NF_MATCH_DSCP is not set # CONFIG_IP_NF_MATCH_AH_ESP is not set CONFIG_IP_NF_MATCH_LENGTH=y # CONFIG_IP_NF_MATCH_TTL is not set CONFIG_IP_NF_MATCH_TCPMSS=y # CONFIG_IP_NF_MATCH_HELPER is not set CONFIG_IP_NF_MATCH_STATE=y # CONFIG_IP_NF_MATCH_CONNTRACK is not set # CONFIG_IP_NF_MATCH_UNCLEAN is not set # CONFIG_IP_NF_MATCH_OWNER is not set CONFIG_IP_NF_FILTER=y CONFIG_IP_NF_TARGET_REJECT=y # CONFIG_IP_NF_TARGET_MIRROR is not set CONFIG_IP_NF_NAT=y CONFIG_IP_NF_NAT_NEEDED=y CONFIG_IP_NF_TARGET_MASQUERADE=y CONFIG_IP_NF_TARGET_REDIRECT=y # CONFIG_IP_NF_NAT_LOCAL is not set # CONFIG_IP_NF_NAT_SNMP_BASIC is not set CONFIG_IP_NF_NAT_FTP=y CONFIG_IP_NF_MANGLE=y CONFIG_IP_NF_TARGET_TOS=y # CONFIG_IP_NF_TARGET_ECN is not set # CONFIG_IP_NF_TARGET_DSCP is not set CONFIG_IP_NF_TARGET_MARK=y CONFIG_IP_NF_TARGET_LOG=y # CONFIG_IP_NF_TARGET_ULOG is not set CONFIG_IP_NF_TARGET_TCPMSS=y # CONFIG_IP_NF_ARPTABLES is not set # # IP: Virtual Server Configuration # # CONF
errors when compiling 2.6 kernel source
Hey Folks, I tried compiling a 2.6 kernel from sources (2.6.11 from kernel.org to be exact), and I got the following: LD init/built-in.o LD .tmp_vmlinux1 drivers/built-in.o(.text+0x4dbb8): In function `do_fd_request': : undefined reference to `sti' drivers/built-in.o(.text+0x4dd50): In function `set_timeout': : undefined reference to `save_flags' drivers/built-in.o(.text+0x4dd58): In function `set_timeout': : undefined reference to `cli' drivers/built-in.o(.text+0x4dd98): In function `set_timeout': : undefined reference to `restore_flags' drivers/built-in.o(.text+0x4eb2c): In function `grab_drive': : undefined reference to `save_flags' drivers/built-in.o(.text+0x4eb30): In function `grab_drive': : undefined reference to `cli' drivers/built-in.o(.text+0x4eb94): In function `grab_drive': : undefined reference to `restore_flags' drivers/built-in.o(.text+0x4ebc8): In function `grab_drive': : undefined reference to `restore_flags' drivers/built-in.o(.text+0x4ebec): In function `release_drive': : undefined reference to `save_flags' drivers/built-in.o(.text+0x4ebf0): In function `release_drive': : undefined reference to `cli' drivers/built-in.o(.text+0x4ec08): In function `release_drive': : undefined reference to `restore_flags' make: *** [.tmp_vmlinux1] Error 1 What's causing this ? cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: errors when compiling 2.6 kernel source
On Wed, 9 Mar 2005, Michael Schmitz wrote: I tried compiling a 2.6 kernel from sources (2.6.11 from kernel.org to be exact), and I got the following: What's causing this ? That looks like the swim3 driver. Last I heard, it hadn't been updated when some of the other parts of the kernel changed. Looks like it hasn't been touched in ages, then. BenH should correct me if I'm way off base, but you basically need to replace, at a minimum: save_flags(flags)+cli() by local_irq_save(flags) restore_flags(flags)by local_irq_restore(flags) sti() by local_irq_enable() for uniprocessor use (are there any SMP capable machines with SWIM3??). If that compiles and runs, you can proceed to using proper spinlock functions (a spinlock has already been added to the driver but isn't currently used). See swim_iop.c and floppy.c for guidance. I will take a look. And to answer your questions, I believe the SWIM3 is the driver for the last floppy drive on Power Mac and Power Mac G3 systems, and matching Mac Clones from that period. The 7300-9600 macs (and their matching clones) do have 2x CPU cards available for them. I actually have a 2x 180 604e card, and there are 2X G4 upgrade systems as well for these machines. Thanks all. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: errors when compiling 2.6 kernel source
On Wed, 9 Mar 2005, Michael Schmitz wrote: Looks like it hasn't been touched in ages, then. BenH should correct me if I'm way off base, but you basically need to replace, at a minimum: save_flags(flags)+cli() by local_irq_save(flags) restore_flags(flags)by local_irq_restore(flags) sti() by local_irq_enable() for uniprocessor use (are there any SMP capable machines with SWIM3??). if that compiles and runs, you can proceed to using proper spinlock functions (a spinlock has already been added to the driver but isn't currently used). This code needs to be SMP safe (see Bug#271517). You'll need proper That's why I asked about SMP machines with SWIM3. locking primitives (ie. spin_lock_irqsave(), etc...). The locks will compile away on UP machines. Sure. Keep in mind that this isn't a simple case of search and replace. You should understand what data needs mutual exclusion. The stray sti() looks like a hack. Enabling interrupts with sti() before return is bad The PC floppy driver does things in much the same way, from a quick glance. Most notably, the spinlock is never used in that driver either. Same for the swim_iop driver. I conclude there's not much need for mutual exclusion here. Yes, I'm aware that the interrupt handling isn't the same for Intel and PowerPC hardware ... practice because your function could be returning to a function that expects interrupts to still be disabled. The driver should make sure this doesn't happen, naturally. Seems to have worked at some time in the past. I'm aware that my suggestions are just a first thing to try. If the machine freezes, or other bad things happen, look more closely. Don't try this on mission critical hardware :-) Well, this machine finished up its last "critical mission" (helping me through my thesis ;-) more than a year ago, so it can be "beaten up" to test things like this. My better half is already smiling at me due to the number of reboots this machine has suffered through in just the last few weeks testing kernels and installers :) I'll take a look at the SWIM3 driver, with all your suggestions in mind, and I'll post back here with the results, good, bad, or desperate pleas when I get lost in kernel code :-) vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: pre-release 2.6.11-0.1 debian/powerpc kernel.
On Thu, 10 Mar 2005, Leandro Guimaraens Faria Corsetti Dutra wrote: Em Tue, 08 Mar 2005 20:10:10 +0100, Sven Luther escreveu: am in the process of uploading a pre-release 2.6.11 debian/powerpc kernel to http://people.debian.org/~luther/powerpc/2.6.11-0.1 Is there a guide somewhere on how to use your packages? I couldn't add it to sources.list and have no idea on how to use the source packages... will try to use the image, but wanted to check a bug workaround with SynCE... Download the .deb file to your system, then do: # dpkg -i file_you_downloaded.deb Check the man/help pages for dpkg. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: HELP! - problems with 2.6.8 kernel on PM G4/350 (PCI graphics)
On Fri, 11 Mar 2005, Daniel E. Jonsen wrote: 1) Lots of complaining about the firewire interface (maybe this machine has a "Rev. 0" pcilynx chip that the 2.6.8 kernel's driver doesn't like?). After booting, I can't access any firewire devices. I need to use a firewire HD for periodic backups. You might try disabling pcilynx support, if you are compiling your own kernel. I think all Apple's firewire controllers were OHCI compatible, but someone else should correct me if I'm wrong. I've used a firewire drive on my old 8500 with a PCI card for backup, and it has served me very well. 2) Complaints about the (lack of) serial ports - "Illegal UART type - Undefined". Is there a "clean" way of telling the kernal that there are no serial ports, e.g., a "dummy" serial port driver I can use? Can I unset CONFIG_SERIAL_CORE and CONFIG_SERIAL_PMACZILOG, or do I need them even though this machine doesn't have any serial ports? Be careful here though - IIRC, this machine came from a time when there were still honest-to-goodness hardware modems in PCs and Macs. If this machine has an internal modem, it probably shows up as a serial port. If you want to keep that functionality, you want to keep serial port support. Check dmesg output for occurences of "ttyS" 3) I'm not sure why, but the kernel loads the MESH module, even though this machine, as far as I know, has no built-in SCSI. I do have an Adaptec (aic7xxx) PCI card installed, but it loads the aic7xxx module for that (and calls it SCSI bus 1, while MESH is SCSI bus 0). Can I leave MESH out of a kernel compiled specifically for this machine? Should be okay. In order to cure these problems and to have a leaner, meaner kernel, I installed the 2.6.8 kernel source deb that came on the Sarge distro CDs. When I attempt to compile, the process exits abnormally as follows: CC init/version.o LD init/built-in.o LD .tmp_vmlinux1 sound/built-in.o(.init.text+0xb68): In function `dmasound_awacs_init': : undefined reference to `pmac_xpram_read' make[1]: *** [.tmp_vmlinux1] Error 1 make[1]: Leaving directory `/usr/src/kernel-source-2.6.8' make: *** [stamp-build] Error 2 g4-debian:/usr/src/kernel-source-2.6.8# AWACS support probably hasn't been updated for this kernel version yet. Take a look at a thread from a couple of days ago about my own problems compiling a 2.6 kernel. In my case, I'm missing floppy drive support ... You can try disabling sound support for the time being. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Hello
> On Friday 11 March 2005 22:33, Michael wrote: > > Sorry. I have tried the Yellow-Dog, Mandrake 9.1. Debian 3.0r2 > "Woody" installations. I have also tried the Rock linux LiveCD > image which appeared to have the same symptoms. > To me, sort of a newbie, it appears as if as soon as the kernel takes > over from the boot loader that it tries to load an incompatible > monitor > driver, or display mode. Having just gotten back from another attempt > to install Woody, with a new problem this time, the only way I could > even boot the disk was to boot install24-safe. I am not sure the diff > between install24 and install24-safe but I think it is important to my > problem. You're not going crazy - I thought I was too for a while ;-) It seems to be problem with recent kernels, and the common thing your case and mine share (from my simplistic viewpoint) is that we both have PCI ATI video cards. I have not been able to successfully boot a recent Debian kernel and have the video work. I filed a bug with debian and CC'ed it to this list. It's in the arcives. The closest I came was with a 2.6 installer that Sven pointed me to recently (again in the list archives) in which the kernel that came with the installer had working video, but the installed kernel did not, though it did allow the machine to boot, and me to log in blindly. At that point, I could start X11 manually, and get video. But that can't work for a text installer. If you're really stuck, contact me off list and we can see about getting you a working kernel. I've been using debian for a while, and linux for a while longer than that, so I've had my own kernels which work for my hardware. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ethernet card question
On Mon, 21 Mar 2005, Rolando Abarca wrote: is there a way to know if there's a cable connected to the ethernet card?... I ask this because I have a lombard powerbook, with a dhcp configured eth0, but if I boot GNU/Linux (ubuntu hoary) it stops waiting for a dhcp server, I would like to modify the networking script, so if there's no cable present on the card, don't configure the eth0 device, only the loopback. Is it possible? Something like OS X does when it's booting. I thought this was alrady the default behaviour in Debian (I run the testing tree). When my 8500 boots up, if I don't have the network cable plugged in, it will pause while the networking scripts try to acquire a connection, but after about a minute or so, it will give up, continue the boot cycle, and just have the loopback interface ... I can't see why this would be different on Ubuntu. I remembered when I was setting up Warty on a PC, the DHCP timeout would be __VERY__ long. But I thought that was just because it was an older machine ... I would recommend you give it 2-3 minutes to see if DHCP can figure out you don't have a network connected. If you have to wait longer than 5 minutes, I'm not sure where the trouble would be ... cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Update to swim3 driver for 2.6 (was: errors when compiling 2.6 kernel source)
On Wed, 9 Mar 2005, Michael Schmitz wrote: Looks like it hasn't been touched in ages, then. BenH should correct if I'm way off base, but you basically need to replace, at a minimum: save_flags(flags)+cli() by local_irq_save(flags) restore_flags(flags)by local_irq_restore(flags) sti() by local_irq_enable() for uniprocessor use (are there any SMP capable machines with SWIM3??). If that compiles and runs, you can proceed to using proper spinlock functions (a spinlock has already been added to the driver but isn't currently used). This code needs to be SMP safe (see Bug#271517). You'll need proper That's why I asked about SMP machines with SWIM3. locking primitives (ie. spin_lock_irqsave(), etc...). The locks will compile away on UP machines. Sure. Keep in mind that this isn't a simple case of search and replace. You should understand what data needs mutual exclusion. The stray sti() looks like a hack. Enabling interrupts with sti() before return is bad The PC floppy driver does things in much the same way, from a quick glance. Most notably, the spinlock is never used in that driver either. Same for the swim_iop driver. I conclude there's not much need for mutual exclusion here. Yes, I'm aware that the interrupt handling isn't the same for Intel and PowerPC hardware ... practice because your function could be returning to a function that expects interrupts to still be disabled. The driver should make sure this doesn't happen, naturally. Seems to have worked at some time in the past. I'm aware that my suggestions are just a first thing to try. If the machine freezes, or other bad things happen, look more closely. Don't try this on mission critical hardware :-) Hi Folks, I tried Michael's suggestions of replacing the older functions with the "local_irq_*" equivalents, and those seemed to work for my 8500. So my next question is there documentation on making this SMP-safe ? I think I can find a copy of the older "linux device drivers" book, but is that worth it ? Or is googling the way to go here ? Sorry it took so long to get to this - was travelling for work. Thanks vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: PowerMac floppy support is broken in 2.6.11.5
On Thu, 24 Mar 2005, Christoph Ewering wrote: Just wanted to let you know, that the floppy support in 2.6.11.5 is broken. I do not know what is with previous kernels. This is what I get when I switch on PowerMac floppy support, if I switch it of, the kernel compiles with out any error. CC init/version.o LD init/built-in.o LD .tmp_vmlinux1 drivers/built-in.o(.text+0x70b90): In function `do_fd_request': : undefined reference to `sti' drivers/built-in.o(.text+0x70d50): In function `set_timeout': : undefined reference to `save_flags' drivers/built-in.o(.text+0x70d58): In function `set_timeout': : undefined reference to `cli' drivers/built-in.o(.text+0x70d98): In function `set_timeout': : undefined reference to `restore_flags' drivers/built-in.o(.text+0x71b68): In function `grab_drive': : undefined reference to `save_flags' drivers/built-in.o(.text+0x71b6c): In function `grab_drive': : undefined reference to `cli' drivers/built-in.o(.text+0x71bd0): In function `grab_drive': : undefined reference to `restore_flags' drivers/built-in.o(.text+0x71c04): In function `grab_drive': : undefined reference to `restore_flags' drivers/built-in.o(.text+0x71c28): In function `release_drive': : undefined reference to `save_flags' drivers/built-in.o(.text+0x71c2c): In function `release_drive': : undefined reference to `cli' drivers/built-in.o(.text+0x71c44): In function `release_drive': : undefined reference to `restore_flags' make: *** [.tmp_vmlinux1] Fehler 1 I'm trying to help fix this. See the thread titled "errors when compiling 2.6 kernel source" and the message I just posted. This is also a reported bug (# 271517). I also just posted a message before the mailing list sent me your post ;-) cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
swim driver question for 2.6
Hi Folks In trying to update the swim3 driver for 2.6, I was also pointed to the swim_iop.c driver (activated by CONFIG_BLK_DEV_SWIM_IOP kernel option). Does anyone know if this driver is still needed for 2.6 on 68k machines ? It doesn't seem that is driver has been updated either (at least in the kernel.org tree). What floppy drivers do folks with 68k macs use ? vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: swim driver question for 2.6
On Sat, Mar 26, 2005, Brad Boyer wrote: > On Sat, Mar 26, 2005 at 10:31:01AM -0600, vinai wrote: > >> In trying to update the swim3 driver for 2.6, I was also pointed >> ot the swim_iop.c driver (activated by CONFIG_BLK_DEV_SWIM_IOP >> kernel option). Does anyone know if this driver is still needed >> for 2.6 on 68k machines ? > > The swim_iop driver never worked, honestly. I was the most recent > person to work on it, but I never got read/write running. The only > thing I got to work was all the ioctl type stuff, like eject and > format detection. The bulk data transfer is poorly documented. > > It's only used on three specific models where there is no direct > access to the SWIM chip without disabling ADB. However, for those > three machines we won't be able to use floppies without this > driver or a replacement for it. > >> It doesn't seem that is driver has been updated either (at least >> in the kernel.org tree). What floppy drivers do folks with 68k >> macs use ? > > It's only recently that anyone got any floppy driver working on a > 68k mac in linux. The swim3 driver isn't quite right for any of > them, since they all have either an older version of SWIM or a > different chip entirely. Okay - thanks for the confirmation. I had a look through the code for both drivers, and they seemed pretty self-contained (no cross dependencies), but I just wanted to make sure. I'm only now just learning how to crawl around in kernel code ... Another question, if folks don't mind: As Micheal suggested in an earlier thread, some older interrupt routines had to be replaced with the 2.6 version (local_irq_**). This was pointed out not to be SMP safe, so I read a little, and started again with substitution, replacing the local_irq_* lines with the spin_lock equivalents. However, I could not find the spin_lock_ equivalent of local_irq_enable. The closest I found was spin_lock_irqsave in floopy.c ... Is this what should be used to replace local_irq_enable ? With all of the other irq handling functions replaced, except that one (local_irq)enable), the machine locked up and I had to do a hard reboot. thanx vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: swim driver question for 2.6
So it looks like I spoke to soon. Replacing all the 2.4 interrupt handling routines from 2.4 to their 2.6 equivalents allowed the swim3 module to compile and load, but not function properly. In trying to mount a floppy as either a user or as root, I get the following: /dev/fd0: Input/output error mount: /dev/fd0: can't read superblock So with the changes to the driver (cli, sti and restore_flags replaced by local_irq equivalents), did I miss something with the initialization of the drive ? Also, is there any way to make this driver module removable ? Once it's loaded, the module cannot be removed, and /proc/modules says that swim3 is [permanent] ... Anything I can do to change this, at least for testing and debugging ? cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apt-build errors
> On Apr 2, 2005, Tommy Trussell <[EMAIL PROTECTED]> wrote: > > ... > > For the record, here is my current apt-build.conf > > --- > > build-dir = /var/cache/apt-build/build > repository-dir = /var/cache/apt-build/repository > Olevel = -O2 > #mcpu = -mcpu=Medium > mcpu = -mcpu=740 > options = " " > make_options = " " Just out of curiousity, what CPU is designated the "740" ? I though G3's were 750, G4's == 74**, and G5 == 970. And the olders 604/604e/ 603/603e/601 used those exact strings ... vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: aty128fb
> I am still unable to start my computer with video=aty128fb and X is > still complaining about no frame buffer devices, even though I have > /dev/fb1 - 9 or 8. I am utterly frustrated at this, and I have found > no help in anything I have read on line I can pass video=aty128fb > argument on the command line, but I don't think it does anything. I > even installed fbset and ran the command to output the configuration > needed for X, but didn't understand it, is someone having similar > problems? Not just a problem for you. Somewhere along the line, I also lost support for my Mach64 video card, even doing text boots into a console. This loss of support was in the Debian kernels - which is why I didn't notice when that happened - as I normally compile my own kernels. The best result I've had is from an installer kernel - it was able to text-boot into d-i and let me install testing a couple of months ago. But the kernel it installed for actually running the system again did not give me any video on boot. So what I do now is boot blindly, login and start X. I have my old X settings from when I had fully supported video, and the video drivers in X still work, thankfully. And I keep around a 2.4 kernel which has full (working) support for all my hardware, just in case I need to do diagnostics on boot. vinai vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: aty128fb
On Wed, 6 Apr 2005, Benjamin Herrenschmidt wrote: On Tue, 2005-04-05 at 21:24 -0500, vinai wrote: I am still unable to start my computer with video=aty128fb and X is still complaining about no frame buffer devices, even though I have /dev/fb1 - 9 or 8. I am utterly frustrated at this, and I have found no help in anything I have read on line I can pass video=aty128fb argument on the command line, but I don't think it does anything. I even installed fbset and ran the command to output the configuration needed for X, but didn't understand it, is someone having similar problems? Not just a problem for you. Somewhere along the line, I also lost support for my Mach64 video card, even doing text boots into a console. This loss of support was in the Debian kernels - which is why I didn't notice when that happened - as I normally compile my own kernels. The best result I've had is from an installer kernel - it was able to text-boot into d-i and let me install testing a couple of months ago. But the kernel it installed for actually running the system again did not give me any video on boot. So what I do now is boot blindly, login and start X. I have my old X settings from when I had fully supported video, and the video drivers in X still work, thankfully. And I keep around a 2.4 kernel which has full (working) support for all my hardware, just in case I need to do diagnostics on boot. Does upstream 2.6.12-rc2 work for you ? Ben. Don't know. I can give it a try tonight. Where do I pull the sources for this kernel from ? vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: kernel 2.6.12-rc2
> ok I did some reading, now the 2.6.12-rc2 patch needs to be > applied > directly to the 2.6.11 kernel source, right? I tried applying it to > the > debian kernel-source-2.6.11 tree but some of the patch failed, and > ultimately make-kpkg failed. So I am guessing the debian tree is a > little different. I also read Linus's directions, do I really have to > Symlink /usr/src/linux/include/asmppc to /usr/include/asm and > /usr/src/linux/include/linux to /usr/include/linux and > /usr/src/linux/include/scsi to /usr/include/scsi having already > symlinked /usr/src/(kernel source) to /usr/src/linux? Because on my > previous compiles I never did that. If so cool, if not I need to > figure > out how to fix those things. > --Mike Actually, it needs to be applied to the 2.6.11.6 sources. Just did exactly that and compiling right now. Got the 2.6.11.6 and patch for 2.6.12-rc2 from kernel.org. vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: kernel 2.6.12-rc2
> vinai wrote: >> >>Actually, it needs to be applied to the 2.6.11.6 sources. Just did >>exactly that and compiling right now. Got the 2.6.11.6 and patch for >>2.6.12-rc2 from kernel.org. >> >>vinai >> > ok I have both the latest stable patch to bring 2.6.11 to 2.6.11.6 > and the rc2 patch, I can just apply the 2.6.11.6 patch to the 2.6.11 > source and the rc-2 patch to that right? And is the Debian 2.6.11 > tree package that installs kernel-source-2.6.11 really different than > the 2.6.11 kernel source from kernel.org? Or was my prorblem just > tryring to apply the patch rc2 before the 2.6.11.6 patch? Not quite sure. I actually just grabbed the entire 2.6.11.6 source tree and applied the rc2 patch. Just because of how much work goes on with the kernel from different groups, I would not try to mix source trees and patches from different places. There's no easy way to reconcile what one group does versus another, unless you use something like CVS or Bitkeeper, and we all know where that's going ... So you should be fine applying the patches just like you said. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: aty128fb
>> > Does upstream 2.6.12-rc2 work for you ? >> > >> > Ben. >> >> Don't know. I can give it a try tonight. Where do I pull the >> sources >> for this kernel from ? > > ftp.kernel.org > > Ben. Pulled the 2.6.11.6 sources and applied the .12-rc2 patch. Still no video after booting. The last thing visible on the screen is: setup_arch: bootmem arch: exit cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: aty128fb
On Thu, 7 Apr 2005, Benjamin Herrenschmidt wrote: On Wed, 2005-04-06 at 20:55 -0500, vinai wrote: Does upstream 2.6.12-rc2 work for you ? Ben. Don't know. I can give it a try tonight. Where do I pull the sources for this kernel from ? ftp.kernel.org Ben. Pulled the 2.6.11.6 sources and applied the .12-rc2 patch. Still no video after booting. The last thing visible on the screen is: setup_arch: bootmem arch: exit Hrm... and aty128fb is built into the kernel ? Very strange ... How many connectors on you card ? Is the "other" connector working ? Ben. No - I do NOT have aty128 compiled in, neither static or as a module. I did not think dependencies work that way (i.e. it'd SEEM more logical if newer hardware depended on older hardware, but not older hardware on the drivers for the newer stuff). I will compile it in and report back tonight. Also - the Mach64 card has 2 ports - a VGA port, and the old Mac-style video port. I haven't had a monitor with that type of connector for a while (the old mac style) and I've been using a VGA LCD panel. But I will see if I can find an adaptor to let me use the VGA monitor on that port and also report the results of that. Thanks again Ben. cheers vinai P.S. The 2.6.12-rc2 fixed some other problems which I'll detail in my next note to deb-ppc. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
2.6.12-rc2 "fixes" 8500 floppy and sound ...
Hi Everyone, Some good news to report back. Yesterday when testing if 2.6.12-rc2 helped with video, I used an older .config in which the swim3 driver and dmasound_pmac were compiled in. The kernel compiled, and both of these components now work! There were still warnings when both of these components were being compiled. I believe the messages for the swim3 driver were about the previously troublesome interrupt handling routines "being deprecated", but the module still compiled and worked - and the code for the module itself is unchanged from 2.6.11, where it did _NOT_ compile. I missed the messages about the errors around dmasound_pmac, but I'll take a closer look tonight and this weekend when recompiling to add aty128 to my kernel. Is there a way to track the output of the kernel compilation process, other than redirection using the ">" or ">&" constructs ? Anything folks want me to look out for, or check ? I'd been stuck in trying to figure out what to do about the non-working swim3. I had put in a lot of debugging code into the routines there to track the event loops, but had a hard time making sense of it and seeing where the problem was Hopefully, this will help. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: aty128fb
>> No - I do NOT have aty128 compiled in, neither static or as a >> module. I did not think dependencies work that way (i.e. it'd >> SEEM more logical if newer hardware depended on older hardware, >> but not older hardware on the drivers for the newer stuff). I >> will compile it in and report back tonight. >> >> Also - the Mach64 card has 2 ports - a VGA port, and the old >> Mac-style video port. I haven't had a monitor with that type >> of connector for a while (the old mac style) and I've been using >> a VGA LCD panel. But I will see if I can find an adaptor to let >> me use the VGA monitor on that port and also report the results >> of that. > > Well, I misunderstood your setup and though you had a r128, sorry. If > atyfb is indeed not working properly for you, you should report that > on linux-fbdev mailing list. I don't know this driver that well, and > those chips are real cans of worms, but somebody there may be able to > help. My apologies for the confusion I caused between you and Micheal. I'd wanted to say there was another person with non-working console video using older ATI graphics cards. I'll try to get over to that list and see if someone can help out. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2.6.12-rc2 "fixes" 8500 floppy and sound ...
> Some good news to report back. Yesterday when testing if 2.6.12-rc2 > helped with video, I used an older .config in which the swim3 driver > and dmasound_pmac were compiled in. The kernel compiled, and both > of these components now work! There were still warnings when both of > these components were being compiled. I believe the messages for the > swim3 driver were about the previously troublesome interrupt handling > routines "being deprecated", but the module still compiled and worked > - and the code for the module itself is unchanged from 2.6.11, where > it did _NOT_ compile. I missed the messages about the errors around > dmasound_pmac, but I'll take a closer look tonight and this weekend > when recompiling to add aty128 to my kernel. Is there a way to track > the output of the kernel compilation process, other than redirection > using the ">" or ">&" constructs ? > > Anything folks want me to look out for, or check ? I'd been stuck in > trying to figure out what to do about the non-working swim3. I had > put in a lot of debugging code into the routines there to track the > event loops, but had a hard time making sense of it and seeing where > the problem was Hopefully, this will help. First off - Thanks to Charles for how to direct standard output to multiple places. Worked like a charm. But I need to revise what I initially posted above. I had applied the 2.6.12-rc2 patch to the 2.6.11.6 sources, which was wrong, but still produced the results above. Tonight - I went back and applied the patch to the 2.6.11 tree, and it (of course) patched without a hitch. However, now, the swim3 driver no longer worked. So I manually changed out the references to the old interrupt handling routines to the newer uni-processor ones, and things seem to work fine again. Sound still works with this patching. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: applying patch
On Fri, 8 Apr 2005, Tamas K Papp wrote: I want to try 2.6.12-rc2. I downloaded and extracted the latest 2.6.11 from kernel.org, and tried to apply the 2.6.12-rc2 patch. I get messages like this: patching file drivers/media/video/adv7170.c Reversed (or previously applied) patch detected! Assume -R? [n] What should I answer here? Or should I just wait for 2.6.12 final? I am trying to get sleep support for my powerbook. When I applied the .12-rc2 patch last night, I did not have to answer any questions. I did when I was trying to patch the 2.6.11.6 sources, but the .12-rc2 patch applied cleanly to 2.6.11. This is what I did: - unpacked linux-2.6.11.tar.gz -> linux-2.6.11 directory - unzipped 2.6.12-rc2-mm2.bz2 in the same directory when the 2.6.11 source directory was. - entered the linux-2.6.11 directory and ran the command: $ patch -p1 < ../[name of patch file] and it applied cleanly, without any error. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: applying patch
On Fri, 8 Apr 2005, vinai wrote: On Fri, 8 Apr 2005, Tamas K Papp wrote: I want to try 2.6.12-rc2. I downloaded and extracted the latest 2.6.11 from kernel.org, and tried to apply the 2.6.12-rc2 patch. I get messages like this: patching file drivers/media/video/adv7170.c Reversed (or previously applied) patch detected! Assume -R? [n] What should I answer here? Or should I just wait for 2.6.12 final? I am trying to get sleep support for my powerbook. When I applied the .12-rc2 patch last night, I did not have to answer any questions. I did when I was trying to patch the 2.6.11.6 sources, but the .12-rc2 patch applied cleanly to 2.6.11. This is what I did: - unpacked linux-2.6.11.tar.gz -> linux-2.6.11 directory - unzipped 2.6.12-rc2-mm2.bz2 in the same directory when the 2.6.11 source directory was. - entered the linux-2.6.11 directory and ran the command: $ patch -p1 < ../[name of patch file] and it applied cleanly, without any error. I may have spoken too soon (doing that too often of late - getting to be a REALLY bad habit ;-) ... I just tried grabbing the linux-2.6.11 tree from kernel.org, and the .12-rc2 patch, and they no longer seem to apply cleanly. One thing I did notice is that all the files in the patch directory are dated today, so they have been updated in the last 24 hours. I'll check when I get back home, but I think the patch file I downloaded (and which worked) is probably very different from the one now available from the kernel.org front page ... The newer one also seems a lot smaller - the one I used last night seemed to be around ~ 18 MB (uncompressed). The one now available is 9 - 10 MB ... I can report back when I have something more definitive. vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: applying patch
> I want to try 2.6.12-rc2. I downloaded and extracted the latest > 2.6.11 from kernel.org, and tried to apply the 2.6.12-rc2 patch. I > get messages like this: > > patching file drivers/media/video/adv7170.c > Reversed (or previously applied) patch detected! Assume -R? [n] > > What should I answer here? Or should I just wait for 2.6.12 final? > I am trying to get sleep support for my powerbook. Just verified Wolfgang's procedure - everything patches cleanly if done in order he mentioned. I haven't complied or boot a kernel from these sources either though. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Kernel-imaage-2.6.11
On Mon, 11 Apr 2005, Sven Luther wrote: On Sun, Apr 10, 2005 at 10:55:49PM -0700, Michael wrote: ... And just to clarify I know it says this because I have booted with video=ofonly, logged in, and then run startx, but the framebuffer I am suposed to use is aty128fb. But my fear is that if the video=ofonly is passed to the kernel at login that that will stop the x-server from being able to access the aty128 framebuffer, could someone tell me if my fears are justified in that? Seems a reasonable conclusion. Since as far as i know, offb doesn't let himself be unloaded for another framebuffer device to take on. So I'm curious here myself - I've been using the "No video driver" option in BootX for the last few years (ever since the 2.4 series) - but I've gotten hardware acceleration in X. And it IS there because when I used my old AppleVision 1710 on the ATI and the 8500 on-board video (which does not have hardware acceleration support) my eyes could tell - judged by how quickly they became strained using X (less than half an hour using X with the unsupported card, versus an entire day when the ATI card drove the monitor). Isn't the "No video driver" option in BootX equivalent to video=ofonly ? Thanx vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
acquiring an interrupt for swim3 floppy driver
Hi Folks, I was wondering what was the best/cleanest (and SMP-safe) way to grab an interrupt ? swim3 currently uses local_irq_enable, but from past discussions on the list, it did not sound like it was SMP safe. All of the other local_irq_ code seems to be easily enough bracketed off by equivalent spin lock functions, but I haven't found one that works for just the irq_enable ... I looked at floppy.c, and tried to backtrack where that driver got an interrupt, and it went back to asm/floppy.h, which seemed to want hard coded values for floppy IRQ and DMA, which I did not think we had to do on PPC ? So that was the dead end that brought me back here ... thanx vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
atyfb gives penguin logo, but no more ...
Hi again folks, I've been trying to get at the lack of console video with my Mach 64 card on my 8500. Per Michael's suggestions, I again tried forcing resolutions through BootX arguments, but with no luck. I did have a little bit of luck elsewhere though. I was looking at my .config, where CONFIG_FRAMEBUFFER_CONSOLE=m. However, I don't use a RAM disk with my kernel. I changed this to =y, recompiled and rebooted with the new kernel. Now, after the arch: exit message, I get the penguin logo !! However, still no text. So ,my question is - what does the rest of the boot process need, in terms of fonts, etc ? It looks like I have video, just nothing to show ... cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Strange frequency reported by the kernel
> I received a G3 upgrade card for a PowerMac 9500 as a gift from one of > the subscribers of this list (hi, Tormod!). > > I replaced the dual 180/MP G04e daughter card with this new G3 upgrade > and the last time that I booted with MacOS 8.1, it reported the card > being having a G3 450MHz. > > I have eliminated MacOS from my system and went Debian only, but one > curious thing I saw with kernels from the 2.2, 2.4 and 2.6 series is > that they report the frequency of the processor as 195MHz. The > bogomips count presented in /proc/cpuinfo indicates 796.67, which > would lead me to believe that there is something slightly wierd with > this setup. > > Here is the complete /proc/cpuinfo on this machine: > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - > processor : 0 > cpu : 740/750 > clock : 195MHz > revision : 131.0 (pvr 0008 8300) > bogomips : 796.67 > machine : Power Macintosh > motherboard : AAPL,9500 MacRISC > detected as : 16 (PowerMac 9500/9600) > pmac flags: > L2 cache : 512K unified > memory: 64MB > pmac-generation : OldWorld > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - > > I also learned that I should enable the L2 cache of this card to get > better performance. Is this indeed necessary? > > I am open to any questions regarding my setup. > > > Thanks in advance for any help, Rogério Brito. Rogerio, >From what I've remembered reading, the speed detection was always problematic on the 1st generation PCI machines with G3 and G4 upgrades, and I think the problem was never cleanly resolved as it did not cause any real trouble ... From the bogomips rating, it looks like you have a ~ 400 MHz CPU... To get the L2 cache activated, I think you need to write a value to the l2cr variable which lives somewhere in the /proc directory. I think the "find" command should point you to it. And if not - google is your friend ;-) I just had a couple of questions, if you didn't mind... What kernel did you use to get SMP support for your 2x 604e card (I tried on my 8500 and was unsuccessful). And what are you using to boot linux without MacOS on this machine ? cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: acquiring an interrupt for swim3 floppy driver
On Wed, 13 Apr 2005, Brad Boyer wrote: On Tue, Apr 12, 2005 at 07:37:16PM -0500, vinai wrote: I was wondering what was the best/cleanest (and SMP-safe) way to grab an interrupt ? swim3 currently uses local_irq_enable, but from past discussions on the list, it did not sound like it was SMP safe. All of the other local_irq_ code seems to be easily enough bracketed off by equivalent spin lock functions, but I haven't found one that works for just the irq_enable ... I just looked at the copy of swim3.c that I have (from 2.6.8.1), and I can't figure out why it's even using local_irq_enable(). It doesn't seem like a safe thing to do anyway. I'd suggest trying to figure out why it was doing that before, since it's possible you can just drop it. I looked at floppy.c, and tried to backtrack where that driver got an interrupt, and it went back to asm/floppy.h, which seemed to want hard coded values for floppy IRQ and DMA, which I did not think we had to do on PPC ? So that was the dead end that brought me back here ... Are you converting it over to being a proper macio driver while you're there? If so, you can just ask the macio layer for the various correct addresses and IRQ values with the functions on the macio_dev object. Otherwise, you would get them out of the device_node just like it does in the old version. The mainline floppy code is pretty stupid and PC specific. It expects to have most of the information compiled into the kernel, which doesn't really work for mac hardware. No - my attempts right now were to get the driver to compile (by getting rid of the old sti/cli/save_flags/restore_flags). I was able to do this by getting rid of these functions and replacing them with the local_irq_ equivalents, and the driver compiled, and even worked ;-) However, in trying to make the driver SMP-safe, I haven't had any luck. Replacing the local_irq functions by their spin_lock equivalents have not resulted in a working kernel for me. I thought this was due to the local_irq_enable. x86 floppy.c did not use such a routine, but the way that driver acquired an interrupt value led me back to defined values for FLOPPY_IRQ and FLOPPY_DMA - which did not look right for PowerPC... Thanks for the pointers - I'll look into device_node and see what's in there ;-) cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Compile errors Patching 2.6.11.7 to 2.6.12-rc2
On Thu, 14 Apr 2005, Brian W. Carver wrote: download 2.6.11.7 from kernel.org and put it in /usr/src tar -jxf linux-2.6.11.7.tar.bz2 rm /usr/src/linux ln -s linux-2.6.11.7 linux download patch-2.6.12-rc2.bz2 bunzip2 patch-2.6.12-rc2.bz2 mv patch-2.6.12-rc2 /usr/src/linux cd /usr/src/linux patch -p1 Download benssoundpatch from http://sharealike.net/debian/ppc/benssoundpatch and mv benssoundpatch /usr/src/linux patch -p1 cp /boot/config-2.6.8-powerpc .config make oldconfig #accept the defaults on everything and [1-32] your answer is 1. make-kpkg clean make-kpkg --append-to-version=.050414 kernel_image Take a look at Wolfgang Pfeiffer's answers from a few days ago with the thread named "applying patch". Basically, a lot of the patches on the kernel.org page are not meant to be applied to the 2.6.11.? realeases, just the vanilla 2.6.11. I made the exact same mistakes. Wolfgang's message has the URL's as to where to get the proper patches that apply cleanly. After applying those, you should be able to get the various new hardware patches to apply more cleanly. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Recompile a driver, without recompiling the whole kernel
On Thu, 14 Apr 2005, Wolfgang Pfeiffer wrote: Hi All Often I see patches for kernel drivers here: Is there a more or less painless way to recompile a driver for some kernel/kernel sources, without the need to recompile the whole kernel? ... The solutions I googled so far don't look very promising ... All I want is compiling the new driver and copy it to somewhere below /lib/modules/`uname -r`/ Impossible desires? Well, you're going to have to do a compile at least once. And once you don't clean the kernel tree, when you work on the driver, you should just be able to recompile the driver, and move it over to /lib/modules. This is from my (very limited) experience. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: problem with powerbook g3: black screen on boot
On Mon, 18 Apr 2005, Rainer Gutkas wrote: Maximilian Gerlach schrieb: Hi everyone, I bought a Powerbook G3 on ebay. When I start it I hear the apple sound but the display stays black - any suggestions? :/ Thanks, Maxi Which kind of G3 Powerbook, Wallstreet, Lombard, Pismo, or if you don't know that which connections has it got on the back: Wallstreet got SCSI connections, Pismo got 2 USB and 2 Firewire, Lombard to be honest I don't know, Wallstreets have SCSI + ADB ports Lombards have SCSI + USB Pismos have Firewire + USB HTH vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apple dvd pccard
On Wed, 20 Apr 2005, Rolando Abarca wrote: does anyone knows if there's a driver for the apple dvd pccard (the mpeg2 decoder)? that one that was for wallstreet and lombards... regards, Last I checked, there was no driver for this hardware. I did own a Lombard at one point, and I'm not sure if this device even showed up on an lspci scan. The only MPEG decoders I know of with good linux support are the ones in the Hauppauge line of TV encoder/decoder cards, and the HDTV card specfically sold for linux... There are most likely others, just not what I've come across when I was searching a while back... cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apple dvd pccard
On Wed, 20 Apr 2005, Dean Hamstead wrote: sadly the mpeg decoder on the ati cards in imac dv's (and my ibook dual usb) are unsupported thats the sole reason my old imac dv runs osx man id be willing to throw cash at someone to reverse engineer it. pity ben is a busy man as it is. someone clone him. Well, that's a little bit of a different situation. On the older Mac Powerbooks, with 64-bit ATI graphics chips, there was a dedicated MPEG-2 hardware decoder, either a PCMCIA card, or a chip on the laptop motherboard. I don't know if the 64-bit chip lacked the horsepower, or just basicfunctionality, to play MPEG-2 streams at the appropriate frame rate, but additional hardware was needed. On your macs, you most likely have at least 128-bit graphics hardware (the lombard was the last laptop to have 64-bit, and either the old Beige G3 tower or Blue and White G3 tower was the last desktop to come with 64-bit). In 128-bit chips and above, MPEG-2 decoding is done in the video card hardware. I don't think ATI ever published info about this function for any of their cards, which is why you need a machine with enough CPU power to decompress DVDs. However, when I did have my lombard (G3 400), I thought either Xine, mplayer, or VideoLan (most likely the last one, as that's what I've been using the most) would play DVDs _JUST_ at the level of "watchability" (i.e. if there was a lot of rapid action or scene changes, video would stutter. But for a less variable scene, it'd be fine). And that was quite a few years ago, on OS X... Having a G4 helps, as I think the writers of some of these packages have been able to utilize Altivec to assist with the MPEG-2 decoding. But this is where things stand (AFAIK ;-) cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Question on Sonnet's Tempo Trio Card
> Since I have recently received a G3 upgrade card for an old PowerMac > 9500 that my uncle once lent me and since now I have 192MB of memory, > the system is a bit more usable than it was ealier. > > Since the disk is now a bottleneck, I was thinking of upgrading the > system with a Sonnet's Tempo Trio card, that has information available > at <http://www.sonnettech.com/product/tempo_trio.html>. Rogerio, Take a look at this page: http://www.xlr8yourmac.com/OSX/sonnet_ide_cards_and_10_2.html According to the information there, the Trio has a Promise chipset, and I believe the linux kernel does indeed have support for this family of silicon. Just as an aside bit of info - I tried a dual-function PCI card (USB and firewire) in my 8500, which I could never get to work. I don't know if it was my 8500, the card, or the kernel (this was with a 2.4.2? kernel), but I always had problems with the firewire functions on that card. And the problems were not confined to a single card. I asked my vendor if I could try another brand of multi-function card, and there was the same effect. I will also admit that all my slots on this machine are filled - max RAM, SATA card (which works really well) and an ATI card, and even moving things around (swapping card between slots) did not help either. Hope this info helps. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Question on Sonnet's Tempo Trio Card
Rogerio, > Thanks for letting me know that there may be problems ahead. I thought > that I would import one of these as it sounds like a good deal (less > expensive than getting 3 cards, not to mention higher availability of > slots and, the main factor, PCI IDE cards here aren't easy to find). This was my rationale as well, though you don't have the limit on slots that I do ;-) I wanted to switch over to a USB mouse and keyboard, and keep the firewire functionality, but was unable to do so. I had been using a firewire only PCI card to back up my internal drives, and that worked very well and stably, after the firewire was supported was put into the kernel. You might still want to look at individual cards. The USB and firewire cards are relatively cheap these days (even USB cards with USB 2.0/high speed, or whatever they're calling it these days). Your expense would be with the hard drive card. Plus, if any of the functions break, you would just have to replace a single card ... >> I will also admit that all my slots on this machine are filled - max >> RAM, SATA card (which works really well) and an ATI card, and even >> moving things around (swapping card between slots) did not help >> either. > > Just curious, how much RAM do you have on your machine? And which SATA > card do you use? SATA devices aren't common here, but they may become > easier to find as times goes on. And does your card have support for > MacOS X? Does it perform well under Linux? This was the machine I did the bulk of my dissertation work on - so it has 1 G RAM. The SATA card I use is the one shown here: http://eshop.macsales.com/item/Firmtek/FTST1S2/ (My pologies for the vendor link - it was the most convenient one). It is supported by linux, and with the 10K rpm Raptor drive, I get rates of 17-18 MB/s (using the hdparm -tT test) whereas before (with my older SCSI-2 drives), I used to get around 4-5 MB/s. Boot into linux is much faster, and the drive is no longer the bottleneck in this system ;-) So yes - linux performance is great ! The page said the card IS supported in Mac OS X. Another caveat however (you had to see this coming - didn't you ;-) was that I could not get this drive to be bootable with Mac OS 9 on my 8500. The versions of drive setup that came with 9.1 (last version to work on this machine officially) as well as 9.2.2 can't initialize the drive. I think this has more to do with the versions of OS 9 partitioning utility not being able to recognize the geometry of the Raptor disk, rather than anything to do with the Firmtek SATA card. I was able to partition the drive just fine in linux with mac-fdisk. What I thought about trying, but never got around to it, was to put the drive in a mac running OS X, use OS X's Disk Utility to initialize and partition the drive with Mac OS 9 support. But I never did get the opportunity. So what I have to do now is boot Mac OS 9 from an old SCSI drive and use BootX to lauch my linux installation on the Raptor drive. It's been chugging along for the better part of the last couple of years like this, and seems to work pretty well. HTH cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2.6.12-rc4 lots faster than any kernel before?
> On Sun, 2005-05-15 at 13:40 +0100, Erik Chakravarty wrote: > >> Never mind >> >> I see you have disabled CONFIG_MOUSE_PS2 > > That is bogus on a mac. > > Ben. I know PS2 ports don't exist on macs, but don't pegasos/genesei boards have PS2 ports ? vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Powerbook G3 Pismo - resolution change problem
>> First, isn't (PB G3 Firewire)==(PB G3 Pismo)? > > Don't think so, but I could have gotten that wrong in my head. It is - the Pismo (Y2K powerbook, last "black powerbook") IS the PB G3 firewire - I'm typing this on one right now :) It is the only G3 PB with Firewire. Everything before had SCSI, and all later powerbooks have G4 CPUs. >> Anyway, correct me if I'm wrong, but before I removed mac os (9) >> from my laptop's hdd, there was some nifty little panel applet >> there that allowed me to change the resolution to a lower one, >> and it did in fact change it. Same thing with games on that mac. >> Or is it faulty eprom in my head? > > With my machine here, I've never been able to change the resolution > from 1024x768. Before I upgraded to 10.2, I ran 9.x. When I used the > 'panel applet' to change the resolution, it would always stay in > 1024x768, but center the 800x600 or 640x480 and the 'unused' pixels > would look like a black border. I can't speak for OS 9 as it's been years since I used it on this PB, but even now, under OS X 10.3, when you select 640x480 or 800x600, you use all the pixels (i.e. the entire LCD area) - the 1024 x 768 pixels are interpolated, so the LCD shows just 640x480 or 800x600 worth of info. But it does use the entire LCD area - there is no dark border of unused pixels. I don't know if the smoothing is done in the graphics hardware, driver or the OS itself though. But lower, "full screen" resolutions are possible ... They just don't look as nice... cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: X Window Server setup: video card Bus ID.
> I'm making the switch from Ubuntu to Debian 'Sarge' (Ubuntu's too > unstable for my liking). After doing a netinstall of Sarge and using > apt-get to retrieve x-window-system, kde, and kdm, I set up the > x-window-system preferences. It says I need to enter a Bus ID for my > video card (because I'm on PPC, even though I'm only using one vard). > I used lspci and found my ATI Rage 128 AGP (8MB) with a Bud ID of > :00:10.0, but it says I have to convert that to hexadecimal to > put it in. Am I doing this right so far, and, if so, how do I convert > this to hexadecimal. I've scoured Google for advice, but PPC support > isn't very strong in the Linux community from what I've seen. Figure > this is the best place to ask. Any help? On my 8500, with the on-board video chip (control) and a PCI ATI Mach 64 card, lspci (just taking the output for the video cards) gives: :00:0e.0 VGA compatible controller: ATI Technologies Inc 3D Rage Pro (rev 5c) 0001:01:0b.0 Non-VGA unclassified device: Apple Computer Inc. Control Video And from my /etc/X11/XF86Config-4, the accompanying BusID lines are: BusID "PCI:0:14:0" and BusID "PCI:1:11:0" I think the instructions you have were slightly reversed - you don't convert the lspci output to hex. From my example above, that output is already in hex. Convert the hex values to decimal to put in your XF86Config file. HTH cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Oops in snd-powermac on powerbook 3400
> On Thu, 2005-05-05 at 23:35 -0400, Keenan Pepper wrote: >> I fixed it! >> >> I changed it to use the same device node for sound as for >> chip->node, and it started beeping at me! Apparently sometime >> between 2.6.8 and 2.6.12-rc3 the KIOCSOUND ioctl was implemented, >> because it sure didn't beep before. >> >> BTW, this is the first time I've ever fixed a kernel bug. > > Ok, I think your patch was whitespace damaged, so I re-did it by hand > ;) > > Here's the result. Please test. Also, other people around, please test > if with this patch, snd-powermac works, especially on oldworld > machines! > > I'll submit upstream tomorrow hopefully. > > Ben. My apologies for only now replying to this, but I can confirm the patch that originally accompanied the above e-mail also worked fine on my 8500, I'm listening to an audio CD as I type this ;-) I just got back to playing around, and patched 2.6.11 up to 2.6.12-rc4 and did not see this patched included, so I thought I'd try to include it myself locally. Nice work all :) cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Antialiased fonts in gtk1.x applications
> I already know the answer: gdkxft :) > > ... > > The output I get while compiling the 1.5 package is this: > > [cut coming right after the end of ./configure] > > In file included from /usr/include/X11/Xft/Xft.h:41, > from gdkxft.c:29: > /usr/include/ft2build.h:56:38: freetype/config/ftheader.h: No such > file or directory > > ... > > I have the ftheader.h file and is positioned in > > /usr/include/freetype2/freetype/config/ftheader.h > > I kind of see the solution (an "include" problem) but I don't know > how to do this. In the Makefile, check to see that /usr/include/freetype2 is part of the search path. If not, you can modify where the "inlcude PATH" is to include that subdirectory. Alternatively, /usr/include is pretty much default for include PATHs. Where you have the: #include that's causing you trouble, try changing it to: #include as now, it'll just have to look inside /usr/include (which it should already do) to find the necessary file. HTH. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: On getting quik to work with oldworld machines (was: Re: Strange frequency reported by the kernel)
On Wed, 25 May 2005, [iso-8859-1] Rogério Brito wrote: On May 25 2005, Peter 'p2' De Schrijver wrote: quik doesn't know about ext3 journaling. So better use a normal ext2 partition for /boot or be very sure that the /boot partition is unmounted properly before rebooting. Yes, that was the point of my message. OTOH, the sarge d-i says that quik won't work with a /boot partition separated from a root partition. This is, anyway, a major annoyance regarding the way quik works. It would be so nice if it were possible to have a unified bootloader for both oldworld and newworld machines that understood symlinks (the config file for quik mentions that it doesn't know how to follow symlinks). I'll try your suggestions this weekend. Right now, I have a SCSI disk in my 8500 which boots Mac OS 9 + Linux via BootX. To test this, I'll probably have to remove the SCSI disk, and set up my linux drive to be able to boot via quik. On the bright side, I have a couple of spare SCSI disks that need to be formatted, and I can probably borrow them for a weekend to be able to test this out more completely, without me (completely) destroying my old installation :) I'll report back when I have some test results - good or bad. Thanks Rogerio. vinai
Re: bootable cds on oldworld [was Re: Backing an HFS partition]
On Fri, 27 May 2005, Hans Ekbrand wrote: Can the mklinux CD:s boot oldworld macs? If so, you could perhaps copy the partitions on the mklinux CD instead of copywrited Apple ones? Yes - they most certainly can. I started off playing with linux on the power pc with Mk Linux on a 7500. I don't know if it will boot a NuBus machine though. The other issue is, I think the drivers on those CDs are still the property of Apple. Remember Mk Linux came from, and for a while, was supported by Apple. I think in the realm of truly free loaders (for Mac machines) has only BootX (though that need MacOS to run), quik, and yaboot. The verdict is still out on miBoot. Can anyone comment on the status of that piece piece of software ? cheers vinai. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bootable cds on oldworld [was Re: Backing an HFS partition]
On Fri, 27 May 2005, Rainer Gutkas wrote: I try to get miboot working on a G3 Powerbook Wallstreet for a while now. So this Pb has got a 2 Gig harddisk and spending 100-200 MB's for a MacOs Bottloader seems like a big waste for me. Actually evry MB I can save is of great worth. But actually I didn't manage to get it right until now, so if you can give me suggestions what I could do to get miboot up and running (for a linux only booting system) it would be great! I don't have any experience with miBoot - just with BootX, and a little with yaboot. I've been trying to get my 8500 "off mac os 9" (similar to your case) without success. I also heard/read somewhere a while back that there were licensing issues with miBoot, which prevented it being included with Debian. I was just wondering what the source of those issues were, and if they were any closer to being resolved ... ? vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bootable cds on oldworld [was Re: Backing an HFS partition]
On Fri, 27 May 2005, Rainer Gutkas wrote: Could you discuss 4) with me? And if it is off list it's ok too. I let's say hypotheticall that I got a laptop which I wan't a good operationg system to run on and for me that's linux. All I need is 4) the theoretical howto as practical as possible. I don't know of how much help I can be. My only real experience is with BootX, which it also sounds you're trying to get away from. Have you seen this page: http://penguinppc.org/bootloaders/ Details are a little scarce, but between these and a google search, you may find enough info to get you going ... I myself, am trying to get rid of OS 9 + BootX, and will mostly likely end up using quik. Another person on this list with a machine similar to mine has reported some success with this combo. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Kernel image not recognized
On Thu, 2 Jun 2005, Federico 'Pain' Pistono wrote: Hi, I've dowloaded the kernel linux-2.6.11.10. then: $: cd /usr/src/ $: tar jxvf linux-2.6.11.10.tar.bz2 $: ln -sf linux-2.6.11.10 linux $: cd linux $: make menuconfig $: make $: make modules $: make modules_install $: make bzImage $: cp /arch/ppc/boot/images/vmlinux.elf-pmac /boot/vmlinux2.6.11.10 If you are doing this on a Macintosh, you don't need to "make bzImage" - there is no size restriction on the loadable kernel on macs like there is for older x86 machines (don't know about newer ones). Just doing a "make vmlinux" should be fine (and the "vmlinux" should be from the root of the linux kernel source code directory). If you're trying this on a PReP/CHRP/other non-Apple PowerPC machine, the situation may be different. Then, this list's archives and Debian documentation are your best resources :) cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: accelerator boards
On Thu, 2 Jun 2005, F. Heitkamp wrote: Does Debian work with the accelerator boards for Macintosh computers? Has anyone tested or had experience/problems with any of them? It's not a matter of Debian working - it's whether the kernel supports them or not. For "newer macs" (i.e. anything with a built-in USB port), you should be able to use G4 upgrades just fine. The company making the upgrade card may have firmware updates for you to install, which may require you to run from OS X, but after that, you should be able to use them in linux without any problems. If you're like me, and you have an older mac (machines that originally came with 601/603(e)/604(e) CPUs), then it becomes more complicated. The ROM on these older macs don't know about the L2 and L3 caches on the CPUs, so the "drivers" that come with these cards turns on these CPU caches. If you boot directly into linux without these caches on, the machine will run, but you just won't get the speed and performance this memory provides. There are ways to turn on the cache memory on these cards, without the use of their MacOS drivers. Do a search on the archives for "l2cr". It is a variable that lives in the /proc directory, and if manipulated properly, will turn on the L2 / L3 cache memories on G3 and G4 CPUs. So yes - it is possible, but depending on your mac and the upgrade card you want to use, you will run into varying levels of difficulty :) cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: iMac G5 Status?
> well, at least CONFIG_ADB and CONFIG_THERM_PM72 aren't set in > /boot/config-2.6.11-pseries, i guess that would help. The other options you mentioned in your other email (I deleted it before I could reply to that note): > there are quite a few differences comparing to the -power4-smp config. > is it normal that CONFIG_ADB, CONFIG_DMASOUND_PMAC, CONFIG_SND_POWERMAC > and CONFIG_SND_BIT32_EMUL are not set? are for hardware that do not exist on this (G5) generation of macintosh machines. They are relevant for me - I have a 10-year old, first gen PCI PowerMacintosh :) cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Missing Scsi Card in Power Mac G3 Server
> Anybody has any idea on what belong in a Power Mac G3 Server for a > Scsi card, cause i want to make a server with it on Debian, but the > scsi card to connect the hard drive is missing, if anybody does know > please let me know so i can build a nice machine to make a webserver > or something of this please No experience with Mac SCSI cards personally (the macs I've had with SCSI have had on-board controllers) but this page might be of help: http://www.xlr8yourmac.com/scsi_.html cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: accelerator boards
On Fri, 3 Jun 2005, F. Heitkamp wrote: On Thu, 2 Jun 2005, vinai wrote: On Thu, 2 Jun 2005, F. Heitkamp wrote: Does Debian work with the accelerator boards for Macintosh computers? Has anyone tested or had experience/problems with any of them? It's not a matter of Debian working - it's whether the kernel supports them or not. For "newer macs" (i.e. anything with a built-in USB port), you should be able to use G4 upgrades just fine. The company making the upgrade card may have firmware updates for you to install, which may require you to run from OS X, but after that, you should be able to use them in linux without any problems. I have a "digital audio" G4 Mac with Dual 533 PPC 7410s. I was considering getting a dual 1.7GHz PowerLogix board. I say "considering" because I may just wait several months and get a G5. I think these should work just fine. As to whether or not they're worth the $$ for the performance increase, that's something you need to figure out yourself for the jobs you want to do with this machine. Just as an example - if you're going to be doing something that throws around a lot of data, where memory bandwidth is key, you definitely want to save up $ for a G5. There are a few benchmarking tests available. They would be better able to help you make up your mind than anything else :) Also keep in mind, I think Powerlogix usually wants some software installed (either firmware updates or drivers for their cards), and that'll most likely require you to keep OS X around ... That being said, at work, I have a machine not too different from yours, a 667 G4. I'm also contemplating upgrading to dual 1.8 G4s and getting a slight nicer video card. It'll be cheaper than a G5, and that should more than give me the performance boost I'd need. cheers vinai -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: MOL works on PB3400
On Mon, 25 Oct 1999, Mike James wrote: ->MOL on PB3400 confirmed: -> ->Mac-on-Linux (mol-0.9.21) boots Mac OS 8.6 from an HFS partition on a ->PowerBook 3400 (603e) running Debian potato. (No, I haven't confirmed ->that it works under LinuxPPC ... Just checked out 0.9.22 with LinuxPPC, and it boots great with a 2.2.12 kernel, unpatched !! Sam - tremendous job ;-) !! Just to add details - this is on a 7500 (though in MOL under the System Profiler, it shows up as a 7300/7350 WGS) and speed seems to be comparable to MacOS running natively (but I guess that was the whole point ;-) and the OS version is 7.6.1 I did not do a whole lot of checking, but as I said, System Profiler seemed to work okay and detect everything (software wise, not hardware though) and Claris seemed okay as well ... (Opened up a spreadsheet okay ...) Again Sam, great piece of work ! My (and I am sure many others) thanks to you. cheers vinai
Re: PPC
Also, at: http://www.linuxiso.org/debian.html you can download debian 2.2r3 for PPC. You will have to roll your own 2.4 kernel though ... cheers vinai On Tue, 5 Jun 2001, Andrew D Dixon wrote: }"Heyliger, Brian" wrote: }> }> Where can I buy a compiled copy of the PPC version of Debian? Preferably }> with the 2.4 kernel. }> }Um, you can download one off of the internet. You guys at Cisco do have }the internet dont you;) } }But seriously, check out: } http://www.debian.org/distrib/ }there's a link to vendors as well as instructions for downloading.
Re: Powerbook G3 1999 xfree 4.0.2 accel?
On Thu, 7 Jun 2001, Matthew Bernhardt wrote: }I may have missed a sentence somewhere saying that there is no way to }use the r128 driver on a 1999 Powerbook G3. } }I grabbed 4.0.2's from http://penguinppc.org/~puetzk/, but I still get a }"No Screens Found". I might be wrong, but doesn't the 1999 PB G3 have a Mach 64 graphics chip ? vinai
Re: unsupported display? (dual scan on PowerBook 2400 & G3)
On Mon, 25 Jun 2001, Leif Halvard Silli wrote: }Of the G3-powerbooks, the 233mhz wallstreet is the onlyone wiht dual scan }display. There is also one version of the 2400 PowerBook which has dual }scan. LinuxPPC claims to support both the 2400's and the G3's. But since }dual scan is relatively seldom on the PCI-macs, I fear that they have been }left out. The rest of the dual scan power books are also few and counts for }pb's which generally aren't considered to work with Linux (like 1400 }series). I am not sure if being dual-scan or not makes a difference or not ... I think what's more important is your kernel have support for the graphics chip in these powerbooks (which is the CT6555* for the 3400s and 2400s models). Ever since linuxPPC 1999 (which I know are dirty words around here :) and the 2.2.12 kernels - I've gotten Linux running without any major graphics problems. In the text console, there were no issue that I could remember. For X - I was able to get things up and running using Xpmac.9 or .10. cheers vinai
Re: ata/100
On Thu, 28 Jun 2001, Noah John wrote: }are there any ata/100 drivers available for debain on PPC? also, are there }any ata/100 drivers available anywhere for macos? I'm planning on getting a }pm9600 and putting an ATA controller card in it. My question is, can i use }Maxtor's ata/100 or am I limited to an ata/66? I want to run macos and }debian on the 9600... Take a look at www.xlr8yourmac.com today (June 28). There was a notice about the first ATA-100 controller for MacOS based machines ... vinai
Re: Xfree4(.1) on Lombard?
On Tue, 3 Jul 2001, Colin Walters wrote: }Christian Jaeger <[EMAIL PROTECTED]> writes: } }> Of course it didn't work out, after several hours of fiddling around }> I see there are too many failed hunks. Maybe I would do it when I }> know that I am really the first person building xfree4.1. Am I? } }http://people.debian.org/~branden> } }I don't think you need to go to XFree 4.1 to get X running, though; I }thought I've seen people report success with earlier XFree versions. Been running XFree 3.3.6 on my Lombard for the last few months with Potato 2.2r2. This is the machine I am doing a lot of my thesis work on, so I have not had the time or courage to do any further upgrades after I got everything working ... But if you need a working XF86Config file for 3.3.6 on a Lombard, let me know and I can send it off to you. I found this one on the web somewhere, and I modified it slightly for my purposes. cheers vinai
where to add MOL init stuff in Potato
Hi All, I just compiled MOL and got it running on my Lombard running Potato 2.2r2. I want to run a script at boot time so MOL's modules are loaded. This allows non-root users to start MOL. Where would I run the script from ? I have it in /usr/bin/startmol_at_boot. In LinuxPPC's installation, I would just add that line to the end of rc.sysinit. But where would be the cleanest place to put such a call on a Potato system ? Thanks vinai
Re: MOL with stock potato?
On Tue, 10 Jul 2001, Matthew Bernhardt wrote: } I have a Powerbook '99 running smoothly, if I have to do 2.4 I assume }using kernel.org's is not a good idea with PPC. } } Can I hack out the MOL compile or should I just hit 2.4? Have compiled and run MOL on 2.2.19 and happy with it. This is an older version of MOL though (0.9.55, and I think it is up to 0.9.61 now ?) but I've gotten it up and running on both a PB 1999 and an 8500 ... vinai
package with netatalk startup script ?
Hey Folks, I was wondering if someone could tell me which package contains the startup scripts in /etc/init.d ? I messed up my netatalk script without first making a backup and I would like to get that script back. Barring that - if someone would be willing to e-mail me (off-list) a copy of their copy of this file, I would also be most appreciative. Feeling very silly ... Thanks all. vinai
Re: Video card support
On Sat, 14 Jul 2001, samuel rose wrote: }Before I go and buy a new video card for my PPC 7500, I was wondering if }there is a definitive resource for checking compatibility with debianppc. }There is some stuff on penguineppc.org. Also, for XF86, I am guessing that }I should check their website for compatibility or for a server? Pretty much your only option right now would be an ATI card. Support for the Radeon was recently added by Ani Joshi to the XFree PowerPC tree, and Rage 128 support has been present for a long time. Mach 64 support is also present, but it might be harder to come by one of these cards. As far as other Mac video cards (ixMicro, 3dfx cards) - I have seen reports of these working, though their performance does not seem to be as good as ATI cards, and support may not be as mature. I am not sure if nVidia makes a PCI graphics card for Macs, but I don't think there is any support for these cards on PPC XFree right now, and people don't seem very optimisitic about that situation improving ... cheers vinai
package with netatalk script ? - SOLVED
I found what was wrong with my setup ... I tried: apt-get remove netatalk apt-get install netatalk But it would not install as it could not find the /etc/init.d/netatalk from the previous install (as I removed it as I messed it up). What I had to do was to: apt-get remove netatalk dpkg --purge netalk apt-get install netatalk which worked and now I have netatalk up and running ... Thanks again Daniel. vinai On Sat, 14 Jul 2001, Daniel Jacobowitz wrote: }On Sat, Jul 14, 2001 at 03:49:27PM -0500, Vinai wrote: } }> I was wondering if someone could tell me which package contains the }> startup scripts in /etc/init.d ? I messed up my netatalk script }> without first making a backup and I would like to get that script }> back. }> }> Barring that - if someone would be willing to e-mail me (off-list) a }> copy of their copy of this file, I would also be most appreciative. } }They're in the packages which need them - netatalk's script will be in }netatalk.
Re: Ethernet Powerbook3400c
On Tue, 7 Aug 2001, Pedro Alves wrote: }Hi again } }How do I configure ethernet settings, after debian installed? }I have a modem/ethernet card built-in in the powerbook. I had this stashed away. Thought it would come in useful to someone else again :) Hope you are comfortable building your own kernels... cheers vinai *** Date: Sun, 12 Mar 2000 23:02:41 -0600 From: Michael Coyle <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Subject: note on kernel for 3400's > Does the 3400 still need that polling ethernet driver patch? No, it's part of the kernel source now. Here is the comment from Maio: "In the kernel configuration, under Ethernet (10 or 100Mbit), I put NO for "Generic DECchip & DIGITAL EtherWORKS PCI/EISA", and YES to "DECchip Tulip (dc21x4x) PCI support".
Re: Is it possible to burn MacOS bootable CD's on Linux
On Fri, 10 Aug 2001, Ethan Benson wrote: }> also it is possible to have toast create an old+new world bootable cd }> image with a size of 650M and from linux repopulate it and burn the }> result. sorry if i have the details missing here but i did get this to }> work. } }no toast won't create hybrid filesystems AFAIK, hybrid filesystem is }required since part of the debian archive is on CD1. Toast _CAN_ be used to make a CD from the images. I used Toast (a 4.x version) to burn CD's from the debian images available at www.linuxiso.org. I don't use burning software under linux, so I know for sure I used Toast. What I ended up doing was burning these as disc images. I can't remember for sure now, but I don't think I touched the option where it asked if I wanted to make a CD bootable or not, and I did _NOT_ mount the images at all. And all of the CD's work fine on my Lombard, including a bootable CD #1. HTH vinai
