Re: yaboot-installer bugs
On Wed, 2002-09-18 at 22:31, Matt Kraai wrote: I've found two bugs in yaboot-installer: Sweet. Your patch looks good to me, if you want to just go ahead and commit it I'd be happy with that. So I take it from your message that you've actually tried testing it on a live system? I am planning to do that very soon, I've just gotten caught up temporarily with other things. The latest version of MOL will apparently soon hit main, which means testing will be a whole lot easier. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: PowerPC kernel repository (was: Woody on IBM RS/6000 7025 F50)
On Tue, 2002-09-03 at 23:44, Chris Tillman wrote: Since the consensus is that the current prep and chrp kernels which we supply as part of boot-floppies are badly broken and unusable, would something like this be considered for a stable point release? I think it should be. I've only heard negative feedback about the kernels we supply. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#158947: Installation on new Ibooks
On Sat, 2002-08-31 at 17:56, Chris Tillman wrote: I _thought_ I was missing something. Makes sense to me... Well, I am very hesitant to make such a potentially intrusive change for boot-floppies. I guess it would be up to the main b-f maintainer to decide whether or not this change should be made...which would be...Eduard? However, I will soon make this change for at least yaboot-installer for d-i. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#158947: Installation on new Ibooks
On Sat, 2002-08-31 at 13:29, Chris Tillman wrote: If not, please get back to us. Otherwise, this bug can be closed. Thank you for your report. Well, there is still the bug that booting from install24-safe doesn't write out the append=video=ofonly to yaboot.conf. I suppose we could do some hack with looking at /proc/cmdline, and seenig if it contains 'video=ofonly'... Actually though, the more I think about it, the more it makes sense to by default write out an 'append=' line in the installed bootloader with the arguments used to boot the installer. For all architectures. Can anyone think of a case where this would break? We would have to ignore a few arguments like initrd=, of course. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: debian-installer patch followup #2
On Wed, 2002-08-28 at 11:02, Wartan Hachaturow wrote: Some time ago I've synced autopartkit with 1.6. It's pretty easy to do this with 1.4, since API is different mostly in function names ;) I've also synced partkit, but there has been some changes ;) since then. I could do a merge and commit it, if it's needed.. Please do; I think we should be using the latest version of parted. We're probably going to have to get used to using bleeding-edge versions of parted, as we might need to port it to new architectures (e.g. sparc). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: shell not working
On Fri, 2002-08-30 at 02:44, Michael Cardenas wrote: moshez mentioned that cdebconf expects that it has control, so any gtk frontend might have to some special thread handling. is this the case? Right. is it also the case for the shell? I don't think so; debconf will just be blocked until the shell exits, which is fine. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: cvs commit to debian-installer/tools/grub-installer/debian bywart
On Fri, 2002-08-30 at 15:50, Debian Boot CVS Master wrote: Repository: debian-installer/tools/grub-installer/debian who:wart time: Fri Aug 30 13:50:38 MDT 2002 Log Message: I'm a shithead, I know. Fixing misplaced commit. Heh...the poor grub-installer has been murdered twice now :) First by me, now you... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [d-i] Yet Another bug with pre-sprintf malloc:ing
On Mon, 2002-08-26 at 06:51, Martin Sjögren wrote: I found Yet Another bug with pre-sprintf malloc:ing in anna (in retriever.c, get_packages), a patch is attached. This kind of thing is exactly why GNU came up with the asprintf extension. Since we're using glibc and gcc, I suggest we have a policy of allowing GNU extensions, such as glibc library functions, and GNU C. I guess this is up to Mithrandir...what do you think? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: some new install documentation: booting from a hard disk usingLILO
On Wed, 2002-07-31 at 17:20, [EMAIL PROTECTED] wrote: Comments welcome. Should I send the SGML patch to this list ? A patch would be nice, yes. I can also provide a french translation in short time if asked. I'm sure the french translation team would appreciate that :) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: cvs commit to debian-installer/tools/grub-installer/debian bywalters
On Mon, 2002-07-29 at 02:48, Debian Boot CVS Master wrote: Repository: debian-installer/tools/grub-installer/debian Ugh. I don't know how that happened. I'll fix that ASAP. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: cvs commit to debian-installer/tools/grub-installer/debian bywalters
On Mon, 2002-07-29 at 03:26, Colin Walters wrote: On Mon, 2002-07-29 at 02:48, Debian Boot CVS Master wrote: Repository: debian-installer/tools/grub-installer/debian Ugh. I don't know how that happened. I'll fix that ASAP. Ok, turns out that since I just did 'cp -a grub-installer yaboot-installer' in order to get a skeleton directory, it copied the CVS directories too, so when I tried to commit the yaboot-installer stuff CVS thought I was committing to grub-installer. Confusing. Anyways, this should be fixed. Sorry for the extra commit message spam :) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: i18n second stage...
On Wed, 2002-06-12 at 01:59, Junichi Uekawa wrote: For Japanese, we at least need some kind of debconf fix for utf-8 character conversion support, or a working japanese character terminal (jfbterm for all arches?). Couldn't debconf just use iconv, or am I missing something? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#142481: boot-floppies: tries to mount partition as XFS evenafter reinitialisation
On Sat, 2002-04-13 at 13:47, Eduard Bloch wrote: See the first patch. It should work, but I have too bad access to my development machine now to test extensively and find the bug. Eduard, I think your patch looks good, but my vote is going to be against committing it. This is really a corner case, especially since none of the installation kernel images we distribute have XFS enabled. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 2.4 kernel as default boot kernel on CD #1 ??
On Sat, 2002-04-06 at 02:55, Raphael Hertzog wrote: We also have the MULTIBOOT feature, that let us select a flavor to boot on but it hasn't been widely tested afaik... This seems to me to be the best option of them all. What are the disadvantages of using this? For example, for the powerpc installer, I made it so the user has a choice on bootup of 2.2 versus 2.4, using yaboot. I assume MULTIBOOT is something similar using syslinux or whatever on i386? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: more on pcmcia-modules on PowerPC
On Sat, 2002-04-06 at 14:20, Adam Di Carlo wrote: Colin Walters [EMAIL PROTECTED] writes: We should be making this build against a version of pcmcia-cs which fixes #139398. I've uploaded a NMU to DELAYED/2-day. But I suppose I can use this NMU locally to build a new version of the modules package now. Yes, certainly. Let me know if I should do another build of the PowerPC boot-floppies to grab these new pcmcia modules and pcmcia-cs... If you can, sure. But looking at the changelog.gz in pcmcia-source, I don't think it's absolutely critical. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: clock is ticking on arm, mips, mipsel
On Wed, 2002-04-03 at 11:39, Adam Di Carlo wrote: Please do builds/uploads of boot-floppies 3.0.22 for arm, mips, and mipsel within the next 48 hours. If not, you will probably not have up-to-date boot-floppies for the Woody release. Thanks. We are doing one last 3.0.23 release, right? There's one bit of code (already committed) that I would like to get into the final woody b-f for powerpc, although it's not strictly release-critical. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: more on pcmcia-modules on PowerPC
On Tue, 2002-04-02 at 13:54, Adam Di Carlo wrote: Colin, there are some problems with your pcmcia-modules-2.4.18_3.1.29-4+newpmac.1_powerpc.deb upload. 1) The package name should be pcmcia-modules-2.4.18-newpmac -- e.g., subarch in pkg name, please. Otherwise PowerPC build breaks. Yes, I will fix this. 2) Could you please port/upload the latest pcmcia-cs in unstable for PowerPC and build against tht? We should be making this build against a version of pcmcia-cs which fixes #139398. I've uploaded a NMU to DELAYED/2-day. But I suppose I can use this NMU locally to build a new version of the modules package now. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Help - problem after installing a new kernel
On Sat, 2002-03-30 at 16:57, Chris Tillman wrote: It built it, believe it or not. But I don't use pcmcia :-) Right, but I don't think the modules would have worked. The image was in there, but it couldn't find it; maybe the upper level folder was changed? I tried turning debug on during the build to see if I could tell what archive path it was trying to construct but I couldn't find anything useful. Last time I ran into this problem I ended up purging my /archive directory and starting again. However, I think it may suffice to clear out the /archive/debian/download/var/lib/apt/lists/* Packages files. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [ppc] new-powermac kernel lacks iso9660 filesystem?
On Thu, 2002-03-28 at 02:25, Chris Lawrence wrote: I have a report from a user (David Stanaway, CC'd) that he had a bit of trouble with the new-powermac flavor because the included kernel in the 3.0.19 boot-floppies lacked the iso9660 filesystem. Ok, I see that the new kernel we needed has made it into the archive. I just updated boot-floppies to use it, so hopefully this issue will be fixed with the next b-f release. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Still unable to 'Initialize a Linux Partition' with bf 3.0.21
On Sat, 2002-03-16 at 18:32, Andreas Wüst wrote: I wonder about this, because the 'Initialize..' part worked perfectly well with an 2.4.9 apus kernel (but this isn't much use to me, as 2.4.9 doesn't support dhcp, and so I can't install the rest of the system after the reboot *snif*). By 2.4.9 doesn't support dhcp, I'm assuming you mean it was compiled without CONFIG_FILTER or CONFIG_PACKET; I'll look into fixing that if so. Anyways, have you tried diffing the .configs between the 2.4.9 apus kernel and your 2.4.17 kernel? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Still unable to 'Initialize a Linux Partition' with bf 3.0.21
On Sun, 2002-03-17 at 14:09, Andreas Wüst wrote: Anyways, have you tried diffing the .configs between the 2.4.9 apus kernel and your 2.4.17 kernel? Nope! What should I be looking for? I don't know; like important platform-specific config options that you might have missed. I am now going to file a bug against bf's, as Chris advised me. OK, but as far as I know no one is actively working on and testing apus support in the boot-floppies, so it is unlikely this bug will be fixed (ever, since the b-f code base is being terminated after woody). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Missing filesystems for 2.4 kernel on testing cds
[ Dropping Brian from the CCs as he is likely not interested ] On Mon, 2002-03-11 at 12:23, Matt Zimmerman wrote: I have done this, using a previous version of the netinst CDs, and it worked fine. I didn't use NFS, but I was able to install the kernel and such from the CD. I'm pretty sure that I even posted a report to debian-boot at the time. Er...ah, yes, I see it. But you were using a pre-3.0.19 version of b-f, no? If so, that means you likely weren't using the new-powermac installer and/or the modifications I made to debian-cd to support it; this is what needed testing, and Brian was the first person to confirm it worked. This would explain why you were able to install from the CD, since you were using the powermac (not new-powermac) kernel. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Missing filesystems for 2.4 kernel on testing cds
Package: kernel-image-2.4.16-newpmac Severity: serious [ Dan: basically the gist of this is that the newpmac kernel needs to have NFS and iso9660 built in. Also, (on an unrelated note), we need MESH scsi built in. ] On Fri, 2002-03-08 at 13:30, Bradford Powell wrote: I've downloaded and tried the unofficial woody cd (just disc 1) from powerpc.trasno.net (downloaded on March 1). Well, you are the first person I've heard from who has actually tested installing woody on a PowerPC machine from CD. So, I boot from the cd, choose 'linux24' from yaboot, and start installation. It's good to know this part works. When it gets to the point of mounting root over nfs, (after network was properly setup by dhcp: I tested this by pinging the nfs server), the install script was unable to do so. Dropping to the shell in the other console, catting /proc/filesystems does not list nfs. NFS is compiled as a module; I guess my thought was that most people will be either doing a full network install (in which case they'll download the drivers.tgz along with the kernel), or they'll be doing a full CD install (and drivers.tgz comes on the CD). However, since the new-powermac kernel is already too big to fit on a 1.44 floppy, I don't think we'll lose anything by compiling NFS (+NFS root) in. Then I decided to try nfs root install under the old kernel. This time the root fs was mounted, base system was installed (I think), but before base system install was complete, debootstrap reported unable to run chroot /target dpkg. Strange; I remember hearing about this problem before, but I don't remember what the cause was. I think it's been fixed. Repartitioning went as planned. Again, I chose linux24 from the yaboot selector. Now after formatting the root partition and mounting it, the time came to install the base system. So I selected install from CD. Unable to mount cd... I checked /proc/filesystems: no iso9660. Oops! We will fix this right away. I then proceeded to install to the hard disk using the 2.2 kernel and updated to 2.4 with my first deselect. Summary: the linux24 kernel on the install cd should probably have the filesystems iso9660 and nfs compiled in. I'm not sure what went wrong with the install to an nfs root using the 2.2 kernel (unable to run chroot /target dpkg) But, I did get it installed (but haven't done much with it yet). That's good to hear! Thank you very much for testing and sending this report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#137170: boot-floppies: [powerpc new-powermac] scsi disk not detected
On Thu, 2002-03-07 at 20:32, Chris Tillman wrote: Maybe. But before waking Herbert up I wanted to see if maybe Colin did something strange to the kernel he got, or if it's straight out of the archive. The only relevant difference I can see with the -powerpc flavor is that CONFIG_SCSI_MESH is disabled. Is this the kind of SCSI which you're using? I don't recall why I disabled it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#110717: Refer to 135504
On Mon, 2002-03-04 at 23:42, Chris Tillman wrote: Earlier I posted to this bug, thinking my issues on powerpc might be related. Now I'm quite sure that the problems I was seeing were due to the zombie process described in 135504. So, disregard my ramblings in this bug. Could you merge them then? There are instructions on how to do that here: http://www.debian.org/Bugs/server-control -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#136818: powerpc/new-powermac] will not mount /boot partition
On Tue, 2002-03-05 at 11:55, Matt Kraai wrote: It worked in my testing (an iBook2). I say commit it. Done. Thanks for testing this. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#136818: powerpc/new-powermac] will not mount /boot partition
On Mon, 2002-03-04 at 13:57, Matt Kraai wrote:
Package: boot-floppies
Version: N/A; reported 2002-03-04
The boot-floppies do not permit a /boot partition to be
mounted They complain that its file system is not supported by
the boot loader
Ok Here's my suggested patch It is not yet tested, but it *should*
work I think
Index: partition_config.c
===
RCS file: /cvs/debian-boot/boot-floppies/utilities/dbootstrap/partition_config.c,v
retrieving revision 1.146
diff -u -u -r1.146 partition_config.c
--- partition_config.c 2002/02/17 22:00:23 1.146
+++ partition_config.c 2002/03/04 20:40:55
-119,40 +119,40
#if #cpu(powerpc)
int
verify_powerpc_mount(char *mount_point) {
-/* yaboot can read ext2,ext3,xfs,reiserfs, but crippleware filesystems
- like hfs will break make-bootable.
- quik can only read ext2. */
-if ((!strcmp(mount_point, /boot)) || (!strcmp(mount_point, /boot/))) {
- if (strstr(Arch2, PowerMac)) {
- int is_newworld = !strcmp(Arch3, NewWorld);
- if (!is_fstype(/boot, ext2)
- ||
- (is_newworld ? !is_fstype(/boot, ext3)
-|| !is_fstype(/boot, xfs)
-/* Should someone WANT to lose their data...*/
- || !is_fstype(/boot, reiserfs)
-: 0))
- {
- snprintf(prtbuf, sizeof(prtbuf),
-_(The file system type you have chosen for /boot
- is not compatible with the %s bootloader. Please
- choose one of: ext2%s%s),
-is_newworld ? yaboot : quik,
-is_ext3 ? ext3 : ,
-is_xfs ? xfs : );
- problemBox(prtbuf,
- _(Choose another file system));
- return 1;
- }
- else if (!(((Arch3 != NULL) (0 == strcmp(Arch3, NewWorld))) || (0 ==
strcmp(Arch2, chrp
- if (DLG_NO == yesNoBox(_(Quik may not work correctly if a separate
- /boot partition is used. Do you wish to continue
- mounting /boot?),
-_(Quik problems with /boot not on /)))
- return 1;
- }
+ /* yaboot can read ext2,ext3,xfs,reiserfs, but crippleware filesystems
+ like hfs will break make-bootable.
+ quik can only read ext2. */
+ if (!strncmp(mount_point, /target/boot, 12)) {
+if (strstr(Arch2, PowerMac)) {
+ int is_newworld = !strcmp(Arch3, NewWorld);
+ if ((!is_newworld
+ !is_fstype(/boot, ext2))
+ || (is_newworld
+ !is_fstype(/boot, ext2)
+ !is_fstype(/boot, ext3)
+ !is_fstype(/boot, xfs)
+ !is_fstype(/boot, reiserfs)))
+ {
+ snprintf(prtbuf, sizeof(prtbuf),
+ _(The file system type you have chosen for /boot
+is not compatible with the %s bootloader. Please
+choose one of: ext2%s%s),
+ is_newworld ? yaboot : quik,
+ is_ext3 ? ext3 : ,
+ is_xfs ? xfs : );
+ problemBox(prtbuf, _(Choose another file system));
+ return 1;
+ }
+ else if (!(((Arch3 != NULL) (0 == strcmp(Arch3, NewWorld)))
+|| (0 == strcmp(Arch2, chrp
+ if (DLG_NO == yesNoBox(_(Quik may not work correctly if a separate
+/boot partition is used. Do you wish to continue
+mounting /boot?),
+ _(Quik problems with /boot not on /)))
+ return 1;
}
-return 0;
+ }
+ return 0;
}
#endif
-639,16 +639,16
else {
mount_point = strdup(mount_points[status].tag);
}
-#if #cpu(powerpc)
-if (verify_powerpc_mount(mount_point))
- return 1;
-#endif
if (NULL == mount_point) return 1;
real_mount_point = (char *)malloc(9 + strlen(mount_point));
strcpy(real_mount_point, /target);
if (mount_point[0] != '/')
strcat(real_mount_point, /);
strcat(real_mount_point, mount_point);
+#if #cpu(powerpc)
+if (verify_powerpc_mount(real_mount_point))
+ return 1;
+#endif
}
if (! NAME_ISDIR( real_mount_point, statbuf ) ) {
Bug#136818: powerpc/new-powermac] will not mount /boot partition
On Mon, 2002-03-04 at 16:19, Matt Kraai wrote:
No The file system must be mounted for is_fstype, and hence
verify_powerpc_mount, to work correctly Since it isn't
mounted, is_fstype always returns false, so verify_powerpc_mount
always decides that the /boot partition file system is not
supported
Ah, right That's rather a bummer I think I see how to solve this
though; we just pass verify_powerpc_mount the type of the partition we
want to mount it as Maybe this will work:
Index: partition_config.c
===
RCS file: /cvs/debian-boot/boot-floppies/utilities/dbootstrap/partition_config.c,v
retrieving revision 1.146
diff -u -u -r1.146 partition_config.c
--- partition_config.c 2002/02/17 22:00:23 1.146
+++ partition_config.c 2002/03/04 22:35:09
-118,41 +118,41
#if #cpu(powerpc)
int
-verify_powerpc_mount(char *mount_point) {
-/* yaboot can read ext2,ext3,xfs,reiserfs, but crippleware filesystems
- like hfs will break make-bootable.
- quik can only read ext2. */
-if ((!strcmp(mount_point, /boot)) || (!strcmp(mount_point, /boot/))) {
- if (strstr(Arch2, PowerMac)) {
- int is_newworld = !strcmp(Arch3, NewWorld);
- if (!is_fstype(/boot, ext2)
- ||
- (is_newworld ? !is_fstype(/boot, ext3)
-|| !is_fstype(/boot, xfs)
-/* Should someone WANT to lose their data...*/
- || !is_fstype(/boot, reiserfs)
-: 0))
- {
- snprintf(prtbuf, sizeof(prtbuf),
-_(The file system type you have chosen for /boot
- is not compatible with the %s bootloader. Please
- choose one of: ext2%s%s),
-is_newworld ? yaboot : quik,
-is_ext3 ? ext3 : ,
-is_xfs ? xfs : );
- problemBox(prtbuf,
- _(Choose another file system));
- return 1;
- }
- else if (!(((Arch3 != NULL) (0 == strcmp(Arch3, NewWorld))) || (0 ==
strcmp(Arch2, chrp
- if (DLG_NO == yesNoBox(_(Quik may not work correctly if a separate
- /boot partition is used. Do you wish to continue
- mounting /boot?),
-_(Quik problems with /boot not on /)))
- return 1;
- }
+verify_powerpc_mount(char *mount_point, char *type) {
+ /* yaboot can read ext2,ext3,xfs,reiserfs, but crippleware filesystems
+ like hfs will break make-bootable.
+ quik can only read ext2. */
+ if (!strncmp(mount_point, /target/boot, 12)) {
+if (strstr(Arch2, PowerMac)) {
+ int is_newworld = !strcmp(Arch3, NewWorld);
+ if ((!is_newworld
+ strcmp(type, ext2))
+ || (is_newworld
+ strcmp(type, ext2)
+ strcmp(type, ext3)
+ strcmp(type, xfs)
+ strcmp(type, reiserfs)))
+ {
+ snprintf(prtbuf, sizeof(prtbuf),
+ _(The file system type you have chosen for /boot
+is not compatible with the %s bootloader. Please
+choose one of: ext2%s%s),
+ is_newworld ? yaboot : quik,
+ is_ext3 ? ext3 : ,
+ is_xfs ? xfs : );
+ problemBox(prtbuf, _(Choose another file system));
+ return 1;
+ }
+ else if (!(((Arch3 != NULL) (0 == strcmp(Arch3, NewWorld)))
+|| (0 == strcmp(Arch2, chrp
+ if (DLG_NO == yesNoBox(_(Quik may not work correctly if a separate
+/boot partition is used. Do you wish to continue
+mounting /boot?),
+ _(Quik problems with /boot not on /)))
+ return 1;
}
-return 0;
+ }
+ return 0;
}
#endif
-639,10 +639,6
else {
mount_point = strdup(mount_points[status].tag);
}
-#if #cpu(powerpc)
-if (verify_powerpc_mount(mount_point))
- return 1;
-#endif
if (NULL == mount_point) return 1;
real_mount_point = (char *)malloc(9 + strlen(mount_point));
strcpy(real_mount_point, /target);
-710,7 +706,7
if (!(strcmp(xfsmagic, XFSB))) { /* XFS */
#if #cpu(powerpc)
- if (verify_powerpc_mount(mount_point))
+ if (verify_powerpc_mount(mount_point, xfs))
return 1;
#endif
INFOMSG(Mounting XFS partition %s on %s, partition-name, mount_point);
-722,7 +718,7
else if (!(strcmp(shfsmagic, ReIsErFs) strcmp(shfsmagic, ReIsEr2Fs))) /*
reiserfs */
{
#if #cpu (powerpc)
- if (verify_powerpc_mount(mount_point))
+ if (verify_powerpc_mount(mount_point, reiserfs))
return 1;
#endif
snprintf(prtbuf, sizeof(prtbuf), _(The \notail\ option is required on the
/boot
Re: b-f 3.0.19 on PowerBook G4
On Sat, 2002-03-02 at 13:30, Michel Dänzer wrote: The short story is that the new-powermac flavor basically worked fine. :) Cool. It claimed yaboot couldn't handle an ext2 /boot, which obviously isn't true. I already reported this to Colin Walters on IRC and I think he's working on this or has even already fixed it. Matt Kraai came up with much better logic for the code; I just need to find a way to test it now. Michel, if you wouldn't mind testing the partitioning code in the installer at http://penguinppc.org/~walters/debian/installer-cvs on your pismo, I'd appreciate it. The keymap I chose wasn't for a Mac. I think the new-powermac flavour should at least offer Mac keymaps by default. This is very strange. I just picked the default keymap (I can't remember what it was called), and it worked fine. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: b-f 3.0.19 on PowerBook G4
On Sun, 2002-03-03 at 18:31, Michel Dänzer wrote: I hope I'll find time to test it tomorrow or at least sometime this week. Cool; that would be much appreciated. BTW, I mistyped the original URL; it's actually at http://penguinppc.org/~walters/debian/installer-current/ The basic layout was fine, but I couldn't type 'special' symbols like |. I suspect it was the sg map (for a PC keyboard), not mac-usb-de_CH. Ahh. If your keyboard isn't in the US layout, that would explain the difference. My knowledge about console keymaps and such is very close to zero, so I hope someone else can fix this... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: When SGML formatting fixes hit translators
On Sun, 2002-03-03 at 13:50, Claus Hindsgaul wrote: Please think of us for a moment before submitting such patches (and reformatting large paragraphs after deleting a word). I am sure that robster had all the best intentions and were not at all aware of this side effect when submitting the patch. Of course the SGML source should not be allowed to grow completely ugly and unreadable. Any suggestions on how to handle this problem (e.g. making doc-check more intelligent) are of course very welcome, Well, I'm not sure it makes sense for translators to follow re-filling of text, since word lengths are not consistent. Why not just use 'diff -b'? Or is there something I'm missing? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: `host name' vs. `hostname'
On Thu, 2002-02-28 at 14:50, Matt Kraai wrote: Howdy, I am pleased to report that earlier today I was able to install Debian Woody on my iBook2 using the new-powermac flavor without any serious problems. However, I did find some minor issues that I'd like to fix. Did you make a separate /boot partition? I've gotten one report that the code which tries to prevent you from making a /boot with an incorrect filesystem type is itself incorrect. It would be great if you could try beating on that part of the installer a bit; try making /boot ext3, HFS (shouldn't work), etc. I really wish UML worked on PowerPC; then it would be a lot easier to test the partitioning code. Which term is preferred, `host name' or `hostname'. The boot-floppies use both on the same screen, and some consistency would be nice. FOLDOC lists it as one word[1,2], but the hostname manual page uses two. Any preference? I think you'll have people telling you both are correct. It's a lot like filesystem versus file system. The former is very popular in common usage, but you'll usually only see the latter in formal works. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: language chooser clarity report
On Fri, 2002-03-01 at 19:11, Junichi Uekawa wrote: en - Proceed installation in English ^ with Continue might be slightly better than Proceed, too. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Floppy boot for powerpc: add video=ofonly?
On Sat, 2002-02-16 at 00:22, Chris Tillman wrote: Many oldworld powerpc computers need the video=ofonly boot argument in order to use the installer. The hfs-boot-floppy image we currently provide does not supply this argument. I used a binary editor (beav) to add video=ofonly into the miBoot System.bin source file in boot-floppies, and verified that on my oldworld machines the modified file enables the floppy to work, where before the monitor couldn't sync up. (I replaced spaces in the file, and didn't change the file size). The only issue I can think of is that while it might let the boot-floppies work initially, the installed system won't have the video=ofonly, so the user is left wondering why they don't have a bootable system. Actually, I think the debian-safe option in the current b-f has the same problem. We'll have to add some hackery to check whether video=ofonly was given, and output that to the generated yaboot.conf. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Testing boot-floppies on iMac
On Sat, 2002-02-16 at 15:27, Matt Zimmerman wrote: Yesterday, I got an iMac (NewWorld) to play with at work, and decided to install woody on it. It has no floppy drive, so the only bootable woody installation media that I could find were ISO images at: http://people.debian.org/~ieure/netinst/ which I burned onto CD-R. If you have another machine on the local network, netbooting is the best option, in my opinion. http://http.us.debian.org/debian/dists/woody/main/disks-powerpc/current/doc/ch-install-methods.en.html#s-install-tftp I couldn't figure out how to tag a partition as swap, this being my first PowerPC installation. the mac-fdisk help was not helpful). I continued without one, and have since learned (from http://penguinppc.org/projects/yaboot/doc/mac-fdisk-basics.shtml linked from the release notes) that it should be named 'swap'. I assume this would have worked; if so, it should be mentioned directly in the release notes. The entire mac-fdisk-basics document is really required reading for installing on PowerPC. Also, how can I help to test future PowerPC boot-floppies? Is there some way to network-boot this beast, or do I have to continue to use CD-Rs? We definitely need more testers for the PowerPC boot-floppies! If you have trouble netbooting, someone is almost always on openprojects.net #debianppc (I'm walters on IRC). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Hardware detection w/example code.
On Thu, 2002-02-14 at 12:30, Karl M. Hegbloom wrote: Any reason not to put hardware detection into the boot-floppies? As far as I know, this feature is going to be part of debian-installer, but not boot-floppies. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Boot-floppies 3.0.19 question
On Thu, 2002-02-14 at 15:34, Ian Eure wrote: Hello. I maintain the unofficial Debian woody netinst CD image. When I built new images against b-f 3.0.19 earlier this week, I was very surprised to see that the installer no longer supports downloading the base .debs from a Debian mirror, and now requires them on some sort of local media. Do you mean downloading individual .debs via HTTP? I did a test install of 3.0.19 new-powermac on powerpc yesterday over HTTP, and it worked correctly. What architecture and flavor were you testing? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Malformed release file
On Thu, 2002-02-14 at 15:50, Nancy Davis wrote: Hi, In trying to install the base system for woody via http, I receive the following error: Malformed release file http://http.us.deibna.org:80/debian/dists/woody/Release What version of the boot-floppies are you using? This was a symptom of a bug that should have been fixed in 3.0.19. Also, what architecture and boot-floppies flavor are you using? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: won't boot up
On Tue, 2002-02-12 at 13:43, rob mcrae wrote: I have a 486 66MHz it wont boot up something must be badly wrong with it. boot disk will not work it always says operating system not found. That almost certainly means either the floppy disk you're using to boot is bad, or the disk drive itself is bad. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: bf-3.0.19 status, pending release
On Tue, 2002-02-05 at 22:41, Chris Tillman wrote: No, I can't. The build wants glibc-2.2.5-3, which is only built for sparc and i386 so far. I think you have an out of date mirror: walters@auric madison libc6 ~ libc6 | 2.1.3-19 |stable | arm, i386, m68k, powerpc, sparc libc6 | 2.1.3-20 | proposed-updates | arm, i386, m68k, powerpc, sparc libc6 |2.2.4-7 | testing | arm, hppa, i386, m68k, mips, mipsel, powerpc, sparc libc6 | 2.2.4-7.0.2 | testing | s390 libc6 |2.2.5-1 | unstable | arm libc6 |2.2.5-3 | unstable | hppa, i386, m68k, mips, mipsel, powerpc, s390, sparc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Unable to install from hard disk or mounted partition
On Mon, 2002-01-28 at 05:53, christian mock wrote: I'd suggest fixing the SEGV in utilities/dbootstrap/baseconfig.c:155, as described in bug #126370; it's an easy fix, I'd suggest: if (bootargs.console[5]==',') ttyspeed=atoi(bootargs.console+6); if (bootargs.console bootargs.console[5]==',') ttyspeed=atoi(bootargs.console+6); and it fixes a severe problem for installations over a serial console. Your patch looks sane to me; I've applied it. Thanks! (However, I'd like to wait to close the bug until we have confirmation from the bug submitters that it fixes their problem.) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Fix problem report, in progress on hard disk install problems
On Fri, 2002-01-25 at 22:03, Chris Tillman wrote: I started in on getting the hard disk installation to work again. I wonder when it broke? Anyways, would you mind postponing these changes from being applied to CVS until after 3.0.19 is released? There are a lot of show-stoppers in 3.0.18, such as the debootstrap problems, lilo on i386, the fact that mac-fdisk is b0rked on the powerpc disks. Any changes to choose_medium.c will require a lot of testing; that code is a nasty, quivering logical mess. (Your changes do look sane though at first glance). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Processed: Re: Bug#131030: boot-floppies: DHCP doesn't workbecause kernel hasn't got CONFIG_PACKET and CONFIG_FILTER
On Sat, 2002-01-26 at 18:03, Debian Bug Tracking System wrote: Processing commands for [EMAIL PROTECTED]: reassign 131030 boot-floppies Bug#131030: boot-floppies: DHCP doesn't work because kernel hasn't got CONFIG_PACKET and CONFIG_FILTER Bug reassigned from package `kernel-image-2.2.20-prep' to `boot-floppies'. Ok then, I'll take care of this. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Debian Linus and Macs
On Sun, 2002-01-20 at 18:29, Harry Moses wrote: In the Debian home pages you mention several architectures, none of which seem to indicate that a download is included for the Mac. Perhaps the Power PC includes the Mac. I have an IMac system 9.2. Can the Debian download work on my computer? Indeed it can. We have a list of supported hardware here: http://www.debian.org/ports/powerpc/inst/install And an installation summary here, as well as links to the full installation manual: http://www.debian.org/ports/powerpc/inst/pmac Good luck! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#129790: install-doc: Missing alpha, mipsel and sparc packages keep it out of Woody
severity 129790 wishlist tag 129790 wontfix thanks On Fri, 2002-01-18 at 03:40, Petter Reinholdtsen wrote: Reading URL:http://ftp-master.debian.org/testing/update_excuses.html.gz, it seems like the missing support for alpha, mipsel and sparc keeps boot-floopies out of Woody. It can also be the high bug count. It is not possible to autobuild boot-floppies yet, and this won't be fixed. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#129837: System Hardware Requirements misleading
tag 129837 pending thanks Section 2.6.1 of the Installation Manual for i386 talks about avoiding hardware for which the vendor refuses to provide information that would allow drivers to be developed or released under an open-source license. The example given is the IBM laptop DSP sound system used in recent ThinkPad systems. This is now out-of-date information and to continue to cite it is both misleading and does an injustice to IBM which has become a significant supporter of Linux and Open Source in general. Hello Mark, This was fixed in CVS a few days ago by Chris Tillman, and a new version of install-doc will come with the official release of boot-floppies 3.0.19. Thanks for the report! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [i386] introducing a kernel 2.4 installation flavor
On Mon, 2002-01-14 at 19:02, Eduard Bloch wrote: So what do you think? IMHO most users will agree. I think it is a very, very good idea! You didn't mention my favorite thing about 2.4, which is iptables. Almost a must for installing a firewall. If there is a way we could make it into a boot-time choice on the CD, then I can't imagine how anyone would object to that. You should work both quickly and carefully; I think the release manager is using the stability of boot-floppies as at least one gauge for when to push woody into deeper freeze. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
what's holding up 3.0.19 ?
So, I really think we should release 3.0.19 soon, because it needs more testing. So, what's holding it back? Looking at a selection of the bug list: #126489, #127955: I've just tagged these pending, 3.0.19 will fix them #120386: looks like something we need to fix, but I don't know much about it. #123948: this one can be a priority for 3.0.20 or later #128258: PReP work is still ongoing, this one can be delayed for 3.0.20 or later #127971: I am not sure about this one, but it can probably wait. I've certainly never seen it. #110717: This one can wait for 3.0.20 or later. #116093: This one is definitely present in potato (I just tried it a few days ago), but I'm not sure about woody. It's important, but it can probably wait for 3.0.20 too. We are also waiting on #127258 in order to close #126205, although it's not absolutely critical. If the maintainer doesn't fix that one soon, I can just upload the package I've been using to test. Any other thoughts/opinions? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: powerpc b-f build compiles the kernel
On Sat, 2002-01-05 at 18:07, Adam Di Carlo wrote: Oh, so everything works well w/o building the kernel on PowerPC? If so, lets just kill it. Killing bootprep.sh means that the PReP kernel won't be built. Honestly, I have no idea if a PReP install even works... What annoys me about bootprep.sh is that it builds a kernel as part of the boot-floppies build process; I don't see why it can't just use a kernel-image package like every other arch. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
PReP hackers wanted
On Sun, 2002-01-06 at 14:15, Adam Di Carlo wrote: Well, the only reason I could see is if it either needs a kernel config option enabled, not provided by the kernel-image-*-prep package, or else if it needs a file from the kernel build not normally provided by the package. It should be pretty easy to read bootprep.sh Probably. But I have basically zero knowledge of the PReP architecture, and no way to test the resulting kernel image. So, I'd like to take this opportunity to call for anyone with a working knowledge of the PReP architecture, shell script, and preferably C to help work on the Debian boot-floppies for PReP. Right now it doesn't even build because the root disk is too large. Are you interested in seeing the Universal Operating System run on your favorite PowerPC architecture, the PReP? Then see instructions for getting the boot-floppies CVS from http://cvs.debian.org, and send us patches! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: powerpc b-f build compiles the kernel
[Incidentally, I'm on both -boot and -powerpc, so CCs are not necessary]
On Sun, 2002-01-06 at 15:54, Tom Rini wrote:
Can someone post that? I'm sort-of curious what needs to be done
'special' for a prep box..
As far as I can tell, this script just builds a kernel at boot-floppies
build time, it doesn't do anything magic. But perhaps I'm wrong. This
is 'bootprep.sh' in the boot-floppies CVS, available at:
http://cvs.debian.org/boot-floppies
#! /bin/bash
# PReP Boot Builder. Copyright 1999, Matt Porter [EMAIL PROTECTED]
# This is free software under the GNU General Public License.
. ./common.sh
# Usage message
if [ $# -lt 2 ]; then
echo usage: $0 archive kver 12
cat 12 EOF
archive: the directory where Debian binary packages will be downloaded
kver: kernel version
EOF
exit -1
fi
# set this to the location of the package archive
archive=$1
# set this to the kernel version
kver=$2
#---
# build PReP bootable images
#---
# Save the current directory
top=`pwd`
# Create a tmp directory
builddir=${tmpdir}/boot-floppies/bootprep-tmp-$$
rm -rf $builddir
make_tmpdir $builddir
# Extract kernel source and patches
THEPKG=$(grab_paths kernel-source-$kver)
dpkg --extract $THEPKG $builddir
THEPKG=$(grab_paths kernel-patch-$kver-powerpc)
dpkg --extract $THEPKG $builddir
# Unpack kernel source
cd $builddir/usr/src
if [ -f kernel-source-${kver}.tar.gz ]; then
tar zxf kernel-source-$kver.tar.gz
else
tar x --bzip2 -f kernel-source-$kver.tar.bz2
fi
# Patch kernel source - FIXME: allow for other patches
cd kernel-source-$kver
# this patch may be getting applied twice. Comment
# this out if you get messages about patches being already
# applied.
# FIXME: there are better ways to apply kernel patches!
zcat $builddir/usr/src/kernel-patches/powerpc/*.diff.gz \
| patch -l -s -p1
# Add rootprep.bin as the ramdisk
cp $top/rootprep.bin ./arch/ppc/coffboot/ramdisk.image.gz
# Fix .config for ramdisk rooting and kernel ip autoconfiguration.
# This image is 1440 KB and is suitable for CD/net booting.
sed 's:# CONFIG_CMDLINE_BOOL is not set:CONFIG_CMDLINE_BOOL=y \
CONFIG_CMDLINE=root=/dev/ram ip=off:' \
$builddir/usr/src/kernel-patches/powerpc/config.prep .config.param
# Enable kernel ip autoconfiguration
sed 's:# CONFIG_IP_PNP is not set:CONFIG_IP_PNP=y \
CONFIG_IP_PNP_ENABLE=y \
CONFIG_IP_PNP_DHCP=y \
CONFIG_IP_PNP_BOOTP=y \
CONFIG_IP_PNP_RARP=y:' \
.config.param .config.ipauto
# Enable nfsroot
sed 's:CONFIG_NFS_FS=y:CONFIG_NFS_FS=y \
CONFIG_ROOT_NFS=y:' \
.config.ipauto .config
# Build CD/net kernel boot image
make oldconfig make dep make clean make zImage.initrd
cp ./arch/ppc/boot/zImage.initrd $top/bootprepfull.bin
# Copy config.prep-.config and fix up for floppy/ramdisk rooting.
# This image is =1440 KB and is suitable for floppy booting.
sed 's:# CONFIG_CMDLINE_BOOL is not set:CONFIG_CMDLINE_BOOL=y\
CONFIG_CMDLINE=root=/dev/fd0 load_ramdisk=1:' \
$builddir/usr/src/kernel-patches/powerpc/config.prep .config
# Build floppy-sized kernel boot image
make oldconfig make zImage
cp ./arch/ppc/boot/zImage $top/bootprep.bin
# Clean up
rm -rf $builddir
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#127955: more info please
tag 127955 moreinfo thanks Hello Erik, I believe the mac-fdisk problems are fixed in the CVS version of the debian installer. As for the problems with the boot process hanging, what exactly do you mean? At which point does it hang? Also, if you're installing on an ibook2, could you try the new-powermac installer? It's available at: http://people.debian.org/~walters/debian/installer/new-powermac It uses a 2.4 kernel, and will support your ethernet card, video, sleep, etc. The new-powermac version will be part of the next release of the debian installer. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: URGENT! Install Base System via network: settings wrong
On Fri, 2002-01-04 at 07:02, Rory Campbell-Lange wrote: And get the message Malformed release file http://ftp.uk.debian.org:80/debian/dists/woody/Release. What version of boot-floppies is this? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#126208: boot-floppies: First choose DHCP-Configure, then change to static
On Sun, 2001-12-23 at 00:46, Colin Walters wrote: tag 126208 patch thanks On Sat, 2001-12-22 at 16:51, [EMAIL PROTECTED] wrote: Looks like configure_static_network() needs to take steps to kill off the DHCP daemons (pump and/or dhclient) if they're running. How about the following patch? Well, I thought the Debian busybox package had pidof enabled, but apparently it doesn't. Does anyone object to adding this? It will increase the size of the busybox binary by 300 bytes. Are any architectures so tight on space that this will push them over? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#126208: boot-floppies: First choose DHCP-Configure, thenchange to static
On Sun, 2001-12-30 at 19:01, Erik Andersen wrote: If they are that tight, removing the library symlinks (and just using the correct names) will surely be enough to recover that space. Ok. Could you go ahead and enable pidof in that case? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
punting on 1440kb images for new-powermac
So, I've been trying to cram a full 2.4 kernel into a 1440kb floppy image for new-powermac, but I have not been successful. Does anyone have any hints for things I can safely take out? (My domain of PowerPC hardware knowledege is mainly limited to NewWorld PowerMacs). The current .config I'm using is at http://people.debian.org/~walters/debian/installer/newpmac So, unless anyone has any ideas, I am probably just going to punt on 1440kb floppy support for new-powermac. I'll generate 2880kb images instead. I imagine most people will either install from an existing operating system, use the CD, or netboot, none of which are restricted to 1440kb images. Any thoughts/words of wisdom/inspiration? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
ppc root floppy way too large
Does anyone have any ideas on this? E: rootpmac.bin is larger than root1440pmac.bin (1735885 1474560) I think this is because of the recently enabled internationalization, but I have no idea how to go about saving over 250k. One thing I did notice is that the slang libraries seem to have not been reduced at all: space-ghost:/var/tmp/boot-floppies/root-tmp-4578/lib# ls -al libslang.so* -rw-r--r--2 root src443928 Dec 24 17:28 libslang.so.1 -rw-r--r--2 root src443928 Dec 24 17:28 libslang.so.1.4.4 Anyone have any ideas? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: ppc root floppy way too large
On Wed, 2001-12-26 at 14:30, Philip Blundell wrote: Do you have slang1-utf8-pic installed? PowerPC is presumably building OK for other people, so it sounds like this has to be some kind of environment problem on your machine. Indeed I did not, and installing it fixed the problem! Thank you very much! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: cvs commit to boot-floppies by walters
On Wed, 2001-12-26 at 20:52, Chris Tillman wrote: Cool! I had gotten tired of waiting for all that kernel building and had commented them out in my copy. Yeah, I don't understand why bootprep.sh has to make a kernel image at boot-floppies build time; it seems far easier to get an actual kernel-image package into the Debian archive. But, since I don't understand it and I don't have time to fix it, commenting it out in the Makefile in my local copy is easiest. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: ppc root floppy way too large
On Wed, 2001-12-26 at 19:23, Philip Blundell wrote: Hmm, I wonder why dpkg-checkbuilddeps (part of make check_depends) didn't complain about this. Did you have plain old slang1-pic installed already? I don't have it installed now; I'm not sure whether I did then. Maybe we need a complementary rule to check_non_i18n, to make sure that these packages _are_ installed when doing a LC build. That would be very helpful, I think. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#126208: boot-floppies: First choose DHCP-Configure, then change to static
tag 126208 patch
thanks
On Sat, 2001-12-22 at 16:51, [EMAIL PROTECTED] wrote:
Looks like configure_static_network() needs to take steps to kill off the
DHCP daemons (pump and/or dhclient) if they're running.
How about the following patch?
--- netconfig.c.~1.78.~ Tue Dec 18 19:23:04 2001
+++ netconfig.c Sun Dec 23 00:41:03 2001
@@ -845,6 +845,15 @@
/* Down the interface to clear old gateways. */
snprintf(cmdline, MAXLINE, /sbin/ifconfig %s down, netinterface);
execlog(cmdline, LOG_INFO);
+
+ /* Kill any dhcp client daemon that may have been running if the
+user previously chose DHCP */
+ if (NAME_ISEXE(/sbin/pump, nc_statbuf))
+ execlog(/sbin/pump -k, LOG_DEBUG);
+ else if (NAME_ISEXE(NC_DHCLIENT_FILE, nc_statbuf)) {
+ snprintf(cmdline, MAXLINE, kill `pidof %s`, NC_DHCLIENT_FILE);
+ execlog(cmdline, LOG_DEBUG);
+ }
snprintf(cmdline, MAXLINE-5, /sbin/ifconfig %s %s netmask ,
netinterface, IP4tostr(prtbuf, ipaddr));
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Basedebs.tgz documentation
On Thu, 2001-12-20 at 22:35, Chris Tillman wrote: A few months ago I put together a Debian floppy-maker script for usage on powerpc. The script depends on the MacOS-provided Applescript 'shell' and the scriptable Disk Copy program. The point was made before, that since these MacOS programs are not free (they come with all MacOS distributions), a script depending on them would not be Debian-distributable. Is the software that would be in the Debian archives itself free software, disregarding its dependency on non-free software? If so, I don't see what the problem would be with distributing it. (However, I don't speak for the ftp-masters, who are presumably the ones who would ultimately decide). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: new powerpc subarch: newpmac (using 2.4 kernel)
On Tue, 2001-12-18 at 19:17, Adam Di Carlo wrote: So is this stuff testing and working? If so, can we commit it for 3.0.19 ? We still need a kernel in the archive for it. Other than that it works. I've just committed the code, except for some changes to partition_config.c that I want to test some more. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Booting my laptop
On Fri, 1904-01-01 at 03:07, Chris Tillman wrote: I don't know how to generate it, although I understand it's not too hard, but there's one you can download at http://people.debian.org/~walters/debian/basedebs.tgz U, those basedebs are for PowerPC, which will likely not work on his laptop :) The way to generate basedebs.tgz is by using debootstrap, like: cd /tmp mkdir woody debootstrap --download-only woody woody http://http.us.debian.org cd woody tar czvf basedebs.tgz var debootstrap is a shell script that should run on RedHat; you can get it from: http://packages.debian.org/debootstrap Good luck! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: new ibook, b-f, kernel 2.4
Florian Friesdorf [EMAIL PROTECTED] writes: I tried to install woody on my new ibook (G3/600/384MB) with current boot-floppies 3.0.17 over network (in OF: boot enet:bootp,ip,yaboot). The kernel is booting fine and the installer gets started. The problem is, I need the network to install, but the gmac driver in 2.2.19 is to old to handle my builtin 10/100 network adapter. So I tweaked around with b-fs and managed to include kernel and modules from kernel-image-2.4.16-powerpc_2.4.16-1_powerpc.deb. I have been working on a new-powermac subarchitecture which will use benh's kernel tree. You can find my current work at: http://people.debian.org/~walters/debian/installer I plan to announce it to the -powerpc list in a few days for testing after I iron out one last bug. Florian, I would appreciate it if you could test it before it's announced for mass testing :) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: More feedback
Chris Tillman [EMAIL PROTECTED] writes: I tried the new-powermac installer again, and found it can't mount hfs disks. I wonder if busybox code is incompatible with the 2.4 kernel on this, or maybe it was because I was trying an ext3 install. Mmm...I built HFS support as a module in that kernel. Could that have been the problem? I tried hmount from another partition, but it didn't work because of the library reduction. If the h* tools use the kernel HFS support at all, and the module wasn't loaded, this could have been the problem. Or do you know for sure the library reduction busted the h* tools somehow? That made me think about the recent suggestions for a small CD-ROM iso with the installer system on it, and how that would free us from the need to reduce libraries so drastically etc. Yeah. At least for NewWorld PowerMacs, pretty much no one uses floppies. I'm sure a bigger-than-floppy-size root.bin could be booted straight from the hard disk on powermacs; so you wouldn't really even need to shoehorn stuff into an iso. Download and go. Yeah. But personally, I don't like the assumption that one already has an operating system on the machine, or that if there is one that people know how to use it. Still though, it is generally a practical assumption to make. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: new powerpc subarch: newpmac (using 2.4 kernel)
We also want this patch to release.sh
--- release.sh.~1.151.~ Tue Dec 4 14:25:40 2001
+++ release.sh Sat Dec 8 02:59:09 2001
@@ -734,8 +734,12 @@
# PMac install kit: HFS floppy image
#
mkdir -p $release/powermac/images-1.44
+ mkdir -p $release/new-powermac/images-1.44
+
if [ -f linuxpmac.gz ]; then zcat linuxpmac.gz $release/powermac/linux; fi
if [ -f sys_mappmac.gz ]; then cp sys_mappmac.gz $release/powermac/sysmap.gz ; fi
+ if [ -f linuxnewpmac.gz ]; then zcat linuxnewpmac.gz $release/new-powermac/linux;
+fi
+ if [ -f sys_mapnewpmac.gz ]; then cp sys_mapnewpmac.gz
+$release/new-powermac/sysmap.gz ; fi
sh powerpc-specials/miBoot/mkboot.sh
$release/powermac/images-1.44/boot-floppy-hfs.img
@@ -747,17 +751,34 @@
dnum=`echo $bdfile | cut -d - -f 2 | cut -c 1`
mv $release/images-1.44/$bdfile $release/powermac/images-1.44/driver-$dnum.bin
done
+ mv $release/images-1.44/rescuenewpmac.bin
+$release/new-powermac/images-1.44/rescue.bin
+ mv $release/images-1.44/rootnewpmac.bin $release/new-powermac/images-1.44/root.bin
+ mv $release/driversnewpmac.tgz $release/new-powermac/drivers.tgz
+ for dfile in $release/images-1.44/drivernewpmac-?.bin; do
+bdfile=`basename $dfile`
+dnum=`echo $bdfile | cut -d - -f 2 | cut -c 1`
+mv $release/images-1.44/$bdfile $release/new-powermac/images-1.44/driver-$dnum.bin
+ done
+
#$MAKE -C powerpc-specials/coffboot ../../rescue-pmac.coff
#cp rescue-pmac.coff $release/powermac/rescue-pmac.coff
ar -p $(grab_paths yaboot) data.tar.gz | tar -zxp -C $tmpdir/boot-floppies
./usr/lib/yaboot/yaboot
mv $tmpdir/boot-floppies/usr/lib/yaboot/yaboot $release/powermac/
+ ln -s $release/powermac/yaboot $release/new-powermac
(cd $tmpdir/boot-floppies rmdir -p usr/lib/yaboot)
cp $release/powermac/images-1.44/root.bin $release/powermac/ramdisk.image.gz
+ ln -s $release/powermac/ramdisk.image.gz $release/new-powermac/ramdisk.image.gz
cp ${tools_dir}/BootX*.sit $release/powermac/
cp ${tools_dir}/bootvars*sit.hqx $release/powermac/
+
cp powerpc-specials/bootargs-pmac $release/powermac/bootargs
cp powerpc-specials/install-pmac.en.txt $release/powermac/install.txt
cp powerpc-specials/yaboot.conf $release/powermac/yaboot.conf
+ ln -s $release/powermac/bootargs $release/new-powermac/
+ cp powerpc-specials/install-pmac.en.txt $release/powermac/install.txt
+ ln -s $release/powermac/install.txt $release/new-powermac/
+ cp powerpc-specials/yaboot.conf $release/powermac/yaboot.conf
+ ln -s $release/powermac/yaboot.conf $release/new-powermac/
#
# PReP install kit
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: cannot resume installation after abort
Adam Di Carlo [EMAIL PROTECTED] writes: The problem is more general -- all 'install base' steps from a cvs boot-floppies (i18n or not) is showing Packages.gz as corrupt and the same sequence of error. This must be a different problem that I'm having. I get this too, now. This bug seems to have somehow gotten worse :/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: new powerpc subarch: newpmac (using 2.4 kernel)
Adam Di Carlo [EMAIL PROTECTED] writes: I'm intending to apply these after we fix the base install problems and release 3.0.18. In short, they will go in for 3.0.19. If you could, please resend the patch after 3.0.18 is released, from CVS head. Ok, I will do that. I hope we can figure out these base install problems soon; there isn't much point in calling for people to test new-powermac until the base install works... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#122738: retitle
retitle 122738 libfdisk doesn't grok /proc/partitions with DevFS
thanks
The problem is that even if DevFS is enabled, but not mounted,
/proc/partitions is in DevFS format, and libfdisk doesn't like it.
See the function get_part_info in
boot-floppies/utitlities/libfdisk/fdisk.c.
I have a function to work around the problem for IDE disks, but it
doens't handle SCSI at all:
static void maybe_translate_devfs_name(char *name)
{
int is_ide = !strncmp(ide/, name, 4);
/* If it's not a DevFS-style IDE or SCSI partition name, ignore it */
if (is_ide || !(strncmp(scsi/, name, 5))) {
int host, bus, target, lun;
int len = strlen(name);
int is_disc = len 4 !(strcmp(name+len-4, disc));
char *partstr = ;
if (is_disc)
sscanf(name + is_ide ? 4 : 5, host%d/bus%d/target%d/lun%d,
host, bus, target, lun);
else
sscanf(name + is_ide ? 4 : 5, host%d/bus%d/target%d/lun%d/part%s,
host, bus, target, lun, partstr);
sprintf(name, %cd%c%s, is_ide ? 'h' : 's',
97 + bus*2 + target,
is_disc ? : partstr);
}
}
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
new powerpc subarch: newpmac (using 2.4 kernel)
So, currently woody (using a 2.2 kernel on powerpc) doesn't support most newer PowerPC Apple Macintosh hardware. So, here are the initial patches to create a new subarchitecture, which I modeled after the sun4u stuff. We still need to get a special 2.4 kernel image (as well as the pcmcia modules) into woody if this is going to work. I'm sorry I didn't do this earlier, but it took me a LONG time to realize that the reason my loopback device didn't work was because I use tmpfs on /tmp. Comments appreciated; I am just getting my feet wet with b-f. (Note; this patch also removes the pointless special PowerPC losetup stuff, as well as changing the tempfile calls to make temporary files in /var/tmp, which is where the rest of the b-f stuff happens anyways, so it seems like a good idea to me.) Index: config === RCS file: /cvs/debian-boot/boot-floppies/config,v retrieving revision 1.117 diff -u -u -r1.117 config --- config 2001/12/07 22:46:08 1.117 +++ config 2001/12/08 04:58:33 @@ -26,6 +26,8 @@ kver := 2.2.19 pcmcia_kver := 2.2.19-pmac apuskver := 2.2.10 +# NewWorld powermacs really need a 2.4 kernel +kver_newpmac := 2.4.17-pre5-benh20011207 endif ifeq $(architecture) sparc kver := 2.2.20 Index: kernel.sh === RCS file: /cvs/debian-boot/boot-floppies/kernel.sh,v retrieving revision 1.37 diff -u -r1.37 kernel.sh --- kernel.sh 2001/11/19 07:57:42 1.37 +++ kernel.sh 2001/12/08 04:42:44 @@ -28,7 +28,7 @@ shift 3 # whether this arch supports PCMCIA -if [ $arch = i386 -o $subarch = pmac ]; then +if [ $arch = i386 -o $subarch = pmac -o $subarch = newpmac ]; then HAS_PCMCIA=true else HAS_PCMCIA=false @@ -103,18 +103,20 @@ fi info creating 'linux$subarch' -if [ $subarch = pmac ]; then -# vmlinux-version is the 'work' kernel -# vmlinux.coff-version is the kernel for OF booting -cp $extractdir/boot/vmlinux-* linux$subarch -cp $extractdir/boot/vmlinux.coff-* linux$subarch.coff -else -if [ $arch = hppa ]; then +case $arch/$subarch in +*/*pmac) + # vmlinux-version is the 'work' kernel + # vmlinux.coff-version is the kernel for OF booting + cp $extractdir/boot/vmlinux-* linux$subarch + cp $extractdir/boot/vmlinux.coff-* linux$subarch.coff + ;; +hppa/*) cp $extractdir/boot/vmlinux* linux$subarch -else + ;; +*) cp $extractdir/boot/vmlinuz* linux$subarch -fi -fi + ;; +esac info creating 'sys_map$subarch.gz' cat $extractdir/boot/System.map-* | gzip --no-name -9 sys_map$subarch.gz mount -o loop $fsopts -t $fstype $floppy $mnt @@ -509,7 +496,7 @@ relax -if [ $arch = powerpc -a $subarch = pmac ]; then +if [ $arch = powerpc -a \( $subarch = pmac -o $subarch = newpmac \) ]; then info compressing and installing kernel 'linux$subarch' gzip -9f linux$subarch cp linux$subarch.gz $mnt/linux.gz @@ -615,7 +602,7 @@ # no bootloader yet, write out install.sh like m68k write_m68kinfo ;; - pmac) + pmac|newpmac) # boot from vmlinux.coff installer for now. # write_powerpc write_m68kinfo Index: rootdisk.sh === RCS file: /cvs/debian-boot/boot-floppies/rootdisk.sh,v retrieving revision 1.299 diff -u -r1.299 rootdisk.sh --- rootdisk.sh 2001/12/05 09:12:50 1.299 +++ rootdisk.sh 2001/12/08 04:55:24 @@ -901,21 +901,6 @@ # make an ext2 filesystem mke2fs -O none -F -q -N $inodes -m 0 $floppy -case $arch in - powerpc) -relax -sync - -info doing special PowerPC losetup stuff (why?) -losetup /dev/loop7 $floppy -mke2fs -O none -F -q -N $inodes -m 0 /dev/loop7 -losetup -d /dev/loop7 - -relax -sync -;; -esac - grep -q loop /proc/devices || insmod loop info mounting loop file $floppy on '$mnt' Index: make/powerpc.rules === RCS file: /cvs/debian-boot/boot-floppies/make/powerpc.rules,v retrieving revision 1.2 diff -u -r1.2 powerpc.rules --- make/powerpc.rules 2001/05/09 21:51:14 1.2 +++ make/powerpc.rules 2001/12/08 04:56:56 @@ -13,6 +13,14 @@ modulesapus.tgz sys_mapapus.gz configapus.gz linuxapus modcontapus:kernel.sh $(ROOTCMD) ./kernel.sh $(archive) $(apuskver) apus +# The pmacnew images use a different kernel version than the pmac ones +modulesnewpmac.tgz sys_mapnewpmac.gz confignewpmac.gz linuxnewpmac modcontnewpmac: +kernel.sh + $(ROOTCMD) ./kernel.sh $(archive) $(kver_newpmac) newpmac +$(packages_with_modules) + +resc1440newpmac.bin: $(root_for_rescue) rescue.sh $(arch_rescue_deps) linuxnewpmac + $(ROOTCMD) ./rescue.sh $(archive) newpmac $(kver_newpmac) rootpmac.bin \ + 1440
Bug#122738: boot-floppies can't handle devfs
Adam Di Carlo [EMAIL PROTECTED] writes: Unfortunately boot-floppies doens't currently support devfs. Colin found an issue with parsing of /proc/partitions, but I'm afraid that's probably just the tip of the iceburg. Getting devfs support into boot-floppies would set us back at least 2 months. Just to clarify, this issue appears even devfs=nomount, because the /proc/partitions format changes when devfs is compiled into the kernel, even if it isn't mounted. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Building custom boot-floppies
S. Salman Ahmed [EMAIL PROTECTED] writes: I am interested in building b-f from CVS. Are there instructions somewhere on how to do this ? Checkout instructions: URL:http://cvs.debian.org Build instructions are in the README file. Also, is it possible to customize b-f by putting a different kernel or hacking in support for things like ReiserFS, XFS, etc. ? Yes. Are there are instructions on how to do this ? None that I know of; however I believe it's just a matter of making a kernel-image .deb and dropping it in /archive/debian/incoming, and frobbing the config file in the toplevel before building. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#121374: [David Weingart dweingart@pobox.com] Base system stalls during boot
Package: boot-floppies Severity: important ---BeginMessage--- My system: PowerMac G4/400 PCI Graphics, 160 MB RAM, 2 internal IDE drives - 10 GB Western Digital and a 40 GB Maxtor. My problem: I'm trying to install Woody, so I downloaded the files needed to create an boot/install CD (as per http://people.debian.org/~branden/ibook.html), intending to perform the install via FTP. I went through the install, and everything seemed to work without incident. I chose to make the system bootable from the hard disk, and rebooted the system. This part worked as well - I'm getting the yaboot prompt, and the linux system starts to boot from the hard drive. However, it stalls with error message: 15TZconfig exited with return value 1 I noticed that I wasn't asked to set the time zone from the installer (which is different from the 2.2 installer). If I reboot, the system stalls with a different error: Id 1 respawning too fast I've wiped the partition clean and re-installed twice with the same results. Am I doing something wrong? Thanks! -dave -- David Weingart [EMAIL PROTECTED] The foot can split wood, but it can't split a watermelon. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] ---End Message---
devfs-enabled kernels and /proc/partitions
Currently, the boot-floppies can't deal with a kernel which has
CONFIG_DEVFS_FS=y, even if CONFIG_DEVFS_MOUNT is not set. I tracked
this down to libfdisk not liking the devfs /proc/partitions.
I started to write a little C function which translated the names,
and this is fairly trivial for IDE. However, it is not so trivial
for SCSI.
So, we could just try to ensure that b-f will never be booted with a
DevFS-enabled kernel, but I think this is a bad idea.
I found a function to do the conversion, but it is written in shell;
popening /bin/sh in the middle of libfdisk seems really gross to
me...does anyone know of a C function somewhere we could snarf? If
not, I guess I can just send a patch to do the former.
For reference, here's the shell function in question:
## convert devfs names into normal short ones, written by Tom Rini.
fixdevfs()
{
## get partition number, if any
local PARTNUM=${1##*[a-z]}
## Find the bus type.
local TYPE=$(v=${1#/dev/} ; echo ${v%/host*})
## Find the host number.
local HOST=$(v=${1#/dev/*/host} ; echo ${v%/bus*})
## Find the bus number.
local BUS=$(v=${1#/dev/*/bus} ; echo ${v%/tar*})
## Find the target.
local TARGET=$(v=${1#/dev/*/target} ; echo ${v%/lun*})
case $TYPE in
ide)
case $HOST in
0)
case $TARGET in
0)
local DEV=hda
;;
1)
local DEV=hdb
;;
esac
;;
1)
case $TARGET in
0)
local DEV=hdc
;;
1)
local DEV=hdd
;;
esac
;;
*)
echo 12 $PRG: $1: Unable to translate this device, try again
without devfs.
return 1
esac
local DEV=${DEV}${PARTNUM}
echo /dev/$DEV
return 0
;;
scsi)
local LUN=$(v=${1#/dev/*/lun} ; echo ${v%/*})
## In this case, we need to figure out what number our device is
local DEVCOUNT=0
## copy scsi file into a variable removing Attached Devices
## which is the first line. this avoids a lot of
## [incmopatible] crap later, and improves readability.
## find number of lines once and recycle that number, to save
## some time (linecount is a bit slow). subtract one line
## to scrap Attached Devices:
local SCSILINES=$(($(linecount /proc/scsi/scsi) - 1))
local PROCSCSI=$(cat /proc/scsi/scsi | tail -n $SCSILINES)
for i in $(smallseq $(($SCSILINES / 3))) ; do
## put every scsi device into one single line
local DEVINFO=$(echo $PROCSCSI | head -n $(($i * 3)) | tail -n 3)
[ $DEBUG = 1 ] echo 12 $PRG: DEBUG: fixdevfs: DEVINFO=$DEVINFO
## cut the type field, expect Direct-Access later.
local DEVTYPE=$(v=$(echo ${DEVINFO##*Type: }) ; echo ${v%% *})
[ $DEBUG = 1 ] echo 12 $PRG: DEBUG: fixdevfs: DEVTYPE=$DEVTYPE
if [ $DEVTYPE = Direct-Access ] ; then
## Lets find out some more information
## get the device id.
local DEVID=$(v=$(echo ${DEVINFO##*Id: }) ; n=$(echo ${v%% *}) ; echo
${n#*0})
[ $DEBUG = 1 ] echo 12 $PRG: DEBUG: fixdevfs: DEVID=$DEVID
## get the device lun.
local DEVLUN=$(v=$(echo ${DEVINFO##*Lun: }) ; n=$(echo ${v%% *}) ;
echo ${n#*0})
[ $DEBUG = 1 ] echo 12 $PRG: DEBUG: fixdevfs: DEVLUN=$DEVLUN
## get the device channel.
local DEVCHAN=$(v=$(echo ${DEVINFO##*Channel: }) ; n=$(echo ${v%% *})
; echo ${n#*0})
[ $DEBUG = 1 ] echo 12 $PRG: DEBUG: fixdevfs: DEVCHAN=$DEVCHAN
## get the scsi host id.
local DEVHOST=$(v=$(echo ${DEVINFO##*Host: scsi}) ; echo ${v%% *})
[ $DEBUG = 1 ] echo 12 $PRG: DEBUG: fixdevfs: DEVHOST=$DEVHOST
local DEVCOUNT=$(($DEVCOUNT + 1))
[ $DEBUG = 1 ] echo 12 $PRG: DEBUG: fixdevfs:
DEVCOUNT=$DEVCOUNT
if [ $DEVHOST = $HOST -a $DEVCHAN = $BUS -a \
$DEVID = $TARGET -a $DEVLUN = $LUN ] ; then
local DEV=sd$(smalltr $DEVCOUNT)${PARTNUM}
echo /dev/$DEV
return 0
fi
fi
done
echo 12 $PRG: $1: Unable to translate this device, try again without devfs.
return 1
;;
*)
echo 12 $PRG: Unknown bus $TYPE
return 1
;;
esac
## we should never get here
return 1
}
Re: Bug#121110: boot-floppies: TFTP boot on Alpha broken
Falk Hueffner [EMAIL PROTECTED] writes: rgh mellum: re init q, when I had that problem ia64, it was because I forgot to boot with root=/dev/ram rgh mellum: without that, the kernel wants to do it's initrd magic, and runs /linuxrc before starting sbin/init. linuxrc is linked to init, so effectively init gets run as pid 12 (or something 1, anyway) Can we just link /linuxrc to /bin/true? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: tasksel bug-scrub for woody; translations; plans for 1.13
Randolph Chung [EMAIL PROTECTED] writes: 108061: java doesn't work on many of the debian architectures yet, so i'd rather not do this for now. gcj-3.0 and kaffe combined run on at least an important subset of them (specifically alpha, i386, m68k, powerpc, sparc, arm). I think it would be worthwhile to have a Java task with these for woody. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [again] basedebs.tgz: Supported or not?
Chris Tillman [EMAIL PROTECTED] writes: I understand there is a capability provided to use a basedebs.tgz file to install base from the hard disk in lieu of a network install. This was particularly needed for those who connect via a modem, because ppp is not available in the pre-base installer system. But there still isn't any basedebs.tgz in the built distribution yet, AFAIK. And as a result, I haven't documented how to use it. That is a big hole in boot-floppies, and in the docs, in fact I read here this week that it was unacceptable and I agree. I don't think we need to regenerate the basedebs.tgz at every build of b-f, especially now that base is frozen. As far as I understand things, the basedebs.tgz is essentially independent of b-f. It's just a snapshot of woody at a particular time. I guess what we will have to do is generate it occasionally and dump it in the current/ directory of disks-$ARCH on the mirrors. It kind of sucks to add a large tarball to the mirrors of what is essentially just the .debs they are already mirroring, plus a little extra data, but I don't see a way around it. Opinions/corrections? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Debian por PowerPC
[EMAIL PROTECTED] (Alejo Escobar) writes: [...] The problem is with the mouse. One of the very last messages in the installation is about the mouse. It says that if the mouse is not recognized in X environment, i can delete /dev/mouse and hope the system hardware scan could guess the conection. (more or less). The mouse is not recognized. I power-off the system and deleted the /dev/mouse file but it didn't work. Try using /dev/input/mice. This question is better suited for debian-powerpc or debian-user, by the way. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: teamwork (was Re: #103302 ask permission blah blah)
Adam Heath [EMAIL PROTECTED] writes: How can we trust someone who isn't one of us? Adam, this is ridiculous. We implicitly trust a heck of a lot of people who aren't one of us. Have you audited all the code of all the non Debian-native programs you maintain (whose authors were not Debian developers)? Do you expect other Debian developers to? I don't see how you have any cause at all to imply Ethan is sabotaging the boot-floppies. He does a lot of good work for this project, and all I can say is that I hope we can get over these differences. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: #103302 ask permission before erasing /var/cache/apt/archives
Ethan Benson [EMAIL PROTECTED] writes: Joey, you SHOULD quit this project. You are of no use to it any longer, you are publically displaying an absolute intent to force YOUR system administration policies down upon ALL of debian's users. Ethan, I think losing Joey would be a heck of a lot worse than this bug not getting fixed. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: #103302 ask permission before erasing /var/cache/apt/archives
Anthony Towns [EMAIL PROTECTED] writes: Uh. There doesn't appear to be a patch in the BTS for the 0.1.15.8 upload. There should be. Ok, I added it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
race condition in extract_base.c
I think there is a race condition when we're executing debootstrap
that will lead to random errors and zombie debootstrap processes.
Could someone sanity check this patch? I haven't been able to test
it, but it looks like it should be right.
--- extract_base.c.~1.61.~ Sat Nov 3 10:28:01 2001
+++ extract_base.c Fri Nov 9 08:44:52 2001
@@ -220,7 +220,7 @@
boxProgress(PROGRESS_CLOSE, NULL);
-if (waitpid(pid, status, WNOHANG) (WIFEXITED(status) != 0 ))
+if (waitpid(pid, status, 0) (WIFEXITED(status) != 0 ))
{
rv = WEXITSTATUS(status);
if (rv != 0 ) {
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: debiam installation
daniele-rosa [EMAIL PROTECTED] writes: i'm daniele rosa. i'd like to install the debian version of linux. it's my first contact with this OS. how can i have the boot floppies?. There are installation instructions here: URL:http://www.debian.org/releases/stable/#new-inst -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: boot-floppie 3.0.13 success, of sorts
John H. Robinson, IV [EMAIL PROTECTED] writes: major frownies: (again, who to bug? netbase?) smtp is probably exim, which should be secure enough (although I prefer Postfix personally). The others are from netkit-inetd. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Still no base tarball
Anthony Towns [EMAIL PROTECTED] writes: And you do realise we're still months away from release, and that without -testing doing anything, it's just getting further away? AFAIK, absolutely *no one* has even tried a basedebs install yet, eg. I've done an install using basedebs.tgz over NFS on PowerPC. Worked fine. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: New cvs snapshot of woody boot-floppies (supports basedebs.tgz)
Russell Hires [EMAIL PROTECTED] writes: Irritating is, it says searching for Release when in fact its searching for basedebs.tgz. Yeah! I second that...of course, after unpacking the basedebs, then it's getting the Release file... My fault. I think this patch should do the trick. Could someone apply it? --- choose_medium.c.~1.108.~Fri Jul 13 12:38:38 2001 +++ choose_medium.c Tue Jul 17 00:09:16 2001 @@ -422,9 +422,15 @@ free (mountpoint); mountpoint = strdup (buffer); } - snprintf(prtbuf, sizeof(prtbuf), -_(The installation program is building a list of all directories containing the file '%s' that can be used to install %s.), -pattern, descr); + if (backuppattern) + snprintf(prtbuf, sizeof(prtbuf), +_(The installation program is building a list of all +directories containing a file '%s' or '%s' that can be used to install %s.), +pattern, backuppattern, descr); + else + snprintf(prtbuf, sizeof(prtbuf), +_(The installation program is building a list of all +directories containing a file '%s' that can be used to install %s.), +pattern, descr); + pleaseWaitBox(prtbuf); dirc = 0; dirv = (char **) malloc(5 * sizeof(char *)); It finds it but then says there was a problem unpacking the tarball (the acutal error message is overwriten by that window, console-3 only tells me there as a problem, like in the error window). Didn't have this problem at all. And it does not look as if I had run out of space, the partition is only ~100MB, but there are still 68MB available. I have tons of space... :-) But a slow disk. :-\ But maybe other testers have more luck... Christian Well, I'm stuck on the 4403980 4403980 Downloading packages message. For at least 30 minutes. I don't think I crashed, because the cursor is blinking to the right, and there is occasional disk activitiy --- this is all on my LCII (m68K mac). I don't know what's going on here...This machine is not connected to anything internet related (yet), though it is connected via a serial cable to my Powermac... Russell -- 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: New cvs snapshot of woody boot-floppies (supports basedebs.tgz)
Colin Walters [EMAIL PROTECTED] writes:
My fault. I think this patch should do the trick. Could someone
apply it?
Might as well fix the other things I broke while I'm in there. This
patch supersedes the previous one. Lightly tested.
Index: utilities/dbootstrap/choose_medium.c
===
RCS file: /cvs/debian-boot/boot-floppies/utilities/dbootstrap/choose_medium.c,v
retrieving revision 1.108
diff -u -r1.108 choose_medium.c
--- utilities/dbootstrap/choose_medium.c2001/07/13 02:33:02 1.108
+++ utilities/dbootstrap/choose_medium.c2001/07/17 04:56:30
@@ -422,9 +422,15 @@
free (mountpoint);
mountpoint = strdup (buffer);
}
- snprintf(prtbuf, sizeof(prtbuf),
-_(The installation program is building a list of all directories
containing the file '%s' that can be used to install %s.),
-pattern, descr);
+ if (backuppattern)
+ snprintf(prtbuf, sizeof(prtbuf),
+_(The installation program is building a list of all
+directories containing a file '%s' or '%s' that can be used to install %s.),
+pattern, backuppattern, descr);
+ else
+ snprintf(prtbuf, sizeof(prtbuf),
+_(The installation program is building a list of all
+directories containing a file '%s' that can be used to install %s.),
+pattern, descr);
+
pleaseWaitBox(prtbuf);
dirc = 0;
dirv = (char **) malloc(5 * sizeof(char *));
@@ -503,8 +509,12 @@
snprintf(prtbuf, sizeof(prtbuf),
_(The installation program couldn't find any directory
containing the files %s.), buffer);
} else {
- snprintf(prtbuf, sizeof(prtbuf),
-_(The installation program couldn't find any directory
containing the file %s.), pattern);
+ if (backuppattern)
+ snprintf(prtbuf, sizeof(prtbuf),
+_(The installation program couldn't find any
+directory containing a file %s or %s.), backuppattern, pattern);
+ else
+ snprintf(prtbuf, sizeof(prtbuf),
+_(The installation program couldn't find any
+directory containing the file %s.), pattern);
}
problemBox(prtbuf, _(File not found!));
status = -1;
@@ -521,15 +531,27 @@
else if (status != DLG_OKAY) {
status = -1;
} else {
+ int found;
Archive_Dir = strdup(buffer);
snprintf(prtbuf, sizeof(prtbuf), %s/%s, Archive_Dir,
pattern);
- /* FIXME: add support for backuppattern here */
- if (!NAME_ISREG(prtbuf, statbuf)) {
+ found = NAME_ISREG(prtbuf, statbuf);
+ if (!found backuppattern) {
+ snprintf(prtbuf, sizeof(prtbuf), %s/%s, Archive_Dir,
+backuppattern);
+ found = NAME_ISREG(prtbuf, statbuf);
+ }
+
+ if (!found) {
INFOMSG(manual dir specified for %s, but %s not found, descr,
prtbuf);
- snprintf(prtbuf, sizeof(prtbuf),
-_(%s does not contain the file %s that is needed to
install %s),
-Archive_Dir, pattern, descr);
+ if (backuppattern)
+ snprintf(prtbuf, sizeof(prtbuf),
+_(%s does not contain a file %s or %s that is needed
+to install %s),
+Archive_Dir, pattern, backuppattern, descr);
+ else
+ snprintf(prtbuf, sizeof(prtbuf),
+_(%s does not contain the file %s that is needed to
+install %s),
+Archive_Dir, pattern, descr);
problemBox(prtbuf, _(File not found!));
status = -1;
} else {
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
possibly working basedebs support
Ok,
This patch appears to work for me. A few things still need to be
fixed, but I'm posting this here now because it's my first patch for
b-f that is somewhat intrusive, so if I have the totally wrong
approach or anything, I'd like to know, before I finish shaking out
any bugs.
I still need to test if this breaks anything else...
Index: utilities/dbootstrap/choose_medium.c
===
RCS file: /cvs/debian-boot/boot-floppies/utilities/dbootstrap/choose_medium.c,v
retrieving revision 1.106
diff -u -r1.106 choose_medium.c
--- utilities/dbootstrap/choose_medium.c2001/07/05 14:13:49 1.106
+++ utilities/dbootstrap/choose_medium.c2001/07/12 07:10:36
@@ -23,6 +23,7 @@
int dirc = 0;
char **dirv;
char *pattern;
+char *backuppattern;
/* This and several other globals probably belong in a header also once more of
this is librified. */
char *cdrom_device = NULL;
@@ -212,7 +213,7 @@
{
char buffer[PATH_MAX+1];
char buffer2[PATH_MAX+1];
-int i;
+int i, found;
char *mountpoint, *descr = , *def;
static char *preventry = NULL;
/* glob_t globbuf; */
@@ -284,6 +285,7 @@
break;
case debootstrap:
pattern = DEBIAN_ARCHIVE_PREFIX ARCHNAME /Release;
+ backuppattern = BASETGZ;
descr = _(the Base system);
break;
#if #cpu(alpha)
@@ -312,33 +314,38 @@
else
snprintf(prtbuf, sizeof(prtbuf),
%s/ DEBIAN_ARCHIVE_PREFIX ARCHNAME /Release, mountpoint);
-
-if (NAME_ISREG(prtbuf, statbuf)) {
- DEBUGMSG(located file at %s, prtbuf);
- /* do an additional test when installing kernel and modules,
- since we need drivers disk */
- if (disqtype == kernel) {
- snprintf(prtbuf, sizeof(prtbuf),%s/ ARCHIVE_LOCATION /%s,
-mountpoint, ARCHNAME, drivers_path);
- if (NAME_ISREG(prtbuf, statbuf)) {
- /* success */
- snprintf(prtbuf, sizeof(prtbuf), %s/ ARCHIVE_LOCATION,
- mountpoint, ARCHNAME);
- def = strdup(prtbuf);
- } else {
- INFOMSG(found kernel, but no drivers at %s, prtbuf);
- }
- } else {
- /* success, the file is where we expect it */
- if (disqtype != debootstrap ) {
- snprintf(prtbuf, sizeof(prtbuf), %s/ ARCHIVE_LOCATION,
- mountpoint, ARCHNAME);
- def = strdup(prtbuf);
+
+found = NAME_ISREG(prtbuf, statbuf);
+if (!found backuppattern) {
+ snprintf(prtbuf, sizeof(prtbuf), %s/%s,
+mountpoint, backuppattern);
+ found = NAME_ISREG(prtbuf, statbuf);
+}
+
+if (found) {
+ DEBUGMSG(located file at %s, prtbuf);
+ /* do an additional test when installing kernel and modules,
+ since we need drivers disk */
+ if (disqtype == kernel) {
+ snprintf(prtbuf, sizeof(prtbuf),%s/ ARCHIVE_LOCATION /%s,
+mountpoint, ARCHNAME, drivers_path);
+ if (NAME_ISREG(prtbuf, statbuf)) {
+ /* success */
+ snprintf(prtbuf, sizeof(prtbuf), %s/ ARCHIVE_LOCATION,
+mountpoint, ARCHNAME);
+ def = strdup(prtbuf);
+ } else {
+ INFOMSG(found kernel, but no drivers at %s, prtbuf);
+ }
+ } else if (disqtype == debootstrap) {
+ /* If we found basedebs.tgz, tell debootstrap about it */
+ def = strdup(strstr(prtbuf, BASETGZ) ? prtbuf : mountpoint);
+ DEBUGMSG(disqtype = debootstrap, def: %s, def ? def : (null));
} else {
- /* special handling for debootstrap, it wants the mountpoint */
- def = strdup(mountpoint);
+ snprintf(prtbuf, sizeof(prtbuf), %s/ ARCHIVE_LOCATION,
+mountpoint, ARCHNAME);
+ def = strdup(prtbuf);
}
- }
} else if (disqtype == kernel) {
/* try a flat dir structure */
INFOMSG(_(no file '%s' found, kernel not in default location), prtbuf);
@@ -363,21 +370,34 @@
INFOMSG(_(no file '%s' found, Debian archive not in default location),
prtbuf);
}
+
+DEBUGMSG(def: %s, def ? def : (null));
+
if (def ! bootargs.isverbose) {
/* there's a default, lets go with it */
status = 0;
} else {
if (def) {
- snprintf(prtbuf, sizeof(prtbuf),
-_(Please select the directory containing a file %s that you will
use to install %s.\nUnless you really know what you need choose the default.),
-pattern, descr);
+ if (backuppattern)
+ snprintf(prtbuf, sizeof(prtbuf),
+_(Please select the directory containing a file %s or %s
+that you will use to install %s.\nUnless you really know what you need choose the
+default.),
+pattern, backuppattern, descr);
+ else
+ snprintf(prtbuf,
Re: status of basedebs support in boot-floppies?
Anthony Towns [EMAIL PROTECTED] writes: How's this going now? Getting a 3.0.8 out that supports basedebs properly (ie, so you don't have to switch to VC2) would be helpful for getting a bit better testing (hopefully). I'm working on it (Ethan conned me into it :) ), albeit very slowly at the moment. I expect to have something working by this weekend, unless someone else beats me to it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#99652: dialog boxes confusing - active choice not clear
David Whedon [EMAIL PROTECTED] writes: An easy option which may or may not be any better is to not use the 'compact' version of buttons for yes/no boxes. An example of this can be found here: http://people.debian.org/~dwhedon/yesno_proposed.png I think that this is much clearer, and it would certainly alleviate my concerns. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#99652: dialog boxes confusing - active choice not clear
Package: boot-floppies Severity: normal The (ncurses?) dialog boxes for the boot-floppies are confusing, because unless one knows in advance that the blue item is the active one in a Yes/No question, one doesn't know which choice one is picking, especially since the default varies per question. I helped a friend with a B.S. in Computer Science do a Debian install yesterday, and he was quite confused at first until I told him that the blue choice was active. I can think of a couple of ways to fix this. The easiest would probably be to not make movement between the Yes/No choices circular; i.e. if one hits the right arrow when the active choice is No, it just stays there. That way, one could, by using a bit of deduction, figure out what the active choice is. Another alternative might be to put some sort of indicator under the active choice, like '^', '*', or '---^---', etc. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
