10 years agoMerge branch 'devel'
Tony Breeds [Fri, 16 Jul 2010 05:11:38 +0000 (15:11 +1000)]
Merge branch 'devel'

10 years agoextract_netinfo_args() should be a void function.
Tony Breeds [Sat, 6 Mar 2010 03:59:04 +0000 (14:59 +1100)]
extract_netinfo_args() should be a void function.

If there is no "netinfo" packet, extract_netinfo_args() will fail and cause
parse_device_path() to abort.  This basically meant that yaboot will fail to
load any kernel/initrd under those circumstances.

This fix changes extract_netinfo_args() to be a void function.

Signed-off-by: Tony Breeds <tonyb@au1.ibm.com>
10 years agoMerge branch 'netboot'
Tony Breeds [Fri, 16 Jul 2010 04:56:12 +0000 (14:56 +1000)]
Merge branch 'netboot'

10 years agoYaboot 1.3.16 yaboot-1.3.16
Tony Breeds [Fri, 9 Jul 2010 02:18:17 +0000 (12:18 +1000)]
Yaboot 1.3.16

10 years agoAdd "Developers Certificate of Origin"
Tony Breeds [Fri, 9 Jul 2010 02:16:35 +0000 (12:16 +1000)]
Add "Developers Certificate of Origin"

All further patch contributions will need to adhere to this document,
based on the linux kernel's process.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
10 years agoVersion 1.3.15 yaboot-1.3.15
Tony Breeds [Fri, 9 Jul 2010 02:14:10 +0000 (12:14 +1000)]
Version 1.3.15

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoCheck each partition for swapspace before other filesystems.
Tony Breeds [Fri, 23 Oct 2009 03:05:50 +0000 (14:05 +1100)]
Check each partition for swapspace before other filesystems.

Swapspace isn't guaranteed to be on a partition tagged as such, add a
little defensive programming in that case.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoDo not of_open() LINUX_NATIVE partitions.
Tony Breeds [Thu, 22 Oct 2009 05:20:47 +0000 (16:20 +1100)]
Do not of_open() LINUX_NATIVE partitions.

If for some reason a partition is LINUX_NATIVE but is not handled by one
of the existing filesystem drivers, of_open() /may/ appear to succeed,
but infact the ihandle is garbage.

Make sure we skip LINUX_NATIVE partitions.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoMerge branch 'birecs' into devel
Tony Breeds [Thu, 22 Oct 2009 05:17:10 +0000 (16:17 +1100)]
Merge branch 'birecs' into devel

11 years agoHandle ipv6 boot parameters for POWER architecture.
Tony Breeds [Thu, 22 Oct 2009 04:53:00 +0000 (15:53 +1100)]
Handle ipv6 boot parameters for POWER architecture.

This is implementation derived.

This follows the semantics defined in section 4.3.1 of
(It is under the Members area of TSC - Platform Architecture committee).

[ Fixed merge conflicts in second/file.c and second/fs_of.c ]

Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoParse the subnetmask from bootpath and DHCP options
Tony Breeds [Thu, 22 Oct 2009 04:46:37 +0000 (15:46 +1100)]
Parse the subnetmask from bootpath and DHCP options

Some power6 machines need to have the subnetmask to correctly open()
the network device.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoUse an empty server address for "older" firmwares.
Tony Breeds [Thu, 7 May 2009 03:07:21 +0000 (13:07 +1000)]
Use an empty server address for "older" firmwares.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoCheck for valid "looking" IPs before saving them.
Tony Breeds [Wed, 6 May 2009 06:05:46 +0000 (16:05 +1000)]
Check for valid "looking" IPs before saving them.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoAdd parseing for Vendor options aka DHCP options.
Tony Breeds [Tue, 5 May 2009 06:14:20 +0000 (16:14 +1000)]
Add parseing for Vendor options aka DHCP options.

If we have DHCP options process them.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
11 years agoRemove Bootinfo (BIRECS).
Tony Breeds [Fri, 2 Oct 2009 05:55:33 +0000 (15:55 +1000)]
Remove Bootinfo (BIRECS).

The are not used and can scribble on kernel memory.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agoUp LOAD_BUFFER_SIZE to 25Mb.
Tony Breeds [Tue, 5 May 2009 00:52:02 +0000 (10:52 +1000)]

20MB is already too small for some distro installers.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agoEnhance the parseing of IPv4 information.
Tony Breeds [Thu, 5 Mar 2009 04:04:47 +0000 (15:04 +1100)]
Enhance the parseing of IPv4 information.

Pull information out of the bootp-reply (or similar) property.  This
successfully netboots in a wider variety of OF "boot" lines.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agoMove ipv4 arguments handling to a separate function. No fucntional change.
Chandra Seetharaman [Fri, 9 Jan 2009 13:59:18 +0000 (05:59 -0800)]
Move ipv4 arguments handling to a separate function. No fucntional change.

This is just in preparation for the next patch which adds supports for

Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>

12 years agoFix netboot fucntionality to use the parameters specified.
Chandra Seetharaman [Fri, 9 Jan 2009 13:57:44 +0000 (05:57 -0800)]
Fix netboot fucntionality to use the parameters specified.

Currently, yaboot works properly when there is a tftp boot server serving
the broadcast requests sent in the network.

If the user specifies siaddr, ciaddr, and other arguments in the command
line, it is not handled appropriately.

This patch changes that behavior and make yaboot work properly in both cases.

The semantics specified in section 4.1 of
http://playground.sun.com/1275/practice/obp-tftp/tftp1_0.pdf is followed.

Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>

12 years agoPrepare for netboot fix(following patch). No Functional change.
Chandra Seetharaman [Fri, 9 Jan 2009 13:56:40 +0000 (05:56 -0800)]
Prepare for netboot fix(following patch). No Functional change.

This patch just replaces the arguments "char *dev_name" and "char *file_name"
with "struct boot_fspec_t fspec".

Signed-off-by: Chandra Seetharaman <sekharan@us.ibm.com>

12 years agoIncrease the LOAD_BUFFER_SIZE to 20MiB.
Tony Breeds [Tue, 17 Feb 2009 23:20:12 +0000 (10:20 +1100)]
Increase the LOAD_BUFFER_SIZE to 20MiB.

As installers add more features initrd's get bigger.  Lets try to
accommodate them.

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agolong append="..." causes yaboot to reject entire yaboot.conf (MAX_TOKEN too small) 1.3.15-merge yaboot-1.3.15-rc
Will Woods [Thu, 13 Nov 2008 00:59:43 +0000 (00:59 +0000)]
long append="..." causes yaboot to reject entire yaboot.conf (MAX_TOKEN too small)

yaboot's second/cfg.c has:
  #define MAX_TOKEN 200

When yaboot parses its config file, if it encounters a quoted string (like
append="...") longer than MAX_TOKEN, it rejects the entire config.

Typical preupgrade usage uses ~250 chars, which means preupgrade currently
renders ppc systems unbootable without this patch.

Suggest increasing MAX_TOKEN to.. let's say 511. cfg.c does "char
buf[MAX_TOKEN+1]", and everyone loves powers of 2.

Signed-off-by: Roman Rakus <rrakus@redhat.com>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agoybin --bootonce doesn't work
Will Woods [Fri, 14 Nov 2008 04:55:53 +0000 (04:55 +0000)]
ybin --bootonce doesn't work

1) It inverts the logic of a test to see if nvram is available, and
2) It assumes the output of 'nvsetenv VARNAME' will be 'VARNAME=VALUE', rather
than just 'VALUE' (as it is when using the "nvram" utility).

From bugzilla:

The attached patch fixes these problems and makes --bootonce work as expected.

Signed-off-by: Roman Rakus <rrakus@redhat.com>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agoRevert "Let prom_claim_chunk do the work for tftp images to support larger images."
Tony Breeds [Fri, 13 Feb 2009 02:51:03 +0000 (13:51 +1100)]
Revert "Let prom_claim_chunk do the work for tftp images to support larger images."

This reverts commit c8b04c614770870d75825c19bcd4729cc779b84a.

This fixes the regression described in:

Signed-off-by: Roman Rakus <rrakus@redhat.com>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agoybin succeeds but returns error code
Anton Blanchard [Thu, 5 Feb 2009 14:19:26 +0000 (14:19 +0000)]
ybin succeeds but returns error code


On a System p box, ybin returns an error even though it succeeds:

# ybin
# echo $?

The patch below fixes it:

# ybin
# echo $?

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
12 years agoAllow yaboot to load relocatable kernel
Mohan Kumar M [Fri, 12 Dec 2008 03:31:38 +0000 (03:31 +0000)]
Allow yaboot to load relocatable kernel

PPC64 relocatable kernels (built with CONFIG_RELOCATABLE=y) have the type of
ET_DYN. But yaboot code won't load the kernel if the ELF type is not
ET_EXEC. Attached patch adds support to yaboot to load relocatable kernels
also (ie load ET_DYN type also)

Signed-off-by: M. Mohan Kumar <mohan@in.ibm.com>
Acked-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
13 years agoPegasos and partition numbering
David Woodhouse [Fri, 28 Mar 2008 00:04:02 +0000 (00:04 +0000)]
Pegasos and partition numbering

The whole bplan partition numbering thing is a clusterfuck, but it
shouldn't be as crappy as it is.

We can _cope_ with the fact that firmware is broken and has an
off-by-one in its partition numbering. All we need to do in yaboot is
add 1 to the partition numbers we infer from in /chosen/boot-device, if
we detect that we're on an afflicted machine. That's all we need to do,
and all we ever _should_ have done. And is already in yaboot. Yaboot
does its own block device handling, including partitions, and doesn't
ever care about the firmware's problems (in that respect, at least).

It doesn't affect the installer at _all_, because the installer can't
set up the firmware's boot-device anyway and it has to be done by hand.
All it affects is the release notes telling the user how to make it

Unfortunately, we didn't make it that simple when we first supported the
Pegasos. For reasons which aren't entirely clear to me, we ended up with
Amiga partition table support in yaboot with the _same_ off-by-one
error, to match the firmware. And thus we have hacks in the installer to
use amiga partitions for Pegasos, and to cope with the off-by-one crap.

It's only after avoiding all this crap purely by accident on Efika, by
using DOS partition tables, that I realise how stupid I was to blindly
copy the crap that other people were doing, and to believe that Pegasos
would only work with Amiga partition tables.

I'd like to get rid of the off-by-one bug in yaboot's Amiga partition
handling. At the moment, our simple 'if bplan, partition++' in
yaboot_main() is wrong when we have Amiga partitions, although it's fine
for other partition types.

Actually, I'd also like to make that same increment conditional on
!conf_given, so that if someone specifies 'conf=hd:1,/yaboot.conf' on
the command line, that partition number _isn't_ incremented.

So any time you see a proper path specified as 'dev:part,/path/name' you
know it's a real one with proper partition numbers. Remember, the bplan
firmware doesn't allow that form, and takes a space between the
'dev:part' bit and the filename:
boot hd:0 /yaboot/yaboot conf=hd:1,/yaboot/yaboot.conf

Fixing the off-by-one bug in the Amiga partition handling means that
upgrades might break. I suppose we could have a 'noamigaoffbyone'
configuration option which all newly-written yaboot.conf files would
have, which controls this behaviour. But to be honest I just don't think
it's worth it.

13 years agouse public interface to detect CAS reboots (take 2)
Mike Wolf [Tue, 15 Jan 2008 20:33:51 +0000 (14:33 -0600)]
use public interface to detect CAS reboots (take 2)

This is the second pass of this patch.  The fw variable name is changed
to use ibm,client-architecture-support-reboot field which is already present
in the firmware.  Please disregard the previous patch.

The firmware field used to detect CAS reboots (ibm,fw-nbr-reboots) is
really a private field that could change without warning.  A new field
ibm,#reconfig-reboots will be added as a public interface intended to
be used for this detection.  The patch will first check for
ibm,client-architecture-support-reboot if that is not found it will see
if ibm,fw-nbr-reboots is present and will use that instead.


13 years agoThe CAS (Client-Architecture Support) call tells firmware what capabilities the
Paul Nasrat [Fri, 4 Jan 2008 13:00:19 +0000 (13:00 +0000)]
The CAS (Client-Architecture Support) call tells firmware what capabilities the
OS has.  These capabilities result in different modes which the device-tree is
configured in, as well as what processor capabilities are presented.  So, if
the capabilities are different from what was previously booted, firmware has to
reboot to reconfigure the device-tree.  The second boot will have the updated
device-tree and we can boot as normal.

When this firmware initiated reboot occurs yaboot will now boot the same kernel
as the previous boot attempt with no action by the user needed.  I have
successfully booted on POWER5 and POWER6 machines using various levels of the

Patch from: Mike Wolf <mjw@linux.vnet.ibm.com>

13 years agoRevert "The attached patch adds support for writing the nvram using the nvram"
Paul Nasrat [Thu, 3 Jan 2008 19:21:44 +0000 (19:21 +0000)]
Revert "The attached patch adds support for writing the nvram using the nvram"

This reverts commit ac8fbd5ad52b43b820973f1828a29f85bed13a1d.

13 years agoRevert "Fix shell syntax"
Paul Nasrat [Thu, 3 Jan 2008 19:21:22 +0000 (19:21 +0000)]
Revert "Fix shell syntax"

This reverts commit 24679e706305e0910f181a9640b8b88a30662d03.

13 years agoFix shell syntax
Paul Nasrat [Thu, 3 Jan 2008 19:07:50 +0000 (19:07 +0000)]
Fix shell syntax

13 years agoThe attached patch adds support for writing the nvram using the nvram
Paul Nasrat [Tue, 4 Dec 2007 08:43:32 +0000 (08:43 +0000)]
The attached patch adds support for writing the nvram using the nvram
utility from the new powerpc-utils pkg from IBM; it also simplify a bit
the code.

Patch from: Emanuele Giaquinta (exg@gentoo.org)

13 years agoForce build to be ppc32
Paul Nasrat [Mon, 3 Dec 2007 14:21:11 +0000 (14:21 +0000)]
Force build to be ppc32

13 years agoFix typo
Paul Nasrat [Mon, 3 Dec 2007 14:14:42 +0000 (14:14 +0000)]
Fix typo

13 years agoFix typo in patch porting
Paul Nasrat [Mon, 3 Dec 2007 14:13:54 +0000 (14:13 +0000)]
Fix typo in patch porting

13 years agoFix partition numbering handling with bplan firmware
Paul Nasrat [Mon, 3 Dec 2007 14:05:48 +0000 (14:05 +0000)]
Fix partition numbering handling with bplan firmware
Patch based on Fedora patch by dwmw2@infradead.org

13 years agoTry harder to allocate malloc region
Paul Nasrat [Mon, 3 Dec 2007 13:17:33 +0000 (13:17 +0000)]
Try harder to allocate malloc region
Patch from dwmw2@infradead.org

13 years agoPrint version-release of yaboot
Paul Nasrat [Mon, 3 Dec 2007 13:13:23 +0000 (13:13 +0000)]
Print version-release of yaboot
Patch from dwmw2@infradead.org

13 years agoLet prom_claim_chunk do the work for tftp images to support larger images.
Paul Nasrat [Mon, 3 Dec 2007 13:04:19 +0000 (13:04 +0000)]
Let prom_claim_chunk do the work for tftp images to support larger images.

Patch from: Leonardo Rangel <lrangel@linux.vnet.ibm.com>

13 years agoThese patch provides the function of load a alternative initrd file
Paul Nasrat [Mon, 3 Dec 2007 12:41:42 +0000 (12:41 +0000)]
These patch provides the function of load a alternative initrd file
specified by the user via Yaboot prompt. To load the new file, just use
the "initrd=<name_of_file>" command. The <name_of_file> could include
all the directory path to the file.

Using only the "initrd" command, Yaboot will load the default kernel
specified in the configuration file and the initrd file specified by the
user. The user can, also, specify an initrd file to any other label of
the configuration file, just adding the "initrd" command before the
label's name.

Signed-off-by: Paulo Ricardo Paz Vital <vital@br.ibm.com>
13 years agomove the tftp load space and use prom_claim_chunk to claim the memory.
Paul Nasrat [Mon, 3 Dec 2007 12:39:06 +0000 (12:39 +0000)]
move the tftp load space and use prom_claim_chunk to claim the memory.
Growing kernels mean that the 6MB that was allotted is not enough.

Patch from Scott Moser

13 years agoThe patch from Scott Moser moves the "search-for-prom_claim'able" routine that
Paul Nasrat [Mon, 3 Dec 2007 12:37:10 +0000 (12:37 +0000)]
The patch from Scott Moser moves the "search-for-prom_claim'able" routine that
was present in load_elf64 and load_elf32 to a function named
prom_claim_chunk.  This reduces the code-snippit duplication and makes
the function available for of_net_open.

Scott Moser ssmoser@us.ibm.com

13 years agoRelease 1.3.14 yaboot-1.3.14
Paul Nasrat [Fri, 17 Aug 2007 14:29:26 +0000 (15:29 +0100)]
Release 1.3.14

13 years agoOn some of the larger kernels we are starting to see a space squeeze.
Paul Nasrat [Fri, 17 Aug 2007 14:22:00 +0000 (15:22 +0100)]
On some of the larger kernels we are starting to see a space squeeze.
The kernel is currently being put at 20MBs and on some of the newer
pSeries systems we are getting real-bases at 32MBs (plus AIX 5.3 has
real-base at 32MB, so if AIX is ever booted, everything gets shifted).

If the uncompressed kernel+initrd is larger than 12MBs then we see
truncation (typical on install kernels).

This patch moves the KERNELADDR to 0x00C00000 (12MB).  If real-base is
12MB and is loaed there, yaboot will keep going up by a megabyte until
prom_claim finds space.  And the uncompressed kernel should be
indifferent where it's loaded because there's nothing at 32MB (ie it
will operate as it does today).  If OF is loaded at 32MB then the 12MB
region is free and the uncompressed kernel will have more space to grow.

I have tested this on numerous pSeries boxes.  I have not tested on
anything else.

Signed-off-by: Jake Moilanen <moilanen@austin.ibm.com>
13 years agoFix typo courtesy of Kenichi Nagai
Paul Nasrat [Thu, 26 Jul 2007 08:27:58 +0000 (09:27 +0100)]
Fix typo courtesy of Kenichi Nagai

13 years agoPrepare 1.3.14
Paul Nasrat [Thu, 19 Jul 2007 13:16:30 +0000 (14:16 +0100)]
Prepare 1.3.14

14 years agoNetboot fixes:
Paul Nasrat [Thu, 17 May 2007 12:54:20 +0000 (13:54 +0100)]
Netboot fixes:
- the ARRAY_SIZE macro is no more needed here,
- use of cfgpath to have the actual config path to apply for mac *and* ip
- no need to free a buffer big enough, and besides malloc(9) was not
enough ("/etc/" missed),
- use of intermediate length variables to avoid unneeded calls to strlen()
and strrchr().
benoit.guillon <benoit.guillon@tele2.fr>

14 years agoCorrect return type
Paul Nasrat [Thu, 17 May 2007 10:30:59 +0000 (11:30 +0100)]
Correct return type
Leonardo Rangel <lrangel@linux.vnet.ibm.com>

14 years agoAdd missing includes
Paul Nasrat [Thu, 17 May 2007 10:29:29 +0000 (11:29 +0100)]
Add missing includes
Leonardo Rangel <lrangel@linux.vnet.ibm.com>

14 years agoFix compilation warnings by using return not exit.
Paul Nasrat [Thu, 17 May 2007 10:28:19 +0000 (11:28 +0100)]
Fix compilation warnings by using return not exit.
Leonardo Rangel <lrangel@linux.vnet.ibm.com>

14 years agoNow we stop truncating characters for pxelinux style boot at the directory seperator.
Paul Nasrat [Thu, 17 May 2007 10:27:23 +0000 (11:27 +0100)]
Now we stop truncating characters for pxelinux style boot at the directory seperator.
Leonardo Rangel <lrangel@linux.vnet.ibm.com>

14 years agoUpdate ChangeLog from git yaboot-1.3.14rc2
Paul Nasrat [Wed, 16 May 2007 09:22:59 +0000 (10:22 +0100)]
Update ChangeLog from git

14 years agouse of cfg_set_default_by_mac() function in yaboot.c
Paul Nasrat [Tue, 15 May 2007 13:28:02 +0000 (14:28 +0100)]
use of cfg_set_default_by_mac() function in yaboot.c

This last part calls the cfg_set_default_by_mac() function in yaboot.c
by load_config_file() function.

Signed-off-by: Paulo Ricardo Paz Vital <vital@br.ibm.com>
14 years agoThis part creates the cfg_set_default_by_mac() function.
Paul Nasrat [Tue, 15 May 2007 13:25:55 +0000 (14:25 +0100)]
This part creates the cfg_set_default_by_mac() function.

This function sets the default cf_option if there is an image label if
the same MAC address of the machine is netbooting.

Signed-off-by: Paulo Ricardo Paz Vital <vital@br.ibm.com>
14 years agoload_my_config_file() modifications in yaboot.c
Paul Nasrat [Tue, 15 May 2007 13:24:06 +0000 (14:24 +0100)]
load_my_config_file() modifications in yaboot.c

This second part, deletes the netbooting like PXE algorithm from
yaboot.c (reallocated in prom.c) and sets the use of prom_get_netinfo(),
prom_get_mac (), prom_get_ip () functions in load_my_config_file()
function to netbooting work like before the changes.

Signed-off-by: Paulo Ricardo Paz Vital <vital@br.ibm.com>
14 years agoprom functions
Paul Nasrat [Tue, 15 May 2007 13:23:26 +0000 (14:23 +0100)]
prom functions

This first part creates the prom_get_netinfo (), prom_get_mac (),
prom_get_ip () functions, and some structures to get the necessary
information about the MAC and IP addresses.

Actually, this is a reallocation of netbooting like PXE algorithm wrote
by Benoit Guillon [1], to be used by any other part of Yaboot to get the
netbooting information.

[1] http://ozlabs.org/pipermail/yaboot-devel/2006-December/000062.html

Signed-off-by: Paulo Ricardo Paz Vital <vital@br.ibm.com>
14 years agoThis patch fixes OF netbooting on pSeries
Paul Nasrat [Tue, 15 May 2007 13:18:33 +0000 (14:18 +0100)]
This patch fixes OF netbooting on pSeries
Signed-off-by: Paulo Ricardo Paz Vital <vital@br.ibm.com>
14 years agoUser-specified config file on Yaboot's prompt
Paul Nasrat [Tue, 15 May 2007 13:14:43 +0000 (14:14 +0100)]
User-specified config file on Yaboot's prompt
This patch clears out the kernel image list before loading a new config file.
Leonardo Rangel lrangel at linux.vnet.ibm.com

14 years agoEnable user to supply config from yaboot command line.
Paul Nasrat [Thu, 19 Apr 2007 09:27:05 +0000 (10:27 +0100)]
Enable user to supply config from yaboot command line.

Leonardo Rangel <lrangel@linux.vnet.ibm.com>

14 years agoThe following patch fix reverts removal of some lines for the netboot patch.
Paul Nasrat [Thu, 19 Apr 2007 09:25:28 +0000 (10:25 +0100)]
The following patch fix reverts removal of some lines for the netboot patch.

Signed-off-by: Paulo Ricardo Paz Vital <vital@br.ibm.com>
14 years agoybin userland interface for specifying bootonce
Paul Nasrat [Thu, 19 Apr 2007 09:14:24 +0000 (10:14 +0100)]
ybin userland interface for specifying bootonce

Add an option to ybin called "--bootonce" that takes exactly one
argument, the label of the kernel to boot exactly once, on next reboot.

Do some sanity checking.  Grep through the yaboot config file to make
sure that label=WHATEVER exists somewhere, bail if not.  Try to prune
out comment lines.  Perhaps this might be implemented better.  It worked
will in my testing, though.

Set the boot-once OF environment variable to the --bootonce value.  Due
to the previously mentioned nvsetenv bug, this will fail silently if
you're trying to set boot-once on a system where boot-once doesn't
already exist in nvram.  Thus, immediately after doing the set, try and
read it back out of nvsetenv and make sure it landed there.  Bail if

Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
14 years agoyaboot menu user interface
Paul Nasrat [Thu, 19 Apr 2007 09:13:43 +0000 (10:13 +0100)]
yaboot menu user interface

This patch allows for denoting a kernel label as a "boot-once" label
when printing the menu of labels.

The label flagged as default is marked with an asterisk (*).  If a label
matches the global bootoncelabel, I'm marking it with an ampersand (&).
I'm ambivalent about the symbol used here, or even if it is.  But I
found this bit useful when debugging and testing the functionality, and
it's trivial to add cleanly.  I chose ampersand because it's immediately
left of the asterisk on my keyboard.

14 years agoread boot-once, zero out, set bootoncelabel as default
Paul Nasrat [Thu, 19 Apr 2007 09:12:22 +0000 (10:12 +0100)]
read boot-once, zero out, set bootoncelabel as default

This patch reads the boot-once environment variable from Open Firmware
and stores in a global variable, "bootoncelabel".

It then unconditionally zeros out the boot-once environment variable.

*** Note that according to 1275, setprop will create a specified
environment variable if it doesn't already exist.  A bug in nvsetenv
causes it to fail silently if you try to set an OF environment variable
that does not already exist.  However, simply booting a system running a
yaboot with this patch set compiled in will in fact create the boot-once
environment variable, which nvsetenv can later set and reset.

Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
14 years agoprom setprop primatives
Paul Nasrat [Thu, 19 Apr 2007 09:11:19 +0000 (10:11 +0100)]
prom setprop primatives

This patch creates the prom_set_options() and prom_setprop() functions.
These are precise mirrors of their read-only brothers,
prom_get_options() and prom_getprop(), implemented as defined in the
Open Firmware 1275 spec.

The prom_set_options() function will be used to zero out the boot-once
label unconditionally after reading it.

Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
14 years agoLoading a different config file (confarg feature) origin
Paul Nasrat [Wed, 18 Apr 2007 13:27:59 +0000 (14:27 +0100)]
Loading a different config file (confarg feature)
Origin: Paulo Ricardo Paz Vital <vital@br.ibm.com>

14 years agoRemove devel debugging code
Paul Nasrat [Wed, 7 Mar 2007 11:31:12 +0000 (11:31 +0000)]
Remove devel debugging code
Patch from Benoit Guillon

14 years agoCompile fix
Paul Nasrat [Wed, 7 Mar 2007 08:52:19 +0000 (08:52 +0000)]
Compile fix
Paulo Vital <vital at br.ibm.com>

14 years agoThis is a combination of three patches (the first two for adding
Paul Nasrat [Fri, 2 Mar 2007 16:36:30 +0000 (16:36 +0000)]
This is a combination of three patches (the first two for adding
prom_getproplen and ARRAY_SIZE) for getting the pxelinux config file
search working.

Nathan Lynch (original patch)
Benoit Guillon (update to 1.3.14rc tree)

14 years agoCertain levels of IBM firmware will allow the system to boot from an
Paul Nasrat [Fri, 2 Mar 2007 16:26:18 +0000 (16:26 +0000)]
Certain levels of IBM firmware will allow the system to boot from an
iscsi target.  System OFW accomplishes this by setting up a virtual
disk device with parameters.  These parameters, when passed back to
OFW by yaboot, directs the FW to use virtual device over the ethernet
port that will then access iscsi target as a block device.  This patch
extracts those parameters from the property of the virtual device and
passes them back to OFW to indicate the kernel is to be retrieved via
the iscsi protocol.

Signed-off-by: Doug Maxey <dwm@austin.ibm.com>
Cc: Ben Herrenscmidt <benh@kernel.crashing.org>
Cc: Mike Christie <michaelc@cs.wisc.edu>
14 years agoMerge ability to have per machine type config entries from ubuntu/silo.
Paul Nasrat [Thu, 15 Feb 2007 12:32:26 +0000 (12:32 +0000)]
Merge ability to have per machine type config entries from ubuntu/silo.

14 years agoDetermine last ext3 LBA to fix wild LBA reads
Paul Nasrat [Thu, 15 Feb 2007 12:24:28 +0000 (12:24 +0000)]
Determine last ext3 LBA to fix wild LBA reads
Signed-off-by: Doug Maxey <dwm@austin.ibm.com>
Cc: Ben Herrenschmidt <benh@kernel.crashing.org>
14 years agoRemove binary libext2fs.a and build against system library.
Paul Nasrat [Thu, 15 Feb 2007 12:21:14 +0000 (12:21 +0000)]
Remove binary libext2fs.a and build against system library.

14 years agoRemove binary libext2fs.a and build against system library.
Paul Nasrat [Thu, 15 Feb 2007 12:09:45 +0000 (12:09 +0000)]
Remove binary libext2fs.a and build against system library.

14 years agoDevice-trees don't have symlinks made any more
Paul Nasrat [Tue, 28 Nov 2006 14:36:02 +0000 (14:36 +0000)]
Device-trees don't have symlinks made any more

14 years agoAdd support to ofpath for usb-storage and fix sbp-2 storage
Paul Nasrat [Thu, 16 Nov 2006 12:34:02 +0000 (12:34 +0000)]
Add support to ofpath for usb-storage and fix sbp-2 storage
Patch from Alex Kanavin

14 years agoClean up yabootconfig kernel/initrd code - Joseph Jezak <josejx@gentoo.org>
Paul Nasrat [Tue, 11 Jul 2006 20:25:26 +0000 (16:25 -0400)]
Clean up yabootconfig kernel/initrd code -  Joseph Jezak <josejx@gentoo.org>

14 years agoFix for not all device trees having k2-sata@N/compatible (from gentoo)
Paul Nasrat [Tue, 11 Jul 2006 20:18:09 +0000 (16:18 -0400)]
Fix for not all device trees having k2-sata@N/compatible (from gentoo)

15 years agoMerge ubuntu ofpath fixes
Paul Nasrat [Wed, 10 May 2006 18:48:35 +0000 (14:48 -0400)]
Merge ubuntu ofpath fixes

15 years agoSweep all the files for trailing whitespace.
Doug Maxey [Tue, 25 Apr 2006 03:37:33 +0000 (22:37 -0500)]
Sweep all the files for trailing whitespace.

Signed-off-by: Doug Maxey <dwm@austin.ibm.com>
15 years agoUpdate to point to ozlabs pnasrat@redhat.com,yaboot--devel--1.3
Paul Nasrat [Fri, 21 Apr 2006 14:49:59 +0000 (10:49 -0400)]
Update to point to ozlabs

15 years agoupdate TODO
Paul Nasrat [Fri, 21 Apr 2006 14:47:04 +0000 (10:47 -0400)]
update TODO

15 years agoFix ofpath for G5 with multiple drives yaboot-1.3.14rc1
Paul Nasrat [Fri, 10 Feb 2006 01:29:18 +0000 (01:29 +0000)]
Fix ofpath for G5 with multiple drives

Fix ofpath for G5 with multiple drives

git-archimport-id: pnasrat@redhat.com/yaboot--devel--1.3--patch-3

15 years agopause after claim failed (Nathan Lynch)
Paul Nasrat [Fri, 10 Feb 2006 01:28:01 +0000 (01:28 +0000)]
pause after claim failed (Nathan Lynch)

pause after claim failed (Nathan Lynch)

git-archimport-id: pnasrat@redhat.com/yaboot--devel--1.3--patch-2

15 years agogit-archimport-id: pnasrat@redhat.com/yaboot--devel--1.3--patch-1
Paul Nasrat [Mon, 30 Jan 2006 19:34:23 +0000 (19:34 +0000)]
git-archimport-id: pnasrat@redhat.com/yaboot--devel--1.3--patch-1

15 years agoReplace netboot bootpath parser. erbenson@alaska.net--public,yaboot--devel--1.3 pnasrat@redhat.com,yaboot--devel--1.3--base-0
Ethan Benson [Mon, 26 Sep 2005 06:41:14 +0000 (06:41 +0000)]
Replace netboot bootpath parser.

* second/file.c: Replace bootpath parser for netboot cases, the new
  one should (hopefully) correctly handle most/all netboot
  situations. (Nathan Lynch)
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-88

15 years agoDon't use ELF entry point.
Ethan Benson [Mon, 26 Sep 2005 06:36:30 +0000 (06:36 +0000)]
Don't use ELF entry point.

* second/yaboot.c: Don't use ELF entry point anymore. (Olaf Hering)
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-87

15 years agoFix Cross-compile
Ethan Benson [Mon, 26 Sep 2005 06:32:56 +0000 (06:32 +0000)]
Fix Cross-compile

* Makefile: Fix ability to cross compile.
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-86

15 years agoBugfix to patch-83
Ethan Benson [Wed, 3 Aug 2005 03:12:29 +0000 (03:12 +0000)]
Bugfix to patch-83

* Correct dereference of NULL in RAID partition patch (patch-83).
  (Patch from Dustin).
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-85

15 years agoPretend to allocate/deallocate memory correctly
Ethan Benson [Fri, 29 Jul 2005 04:29:18 +0000 (04:29 +0000)]
Pretend to allocate/deallocate memory correctly

* Pretend like we actually manage memory properly by calling free() on
  things which were malloc()'ed or strdup()'ed.
  (Patch from Dustin Kirkland).
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-84

15 years agohandle RAID partitions on x86 partition tables
Ethan Benson [Sat, 9 Jul 2005 23:51:20 +0000 (23:51 +0000)]
handle RAID partitions on x86 partition tables

* Support reading of type RAID partitions, except from fs_of.
  This only works on x86 partition tables, Pmac partition tables will
  need alternate methods.  (Patch from Dustin Kirkland).
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-83

15 years agoRemove check for BootX broken device-trees
Ethan Benson [Mon, 4 Jul 2005 03:42:18 +0000 (03:42 +0000)]
Remove check for BootX broken device-trees

* ybin/ofpath: Remove check for broken-by-BootX device-tree, it no
  longer works under current 2.6 kernels and nobody attempts to use
  BootX anymore.
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-82

15 years agoMerge Dustin's RAID patch
Ethan Benson [Mon, 4 Jul 2005 03:34:00 +0000 (03:34 +0000)]
Merge Dustin's RAID patch

* ybin/ybin: Allow for multiple bootstrap partitions.  Patch by Dustin
  Kirkland.  May not fully work on PMAC hardware.
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-81

16 years agoamiga partition table support.
Ethan Benson [Fri, 20 May 2005 03:10:03 +0000 (03:10 +0000)]
amiga partition table support.

* Support for amiga partition tables.  (Sven Luther)
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-80

16 years agogcc-3.4 fix
Ethan Benson [Thu, 19 May 2005 04:26:37 +0000 (04:26 +0000)]
gcc-3.4 fix

* GCC 3.4 compilation fix.  (Paul Nasrat)
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-79

16 years agoUpdate addnote for newer pSeries hardware
Ethan Benson [Wed, 2 Feb 2005 05:26:31 +0000 (05:26 +0000)]
Update addnote for newer pSeries hardware

* util/addnote.c: Patch from Paul Mackerras to add second .note
  required by more recent pSeries hardware.

git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-78

16 years agoUpdate penguinppc.org links s/projects/bootloaders/g
Ethan Benson [Sun, 26 Sep 2004 08:10:31 +0000 (08:10 +0000)]
Update penguinppc.org links s/projects/bootloaders/g

* doc/yaboot-howto*.sgml: Update penguinppc.org links;
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-77

16 years agoFix typos in yaboot man page
Ethan Benson [Sat, 24 Jul 2004 20:07:25 +0000 (20:07 +0000)]
Fix typos in yaboot man page

man/yaboot.8: Fix various typos (thanks Helge Kreutzman).
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-76

16 years agoPrep 1.3.13 release yaboot-1.3.13
Ethan Benson [Sun, 11 Jul 2004 20:14:40 +0000 (20:14 +0000)]
Prep 1.3.13 release

* Bump version numbers for 1.3.13 release.
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-75

16 years agoUpdate changelog for 1.3.13
Ethan Benson [Sun, 11 Jul 2004 20:02:24 +0000 (20:02 +0000)]
Update changelog for 1.3.13

* changelog: Update for 1.3.13
git-archimport-id: erbenson@alaska.net--public/yaboot--devel--1.3--patch-74