remove the recently added -O0 for vax & gcc 14, the compiler bug is fixed.

thanks to Kalvis.

vax-gcc14 triggers an ICE here, same one as most new ones, ending in:

in extract_insn, at recog.cc:2818

bsd.own.mk: rename GCC_NO_* to CC_WNO_*

Rename compiler-warning-disable variables from
GCC_NO_warning
to
CC_WNO_warning
where warning is the full warning name as used by the compiler.

GCC_NO_IMPLICIT_FALLTHRU is CC_WNO_IMPLICIT_FALLTHROUGH

Using the convention CC_compilerflag, where compilerflag
is based on the full compiler flag name.

ufs: fixed signed/unsigned bugs affecting large file systems

Apply these commits from FreeBSD:

commit e870d1e6f97cc73308c11c40684b775bcfa906a2
Author: Kirk McKusick <mckusick@FreeBSD.org>
Date: Wed Feb 10 20:10:35 2010 +0000

This fix corrects a problem in the file system that treats large
inode numbers as negative rather than unsigned. For a default
(16K block) file system, this bug began to show up at a file system
size above about 16Tb.

To fully handle this problem, newfs must be updated to ensure that
it will never create a filesystem with more than 2^32 inodes. That
patch will be forthcoming soon.

Reported by: Scott Burns, John Kilburg, Bruce Evans
Followup by: Jeff Roberson
PR: 133980
MFC after: 2 weeks

commit 81479e688b0f643ffacd3f335b4b4bba460b769d
Author: Kirk McKusick <mckusick@FreeBSD.org>
Date: Thu Feb 11 18:14:53 2010 +0000

One last pass to get all the unsigned comparisons correct.


In additional to the changes from FreeBSD, this commit includes quite a few
related changes to appease -Wsign-compare.

branches: 1.39.2;
makefs(8): Needs -lm for ceil in udf as a host tool too.

udf.c uses math library functions.

introduce some common variables for use in GCC warning disables:

GCC_NO_FORMAT_TRUNCATION -Wno-format-truncation (GCC 7/8)
GCC_NO_STRINGOP_TRUNCATION -Wno-stringop-truncation (GCC 8)
GCC_NO_STRINGOP_OVERFLOW -Wno-stringop-overflow (GCC 8)
GCC_NO_CAST_FUNCTION_TYPE -Wno-cast-function-type (GCC 8)

use these to turn off warnings for most GCC-8 complaints. many
of these are false positives, most of the real bugs are already
commited, or are yet to come.


we plan to introduce versions of (some?) of these that use the
"-Wno-error=" form, which still displays the warnings but does
not make it an error, and all of the above will be re-considered
as either being "fix me" (warning still displayed) or "warning
is wrong."

branches: 1.36.26;
Implement `makefs -t udf'.

Formatting options may be enhanced to make it more in line with newfs_udf on
say labeling.

This works well enough to populate plain files in the root dir. creating
directories fails.

add debugging, fix warnings.

ok, we can create the filesystem now.

- add support for parsing different types; not just int.
- add beginning of msdos support.

branches: 1.31.2;
WARNS=5

added chfs support for makefs

branches: 1.29.2;
Re-enable v7fs support in the tools version of "makefs" after the
issues have been fixed by Alan Barrett. Tested under Mac OS X.

Define "USE_V7FS" in the makefile of the directory with the actual source.

Don't include v7fs support into tool version of "makefs" because it
breaks the build.

v7fs(7th edition(V7) file system) support.

Enable WARNS=4 by default, except for:
cpuctl dumplfs hprop ipf iprop-log kadmin kcm kdc kdigest
kimpersonate kstash ktutil makefs ndbootd ntp pppd quot
racoon racoonctl rtadvd sntp sup tcpdchk tcpdmatch tcpdump
traceroute traceroute6 user veriexecgen wsmoused zic
(Mostly third-party applications)

branches: 1.24.2;
Move fs-specific stuff in Makefile to fs/Makefile.inc to clarify
top-level Makefile.

Add support for encoding RISC OS metadata into ISO 9660 filesystems.
This is done by adding the general ability to stuff non-SUSP data into
the end of the System Use field of a Directory Record, which required
some amount of rewriting of the SUSP support. I think the result is
at least as good as what came before, and I've fixed at least one bug
along the way. Tested against RISC OS 3.70 on my Risc PC.

Oh, why do we want it? It should allow us to make acorn{26,32} CDs that
can be booted directly from RISC OS without mucking around copying the
bootloader to a native filing system.

Wrap compiler-specific flags with HAVE_GCC and HAVE_PCC as necessary. Add a few flags for PCC.

branches: 1.21.12;
Don't depend on lubutil if we're building host tools. Fix tools build on
non-NetBSD hosts.

Complete the move of string_to_flags() and flags_to_string() from the
bin/ls sources to libutil:
o Bump libutil minor version number
o Fix uses to include <util.h> to pick up the function definitions
o Fix most uses of flags_to_string() to release the now-malloc()ed result

don't do the -Wno-pointer-sign dance for host programs.

sprinkle some -fno-strict-aliasing and -Wno-pointer-sign with GCC4.

Remove extraneous #PKG_DEVELOPER=1 line that snuck in with the last
commit. (Thanks Masao Uebayashi.)

Cosmetic: alphabetize SRCS. Start a new line on each change of
leading letter.

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

Now that <bsd.prog.mk> DTRT if HOSTPROG is defined (i.e, it is a no-op),
there's no need to special-case .include-ing it.

tweaks for fparseln(3) move from libutil to libc:
- remove #include <util.h> if nothing else needed it
- remove LDFLAGS+=-lutil if nothing else needed it

oops; libutil is (currently) needed by ../mtree/spec.c for fparseln

- remove -lutil; nothing here needs it
- no need to reference ../../bin/dd now that strsuftoll(3) is in libc

convert from strsuftoull() (from ../../bin/dd) to strsuftoll(3) in libc

use NETBSDSRCDIR as appropriate

Roll in fixes to permit cross-compiling from non-NetBSD hosts. This
round has been tested on Solaris/x86 and Linux hosts.

* Add host tools cap_mkdb, ctags, m4, uudecode.
* Protect __RCSID() and __COPYRIGHT() better.
* Reduce the number of places that need to include "config.h", to keep
sources closer to their "vanilla" versions.
* Add more compat #defines and autoconf-checked functions.

Add '-N dbdir', to specify that user & group lookups (via pwcache(3)) are to
be done from dbdir/master.passwd and dbdir/group, using simple text file
parsing routines from ../mtree/getid.c

- pull in <ufs/ufs/dinode.h> before <ufs/ffs/fs.h>
- #include <ufs/[uf]fs/*.h> instead of "ufs/[uf]fs/*.h", and don't bother
with -I../../sys; we'll solve the hostprog problem another way and
unconditionally pulling in NetBSD-current's usr/src/sys on a host system
could be a Bad Thing.

build this as a host tool.

- Use strsuftoull() from dd(1) to parse numbers, rather than home grown
version. The new version supports products, 'g' (GB) and 't' (TB)
suffices, and `b' now means `blocks' instead of `bytes'.
Functionality requested by perry.
- debug is now a u_int instead of an int.
- Ensure that various numbers have sane upper limits (e.g, 99 for %, etc)

reference the ufs/[uf]fs headers in ../../sys rather than using those
in /usr/include. this should make it easier to make this a host tool

fix manual page building

branches: 1.1.1;
Initial revision

s/strucutre/structure/ in comment.

fix various typos in comments and makefs README file.

reflect reality about filesystems supported.

src is too big these days to tolerate superfluous apostrophes. It's
"its", people!

v7fs(7th edition(V7) file system) support.

cd9660 support has been present for a while

Refactor the filesystem specific portions out of makefs.c/makefs.h completely.
Instead of extending fsinfo_t it now holds a void * to file system specific
data. This is then setup/cleaned up by the additional of 2 additional
callbacks. Makes adding new filesystems simpler as almost no code has
to be updated in the generic makefs code now.

minor typo

branches: 1.1.1;
Initial revision

Add a todo item: read files from multiple directories.

branches: 1.6.24;
- clean up some debugging and verbose output
- use fs_cstotal.cs_nifree rather than fs->fs_ncg * fs->fs_ipg to
determine the number of free inodes available in the file system
- don't count '.' (the root node for a level) in inode sizing, since
the parent has already been counted
- some more todo items

-N dbdir allows reading user & group info from a different source

more stuff

typographical cleanups

a possible feature...

branches: 1.1.1;
Initial revision

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

Move date setting into cd9660_populate_iso_dir_record so there is no
path that leaves it unset.

From FreeBSD https://reviews.freebsd.org/D39258

s/filname/filename/

rock_ridge_move_count is only incremented and can never be negative so change
%08i to %08u.
This removes a warning when compiling with tools outside ./build.sh

s/initalize/initialize/ in comment or printf message.

branches: 1.55.12;
Change duplicate 'D' option to 'm'. From Ed Maste @ FreeBSD

makefs(8): add cd9660 eltorito-alt-boot option for EFI boot.

branches: 1.53.2;
PR/51652: Sevan Janiyan: makefs dies due to segmentation fault
Don't dereference NULL when running out of nodes during rename.

branches: 1.52.2;
little size_t/ssize_t...

Fix some more MKREPRO issues in cdrom creation. Now amd64 passes MKREPRO.

Provide a -T option to set timestamps to a consistent value for MKREPRO

more error normalization

improve error messages (remove \n, use __func__, etc)

Now that tools/compat/compat_defs.h defones __USE, there's no
need to treat the tools build as a special case.

avoid using __USE() in the tools version of this.

fix unused variable warnings.

`makefs -t cd9660' fix from FreeBSD:
Submitted by: Thomas Schmitt via marius@freebsd.org
Obtained from: FreeBSD r253707

- Correctly set the Expiration Time in the Primary Volume Descriptor;
according to ISO 9660 8.4.26.1 unspecified date and time are denoted by the
digit 0 in RBP 1 to 16 but the number 0 in RBP 17.

- Rock Ridge TF entries should use a length of 5, because after the 4 bytes of
generic SUSP header there is one byte of flags. See typedef of ISO_RRIP_TF
in iso9660_rrip.h.

macroize to make it more type-safe.

the allow options are 8 bits wide.

make everything use the generic options parser.

handle some of the options by the generic parser.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

- Fix single letter option parsing.
- Instead of returning 1 << index and 0 on error,
return just index and -1 on error for the set_option*() routines.

remove duplicated code, and try to cleanup parsing by using the shared code.
cd9660 needs a lot of work.

- add support for parsing different types; not just int.
- add beginning of msdos support.

branches: 1.35.6;
Patch from Jung-uk Kim (jkim at FreeBSD dot org) to allow contents of multiple
directories to be merged to the current image.

branches: 1.34.2;
more fixes from martin

remove stray brace, minor knf

PR/45285: Martin Matuska: makefs does not properly convert ISO level 1 and 2
filenames (buffer overflow)

makefs does not properly verify the maximum filename length in the
special "." case for both ISO level 1 and ISO level 2 filename
conversion. This creates broken images or causes a buffer overflow
(ISO level 2).

ISO level 1:
If a filename contains only dots or up to 8 characters followed by
dots the 8+3 limit check doesn't work.

ISO level 2:
If a filename contains a dot in the first 30 characters and a dot
on the 30th character, the length limit check doesn't work and the
buffer is overflowed.

$ mkdir level1
$ touch level1/12345............
$ makefs -t cd9660 -o isolevel=1 test.iso level1

$ mkdir level2
$ touch level2/1234567890.2345678901234567.....34567890123456789012345
$ makefs -t cd9660 -o isolevel=2 test.iso level2

PR/45217: Martin Matuska: Use proper lengths to space pad fields and don't
overwrite them according to the spec.

Apply fixes for PR bin/44114 (makefs(8) -t cd9660 -o rockridge creates
corrupted cd9660fs):
- makefs/cd9660.c:
- cd9660_rrip_move_directory() should also update lenght[0] and
name_len[0] in dir->isoDirRecord
- makefs/cd9660/iso9660_rrip.c:
- cd9660_rrip_finalize_node() should check rr_real_parent in node->parent,
not in node itself in RRIP_PL case
- cd9660_rrip_initialize_node() should update only node passed as arg
so handle RRIP_PL in DOTDOT case

Fixes malformed dotdot entries in deep (more than 8 level) directories
moved into .rr_moved dir.

Should be pulled up to netbsd-5.
(no official ISO has such deep dirs, but cobalt restorecd is affected)

From Nathan Whitehorn (nwhitehorn at freebsd dot org):
Add code to generate bootable ISOs on Powermac and CHRP systems.
Synthesize some partition maps (APM and MBR, respectively) pointing
to (a) the whole disk, and (b) relevant El Torito boot images that
have been added by other code. These partition maps are a little
bit funny looking, but they seem to work. FreeBSD has been using
this successfully in their release generation on powerpc, as well
as generating all non-SPARC install media. SPARC support could
probably be added as an extension of this patch.

cd9960 -> cd9660

Handle > 2GB images. Thanks to Greg Troxel for testing!

remove fsopts asserts - the parameter is unused, so we don't care here

Add support for encoding RISC OS metadata into ISO 9660 filesystems.
This is done by adding the general ability to stuff non-SUSP data into
the end of the System Use field of a Directory Record, which required
some amount of rewriting of the SUSP support. I think the result is
at least as good as what came before, and I've fixed at least one bug
along the way. Tested against RISC OS 3.70 on my Risc PC.

Oh, why do we want it? It should allow us to make acorn{26,32} CDs that
can be booted directly from RISC OS without mucking around copying the
bootloader to a native filing system.

If the user specifies a cd9660-specific option that needs an integer value
without supplying a value, complain at them instead of segfaulting.
Or, to put it another way, check that the value is non-NULL before calling
set_option().

Make the allow-multidot option useful.

branches: 1.22.2;
Import changes from the MirBSD Project:

Initialise memory before writing to only part of it,
making sure that the string is NUL-terminated

Allocate the maximum allowed length (including the
terminating NUL byte) for the temporary name string, make sure
the final output is NUL terminated, and copy back the NUL when
overwriting the previously created name (since maxlen is always
at least 1 less than ISO_FILENAME_MAXLENGTH_WITH_PADDING, this
is no problem)

Patches from Thorsten Glaser through freenode.net.

Fix typo in cd9660 functionname cd9660_setup_boot_volume_descritpor ->
cd9660_setup_boot_volume_descriptor.

Interestingly the same typo was used in 3 files!

Add an option to load a generic boot image into the first 32K of the
cd9660 image.

branches: 1.19.10; 1.19.12;
Fix deep directory move rock-ridge extension. Code was there, but it had
bit-rotted.

* Don't include headers twice
* Remove a few trailing whitespaces
* Rearrange and join to one #if for some headers

Patch contributed by Slava Semushin <slava.semushin@gmail.com>
in private mail.

from Anon Ymous: compile with -Wextra

branches: 1.16.2; 1.16.4;
use c99 initializers

implement -x by pruning fsnodes tree before building filesystem rather
than by skipping nodes while building filesystem

make -t cd9660 obey -x option similarly to ffs

Fix fix for Coverity CID 2751:

Move an error check to the right place, and instead of blindly insisting on
an argument for all of "no-emul-boot", "no-boot", "hard-disk-boot" and
"boot-load-segment", only require it for the last one, which is the only
one that requires/uses it.

OK'd by Christos

Coverity CID 2751: Avoid NULL deref.

Fix a bug in cd9660_sort_nodes, which tried to insert `next' before
`node' on a tail-queue, when `node' wasn't on the tail-queue.
Sometimes this led to a segfault. Insert before `cn', since that
is the intention. Closes PR 32194. Thanks, Jeff Rizzo, for reporting
this bug.

Change cd9660's default verbosity level to 0, quieting it down
considerably.

Remove the cd9660_inode hashtable for hard-links support, and re-use
the fsinode for the same purpose, per Luke Mewburn's suggestion on
source-changes@.

Add hard-link support: keep a hashtable that maps inodes for
regular files on the source filesystem to the offset of the first
data block on the target filesystem. Don't allocate blocks for a
file a second time (i.e., if it is already present in the hashtable).
Avoid writing the blocks of a file twice (use the FS_WRITTEN flag
in fsinode->flags for this purpose).

Bug fix: record the actual number number of 512-byte "virtual"
sectors in the Sector Count field of the Initial/Default Entry in
the "El Torito" Boot Catalog, instead of blithely writing 1 (!!)
into the Sector Count field. Ditto the Section Entries.

Remove last remnants of cd9660's unsupported boot-load-size option.

Really use cd9660's boot-load-segment argument.

Change the boot images list from a LIST to a TAILQ.

Re-implement CD9660_BLOCKS() with howmany().

Introduce ELTORITO_DPRINTF and some debug messages.

Zero memory malloc'd in the El Torito code.

Revise some debugging outputs.

Insert a new boot-image on the boot-images list after the last
boot-image for the same architecture. It makes a (trivial) functional
difference.

Miscellaneous cosmetic changes. Remove useless variable-assignment.

Remove invisible whitespace at ends of lines that disrupted my
speedy cursor-movement habits in vi.

Overhaul makefs -t cd9660.

Functional: Do not add a missing period or a version number to
directories, since that doesn't fit the standard (thanks
Daniel Watt).

Add RRIP PX records (POSIX meta-information) to '.'
and '..' entries. Quiets kernel complaint, when we
mount a cd9660 filesystem, that there is a missing PX
entry.

Cosmetic: Reduce height of staircases. Remove an unused member
from struct _cd9660node. Extract a subroutine. KNF.

Use a TAILQ macros for lists of both SUSP/RRIP attributes
and directory entries.

XXX
XXX The new RRIP PX entries are not quite right.
XXX

Cosmetic: change to a standard idiom, *(x++) --> *x++. Flag some
places needing bounds-checking with XXX. Fix indendation.

By default, do not omit the trailing period from ISO9660 filenames
unless given the option '-o omit-trailing-period'. Also, fix a
logic error so that makefs obeys option '-o omit-trailing-period'
when it is given.

Move local variable declaration to start of block (also start of function).
Fixes build problem for vax.

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

little size_t/ssize_t...

make everything use the generic options parser.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

branches: 1.18.6;
Patch from Jung-uk Kim (jkim at FreeBSD dot org) to allow contents of multiple
directories to be merged to the current image.

branches: 1.17.2;
Move inclusion of sys/endian.h from cd9660_eltorito.c to cd9660.h
since actual user is not cd9660_eltorito.c but iso.h and
cd9660_eltorito.h.

Actually, include order/place of sys/endian.h doesn't matter on
netbsd since it is always included by sys/types.h but it's not
true on other system. This should fix cross build breakage on
freebsd introduced by rev. 1.16 of cd9660_eltorito.c.
Problem reported and fix suggested on twitter.

From Nathan Whitehorn (nwhitehorn at freebsd dot org):
Add code to generate bootable ISOs on Powermac and CHRP systems.
Synthesize some partition maps (APM and MBR, respectively) pointing
to (a) the whole disk, and (b) relevant El Torito boot images that
have been added by other code. These partition maps are a little
bit funny looking, but they seem to work. FreeBSD has been using
this successfully in their release generation on powerpc, as well
as generating all non-SPARC install media. SPARC support could
probably be added as an extension of this patch.

Handle > 2GB images. Thanks to Greg Troxel for testing!

- simple fseek->fseeko conversion. probably needs more work.
- use a constant instead of sprinkling 2048 everywhere.

Add support for encoding RISC OS metadata into ISO 9660 filesystems.
This is done by adding the general ability to stuff non-SUSP data into
the end of the System Use field of a Directory Record, which required
some amount of rewriting of the SUSP support. I think the result is
at least as good as what came before, and I've fixed at least one bug
along the way. Tested against RISC OS 3.70 on my Risc PC.

Oh, why do we want it? It should allow us to make acorn{26,32} CDs that
can be booted directly from RISC OS without mucking around copying the
bootloader to a native filing system.

branches: 1.12.4; 1.12.6; 1.12.8;
Fix typo in cd9660 functionname cd9660_setup_boot_volume_descritpor ->
cd9660_setup_boot_volume_descriptor.

Interestingly the same typo was used in 3 files!

Add an option to load a generic boot image into the first 32K of the
cd9660 image.

branches: 1.10.18; 1.10.20;
Stop duplicating the ISO9660 sources in sys/fs/cd9660/ for makefs(8).

Remove the cd9660_inode hashtable for hard-links support, and re-use
the fsinode for the same purpose, per Luke Mewburn's suggestion on
source-changes@.

Add hard-link support: keep a hashtable that maps inodes for
regular files on the source filesystem to the offset of the first
data block on the target filesystem. Don't allocate blocks for a
file a second time (i.e., if it is already present in the hashtable).
Avoid writing the blocks of a file twice (use the FS_WRITTEN flag
in fsinode->flags for this purpose).

Now that I understand what order boot-images and boot-entries are
intended to be kept on the boot-images and -entries lists, keep
them in that order: images grouped and entries grouped by system,
with later images nearer the front of the list.

*Really* apply boot options to the last boot-image specified.

Bug fix: record the actual number number of 512-byte "virtual"
sectors in the Sector Count field of the Initial/Default Entry in
the "El Torito" Boot Catalog, instead of blithely writing 1 (!!)
into the Sector Count field. Ditto the Section Entries.

Remove last remnants of cd9660's unsupported boot-load-size option.

Really use cd9660's boot-load-segment argument.

Change the boot images list from a LIST to a TAILQ.

Re-implement CD9660_BLOCKS() with howmany().

Introduce ELTORITO_DPRINTF and some debug messages.

Zero memory malloc'd in the El Torito code.

Revise some debugging outputs.

Insert a new boot-image on the boot-images list after the last
boot-image for the same architecture. It makes a (trivial) functional
difference.

Miscellaneous cosmetic changes. Remove useless variable-assignment.

Remove invisible whitespace at ends of lines that disrupted my
speedy cursor-movement habits in vi.

Overhaul makefs -t cd9660.

Functional: Do not add a missing period or a version number to
directories, since that doesn't fit the standard (thanks
Daniel Watt).

Add RRIP PX records (POSIX meta-information) to '.'
and '..' entries. Quiets kernel complaint, when we
mount a cd9660 filesystem, that there is a missing PX
entry.

Cosmetic: Reduce height of staircases. Remove an unused member
from struct _cd9660node. Extract a subroutine. KNF.

Use a TAILQ macros for lists of both SUSP/RRIP attributes
and directory entries.

XXX
XXX The new RRIP PX entries are not quite right.
XXX

Fix some wrapping lines.

Add nbtool_config.h include, fixing the tool compile on some platforms.

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

make everything use the generic options parser.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

- Fix single letter option parsing.
- Instead of returning 1 << index and 0 on error,
return just index and -1 on error for the set_option*() routines.

remove duplicated code, and try to cleanup parsing by using the shared code.
cd9660 needs a lot of work.

- add support for parsing different types; not just int.
- add beginning of msdos support.

branches: 1.3.2; 1.3.4;
WARNS=5

- fix compilation isses
- add some error checking
- avoid alloca
- add missing frees

added chfs support for makefs

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

branches: 1.1.2; 1.1.4;
added chfs support for makefs

From FreeBSD D50078 by jrtc27:

Linux's struct stat does not have an st_flags member, but we still want
to be able to create FreeBSD disk images with non-zero st_flags. Add a
replacement member to struct fsinode to store it on Linux.

branches: 1.76.2;
makefs(8): Fix format string for newly 64-bit target inode numbers.

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

ufs: fixed signed/unsigned bugs affecting large file systems

Apply these commits from FreeBSD:

commit e870d1e6f97cc73308c11c40684b775bcfa906a2
Author: Kirk McKusick <mckusick@FreeBSD.org>
Date: Wed Feb 10 20:10:35 2010 +0000

This fix corrects a problem in the file system that treats large
inode numbers as negative rather than unsigned. For a default
(16K block) file system, this bug began to show up at a file system
size above about 16Tb.

To fully handle this problem, newfs must be updated to ensure that
it will never create a filesystem with more than 2^32 inodes. That
patch will be forthcoming soon.

Reported by: Scott Burns, John Kilburg, Bruce Evans
Followup by: Jeff Roberson
PR: 133980
MFC after: 2 weeks

commit 81479e688b0f643ffacd3f335b4b4bba460b769d
Author: Kirk McKusick <mckusick@FreeBSD.org>
Date: Thu Feb 11 18:14:53 2010 +0000

One last pass to get all the unsigned comparisons correct.


In additional to the changes from FreeBSD, this commit includes quite a few
related changes to appease -Wsign-compare.

branches: 1.73.2;
Restore backward compatibility of UFS2 with previous NetBSD releases by
disabling support in UFS2 for extended attributes (including ACLs).
Add a new variant of UFS2 called "UFS2ea" that does support extended attributes.
Add new fsck_ffs operations "-c ea" and "-c no-ea" to convert file systems
from UFS2 to UFS2ea and vice-versa (both of which delete all existing extended
attributes in the process).

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

makefs: fix some typos

PR/52828: Mark Johnston: makefs UFS2 lazy inode initialization is buggy

makefs(8) emulates UFS2 in performing lazy initialization of inode
blocks when allocating and writing inodes. However, it only ever
initializes one inode block at a time, which may be insufficient.
If so, a later initialization may clobber an inode, resulting in
an inconsistent filesystem.

I committed a minimal fix for the problem to FreeBSD:
https://svnweb.freebsd.org/changeset/base/326912

branches: 1.69.4;
stampst.st_ion consistent treatment (non-zero vs zero instead of testing == 1)

no, this is wrong.

zero out the superblock so that it does not contain random stuff in the
spare fields.

branches: 1.66.2; 1.66.4;
more deterministic ffs for reproducible builds.

Add timestamp support to the ffs image creation; needed for in kernel memory
images MKREPRO.

PR/49559: Christian Brueffer: Fix typo maxbpf instead of maxbpg.

branches: 1.63.6;
fsbtodb() -> FFS_FSBTODB(), EXT2_FSBTODB(), or MFS_FSBTODB()
dbtofsb() -> FFS_DBTOFSB() or EXT2_DBTOFSB()

(Christos already did the lfs ones a few days back)

Rename ambiguous macros:
MAXDIRSIZE -> UFS_MAXDIRSIZE or LFS_MAXDIRSIZE
NINDIR -> FFS_NINDIR, EXT2_NINDIR, LFS_NINDIR, or MFS_NINDIR
INOPB -> FFS_INOPB, LFS_INOPB
INOPF -> FFS_INOPF, LFS_INOPF
blksize -> ffs_blksize, ext2_blksize, or lfs_blksize
sblksize -> ffs_blksize

These are not the only ambiguously defined filesystem macros, of
course, there's a pile more. I may not have found all the ambiguous
definitions of blksize(), too, as there are a lot of other things
called 'blksize' in the system.

Stick UFS_ in front of these symbols:
DIRBLKSIZ
DIRECTSIZ
DIRSIZ
OLDDIRFMT
NEWDIRFMT

Part of PR 47909.

- more changes to make -O work
- fix err* calls.

- don't abuse vp->fs to mean struct fs for ffs and struct msdos_opts;
make it always fsinfo_t and change void * to that.
- kill unused structure members.

gc global sectorsize.

make everything use the generic options parser.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

This works well enough to populate plain files in the root dir. creating
directories fails.

make the buffer functions look exactly like the kernel ones and add other
cruft to make the kernel files compile.

- Fix single letter option parsing.
- Instead of returning 1 << index and 0 on error,
return just index and -1 on error for the set_option*() routines.

return the bit of the option set, so that others can act on it.

remove duplicated code, and try to cleanup parsing by using the shared code.
cd9660 needs a lot of work.

- add support for parsing different types; not just int.
- add beginning of msdos support.

Stuff UFS_ in front of a few of ufs's symbols to reduce namespace
pollution. Specifically:
ROOTINO -> UFS_ROOTINO
WINO -> UFS_WINO
NXADDR -> UFS_NXADDR
NDADDR -> UFS_NDADDR
NIADDR -> UFS_NIADDR
MAXSYMLINKLEN -> UFS_MAXSYMLINKLEN
MAXSYMLINKLEN_UFS[12] -> UFS[12]_MAXSYMLINKLEN (for consistency)

Sort out ext2fs's misuse of NDADDR and NIADDR; fortunately, these have
the same values in ext2fs and ffs.

No functional change intended.

branches: 1.48.2;
Add -Z to enable sparse files - for ffs at least.
Based on a patch from shesha at juniper.

WARNS=5

Patch from Jung-uk Kim (jkim at FreeBSD dot org) to allow contents of multiple
directories to be merged to the current image.

branches: 1.45.2;
add support for setting the ufs label. (Nathan Whitehorn)

Create images with 0666 - umask, not 0777 - umask (e.g. without x bits).

Distinguish read(2) failure due to a short read from other causes, and
suggest in the error message a possible cause: the size= attribute in
the metalog (if one is given) may be different from the source file's
actual size.

branches: 1.42.20;
from Anon Ymous: compile with -Wextra

use c99 initializers

implement -x by pruning fsnodes tree before building filesystem rather
than by skipping nodes while building filesystem

Coverity CID 174: Add extra NULL test to appease coverity.

free buf on error path
coverity 190 / run 5

Use the real field name (d_fileno instead of d_ino).

fix compilatio problems on LP64

Fix a sign warning for WARNS=3.

appease gcc -Wuninitialized

Refactor the filesystem specific portions out of makefs.c/makefs.h completely.
Instead of extending fsinfo_t it now holds a void * to file system specific
data. This is then setup/cleaned up by the additional of 2 additional
callbacks. Makes adding new filesystems simpler as almost no code has
to be updated in the generic makefs code now.

Check for fstatvfs and provide/use it only if the host system has it.
Fixes PR#27221

include <sys/statvfs.h> if HAVE_STRUCT_STATVFS_F_IOSIZE
addresses part of PR toolchain/26415

Define FFS_MAXBSIZE to 65536 and use instead of MAXBSIZE.
Deprecate unused MAXINOPB.
Should fix PR [toolchain/25603].

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

Allow for ROOTINO when deciding whether their are enough inodes in the
created filesystem. Otherwise this fails when asked for 63 inodes for
i386 ramdisk-big.
This code really needs to use the current version of newfs/mkfs.c where
the actual inode count can be passed in (instead of the density).
Fixes a build problem when I add /mountroot to mtree.common for sysinst.

Replace the statfs() family of system calls with statvfs().
Retain binary compatibility.

After populating the file system, update superblock->fs_old_cstotal from
superblock->fs_cstotal. Fixes inconsistencies found by FreeBSD's fsck.
Problem noted by Luigi Rizzo <luigi@FreeBSD.org> via Colin Percival.

branches: 1.25.2;
Mention that we should be adding space for bitmaps.

Remove a strange multiplication which appears to be incorrect.

realloc pedant

Revert arc4random usage. makefs is only used for install images,
filehandle randomization is not an issue there, and using arc4random
broke building makefs as a host tool.

use arc4random

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22366, verified by myself.

Add support for UFS2. UFS2 is an enhanced FFS, adding support for
64 bit block pointers, extended attribute storage, and a few
other things.

This commit does not yet include the code to manipulate the extended
storage (for e.g. ACLs), this will be done later.

Originally written by Kirk McKusick and Network Associates Laboratories for
FreeBSD.

Tweak Jason's -x support to pass its state around as part of fsinf_t,
just like all the other "global" options.
Update the usage for -x.
Crank date on man page.

(Thanks to Jason for adding -x; I've been meaning to add that
functionality for a while :-)

Add a -x option which causes makefs(8) to exclude any file system node
not explicitly listed in the specfile.

Bump daddr_t to 64 bits. Replace it with int32_t in all places where
it was used on-disk, so that on-disk formats remain the same.
Remove ufs_daddr_t and ufs_lbn_t for the time being.

d_namlen is a uint8_t, and thus will always be < MAXMANLEN + 1 (256),
so there is no need to assert it.

branches: 1.14.2;
- Clear fs_fmod in superblock before final write. Noted by Erik Anggard.
- Explicitly close the image file descriptor at the end of ffs_makefs().

- clean up some debugging and verbose output
- use fs_cstotal.cs_nifree rather than fs->fs_ncg * fs->fs_ipg to
determine the number of free inodes available in the file system
- don't count '.' (the root node for a level) in inode sizing, since
the parent has already been counted
- some more todo items

Roll in fixes to permit cross-compiling from non-NetBSD hosts. This
round has been tested on Solaris/x86 and Linux hosts.

* Add host tools cap_mkdb, ctags, m4, uudecode.
* Protect __RCSID() and __COPYRIGHT() better.
* Reduce the number of places that need to include "config.h", to keep
sources closer to their "vanilla" versions.
* Add more compat #defines and autoconf-checked functions.

be a little bit more verbose

- crank default cpg from 16 to 65536, and let the auto-sizing code in
mkfs() do its job
- pass cpgflg=1 to mkfs() if the user provided cpg, =0 otherwise
- improve error reporting in ffs_write_file()
- only add superblock & block map slop for one cylinder group.
trying to calculate the number of required cylinder groups caused
too many problems when tweaking parameters for small file systems.
rely upon slop from minfree, and upon ffs_write_file() to complain
if the user is too aggressive about tweaking the parameters

use -1 instead of 0 as the value to indicate if a numeric parameter
hasn't been set and thus needs a default override.
allows minfree and rotdelay to take a value of 0.

add "ffs/ufs_inode.h" with a minimal struct inode, to replace
<ufs/ufs/inode.h>, since the latter has a lot of cruft we don't need
and it #include's way more stuff in <sys/*> (etc) than is needed here.
yet another nail in the "let's make makefs a proper host tool" coffin.

- pull in <ufs/ufs/dinode.h> before <ufs/ffs/fs.h>
- #include <ufs/[uf]fs/*.h> instead of "ufs/[uf]fs/*.h", and don't bother
with -I../../sys; we'll solve the hostprog problem another way and
unconditionally pulling in NetBSD-current's usr/src/sys on a host system
could be a Bad Thing.

fix bugs reported by matt green:
- when calculating file size, round up to frag size not block size
- #define FFS_EI in makefs.h, and include that before [uf]fs/*.h

other stuff:
- round up final size to next block (instead of next sector)
- protect makefs.h from multiple inclusion

rework `fsnode' into two components; `fsnode' contains the name and tree
information and a link to the appropriate fsinode, and `fsinode' contains
the inode information and a reference count. multiple fsnodes may point
to the same fsinode. this replaces the 'dup' pointer in the previous code.

reference the ufs/[uf]fs headers in ../../sys rather than using those
in /usr/include. this should make it easier to make this a host tool

tweak copyright

add __RCSID()

branches: 1.1.1;
Initial revision

Restore backward compatibility of UFS2 with previous NetBSD releases by
disabling support in UFS2 for extended attributes (including ACLs).
Add a new variant of UFS2 called "UFS2ea" that does support extended attributes.
Add new fsck_ffs operations "-c ea" and "-c no-ea" to convert file systems
from UFS2 to UFS2ea and vice-versa (both of which delete all existing extended
attributes in the process).

add support for setting the ufs label. (Nathan Whitehorn)

Refactor the filesystem specific portions out of makefs.c/makefs.h completely.
Instead of extending fsinfo_t it now holds a void * to file system specific
data. This is then setup/cleaned up by the additional of 2 additional
callbacks. Makes adding new filesystems simpler as almost no code has
to be updated in the generic makefs code now.

Document debugging options and allow specifying them by name.

branches: 1.72.2;
Add an option to warn and error out for mtree missing entries.

Restore backward compatibility of UFS2 with previous NetBSD releases by
disabling support in UFS2 for extended attributes (including ACLs).
Add a new variant of UFS2 called "UFS2ea" that does support extended attributes.
Add new fsck_ffs operations "-c ea" and "-c no-ea" to convert file systems
from UFS2 to UFS2ea and vice-versa (both of which delete all existing extended
attributes in the process).

new sentence, new line

Refactor and rewrite of newfs_udf(8) and makefs(8) (-t udf) with a shared
section for fsck_udf(8).

Newfs_udf and makefs support predefined disc image profiles, harddisc
partitions (disklabel and wedges on all generic block devices) and all optical
media types on CD/DVD/BD writers.

Newfs_udf and makefs now also support formatting of UDF 2.50 with a metadata
partition.

Explain cd9660 and udf exceptions on the rule that the default sectorsize is
512.

Add a -L option to follow all symbolic links. Useful if you have symlinks
in a makefs directory tree but want to refer to the actual file.

Add "efi" as a supported boot image type and derive the platform ID for
the validation entry from the default boot image instead of hard-coding
X86.

Indent the makefs(8) options for UDF like the other filesystems described.

Correct documentation of -m and -M.
From Ed Maste @ FreeBSD

branches: 1.63.12;
Whitespace fixes.

Explain the 's' image size better.

file system police.

untorture language.

mention timezone change for -T timestamp

makefs(8): add cd9660 eltorito-alt-boot option for EFI boot.

branches: 1.57.2;
Document msdos-specific options directly.

From jmc@OpenBSD.

Bump date.

Some changes and sorting based on diff sent from jmc@OpenBSD.

branches: 1.55.2;
"file system" as two words.

Provide a -T option to set timestamps to a consistent value for MKREPRO

New sentence, new line.
Use one list instead of two consecutive ones.
Add .El to end list.

Add minimum UDF version specification to makefs(8) -t udf.

Allow for the logical volume label to be specified as well as the physical
volume label. Also allow the volumeset name to be specified if desired. The
syntax follows the newfs_udf(8) syntax.

New sentence, new line.

Update makefs(8) man page to reflect the selection of default sector and disc
sizes based on the disc type.

Note that i added the UDF support to makefs(8) and bump data

Sort. Improve table formatting.

Implement `makefs -t udf'.

Formatting options may be enhanced to make it more in line with newfs_udf on
say labeling.

add a replace flag so we can overlay exiting files when we merge directories.

add and document offset.

mention me

now that msdos works, document it.

branches: 1.41.2;
Add -Z to enable sparse files - for ffs at least.
Based on a patch from shesha at juniper.

New sentence, new line.

added chfs support for makefs

Patch from Jung-uk Kim (jkim at FreeBSD dot org) to allow contents of multiple
directories to be merged to the current image.

branches: 1.37.2;
add support for setting the ufs label. (Nathan Whitehorn)

Re-order `usage' alphabetically;
rename option arguments in the manpage's `SYNOPSIS' section to
match those from `usage' (not the other way around; the `usage'-line
(and other parts of makefs.c) contain the correct names);
minor punctuation improvements.

From Snader_LB.

Bump date for previous, fix typo.

v7fs(7th edition(V7) file system) support.

From Nathan Whitehorn (nwhitehorn at freebsd dot org):
Add code to generate bootable ISOs on Powermac and CHRP systems.
Synthesize some partition maps (APM and MBR, respectively) pointing
to (a) the whole disk, and (b) relevant El Torito boot images that
have been added by other code. These partition maps are a little
bit funny looking, but they seem to work. FreeBSD has been using
this successfully in their release generation on powerpc, as well
as generating all non-SPARC install media. SPARC support could
probably be added as an extension of this patch.

Bump date for last change.

Add support for encoding RISC OS metadata into ISO 9660 filesystems.
This is done by adding the general ability to stuff non-SUSP data into
the end of the System Use field of a Directory Record, which required
some amount of rewriting of the SUSP support. I think the result is
at least as good as what came before, and I've fixed at least one bug
along the way. Tested against RISC OS 3.70 on my Risc PC.

Oh, why do we want it? It should allow us to make acorn{26,32} CDs that
can be booted directly from RISC OS without mucking around copying the
bootloader to a native filing system.

Add an option to load a generic boot image into the first 32K of the
cd9660 image.

branches: 1.29.6; 1.29.8;
Fix Xref.

branches: 1.28.4;
Use IEC 60027-2 prefixes for 2^n based prefixes.

Use periods in tables consistently.

New sentence new line.

bump date

fix typo: explcitly -> explicitly

Sort SEE ALSO.

Add reference to installboot(8), as suggested by Patrick Welche.

Bump date for previous.

boot-load-size, the number of load sectors for the boot image, is
not yet, and will not be, supported.

The system type mac68k is also allowed.

XXX Is that correct?

Synchronize documentation with utility.

Remove unnecessary keeps (Bk/Ek). Sort options. Various minor fixes.
Bump date.

Comment out some currently unimplemented options, and add a few descriptions
for others (from Daniel Watt).

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

Fix a typo in the description of -f. Noted by Jukka Salmi in PR 30713.
Add some missing punctuation.

Spell the plural of suffix "suffixes", not "suffices".
Inspired by PR 24400 by Todd Vierling.

Add support for UFS2. UFS2 is an enhanced FFS, adding support for
64 bit block pointers, extended attribute storage, and a few
other things.

This commit does not yet include the code to manipulate the extended
storage (for e.g. ACLs), this will be done later.

Originally written by Kirk McKusick and Network Associates Laboratories for
FreeBSD.

Tweak Jason's -x support to pass its state around as part of fsinf_t,
just like all the other "global" options.
Update the usage for -x.
Crank date on man page.

(Thanks to Jason for adding -x; I've been meaning to add that
functionality for a while :-)

Add a -x option which causes makefs(8) to exclude any file system node
not explicitly listed in the specfile.

support 1234 and 4321 as valid byte-order arguments for -B

.Nm does not need a dummy argument ("") before punctuation or
for correct formatting of the SYNOPSIS any longer.

The correct capitalisation of 'NetBSD.org' is (you
guessed it) 'NetBSD.org'.

some mdoc fixes.

If an entry is in the specfile but not in the underlying file system, and it's
marked "optional", don't add it. (e.g, don't create a zero-length file).

Add '-N dbdir', to specify that user & group lookups (via pwcache(3)) are to
be done from dbdir/master.passwd and dbdir/group, using simple text file
parsing routines from ../mtree/getid.c

Drop an unnecessary .Pp, fix a typo, use standard headers, sort sections,
mark up author name.

- Use strsuftoull() from dd(1) to parse numbers, rather than home grown
version. The new version supports products, 'g' (GB) and 't' (TB)
suffices, and `b' now means `blocks' instead of `bytes'.
Functionality requested by perry.
- debug is now a u_int instead of an int.
- Ensure that various numbers have sane upper limits (e.g, 99 for %, etc)

tweak copyright

branches: 1.1.1;
Initial revision

Document debugging options and allow specifying them by name.

branches: 1.58.2;
Add an option to warn and error out for mtree missing entries.

Add an options pretty-printer, pass some lint.

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

Add a -L option to follow all symbolic links. Useful if you have symlinks
in a makefs directory tree but want to refer to the actual file.

Don't pass garbage to parsedate, but do use the return value.

Sort options in usage to match man page order.

Provide a -T option to set timestamps to a consistent value for MKREPRO

Implement `makefs -t udf'.

Formatting options may be enhanced to make it more in line with newfs_udf on
say labeling.

add a replace flag so we can overlay exiting files when we merge directories.

add and document offset.

remove bogus err.

while it is studly to play with token pasting, passing directly the type
is a lot more readable.

use "val" not "var"

make everything use the generic options parser.

Allow options without values meaning the empty string for string options
and = 1 to numerics.

deal with options that don't have a single letter argument.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

fix single letter parsing.

- Fix single letter option parsing.
- Instead of returning 1 << index and 0 on error,
return just index and -1 on error for the set_option*() routines.

return the bit of the option set, so that others can act on it.

remove duplicated code, and try to cleanup parsing by using the shared code.
cd9660 needs a lot of work.

- add support for parsing different types; not just int.
- add beginning of msdos support.

branches: 1.35.2;
Add -Z to enable sparse files - for ffs at least.
Based on a patch from shesha at juniper.

usage is dead. name_len should be read as unsigned byte.

- fix compilation isses
- add some error checking
- avoid alloca
- add missing frees

added chfs support for makefs

Patch from Jung-uk Kim (jkim at FreeBSD dot org) to allow contents of multiple
directories to be merged to the current image.

branches: 1.30.2;
Re-order `usage' alphabetically;
rename option arguments in the manpage's `SYNOPSIS' section to
match those from `usage' (not the other way around; the `usage'-line
(and other parts of makefs.c) contain the correct names);
minor punctuation improvements.

From Snader_LB.

Re-enable v7fs support in the tools version of "makefs" after the
issues have been fixed by Alan Barrett. Tested under Mac OS X.

Don't include v7fs support into tool version of "makefs" because it
breaks the build.

v7fs(7th edition(V7) file system) support.

use c99 initializers

implement -x by pruning fsnodes tree before building filesystem rather
than by skipping nodes while building filesystem

add implementation of free_fsnodes() which unlinks a node
from the hierarchy and frees it and any of its children
use it to free memory before a normal exit, which helps diagnostics

use strtoll instead of strsuftoll to parse -d so it can be specified
in hex with a 0x prefix

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

Refactor the filesystem specific portions out of makefs.c/makefs.h completely.
Instead of extending fsinfo_t it now holds a void * to file system specific
data. This is then setup/cleaned up by the additional of 2 additional
callbacks. Makes adding new filesystems simpler as almost no code has
to be updated in the generic makefs code now.

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

branches: 1.19.2;
Homogenize usage messages: make the 'usage' word all lowercase, as this seems
to be the most common practice in our tree.

Add support for UFS2. UFS2 is an enhanced FFS, adding support for
64 bit block pointers, extended attribute storage, and a few
other things.

This commit does not yet include the code to manipulate the extended
storage (for e.g. ACLs), this will be done later.

Originally written by Kirk McKusick and Network Associates Laboratories for
FreeBSD.

Tweak Jason's -x support to pass its state around as part of fsinf_t,
just like all the other "global" options.
Update the usage for -x.
Crank date on man page.

(Thanks to Jason for adding -x; I've been meaning to add that
functionality for a while :-)

Add a -x option which causes makefs(8) to exclude any file system node
not explicitly listed in the specfile.

support 1234 and 4321 as valid byte-order arguments for -B

convert from strsuftoull() (from ../../bin/dd) to strsuftoll(3) in libc

Roll in fixes to permit cross-compiling from non-NetBSD hosts. This
round has been tested on Solaris/x86 and Linux hosts.

* Add host tools cap_mkdb, ctags, m4, uudecode.
* Protect __RCSID() and __COPYRIGHT() better.
* Reduce the number of places that need to include "config.h", to keep
sources closer to their "vanilla" versions.
* Add more compat #defines and autoconf-checked functions.

- crank default cpg from 16 to 65536, and let the auto-sizing code in
mkfs() do its job
- pass cpgflg=1 to mkfs() if the user provided cpg, =0 otherwise
- improve error reporting in ffs_write_file()
- only add superblock & block map slop for one cylinder group.
trying to calculate the number of required cylinder groups caused
too many problems when tweaking parameters for small file systems.
rely upon slop from minfree, and upon ffs_write_file() to complain
if the user is too aggressive about tweaking the parameters

Add '-N dbdir', to specify that user & group lookups (via pwcache(3)) are to
be done from dbdir/master.passwd and dbdir/group, using simple text file
parsing routines from ../mtree/getid.c

add config.h when compiled as host tool.

use -1 instead of 0 as the value to indicate if a numeric parameter
hasn't been set and thus needs a default override.
allows minfree and rotdelay to take a value of 0.

setprogname(argv[0])

rework `fsnode' into two components; `fsnode' contains the name and tree
information and a link to the appropriate fsinode, and `fsinode' contains
the inode information and a reference count. multiple fsnodes may point
to the same fsinode. this replaces the 'dup' pointer in the previous code.

- Use strsuftoull() from dd(1) to parse numbers, rather than home grown
version. The new version supports products, 'g' (GB) and 't' (TB)
suffices, and `b' now means `blocks' instead of `bytes'.
Functionality requested by perry.
- debug is now a u_int instead of an int.
- Ensure that various numbers have sane upper limits (e.g, 99 for %, etc)

Include <limits.h> to get LLONG_{MIN,MAX}.

tweak copyright

cast time_t values to long if printing as %ld to appease the sparc64 port

add __RCSID()

branches: 1.1.1;
Initial revision

From FreeBSD D50078 by jrtc27:

Linux's struct stat does not have an st_flags member, but we still want
to be able to create FreeBSD disk images with non-zero st_flags. Add a
replacement member to struct fsinode to store it on Linux.

Document debugging options and allow specifying them by name.

branches: 1.42.2;
Bump inode number on target FS from 32 bits to 64 bits

Add an options pretty-printer, pass some lint.

Revert, sorry this was not intended to be committed.

When creating disc images, keep the proposed size a multiple of the blockingnr
for good measure; this prevents possible burning/copying issues on packet
media.

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

Add a -L option to follow all symbolic links. Useful if you have symlinks
in a makefs directory tree but want to refer to the actual file.

Provide a -T option to set timestamps to a consistent value for MKREPRO

Implement `makefs -t udf'.

Formatting options may be enhanced to make it more in line with newfs_udf on
say labeling.

add a replace flag so we can overlay exiting files when we merge directories.

add and document offset.

- don't abuse vp->fs to mean struct fs for ffs and struct msdos_opts;
make it always fsinfo_t and change void * to that.
- kill unused structure members.

make everything use the generic options parser.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

remove duplicated code, and try to cleanup parsing by using the shared code.
cd9660 needs a lot of work.

- add support for parsing different types; not just int.
- add beginning of msdos support.

branches: 1.27.2;
Add -Z to enable sparse files - for ffs at least.
Based on a patch from shesha at juniper.

- fix compilation isses
- add some error checking
- avoid alloca
- add missing frees

added chfs support for makefs

Patch from Jung-uk Kim (jkim at FreeBSD dot org) to allow contents of multiple
directories to be merged to the current image.

branches: 1.23.2;
Re-enable v7fs support in the tools version of "makefs" after the
issues have been fixed by Alan Barrett. Tested under Mac OS X.

Don't include v7fs support into tool version of "makefs" because it
breaks the build.

v7fs(7th edition(V7) file system) support.

fix for 64 bit time_t and dev_t.

implement -x by pruning fsnodes tree before building filesystem rather
than by skipping nodes while building filesystem

add implementation of free_fsnodes() which unlinks a node
from the hierarchy and frees it and any of its children
use it to free memory before a normal exit, which helps diagnostics

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

Refactor the filesystem specific portions out of makefs.c/makefs.h completely.
Instead of extending fsinfo_t it now holds a void * to file system specific
data. This is then setup/cleaned up by the additional of 2 additional
callbacks. Makes adding new filesystems simpler as almost no code has
to be updated in the generic makefs code now.

Check for fstatvfs and provide/use it only if the host system has it.
Fixes PR#27221

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

Replace the statfs() family of system calls with statvfs().
Retain binary compatibility.

branches: 1.12.2;
Overhaul how `build.sh tools' are used:

* Rename "config.h" to "nbtool_config.h" and
HAVE_CONFIG_H to HAVE_NBTOOL_CONFIG_H.
This makes in more obvious in the source when we're using
tools/compat/config.h versus "standard autoconf" config.h

* Consistently move the inclusion of nbtool_config.h to before
<sys/cdefs.h> so that the former can provide __RCSID() (et al),
and there's no need to protect those macros any more.

These changes should make it easier to "tool-ify" a program by adding:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
to the top of the source files (for the general case).

Add support for UFS2. UFS2 is an enhanced FFS, adding support for
64 bit block pointers, extended attribute storage, and a few
other things.

This commit does not yet include the code to manipulate the extended
storage (for e.g. ACLs), this will be done later.

Originally written by Kirk McKusick and Network Associates Laboratories for
FreeBSD.

Tweak Jason's -x support to pass its state around as part of fsinf_t,
just like all the other "global" options.
Update the usage for -x.
Crank date on man page.

(Thanks to Jason for adding -x; I've been meaning to add that
functionality for a while :-)

Add a -x option which causes makefs(8) to exclude any file system node
not explicitly listed in the specfile.

remove (now) incorrect comment

Roll in fixes to permit cross-compiling from non-NetBSD hosts. This
round has been tested on Solaris/x86 and Linux hosts.

* Add host tools cap_mkdb, ctags, m4, uudecode.
* Protect __RCSID() and __COPYRIGHT() better.
* Reduce the number of places that need to include "config.h", to keep
sources closer to their "vanilla" versions.
* Add more compat #defines and autoconf-checked functions.

- crank default cpg from 16 to 65536, and let the auto-sizing code in
mkfs() do its job
- pass cpgflg=1 to mkfs() if the user provided cpg, =0 otherwise
- improve error reporting in ffs_write_file()
- only add superblock & block map slop for one cylinder group.
trying to calculate the number of required cylinder groups caused
too many problems when tweaking parameters for small file systems.
rely upon slop from minfree, and upon ffs_write_file() to complain
if the user is too aggressive about tweaking the parameters

fix bugs reported by matt green:
- when calculating file size, round up to frag size not block size
- #define FFS_EI in makefs.h, and include that before [uf]fs/*.h

other stuff:
- round up final size to next block (instead of next sector)
- protect makefs.h from multiple inclusion

rework `fsnode' into two components; `fsnode' contains the name and tree
information and a link to the appropriate fsinode, and `fsinode' contains
the inode information and a reference count. multiple fsnodes may point
to the same fsinode. this replaces the 'dup' pointer in the previous code.

- Use strsuftoull() from dd(1) to parse numbers, rather than home grown
version. The new version supports products, 'g' (GB) and 't' (TB)
suffices, and `b' now means `blocks' instead of `bytes'.
Functionality requested by perry.
- debug is now a u_int instead of an int.
- Ensure that various numbers have sane upper limits (e.g, 99 for %, etc)

tweak copyright

branches: 1.1.1;
Initial revision

makefs/msdos: Fix broken [extra-directory ...] case

"path + cur->name" is not same as "cur->root + cur->path + cur->name"
for extra-directory files, as extra-directory files are
in different location. Do what makefs ffs code does.

PR/57945: Jan-Benedict Glaw: makefs: msdos filesystems with label not
reproducible.
When we specify a label, the volume label is used to create the root dir
and it gets stamped by the timestamp. We were trying to set the timestamp
in the fs-specific part of the options parsing which did not work since
the method was called only for fs_specific options. Move setting of the
timestamp just before we create the fs, where we reconcile the rest of
the options.

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

Return a proper exit status on failure on creating msdosfs.

No particular comment in PR bin/57752.
Should be pulled up to netbsd-10 and netbsd-9.

branches: 1.21.2;
makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

branches: 1.20.14;
leave the size alone and set the create_size to include the offset. It
does not matter anyway, but it makes more sense this way.

for consistency with ffs, don't count the offset into the size.

allow 0 timestamp

start fixing msdosfs for reproducible builds; does not work yet.

branches: 1.16.2; 1.16.4;
Add support to msdosfs and makefs to generate correct Unicode (UCS-2) directory
entries from UTF8 encoded file names.

remove clause 3.

branches: 1.14.4; 1.14.10;
- more changes to make -O work
- fix err* calls.

- don't abuse vp->fs to mean struct fs for ffs and struct msdos_opts;
make it always fsinfo_t and change void * to that.
- kill unused structure members.

gc global sectorsize.

add missing arguments.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

fix recursion error handling. top level directory works.

This works well enough to populate plain files in the root dir. creating
directories fails.

fixed directory entry allocation. Now the file data remains and is currently
broken.

add more msdos stuff.

- Fix single letter option parsing.
- Instead of returning 1 << index and 0 on error,
return just index and -1 on error for the set_option*() routines.

ok, we can create the filesystem now.

return the bit of the option set, so that others can act on it.

remove duplicated code, and try to cleanup parsing by using the shared code.
cd9660 needs a lot of work.

- add support for parsing different types; not just int.
- add beginning of msdos support.

remove clause 3.

branches: 1.2.4; 1.2.10;
add debugging, fix warnings.

add more msdos stuff.

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

When creating disc images, keep the proposed size a multiple of the blockingnr
for good measure; this prevents possible burning/copying issues on packet
media.

Avoid warning by clang of unused variable

While here, use the computated obj_size even though it's always the same as
its information length since we don't use extended attribute files or
stream directories.

Fix endian bug that resulted in UniqueIDs of 0 in file identifiers; note that
the the unique ID in a longad stored in a FID is 32 bit where in the file
descriptor (fe/efe) its 64 bits long.

No need to use ceilf() when ceil() is already used

Fix endian bug in makefs for udf

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

makefs: fix some typos

Refactor and rewrite of newfs_udf(8) and makefs(8) (-t udf) with a shared
section for fsck_udf(8).

Newfs_udf and makefs support predefined disc image profiles, harddisc
partitions (disklabel and wedges on all generic block devices) and all optical
media types on CD/DVD/BD writers.

Newfs_udf and makefs now also support formatting of UDF 2.50 with a metadata
partition.

s/helt/held+s/eroneously/erroneously/+s/splitted/split/+s/recommented/recommended/

Remove unused variable (to fix the build)

Believe the datablocks predictor when determining if data on a node gets
stored internal or not. Also make a note that the datablocks predictor takes
NO extended attributes stored in the node into account

In rare cases it could lead to confusion where the predictor would say it
wouldn't fit internally when it could just have fitted. This would trigger the
assertion. Now it will on rare accasions create a datablock even though it
might have fitted.

- add or adjust /* FALLTHROUGH */ where appropriate
- add __unreachable() after functions that can return but won't in
this case, and thus can't be marked __dead easily

branches: 1.18.12;
If we are using a timestamp from the command line, don't pay attention to
the user timezone, use UTC instead (for reproducible builds).

branches: 1.17.2; 1.17.4;
improve error messages (remove \n, use __func__, etc)

branches: 1.16.4; 1.16.8;
long_ad is a mix of struct and union, so use memset to initialise in the
one place it is used.

initialize dummy_ref

fix unused variable warnings.

when building as a tool, only use struct tm tm_gmtoff if HAVE_STRUCT_TM_TM_GMTOFF is defined -- fixes cygwin build

Fix possible 32 bit issues. In practice hard to reach but better be safe than
sorry.

Oops, used size_t when off_t was meant. This results in the 4G file size
modulation seen on 32 bit machines due to size_t being 32 bit there.

While here, also fix compilation errors he@ discovered while compiling on
NetBSD/vax.

Switch back program version to the version of the newfs_udf/makefs code.

While here, visit the 80 columns police :)

Add minimum UDF version specification to makefs(8) -t udf.

Allow for the logical volume label to be specified as well as the physical
volume label. Also allow the volumeset name to be specified if desired. The
syntax follows the newfs_udf(8) syntax.

Implement auto-setting of sector size and disc size when specifying
-oT=<devtype> to one of the supported disk types.

While here, also fix where the -s size argument would be overriden by the
calculated size.

Fix assert() on allocation type by masking it with the flag allocation bit
mask. Fixes internal allocation writing in nodes with suid/sgid/sticky
accessnode.

Fix memory leaks found by Coverity and fix memoryleaks in the new code

Revert joergs patch and make it compile in both the normal as in the tools
environment by providing the MMC datastructures separately for compat.

Allow building UDF parts of makefs on !NetBSD.

Make memset() usage consequent in using the right types

Implement `makefs -t udf'.

Formatting options may be enhanced to make it more in line with newfs_udf on
say labeling.

make everything use the generic options parser.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

- Fix single letter option parsing.
- Instead of returning 1 << index and 0 on error,
return just index and -1 on error for the set_option*() routines.

remove duplicated code, and try to cleanup parsing by using the shared code.
cd9660 needs a lot of work.

- add support for parsing different types; not just int.
- add beginning of msdos support.

branches: 1.3.2; 1.3.8;
newfs_v7fs(8)'s -v and -P options are obsolete. changed to newfs(8) compatible -V option.

Allow cross-compiling from Linux

v7fs(7th edition(V7) file system) support.

newfs_v7fs(8)'s -v and -P options are obsolete. changed to newfs(8) compatible -V option.

v7fs(7th edition(V7) file system) support.

s/sollowed/followed/ in comment.

From FreeBSD D50078 by jrtc27:

Linux's struct stat does not have an st_flags member, but we still want
to be able to create FreeBSD disk images with non-zero st_flags. Add a
replacement member to struct fsinode to store it on Linux.

PR/58759: Kenichi Hashimoto: failed merge with extra-directory using -r option

branches: 1.40.2;
Add an option to warn and error out for mtree missing entries.

makefs: fix out-of-bounds fsnode count in fsnode_sort

Found by running './makefs img.dat cd9660'.

While here, apply more KNF.

use __func__

make a separate sorting function and KNF (thanks rillig)

pass lint, simplify

makefs: Fix symlink permission bits

Permission bits for symlinks are taken straight from `lstat()`. However, the
actual bits presented to the user are filesystem/kernel specific. For example,
Linux with ext2/3/4 will use 0777 for symlinks, whereas NetBSD/FFS will
show 0755. As far as `makefs` is in the loop, the target filesystem will likely
be FFS, so use 0755.

makefs: Sort directory contents by name (Jan-Benedict Glaw)

`makefs` inserts nodes into its internal data structures in the order as
returned by `readdir()` calls. As this is unpredictable, sort entries by
name before creating the target filesystem.

This is done by first converting the (per-directory) linked list into
a plain array, sort it, finally re-link the list. Special case for the
sorting function: The "." directory entry seems to be ment to be always
at the front, so always check that first.

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

makefs: fix some typos

Add a -L option to follow all symbolic links. Useful if you have symlinks
in a makefs directory tree but want to refer to the actual file.

Provide a -T option to set timestamps to a consistent value for MKREPRO

add a replace flag so we can overlay exiting files when we merge directories.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

branches: 1.26.2;
WARNS=5

Patch from Jung-uk Kim (jkim at FreeBSD dot org) to allow contents of multiple
directories to be merged to the current image.

branches: 1.24.8;
fix for 64 bit time_t and dev_t.

implement -x by pruning fsnodes tree before building filesystem rather
than by skipping nodes while building filesystem

add implementation of free_fsnodes() which unlinks a node
from the hierarchy and frees it and any of its children
use it to free memory before a normal exit, which helps diagnostics

free specfile structure once it has been applied

use new implementation of link_check which i
implemented for du.c revision 1.25 :

Use open address hashtable with multiplicative double hash probing
and exponential hashtable growth for tracking previously seen inodes.
This replaces the O(n^2) behavior of linkchk() with an implementation
that should be O(n*logn)

Stop duplicating the ISO9660 sources in sys/fs/cd9660/ for makefs(8).

64 bit inode changes

Completely rework how tools/compat is done. Purge all uses/references to
_NETBSD_SOURCE as this makes cross building from older/newer versions of
NetBSD harder, not easier (and also makes the resulting tools 'different')

Wrap all required code with the inclusion of nbtool_config.h, attempt to
only use POSIX code in all places (or when reasonable test w. configure and
provide definitions: ala u_int, etc).

Reviewed by lukem. Tested on FreeBSD 4.9, Redhat Linux ES3, NetBSD 1.6.2 x86
NetBSD current (x86 and amd64) and Solaris 9.

Fixes PR's: PR#17762 PR#25944

Use uint64_t instead of uint32_t to store a host inode when checking
for duplicate inodes, as Cygwin's stat(2) has a 64 bit st_ino.
Fix from Ian Lance Taylor in private email.

branches: 1.15.2;
realloc pedant

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22366, verified by myself.

just for safety - use sizeof

Add a -x option which causes makefs(8) to exclude any file system node
not explicitly listed in the specfile.

s/ulong/unsigned long/. ulong is for SysV compatibility, there's no
point in using it in our code. This change makes makefs(8) compile on
host systems without ulong.

use readlink with bufsize - 1; approved thorpej.

branches: 1.9.2;
If an entry is in the specfile but not in the underlying file system, and it's
marked "optional", don't add it. (e.g, don't create a zero-length file).

Roll in fixes to permit cross-compiling from non-NetBSD hosts. This
round has been tested on Solaris/x86 and Linux hosts.

* Add host tools cap_mkdb, ctags, m4, uudecode.
* Protect __RCSID() and __COPYRIGHT() better.
* Reduce the number of places that need to include "config.h", to keep
sources closer to their "vanilla" versions.
* Add more compat #defines and autoconf-checked functions.

Merge create_fsinode() into create_fsnode(), so that a newly created
fsnode always has a valid inode member. Problem noted by Artem Belevich.

rework `fsnode' into two components; `fsnode' contains the name and tree
information and a link to the appropriate fsinode, and `fsinode' contains
the inode information and a reference count. multiple fsnodes may point
to the same fsinode. this replaces the 'dup' pointer in the previous code.

tweak copyright

cast time_t values to long if printing as %ld to appease the sparc64 port

- apply_specdir(): when adding a new symlink from the specfile, do the assert
check on the correct node
- apply_specentry(): if this node is a duplicate of another, apply the
changes to the `master' entry instead of this one.
- fix inotype() to DTRT
- comment out some debugging info that is too verbose

add __RCSID()

branches: 1.1.1;
Initial revision

Remove many HAVE_GCC || HAVE_PCC conditionals as the options also apply
to Clang. Add a few cases of HAVE_LLVM for -fno-strict-aliasing.

branches: 1.2.2;
remove most of the remaining HAVE_GCC tests that are always true in
the modern world.

branches: 1.1.6;
Move fs-specific stuff in Makefile to fs/Makefile.inc to clarify
top-level Makefile.

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

branches: 1.1.8; 1.1.14;
Add support for encoding RISC OS metadata into ISO 9660 filesystems.
This is done by adding the general ability to stuff non-SUSP data into
the end of the System Use field of a Directory Record, which required
some amount of rewriting of the SUSP support. I think the result is
at least as good as what came before, and I've fixed at least one bug
along the way. Tested against RISC OS 3.70 on my Risc PC.

Oh, why do we want it? It should allow us to make acorn{26,32} CDs that
can be booted directly from RISC OS without mucking around copying the
bootloader to a native filing system.

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

Add support for encoding RISC OS metadata into ISO 9660 filesystems.
This is done by adding the general ability to stuff non-SUSP data into
the end of the System Use field of a Directory Record, which required
some amount of rewriting of the SUSP support. I think the result is
at least as good as what came before, and I've fixed at least one bug
along the way. Tested against RISC OS 3.70 on my Risc PC.

Oh, why do we want it? It should allow us to make acorn{26,32} CDs that
can be booted directly from RISC OS without mucking around copying the
bootloader to a native filing system.

We know the target string buffer only needs 18 but gcc can't figure this out.
Fixes compiling with -Werror=format-truncation

If we are using a timestamp from the command line, don't pay attention to
the user timezone, use UTC instead (for reproducible builds).

branches: 1.4.58; 1.4.62;
PR/36006: Takahiro Kambe: makefs(8) create cd9660 file system with wrong time
stamp because localtime(3) and gmtime(3) share the same buffer. Fixed by using
the _r functions.

branches: 1.3.4;
Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

Remove invisible whitespace at ends of lines that disrupted my
speedy cursor-movement habits in vi.

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

fix unused variable warnings.

- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

branches: 1.11.6; 1.11.12;
Handle > 2GB images. Thanks to Greg Troxel for testing!

- simple fseek->fseeko conversion. probably needs more work.
- use a constant instead of sprinkling 2048 everywhere.

Typo: debuging -> debugging.

emasculate debugging mode for the hosttool version of makefs, to avoid various
hairy endian-decoding macros/inlined functions/etc.

revert this change, too.

Due to recent changes in the cd9660 code, crosscompilation of tools/makefs
now fails due to the use of [be]e{16,32}dec. Add a compat header, and force
the inclusion of <sys/endian.h>.

from Anon Ymous: compile with -Wextra

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

Remove invisible whitespace at ends of lines that disrupted my
speedy cursor-movement habits in vi.

Overhaul makefs -t cd9660.

Functional: Do not add a missing period or a version number to
directories, since that doesn't fit the standard (thanks
Daniel Watt).

Add RRIP PX records (POSIX meta-information) to '.'
and '..' entries. Quiets kernel complaint, when we
mount a cd9660 filesystem, that there is a missing PX
entry.

Cosmetic: Reduce height of staircases. Remove an unused member
from struct _cd9660node. Extract a subroutine. KNF.

Use a TAILQ macros for lists of both SUSP/RRIP attributes
and directory entries.

XXX
XXX The new RRIP PX entries are not quite right.
XXX

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

Use EXIT_SUCCESS and EXIT_FAILURE where appropriate per style guide.

The boot catalog pointer is a DWord, but we previously populated it via
cd9660_bothendian_dword which overwrote four unused bytes following it.
See El Torito 1.0 (1995) Figure 7 for details.

Found by Coverity on FreeBSD

makefs(8): Nix trailing whitespace.

Computed by machine, eyeballed by hand, so to speak.

Add "efi" as a supported boot image type and derive the platform ID for
the validation entry from the default boot image instead of hard-coding
X86.

Correctly mark the last El Torito section header.

Pointed out by Benno Rice via DM.

branches: 1.22.2;
Initialize boot_catalog_entry's entry_type properly.

This had been missing but the type was used in cd9660_setup_boot().

From OpenBSD usr.sbin/makefs/cd9660/cd9660_eltorito.c r1.10.

branches: 1.21.4;
makefs(8): add cd9660 eltorito-alt-boot option for EFI boot.

branches: 1.20.12; 1.20.16;
- use emalloc and friends
- kill a bunch of global variables, more work to be done here
- homogenize option parsing. more work for cd9660
- use the new options parsing code to print an fs-specific usage

branches: 1.19.2;
WARNS=5

Pull FreeBSD fix from Andreas Tobbler:

- Write the ISO9660 descriptor after the apm partition entries.
- Fill the needed pmPartStatus flags. At least the OpenBIOS
implementation relies on these flags.

This commit fixes the panic seen on OS-X when inserting a FreeBSD/ppc disc.
Additionally OpenBIOS recognizes the partition where the boot code is located.
This lets us load a FreeBSD/ppc PowerMac kernel inside qemu.

branches: 1.17.2;
Move inclusion of sys/endian.h from cd9660_eltorito.c to cd9660.h
since actual user is not cd9660_eltorito.c but iso.h and
cd9660_eltorito.h.

Actually, include order/place of sys/endian.h doesn't matter on
netbsd since it is always included by sys/types.h but it's not
true on other system. This should fix cross build breakage on
freebsd introduced by rev. 1.16 of cd9660_eltorito.c.
Problem reported and fix suggested on twitter.

include system headers after config.h

From Nathan Whitehorn (nwhitehorn at freebsd dot org):
Add code to generate bootable ISOs on Powermac and CHRP systems.
Synthesize some partition maps (APM and MBR, respectively) pointing
to (a) the whole disk, and (b) relevant El Torito boot images that
have been added by other code. These partition maps are a little
bit funny looking, but they seem to work. FreeBSD has been using
this successfully in their release generation on powerpc, as well
as generating all non-SPARC install media. SPARC support could
probably be added as an extension of this patch.

Handle > 2GB images. Thanks to Greg Troxel for testing!

- simple fseek->fseeko conversion. probably needs more work.
- use a constant instead of sprinkling 2048 everywhere.

Fix typo in cd9660 functionname cd9660_setup_boot_volume_descritpor ->
cd9660_setup_boot_volume_descriptor.

Interestingly the same typo was used in 3 files!

branches: 1.11.20;
Coverity CID 548: Fix resource leak.

Oops, condition another printf on verbose_level > 0.

Condition several cd9660 printfs on verbose_level > 0. Silences
makefs -t cd9660.

Now that I understand what order boot-images and boot-entries are
intended to be kept on the boot-images and -entries lists, keep
them in that order: images grouped and entries grouped by system,
with later images nearer the front of the list.

*Really* apply boot options to the last boot-image specified.

Bug fix: record the actual number number of 512-byte "virtual"
sectors in the Sector Count field of the Initial/Default Entry in
the "El Torito" Boot Catalog, instead of blithely writing 1 (!!)
into the Sector Count field. Ditto the Section Entries.

Remove last remnants of cd9660's unsupported boot-load-size option.

Really use cd9660's boot-load-segment argument.

Change the boot images list from a LIST to a TAILQ.

Re-implement CD9660_BLOCKS() with howmany().

Introduce ELTORITO_DPRINTF and some debug messages.

Zero memory malloc'd in the El Torito code.

Revise some debugging outputs.

Insert a new boot-image on the boot-images list after the last
boot-image for the same architecture. It makes a (trivial) functional
difference.

Miscellaneous cosmetic changes. Remove useless variable-assignment.

Cosmetic: add temporary variables for lengthy, repeated clauses
like 'default_entry->entry_data.IE'. KNF.

Remove invisible whitespace at ends of lines that disrupted my
speedy cursor-movement habits in vi.

Correct strcmp(3) usage (expect equal -> 0).

Use LIST_ macros instead of accessing list structure members lh_first
and le_next directly.

Change from malloc/strcpy idiom to strdup.

Revamp err, errx, warn, warnx usage.

Misc. cosmetic tweaks.

Make -o bootimage=sysname;filename expect NetBSD architecture names
(x86 -> i386; mac -> {macppc, mac68k}; ppc -> powerpc) for sysname.

cd9660 support. Written by Daniel Watt, Walter Deignan, Ryan Gabrys,
Alan Perez-Rathke and Ram Vedam.

Prototyping, KNF, WARNS=3 by me.

Note that this is marked experimental for now. It works well enough to
make NetBSD/i386 bootable images, but does need more work.

makefs(8): add cd9660 eltorito-alt-boot option for EFI boot.

branches: 1.5.28; 1.5.32;
Get rid of the DOS line endings in the file.
No change of the code intended.

Now that I understand what order boot-images and boot-entries are
intended to be kept on the boot-images and -entries lists, keep
them in that order: images grouped and entries grouped by system,
with later images nearer the front of the list.

*Really* apply boot options to the last boot-image specified.

Bug fix: record the actual number number of 512-byte "virtual"
sectors in the Sector Count field of the Initial/Default Entry in
the "El Torito" Boot Catalog, instead of blithely writing 1 (!!)
into the Sector Count field. Ditto the Section Entries.

Remove last remnants of cd9660's unsupported boot-load-size option.

Really use cd9660's boot-load-segment argument.

Change the boot images list from a LIST to a TAILQ.

Re-implement CD9660_BLOCKS() with howmany().

Introduce ELTORITO_DPRINTF and some debug messages.

Zero memory malloc'd in the El Torito code.

Revise some debugging outputs.

Insert a new boot-image on the boot-images list after the last
boot-image for the same architecture. It makes a (trivial) functional
difference.

Miscellaneous cosmetic changes. Remove useless variable-assignment.