CVS log for src/sbin/restore/tape.c

Up to [cvs.NetBSD.org] / src / sbin / restore

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

triple "r" typos, mainly s/interrrupt/interrupt/ in comments and one definition.

apply this commit from FreeBSD:

  commit 9dda00df7e8f9279a43d92758df6a7e10a9aed95
  Author: Chuck Silvers <chs@FreeBSD.org>
  Date:   Mon Dec 12 08:14:17 2022 -0800

      restore: fix restore of NFS4 ACLs

      Changing the mode bits on a file with an NFS4 ACL results in the
      NFS4 ACL being replaced by one matching the new mode bits being set,
      so when restoring a file with an NFS4 ACL, set the owner/group/mode first
      and then set the NFS4 ACL, so that setting the mode does not throw away
      the ACL that we just set.

      Reviewed by:    mckusick
      Differential Revision:  https://reviews.freebsd.org/D37618

apply this change from FreeBSD:

  commit c028393d7072f1f88efd8d6e6c77bb9b15b3f3b6
  Author: Kirk McKusick <mckusick@FreeBSD.org>
  Date:   Fri Apr 11 21:48:14 2008 +0000

      Correctly set file group when restore is run by a user other than root.

include the filename in the panic message for missing data.

instead of merely seeing:
	getfile: lost data
now this is seen:
	getfile: lost data: ./usr/libdata/debug/usr/libexec/cc1.debug

Add external attribute dumping and restoring support from FreeBSD.
Does not fully work yet, attributes are being saved and restored correctly,
but don't appear in the restored files somehow.

s/skiping/skipping/

If we hit an unknown header type (likely to be a corrupt record), and
the user choose to not abort, skip to the next header instead of trying
to use it.
This allowed me to recover files from a corrupt dump, instead of
getting a segfault.

Pull up following revision(s) (requested by nakayama in ticket #1096):
	bin/mv/mv.c: revision 1.44
	bin/cp/utils.c: revision 1.43-1.44
	lib/librumphijack/hijack.c: revision 1.112-1.115
	usr.bin/touch/touch.c: revision 1.33
	sbin/restore/tape.c: revision 1.68
	sbin/restore/dirs.c: revision 1.51
Don't truncate at sub-microsecond while preserving timestamps.

One of motivation of this change is to make the behavior of test(1)
-nt/ot with preserved copy (like cp -p) closer to the NetBSD 6.
Of course whether full timestamps are kept or not depends also on
underlying file system.

The ifdef added in mv(1) since existing ifdefs was our local change
to compile it on solaris (though I couldn't test it):
http://mail-index.netbsd.org/tech-userlevel/2014/11/28/msg008831.html

Fix the name of failed function in warning message.

Hijack utimensat(2) so that t_vfs test passes after cp(1)/mv(1) are
changed to use the system call.  Linux also has this system call, but
not tested this on linux.

Also hijack futimens(2) so that t_sh test passes.

Define a generic ATCALL() and use it to implement utimensat()
Make ATCALL() behave for absolute paths too.

Pull up following revision(s) (requested by nakayama in ticket #1096):
	bin/mv/mv.c: revision 1.44
	bin/cp/utils.c: revision 1.43-1.44
	lib/librumphijack/hijack.c: revision 1.112-1.115
	usr.bin/touch/touch.c: revision 1.33
	sbin/restore/tape.c: revision 1.68
	sbin/restore/dirs.c: revision 1.51
Don't truncate at sub-microsecond while preserving timestamps.

One of motivation of this change is to make the behavior of test(1)
-nt/ot with preserved copy (like cp -p) closer to the NetBSD 6.
Of course whether full timestamps are kept or not depends also on
underlying file system.

The ifdef added in mv(1) since existing ifdefs was our local change
to compile it on solaris (though I couldn't test it):
http://mail-index.netbsd.org/tech-userlevel/2014/11/28/msg008831.html

Fix the name of failed function in warning message.

Hijack utimensat(2) so that t_vfs test passes after cp(1)/mv(1) are
changed to use the system call.  Linux also has this system call, but
not tested this on linux.

Also hijack futimens(2) so that t_sh test passes.

Define a generic ATCALL() and use it to implement utimensat()
Make ATCALL() behave for absolute paths too.

Don't truncate at microseconds while preserving timestamps.

One of motivation of this change is to make the behavior of test(1)
-nt/ot with preserved copy (like cp -p) closer to the NetBSD 6.
Of course whether full timestamps are kept or not depends also on
underlying file system.

The ifdef added in mv(1) since existing ifdefs was our local change
to compile it on solaris (though I couldn't test it):
http://mail-index.netbsd.org/tech-userlevel/2014/11/28/msg008831.html

resync with head

sync with head

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.

Use __dead

Add wrapper functions around hash algorithm operations to avoid
undefined behavior arising from illegal function casts. As a side
effect, no longer need -Wno-pointer-sign either.

Simplify silly code and make it closer to type-safe. This causes amd64
gcc to reorder two pairs of instructions for some reason but the object
files are otherwise unchanged.

Sync with HEAD.

Third (and last) commit. See http://mail-index.netbsd.org/source-changes/2009/05/13/msg221222.html

fix sign-compare issues

remove a useless cast.

Sync with HEAD.

Pull up following revision(s) (requested by hannken in ticket #210):
	include/protocols/dumprestore.h: revision 1.15
	sbin/restore/tape.c: revision 1.61
Make restore work for Linux dump volumes by ignoring extended attribute
records on these volumes.  Tested with Centos 5.2.
Reviewed by: Manuel Bouyer <bouyer@netbsd.org>

Make restore work for Linux dump volumes by ignoring extended attribute
records on these volumes.  Tested with Centos 5.2.

Reviewed by: Manuel Bouyer <bouyer@netbsd.org>

sync with HEAD

Sync with HEAD.

Fix some inconsisent/conflicting definitions and missing parameters

Pull up following revision(s) (requested by hannken in ticket #483):
	sbin/restore/tape.c: revision 1.59 via patch
Add support for Solaris ufsdump volumes with more than 512*1024 inodes.
Here the bitmaps are written as
	CLRI or BITS with c_count <= 512
	ADDR* for the remaining blocks
Remove the bitmap handling from getfile(), remove xtrmap() and xtrmapskip().
Add new function getbitmap() modeled after getfile() that does bitmap
allocation, bitmap expansion and sets maxino.
Reviewed by: Manuel Bouyer <bouyer@netbsd.org>

Pull up following revision(s) (requested by hannken in ticket #1674):
	sbin/restore/tape.c: revision 1.59
Add support for Solaris ufsdump volumes with more than 512*1024 inodes.
Here the bitmaps are written as
	CLRI or BITS with c_count <= 512
	ADDR* for the remaining blocks
Remove the bitmap handling from getfile(), remove xtrmap() and xtrmapskip().
Add new function getbitmap() modeled after getfile() that does bitmap
allocation, bitmap expansion and sets maxino.
Reviewed by: Manuel Bouyer <bouyer@netbsd.org>

Add support for Solaris ufsdump volumes with more than 512*1024 inodes.
Here the bitmaps are written as

	CLRI or BITS with c_count <= 512
	ADDR* for the remaining blocks

Remove the bitmap handling from getfile(), remove xtrmap() and xtrmapskip().
Add new function getbitmap() modeled after getfile() that does bitmap
allocation, bitmap expansion and sets maxino.

Reviewed by: Manuel Bouyer <bouyer@netbsd.org>

Convert a couple of "(void)&" constructs to volatile.
Flag various unused parameters in restore and dump so that these will
compile with -Wextra.  (Note: restore uses some stuff from dump.)

find rmd160.h

c99 iniitializer

Use crypto/rmd160.h.

64 bit inode changes

sprinkle const.

Kill __P(), use ANSI function declarations.

Add an option (-M) to write a mtree specification (which needs to be passed
through sort before being feed to mtree) with file flags, instead of restoring
file flags at the same time as other attributes. Fix various issue with
schg, uchg, sappnd or uappnd flags which cause restore to fail in some case.
Discussed on tech-userlevel:
http://mail-index.netbsd.org/tech-userlevel/2004/10/12/0000.html

Add -D flag which computes the digest of each regular files in the archive
using specified algorithm (currently md5, rmd160 or sha1) and write them
to standard output.

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

Patches provided by Joel Baker in PR 22308, 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.

Since "host" is used when RRESTORE is defined, enclose variable definition
with "#ifdef RRESTORE" and "#endif".

__STDC__ is always defined on NetBSD.

Don't forget to update curfile, if tape begins with TS_ADDR type block.
In particular, this means that if one member (say the last member) of a tape
set begins with continuation blocks, restore will not consider that tape
to contain all the inodes (curfile.ino==0 at the beginning of the tape).

Close PR #15545.

fix compile issues now that on-tape times are int32_t instead of time_t

Remove incorrect advice for lint; the control may return from panic().

Pull up revisions 1.40-1.41 (requested by enami):
  Make restore correctly handle large dumps with large inode maps.
  Fixes PR#11188.

Extract file flags of symbolic link.

Add comment and sanity check about why we need not to skip remaining
blocks if we are reading inode bitmaps.

Correctly extract the inode maps bigger than 512 tape blocks.

Pull up revision 1.39 (requested by sommerfeld):
  Print the header even on dumps from a system with an unset
  host name.  Fixes PR#9282.

Fix PR9282: restore doesn't print dump header on dumps from host with
unset hostname.

user friendly enhancments.
reworked from [bin/6710] by Brian Grayson <bgrayson@ece.utexas.edu>

Add an unlink (-u) flag so that we can use this for installs.

Need <time.h> for ctime() prototype.

use static int, not just static

Pull rev 1.34 up from trunk (enami)

Fix the bug introduced in last change; pass IFCHR or IFBLK bit to
mknod(2) via mode.

Pull rev 1.32 and 1.33 up from trunk (enami)

Use file mode 0600 when creating special file or fifo.
Suggested by Charles M. Hannum.

Restore file attributes after extracting regular file.  Fixes PR#4515
from stephen.ma@jtec.com.au:

- call getfile() before altering file attributes.
- open file with mode 0600 instead of 0666 so that file won't remain
  group or world readable/writable even if getfile() terminated.
- also, move skipfile() before altering file attributes in IF{CHR,BLK} and
  IFIFO case for symmetry (suggested by Charles M. Hannum).

Use futimes(2).  Do {f,}utimes(2) *before* {f,}chflags(2), for obvious
reasons.

Restore owner/group/mode/atime/mtime of symbolic links.

resolve conflicts from lite-2 merge

imported from lite-2

* cleanup for WARNS=1
* fix use of .Nm

as inspired by a commit message of Todd Miller <millert@cvs.openbsd.org>:
* cleanup a bit for -Wall
* use __progname instead of "restore"

use _PATH_DEFTAPE from <paths.h> instead of "pathnames.h"
cleanup references/examples to use /dev/rst* instead of /dev/rmt*
minor .Nm usage cleanup in man page

* use mkstemp() instead of mktemp()
* remove "register" qualifiers

Fix typo: PR bin/3059

From trunk:
Add some missing byte-swaps, so this works across endians.

Pullup request from Chris G Demetriou <Chris_G_Demetriou@ux2.sp.cs.cmu.edu>

>Make dump and restore usable on the alpha

Make sure to byte-swap all of the relevant fields; especially c_flags.

When initializing 'fssize,' the size of the data buffer to be used when
writing data to the file system, if the "optimal" file system I/O
operation block size is less than TP_BSIZE, leave fssize alone (i.e.
at its default setting of MAXBSIZE).  This was causing restore's
stack to be trashed, because the end-of-buffer checking/flushing code
around line 680 would never notice that the buffer was full (because
it'd be comparing a buffer segment index, which would always be >= 1, to
fssize / TP_BSIZE, which could be zero in that case), and would keep
filling and filling and filling...

patches from Tom I Helbekkmo <tih@nhh.no> to deal with type-size issues,
so this works (better, at least) on 64-bit machines (e.g. alpha).

Be more paranoid with our effective uid.

stat structure POSIXification

Back out previous change.

typeof(timeval.tv_sec) != time_t

convert to new RCS Id conventions; reduce my headache

Use atexit() rather than tweaking every exit() to call the cleanup function.

Mostly sync with CSRG.

Eliminate uses of some obsolete functions.

Allow extracting FIFOs.

Import 4.4-Lite version.

Update from 4.4-Lite, with local changes.

copyright foo

kill some #ifndefs

just kill the cast completely

bad cast

Check correct header when looking at tape number.
from thomas@mathematik.uni-Bremen.de

new version from CSRG, via BSDI, with fixes

Add RCS identifiers.

Fix message.

restore didn't understand dirs which spanned more than one volume.
fixed by patch from Gene Stark (gene@stark.uucp).

initial import of 386bsd-0.1 sources

Initial revision