]> git.ipfire.org Git - thirdparty/mdadm.git/log
thirdparty/mdadm.git
17 years agoClarify when update=super-minor happens automatcially
Neil Brown [Mon, 9 Oct 2006 21:06:08 +0000 (07:06 +1000)] 
Clarify when update=super-minor happens automatcially

.. in man page

17 years agoFix typo in earlier patch.
Neil Brown [Mon, 9 Oct 2006 20:30:07 +0000 (06:30 +1000)] 
Fix typo in earlier patch.

Thanks Martin Krafft

17 years agoFixed endian problem with bitmap metdata
Paul Clements [Mon, 9 Oct 2006 01:17:22 +0000 (11:17 +1000)] 
Fixed endian problem with bitmap metdata

From: Paul Clements <paul.clements@steeleye.com>

17 years agoFix some endian-ness issues with v1 superblocks.
Neil Brown [Mon, 9 Oct 2006 01:17:12 +0000 (11:17 +1000)] 
Fix some endian-ness issues with v1 superblocks.

17 years agoMake --examine report chunk size when it should
Neil Brown [Mon, 9 Oct 2006 01:16:56 +0000 (11:16 +1000)] 
Make --examine report chunk size when it should

Currently it doesn't for raid6 and raid10

17 years agoMinor man page and comment fixes
Neil Brown [Mon, 9 Oct 2006 01:16:33 +0000 (11:16 +1000)] 
Minor man page and comment fixes

Thanks To: "Scott Weikart" <Scott.W@Benetech.org>

17 years agoAllow symlink creation to be disabled from command line or mdadm.conf
Neil Brown [Fri, 11 Aug 2006 08:00:05 +0000 (18:00 +1000)] 
Allow symlink creation to be disabled from command line or mdadm.conf

17 years agoCreate symlinks from /dev into /dev/md
Neil Brown [Fri, 11 Aug 2006 08:00:00 +0000 (18:00 +1000)] 
Create symlinks from /dev into /dev/md

When (auto)creating devices in /dev/md/, create
an appropriate symlink from /dev to avoid confusion.

17 years agoMinor improvement in name auto-chosen for version-1 superblocks.
Neil Brown [Fri, 11 Aug 2006 07:59:57 +0000 (17:59 +1000)] 
Minor improvement in name auto-chosen for version-1 superblocks.

/dev/md_d0  should have name 'd0' by default to
make /dev/md/d0.

17 years agoRemove indeterminism from names in --examine --brief output.
Neil Brown [Fri, 11 Aug 2006 07:59:53 +0000 (17:59 +1000)] 
Remove indeterminism from names in --examine --brief output.

It is much better for this sort of thing to be predictable
rather than depending on what devices exist, especially as
these days entries in /dev/ often don't pre-exist.
So make it always
   /dev/mdX

for version0 superblocks.

Version1 are always /dev/md/NAME

17 years agoRelease 2.5.3 mdadm-2.5.3
Neil Brown [Mon, 7 Aug 2006 02:00:30 +0000 (12:00 +1000)] 
Release 2.5.3

17 years agoTidyup automatic name choice for v-1 arrays
Neil Brown [Mon, 7 Aug 2006 01:30:02 +0000 (11:30 +1000)] 
Tidyup automatic name choice for v-1 arrays

 /dev/md_d0 now becomes '0', not '_d0'.

17 years agoFix starting of degraded arrays.
Neil Brown [Mon, 7 Aug 2006 01:13:01 +0000 (11:13 +1000)] 
Fix starting of degraded arrays.

Recent change broke assembling of degraded arrays, making
it require --run.  This fixes that.

17 years agoExplicitly state GPL license in man pages.
Neil Brown [Mon, 7 Aug 2006 01:12:25 +0000 (11:12 +1000)] 
Explicitly state GPL license in man pages.

17 years agoDocument usage of version0.91 superblock.
Neil Brown [Mon, 7 Aug 2006 01:11:42 +0000 (11:11 +1000)] 
Document usage of version0.91 superblock.

17 years agoRelease 2.5.2 mdadm-2.5.2
Neil Brown [Tue, 27 Jun 2006 11:31:58 +0000 (21:31 +1000)] 
Release 2.5.2

Description...

17 years agoMore consistent honoring of --configfile
Neil Brown [Mon, 26 Jun 2006 05:11:01 +0000 (15:11 +1000)] 
More consistent honoring of --configfile

 Never use /etc/mdadm.conf if --config file is given (previously
some code used one, some used the other).

17 years agoFix user-after-free bug in error path in --monitor mode.
Neil Brown [Mon, 26 Jun 2006 05:10:55 +0000 (15:10 +1000)] 
Fix user-after-free bug in error path in --monitor mode.

17 years agoImprove ftw handling.
Neil Brown [Mon, 26 Jun 2006 05:10:48 +0000 (15:10 +1000)] 
Improve ftw handling.

If not 'ftw' is available, still allow openning of devices by dev number.

More recent version of uclibc support nftw, so add support to check
for that.

17 years agoFix a recently introduced bug, and make --assemble more resilient to it.
Neil Brown [Mon, 26 Jun 2006 02:26:12 +0000 (12:26 +1000)] 
Fix a recently introduced bug, and make --assemble more resilient to it.

Make -assemble a bit more resilient to finding strange
 information in superblocks.
Don't claim newly added spares are InSync!! (don't know why that
 code was ever in there)

17 years agoWork around bug in --add handling for version-1 superblocks
Neil Brown [Mon, 26 Jun 2006 02:26:09 +0000 (12:26 +1000)] 
Work around bug in --add handling for  version-1 superblocks

In 2.6.17 (and prior), the dev_number is ignored when a device
is added to an active array.  Rather the first free number is used.
So we work around this by making sure we use the first free
number for dev_number.

Description...

17 years agoAdd 'Array Slot' line to --examine for version-1 superblocks
Neil Brown [Mon, 26 Jun 2006 02:26:01 +0000 (12:26 +1000)] 
Add 'Array Slot' line to --examine for version-1 superblocks

to make it a bit easier to see what is happening.

17 years agoFix manpage typo
martin f krafft [Mon, 26 Jun 2006 02:07:12 +0000 (12:07 +1000)] 
Fix manpage typo

From: martin f krafft <madduck@madduck.net>

17 years agoGet the changelog uptodate for recent changes.
Neil Brown [Mon, 26 Jun 2006 02:06:01 +0000 (12:06 +1000)] 
Get the changelog uptodate for recent changes.

17 years agoMove a variable declaration to the declaration area.
Paul Clements [Tue, 20 Jun 2006 00:01:47 +0000 (10:01 +1000)] 
Move a variable declaration to the declaration area.

While declaring variables in the middle of code withs with
newer gcc's it doesn't work with older, and it is arguably
less readable, so just do the right thing.

From: Paul Clements <paul.clements@steeleye.com>

17 years agoStop map_dev from returning [0:0]
Neil Brown [Tue, 20 Jun 2006 00:01:43 +0000 (10:01 +1000)] 
Stop map_dev from returning [0:0]

We sometimes need the NULL when major==minor==0.
So make sure all callers of map_dev can cope with NULL.

17 years agoFix problem with post-increment usage in macro
Paul Clements [Tue, 20 Jun 2006 00:01:23 +0000 (10:01 +1000)] 
Fix problem with post-increment usage in macro

Bad/bad/bad, and cause compiler error on ppc (gcc 3.2.3).

From: Paul Clements <paul.clements@steeleye.com>

17 years agoRelease 2.5.1 mdadm-2.5.1
Neil Brown [Fri, 16 Jun 2006 00:54:14 +0000 (10:54 +1000)] 
Release 2.5.1

17 years agoUpdate Changelog for 2.5.1
Neil Brown [Fri, 16 Jun 2006 00:53:55 +0000 (10:53 +1000)] 
Update Changelog for 2.5.1

17 years agoFix offsetof macro for 64bit hosts
Neil Brown [Fri, 16 Jun 2006 00:53:50 +0000 (10:53 +1000)] 
Fix offsetof macro for 64bit hosts

17 years agoFix memory leak in monitor mode
Neil Brown [Fri, 16 Jun 2006 00:26:12 +0000 (10:26 +1000)] 
Fix memory leak in monitor mode

When rescanning /dev, we didn't free the old list.
Also don't search for device with a number of 0,0

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoFix up some internalbitmap tests
Neil Brown [Fri, 16 Jun 2006 00:21:11 +0000 (10:21 +1000)] 
Fix up some internalbitmap tests

17 years agoAllow CONFFILE2 to be overridden during make invocation. origin
martin f. krafft [Fri, 2 Jun 2006 10:53:02 +0000 (20:53 +1000)] 
Allow CONFFILE2 to be overridden during make invocation.

17 years agoCPPFLAGS gets used automatically on implicit rules.
martin f krafft [Fri, 2 Jun 2006 08:31:48 +0000 (10:31 +0200)] 
CPPFLAGS gets used automatically on implicit rules.

Take it out for cosmetic reasons to prevent duplicate -DDEBIAN.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAvoid misdetection of overlapping partitions...
Neil Brown [Fri, 2 Jun 2006 06:58:39 +0000 (06:58 +0000)] 
Avoid misdetection of overlapping partitions...

As version-0.90 superblock don't record the superblock
offset, it is possible for overlapping partitions,
or a partition that starts on a 64K boundary in the whole device
to result in mis-detection - one partition or device might
be detected where the other was intended.

To avoid this awkward possibility, we reject assembly attempts
which seem to have two devices that are different but have the
same version-0.90 superblock.

To avoid this problem altogether, switch to version-1 metadata.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoRemove libssl dependancy and always use sha1.c code
Neil Brown [Fri, 2 Jun 2006 06:12:19 +0000 (06:12 +0000)] 
Remove libssl dependancy and always use sha1.c code

Apparently there are license issues with openssl, so
just use sha1.c always.   This means we can get rid of
SHA1.c

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoGet --stop to honour --quiet
Neil Brown [Fri, 2 Jun 2006 05:53:22 +0000 (05:53 +0000)] 
Get --stop to honour --quiet

And as a side effect, if --quiet isn't given, stopped devices are
reported.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoYet another include file that someone doesn't like....
Neil Brown [Fri, 2 Jun 2006 05:37:47 +0000 (05:37 +0000)] 
Yet another include file that someone doesn't like....

It seems byteswap.h isn't universally accepted, so
we do it by hand...

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAllow default metadata to be specified in mdadm.conf
Neil Brown [Fri, 2 Jun 2006 05:33:40 +0000 (05:33 +0000)] 
Allow default metadata to be specified in mdadm.conf

CREATE metadata=1
in mdadm.conf will cause version-1 superblocks to be the default.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoDon't try to create a new device when using --manage or --grow
Neil Brown [Fri, 2 Jun 2006 01:53:13 +0000 (01:53 +0000)] 
Don't try to create a new device when using --manage or --grow

... as that's just silly!

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoFinally (I hope) get byteorder stuff portable.
Neil Brown [Fri, 2 Jun 2006 01:49:00 +0000 (01:49 +0000)] 
Finally (I hope) get byteorder stuff portable.

/usr/include/endian.h and /usr/include/byteswap.h together
can give us almost what we need...

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoFix typo in mail message from --monitor
Neil Brown [Fri, 2 Jun 2006 01:34:30 +0000 (01:34 +0000)] 
Fix typo in mail message from --monitor

From: martin f krafft <madduck@madduck.net>

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoFix error code from examine properly..
Neil Brown [Fri, 2 Jun 2006 01:32:45 +0000 (01:32 +0000)] 
Fix error code from examine properly..

The last one was wrong :-(

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoImprove compiling for static binaries.
Neil Brown [Mon, 29 May 2006 04:09:21 +0000 (04:09 +0000)] 
Improve compiling for static binaries.

Have "#ifdef STATIC" in config.c, and the 'rmconf' target
was a mess.
Instead, create 'pwgr.c' with stub routines for those unavailable
when statically compiled, and include that in STATICOBJ

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoMake add_dev for uclibc return something.
Neil Brown [Mon, 29 May 2006 04:02:15 +0000 (04:02 +0000)] 
Make add_dev for uclibc return something.

As it is now 'int', it needs to.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAdd uninstall target to Makefile
Neil Brown [Mon, 29 May 2006 02:18:22 +0000 (02:18 +0000)] 
Add uninstall target to Makefile

From: "Dirk Jagdmann" <jagdmann@gmail.com>

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoNew install-* targets for installing differently compiled mdadms
Neil Brown [Mon, 29 May 2006 02:18:22 +0000 (02:18 +0000)] 
New install-* targets for installing differently compiled mdadms

e.g. install-static to install the statically compiled version, etc.

From: "Dirk Jagdmann" <jagdmann@gmail.com>
Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoMakefile improvements for static linking.
Neil Brown [Mon, 29 May 2006 02:18:22 +0000 (02:18 +0000)] 
Makefile improvements for static linking.

From: "Dirk Jagdmann" <jagdmann@gmail.com>
Signed-off-by: Neil Brown <neilb@suse.de>
17 years agocheck return status of all write/fwrite functions as required by glibc 2.4
Neil Brown [Mon, 29 May 2006 02:06:32 +0000 (02:06 +0000)] 
check return status of all write/fwrite functions as required by glibc 2.4

From: Luca Berra <bluca@vodka.it>

glibc 2.4 is pedantic on ignoring return values from fprintf, fwrite and
write, so now we check the rval and actually do something with it.
in the Grow.c case i only print a warning, since i don't think we can do
anything in case we fail invalidating those superblocks (is should never
happen, but then...)

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoUclibc has deprecated 'random' so use 'rand' instead.
Neil Brown [Mon, 29 May 2006 02:06:32 +0000 (02:06 +0000)] 
Uclibc has deprecated 'random' so use 'rand' instead.

glibc says of rand

       Do not use this function in applications  intended  to  be
       portable when good randomness is needed.

Go figure...

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoadd CFLAGS to mdassemble build and fix a couple of non-returning functions
Neil Brown [Mon, 29 May 2006 02:06:32 +0000 (02:06 +0000)] 
add CFLAGS to mdassemble build and fix a couple of non-returning functions

pass CFLAGS to mdassemble build, enabling -Wall -Werror showed some
issues also fixed by the patch.

From: Luca Berra <bluca@vodka.it>
Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoThis is to avoid gcc warnings when building with strict-aliasing optimization
Neil Brown [Mon, 29 May 2006 02:06:32 +0000 (02:06 +0000)] 
This is to avoid gcc warnings when building with strict-aliasing optimization

fix for another srict-aliasing problem, you can typecast a reference to a
void pointer to anything, you cannot typecast a reference to a
struct.

From: Luca Berra <bluca@vodka.it>
Signed-off-by: Neil Brown <neilb@suse.de>
17 years agosnprintf size should be at most the size of the buffer
Neil Brown [Mon, 29 May 2006 02:06:32 +0000 (02:06 +0000)] 
snprintf size should be at most the size of the buffer

From: Luca Berra <bluca@vodka.it>
Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoRelease 2.5 mdadm-2.5
Neil Brown [Fri, 26 May 2006 07:12:48 +0000 (07:12 +0000)] 
Release 2.5

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoMake sure everything compiles...
Neil Brown [Fri, 26 May 2006 04:28:33 +0000 (04:28 +0000)] 
Make sure everything compiles...

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoSupport --auto-update-homehost
Neil Brown [Fri, 26 May 2006 04:16:07 +0000 (04:16 +0000)] 
Support --auto-update-homehost

This can be used to bootstrape homehost tagging.

If no arrays are found that are tagged, we look for any array
and tag it.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoSet default name for v1 array based on device name.
Neil Brown [Fri, 26 May 2006 03:24:02 +0000 (03:24 +0000)] 
Set default name for v1 array based on device name.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoSupport auto-assembling of stacked devices
Neil Brown [Fri, 26 May 2006 03:11:57 +0000 (03:11 +0000)] 
Support auto-assembling of stacked devices

and assorted bugfixes.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoNever remove device nodes of active devices.
Neil Brown [Fri, 26 May 2006 02:32:15 +0000 (02:32 +0000)] 
Never remove device nodes of active devices.

This is just too risky.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoImprove names reported by --examine --brief
Neil Brown [Fri, 26 May 2006 02:29:12 +0000 (02:29 +0000)] 
Improve names reported by --examine --brief

Instead of depending (too much) on what is in /dev,
we make names based on the content of the superblock.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoChoose better devnumbers and tidy up some issues with finding names.
Neil Brown [Fri, 26 May 2006 01:40:53 +0000 (01:40 +0000)] 
Choose better devnumbers and tidy up some issues with finding names.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoInitial implementation of auto-assembly
Neil Brown [Fri, 26 May 2006 00:50:15 +0000 (00:50 +0000)] 
Initial implementation of auto-assembly

This basically works, but needs various improvements and some tests.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAllow autoassembly to choose it's own name for the array.
Neil Brown [Tue, 23 May 2006 06:46:40 +0000 (06:46 +0000)] 
Allow autoassembly to choose it's own name for the array.

This cannot be used yet, but it is working towards auto-assembly.

When auto-assembling an array, we make a name in /dev/md/
giving a number (from the peferred minor) or name (from set-name).

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoWhen assembly arrays using incomplete detail, prefer arrays built for this host.
Neil Brown [Tue, 23 May 2006 06:33:28 +0000 (06:33 +0000)] 
When assembly arrays using incomplete detail, prefer arrays built for this host.

i.e. if assembling with --name or --super-minor, then if we find two
different arrays with the same apparent identity, and one was built
for 'this' host, then prefer that one instead of giving up in disgust.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoReorganise Assemble code somewhat.
Neil Brown [Tue, 23 May 2006 06:12:40 +0000 (06:12 +0000)] 
Reorganise Assemble code somewhat.

We make sure all devices can are consistent before doing any --update
This saves us from updating some but not all of an array, and then
aborting.

It also means we can backtrack on out decisions, which we might want to
do later.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoMake sure name matching for array assembly understand the possibility of a homehost
Neil Brown [Tue, 23 May 2006 05:24:22 +0000 (05:24 +0000)] 
Make sure name matching for array assembly understand the possibility of a homehost

This "--assemble --name=foo" on host 'bar' will assemble an array named
'foo' or 'bar:foo'.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoSupport --update=homehost to allow updating of homehost information.
Neil Brown [Tue, 23 May 2006 05:16:32 +0000 (05:16 +0000)] 
Support --update=homehost to allow updating of homehost information.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAllow --update=name to update the name during assembly.
Neil Brown [Tue, 23 May 2006 04:57:04 +0000 (04:57 +0000)] 
Allow --update=name to update the name during assembly.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoMake sure homehost is set correctly when --update=uuid
Neil Brown [Tue, 23 May 2006 04:51:39 +0000 (04:51 +0000)] 
Make sure homehost is set correctly when --update=uuid

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoInclude homehost information in --examine as appropriate
Neil Brown [Fri, 19 May 2006 07:19:04 +0000 (07:19 +0000)] 
Include homehost information in --examine as appropriate

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoInclude homehost information in --detail where appropriate.
Neil Brown [Fri, 19 May 2006 07:13:03 +0000 (07:13 +0000)] 
Include homehost information in --detail where appropriate.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoMake homehost information appear in superblock.
Neil Brown [Fri, 19 May 2006 06:56:06 +0000 (06:56 +0000)] 
Make homehost information appear in superblock.

When an array is created, if the homehost is know,
the superblock gets it, either in the uuid, (via sha1)
or in the name field.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAllow homehost to be set on command line or in config file
Neil Brown [Fri, 19 May 2006 05:50:08 +0000 (05:50 +0000)] 
Allow homehost to be set on command line or in config file

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoImprove code for recognising config lines.
Neil Brown [Fri, 19 May 2006 05:33:05 +0000 (05:33 +0000)] 
Improve code for recognising config lines.

No more magic numbers.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoJust updaqte copyright dates and email address
Neil Brown [Fri, 19 May 2006 05:25:11 +0000 (05:25 +0000)] 
Just updaqte copyright dates and email address

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoNew flag --no-degraded
Neil Brown [Fri, 19 May 2006 04:53:22 +0000 (04:53 +0000)] 
New flag --no-degraded

Use to avoid starting arrays if there are
fewer devices available than last time the array was started.
This is only needed with --scan, as with --scan, that behaviour
is the default.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAdd a copy of /proc/mdstat to the mail message send by mdadm --monitor.
Neil Brown [Fri, 19 May 2006 03:58:45 +0000 (03:58 +0000)] 
Add a copy of /proc/mdstat to the mail message send by mdadm --monitor.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoFix incorrect error status from --examine --scan
Neil Brown [Fri, 19 May 2006 03:48:49 +0000 (03:48 +0000)] 
Fix incorrect error status from --examine --scan

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoArrange the 'auto' setting in mdadm.conf can choose default type.
Neil Brown [Tue, 16 May 2006 07:35:06 +0000 (07:35 +0000)] 
Arrange the 'auto' setting in mdadm.conf can choose default type.

So when you say auto=md or auto=part in mdadm.conf, it give a preference
for type of array, but standard name will override.

But --auto=md is more insistant.

FIXME  I'm not at all happy about handling of names that already exist.
  I don't think that should be removed if the device is active.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoWhen updating uuid, update the bitmap as well - external bitmaps
Neil Brown [Tue, 16 May 2006 06:05:27 +0000 (06:05 +0000)] 
When updating uuid, update the bitmap as well - external bitmaps

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoFix a couple of typos in the man page.
Neil Brown [Tue, 16 May 2006 05:53:15 +0000 (05:53 +0000)] 
Fix a couple of typos in the man page.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoWhen updating uuid, update the bitmap as well - internal bitmaps.
Neil Brown [Tue, 16 May 2006 05:52:51 +0000 (05:52 +0000)] 
When updating uuid, update the bitmap as well - internal bitmaps.

Otherwise when you "--update==uuid" an array with an internal
bitmap, it will break, badly.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoDetail checks if array has been started or not and includes that in report.
Neil Brown [Mon, 15 May 2006 06:03:05 +0000 (06:03 +0000)] 
Detail checks if array has been started or not and includes that in report.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAllow default creation info to to be stored in mdadm.conf
Neil Brown [Mon, 15 May 2006 06:02:41 +0000 (06:02 +0000)] 
Allow default creation info to to be stored in mdadm.conf

Default owner, group, mode and 'auto' flag can be given in a 'CREATE' line.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoLimit size of bitmap to 2million chunks.
Neil Brown [Mon, 15 May 2006 04:21:33 +0000 (04:21 +0000)] 
Limit size of bitmap to 2million chunks.

When creating a file bitmap, choose a default size that
results in fewer than 2^21 chunks.  Without this kmalloc
failure in the kernel becomes likely.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoFix problem with sector/KB size confuision for bitmap sizing.
Neil Brown [Mon, 15 May 2006 03:57:53 +0000 (03:57 +0000)] 
Fix problem with sector/KB size confuision for bitmap sizing.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoSupport new offset layout for raid10
Neil Brown [Mon, 15 May 2006 02:46:54 +0000 (02:46 +0000)] 
Support new offset layout for raid10

Requires 2.6.18.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoAssume "DEVICE partitions" if no DEVICE line present.
Neil Brown [Mon, 15 May 2006 01:58:47 +0000 (01:58 +0000)] 
Assume "DEVICE partitions" if no DEVICE line present.

This means the output of "mdadm -Es" can be used as a complete
mdadm.conf file .... if you really want to do that.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoArrange that SparesMissing events generate an email too.
Neil Brown [Mon, 15 May 2006 01:35:47 +0000 (01:35 +0000)] 
Arrange that SparesMissing events generate an email too.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoSupport 'mailfrom' line in mdadm.conf so the From: line in alert emails can be explic...
Neil Brown [Mon, 15 May 2006 01:27:04 +0000 (01:27 +0000)] 
Support 'mailfrom' line in mdadm.conf so the From: line in alert emails can be explicitly set.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years agoUse O_DIRECT to read bitmap files.
Neil Brown [Mon, 15 May 2006 00:56:02 +0000 (00:56 +0000)] 
Use O_DIRECT to read bitmap files.

A pending patch to the kernel causes bitmap file updates
to not go through the page cache, so O_DIRECT is needed to
ensure that we read current data.

Signed-off-by: Neil Brown <neilb@suse.de>
17 years ago[PATCH] Make sure update_super returns correct value.
Neil Brown [Fri, 28 Apr 2006 04:10:38 +0000 (14:10 +1000)] 
[PATCH] Make sure update_super returns correct value.

For 'force' and 'assemble', update_super must return true
if anything was changed.

Also fix a bug with wonly handling in super0.

Signed-off-by: Neil Brown <neilb@suse.de>
### Diffstat output
 ./super0.c |   18 ++++++++++++++----
 ./super1.c |    7 +++++++
 2 files changed, 21 insertions(+), 4 deletions(-)

diff ./super0.c~current~ ./super0.c

17 years ago[PATCH] Get the name for partitioned devices in sysfs correct.
Neil Brown [Fri, 28 Apr 2006 04:09:24 +0000 (14:09 +1000)] 
[PATCH] Get the name for partitioned devices in sysfs correct.

There are 64 partitions per device, not 16.

Signed-off-by: Neil Brown <neilb@suse.de>
### Diffstat output
 ./sysfs.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff ./sysfs.c~current~ ./sysfs.c

18 years agoRelease 2.4.1 mdadm-2.4.1
Neil Brown [Mon, 17 Apr 2006 12:49:28 +0000 (12:49 +0000)] 
Release 2.4.1

Annoying alignment bug in version-1 supers on 64bit..

Signed-off-by: Neil Brown <neilb@suse.de>
18 years agoFix alignment problem in version-1 superblocks.
Neil Brown [Fri, 7 Apr 2006 04:34:45 +0000 (04:34 +0000)] 
Fix alignment problem in version-1 superblocks.

NOTE: This is an incompatable change affecting raid5 reshape.
If you want to reshape a raid5 using version-1 superblocks,
use 2.6.17-rc2 or later, and mdadm-2.4.1 or later.

Signed-off-by: Neil Brown <neilb@suse.de>
18 years agoAdd tests/05r1-grow-internal-1
Neil Brown [Sat, 1 Apr 2006 09:28:36 +0000 (09:28 +0000)] 
Add tests/05r1-grow-internal-1

18 years agofix write mostly for add and re-add
Neil Brown [Mon, 3 Apr 2006 00:11:51 +0000 (00:11 +0000)] 
fix write mostly for add and re-add

The following patch makes it possible to tag a device as write-mostly on
--add and --re-add with a non-persistent superblock array. Previously,
this was not working.

Signed-Off-By: Paul Clements <paul.clements@steeleye.com>
Signed-off-by: Neil Brown <neilb@suse.de>
18 years agoAdd ANNOUNCE-2.4 mdadm-2.4
Neil Brown [Thu, 30 Mar 2006 04:24:09 +0000 (04:24 +0000)] 
Add ANNOUNCE-2.4

18 years agoRelease 2.4
Neil Brown [Mon, 3 Apr 2006 00:09:16 +0000 (00:09 +0000)] 
Release 2.4

Signed-off-by: Neil Brown <neilb@suse.de>
18 years agoAdd tests/05r1-grow-internal-1
Neil Brown [Wed, 29 Mar 2006 05:21:16 +0000 (05:21 +0000)] 
Add tests/05r1-grow-internal-1