Skip to contentSkip to conversation

grub-0.97: btrfs multidevice support [PATCH]

PATCH Edward Shishkin 11 participants 18 messages Sep 24, 2009 – Oct 2, 2009 Btrfs file system

lore
Share & tools
lore
#0 lore raw

Hello everyone.

Please, find the patch for Fedora 10 in the attachment(**). The distro-independent package will be put to kernel.org a bit later.

I. Loading kernels from btrfs volumes

Now you can load kernels and initrds from btrfs volumes composed of many devices.

WARNING!!!
Make sure that all components of your loading btrfs volume(*) are visible to grub. Otherwise, you'll end with unbootable system. The list of available grub devices can be obtained, for example, using tab completion in grub shell.

Number of components of a loading volume is not restricted, however if it is larger then 128, then the boot process will be slowed down because of expensive translations (btrfs-device-id -> grub-device-id) which issue a large number of IOs. We cache only 128 such translations in grub-0.97 because of high memory pressure.

II. Installing grub from btrfs volumes

You can install grub from a btrfs image volume(*) composed of many devices (see above about restrictions). Also you can setup any component of a btrfs boot(*) volume as grub root device.

NOTE!!! Make sure that all components of image and boot volumes(*) are visible to grub, otherwise grub installer will return error.

TECHNICAL NOTE (for grub developers):
The unpleasant surprise was that grub installer overwrites (by default!) the file (stage2), bypassing the file system driver. I can not understand this: it looks like stepping to the clean water with dirty shoe. Hope that grub2 won't afford such things.

In order to install grub from a btrfs image volume use special option (--stage2). This option makes grub installer to rewrite the file with a help of the OS's file system (i.e, via write (2)). Any attempts to install without this option will fail with an error (wrong argument).

The example of possible installation scenario.

Suppose image volume = root volume = loading volume is composed of devices (hd0,4), (hd0,5), (hd1,5), (hd1,7) and is not an OS's root. We want to setup (hd0,4) as grub root device and install grub to the mbr of (hd0).

. build and install grub with btrfs support;
. mount your the "3 in 1" btrfs volume to /mnt;
. create a directory /mnt/grub;
. put the built files stage1, stage2, btrfs_stage1_5, grub.conf,
etc. to /mnt/grub;
. run grub shell;
. grub> root (hd0,4)
. grub> setup --stage2=/mnt/grub/stage2 (hd0)
. have a fun.

Use info(1) grub for more details.

(*) Glossary:
. loading volume:
a btrfs volume that contains kernel image and initrd;

. image volume:
a btrfs volume that contains stage1, stage2, btrfs_stage_1_5, and grub.conf files needed for grub installer;

. boot volume:
a btrfs volume where grub will look for stage2 and grub.conf files in boot time.

(**) Link to the Fedora's grub package:
http://ucho.ignum.cz/fedora/linux/releases/10/Fedora/source/SRPMS/grub-0.97-38.fc10.src.rpm

All comments, bugreports, etc. are welcome as usual.

Thanks,
Edward.

---
 AUTHORS             |    2 
 INSTALL             |    3 
 configure.in        |    7 
 docs/grub.texi      |    3 
 grub/Makefile.am    |    2 
 stage2/Makefile.am  |   38 -
 stage2/btrfs.h      | 1414 ++++++++++++++++++++++++++++++++++++++++
 stage2/builtins.c   |   11 
 stage2/disk_io.c    |    3 
 stage2/filesys.h    |   14 
 stage2/fsys_btrfs.c | 1815 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 stage2/shared.h     |   13 
 12 files changed, 3303 insertions(+), 22 deletions(-)
AUTHORS +2 -0
@@ -41,6 +41,8 @@ Kristoffer Branemyr added VSTa filesyste
 
 Serguei Tzukanov added JFS and XFS support.
 
+Edward Shishkin added Btrfs support.
+
 Jason Thomas added Linux DAC960 support and support for hiding/unhiding
 logical partitions, and did a significant bugfix for the terminal stuff.
 
INSTALL +3 -0
@@ -207,6 +207,9 @@ operates.
 `--disable-reiserfs'
      Omit the ReiserFS support in Stage 2.
 
+`--disable-btrfs'
+     Omit the BtrFS support in Stage 2.
+
 `--disable-vstafs'
      Omit the VSTa filesystem support in Stage 2.
 
configure.in +7 -0
@@ -337,6 +337,13 @@ if test x"$enable_reiserfs" != xno; then
   FSYS_CFLAGS="$FSYS_CFLAGS -DFSYS_REISERFS=1"
 fi
 
+AC_ARG_ENABLE(btrfs,
+  [  --disable-btrfs         disable BtrFS support in Stage 2])
+
+if test x"$enable_btrfs" != xno; then
+  FSYS_CFLAGS="$FSYS_CFLAGS -DFSYS_BTRFS=1"
+fi
+
 AC_ARG_ENABLE(vstafs,
   [  --disable-vstafs        disable VSTa FS support in Stage 2])
 
docs/grub.texi +2 -1
@@ -283,7 +283,7 @@ devices, partitions, and files in a dire
 Support multiple filesystem types transparently, plus a useful explicit
 blocklist notation. The currently supported filesystem types are
 @dfn{BSD FFS}, @dfn{DOS FAT16 and FAT32}, @dfn{Minix fs}, @dfn{Linux
-ext2fs}, @dfn{ReiserFS}, @dfn{JFS}, @dfn{XFS}, and @dfn{VSTa
+ext2fs}, @dfn{ReiserFS}, @dfn{BtrFS}, @dfn{JFS}, @dfn{XFS}, and @dfn{VSTa
 fs}. @xref{Filesystem}, for more information.
 
 @item Support automatic decompression
@@ -1776,6 +1776,7 @@ itself. Usually, this is put in a filesy
 @itemx jfs_stage1_5
 @itemx minix_stage1_5
 @itemx reiserfs_stage1_5
+@itemx btrfs_stage1_5
 @itemx vstafs_stage1_5
 @itemx xfs_stage1_5
 
grub/Makefile.am +1 -1
@@ -8,7 +8,7 @@ endif
 
 AM_CPPFLAGS = -DGRUB_UTIL=1 -DFSYS_EXT2FS=1 -DFSYS_FAT=1 -DFSYS_FFS=1 \
 	-DFSYS_ISO9660=1 -DFSYS_JFS=1 -DFSYS_MINIX=1 -DFSYS_REISERFS=1 \
-	-DFSYS_UFS2=1 -DFSYS_VSTAFS=1 -DFSYS_XFS=1 \
+	-DFSYS_BTRFS=1 -DFSYS_UFS2=1 -DFSYS_VSTAFS=1 -DFSYS_XFS=1 \
 	-DUSE_MD5_PASSWORDS=1 -DSUPPORT_HERCULES=1 \
 	$(SERIAL_FLAGS) -I$(top_srcdir)/stage2 \
 	-I$(top_srcdir)/stage1 -I$(top_srcdir)/lib
stage2/Makefile.am +25 -13
@@ -21,13 +21,13 @@ noinst_LIBRARIES = libgrub.a
 endif
 libgrub_a_SOURCES = boot.c builtins.c char_io.c cmdline.c common.c \
 	disk_io.c fsys_ext2fs.c fsys_fat.c fsys_ffs.c fsys_iso9660.c \
-	fsys_jfs.c fsys_minix.c fsys_reiserfs.c fsys_ufs2.c \
+	fsys_jfs.c fsys_minix.c fsys_reiserfs.c fsys_btrfs.c fsys_ufs2.c \
 	fsys_vstafs.c fsys_xfs.c gunzip.c md5.c serial.c stage2.c \
 	terminfo.c tparm.c graphics.c
 libgrub_a_CFLAGS = $(GRUB_CFLAGS) -I$(top_srcdir)/lib \
 	-DGRUB_UTIL=1 -DFSYS_EXT2FS=1 -DFSYS_FAT=1 -DFSYS_FFS=1 \
 	-DFSYS_ISO9660=1 -DFSYS_JFS=1 -DFSYS_MINIX=1 -DFSYS_REISERFS=1 \
-	-DFSYS_UFS2=1 -DFSYS_VSTAFS=1 -DFSYS_XFS=1 \
+	-DFSYS_BTRFS=1 -DFSYS_UFS2=1 -DFSYS_VSTAFS=1 -DFSYS_XFS=1 \
 	-DUSE_MD5_PASSWORDS=1 -DSUPPORT_SERIAL=1 -DSUPPORT_HERCULES=1
 
 # Stage 2 and Stage 1.5's.
@@ -40,24 +40,26 @@ EXTRA_PROGRAMS = nbloader.exec pxeloader
 if DISKLESS_SUPPORT
 pkgdata_DATA = stage2 stage2_eltorito e2fs_stage1_5 fat_stage1_5 \
 	ffs_stage1_5 iso9660_stage1_5 jfs_stage1_5 minix_stage1_5 \
-	reiserfs_stage1_5 ufs2_stage1_5 vstafs_stage1_5 xfs_stage1_5 \
-	nbgrub pxegrub
+	reiserfs_stage1_5 btrfs_stage1_5 ufs2_stage1_5 vstafs_stage1_5 \
+	xfs_stage1_5 nbgrub pxegrub
 noinst_DATA = pre_stage2 start start_eltorito nbloader pxeloader diskless
 noinst_PROGRAMS = pre_stage2.exec start.exec start_eltorito.exec \
 	e2fs_stage1_5.exec fat_stage1_5.exec ffs_stage1_5.exec \
 	iso9660_stage1_5.exec jfs_stage1_5.exec minix_stage1_5.exec \
-	reiserfs_stage1_5.exec ufs2_stage1_5.exec vstafs_stage1_5.exec \
-	xfs_stage1_5.exec nbloader.exec pxeloader.exec diskless.exec
+	reiserfs_stage1_5.exec btrfs_stage1_5.exec ufs2_stage1_5.exec \
+	vstafs_stage1_5.exec xfs_stage1_5.exec nbloader.exec \
+	pxeloader.exec diskless.exec
 else
 pkgdata_DATA = stage2 stage2_eltorito e2fs_stage1_5 fat_stage1_5 \
 	ffs_stage1_5 iso9660_stage1_5 jfs_stage1_5 minix_stage1_5 \
-	reiserfs_stage1_5 ufs2_stage1_5 vstafs_stage1_5 xfs_stage1_5
+	reiserfs_stage1_5 btrfs_stage1_5 ufs2_stage1_5 vstafs_stage1_5 \
+        xfs_stage1_5
 noinst_DATA = pre_stage2 start start_eltorito
 noinst_PROGRAMS = pre_stage2.exec start.exec start_eltorito.exec \
 	e2fs_stage1_5.exec fat_stage1_5.exec ffs_stage1_5.exec \
 	iso9660_stage1_5.exec jfs_stage1_5.exec minix_stage1_5.exec \
-	reiserfs_stage1_5.exec ufs2_stage1_5.exec vstafs_stage1_5.exec \
-	xfs_stage1_5.exec
+	reiserfs_stage1_5.exec btrfs_stage1_5.exec ufs2_stage1_5.exec \
+        vstafs_stage1_5.exec xfs_stage1_5.exec
 endif
 MOSTLYCLEANFILES = $(noinst_PROGRAMS)
 
@@ -98,7 +100,7 @@ STAGE2_COMPILE = $(STAGE2_CFLAGS) -fno-b
 
 libstage2_a_SOURCES = boot.c builtins.c char_io.c cmdline.c common.c \
 	disk_io.c fsys_ext2fs.c fsys_fat.c fsys_ffs.c fsys_iso9660.c \
-	fsys_jfs.c fsys_minix.c fsys_reiserfs.c fsys_ufs2.c \
+	fsys_jfs.c fsys_minix.c fsys_reiserfs.c fsys_btrfs.c fsys_ufs2.c \
 	fsys_vstafs.c fsys_xfs.c gunzip.c md5.c serial.c stage2.c \
 	terminfo.c tparm.c
 libstage2_a_CFLAGS = $(STAGE2_COMPILE) $(FSYS_CFLAGS)
@@ -112,9 +114,9 @@ STAGE1_5_COMPILE = $(STAGE2_COMPILE) -DN
 pre_stage2_exec_SOURCES = asm.S bios.c boot.c builtins.c char_io.c \
 	cmdline.c common.c console.c disk_io.c fsys_ext2fs.c \
 	fsys_fat.c fsys_ffs.c fsys_iso9660.c fsys_jfs.c fsys_minix.c \
-	fsys_reiserfs.c fsys_ufs2.c fsys_vstafs.c fsys_xfs.c gunzip.c \
-	hercules.c md5.c serial.c smp-imps.c stage2.c terminfo.c tparm.c \
-	graphics.c
+	fsys_reiserfs.c fsys_btrfs.c fsys_ufs2.c fsys_vstafs.c fsys_xfs.c \
+	gunzip.c hercules.c md5.c serial.c smp-imps.c stage2.c terminfo.c \
+	tparm.c graphics.c
 pre_stage2_exec_CFLAGS = $(STAGE2_COMPILE) $(FSYS_CFLAGS)
 pre_stage2_exec_CCASFLAGS = $(STAGE2_COMPILE) $(FSYS_CFLAGS)
 pre_stage2_exec_LDFLAGS = $(PRE_STAGE2_LINK)
@@ -223,6 +225,16 @@ reiserfs_stage1_5_exec_CCASFLAGS = $(STA
 reiserfs_stage1_5_exec_LDFLAGS = $(STAGE1_5_LINK)
 reiserfs_stage1_5_exec_LDADD = @LIBGCC@
 
+# For btrfs_stage1_5 target.
+btrfs_stage1_5_exec_SOURCES = start.S asm.S common.c char_io.c \
+	disk_io.c stage1_5.c fsys_btrfs.c bios.c
+btrfs_stage1_5_exec_CFLAGS = $(STAGE1_5_COMPILE) -DFSYS_BTRFS=1 \
+	-DNO_BLOCK_FILES=1
+btrfs_stage1_5_exec_CCASFLAGS = $(STAGE1_5_COMPILE) -DFSYS_BTRFS=1 \
+	-DNO_BLOCK_FILES=1
+btrfs_stage1_5_exec_LDFLAGS = $(STAGE1_5_LINK)
+btrfs_stage1_5_exec_LDADD = @LIBGCC@
+
 # For vstafs_stage1_5 target.
 vstafs_stage1_5_exec_SOURCES = start.S asm.S common.c char_io.c \
 	disk_io.c stage1_5.c fsys_vstafs.c bios.c
stage2/btrfs.h new file +1414 -0

1,414 lines not on this page · raw message

stage2/builtins.c +11 -0
@@ -2453,6 +2453,16 @@ install_func (char *arg, int flags)
 	  else
 #endif /* GRUB_UTIL */
 	    {
+	      /*
+	       * FIXME: Ugly hack.
+	       * Do not write to btrfs partition
+	       * without a help of the file system!
+	       */
+	      if (!strcmp(fsys_table[fsys_type].name, "btrfs"))
+		{
+		  errnum = ERR_BAD_ARGUMENT;
+		  goto fail;
+		}
 	      if (! devwrite (*saved_sector - part_start, 1, stage2_buffer))
 		goto fail;
 	    }
@@ -4201,6 +4211,7 @@ setup_func (char *arg, int flags)
     {"jfs",      "/jfs_stage1_5"},
     {"minix",    "/minix_stage1_5"},
     {"reiserfs", "/reiserfs_stage1_5"},
+    {"btrfs",    "/btrfs_stage1_5"},
     {"vstafs",   "/vstafs_stage1_5"},
     {"xfs",      "/xfs_stage1_5"}
   };
stage2/disk_io.c +3 -0
@@ -64,6 +64,9 @@ struct fsys_entry fsys_table[NUM_FSYS + 
 # ifdef FSYS_REISERFS
   {"reiserfs", reiserfs_mount, reiserfs_read, reiserfs_dir, 0, reiserfs_embed},
 # endif
+# ifdef FSYS_BTRFS
+  {"btrfs", btrfs_mount, btrfs_read, btrfs_dir, 0, btrfs_embed},
+# endif
 # ifdef FSYS_VSTAFS
   {"vstafs", vstafs_mount, vstafs_read, vstafs_dir, 0, 0},
 # endif
stage2/filesys.h +12 -2
@@ -77,6 +77,16 @@ int reiserfs_embed (int *start_sector, i
 #define FSYS_REISERFS_NUM 0
 #endif
 
+#ifdef FSYS_BTRFS
+#define FSYS_BTRFS_NUM 1
+int btrfs_mount (void);
+int btrfs_read (char *buf, int len);
+int btrfs_dir (char *dirname);
+int btrfs_embed (int *start_sector, int needed_sectors);
+#else
+#define FSYS_BTRFS_NUM 0
+#endif
+
 #ifdef FSYS_VSTAFS
 #define FSYS_VSTAFS_NUM 1
 int vstafs_mount (void);
@@ -127,8 +137,8 @@ int iso9660_dir (char *dirname);
 #ifndef NUM_FSYS
 #define NUM_FSYS	\
   (FSYS_FFS_NUM + FSYS_FAT_NUM + FSYS_EXT2FS_NUM + FSYS_MINIX_NUM	\
-   + FSYS_REISERFS_NUM + FSYS_VSTAFS_NUM + FSYS_JFS_NUM + FSYS_XFS_NUM	\
-   + FSYS_TFTP_NUM + FSYS_ISO9660_NUM + FSYS_UFS2_NUM)
+   + FSYS_REISERFS_NUM + FSYS_BTRFS_NUM + FSYS_VSTAFS_NUM + FSYS_JFS_NUM \
+   + FSYS_XFS_NUM + FSYS_TFTP_NUM + FSYS_ISO9660_NUM + FSYS_UFS2_NUM)
 #endif
 
 /* defines for the block filesystem info area */
stage2/fsys_btrfs.c new file +1815 -0

1,391 lines not on this page · raw message

stage2/shared.h +8 -5

Lines of this file are not shown. View the raw message.

The rest of this diff is not shown. View the raw message.

_______________________________________________
Grub-devel mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/grub-devel

Part of the diff is not shown here. View the raw message.

#1 lore raw

Hi Edward,

I'm sorry but GRUB Legacy is not maintained. At least not by us; we've deprecated it in favour of GRUB 2.

It is also being abandoned by distributors, so I wouldn't recommend that you put any effort in developing for it.

Signature · 5 lines
Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."
#2 lore raw
Quoted 8 lines · Robert Millan

Hi Edward,

I'm sorry but GRUB Legacy is not maintained. At least not by us; we've deprecated it in favour of GRUB 2.

It is also being abandoned by distributors, so I wouldn't recommend that you put any effort in developing for it.

You've been spouting this line for years, and yet my Ubuntu 10.4 machine uses, guess what, GRUB 1. Mainly because you deprecated GRUB Legacy[tm] well before GRUB 2 was usable.

(you in the generic sense of the GRUB project here)

Edward - please do continue to develop patches for GRUB 1 (the one that still actually works plenty well enough for lots of people) and ignore the naysayers who are happy to throw out backwards compatibility.

Sometimes you have to maintain crappy code because people out there depend on it. And we thank those who step up and do it rather than throw their hands up and pretend it doesn't need doing!

Ta,

Bron.

#3 lore raw
Quoted 16 lines · Bron Gondwana
Quoted 8 lines · Robert Millan

Hi Edward,

I'm sorry but GRUB Legacy is not maintained. At least not by us; we've deprecated it in favour of GRUB 2.

It is also being abandoned by distributors, so I wouldn't recommend that you put any effort in developing for it.

You've been spouting this line for years, and yet my Ubuntu 10.4 machine uses, guess what, GRUB 1. Mainly because you deprecated GRUB Legacy[tm] well before GRUB 2 was usable.

(you in the generic sense of the GRUB project here)

Quoted 7 lines

Edward - please do continue to develop patches for GRUB 1 (the one that still actually works plenty well enough for lots of people) and ignore the naysayers who are happy to throw out backwards compatibility.

Sometimes you have to maintain crappy code because people out there depend on it. And we thank those who step up and do it rather than throw their hands up and pretend it doesn't need doing!

I'm not telling anyone what should or shouldn't work with. But if you want to work on a branch we consider deprecated, please do it elsewhere. Discussion regarding GRUB Legacy is off-topic in this list.

Thanks

Signature · 5 lines
Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."
#4 lore raw
Quoted 25 lines · Robert Millan
Quoted 16 lines · Bron Gondwana
Quoted 8 lines · Robert Millan

Hi Edward,

I'm sorry but GRUB Legacy is not maintained. At least not by us; we've deprecated it in favour of GRUB 2.

It is also being abandoned by distributors, so I wouldn't recommend that you put any effort in developing for it.

You've been spouting this line for years, and yet my Ubuntu 10.4 machine uses, guess what, GRUB 1. Mainly because you deprecated GRUB Legacy[tm] well before GRUB 2 was usable.

(you in the generic sense of the GRUB project here)

Quoted 7 lines

Edward - please do continue to develop patches for GRUB 1 (the one that still actually works plenty well enough for lots of people) and ignore the naysayers who are happy to throw out backwards compatibility.

Sometimes you have to maintain crappy code because people out there depend on it. And we thank those who step up and do it rather than throw their hands up and pretend it doesn't need doing!

Quoted 6 lines

I'm not telling anyone what should or shouldn't work with. But if you want to work on a branch we consider deprecated, please do it elsewhere. Discussion regarding GRUB Legacy is off-topic in this list.

Thanks

That would be rather unusual - we certainly discuss ext2 and ancient file systems in our ext development lists.

A lot of distros (and their users) still use grub, just like a lot of users still use ext2.

Working on grub2 is great, but until it meets the needs and wins over the distros, you will unfortunately still need to put up with contributions to the old code base.

Have you looked at what it would take you (or others) to move Edward's patches into grub2?

Regards,

Ric

#5 lore raw
Quoted 6 lines · Ric Wheeler

That would be rather unusual - we certainly discuss ext2 and ancient file systems in our ext development lists.

A lot of distros (and their users) still use grub, just like a lot of users still use ext2.

We support ext2 too. And even awfully old things like fat (which is often used with supposedly modern *cough* EFI).

Signature · 5 lines
Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."
#6 lore raw
In reply to Robert Millan
Quoted 10 lines · Robert Millan
Quoted 9 lines · Bron Gondwana
Quoted 5 lines · Robert Millan

I'm sorry but GRUB Legacy is not maintained. At least not by us; we've deprecated it in favour of GRUB 2.

It is also being abandoned by distributors, so I wouldn't recommend that you put any effort in developing for it.

You've been spouting this line for years, and yet my Ubuntu 10.4 machine uses, guess what, GRUB 1.

I assume you typoed, since there's no such thing as Ubuntu 10.4 yet. When there is (well, 10.04 anyway), it will use GRUB 2 by default.

Quoted 3 lines
Quoted 3 lines

Edward - please do continue to develop patches for GRUB 1 (the one that still actually works plenty well enough for lots of people) and ignore the naysayers who are happy to throw out backwards compatibility.

It would be great if somebody could take up Edward's work and port it to GRUB 2. If nobody else does then I'd be interested in doing so myself, although I will not be able to start for a month or two from now.

Robert is working hard on making GRUB 2 usable, and is just advising Edward that, right now, there is no upstream for GRUB Legacy who could either accept or usefully comment on his patch. It would of course be possible for some people (presumably mostly the distributors who rely on it) to get together and declare themselves the new upstream for GRUB Legacy, but most of the people who might be interested in such things seem to have either lost interest or thrown their weight behind GRUB 2 upstream. Certainly this distributor right here is in the latter camp as it seems much more likely to produce a result that meets our needs in the end. (Plus, I think such a revitalised upstream would be a caretaker at best, and wouldn't really be able to effectively work on some of the major issues that have dogged distributors of GRUB Legacy for years without reinventing the wheel of GRUB 2.)

This isn't naysaying those people who post patches for GRUB Legacy - but given the reality that nobody is maintaining GRUB Legacy upstream right now, which is better, to have your patch ignored or to receive a note saying that it's against an unmaintained target? I'd go for not being ignored any day.

Signature · 1 line
Colin Watson                                       [[email protected]]
#7 lore raw

Am Freitag, den 25.09.2009, 16:01 +0100 schrieb Colin Watson:

Quoted 11 lines · Robert Millan
Quoted 6 lines · Bron Gondwana
Quoted 5 lines · Robert Millan

Edward - please do continue to develop patches for GRUB 1 (the one that still actually works plenty well enough for lots of people) and ignore the naysayers who are happy to throw out backwards compatibility.

It would be great if somebody could take up Edward's work and port it to GRUB 2. If nobody else does then I'd be interested in doing so myself, although I will not be able to start for a month or two from now.

Whoever wants to do this, needs also solve the problem that st.stat_rdev never matches st.st_dev of the real device.
So `grub-probe -t device /' is broken with btrfs.

This seems to be wanted:
http://thread.gmane.org/gmane.comp.file-systems.btrfs/3609/focus=3610
But I haven't read the other mails in that thread.

Signature · 2 lines
Felix Zielcke
Proud Debian Maintainer
#8 lore raw
In reply to Colin Watson
Quoted 5 lines · Colin Watson

This isn't naysaying those people who post patches for GRUB Legacy - but given the reality that nobody is maintaining GRUB Legacy upstream right now, which is better, to have your patch ignored or to receive a note saying that it's against an unmaintained target? I'd go for not being ignored any day.

Thankfully grub2 is starting to see distro work. We did ocfs2
grub2 support years ago and ran into this problem. Do the work against grub and no one will take the patch. Do the work against grub2 and upstream will take the patch, but zero users see it because grub2 was unusable by anyone.
Here's the original ocfs2 work from Mark Fasheh:
http://osdir.com/ml/boot-loaders.grub.devel/2005-11/msg00025.html It's sat idle for four years because of the grub2 vs distros confusion. Hopefully we can get it upstream again!

Joel

Signature · 8 lines
"The only way to get rid of a temptation is to yield to it."
         - Oscar Wilde 

Joel Becker
Principal Software Developer
Oracle
E-mail: [email protected]
Phone: (650) 506-8127
#9 lore raw
In reply to Colin Watson
Quoted 6 lines · Colin Watson
Quoted 5 lines · Robert Millan
Quoted 4 lines · Bron Gondwana
Quoted 3 lines · Robert Millan

I'm sorry but GRUB Legacy is not maintained. At least not by us; we've deprecated it in favour of GRUB 2.

Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
Quoted 2 lines

It is also being abandoned by distributors, so I wouldn't recommend that you put any effort in developing for it.

You've been spouting this line for years, and yet my Ubuntu 10.4 machine uses, guess what, GRUB 1.

I assume you typoed, since there's no such thing as Ubuntu 10.4 yet. When there is (well, 10.04 anyway), it will use GRUB 2 by default.

Yeah, I meant 9.04 of course. Sorry. The one that's currently stable! I am running GRUB 2 on my 9.10 laptop, as well as my Debian "testing" box. It works fine, though /boot/grub is full of a LOT of files. Still, it's simpler all together.

Quoted 11 lines
Quoted 3 lines
Quoted 3 lines

Edward - please do continue to develop patches for GRUB 1 (the one that still actually works plenty well enough for lots of people) and ignore the naysayers who are happy to throw out backwards compatibility.

It would be great if somebody could take up Edward's work and port it to GRUB 2. If nobody else does then I'd be interested in doing so myself, although I will not be able to start for a month or two from now.

Robert is working hard on making GRUB 2 usable, and is just advising Edward that, right now, there is no upstream for GRUB Legacy who could either accept or usefully comment on his patch.

Just out of interest, how long has it been that there was no stable supported version of GRUB? My vague memory is that it's been about 6 years now since GRUB 1 was declared deprecated in favour of the nebulous future.

Quoted 5 lines

This isn't naysaying those people who post patches for GRUB Legacy - but given the reality that nobody is maintaining GRUB Legacy upstream right now, which is better, to have your patch ignored or to receive a note saying that it's against an unmaintained target? I'd go for not being ignored any day.

Well, yeah. True. There's no really good path out of here other than everyone moving to GRUB 2 before running any filesystem that's changed in the past few years as root.

Bron.

#10 lore raw
In reply to Colin Watson

On Fri, Sep 25, 2009 at 11:01 PM, Colin Watson <[email protected]> wr= ote:

Quoted 4 lines · Robert Millan
Quoted 3 lines · Bron Gondwana
Quoted 2 lines · Robert Millan
Quoted 1 line

I'm sorry but GRUB Legacy is not maintained. =C2=A0At least not =

by us; =C2=A0we've

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

deprecated it in favour of GRUB 2.

It is also being abandoned by distributors, so I wouldn't recomm=

end that you

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 1 line

put any effort in developing for it.

You've been spouting this line for years, and yet my Ubuntu 10.4 m=

achine

Quoted 6 lines
Quoted 1 line
Quoted 1 line

uses, guess what, GRUB 1.

I assume you typoed, since there's no such thing as Ubuntu 10.4 yet. When there is (well, 10.04 anyway), it will use GRUB 2 by default.

Quoted 1 line
Quoted 1 line

Edward - please do continue to develop patches for GRUB 1 (the one=

that

Quoted 1 line
Quoted 1 line
Quoted 1 line

still actually works plenty well enough for lots of people) and ig=

nore the

Quoted 3 lines
Quoted 1 line
Quoted 1 line

naysayers who are happy to throw out backwards compatibility.

It would be great if somebody could take up Edward's work and port it=

to

Quoted 1 line

GRUB 2. If nobody else does then I'd be interested in doing so myself=

,

Quoted 1 line

although I will not be able to start for a month or two from now.

Is there any guild lines for porting GPLv2 code to GRUB2? I've looked at the GRUB2 wiki but very few things are documented there
(http://grub.enbug.org/). I'd like to see what it would take to port the patches. If I can afford it, I'd like to try.

Quoted 4 lines

Robert is working hard on making GRUB 2 usable, and is just advising Edward that, right now, there is no upstream for GRUB Legacy who coul=

d

Quoted 2 lines

either accept or usefully comment on his patch. It would of course be possible for some people (presumably mostly the distributors who rely=

on

Quoted 3 lines

it) to get together and declare themselves the new upstream for GRUB Legacy, but most of the people who might be interested in such things seem to have either lost interest or thrown their weight behind GRUB =

2

Quoted 1 line

upstream. Certainly this distributor right here is in the latter camp=

as

Quoted 2 lines

it seems much more likely to produce a result that meets our needs in the end. (Plus, I think such a revitalised upstream would be a careta=

ker

Quoted 1 line

at best, and wouldn't really be able to effectively work on some of t=

he

Quoted 4 lines

major issues that have dogged distributors of GRUB Legacy for years without reinventing the wheel of GRUB 2.)

This isn't naysaying those people who post patches for GRUB Legacy - =

but

Quoted 1 line

given the reality that nobody is maintaining GRUB Legacy upstream rig=

ht

Quoted 6 lines

now, which is better, to have your patch ignored or to receive a note saying that it's against an unmaintained target? I'd go for not being ignored any day.

--
Colin Watson =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =

=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 [[email protected]]

Quoted 2 lines

--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs=

" in

Quoted 2 lines

the body of a message to [email protected]
More majordomo info at =C2=A0http://vger.kernel.org/majordomo-info.ht=

ml

Quoted 1 line

--=20
Cheers,
Peng Tao
State Key Laboratory of Networking and Switching Technology Beijing Univ. of Posts and Telecoms.
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" = in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#11 lore raw
Quoted 8 lines · Peng Tao
Quoted 4 lines

It would be great if somebody could take up Edward's work and port it to GRUB 2. If nobody else does then I'd be interested in doing so myself, although I will not be able to start for a month or two from now.

Is there any guild lines for porting GPLv2 code to GRUB2? I've looked at the GRUB2 wiki but very few things are documented there
(http://grub.enbug.org/). I'd like to see what it would take to port the patches. If I can afford it, I'd like to try.

I assume you mean GPLv2-only code (as opposed to GPLv2-or-later). First step would be to contact the copyright holders and ask them to relicense under v3-compatible terms (e.g. GPLv2-or-later). Chances are they didn't chose these terms as an act of hostility, but were simply being zealous about allowing something before they knew what it is.

If that doesn't work, we'll always have Par^W clean room
(http://en.wikipedia.org/wiki/Clean_room_design).

Signature · 5 lines
Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."
#12 lore raw
Quoted 3 lines · Robert Millan
Quoted 2 lines · Peng Tao
Quoted 2 lines

It would be great if somebody could take up Edward's work and port=

it to

Quoted 1 line
Quoted 1 line
Quoted 1 line

GRUB 2. If nobody else does then I'd be interested in doing so mys=

elf,

Quoted 2 lines
Quoted 2 lines
Quoted 1 line

although I will not be able to start for a month or two from now.

Is there any guild lines for porting GPLv2 code to GRUB2? I've looke=

d

Quoted 5 lines
Quoted 3 lines

at the GRUB2 wiki but very few things are documented there
(http://grub.enbug.org/). I'd like to see what it would take to port the patches. If I can afford it, I'd like to try.

I assume you mean GPLv2-only code (as opposed to GPLv2-or-later). Fir=

st

Quoted 1 line

step would be to contact the copyright holders and ask them to relice=

nse

Quoted 1 line

under v3-compatible terms (e.g. GPLv2-or-later). =C2=A0Chances are th=

ey didn't

Quoted 1 line

chose these terms as an act of hostility, but were simply being zealo=

us

Quoted 4 lines

about allowing something before they knew what it is.

If that doesn't work, we'll always have Par^W clean room
(http://en.wikipedia.org/wiki/Clean_room_design).

Edward's patch (stage2/fsys_btrfs.c) is declared GPLv2-or-later. But stage2/btrfs.h (which is extracted from btrfs-progs) is GPLv2-only. At the point, we only need a clean room for btrfs.h, right? And I'm not sure what a clean room design in GRUB2 looks like. Is there an example?

Quoted 5 lines

--
Robert Millan

=C2=A0The DRM opt-in fallacy: "Your data belongs to us. We will decid=

e when (and

Quoted 1 line

=C2=A0how) you may access your data; but nobody's threatening your fr=

eedom: we

Quoted 8 lines

=C2=A0still allow you to remove your data and not access it at all."

_______________________________________________
Grub-devel mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/grub-devel

--=20
Cheers,
Peng Tao
State Key Laboratory of Networking and Switching Technology Beijing Univ. of Posts and Telecoms.
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" = in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#13 lore raw
Quoted 19 lines · Peng Tao
Quoted 17 lines · Robert Millan
Quoted 9 lines · Peng Tao
Quoted 4 lines

It would be great if somebody could take up Edward's work and port it to GRUB 2. If nobody else does then I'd be interested in doing so myself, although I will not be able to start for a month or two from now.

Is there any guild lines for porting GPLv2 code to GRUB2? I've looked at the GRUB2 wiki but very few things are documented there
(http://grub.enbug.org/). I'd like to see what it would take to port the patches. If I can afford it, I'd like to try.

I assume you mean GPLv2-only code (as opposed to GPLv2-or-later). First step would be to contact the copyright holders and ask them to relicense under v3-compatible terms (e.g. GPLv2-or-later). Chances are they didn't chose these terms as an act of hostility, but were simply being zealous about allowing something before they knew what it is.

Quoted 7 lines
Quoted 3 lines

If that doesn't work, we'll always have Par^W clean room
(http://en.wikipedia.org/wiki/Clean_room_design).

Edward's patch (stage2/fsys_btrfs.c) is declared GPLv2-or-later. But stage2/btrfs.h (which is extracted from btrfs-progs) is GPLv2-only. At the point, we only need a clean room for btrfs.h, right?

Doing any clean room is needed only if other ways fail. And I hope Oracle and btrfs contributors could agree to license under GPLv3-compatible terms

Quoted 21 lines

And I'm not sure what a clean room design in GRUB2 looks like. Is there an example?

Quoted 14 lines

--
Robert Millan

The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and how) you may access your data; but nobody's threatening your freedom: we still allow you to remove your data and not access it at all."

_______________________________________________
Grub-devel mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/grub-devel

#14 lore raw
Quoted 23 lines · Vladimir 'phcoder' Serbinenko
Quoted 22 lines · Peng Tao
Quoted 20 lines · Robert Millan
Quoted 9 lines · Peng Tao
Quoted 4 lines

It would be great if somebody could take up Edward's work and port it to GRUB 2. If nobody else does then I'd be interested in doing so myself, although I will not be able to start for a month or two from now.

Is there any guild lines for porting GPLv2 code to GRUB2? I've looked at the GRUB2 wiki but very few things are documented there
(http://grub.enbug.org/). I'd like to see what it would take to port the patches. If I can afford it, I'd like to try.

I assume you mean GPLv2-only code (as opposed to GPLv2-or-later). First step would be to contact the copyright holders and ask them to relicense under v3-compatible terms (e.g. GPLv2-or-later). Chances are they didn't chose these terms as an act of hostility, but were simply being zealous about allowing something before they knew what it is.

If that doesn't work, we'll always have Par^W clean room
(http://en.wikipedia.org/wiki/Clean_room_design).

Quoted 7 lines
Quoted 4 lines

Edward's patch (stage2/fsys_btrfs.c) is declared GPLv2-or-later. But stage2/btrfs.h (which is extracted from btrfs-progs) is GPLv2-only. At the point, we only need a clean room for btrfs.h, right?

Doing any clean room is needed only if other ways fail. And I hope Oracle and btrfs contributors could agree to license under GPLv3-compatible terms

We can provide a GPLv3 version for this file.

-chris

#15 lore raw
In reply to Vladimir 'phcoder' Serbinenko
Quoted 23 lines · Vladimir 'phcoder' Serbinenko
Quoted 22 lines · Peng Tao
Quoted 20 lines · Robert Millan
Quoted 9 lines · Peng Tao
Quoted 4 lines

It would be great if somebody could take up Edward's work and port it to GRUB 2. If nobody else does then I'd be interested in doing so myself, although I will not be able to start for a month or two from now.

Is there any guild lines for porting GPLv2 code to GRUB2? I've looked at the GRUB2 wiki but very few things are documented there
(http://grub.enbug.org/). I'd like to see what it would take to port the patches. If I can afford it, I'd like to try.

I assume you mean GPLv2-only code (as opposed to GPLv2-or-later). First step would be to contact the copyright holders and ask them to relicense under v3-compatible terms (e.g. GPLv2-or-later). Chances are they didn't chose these terms as an act of hostility, but were simply being zealous about allowing something before they knew what it is.

If that doesn't work, we'll always have Par^W clean room
(http://en.wikipedia.org/wiki/Clean_room_design).

Quoted 7 lines
Quoted 4 lines

Edward's patch (stage2/fsys_btrfs.c) is declared GPLv2-or-later. But stage2/btrfs.h (which is extracted from btrfs-progs) is GPLv2-only. At the point, we only need a clean room for btrfs.h, right?

Doing any clean room is needed only if other ways fail. And I hope Oracle and btrfs contributors could agree to license under GPLv3-compatible terms

Yes, this should only be our last-ressort option.

Signature · 5 lines
Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."
#16 lore raw
In reply to Bron Gondwana
Quoted 8 lines · Bron Gondwana

Edward - please do continue to develop patches for GRUB 1 (the one that still actually works plenty well enough for lots of people) and ignore the naysayers who are happy to throw out backwards compatibility.

Sometimes you have to maintain crappy code because people out there depend on it. And we thank those who step up and do it rather than throw their hands up and pretend it doesn't need doing!

Btw, I just wanted to add that although we decide to focus on the codebase that has a future, it doesn't bother me at all that GRUB Legacy is useful to others. We made it free software so that it enables as many users as possible to use their computer in freedom, and so that distributors can adopt it despite our disrecommendation.

I don't regret that they have the practical means to disagree with us and take a different path. It's precisely the freedom we wanted them to benefit from!

Nevertheless, grub-devel is a mailing list for coordination of GRUB 2 development. Most of us are volunteers and our time is quite limited, so bringing up GRUB Legacy development here is at best a distraction. I'll appreciate if you won't do that anymore.

Thanks for listening

Signature · 5 lines
Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."
#17 lore raw
In reply to Robert Millan

Hello!
Robert I tend to agree with your statements except for these:
"At least not by us; we've deprecated it in favour of GRUB 2. It is also being abandoned by distributors, so I wouldn't recommend that you put any effort in developing for it."

Edward this simply means that your efforts here are only being commented about. Robert is indeed correct. We are indeed shifting our directions towards the next generation of GRUB now called GRUB 2. Outside of the Debian family of distributions, I am not at all sure if any of the others are indeed switching. For example Slackware still ships with it in their extras directories. (But only for the 12.2 and 13.0 releases and of course just for the regular 32 bit processors. It may be using GRUB 2 for booting IA64 bit versions of Slackware.)

Ideally if you feel that this is necessary, then go ahead and do so, also check with your distribution's management team and see if they are interested with your work. This might also include those who do indeed use BTRFS for their work as well.

This is the nature and the wonder of the open source world remember. It can be fun, it certainly is surprising.
--
Gregg C Levine [email protected]
"The Force will be with you always." Obi-Wan Kenobi

Quoted 2 lines

-----Original Message-----
From: [email protected]

[mailto:grub-devel-

Quoted 7 lines

[email protected]] On Behalf Of Robert Millan
Sent: Thursday, September 24, 2009 4:22 PM
To: The development of GRUB 2
Cc: Ric Wheeler; Chris Mason; The development of BTRFS
Subject: Re: grub-0.97: btrfs multidevice support [PATCH]

Quoted 6 lines

Hi Edward,

I'm sorry but GRUB Legacy is not maintained. At least not by us; we've deprecated it in favour of GRUB 2.

It is also being abandoned by distributors, so I wouldn't recommend that

you

Quoted 6 lines

put any effort in developing for it.

--
Robert Millan

The DRM opt-in fallacy: "Your data belongs to us. We will decide when

(and

Quoted 8 lines

how) you may access your data; but nobody's threatening your freedom: we still allow you to remove your data and not access it at all."

_______________________________________________
Grub-devel mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/grub-devel