Add a parameter 'action' to mpfree() so it can optionally discard
or keep the memory pointed at by the char ** parameter.

Allows mpfree() to be used against the global mountpoints[] and
eliminates a couple of manual loops to free the current contents
of mountpoints[].

No intentional functional change.

Use 'const struct disklabel *' parameters where the disklabel is
not altered. And 'const struct partition *' for pointers into
newly const disklabels.

No intentional functional change.

Move prototypes for checklabel(), readlabel() and parselabel()
into disklabel.c since editor.c and dkcksum.c have no need to
know.

No need for makebootarea() prototype as makebootarea() perished with NUMBOOT in
2015 (r1.211).

Enhance prompt shown during -E label editing. Always
display the disk device whose label is being
edited. Change '>' to '#' when in expert mode.

No need for print_unit to be a global.

ok otto@

don't be silent on auto-allocation failure in write mode. ok millert@ krw@

Allow R (resize auto-partition) after A as well as when started with -A.
This ability was lost in editor.c 1.264; noted by ajacoutot@; ok krw@

Split readlabel() into two functions, readlabel() and parselabel(), and
pledge in-between the two funtion calls. The new readlabel() only does
a DIOC*DINFO ioctl, making sure the fd points to a disk, while
parselabel() does the scary string manipulations. In the makelabel()
code path do an early DIOCGDINFO before pledge. This fixes yet another
pledge problem (disklabel -w /dev/tty floppy576).

Some of this was discussed with beck.

ok semarie, earlier version ok deraadt

NUMBOOT is dead! Nuke the variables and abstractions that were used
to build boot blocks.

ok miod@

support passing a template file for the auto-allocation to disklabel.
template gives mountpoints, min-max size ranges and percentage of disk
foremost intended for autoinstalls, installer bits to follow soon.
with input from many, ok theo

Forward declarations for mpcopy & mpfree.
OK miod@, deraadt@

Replace a '512' with DEV_BSIZE. Calculate physmem size in blocks and
not sectors, since the values being tested/manipulated in the auto_alloc
tables are blocks at that point. Use MEG(256) instead of hand
expanding it when testing physmem.

Nuke unused parameter (f) to makebootarea(). Unused since r1.123
(2008), when '-r' support was removed. Spotted by jsg@ with the
assistance of clang.

Add '-F'. Same as '-f' but uses DUIDs when writing the mount point
information to the specified file. After every label write, read the
label to get current UID info.

Some internal code cleanup with no intended functional change.

ok deraadt@

When dealing with disklabel UIDs (DUIDs) be specific and consistent with
naming.

ok krw@

Print disklabel UIDs using printf/fprintf instead of using the hand rolled
uid_print() function.

ok krw@

fix a few warnings. ok jsing

move editor_allocspace & mpsave out to extern.h so that disklabel.c
gets the declarations it uses. ok krw@

Recycle four ancient fields in the disklabel structure, replacing them with
bounds information, ie. the zone of the disk that OpenBSD can use. Have each
pre-disklabel parser (MBR, DPME, or per-arch MD disklabel parsers) figure out
this area and pass it up to userland. Then, delete all the same disk parsing
code from disklabel(8) since the kernel passes it up. Lots and lots of - signs
in the disklabel(8) code.
Tested on as many platforms as possible, the fallout will be repaired as time
goes on. To test, use disklabel -d <drive> and validate that the bounds do not
overlap any boot blocks. This same information is used by disklabel -A...
OK for the concept from krw, miod, and drahn

Add '-h' flag, and '*' as a unit specifier for 'p' Editor command.
Both cause partition sizes to be displayed using a human readable
format with the units automatically chosen by looking at the smallest
partition in the disk label. Remove forceable humanization in 'A'
code and use '-h' in install scripts.

Prodded & ok deraadt@, verbiage tweaks from jmc@

Alexander Hall points out that disklabel -AE <disk> followed by quitting
without making manual changes does not ask whether you'd like to save the
new automatic disklabel. Change to follow the behaviour of -d and ask
every time. ok krw@.

When auto-allocating, check the size of the disk and display
offset/size information in MB if the disk is <10G, and in GB
otherwise. '-p' will override this default.

readlabel() either errored out or returned a pointer to the global
'lab'. So there is no need for the return value. Just use &lab
wherever it was used. No functional change.

Slightly different version read ok to blambert@

Re-work '-A' so that it sets a flag that -E and -w use, and which will
just display the auto-allocation plan otherwise. Don't overwrite any
spoofed partitions, insert auto allocated partitions where we can.

This is much more install script and weird architecture friendly. I'm
talking about you, macppc.

Many suggestions and code from otto@, feedback from many, cattle
prod applied by deraadt@.

ok deraadt@
"if it works for >8GB ... put it in" miod@
"to be clear, ok" todd@

otto points out that the -v printing changes are incompatible with the
expectation that -R will read the output of disklabel. I suspect we
will find another way to do this; ok otto

Move some output under the control of the '-v' flag. In particular only
display the partition info by default as is done in the E(ditor). The
physical info is now only displayed if '-v' is specified.

ok deraadt@

Add 'A' command to Editor mode, and -A flag to automatically execute it
on disks without an existing label. The 'A' command allocates all space
on the disk into a reasonable partition scheme for a root disk.

Feedback from several, time to work on it in-tree.

Prodded (repeatedly) by and ok deraadt@

On *ppc disks shared with MacOS, walk the Apple partition map to find out
the bounds of the OpenBSD area. Should prevent users from shooting themselves
in the feet.

ok krw@

In interactive mode, always suggest offsets and sizes inside the
OpenBSD area and change the 'p' command to only print the boundaries
and size of the OpenBSD area and the free space within that area.
Introduce a 'l' command to print the header only. With krw@; ok krw@
deraadt@; jmc@ for the man page bits

Make disklabel(8) aware of the new fields, allowing for larger than
2T disks, partitions and offsets. ok deraadt@

If a default label has been created with the -d command line option or
the 'D' editor command and has not been written to disk when the user
issues the 'q' editor command, ask if the label should be written to
disk before exiting.

Print disk size and free space in requested unit. Some refactoring to
avoid code duplication as well. ok marco@ deraadt@

add some missing $, ok djm@ 'That looks fine to me' millert@

Tidy disklabel output for todays big disks. Based on a diff from f
at obiit dot org in PR 3847. ok tedu@ deraadt@ millert@

missing protos

Use 'const struct disklabel *' parameters where the disklabel is
not altered. And 'const struct partition *' for pointers into
newly const disklabels.

No intentional functional change.

Move prototypes for checklabel(), readlabel() and parselabel()
into disklabel.c since editor.c and dkcksum.c have no need to
know.

No need for makebootarea() prototype as makebootarea() perished with NUMBOOT in
2015 (r1.211).

Enhance prompt shown during -E label editing. Always
display the disk device whose label is being
edited. Change '>' to '#' when in expert mode.

No need for print_unit to be a global.

ok otto@

don't be silent on auto-allocation failure in write mode. ok millert@ krw@

Allow R (resize auto-partition) after A as well as when started with -A.
This ability was lost in editor.c 1.264; noted by ajacoutot@; ok krw@

Split readlabel() into two functions, readlabel() and parselabel(), and
pledge in-between the two funtion calls. The new readlabel() only does
a DIOC*DINFO ioctl, making sure the fd points to a disk, while
parselabel() does the scary string manipulations. In the makelabel()
code path do an early DIOCGDINFO before pledge. This fixes yet another
pledge problem (disklabel -w /dev/tty floppy576).

Some of this was discussed with beck.

ok semarie, earlier version ok deraadt

NUMBOOT is dead! Nuke the variables and abstractions that were used
to build boot blocks.

ok miod@

support passing a template file for the auto-allocation to disklabel.
template gives mountpoints, min-max size ranges and percentage of disk
foremost intended for autoinstalls, installer bits to follow soon.
with input from many, ok theo

Forward declarations for mpcopy & mpfree.
OK miod@, deraadt@

Replace a '512' with DEV_BSIZE. Calculate physmem size in blocks and
not sectors, since the values being tested/manipulated in the auto_alloc
tables are blocks at that point. Use MEG(256) instead of hand
expanding it when testing physmem.

Nuke unused parameter (f) to makebootarea(). Unused since r1.123
(2008), when '-r' support was removed. Spotted by jsg@ with the
assistance of clang.

Add '-F'. Same as '-f' but uses DUIDs when writing the mount point
information to the specified file. After every label write, read the
label to get current UID info.

Some internal code cleanup with no intended functional change.

ok deraadt@

When dealing with disklabel UIDs (DUIDs) be specific and consistent with
naming.

ok krw@

Print disklabel UIDs using printf/fprintf instead of using the hand rolled
uid_print() function.

ok krw@

fix a few warnings. ok jsing

move editor_allocspace & mpsave out to extern.h so that disklabel.c
gets the declarations it uses. ok krw@

Recycle four ancient fields in the disklabel structure, replacing them with
bounds information, ie. the zone of the disk that OpenBSD can use. Have each
pre-disklabel parser (MBR, DPME, or per-arch MD disklabel parsers) figure out
this area and pass it up to userland. Then, delete all the same disk parsing
code from disklabel(8) since the kernel passes it up. Lots and lots of - signs
in the disklabel(8) code.
Tested on as many platforms as possible, the fallout will be repaired as time
goes on. To test, use disklabel -d <drive> and validate that the bounds do not
overlap any boot blocks. This same information is used by disklabel -A...
OK for the concept from krw, miod, and drahn

Add '-h' flag, and '*' as a unit specifier for 'p' Editor command.
Both cause partition sizes to be displayed using a human readable
format with the units automatically chosen by looking at the smallest
partition in the disk label. Remove forceable humanization in 'A'
code and use '-h' in install scripts.

Prodded & ok deraadt@, verbiage tweaks from jmc@

Alexander Hall points out that disklabel -AE <disk> followed by quitting
without making manual changes does not ask whether you'd like to save the
new automatic disklabel. Change to follow the behaviour of -d and ask
every time. ok krw@.

When auto-allocating, check the size of the disk and display
offset/size information in MB if the disk is <10G, and in GB
otherwise. '-p' will override this default.

readlabel() either errored out or returned a pointer to the global
'lab'. So there is no need for the return value. Just use &lab
wherever it was used. No functional change.

Slightly different version read ok to blambert@

Re-work '-A' so that it sets a flag that -E and -w use, and which will
just display the auto-allocation plan otherwise. Don't overwrite any
spoofed partitions, insert auto allocated partitions where we can.

This is much more install script and weird architecture friendly. I'm
talking about you, macppc.

Many suggestions and code from otto@, feedback from many, cattle
prod applied by deraadt@.

ok deraadt@
"if it works for >8GB ... put it in" miod@
"to be clear, ok" todd@

otto points out that the -v printing changes are incompatible with the
expectation that -R will read the output of disklabel. I suspect we
will find another way to do this; ok otto

Move some output under the control of the '-v' flag. In particular only
display the partition info by default as is done in the E(ditor). The
physical info is now only displayed if '-v' is specified.

ok deraadt@

Add 'A' command to Editor mode, and -A flag to automatically execute it
on disks without an existing label. The 'A' command allocates all space
on the disk into a reasonable partition scheme for a root disk.

Feedback from several, time to work on it in-tree.

Prodded (repeatedly) by and ok deraadt@

On *ppc disks shared with MacOS, walk the Apple partition map to find out
the bounds of the OpenBSD area. Should prevent users from shooting themselves
in the feet.

ok krw@

In interactive mode, always suggest offsets and sizes inside the
OpenBSD area and change the 'p' command to only print the boundaries
and size of the OpenBSD area and the free space within that area.
Introduce a 'l' command to print the header only. With krw@; ok krw@
deraadt@; jmc@ for the man page bits

Make disklabel(8) aware of the new fields, allowing for larger than
2T disks, partitions and offsets. ok deraadt@

If a default label has been created with the -d command line option or
the 'D' editor command and has not been written to disk when the user
issues the 'q' editor command, ask if the label should be written to
disk before exiting.

Print disk size and free space in requested unit. Some refactoring to
avoid code duplication as well. ok marco@ deraadt@

add some missing $, ok djm@ 'That looks fine to me' millert@

Tidy disklabel output for todays big disks. Based on a diff from f
at obiit dot org in PR 3847. ok tedu@ deraadt@ millert@

missing protos

Move prototypes for checklabel(), readlabel() and parselabel()
into disklabel.c since editor.c and dkcksum.c have no need to
know.

No need for makebootarea() prototype as makebootarea() perished with NUMBOOT in
2015 (r1.211).

Enhance prompt shown during -E label editing. Always
display the disk device whose label is being
edited. Change '>' to '#' when in expert mode.

No need for print_unit to be a global.

ok otto@

don't be silent on auto-allocation failure in write mode. ok millert@ krw@

Allow R (resize auto-partition) after A as well as when started with -A.
This ability was lost in editor.c 1.264; noted by ajacoutot@; ok krw@

Split readlabel() into two functions, readlabel() and parselabel(), and
pledge in-between the two funtion calls. The new readlabel() only does
a DIOC*DINFO ioctl, making sure the fd points to a disk, while
parselabel() does the scary string manipulations. In the makelabel()
code path do an early DIOCGDINFO before pledge. This fixes yet another
pledge problem (disklabel -w /dev/tty floppy576).

Some of this was discussed with beck.

ok semarie, earlier version ok deraadt

NUMBOOT is dead! Nuke the variables and abstractions that were used
to build boot blocks.

ok miod@

support passing a template file for the auto-allocation to disklabel.
template gives mountpoints, min-max size ranges and percentage of disk
foremost intended for autoinstalls, installer bits to follow soon.
with input from many, ok theo

Forward declarations for mpcopy & mpfree.
OK miod@, deraadt@

Replace a '512' with DEV_BSIZE. Calculate physmem size in blocks and
not sectors, since the values being tested/manipulated in the auto_alloc
tables are blocks at that point. Use MEG(256) instead of hand
expanding it when testing physmem.

Nuke unused parameter (f) to makebootarea(). Unused since r1.123
(2008), when '-r' support was removed. Spotted by jsg@ with the
assistance of clang.

Add '-F'. Same as '-f' but uses DUIDs when writing the mount point
information to the specified file. After every label write, read the
label to get current UID info.

Some internal code cleanup with no intended functional change.

ok deraadt@

When dealing with disklabel UIDs (DUIDs) be specific and consistent with
naming.

ok krw@

Print disklabel UIDs using printf/fprintf instead of using the hand rolled
uid_print() function.

ok krw@

fix a few warnings. ok jsing

move editor_allocspace & mpsave out to extern.h so that disklabel.c
gets the declarations it uses. ok krw@

Recycle four ancient fields in the disklabel structure, replacing them with
bounds information, ie. the zone of the disk that OpenBSD can use. Have each
pre-disklabel parser (MBR, DPME, or per-arch MD disklabel parsers) figure out
this area and pass it up to userland. Then, delete all the same disk parsing
code from disklabel(8) since the kernel passes it up. Lots and lots of - signs
in the disklabel(8) code.
Tested on as many platforms as possible, the fallout will be repaired as time
goes on. To test, use disklabel -d <drive> and validate that the bounds do not
overlap any boot blocks. This same information is used by disklabel -A...
OK for the concept from krw, miod, and drahn

Add '-h' flag, and '*' as a unit specifier for 'p' Editor command.
Both cause partition sizes to be displayed using a human readable
format with the units automatically chosen by looking at the smallest
partition in the disk label. Remove forceable humanization in 'A'
code and use '-h' in install scripts.

Prodded & ok deraadt@, verbiage tweaks from jmc@

Alexander Hall points out that disklabel -AE <disk> followed by quitting
without making manual changes does not ask whether you'd like to save the
new automatic disklabel. Change to follow the behaviour of -d and ask
every time. ok krw@.

When auto-allocating, check the size of the disk and display
offset/size information in MB if the disk is <10G, and in GB
otherwise. '-p' will override this default.

readlabel() either errored out or returned a pointer to the global
'lab'. So there is no need for the return value. Just use &lab
wherever it was used. No functional change.

Slightly different version read ok to blambert@

Re-work '-A' so that it sets a flag that -E and -w use, and which will
just display the auto-allocation plan otherwise. Don't overwrite any
spoofed partitions, insert auto allocated partitions where we can.

This is much more install script and weird architecture friendly. I'm
talking about you, macppc.

Many suggestions and code from otto@, feedback from many, cattle
prod applied by deraadt@.

ok deraadt@
"if it works for >8GB ... put it in" miod@
"to be clear, ok" todd@

otto points out that the -v printing changes are incompatible with the
expectation that -R will read the output of disklabel. I suspect we
will find another way to do this; ok otto

Move some output under the control of the '-v' flag. In particular only
display the partition info by default as is done in the E(ditor). The
physical info is now only displayed if '-v' is specified.

ok deraadt@

Add 'A' command to Editor mode, and -A flag to automatically execute it
on disks without an existing label. The 'A' command allocates all space
on the disk into a reasonable partition scheme for a root disk.

Feedback from several, time to work on it in-tree.

Prodded (repeatedly) by and ok deraadt@

On *ppc disks shared with MacOS, walk the Apple partition map to find out
the bounds of the OpenBSD area. Should prevent users from shooting themselves
in the feet.

ok krw@

In interactive mode, always suggest offsets and sizes inside the
OpenBSD area and change the 'p' command to only print the boundaries
and size of the OpenBSD area and the free space within that area.
Introduce a 'l' command to print the header only. With krw@; ok krw@
deraadt@; jmc@ for the man page bits

Make disklabel(8) aware of the new fields, allowing for larger than
2T disks, partitions and offsets. ok deraadt@

If a default label has been created with the -d command line option or
the 'D' editor command and has not been written to disk when the user
issues the 'q' editor command, ask if the label should be written to
disk before exiting.

Print disk size and free space in requested unit. Some refactoring to
avoid code duplication as well. ok marco@ deraadt@

add some missing $, ok djm@ 'That looks fine to me' millert@

Tidy disklabel output for todays big disks. Based on a diff from f
at obiit dot org in PR 3847. ok tedu@ deraadt@ millert@

missing protos

Enhance prompt shown during -E label editing. Always
display the disk device whose label is being
edited. Change '>' to '#' when in expert mode.

No need for print_unit to be a global.

ok otto@

don't be silent on auto-allocation failure in write mode. ok millert@ krw@

Allow R (resize auto-partition) after A as well as when started with -A.
This ability was lost in editor.c 1.264; noted by ajacoutot@; ok krw@

Split readlabel() into two functions, readlabel() and parselabel(), and
pledge in-between the two funtion calls. The new readlabel() only does
a DIOC*DINFO ioctl, making sure the fd points to a disk, while
parselabel() does the scary string manipulations. In the makelabel()
code path do an early DIOCGDINFO before pledge. This fixes yet another
pledge problem (disklabel -w /dev/tty floppy576).

Some of this was discussed with beck.

ok semarie, earlier version ok deraadt

NUMBOOT is dead! Nuke the variables and abstractions that were used
to build boot blocks.

ok miod@

support passing a template file for the auto-allocation to disklabel.
template gives mountpoints, min-max size ranges and percentage of disk
foremost intended for autoinstalls, installer bits to follow soon.
with input from many, ok theo

Forward declarations for mpcopy & mpfree.
OK miod@, deraadt@

Replace a '512' with DEV_BSIZE. Calculate physmem size in blocks and
not sectors, since the values being tested/manipulated in the auto_alloc
tables are blocks at that point. Use MEG(256) instead of hand
expanding it when testing physmem.

Nuke unused parameter (f) to makebootarea(). Unused since r1.123
(2008), when '-r' support was removed. Spotted by jsg@ with the
assistance of clang.

Add '-F'. Same as '-f' but uses DUIDs when writing the mount point
information to the specified file. After every label write, read the
label to get current UID info.

Some internal code cleanup with no intended functional change.

ok deraadt@

When dealing with disklabel UIDs (DUIDs) be specific and consistent with
naming.

ok krw@

Print disklabel UIDs using printf/fprintf instead of using the hand rolled
uid_print() function.

ok krw@

fix a few warnings. ok jsing

move editor_allocspace & mpsave out to extern.h so that disklabel.c
gets the declarations it uses. ok krw@

Recycle four ancient fields in the disklabel structure, replacing them with
bounds information, ie. the zone of the disk that OpenBSD can use. Have each
pre-disklabel parser (MBR, DPME, or per-arch MD disklabel parsers) figure out
this area and pass it up to userland. Then, delete all the same disk parsing
code from disklabel(8) since the kernel passes it up. Lots and lots of - signs
in the disklabel(8) code.
Tested on as many platforms as possible, the fallout will be repaired as time
goes on. To test, use disklabel -d <drive> and validate that the bounds do not
overlap any boot blocks. This same information is used by disklabel -A...
OK for the concept from krw, miod, and drahn

Add '-h' flag, and '*' as a unit specifier for 'p' Editor command.
Both cause partition sizes to be displayed using a human readable
format with the units automatically chosen by looking at the smallest
partition in the disk label. Remove forceable humanization in 'A'
code and use '-h' in install scripts.

Prodded & ok deraadt@, verbiage tweaks from jmc@

Alexander Hall points out that disklabel -AE <disk> followed by quitting
without making manual changes does not ask whether you'd like to save the
new automatic disklabel. Change to follow the behaviour of -d and ask
every time. ok krw@.

When auto-allocating, check the size of the disk and display
offset/size information in MB if the disk is <10G, and in GB
otherwise. '-p' will override this default.

readlabel() either errored out or returned a pointer to the global
'lab'. So there is no need for the return value. Just use &lab
wherever it was used. No functional change.

Slightly different version read ok to blambert@

Re-work '-A' so that it sets a flag that -E and -w use, and which will
just display the auto-allocation plan otherwise. Don't overwrite any
spoofed partitions, insert auto allocated partitions where we can.

This is much more install script and weird architecture friendly. I'm
talking about you, macppc.

Many suggestions and code from otto@, feedback from many, cattle
prod applied by deraadt@.

ok deraadt@
"if it works for >8GB ... put it in" miod@
"to be clear, ok" todd@

otto points out that the -v printing changes are incompatible with the
expectation that -R will read the output of disklabel. I suspect we
will find another way to do this; ok otto

Move some output under the control of the '-v' flag. In particular only
display the partition info by default as is done in the E(ditor). The
physical info is now only displayed if '-v' is specified.

ok deraadt@

Add 'A' command to Editor mode, and -A flag to automatically execute it
on disks without an existing label. The 'A' command allocates all space
on the disk into a reasonable partition scheme for a root disk.

Feedback from several, time to work on it in-tree.

Prodded (repeatedly) by and ok deraadt@

On *ppc disks shared with MacOS, walk the Apple partition map to find out
the bounds of the OpenBSD area. Should prevent users from shooting themselves
in the feet.

ok krw@

In interactive mode, always suggest offsets and sizes inside the
OpenBSD area and change the 'p' command to only print the boundaries
and size of the OpenBSD area and the free space within that area.
Introduce a 'l' command to print the header only. With krw@; ok krw@
deraadt@; jmc@ for the man page bits

Make disklabel(8) aware of the new fields, allowing for larger than
2T disks, partitions and offsets. ok deraadt@

If a default label has been created with the -d command line option or
the 'D' editor command and has not been written to disk when the user
issues the 'q' editor command, ask if the label should be written to
disk before exiting.

Print disk size and free space in requested unit. Some refactoring to
avoid code duplication as well. ok marco@ deraadt@

add some missing $, ok djm@ 'That looks fine to me' millert@

Tidy disklabel output for todays big disks. Based on a diff from f
at obiit dot org in PR 3847. ok tedu@ deraadt@ millert@

missing protos

No need for print_unit to be a global.

ok otto@

don't be silent on auto-allocation failure in write mode. ok millert@ krw@

Allow R (resize auto-partition) after A as well as when started with -A.
This ability was lost in editor.c 1.264; noted by ajacoutot@; ok krw@

Split readlabel() into two functions, readlabel() and parselabel(), and
pledge in-between the two funtion calls. The new readlabel() only does
a DIOC*DINFO ioctl, making sure the fd points to a disk, while
parselabel() does the scary string manipulations. In the makelabel()
code path do an early DIOCGDINFO before pledge. This fixes yet another
pledge problem (disklabel -w /dev/tty floppy576).

Some of this was discussed with beck.

ok semarie, earlier version ok deraadt

NUMBOOT is dead! Nuke the variables and abstractions that were used
to build boot blocks.

ok miod@

support passing a template file for the auto-allocation to disklabel.
template gives mountpoints, min-max size ranges and percentage of disk
foremost intended for autoinstalls, installer bits to follow soon.
with input from many, ok theo

Forward declarations for mpcopy & mpfree.
OK miod@, deraadt@

Replace a '512' with DEV_BSIZE. Calculate physmem size in blocks and
not sectors, since the values being tested/manipulated in the auto_alloc
tables are blocks at that point. Use MEG(256) instead of hand
expanding it when testing physmem.

Nuke unused parameter (f) to makebootarea(). Unused since r1.123
(2008), when '-r' support was removed. Spotted by jsg@ with the
assistance of clang.

Add '-F'. Same as '-f' but uses DUIDs when writing the mount point
information to the specified file. After every label write, read the
label to get current UID info.

Some internal code cleanup with no intended functional change.

ok deraadt@

When dealing with disklabel UIDs (DUIDs) be specific and consistent with
naming.

ok krw@

Print disklabel UIDs using printf/fprintf instead of using the hand rolled
uid_print() function.

ok krw@

fix a few warnings. ok jsing

move editor_allocspace & mpsave out to extern.h so that disklabel.c
gets the declarations it uses. ok krw@

Recycle four ancient fields in the disklabel structure, replacing them with
bounds information, ie. the zone of the disk that OpenBSD can use. Have each
pre-disklabel parser (MBR, DPME, or per-arch MD disklabel parsers) figure out
this area and pass it up to userland. Then, delete all the same disk parsing
code from disklabel(8) since the kernel passes it up. Lots and lots of - signs
in the disklabel(8) code.
Tested on as many platforms as possible, the fallout will be repaired as time
goes on. To test, use disklabel -d <drive> and validate that the bounds do not
overlap any boot blocks. This same information is used by disklabel -A...
OK for the concept from krw, miod, and drahn

Add '-h' flag, and '*' as a unit specifier for 'p' Editor command.
Both cause partition sizes to be displayed using a human readable
format with the units automatically chosen by looking at the smallest
partition in the disk label. Remove forceable humanization in 'A'
code and use '-h' in install scripts.

Prodded & ok deraadt@, verbiage tweaks from jmc@

Alexander Hall points out that disklabel -AE <disk> followed by quitting
without making manual changes does not ask whether you'd like to save the
new automatic disklabel. Change to follow the behaviour of -d and ask
every time. ok krw@.

When auto-allocating, check the size of the disk and display
offset/size information in MB if the disk is <10G, and in GB
otherwise. '-p' will override this default.

readlabel() either errored out or returned a pointer to the global
'lab'. So there is no need for the return value. Just use &lab
wherever it was used. No functional change.

Slightly different version read ok to blambert@

Re-work '-A' so that it sets a flag that -E and -w use, and which will
just display the auto-allocation plan otherwise. Don't overwrite any
spoofed partitions, insert auto allocated partitions where we can.

This is much more install script and weird architecture friendly. I'm
talking about you, macppc.

Many suggestions and code from otto@, feedback from many, cattle
prod applied by deraadt@.

ok deraadt@
"if it works for >8GB ... put it in" miod@
"to be clear, ok" todd@

otto points out that the -v printing changes are incompatible with the
expectation that -R will read the output of disklabel. I suspect we
will find another way to do this; ok otto

Move some output under the control of the '-v' flag. In particular only
display the partition info by default as is done in the E(ditor). The
physical info is now only displayed if '-v' is specified.

ok deraadt@

Add 'A' command to Editor mode, and -A flag to automatically execute it
on disks without an existing label. The 'A' command allocates all space
on the disk into a reasonable partition scheme for a root disk.

Feedback from several, time to work on it in-tree.

Prodded (repeatedly) by and ok deraadt@

On *ppc disks shared with MacOS, walk the Apple partition map to find out
the bounds of the OpenBSD area. Should prevent users from shooting themselves
in the feet.

ok krw@

In interactive mode, always suggest offsets and sizes inside the
OpenBSD area and change the 'p' command to only print the boundaries
and size of the OpenBSD area and the free space within that area.
Introduce a 'l' command to print the header only. With krw@; ok krw@
deraadt@; jmc@ for the man page bits

Make disklabel(8) aware of the new fields, allowing for larger than
2T disks, partitions and offsets. ok deraadt@

If a default label has been created with the -d command line option or
the 'D' editor command and has not been written to disk when the user
issues the 'q' editor command, ask if the label should be written to
disk before exiting.

Print disk size and free space in requested unit. Some refactoring to
avoid code duplication as well. ok marco@ deraadt@

add some missing $, ok djm@ 'That looks fine to me' millert@

Tidy disklabel output for todays big disks. Based on a diff from f
at obiit dot org in PR 3847. ok tedu@ deraadt@ millert@

missing protos

don't be silent on auto-allocation failure in write mode. ok millert@ krw@

Allow R (resize auto-partition) after A as well as when started with -A.
This ability was lost in editor.c 1.264; noted by ajacoutot@; ok krw@

Split readlabel() into two functions, readlabel() and parselabel(), and
pledge in-between the two funtion calls. The new readlabel() only does
a DIOC*DINFO ioctl, making sure the fd points to a disk, while
parselabel() does the scary string manipulations. In the makelabel()
code path do an early DIOCGDINFO before pledge. This fixes yet another
pledge problem (disklabel -w /dev/tty floppy576).

Some of this was discussed with beck.

ok semarie, earlier version ok deraadt

NUMBOOT is dead! Nuke the variables and abstractions that were used
to build boot blocks.

ok miod@

support passing a template file for the auto-allocation to disklabel.
template gives mountpoints, min-max size ranges and percentage of disk
foremost intended for autoinstalls, installer bits to follow soon.
with input from many, ok theo

Forward declarations for mpcopy & mpfree.
OK miod@, deraadt@

Replace a '512' with DEV_BSIZE. Calculate physmem size in blocks and
not sectors, since the values being tested/manipulated in the auto_alloc
tables are blocks at that point. Use MEG(256) instead of hand
expanding it when testing physmem.

Nuke unused parameter (f) to makebootarea(). Unused since r1.123
(2008), when '-r' support was removed. Spotted by jsg@ with the
assistance of clang.

Add '-F'. Same as '-f' but uses DUIDs when writing the mount point
information to the specified file. After every label write, read the
label to get current UID info.

Some internal code cleanup with no intended functional change.

ok deraadt@

When dealing with disklabel UIDs (DUIDs) be specific and consistent with
naming.

ok krw@

Print disklabel UIDs using printf/fprintf instead of using the hand rolled
uid_print() function.

ok krw@

fix a few warnings. ok jsing

move editor_allocspace & mpsave out to extern.h so that disklabel.c
gets the declarations it uses. ok krw@

Recycle four ancient fields in the disklabel structure, replacing them with
bounds information, ie. the zone of the disk that OpenBSD can use. Have each
pre-disklabel parser (MBR, DPME, or per-arch MD disklabel parsers) figure out
this area and pass it up to userland. Then, delete all the same disk parsing
code from disklabel(8) since the kernel passes it up. Lots and lots of - signs
in the disklabel(8) code.
Tested on as many platforms as possible, the fallout will be repaired as time
goes on. To test, use disklabel -d <drive> and validate that the bounds do not
overlap any boot blocks. This same information is used by disklabel -A...
OK for the concept from krw, miod, and drahn

Add '-h' flag, and '*' as a unit specifier for 'p' Editor command.
Both cause partition sizes to be displayed using a human readable
format with the units automatically chosen by looking at the smallest
partition in the disk label. Remove forceable humanization in 'A'
code and use '-h' in install scripts.

Prodded & ok deraadt@, verbiage tweaks from jmc@

Alexander Hall points out that disklabel -AE <disk> followed by quitting
without making manual changes does not ask whether you'd like to save the
new automatic disklabel. Change to follow the behaviour of -d and ask
every time. ok krw@.

When auto-allocating, check the size of the disk and display
offset/size information in MB if the disk is <10G, and in GB
otherwise. '-p' will override this default.

readlabel() either errored out or returned a pointer to the global
'lab'. So there is no need for the return value. Just use &lab
wherever it was used. No functional change.

Slightly different version read ok to blambert@

Re-work '-A' so that it sets a flag that -E and -w use, and which will
just display the auto-allocation plan otherwise. Don't overwrite any
spoofed partitions, insert auto allocated partitions where we can.

This is much more install script and weird architecture friendly. I'm
talking about you, macppc.

Many suggestions and code from otto@, feedback from many, cattle
prod applied by deraadt@.

ok deraadt@
"if it works for >8GB ... put it in" miod@
"to be clear, ok" todd@

otto points out that the -v printing changes are incompatible with the
expectation that -R will read the output of disklabel. I suspect we
will find another way to do this; ok otto

Move some output under the control of the '-v' flag. In particular only
display the partition info by default as is done in the E(ditor). The
physical info is now only displayed if '-v' is specified.

ok deraadt@

Add 'A' command to Editor mode, and -A flag to automatically execute it
on disks without an existing label. The 'A' command allocates all space
on the disk into a reasonable partition scheme for a root disk.

Feedback from several, time to work on it in-tree.

Prodded (repeatedly) by and ok deraadt@

On *ppc disks shared with MacOS, walk the Apple partition map to find out
the bounds of the OpenBSD area. Should prevent users from shooting themselves
in the feet.

ok krw@

In interactive mode, always suggest offsets and sizes inside the
OpenBSD area and change the 'p' command to only print the boundaries
and size of the OpenBSD area and the free space within that area.
Introduce a 'l' command to print the header only. With krw@; ok krw@
deraadt@; jmc@ for the man page bits

Make disklabel(8) aware of the new fields, allowing for larger than
2T disks, partitions and offsets. ok deraadt@

If a default label has been created with the -d command line option or
the 'D' editor command and has not been written to disk when the user
issues the 'q' editor command, ask if the label should be written to
disk before exiting.

Print disk size and free space in requested unit. Some refactoring to
avoid code duplication as well. ok marco@ deraadt@

add some missing $, ok djm@ 'That looks fine to me' millert@

Tidy disklabel output for todays big disks. Based on a diff from f
at obiit dot org in PR 3847. ok tedu@ deraadt@ millert@

missing protos