The bootblocks are 32-bit binaries which use the ENTRY() macros which now
include endbr64 macros. That's not going to work, so use the simple method
of -Dendbr64= to remove them from the instruction stream.
ok kettenis guenther

sys/lib/libz isn't configured for lazy optimizations to save space because
those options are incompatible with the kernel anymore. Set DYNAMIC_CRC_TABLE
and BUILDFIXED for these bootblocks, to save space on the media
ok tb mlarkin

additional files from libkern will be needed by clang10
from mortimer

rewrite mdrandom() in C. previously this XOR'd against rdrand if available,
and alternatively XOR'd against TSC. now always run both sequences, and
also support rdseed as a third procedure.
ok kettenis naddy

Allow amd64 boot(8) and friends to read from an ffs2 filesystem.
Also enable ffs2 for floppy, so that its kernel kan access ffs2. ok deraadt@

Implement a hexdump command in the boot loader. This helps to
inspect the memory layout that the firmware has created. It is
especially useful for UEFI debugging.
OK deraadt@ kettenis@

Use arc4 to bit-spread the 512-byte random buffer over the .openbsd.randomdata
section, which has grown a fair bit with the introduction of retguard.
Mortimer discovered the repeated 512-byte sequence as retguard keys, and
this resolves the issue. (Chacha does not fit on the media, so 1.5K early
drop RC4 is hopefully sufficient in our KARL link universe)
Version crank the bootblocks. sysupgrade -s will install new bootblocks.
ok djm mortimer

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

sys/lib/libz isn't configured for lazy optimizations to save space because
those options are incompatible with the kernel anymore. Set DYNAMIC_CRC_TABLE
and BUILDFIXED for these bootblocks, to save space on the media
ok tb mlarkin

additional files from libkern will be needed by clang10
from mortimer

rewrite mdrandom() in C. previously this XOR'd against rdrand if available,
and alternatively XOR'd against TSC. now always run both sequences, and
also support rdseed as a third procedure.
ok kettenis naddy

Allow amd64 boot(8) and friends to read from an ffs2 filesystem.
Also enable ffs2 for floppy, so that its kernel kan access ffs2. ok deraadt@

Implement a hexdump command in the boot loader. This helps to
inspect the memory layout that the firmware has created. It is
especially useful for UEFI debugging.
OK deraadt@ kettenis@

Use arc4 to bit-spread the 512-byte random buffer over the .openbsd.randomdata
section, which has grown a fair bit with the introduction of retguard.
Mortimer discovered the repeated 512-byte sequence as retguard keys, and
this resolves the issue. (Chacha does not fit on the media, so 1.5K early
drop RC4 is hopefully sufficient in our KARL link universe)
Version crank the bootblocks. sysupgrade -s will install new bootblocks.
ok djm mortimer

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

additional files from libkern will be needed by clang10
from mortimer

rewrite mdrandom() in C. previously this XOR'd against rdrand if available,
and alternatively XOR'd against TSC. now always run both sequences, and
also support rdseed as a third procedure.
ok kettenis naddy

Allow amd64 boot(8) and friends to read from an ffs2 filesystem.
Also enable ffs2 for floppy, so that its kernel kan access ffs2. ok deraadt@

Implement a hexdump command in the boot loader. This helps to
inspect the memory layout that the firmware has created. It is
especially useful for UEFI debugging.
OK deraadt@ kettenis@

Use arc4 to bit-spread the 512-byte random buffer over the .openbsd.randomdata
section, which has grown a fair bit with the introduction of retguard.
Mortimer discovered the repeated 512-byte sequence as retguard keys, and
this resolves the issue. (Chacha does not fit on the media, so 1.5K early
drop RC4 is hopefully sufficient in our KARL link universe)
Version crank the bootblocks. sysupgrade -s will install new bootblocks.
ok djm mortimer

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

rewrite mdrandom() in C. previously this XOR'd against rdrand if available,
and alternatively XOR'd against TSC. now always run both sequences, and
also support rdseed as a third procedure.
ok kettenis naddy

Allow amd64 boot(8) and friends to read from an ffs2 filesystem.
Also enable ffs2 for floppy, so that its kernel kan access ffs2. ok deraadt@

Implement a hexdump command in the boot loader. This helps to
inspect the memory layout that the firmware has created. It is
especially useful for UEFI debugging.
OK deraadt@ kettenis@

Use arc4 to bit-spread the 512-byte random buffer over the .openbsd.randomdata
section, which has grown a fair bit with the introduction of retguard.
Mortimer discovered the repeated 512-byte sequence as retguard keys, and
this resolves the issue. (Chacha does not fit on the media, so 1.5K early
drop RC4 is hopefully sufficient in our KARL link universe)
Version crank the bootblocks. sysupgrade -s will install new bootblocks.
ok djm mortimer

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

Allow amd64 boot(8) and friends to read from an ffs2 filesystem.
Also enable ffs2 for floppy, so that its kernel kan access ffs2. ok deraadt@

Implement a hexdump command in the boot loader. This helps to
inspect the memory layout that the firmware has created. It is
especially useful for UEFI debugging.
OK deraadt@ kettenis@

Use arc4 to bit-spread the 512-byte random buffer over the .openbsd.randomdata
section, which has grown a fair bit with the introduction of retguard.
Mortimer discovered the repeated 512-byte sequence as retguard keys, and
this resolves the issue. (Chacha does not fit on the media, so 1.5K early
drop RC4 is hopefully sufficient in our KARL link universe)
Version crank the bootblocks. sysupgrade -s will install new bootblocks.
ok djm mortimer

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

Implement a hexdump command in the boot loader. This helps to
inspect the memory layout that the firmware has created. It is
especially useful for UEFI debugging.
OK deraadt@ kettenis@

Use arc4 to bit-spread the 512-byte random buffer over the .openbsd.randomdata
section, which has grown a fair bit with the introduction of retguard.
Mortimer discovered the repeated 512-byte sequence as retguard keys, and
this resolves the issue. (Chacha does not fit on the media, so 1.5K early
drop RC4 is hopefully sufficient in our KARL link universe)
Version crank the bootblocks. sysupgrade -s will install new bootblocks.
ok djm mortimer

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

Use arc4 to bit-spread the 512-byte random buffer over the .openbsd.randomdata
section, which has grown a fair bit with the introduction of retguard.
Mortimer discovered the repeated 512-byte sequence as retguard keys, and
this resolves the issue. (Chacha does not fit on the media, so 1.5K early
drop RC4 is hopefully sufficient in our KARL link universe)
Version crank the bootblocks. sysupgrade -s will install new bootblocks.
ok djm mortimer

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

In the bootblocks, after discovering and opening /bsd.upgrade, fchmod -x
so the file cannot be re-executed upon the next boot. This provides a
stronger one-shot-upgrade model than the upgrade script's rm /bsd.upgrade.
Now various forms of upgrade failure will reboot into /bsd, which is probably
more recoverable. Performing fchmod -x depends on (1) use of MI boot.c
(not alpha/macppc/sparc64/sgi/octeon) and (2) "can write blocks" functionality
in the IO layer. Most architectures have this support now.

Two diagnostics "fchmod a-x %s: failed" and "/bsd.upgrade is not u+x" will
remain in the tree while refinements happen for some of the laggard
architectures.

based upon a discussion florian
tested in snapshots for more than a week without any complaints

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@

Work arounda behaviour change in binutils-2.14, now -m<...> -nostdlib
removes /usr/libdata from the script search path. add it back with a -L

branches: 1.1.2;
das boot; das cloned das from das i386

Add support to the BIOS bootloader for random kernel base VA

This diff adds support to be able to load a randomly linked kernel VA
(subject to some range restrictions). This change has been in snaps for
a few days without any fallout.

ok deraadt@

libsa's memcpy() is actually memmove(). make a proper memmove(), and give
memcpy() correct behaviour. This also brings the bcopy() macro into line.

remove the remaining references to .depend files since nothing creates them
any longer; ok espie@ deraadt@

As early as possible, create a link /bsd.booted to the /bsd kernel we
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.

In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.

with yasuoka, lots of discussion with mlarkin, ok tom

Allow building the bootstrap with clang:
* skip the integrated assembler for assym.h
* use as(1) to build biosboot.S and the various versions of srt0.S
ok deraadt@

Add bcrypt pbkdf support to the softraid crypto boot loader code.

Based on a diff from djm@

Switch amd64 to libsa MI softraid.

Rename softraid boot files, which are currently in an MD location. This
will allow us to bring in a MI softraid.{c,h} in lib/libsa.

Rename libsa pbkdf2.c to pkcs5_pbkdf2.c so that we match libutil.

Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

for some time now mandoc has not required MLINKS to function
correctly - logically complete that now by removing MLINKS from base;

authors need only to ensure there is an entry in NAME for any function/
util being added. MLINKS will still work, and remain for perl to ease
upgrades;

ok nicm (curses) bcook (ssl)
ok schwarze, who provided a lot of feedback and assistance
ok tb natano jung

Initial commit of uefi boot loader of OpenBSD.

ok deraadt yuo

fix even more trailing tabs

move putchar() into libsa

"sure" miod@

move getchar() into libsa where applicable

ok miod@

Further clean up and unification of the amd64 and i386 boot(8) makefiles.

Try to load entropy data from disk:/etc/random.seed, and additionally
use a MD-supplied random function. Then, insert this into the ELF
openbsd.randomdata of the kernel, so that it has entropy right from
the start. Some help from jsing for the softraid aspects.
Also tested by phessler

Revert previous: libsa/ufs.c was fixed to not use big types when not
necessary. If something crawls in again that needs these we want to know

ok deraadt@

Include udivdi3/umoddi3 when compiling the amd64/i386 boot blocks - they
are needed for the upcoming type changes.

ok deraadt@

Remove softraid boot support from fdboot (again).

fix bogus dependencies, and allow mkdep to run
checked on amd64 by rpe@
okay jsing@

Fix the makefile so it works on other architectures.

Found the hard way by bluhm@

Generate assym.h where we need it, instead of reaching around into other
object directories.

Discussed with deraadt@

Clean up/standardise makefiles for amd64/stand.

Move the code for softraid boot into separate files, leaving only minimal
hooks that are called from the normal code path. This allows softraid boot
support to be completely enabled or disabled at compile time. For boot(8)
all softraid boot support is now enabled by default, including support for
booting from crypto volumes. Disable softraid boot support for fdboot,
which further shrinks the binary.

Only build fdboot on the appropriate architectures. Also fix the path to
check-boot.pl so that it works with both boot and fdboot.

Add a script that checks the file offsets for boot(8)'s .text and .data
sections match up with their LMAs. This is necessary since biosboot does
not perform relocation.

Discussed with deraadt@

Build a 'fdboot' that will be used for floppy disk media. This can be made
smaller than the normal 'boot', however for now it is identical.

ok deraadt@

Teach amd64 boot(8) how to access a softraid crypto volume. This allows for
full disk encryption since the kernel can now be loaded directly from the
crypto volume, instead of needing to be on FFS.

This is currently disabled by default, however can be enabled by compiling
with boot(8) 'make -DBOOT_CRYPTO'.

Remove -noinhibit-exec now that we no longer have to deal with
"relocation truncated" linker errors.

Reduce the diff between amd64/stand and i386/stand, requested by deraadt@.
These create essentially the same bootblocks, so the build system should not be
diverging too much, or at least easily diffable.
There is still a lot of work to be done here, but this is the low-hanging fruit.

ok jsing@

Get rid of all NOPIE and NOPIE_FLAGS in amd64/stand and use -fno-pie and -nopie
directly instead. This will be the approach taken by every other arch too
in the near future. It is easier than relying on bsd.own.mk to set these
correctly.

discussed with and ok deraadt@

Add NOPIE= bits for sys/arch/*/stand to ensure that bootblocks will always be
built with -fno-pie. This gets the hairiest part of PIE out of the way ...

ok deraadt@

libkern, begone. Move to a new mechanism where config(8)'s "file"
directive can select between MI and MD versions of these files. At
the same time, adjust the boot programs to pick exactly what they need,
instead of the 7 or 8 mechanisms previously used.

There will be some fallout from this, but testing it all by myself is a
ridiculously slow process; it will be finished in-tree.

Various developers were very nice and avoided making fun of me when I
was gibbering in the corner..

Pull out the ELF loadfile pieces from the standalone libraries, so that
both 32- and 64-bit versions can be created (previously only one or the
other could be built for a given boot loader).

Use this to allow the i386 and amd64 boot blocks to boot both ELF32 and
ELF64 kernels (i.e. amd64 boot blocks can now load i386 kernels, and
vice versa). Obviously the system must support LONG mode in order to
successfully run the amd64 kernel once it is loaded.

Advice and discussions from/with dale@ (going back three years). Much
testing nick@ and todd@; thanks.

install boot.conf as a section 5 page, not section 8;

from mikolaj kucharski;
ok mk deraadt krw

boot(8) ``machine memory'' support for > 4G.
Convert parser to strtoll(3) and use 64bit printf.
With help from mickey@
i386 and amd64 tests by me, alpha test by mickey@
OK mickey@, miod@, deraadt@

Need moddi3.c qdivrem.c for libz

The return value from getEBDAaddr() (info) is not used in bios_E820(),
so nuke it. amd64 no longer needs biosprobe.c listed in SRCS. Trims
100 bytes from the boot blocks.

Bump versions on boot, cdboot and pxeboot, as I'm getting cautious in
my old age.

ok weingart@