Merge duplicate MD code into MI sr_open_chunk()

It does not have the prettiest signature, but nicely folds identical copies
into softraid.c, which then allows us to reuse sr_open_chunk() yet again i

Merge duplicate MD code into MI sr_open_chunk()

It does not have the prettiest signature, but nicely folds identical copies
into softraid.c, which then allows us to reuse sr_open_chunk() yet again in
an upcoming diff to make -p softraid aware (fixes arm64 installations).

Regress keeps passing.
Works fine on amd64, arm64 and sparc64.
"looks fine" mlarkin for whom this unbreaks a fresh arm64 softraid install

show more ...

Skip softraid(4) keydisks

Keydisks appear as chunks internally (with special properties) and
installboot(8) thus treated them like actual data chunks.

Most users probably don't hit this as their ke

Skip softraid(4) keydisks

Keydisks appear as chunks internally (with special properties) and
installboot(8) thus treated them like actual data chunks.

Most users probably don't hit this as their keydisk is detached and thus
appears "offline" and gets skipped.

Installing to online keydisks may work but is neither expected nor intended
to work, so properly skip them.

Odd setups like keydisk and CRYPTO chunk on the same physical disk would
end up installing getting bootblocks installed twice.

Pointed out by Mikolaj Kucharski <mikolaj AT kucharski DOT name> who also
provided the actual diff (minor wording tweaks by me)

OK jsing

show more ...

softraid: make verbose messages -n/nowrite aware

Using installboot(8) in dry-run mode aka. -vn (verbose + nowrite) and first
reading "would install" followed by an "installing" made me uneasy.

Luck

softraid: make verbose messages -n/nowrite aware

Using installboot(8) in dry-run mode aka. -vn (verbose + nowrite) and first
reading "would install" followed by an "installing" made me uneasy.

Luckily, nothing was written in dry-run mode, it's just that the
softraid specific code fails to honour -n/nowrite.

So make sure it does and pull the printf out of `if (!nowrite)' so we
see the "would install" messages in dry-run mode in the first place.

Output diff for root on CRYPTO softraid on biosboot(8/amd64) to illustrate:

-# installboot -nv sd1
+# ./obj/installboot -nv sd1
would install bootstrap on /dev/rsd1c
using first-stage /usr/mdec/biosboot, second-stage /usr/mdec/boot
sd1: softraid volume with 1 disk(s)
+sd1: would install boot loader on softraid volume
/usr/mdec/boot is 6 blocks x 16384 bytes
-sd0a: installing boot blocks on /dev/rsd0c, part offset 144
+sd0a: would install boot blocks on /dev/rsd0c, part offset 144
master boot record (MBR) at sector 0
partition 3: type 0xA6 offset 64 size 488392001
/usr/mdec/biosboot will be written at sector 64

OK jsing

show more ...

For open/openat, if the flags parameter does not contain O_CREAT, the
3rd (variadic) mode_t parameter is irrelevant. Many developers in the past
have passed mode_t (0, 044, 0644, or such), which mig

For open/openat, if the flags parameter does not contain O_CREAT, the
3rd (variadic) mode_t parameter is irrelevant. Many developers in the past
have passed mode_t (0, 044, 0644, or such), which might lead future people
to copy this broken idiom, and perhaps even believe this parameter has some
meaning or implication or application. Delete them all.
This comes out of a conversation where tb@ noticed that a strange (but
intentional) pledge behaviour is to always knock-out high-bits from
mode_t on a number of system calls as a safety factor, and his bewilderment
that this appeared to be happening against valid modes (at least visually),
but no sorry, they are all irrelevant junk. They could all be 0xdeafbeef.
ok millert

show more ...

convert macppc, octeon, and loongson to use MI installboot, removing
special case scripting in install.md.
(macppc still requires manual steps for HFS bootmode)
tested by krw, visa, gkoehler

Provide clear errors when trying to install oversized boot loader

sparc64 installboot(8) on softraid(4) with too large files, e.g. unstripped
builds, fails poorly with "installboot: softraid install

Provide clear errors when trying to install oversized boot loader

sparc64 installboot(8) on softraid(4) with too large files, e.g. unstripped
builds, fails poorly with "installboot: softraid installboot failed".

This is due to the BIOCINSTALLBOOT ioctl(2) returing the default EINVAL
rather than using softraid's sr_error() interface properly; additionally,
installboot does not check for such message from the bio(4) layer.


Make the kernel generate "boot block too large" and "boot loader too large"
messages for softraid devices and have installboot act upon them analogous
to bioctl(8), by adapting its bio_status() into the new sr_status() helper.

Input, reminder to look at bioctl, same kernel diff from, OK jsing

show more ...

Recommit, now that we found out how to fix the BIOS related issues: go
back to a 4 byte add instruction.
We do not know *why* though, so if somebody likes a challenge...
Lots of help from semarie@ wh

Recommit, now that we found out how to fix the BIOS related issues: go
back to a 4 byte add instruction.
We do not know *why* though, so if somebody likes a challenge...
Lots of help from semarie@ who has a few systems showing the issue.
ok deraadt@

show more ...

Revert previous, there are BIOS that do not like these changes.
Sorry for the inconvenience. With help from semarie@.

Adapt biosboot(8) so it can read boot(8) from an ffs2 filesystem.
To do this, installboot(8) patches an extra value into biosboot(8).
Code originally from Pedro Martelletto with a twist from myself a

Adapt biosboot(8) so it can read boot(8) from an ffs2 filesystem.
To do this, installboot(8) patches an extra value into biosboot(8).
Code originally from Pedro Martelletto with a twist from myself and
kettenis@; ok jsing@ kettenis@

show more ...

Fix copyright lines, the "one-inode file system" in i386_softraid.c is my
invention; and I have had little to do with the i386_installboot.[ch] files.
This was mixed up in the giant reshuffle a few y

Fix copyright lines, the "one-inode file system" in i386_softraid.c is my
invention; and I have had little to do with the i386_installboot.[ch] files.
This was mixed up in the giant reshuffle a few years back. ok jsing@

show more ...

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if

When system calls indicate an error they return -1, not some arbitrary
value < 0. errno is only updated in this case. Change all (most?)
callers of syscalls to follow this better, and let's see if this strictness
helps us in the future.

show more ...

Plug a couple of leaks of input buffers.

Pointed out by David Carlier.

Don't need <sys/types.h> and <sys/param.h>. Nuke the former when DEV_BSIZE
is needed.

Backout the backout of installboot softraid non-gpt case fix.
krw and I crossed streams

Revert GPT change. Breaks non-GPT install/upgrade.

Found the hard way by reyk@

In sr_install_bootblk for i386, don't clobber 'part' in non-GPT case.
Found the hard way by reyk@.
stupid mistake stsp@
ok reyk@

Fix installboot(8) on i386/amd64 when softraid is on top of GPT.
ok krw deraadt

Nuke trailing whitespace to avoid cluttering possible upcoming diffs.

first batch of cleanup to programs based upon the namespace cleanups
in net/pfvar.h sys/proc.h sys/ucred.h arpa/nameser.h
change to PATH_MAX, reduce use of MIN() and MAX(), HOST_NAME_MAX+1,
LOGIN_NAM

first batch of cleanup to programs based upon the namespace cleanups
in net/pfvar.h sys/proc.h sys/ucred.h arpa/nameser.h
change to PATH_MAX, reduce use of MIN() and MAX(), HOST_NAME_MAX+1,
LOGIN_NAME_MAX, etc etc
ok millert guenther, some review by doug

show more ...

Use calloc() instead of malloc()/memset().

From Benjamin Baier.

Rework installboot and use a single directory with a single makefile. The
directory per machine model is arguably cleaner, however it does not play
well with distrib/special and instbin.

Discussed w

Rework installboot and use a single directory with a single makefile. The
directory per machine model is arguably cleaner, however it does not play
well with distrib/special and instbin.

Discussed with deraadt@

show more ...