| Age | Commit message (Collapse) | Author |
|---|
|
|
|
riscv64 efiboot already supports booting from softraid volumes.
These installboot bits make sure that the boot loader will be installed on
chunk devices rather than the volume for root on softraid installations,
i.e. full boot support for riscv64, just like amd64, arm64 and sparc64.
regress is happy.
OK kettenis
|
|
aspath->data should not be touched so just exit early.
While there also initalize source_as correctly in the rev case. This
does not matter here but is more correct.
Problem noticed by anton@
OK tb@
|
|
|
|
|
|
redirect location to be https.
|
|
RRDP file and so a few files have bad unique ids, fix them up to be valid.
|
|
From tb@
|
|
entries from sshd causing intermittent failures.
Discussed with bluhm@ back in August this year.
|
|
|
|
|
|
|
|
|
|
|
|
It is now possible to filter by process name, like:
syscall:mmap:entry
/comm == "ld"/
{
...
}
Currently the parser treats C-string like any other expression member even
if arithmetic operations do no apply to strings.
|
|
|
|
|
|
|
|
On EFI platforms, 'installboot -p' on a softraid volume will only prepare
the filesysem inside the volume and leave physical softraid chunks untouched
which leaves you with unbootable chunks.
The current workaround is to prepare chunks manually (see regress).
Fix it in the same spirit the actual "install" already works in softraid.c.
This is what mlarkin has already been tested in a combined diff with the
MD -> MI softraid merge bits from the previous commit.
Works fine on amd64, arm64 and sparc64 upgrades and installations.
OK jsing
|
|
diff.
|
|
report
|
|
Change regex for kdump output to execve JUSTRETURN.
from anton@
|
|
Also add some scaffolding for IPv6 testing but currently IPv6 is broken
because mpe(4) is unable to configure a local IPv6 address.
|
|
is a bug in the kroute code preventing the VPN route to be added properly.
|
|
|
|
|
|
Running make clean without obj should not remove any non-temp files
like *.conf.
|
|
|
|
to do and I'm still not clear what it was supposed to fix.
|
|
respond with the HTTP protocol in the response location header although
HTTPS is being used in the corresponding request. Flag the test as
expected to fail for now.
|
|
regress should be resilient against partially cleaned obj/.
Feedback bluhm
|
|
|
|
|
|
The list of not yet tested archs is smaller, so follow bsd.regress.mk(5)
advise and just print SKIPPED on those.
|
|
|
|
|
|
|
|
Testing with three softraid chunks now means NDISKS=3 as one would expect
and not NDISKS='1 2 3'.
This uses the powerful jot(1) -w and rs(1) -T commands and allows for more
simplifications in the Makefile.
|
|
Treat keydisks like real chunks until installboot properly skips it and
does not touch/install to them anymore.
|
|
Those that still fail (softraid+keydisk or explicit-stage-files) have fixes on tech@.
|
|
Still fails due to the disklabel/vnd kernel race but running into failing
'installboot -p' helps exercising the fsck/newfs exit code fix.
REGRESS_EXIT_EARLY=1 runs past the prepare target without the diff but
properly stops with it as installboot now exits 1 if system(3) fails.
|
|
|
|
|
|
Tiny bit less noise. The many REGRESS_SETUP_ONCE targets could be merged
into a single one and then be added to .SILENT unless VERBOSE=yes to focus
on actual installboot output and ditch the dd/fdisk/disklabel noise by default.
|
|
|
|
Keydisks requires special treatment and should not be touched by installboot.
In real setups they are probably offline (temporary USB stick, only present
during boot to unlock), but when they are around (keydisk chunk on same disk
as crypto volume)...
installboot currently installs treats keydisk like any other chunk, which
must be fixed.
Thus change the tests into a should-work-but-currently-broken setup so an
installboot fix can be immediately tested against it.
|
|
Not all architectures did what the installer does with empty disks.
|
|
Running installboot(8) on softraid(4) volumes means installing stages on
every softraid chunk.
The overall idea is the same, but MD implementations differ.
sparc64_softraid.c's sr_install_bootblk() reuses sparc64_installboot.c's
md_installboot() for this.
For sparc64, md_installboot() does the copy of stage 2, usually
/usr/mdec/ofwboot to /ofwboot, so when `-r root' is passed, it prefixes the
file path with "root".
For single-disk installations (plain-disk and single-chunk softraid) this is
fine, but as soon as multiple chunks are used, md_installboot() currently
prefixes the path each time, obviously resulting in invalid paths starting
with the second run.
Other architectures do reuse md_installboot() as well but either don't do
such a copy or implement the prefixing differently -- plus they must support
softraid in the firt place to be able to hit this type of bug.
With this fixed, regress/usr.sbin/installboot finally passes on sparc64 and
installboot no longer fails at the end of a fresh installation onto softraid
with multiple chunks.
"looks correct" miod
|
|
|
|
These still fail early on due to the opendev(3)/diskmap(4) race condition,
so only hook them up after the kernel is fixed.
Note to self: some distrib/${MACHINE}/ramdisk/install.md pass explicit
newfs(8) flags -- this might be needed here; test once all the relevant
kernel and installboot(8) bugs are squashed and this regress suite can be
run normally.
|
