Commit Graph

213 Commits

Author SHA1 Message Date
Toomas Soome
6c7a932d0b loader: start kernel in text mode when there is no vbefb vt driver
If kernel is built without VT vbefb driver, make sure
we start kernel in text mode.
2021-01-27 01:07:34 +02:00
Toomas Soome
6c789c55c4 loader: create built in font from bold font face
We did replace full version of default font 8x16v with bold, also
use bold version for built in font.
2021-01-23 10:53:30 +02:00
Toomas Soome
52e3a7300d loader: handle malloc failures in vbe_init, use consistent naming
Handle malloc failures in vbe_init().

If it should so happen and we do get malloc failure in vbe_init(),
use original mode list.

Replace nitems with nentries to have naming consistency and avoid
confusion with nitems() macro.

Reported by: yuripv, rpokala
2021-01-19 21:58:46 +02:00
Toomas Soome
ad1ebbe5ce loader: create local copy of mode list provided by vbeinfoblock
Apparently some systems do corrupt mode list memory area, so we need
to use local copy instead.
2021-01-16 12:23:22 +02:00
Toomas Soome
0974bfa3a8 loader: do not update palette in text mode (real fix)
Apparently palette update while in text mode, will cause
some adapters to end up with blank display. Previous update had
condition reversed.
2021-01-15 15:43:14 +02:00
Toomas Soome
89632acb50 loader: term_image_display() should test screen_buffer
Make sure screen_buffer is not NULL.
2021-01-10 21:59:52 +02:00
Toomas Soome
1790f5e654 loader: do not update palette in text mode
Apparently palette update while in text mode, will cause
some adapters to end up with blank display.
2021-01-07 17:07:53 +02:00
Toomas Soome
babda0952f loader: instead of hw.vga.textmode, use screen.textmode
hw.vga.textmode is directing VT VGA backend to use text mode.

The default screen mode for BIOS loader is text, and default
screen mode for VT VGA backend is graphics (unless we are running on
hypervisor or hw.vga.textmode is set to 1). Using hw.vga.textmode
for loader does remove possibility to have graphical mode VT VGA with
text mode loader.

screen.textmode can have possible values "0" to disable text mode,
and "1" to set text mode.
2021-01-06 14:38:55 +02:00
Toomas Soome
31c2bcad7e loader: remove left over call to unsetenv()
We do not touch kern.vt.fb.default_mode.
2021-01-06 14:35:26 +02:00
Toomas Soome
99870c70ba loader: rewrite vidc_install_font
Instead of trying to set reasonable register values, save significant
register values, then prepare for font upload and then restore
registers from saved data.

This seems to fix text mode for most cases where text mode breakage
was reported.
2021-01-06 10:20:04 +02:00
Toomas Soome
58661b3ba9 loader: only use vidc_install_font() with vga controller
If the controller does not support VGA, we should not
touch VGA registers.
2021-01-05 00:49:18 +02:00
Toomas Soome
8c1dda1171 loader: vidc_load_palette() does not need arguments
Remove left over argument from vidc_load_palette.
2021-01-05 00:49:08 +02:00
Toomas Soome
3630506b9d loader: implement framebuffer console
Draw console on efi.
Add vbe framebuffer for BIOS loader (vbe off, vbe on, vbe list,
vbe set xxx).
autoload font (/boot/fonts) based on resolution and font size.
Add command loadfont (set font by file) and
variable screen.font (set font by size). Pass loaded font to kernel.

Export variables:
screen.height
screen.width
screen.depth

Add gfx primitives to draw the screen and put png image on the screen.
Rework menu draw to iterate list of consoles to enamble device specific
output.

Probably something else I forgot...

Relnotes: yes
Differential Revision: https://reviews.freebsd.org/D27420
2021-01-02 21:41:36 +02:00
Kyle Evans
ac5f382a9d stand: properly declare subdir deps or .WAIT, do parallel build
buildworld already runs the stand build in parallel[1], so make it easier to
identify ordering issues by properly establishing dependencies or adding
.WAIT where needed.

Everything in stand/ relies on libsa, either directly or indirectly, because
libsa build is where the stand headers get installed and it gets linked in
most places.

Interpreters depend on their libs, machine dirs usually depend on top-level
libs that are getting built and at least one of the interpreter flavors.

For i386, order btx/libi386/libfirewire before everything else using a
big-ol-.WAIT hammer. btx is the most common dependency, but the others are
used sporadically. This seems to be where the race reporting on the mailing
list is- AFAICT, the following sequence is happening:

1.) One of the loaders gets built based on stale btx/btxldr
2.) btx/btxldr gets rebuilt
3.) installworld triggers loader rebuild because btx was rebuilt after

This seems like the most plausible explanation, as they've verified system
time and timestamps.

While we're here, let's switch stand/ over to a completely parallel build so
we can work out these kinds of issues in isolation rather than in the middle
of a larger build.

Reviewed by:	bdragon, sjg, tsoome
Tested by:	bdragon (-j1024, no failures, significant speed improvement)
Differential Revision:	https://reviews.freebsd.org/D23411
2020-12-31 11:15:45 -06:00
Toomas Soome
fc7cf7241f zfsboot: add prototype for main()
Some compilers are complaining about missing prototype.

PR:		251150
Reported by:	markiyan.kushnir@gmail.com
2020-11-15 14:04:27 +00:00
Toomas Soome
c50f409c15 loader: edd_device_path_v3 is too small
The EDD v3[1], see table 13, page 33, does define device path as double
qword, that is, 16 bytes, we have only qword.

Also remove edd_device_path_v4 and edd_params_v4 because those are not used,
and there is no size difference in v3 versus v4.

[1] http://www.t13.org/documents/UploadedDocuments/docs2004/d1572r3-EDD3.pdf

MFC after:	2 weeks
2020-10-12 09:34:50 +00:00
Toomas Soome
1dc762d4fb loader: fix non-zfs build
We can not include zfs headers while building without zfs.

Reported by:	Oscar Holmlund
2020-09-23 19:15:22 +00:00
Warner Losh
a5ebda464e Work around cp breakage in current from last week
There was a small window cp was broken. Work around this by using :>
instead of cp /dev/null. Ideally, we'd keep the cp /dev/null in the
build as a regression test, but doing so breaks people that upgraded
during the cp breakage and this is simpler than bootstrapping a
working cp since there's no good __FreeBSD_version sign posts for
that.

Suggested by: lots of people
Too stubborn for his own good: imp
2020-09-23 01:04:25 +00:00
Toomas Soome
e307eb94ae loader: zfs should support bootonce an nextboot
bootonce feature is temporary, one time boot, activated by
"bectl activate -t BE", "bectl activate -T BE" will reset the bootonce flag.

By default, the bootonce setting is reset on attempt to boot and the next
boot will use previously active BE.

By setting zfs_bootonce_activate="YES" in rc.conf, the bootonce BE will
be set permanently active.

bootonce dataset name is recorded in boot pool labels, bootenv area.

in case of nextboot, the nextboot_enable boolean variable is recorded in
freebsd:nvstore nvlist, also stored in boot pool label bootenv area.
On boot, the loader will process /boot/nextboot.conf if nextboot_enable
is "YES", and will set nextboot_enable to "NO", preventing /boot/nextboot.conf
processing on next boot.

bootonce and nextboot features are usable in both UEFI and BIOS boot.

To use bootonce/nextboot features, the boot loader needs to be updated on disk;
if loader.efi is stored on ESP, then ESP needs to be updated and
for BIOS boot, stage2 (zfsboot or gptzfsboot) needs to be updated
(gpart or other tools).

At this time, only lua loader is updated.

Sponsored by:	Netflix, Klara Inc.
Differential Revision:	https://reviews.freebsd.org/D25512
2020-09-21 09:01:10 +00:00
Gleb Smirnoff
c7dd069c70 When we have a command returned by zfs_nextboot() that is longer
than command in the loader.conf, the latter needs to be nul terminated,
otherwise garbage trailer left from zfs_nextboot() will be passed to
parse_cmd() together with loader.conf command.

While here, reset cmd to empty string if read() returns error.

Reviewed by:	tsoome
2020-08-20 20:31:47 +00:00
Mariusz Zaborski
277f38abff zfs: add an option to the bootloader to rewind the ZFS checkpoint
The checkpoints are another way of keeping the state of ZFS.
During the rewind, the pool has to be exported.
This makes checkpoints unusable when using ZFS as root.
Add the option to rewind the ZFS checkpoint at the boot time.
If checkpoint exists, a new option for rewinding a checkpoint will appear in
the bootloader menu.
We fully support boot environments.
If the rewind option is selected, the boot loader will show a list of
boot environments that existed before the checkpoint.

Reviewed by:	tsoome, allanjude, kevans (ok with high-level overview)
Differential Revision:	https://reviews.freebsd.org/D24920
2020-08-18 19:48:04 +00:00
Alex Richardson
1a18ab420b Allow overriding the tool used for stripping binaries
Since the make variable STRIP is already used for other purposes, this
uses STRIPBIN (which is also used for the same purpose by install(1).
This allows using LLVM objcopy to strip binaries instead of the in-tree
elftoolchain objcopy. We make use of this in CheriBSD since passing
binaries generated by our toolchain to elftoolchain strip sometimes results
in assertion failures.

This allows working around https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=248516
by specifying STRIPBIN=/path/to/llvm-strip

Obtained from:	CheriBSD
Reviewed By:	emaste, brooks
Differential Revision: https://reviews.freebsd.org/D25988
2020-08-11 16:46:27 +00:00
Simon J. Gerraty
f2be828f97 Revert that! 2020-07-19 23:56:19 +00:00
Simon J. Gerraty
e17f5b1d30 Oops missed Makefile.config 2020-07-19 23:54:00 +00:00
Toomas Soome
97dd57e66c loader: cstyle cleanup
No functional changes intended.

Sponsored by:	Netflix, Klara Inc.
2020-07-19 06:59:09 +00:00
Toomas Soome
de776da323 loader: implement GELI writes
Bug: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=247482

This patch is based on initial work from allanjude.

PR:		247482
Obtained from:	https://reviews.freebsd.org/D10236
Differential Revision:	https://reviews.freebsd.org/D25605
2020-07-11 06:51:42 +00:00
Baptiste Daroussin
5b990a9463 Revert r362466
Such change should not have happen without prior discussion and review.

With hat:	transitioning core
2020-06-22 07:46:24 +00:00
Hans Petter Selasky
7747001b12 Improve wording to be more precise and clear.
No functional change intended.

s/Master Boot/Main Boot/ (also called MBR)

MFC after:	1 week
Sponsored by:	Mellanox Technologies
2020-06-21 13:34:08 +00:00
Toomas Soome
3830659e99 loader: create single zfs nextboot implementation
We should have nextboot feature implemented in libsa zfs code.
To get there, I have created zfs_nextboot() implementation based on
two sources, our current simple textual string based approach with added
structured boot label PAD structure from OpenZFS.

Secondly, all nvlist details are moved to separate source file and
restructured a bit. This is done to provide base support to add nvlist
add/update feature in followup updates.

And finally, the zfsboot/gptzfsboot disk access functions are swapped to use
libi386 and libsa.

Sponsored by:	Netflix, Klara Inc.
Differential Revision:	https://reviews.freebsd.org/D25324
2020-06-20 06:23:31 +00:00
Toomas Soome
aeea57c709 loader: vidc_init should also erase the screen
Inject \e[J to erase the initial loader screen. We have two options,
find where out cursor is and use BIOS scroll for data from boot2 or erase the
display and start from origin. Erasing the screen is easier and we also
get the screen buffer initialized.

Sponsored by:	Netflix, Klara Inc.
2020-06-17 08:08:57 +00:00
Toomas Soome
6469d2b422 boot2: need to expand tab output and mask getchar
The BIOS ouput char function does not expand tab.
Mask getchar with 0xFF.

Sponsored by:	Netflix, Klara Inc.
2020-06-16 20:35:00 +00:00
Toomas Soome
dfdeb45425 zfsboot: cstyle cleanup
No functional changes intended.

Sponsored by:	Netflix, Klara Inc.
2020-06-14 20:57:24 +00:00
Eric van Gyzen
fac6dee9eb Remove tests for obsolete compilers in the build system
Assume gcc is at least 6.4, the oldest xtoolchain in the ports tree.
Assume clang is at least 6, which was in 11.2-RELEASE.  Drop conditions
for older compilers.

Reviewed by:	imp (earlier version), emaste, jhb
MFC after:	2 weeks
Sponsored by:	Dell EMC Isilon
Differential Revision:	https://reviews.freebsd.org/D24802
2020-05-12 15:22:40 +00:00
Alex Richardson
87c5e64c64 Avoid using non-portable dd conv=osync option
The same thing can be achieved using the bs=$SIZE and conv=sync options.
Unlike conv=osync, conv=sync is support on both Linux and macOS.

Reviewed By:	imp, jhb (earlier version)
Obtained from:	CheriBSD
Differential Revision: https://reviews.freebsd.org/D24788
2020-05-11 08:40:33 +00:00
Mark Johnston
5c17092568 Stop setting PG_U in bootstrap mappings.
These mappings are never visible to userspace as they get replaced when
the amd64 pmap is bootstrapped, but there is no need to set PG_U in the
first place.

Reviewed by:	alc, kib
MFC after:	1 week
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D24547
2020-04-24 13:53:40 +00:00
Toomas Soome
27bae6150a loader: remove libsa/crc32.c and use version from zlib
we have crc32(const void *, size_t) in libsa. Unfortunately zlib has
crc32(long, const unigned char *, unsigned) and we have conflict.

Since we do build libsa with zlib, we can use zlib version instead.

Reviewed by:	allanjude
Differential Revision:	https://reviews.freebsd.org/D24068
2020-03-19 21:05:11 +00:00
Toomas Soome
7f505e7f79 test if port does exist via using scratch register
The SCR, scratch register was not present on the 8250 and 8250B UART, so we
can use to test if we actually do have serial port.

We need this test because some systems will get long delays while attempting
to write to non-existing port and this will slow down the console IO
to extreme.

MFC after:	1 week
2020-03-12 06:45:08 +00:00
Simon J. Gerraty
91cd69ee2c veloader use vectx API for kernel and modules
The vectx API, computes the hash for verifying a file as it is read.
This avoids the overhead of reading files twice - once to verify, then
again to load.

For doing an install via loader, avoiding the need to rewind
large files is critical.

This API is only used for modules, kernel and mdimage as these are the
biggest files read by the loader.
The reduction in boot time depends on how expensive the I/O is
on any given platform.  On a fast VM we see 6% improvement.

For install via loader the first file to be verified is likely to be the
kernel, so some of the prep work (finding manifest etc) done by
verify_file() needs to be factored so it can be reused for
vectx_open().

For missing or unrecognized fingerprint entries, we fail
in vectx_open() unless verifying is disabled.

Otherwise fingerprint check happens in vectx_close() and
since this API is only used for files which must be verified
(VE_MUST) we panic if we get an incorrect hash.

Reviewed by:	imp,tsoome
MFC after:	1 week
Sponsored by:	Juniper Networks
Differential Revision:	https://reviews.freebsd.org//D23827
2020-03-09 16:02:54 +00:00
Simon J. Gerraty
afc571b1a6 veloader use vectx API for kernel and modules
The vectx API, computes the hash for verifying a file as it is read.
This avoids the overhead of reading files twice - once to verify, then
again to load.

For doing an install via loader, avoiding the need to rewind
large files is critical.

This API is only used for modules, kernel and mdimage as these are the
biggest files read by the loader.
The reduction in boot time depends on how expensive the I/O is
on any given platform.  On a fast VM we see 6% improvement.

For install via loader the first file to be verified is likely to be the
kernel, so some of the prep work (finding manifest etc) done by
verify_file() needs to be factored so it can be reused for
vectx_open().

For missing or unrecognized fingerprint entries, we fail
in vectx_open() unless verifying is disabled.

Otherwise fingerprint check happens in vectx_close() and
since this API is only used for files which must be verified
(VE_MUST) we panic if we get an incorrect hash.

Reviewed by:	imp,tsoome
MFC after:	1 week
Sponsored by:	Juniper Networks
Differential Revision:	https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
Dimitry Andric
a2037dba7e Link stand/i386 components using a linker script
LLD 10.0.0 changed the behavior of the -Ttext option, so that using
-Ttext=0x0 now causes linking of the loaders to fail with:

ld: error: output file too large: 18446744073707016908 bytes

I reported this in https://bugs.llvm.org/show_bug.cgi?id=44715, and
initially reverted the upstream change in r357259 to work around it.

However, after some discussion with Fangrui Song in the upstream ticket,
I think we can classify this as an unfortunate interaction between using
-Ttext=0 in combination with --no-rosegment.  (We added the latter
in r332090, because btxld does not correctly handle input with more
than 2 PT_LOAD segments.)

Fangrui suggested to use a linker script instead, and Warner was already
attempting this in r305353, but had to revert it due to "crypto-using
boot problems" (not sure what those were :).

This review updates the stand/i386/boot.ldscript to handle more
sections, inserts some symbols like _edata and such that we use in
libsa, and also discards any .interp section.

It uses ORG which is defined on the linker command line using
--defsym ORG=value to set the start of all the sections.

Reviewed by:	imp
MFC after:	2 weeks
Differential Revision: https://reviews.freebsd.org/D23952
2020-03-04 21:01:22 +00:00
Toomas Soome
ca25195805 loader: crc32 is provided by libsa
Seems like leftover from moving crc32.c to libsa.
2020-03-04 18:38:09 +00:00
Warner Losh
a64f0b83e5 Move smbios.c to libsa.
smbios used to be an i386 only kinda weird quirk to the x86
architecture. But UEFI picked it up, dusted it off and now it's many
other locations. Make it base technology by moving it to libsa and
fixing up the compliation. The code has issues with unaligned access
still, but that will be addressed in a followup commit.

Differential Revision: https://reviews.freebsd.org/D23660
2020-02-20 00:46:22 +00:00
Warner Losh
ed2a65769a Create ptov() function.
Create a ptov() function. It's basically the same as the btx PTOV
macro, but works everywhere. smbios needs this to translate addresses,
but the translation differs between BIOS booting and EFI booting. Make
it a function so one smbios.o can be used everywhere. Provide
definitions for it in the two loaders affected.

Differential Revision: https://reviews.freebsd.org/D23660
2020-02-20 00:46:16 +00:00
Toomas Soome
7503958e4c zfsboot: vdev_read_pad2 does allocate buffer with wrong size
vdev_read_pad2() does read VDEV_PAD_SIZE of data, and will copy size bytes
of it, hence, we need buffer of VDEV_PAD_SIZE bytes.

Issue introduced in r357497.

Reported by:	se
2020-02-05 11:02:00 +00:00
Toomas Soome
a1746b2583 vdev_read_pad2: freeing wrong pointer
Shoud free previously allocated tmp buffer instead. Issue introduced in
r357497.

Reported by:	rpokala
2020-02-05 05:42:42 +00:00
Toomas Soome
4d297e7035 loader: rewrite zfs reader zap code to use malloc
First step on removing zfs_alloc.

Reviewed by:	delphij
Differential Revision:	https://reviews.freebsd.org/D23433
2020-02-04 07:37:55 +00:00
Toomas Soome
b0af1e20e6 loader: bc_add can not use any other probes than ah=0x4b
CD boot is broken for some systems since bioscd and biosdisk merge. The issue is that we can not use anything else than int 13 ah=0x4b to query cd information.

The patch does restore the same probe as was originally used in bioscd.c. Additionally extra buffer padding is used to avoid memory corruption caused by some systems.

PR:		234031
Reported by:	ultramage and others
MFC after:	1 day
2020-02-03 11:33:33 +00:00
Dimitry Andric
ae55765309 Work around assembler error from clang 10.0.0 in gptboot:
stand/i386/gptboot/gptldr.S:141:3: error: value of 36878 is too large for field of 2 bytes.
  jmp MEM_JMP # Start BTX
  ^

Use the same construct as in stand/i386/boot2/boot1.S, which ensures the
jump distance does not become too large.

MFC after:	3 days
2020-01-28 21:41:37 +00:00
Konstantin Belousov
2ee49fac82 Add support for Hygon Dhyana Family 18h processor.
As a new x86 CPU vendor, Chengdu Haiguang IC Design Co., Ltd (Hygon)
is a joint venture between AMD and Haiguang Information Technology Co.,
Ltd., aims at providing x86 processors for China server market.

The first generation Hygon processor(Dhyana) shares most architecture
with AMD's family 17h, but with different CPU vendor ID("HygonGenuine")
and PCI vendor ID(0x1d94) and family series number 18h(Hygon negotiated
with AMD to confirm that only Hygon use family 18h).

To enable Hygon Dhyana support in FreeBSD, add new definitions
HYGON_VENDOR_ID("HygonGenuine") and X86_VENDOR_HYGON(0x1d94) to identify
Hygon Dhyana CPU.

Initialize the CPU features(topology, local APIC ext, MSI, TSC, hwpstate,
MCA, DEBUG_CTL, etc) for amd64 and i386 mode by sharing the code path of
AMD family 17h.

The changes have been applied on FreeBSD 13.0-CURRENT and tested
successfully on Hygon Dhyana processor.

References:
[1] Linux kernel patches for Hygon Dhyana, merged in 4.20:

https://git.kernel.org/tip/c9661c1e80b609cd038db7c908e061f0535804ef

[2] MSR and CPUID definition:

https://www.amd.com/system/files/TechDocs/54945_PPR_Family_17h_Models_00h-0Fh.pdf

Submitted by:	Pu Wen <puwen@hygon.cn>
MFC after:	1 week
Differential revision:	https://reviews.freebsd.org/D23163
2020-01-21 13:22:35 +00:00
Toomas Soome
aaeffe5b70 Backout 356693. The libsa malloc does provide necessary alignment and
memalign by 4 will reduce alignment for some platforms. Thanks for Ian for
pointing this out.
2020-01-13 20:02:27 +00:00