jsg [Wed, 25 Jan 2023 02:01:03 +0000 (02:01 +0000)]
drm/amd/display: Calculate output_color_space after pixel encoding adjustment
From Joshua Ashton
87e84d0ff97cf8d508910d793866083a918d8cab in linux-6.1.y/6.1.8
79601b894849cb6f6d6122e6590f1887ac4a66b3 in mainline linux
jsg [Wed, 25 Jan 2023 01:58:44 +0000 (01:58 +0000)]
drm/amd/display: Fix set scaling doesn's work
From hongao
032cebdff71e92d6aa6fd2a8550eb7c9898553e0 in linux-6.1.y/6.1.8
040625ab82ce6dca7772cb3867fe5c9eb279a344 in mainline linux
jsg [Wed, 25 Jan 2023 01:56:39 +0000 (01:56 +0000)]
drm/i915: Remove unused variable
From Nirmoy Das
92c39d6995923103f952c94ba0e9b859e182d2a8 in linux-6.1.y/6.1.8
2293a73ad4f3b6c37c06713ff1b67659d92ef43d in mainline linux
jsg [Wed, 25 Jan 2023 01:55:14 +0000 (01:55 +0000)]
drm/i915: Allow switching away via vga-switcheroo if uninitialized
From Thomas Zimmermann
4397bcbd729dd79af4cc53615bdca59f80e74643 in linux-6.1.y/6.1.8
a273e95721e96885971a05f1b34cb6d093904d9d in mainline linux
jsg [Wed, 25 Jan 2023 01:53:30 +0000 (01:53 +0000)]
drm/i915/display: Check source height is > 0
From Drew Davenport
597747eb58a7c8deb131f89aad7e136dbb63ef7a in linux-6.1.y/6.1.8
8565c502e7c156d190d8e6d36e443f51b257f165 in mainline linux
jsg [Wed, 25 Jan 2023 01:51:59 +0000 (01:51 +0000)]
drm/i915: re-disable RC6p on Sandy Bridge
From Sasa Dragic
38a9b17d3b58a3163bc031e96c94fc2d800e02b7 in linux-6.1.y/6.1.8
67b0b4ed259e425b7eed09da75b42c80682ca003 in mainline linux
jsg [Wed, 25 Jan 2023 01:50:20 +0000 (01:50 +0000)]
drm/amdgpu: Correct the power calcultion for Renior/Cezanne.
From jie1zhan
8dabe4e0daa999b2f11e354a0f67934420e095b4 in linux-6.1.y/6.1.8
c7bae4aaa5609c1fa9761c35dbcc5fcc92915222 in mainline linux
jsg [Wed, 25 Jan 2023 01:48:33 +0000 (01:48 +0000)]
drm/amdgpu: allow multipipe policy on ASICs with one MEC
From Lang Yu
d693fdf73eb296081523e18d1ac3bc38787e7cab in linux-6.1.y/6.1.8
dc88063b87775971be564d79dc1b05f7b8b5c135 in mainline linux
jsg [Wed, 25 Jan 2023 01:46:56 +0000 (01:46 +0000)]
drm/amdgpu: fix amdgpu_job_free_resources v2
From Christian Koenig
af4a3c274a92ee0545d65b9102b398fb75d7cc9a in linux-6.1.y/6.1.8
74ea8e78ab349514c9f4df0be1189d91267d750d in mainline linux
deraadt [Tue, 24 Jan 2023 23:16:14 +0000 (23:16 +0000)]
permit -execute-only on ppc64 (not default)
jan [Tue, 24 Jan 2023 22:35:46 +0000 (22:35 +0000)]
consolidate mbuf header parsing on device driver layer
with tweaks from mvs@, mpi@ and dlg@
ok mvs@, dlg@
jcs [Tue, 24 Jan 2023 21:48:04 +0000 (21:48 +0000)]
Parallels Desktop 17 emulates an azalia audio device, add it
azalia0 at pci0 dev 1 function 0 "Intel 82801I HD Audio" rev 0x00: msi
azalia0: codecs: Parallels/0x0001
ok patrick
claudio [Tue, 24 Jan 2023 20:06:16 +0000 (20:06 +0000)]
Refactor nd6_options() a bit more. Rewrite the loop to be a proper loop
and not some endless loop with some gotos.
OK kn@
kettenis [Tue, 24 Jan 2023 16:51:05 +0000 (16:51 +0000)]
Make sure pmap_page_protect() does the right thing for execute-only
mappings and enforce this with a KASSERT like we do on amd64. Bring
the pmap_protect() inline in line with the amd64 version.
ok miod@, deraadt@
stsp [Tue, 24 Jan 2023 16:18:22 +0000 (16:18 +0000)]
Fix a crash in iwx(4) when connecting to WEP networks via ifconfig join.
In if_iwx.c r1.63 / CVS commit ID QzTzJtyPl10L63Yi, I fixed a problem with
connecting to WPA1/TKIP networks. This change introduced a deference of
the 'ni' pointer passed to ic_set_key(). This pointer will be passed in
as NULL in certain situations, such as when WEP is used. Add the missing
NULL check.
reported by and fix tested by matthieu@
job [Tue, 24 Jan 2023 15:52:46 +0000 (15:52 +0000)]
Add ASPA to STANDARDS section
job [Tue, 24 Jan 2023 15:50:10 +0000 (15:50 +0000)]
Fix typo
krw [Tue, 24 Jan 2023 15:47:10 +0000 (15:47 +0000)]
Use getpartno() in editor_add(), enhancing getpartno() to find
the default partition and allow selection of partitions up to
MAXPARTITION.
No intentional functional change.
claudio [Tue, 24 Jan 2023 14:14:15 +0000 (14:14 +0000)]
Implmement `bgpctl show rib avs invalid` and firends.
This will show all invalid ASPA paths.
OK tb@
claudio [Tue, 24 Jan 2023 14:13:11 +0000 (14:13 +0000)]
Implement filter and control message matching for ASAP.
This adds avs (ASPA validation state) which can be 'unknown', 'valid'
or 'invalid'. It behaves similar to ovs but the ASPA validation state
of paths from iBGP sessions is 'unknown' and the role of the ebgp session
is important to get the right validation state.
OK tb@
jca [Tue, 24 Jan 2023 13:31:47 +0000 (13:31 +0000)]
sync
jca [Tue, 24 Jan 2023 13:30:44 +0000 (13:30 +0000)]
Provide /dev/ujoy/[0-3]
ok deraadt@ miod@
jca [Tue, 24 Jan 2023 13:29:51 +0000 (13:29 +0000)]
Enable access to usb(4), ugen(4), ulpt(4), ucom(4) and ujoy(4)
Input kettenis@, ok deraadt@ miod@
jca [Tue, 24 Jan 2023 13:28:31 +0000 (13:28 +0000)]
Hook up the same USB device drivers as arm64/conf/GENERIC
Input from kettenis@, ok deraadt@ miod@
claudio [Tue, 24 Jan 2023 11:32:13 +0000 (11:32 +0000)]
Adjust various test results to the new bgpctl show rib output including
ASPA validation state.
OK tb@
claudio [Tue, 24 Jan 2023 11:31:13 +0000 (11:31 +0000)]
Adjust the ASPA test to match the new code which is AFI independent.
OK tb@
claudio [Tue, 24 Jan 2023 11:29:34 +0000 (11:29 +0000)]
Print the ASPA validation state (avs) in various show rib outputs.
OK tb@
claudio [Tue, 24 Jan 2023 11:28:41 +0000 (11:28 +0000)]
Implement ASPA validation and reload logic on ASPA set changes.
For this use the validation state (vstate) in struct prefix and
struct filterstate to store both the ASPA and ROA validity.
Introduce helper functions to set and get the various states for
struct prefix and make sure struct filterstate is also setup properly.
Change the ASPA state in rde_aspath to be AFI/AID and role independent
by storing all 4 possible outcomes. Also add a ASPA generation count
which is used to update the rde_aspath ASPA state cache on reloads.
Rework the rde_aspa.c code to be AFI/AID and role independent. Doing
this for roles is trivial but AFI switch goes deep and is so unnecessary.
The reload is combined with the ROA reload logic and renamed to RPKI
softreload.
OK tb@
job [Tue, 24 Jan 2023 09:48:57 +0000 (09:48 +0000)]
Check return value of X509_digest
OK tb@
miod [Tue, 24 Jan 2023 07:26:33 +0000 (07:26 +0000)]
Add and use a specific routine to fetch instructions from userland when
doing floating-point emulation, as copyin will fail on exec-only mappings
now.
ok deraadt@ kettenis@
jsg [Tue, 24 Jan 2023 01:40:22 +0000 (01:40 +0000)]
match unknown ATI display devices as amdgpu in fw_update
For recent devices amdgpu matches via the hardware ip discovery table,
not with a table of pci vendor and product ids.
So amdgpu_devlist.h and pcidevs do not cover all devices that amdgpu
may match.
in dmesg amdgpu with an unknown product takes the form:
ramdisk kernel, bios/mbr:
vga1 at pci12 dev 0 function 0 vendor "ATI", unknown product 0x687f rev 0xc3
ramdisk kernel, efi or non-x86 arch:
vendor "ATI", unknown product 0x687f (class display subclass VGA, rev 0x03) at pci12 dev 0 function 0 not configured
non-ramdisk kernel:
amdgpu0 at pci12 dev 0 function 0 vendor "ATI", unknown product 0x687f rev 0xc3
ok deraadt@ on an earlier version
deraadt [Tue, 24 Jan 2023 00:16:08 +0000 (00:16 +0000)]
oops, a silly typo
deraadt [Tue, 24 Jan 2023 00:14:30 +0000 (00:14 +0000)]
uvm_map_extract() UVM_EXTRACT_FIXPROT alias mappings are only used for
read/write operations, so mask out PROT_EXEC to avoid creating an pointless
exec mapping in the kernel.
We probably need this masking upon minprot (for the non-UVM_EXTRACT_FIXPROT
case) also, but I haven't done a test yet.
ok kettenis
deraadt [Tue, 24 Jan 2023 00:12:03 +0000 (00:12 +0000)]
ptrace reads/writes memory using uvm_io, which generates an temporary
alias mapping using uvm_map_extract. With xonly now operational, this
alias mapping is created with minprot, which for text will be xonly, and
the kernel cannot read it (unless the architecture has implied read for
exec from kernel pov).
Pass UVM_IO_FIXPROT to create the alias with maxprot instead.
ok kettenis
miod [Mon, 23 Jan 2023 19:31:41 +0000 (19:31 +0000)]
Make sure pmap_extract() doesn't incorrectly return success when invoked
on behalf of an invalid but not page-aligned address.
ok kettenis@
mvs [Mon, 23 Jan 2023 18:35:13 +0000 (18:35 +0000)]
Remove "sb == &so->so_rcv || sb == &so->so_snd" assertion from
sb_notify() and sbspace(). Now it's overkilling.
ok bluhm@
mvs [Mon, 23 Jan 2023 18:34:24 +0000 (18:34 +0000)]
Move SS_ISSENDING flag to `sb_state'. It should belong to the send
buffer as the SS_CANTSENDMORE flag.
ok bluhm@
mvs [Mon, 23 Jan 2023 18:33:34 +0000 (18:33 +0000)]
In somove() rename `state' variable to `rcvstate' to make code more
readable. No functional changes.
Proposed by and ok bluhm@
jsing [Mon, 23 Jan 2023 18:26:56 +0000 (18:26 +0000)]
Change include from _internal_s2n_bignum.h to s2n_bignum_internal.h.
jsing [Mon, 23 Jan 2023 18:25:36 +0000 (18:25 +0000)]
Include the ISC license from s2n-bignum's LICENSE file.
jsing [Mon, 23 Jan 2023 18:22:15 +0000 (18:22 +0000)]
Bring in various s2n-bignum functions for amd64.
This brings in bignum_add(), bignum_cmadd(), bignum_cmul(), bignum_mul()
and bignum_sub(), along with bignum_{mul,sqr}_4_8_alt() and
bignum_{mul,sqr}_8_16_alt().
Discussed with tb@
kn [Mon, 23 Jan 2023 14:49:16 +0000 (14:49 +0000)]
more comments, space and -- nits
jsing [Mon, 23 Jan 2023 12:17:57 +0000 (12:17 +0000)]
Move bn_mul_add_words() and bn_mul_words() from bn_asm.c to bn_mul.c.
These are wrapped with #ifndef HAVE_BN_ADD_MUL_WORDS/HAVE_BN_MUL_WORDS,
which are defined for architectures that provide their own assembly
versions.
jsing [Mon, 23 Jan 2023 12:09:06 +0000 (12:09 +0000)]
Move bn_sqr_words from bn_asm.c to bn_sqr.c.
This is wrapped with #ifndef HAVE_BN_SQR_WORDS, which is then defined for
architectures that provide their own assembly versions.
jsing [Mon, 23 Jan 2023 12:02:48 +0000 (12:02 +0000)]
Move bn_div_words from bn_asm.c to bn_div.c.
This is wrapped with #ifndef HAVE_BN_DIV_WORDS, which are defined for
architectures that provide their own assembly versions.
jsing [Mon, 23 Jan 2023 10:34:21 +0000 (10:34 +0000)]
Include bn_arch.h now that we're using defines from it.
jsing [Mon, 23 Jan 2023 10:31:03 +0000 (10:31 +0000)]
Move bn_add_words() and bn_sub_words from bn_asm.c to bn_add.c.
These are wrapped with #ifndef HAVE_BN_ADD_WORDS/HAVE_BN_SUB_WORDS, which
are defined for architectures that provide their own assembly versions.
jsg [Mon, 23 Jan 2023 10:14:15 +0000 (10:14 +0000)]
regen
jsg [Mon, 23 Jan 2023 10:13:46 +0000 (10:13 +0000)]
add amdgpu device ids newly matched with 6.1 drm
another Cyan Skillfish and Aldebaran/MI200
nicm [Mon, 23 Jan 2023 09:38:03 +0000 (09:38 +0000)]
Regen.
nicm [Mon, 23 Jan 2023 09:36:39 +0000 (09:36 +0000)]
Add Backtab key, based on a change from Crystal Kolipe kolipe.c at
exoticsilicon dot com.
ok miod
nicm [Mon, 23 Jan 2023 09:33:51 +0000 (09:33 +0000)]
Too many \s in example, GitHub issue 3445.
jsg [Mon, 23 Jan 2023 05:19:44 +0000 (05:19 +0000)]
regen
jsg [Mon, 23 Jan 2023 05:19:10 +0000 (05:19 +0000)]
change ATI "Yellow Carp" string to "Rembrandt"
Yellow Carp is the codename used by drm and linux amdgpu firmware.
In product specifications AMD uses Rembrandt.
Mesa renamed last year in
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/16604
Ryzen 6000 Series / Rembrandt graphics are marketed as Radeon 660M and
Radeon 680M. These share the same device id of 0x1681.
Ryzen 7035 Series / Rembrandt-R also use Radeon 660M and Radeon 680M.
It is unclear if they use a device id of 0x1681 or 0x164d.
jsg [Mon, 23 Jan 2023 03:01:52 +0000 (03:01 +0000)]
add a placeholder for GC 11.0.4 device ids
GC 11.0.4 support will be backported to linux 6.1.y at the request of
Mario Limonciello of AMD on the linux stable list.
Known to be an APU from amdgpu_discovery_set_ip_blocks().
Firmware not currently in linux-firmware git.
yasuoka [Sun, 22 Jan 2023 23:05:51 +0000 (23:05 +0000)]
Fix pf_anchor_stackframe commit to revert pf rule matching to the
previous behavior that stops when any rule matches within quick
anchors.
ok sasha kn
tb [Sun, 22 Jan 2023 22:38:23 +0000 (22:38 +0000)]
Fix list of possible values of MODRUBY_REV
With Ruby 2.7 in the attic, MODRUBY_REV can no longer be set to 2.7.
It has been possible to set it to 3.2 for a while.
dv [Sun, 22 Jan 2023 22:18:40 +0000 (22:18 +0000)]
vmd(8): don't remove known vm's from the config on error.
Multiple error paths, specifically the one related to if a guest
cannot allocate memory at start, resulted in a known vm (via
vm.conf(5)) being removed from the vm list. Adjust the error paths
to check if the failing vm is defined in the config before tearing
it down.
Tested with help from beck@ and Mischa Peters.
ok beck@
jeremy [Sun, 22 Jan 2023 20:57:21 +0000 (20:57 +0000)]
Update documentation for removal of ruby27 FLAVOR
krw [Sun, 22 Jan 2023 19:57:25 +0000 (19:57 +0000)]
Fix incorrect comparison. Restores ability to edit 'p'.
cheloha [Sun, 22 Jan 2023 18:36:38 +0000 (18:36 +0000)]
gptimer(4): remove dead MD microtime(9) implementation
anton [Sun, 22 Jan 2023 16:38:36 +0000 (16:38 +0000)]
Cope with xonly on amd64 by moving the testfly routine to the rodata
section.
kettenis [Sun, 22 Jan 2023 13:14:21 +0000 (13:14 +0000)]
Simplify the (so far unused) apm_record_event() function a bit and use it
to send a resume event to userland when we come out of suspend.
ok deraadt@
mvs [Sun, 22 Jan 2023 12:05:44 +0000 (12:05 +0000)]
Move SS_CANTRCVMORE and SS_RCVATMARK bits from `so_state' to `sb_state' of
receive buffer. As it was done for SS_CANTSENDMORE bit, the definition
kept as is, but now these bits belongs to the `sb_state' of receive
buffer. `sb_state' ored with `so_state' when socket data exporting to the
userland.
ok bluhm@
miod [Sat, 21 Jan 2023 19:39:28 +0000 (19:39 +0000)]
Remove dead/unused prototypes
mvs [Sat, 21 Jan 2023 17:35:01 +0000 (17:35 +0000)]
Introduce `rt_lock' rwlock(9) and use it instead of kernel lock to
serialize arpcache() and arpresolve(). In fact, net stack already has
sleep points, so the rwlock(9) is better here because we avoid
intersection with the rest of kernel locked paths. Also this new lock
assumed to use to route layer protection instead of netlock.
Hrvoje Popovski had tested this diff and found no visible performance
impact.
ok bluhm@
jsing [Sat, 21 Jan 2023 17:30:50 +0000 (17:30 +0000)]
Enable bn_sqr() on amd64.
ok tb@
jsing [Sat, 21 Jan 2023 17:29:56 +0000 (17:29 +0000)]
Provide an implementation of bn_sqr() that calls s2n-bignum's bignum_sqr().
ok tb@
jsing [Sat, 21 Jan 2023 16:35:18 +0000 (16:35 +0000)]
Fix include.
jsing [Sat, 21 Jan 2023 16:30:57 +0000 (16:30 +0000)]
Include the ISC license from s2n-bignum's LICENSE file.
jsing [Sat, 21 Jan 2023 16:29:52 +0000 (16:29 +0000)]
Bring in s2n-bignum's bignum_sqr() for amd64.
ok tb@
krw [Sat, 21 Jan 2023 16:20:25 +0000 (16:20 +0000)]
editor_resize(), editor_change(), editor_modify() and
editor_name() use identical logic to choose a partition to act
on, with some minor verbiage differences.
Extract getpartno() to contain/tighten this logic and shrink the
editor_* functions by using it.
No intentional functional change but for verbiage tweaks.
jsing [Sat, 21 Jan 2023 15:53:54 +0000 (15:53 +0000)]
Include the ISC license from s2n-bignum's LICENSE file.
jsing [Sat, 21 Jan 2023 15:51:17 +0000 (15:51 +0000)]
Bring in the internal and "public" headers for s2n-bignum.
s2n-bignum provides a collection of bignum routines that are written in
pure machine code. Each function is written in constant-time style and
has a formal proof. We intend on making use of these for libcrypto's
bignum implementation on aarch64 and amd64.
ok tb@
jsing [Sat, 21 Jan 2023 15:40:13 +0000 (15:40 +0000)]
Refactor BN_mul().
This splits BN_mul() into two parts, one of which is a separate bn_mul()
function. This makes the code more readable and managable, while also
providing a better entry point for assembly optimisation. A separate
bn_mul() is provided for the BN_RECURSION implementation, to reduce
complexity.
This also enables bn_mul_comba4() for four word long bignums - this was
disabled for unknown reasons.
ok tb@
deraadt [Sat, 21 Jan 2023 15:38:05 +0000 (15:38 +0000)]
enable --exec-only as default on AMD64. There will be a some ports fallout
for a little while more (a few more data tables inside text), but they are
being handled quickly
ok tb naddy
jsing [Sat, 21 Jan 2023 14:10:46 +0000 (14:10 +0000)]
Refactor BN_sqr().
This splits BN_sqr() into two parts, one of which is a separate bn_sqr()
function. This makes the code more readable and managable, while also
providing a better entry point for assembly optimisation.
ok tb@
jsing [Sat, 21 Jan 2023 13:24:39 +0000 (13:24 +0000)]
Provide basic benchmarks for BN_mul() and BN_sqr().
mvs [Sat, 21 Jan 2023 11:23:23 +0000 (11:23 +0000)]
Introduce per-sockbuf `sb_state' to use it with SS_CANTSENDMORE.
This time, socket's buffer lock requires solock() to be held. As a part of
socket buffers standalone locking work, move socket state bits which
represent its buffers state to per buffer state.
Opposing the previous reverted diff, the SS_CANTSENDMORE definition left
as is, but it used only with `sb_state'. `sb_state' ored with original
`so_state' when socket's data exported to the userland, so the ABI kept as
it was.
Inputs from deraadt@.
ok bluhm@
kettenis [Sat, 21 Jan 2023 10:34:49 +0000 (10:34 +0000)]
The Qualcomm UEFI Secure Application that handles EFI variables seems to
insist on shared memory in the low 4G of the physical address space.
Use bus_dmamem_alloc_range(9) to make sure this is the case.
Makes setting and getting the RTC time reliable.
ok patrick@, deraadt@, phessler@
kettenis [Sat, 21 Jan 2023 10:30:11 +0000 (10:30 +0000)]
Make bus_dmamem_alloc_range(9) API available.
ok patrick@, deraadt@, phessler@
jsing [Sat, 21 Jan 2023 09:21:11 +0000 (09:21 +0000)]
Reorder functions and drop unnessary static prototypes.
No functional change.
kn [Sat, 21 Jan 2023 09:01:55 +0000 (09:01 +0000)]
sync EXTRACT_CASES and PATCH_CASES with reality
xz handles .lz now, we deal with zstd patches by default, bzip2 and gzip
have command variables.
job [Fri, 20 Jan 2023 22:00:47 +0000 (22:00 +0000)]
Refactor x509v3_cache_extensions
Simplify x509v3_cache_extensions() by using a wrapper to avoid
duplication of code for locking and checking the EXFLAG_INVALID flag.
OK tb@
nicm [Fri, 20 Jan 2023 21:36:00 +0000 (21:36 +0000)]
Add a flag to display-menu to select the manu item chosen first, GitHub
issue 3442.
krw [Fri, 20 Jan 2023 17:47:07 +0000 (17:47 +0000)]
editor_name() and get_mp() don't change the disklabel so
const'ify some parameters and local variables.
jsing [Fri, 20 Jan 2023 17:34:52 +0000 (17:34 +0000)]
Remove various #ifndef BN_SQR_COMBA.
bn_sqr_comba{4,8}() is now always available.
ok tb@
jsing [Fri, 20 Jan 2023 17:31:52 +0000 (17:31 +0000)]
Move bn_{mul,sqr}_comba{4,8}() from bn_asm.c to bn_mul.c/bn_sqr.c.
Wrap these in HAVE_BN_{MUL,SQR}_COMBA{4,8} defines. Add these defines to
bn_arch.h where the architecture currently provides its own version.
ok tb@
jsing [Fri, 20 Jan 2023 17:26:03 +0000 (17:26 +0000)]
Move {mul,sqr}_add_c{,2} macros from bn_asm.c to bn_local.h.
These depend on other macros that are in already in bn_local.h and this
makes them available to other source files. A lot more clean up will be
needed in the future.
Of course x86_64-gcc.c makes use of the same macro names - sprinkle some
undef in there for the time being.
ok tb@
millert [Fri, 20 Jan 2023 17:21:12 +0000 (17:21 +0000)]
The locale is ignored on OpenBSD, match the wording used in isalpha.3.
OK schwarze@
cheloha [Fri, 20 Jan 2023 17:18:08 +0000 (17:18 +0000)]
hppa: simplify itmr_rearm()
The nest of branches in itmr_rearm() can be simplified to:
if (cycles <= t1 - t0) {
/* we probably missed */
}
We're doing modular unsigned 32-bit, so the rollover case in the
current code is superfluous.
Tested by miod@.
"Works for me." miod@
kn [Fri, 20 Jan 2023 17:15:22 +0000 (17:15 +0000)]
Sync usage with SYNOPSIS, mention that -x may use stdout
Feedback from jmc and Crystal Kolipe
OK jmc
deraadt [Fri, 20 Jan 2023 16:03:14 +0000 (16:03 +0000)]
amd64 now has xonly support via the PKU feature. Marking ld.so exec-only
is no longer a NOP on those systems, let's do it.
deraadt [Fri, 20 Jan 2023 16:01:04 +0000 (16:01 +0000)]
On cpu with the PKU feature, prot=PROT_EXEC pages now create pte which
contain PG_XO, which is PKU key1. On every exit from kernel to userland,
force the PKU register to inhibit data read against key1 memory. On
(some) traps into the kernel if the PKU register is changed, abort the
process (processes have no reason to change the PKU register). This
provides us with viable xonly functionality on most modern intel & AMD
cpus. I started with a xsave-based diff from dv@, but discovered the
fpu save/restore logic wasn't a good fit and went to direct register management.
Disabled on HV (vm) systems until we know they handle PKU correctly.
ok kettenis, dv, guenther, etc
claudio [Fri, 20 Jan 2023 15:43:25 +0000 (15:43 +0000)]
Update config regress test now that aspa-set syntax changed.
OK tb@
claudio [Fri, 20 Jan 2023 15:42:34 +0000 (15:42 +0000)]
Remove 'allow' from aspa-set provider-set lists. That keyword is no longer
needed.
OK tb@
claudio [Fri, 20 Jan 2023 15:41:33 +0000 (15:41 +0000)]
Document the aspa-set table. While there remove the superfluous 'allow'
keyword.
OK tb@
kettenis [Fri, 20 Jan 2023 15:11:44 +0000 (15:11 +0000)]
When ppb(4) detaches, make its bus range available again even if it was
configured by the firmware and pre-allocated by our parent pci(4) bus.
Fixes unplugging and replugging thunderbolt devices that were plugged in
when the machine was booted.
ok mlarkin@, deraadt@, miod@
krw [Fri, 20 Jan 2023 14:35:43 +0000 (14:35 +0000)]
char * -> const char * parameters where clang doesn't complain.
jsing [Fri, 20 Jan 2023 12:16:46 +0000 (12:16 +0000)]
Remove unused code.
ok tb@