kevans [Thu, 30 Jan 2020 03:01:00 +0000 (03:01 +0000)]
pkgbase: generate dependencies in generate-ucl.sh
This cuts out a large chunk of duplicated *.ucl files that just needed
dependencies tacked on, and gives generate-ucl.sh some leeway in case a
future pkg may need more than one dependency.
Run-through to determine which ones could simply be removed done with for i
in *.ucl; do diff -U3 template.ucl ${i}; done | less and inspecting for any
differences beyond just adding deps at the end. The remaining ucl files are
basically all differently-licensed, require scripts, or are marked vital.
I've opted to remove %PKGDEPS% entirely without regard for third-party ucl,
as pkgbase is not yet considered production. However, I do not hold a strong
position on this and there is approximately 0 chance it will return.
clibs should have been added previously in generate-ucl.sh as one that
doesn't have any dependencies, but do so now that we would otherwise be
tacking on the runtime dependency.
mjg [Thu, 30 Jan 2020 02:12:52 +0000 (02:12 +0000)]
vfs: unlazy before dooming the vnode
With this change having the listmtx lock held postpones dooming the vnode.
Use this fact to simplify iteration over the lazy list. It also allows
filters to safely access ->v_data.
kevans [Wed, 29 Jan 2020 22:40:13 +0000 (22:40 +0000)]
config.mk: #define MAC as well
MAC is also almost universally a default; every GENERIC includes it, and
it's std.armv[67]. mips is again the oddball here with it only being
included in ERL/OCTEON1.
The only module currently working around this one is mac_veriexec, but it
looks like nothing it builds actually uses the MAC definition. Downstream
consumers enabling MAC in mips using mac_veriexec may be advised to do
something differently here in config.mk.
dim [Wed, 29 Jan 2020 21:07:46 +0000 (21:07 +0000)]
Merge r357267 from the clang1000-import branch:
Fix the following -Werror warning from clang 10.0.0 in tip:
usr.bin/tip/tip/tip.c:428:4: error: misleading indentation; statement is not part of the previous 'if' [-Werror,-Wmisleading-indentation]
if (gch == EOF)
^
usr.bin/tip/tip/tip.c:426:5: note: previous statement is here
} else if (!cumode && gch == character(value(FORCE)))
^
The intent was to have the EOF check grouped with the getchar() call
just above it. This was accidentally introduced in r354624.
dim [Wed, 29 Jan 2020 21:06:22 +0000 (21:06 +0000)]
Merge r357260 from the clang1000-import branch:
Fix the following -Werror warning from clang 10.0.0 in procstat:
usr.bin/procstat/procstat_sigs.c:79:3: error: misleading indentation; statement is not part of the previous 'else' [-Werror,-Wmisleading-indentation]
xo_close_container(name);
^
usr.bin/procstat/procstat_sigs.c:77:4: note: previous statement is here
} else
^
The intent was to group the xo_close_container() call to the previous
snprintf() call.
kevans [Wed, 29 Jan 2020 18:50:55 +0000 (18:50 +0000)]
kmod build: start generating opt_global.h, include it
For untied module builds, we'll generate opt_foo headers if they're included
in SRCS. However, options that would normally be represented in opt_global.h
aren't properly represented.
Start generating opt_global.h with #define VIMAGE for !mips since it's
almost universally a project default and right now kmods must hack it in
themselves in order to be properly compiled for the default kernel. For
example, ^/sys/modules/pf/Makefile
kevans [Wed, 29 Jan 2020 18:47:08 +0000 (18:47 +0000)]
pkgbase: fix caroot packaging and add post-install script
The original intention for caroot was to be packaged separately, perhaps so
that users can have a more/less conservative upgrade policy for this
separated from the rest of base.
secure/caroot/Makefile doesn't have anything interesting to package, but its
subdirectories might. Move the PACKAGE= to Makefile.inc so both blacklisted
and trusted get packaged consistently into the correct one rather than the
default -utilities. Also tag the directories for package=caroot, as they
could also be empty; blacklisted is empty by default, but trusted is not.
Add a post-install script to do certctl rehash, along with a note should we
eventually come up with a way to detect that files have been added or
removed that requires a rehash.
-caroot gets a dependency on -utilities, as that's where we provide certctl
at the moment. We can perhaps reconsider this and put certctl into this
package in the future, but there are some bits within -utilities that
unconditionally invoke certctl so let's hold off for now.
Reviewed by: manu (earlier version, before -utilities dep added)
Differential Revision: https://reviews.freebsd.org/D23352
luporl [Wed, 29 Jan 2020 18:13:44 +0000 (18:13 +0000)]
[PPC64] Fix NUMA on POWER8
On some POWER8 machines, 'ibm,associativity' property may have 6
cells, which would overflow the 5 cells buffer being used.
There was also an issue with the "check if node is root" part,
that have been fixed too.
kevans [Wed, 29 Jan 2020 16:24:16 +0000 (16:24 +0000)]
pkgbase: kill off clibs.ucl and docs.ucl
These are just direct copies of the template. If they don't exist but are
attempted to be used, generate-ucl.sh will simply fallback to the template
anyways. Let it do so to reduce potential maintenance burden.
mhorne [Wed, 29 Jan 2020 15:50:48 +0000 (15:50 +0000)]
Fix definition of SSTATUS_SD
The SD bit is defined as the MSB of the sstatus register, meaning its
position will vary depending on the CSR's length. Previously, there were
two (unused) defines for this, for the 32 and 64-bit cases, but their
definitions were swapped.
Consolidate these into one define: SSTATUS_SD, and make the definition
dependent on the value of __riscv_xlen.
Reviewed by: br
MFC after: 3 days
Differential Revision: https://reviews.freebsd.org/D23402
ganbold [Wed, 29 Jan 2020 12:10:42 +0000 (12:10 +0000)]
Add an overlay for RK3328 USB3 node.
It doesn't exist in mainline dts due to the issues related
with detaching and reattaching USB3 devices as mentioned in
https://patchwork.kernel.org/patch/10853381/
In case of FreeBSD, as a temporary workaround "usbconfig reset"
command can fix the problem.
cem [Wed, 29 Jan 2020 05:59:22 +0000 (05:59 +0000)]
ypldap(8): Fix absent va_end() in 'goto fail' path
In this path, we used va_start() without pairing it with va_end(). Add the
va_end(). (va_start() without paired va_end() is undefined behavior per the C
standard.)
"In many implementations, [va_end] is a do-nothing operation; but those
implementations that need it probably need it badly." - Rationale for the ANSI
C Programming Language, § 4.8.1.3.
cem [Wed, 29 Jan 2020 05:31:40 +0000 (05:31 +0000)]
ixgbe(4): Eliminate bogus sizeof() expressions
All of these uses of sizeof() were on the wrong type in relation to the pointer
passed to SYSCTL_ADD_PROC as arg1. Fortunately, none of the handlers actually
use arg2. So just don't pass a (non-zero) arg2.
cem [Wed, 29 Jan 2020 05:25:20 +0000 (05:25 +0000)]
qmath(3) test: Replace overcomplicated abomination with arc4random(3)
The horrific GENRAND construction bent over backwards to construct 64-bit
signed integers from the 31-bit output of random(3) for about 20 numbers per
test. Reproducibility wasn't a goal: random(3) was seeded with
srandomdev(3). Speed is not a factor for generating 20 integers with
arc4random(3). Range is not a factor: all uses did not bound the range
beyond that of the full [INT64_MIN, INT64_MAX]. Just use arc4random(3).
Reported by: Coverity
CIDs: 1404809, 1404817, 1404838, 1404840 and about 6x other
identical reports of dubious code relating to the
construction
jhibbits [Wed, 29 Jan 2020 04:32:06 +0000 (04:32 +0000)]
powerpc64/csu: Rename dummy asm file to force use of 'cc'
Implicit make rules build .S asm files with the compiler, rather than the
assembler. r356889 removed GNU as from the build for powerpc targets,
causing '.s' asm files to fail to build, due to a missing 'as'. Rename the
one dummy asm file we have to a '.S' to force the implicit rules to build
with the compiler rather than the assembler.
asomers [Tue, 28 Jan 2020 23:07:31 +0000 (23:07 +0000)]
Speed up "zpool import" in the presence of many zvols
By default, zpools may not be backed by zvols (that can be changed with the
"vfs.zfs.vol.recursive" sysctl). When that sysctl is set to 0, the kernel
does not attempt to read zvols when looking for vdevs. But the zpool command
still does. This change brings the zpool command into line with the kernel's
behavior. It speeds "zpool import" when an already imported pool has many
zvols, or a zvol with many snapshots.
kp [Tue, 28 Jan 2020 22:46:51 +0000 (22:46 +0000)]
tests: Test for an epair panic
if_epair abused the ifr_data field to insert its second interface in
IFC_IFLIST. If userspace provides a value for ifr_data it would get
dereferenced by the kernel leading to a panic.
Reported by: Ilja Van Sprundel <ivansprundel@ioactive.com>
MFC after: 3 days
kp [Tue, 28 Jan 2020 22:44:24 +0000 (22:44 +0000)]
epair: Do not abuse params to register the second interface
if_epair used the 'params' argument to pass a pointer to the b interface
through if_clone_create().
This pointer can be controlled by userspace, which means it could be abused to
trigger a panic. While this requires PRIV_NET_IFCREATE
privileges those are assigned to vnet jails, which means that vnet jails
could panic the system.
Reported by: Ilja Van Sprundel <ivansprundel@ioactive.com>
MFC after: 3 days
dim [Tue, 28 Jan 2020 20:12:29 +0000 (20:12 +0000)]
Merge r357225 from the clang1000-import branch:
Fix the following -Werror warning from clang 10.0.0 in newfs_msdos:
sbin/newfs_msdos/newfs_msdos.c:181:2: error: misleading indentation; statement is not part of the previous 'if' [-Werror,-Wmisleading-indentation]
if (o.align) {
^
sbin/newfs_msdos/newfs_msdos.c:179:5: note: previous statement is here
if (argc < 1 || argc > 2)
^
dim [Tue, 28 Jan 2020 20:11:27 +0000 (20:11 +0000)]
Merge r357224 from the clang1000-import branch:
Fix the following -Werror warning from clang 10.0.0 in binutils:
contrib/binutils/bfd/peicode.h:1356:3: error: misleading indentation; statement is not part of the previous 'if' [-Werror,-Wmisleading-indentation]
if (efi)
^
contrib/binutils/bfd/peicode.h:1353:8: note: previous statement is here
if (pe_arch (bfd_target_efi_arch (*target_ptr)) != arch)
^
contrib/binutils/bfd/peicode.h:1370:3: error: misleading indentation; statement is not part of the previous 'if' [-Werror,-Wmisleading-indentation]
if (!efi)
^
contrib/binutils/bfd/peicode.h:1367:8: note: previous statement is here
if (pe_arch (bfd_target_pei_arch (*target_ptr)) != arch)
^
mjg [Tue, 28 Jan 2020 17:48:17 +0000 (17:48 +0000)]
amd64: revamp memcmp
Borrow the trick from memset and memmove and use the scale/index/base addressing
to avoid branches.
If a mismatch is found, the routine has to calculate the difference. Make sure
there is always up to 8 bytes to inspect. This replaces the previous loop which
would operate over up to 16 bytes with an unrolled list of 8 tests.
Speed varies a lot, but this is a net win over the previous routine with probably
a lot more to gain.
trasz [Tue, 28 Jan 2020 13:51:53 +0000 (13:51 +0000)]
Add compat.linux.ignore_ip_recverr sysctl. This is a workaround
for missing IP_RECVERR setsockopt(2) support. Without it, DNS
resolution is broken for glibc >= 2.30 (glibc BZ #24047).
From the user point of view this fixes "yum update" on recent
CentOS 8.
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D23234
kib [Tue, 28 Jan 2020 11:29:06 +0000 (11:29 +0000)]
Save lower root vnode in nullfs mnt data instead of upper.
Nullfs needs to know the root vnode of the lower fs during the
operation. Currently it caches the upper vnode of it, which is also
the root of the nullfs mount. On unmount, nullfs calls vflush() with
rootrefs == 1, and aborts non-forced unmount if there are any more
vnodes instantiated during vflush(). This means that the reference to
the root vnode after failed non-forced unmount could be lost and
nullm_rootvp points to the freed memory.
Fix it by storing the reference for lower vnode instead, which is kept
intact during vflush(). nullfs_root() now instantiates the upper
vnode of lower root. Care about VV_ROOT flag in null_nodeget().
Reported and tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
kevans [Tue, 28 Jan 2020 03:47:29 +0000 (03:47 +0000)]
netbsd-tests: libc: use correct modes in O_SEARCH tests
The current code clearly intended for these to be octal based on the values
used, but the octal prefix was forgotten. Add it now for correctness, but
note that we don't currently execute these tests.
This has been submitted upstream as misc/54902, so I've omitted the standard
FreeBSD markers that we tend to put into netbsd-tests for upstream-candidate
identification.
kevans [Tue, 28 Jan 2020 03:02:18 +0000 (03:02 +0000)]
caroot: blacklisted: automatically pick up *.pem in the tree
This kind of automagica got picked up in trusted/ prior to the initial
commit, but never got applied over in blacklisted. Ideally no one will be
using blacklisted/ to store arbitrary certs that they don't intend to
blacklist, so we should just install anything that's in here rather than
force consumer to first copy cert into place and then modify the file
listing in the Makefile.
Wise man once say: "it is better to restrict too much, than not enough.
sometimes."
cem [Tue, 28 Jan 2020 01:39:50 +0000 (01:39 +0000)]
amdtemp(4): Add support for Family 17h CCD sensors
Probe Family 17h CPUs for up to 4 (Zen, Zen+) or 8 (Zen2) CCD temperature
sensors. These were discovered by Ondrej Čerman
(https://github.com/ocerman) and collaborators experimentally, and are not
currently documented in any datasheet I have access to.
imp [Tue, 28 Jan 2020 00:36:46 +0000 (00:36 +0000)]
Remove old device list
The device list hasn't aged well. All these devices are over a decade old. umass
supports thunb drives almost universally, and the list is too long to try to
list here.
Remove some obsolete advice as well. This isn't the place to talk about how to
create FAT filesystems, nor now to mount them. The only advice that's still
useful is the rescanning of a multi-slot flash adapater.
imp [Mon, 27 Jan 2020 22:40:03 +0000 (22:40 +0000)]
multi-boot for openstack/qcow images
Make stock FreeBSD more useful for people wishing to use them. The
QEMU folks suggested this change. It adds a serial console which
allows them to interact with FreeBSD from the earliest moments. This
allows them to configure FreeBSD via the serial port to set it up for
CI use.
imp [Mon, 27 Jan 2020 22:36:54 +0000 (22:36 +0000)]
Make mqueue objects work across a fork again.
In r110908 (2003) alfred added DFLAG_PASSABLE to tag those types of FD
that can be passed via unix pipes, but mqueuefs didn't exist
yet. Later, in r152825 (2005) davidxu neglected to include
DFLAG_PASSABLE since people don't normally pass these things via unix
sockets (it's a FreeBSD implementation detail that it's a file
descriptor, nobody noticed). Then r223866 (2011) by jonathan used the
new flag in fdcopy, which fork uses. Due to that, mqueuefs actually
broke mqueue objects being propagated by fork. No mention of mqueuefs
was made in r223866, so I think it was an unintended consequence.
Fix this by tagging mqueuefs as passable as well. They were prior to
alfred's change (and it's clear there's no intent in his change to
change this behavior), and POSIX requires this to be the case as well.
imp [Mon, 27 Jan 2020 22:19:55 +0000 (22:19 +0000)]
Create a convenince wrapper to fill in a CAM_PATH_INQ request for MMC sims. Pass
in the parameters needed for the different sims, but it's almost all identical.
dougm [Mon, 27 Jan 2020 15:09:13 +0000 (15:09 +0000)]
Correct the use of RB_AUGMENT in the RB_TREE macros so that is invoked
at the root of every subtree that changes in an insert or delete, and
only once, and ordered from the bottom of the tree to the top. For
intel_gas.c, the only user of RB_AUGMENT I can find, change the
augmenting routine so that it does not climb from entry to tree root
on every call, and remove a 'tree correcting' function that can be
supplanted by proper tree augmentation.
kib [Mon, 27 Jan 2020 13:15:16 +0000 (13:15 +0000)]
Fix aggregating geoms for BIO_SPEEDUP.
If the bio was split into several bios going down, completion computes
bio_completed of the original bio as sum of the bio_completes of the
splits. For BIO_SETUP, bio_length means something different than the
length. it is the requested speedup amount, and is duplicated into the
splits, which is in fact reasonable, since we cannot know how the
previous activity was distributed among subordinate geoms. Obviously,
the sum of n bio_length is greater than bio_length for n > 1, which
triggers assert that bio_length >= bio_completed for e.g. geom_stripe
and geom_raid3.
Fix this by reassigning bio_completed from bio_length for completed
BIO_SPEEDED, I do not think it really mattters what we return in
bio_completed.
arichardson [Mon, 27 Jan 2020 12:02:47 +0000 (12:02 +0000)]
Build hard-float lib32 for mips64hf/mips64elhf
This should fix linker errors when building with clang+lld.
After this change the lib32 compat libraries are now buildt with
-mhard-float instead of -msoft-float
arichardson [Mon, 27 Jan 2020 12:02:41 +0000 (12:02 +0000)]
Allow bootstrapping makefs on older FreeBSD hosts and Linux/macOS
In order to do so we need to install the msdosfs headers to the bootstrap
sysroot and avoid includes of kernel headers that may not exist on every
host (e.g. sys/lockmgr.h). This change should allow bootstrapping of makefs
on FreeBSD 11+ as well as Linux and macOS.
We also have to avoid using the IO_SYNC macro since that may not be
available. In makefs it is only used to switch between calling
bwrite() and bdwrite() which both call the same function. Therefore we
can simply always call bwrite().
For our CheriBSD builds we always bootstrap makefs by setting
LOCAL_XTOOL_DIRS='lib/libnetbsd usr.sbin/makefs' and use the makefs binary
from the build tree to create a bootable disk image.
rmacklem [Sun, 26 Jan 2020 17:59:05 +0000 (17:59 +0000)]
Fix a crash in the NFSv4 server.
The PR reported a crash that occurred when a file was removed while
client(s) were actively doing lock operations on it.
Since nfsvno_getvp() will return NULL when the file does not exist,
the bug was obvious and easy to fix via this patch. It is a little
surprising that this wasn't found sooner, but I guess the above
case rarely occurs.
jhb [Sun, 26 Jan 2020 14:20:57 +0000 (14:20 +0000)]
Fix some misleading indentation warnings reported by recent clang.
These should not be any functional change. While the change in
emul10kx-pcm.c looks like a real bug fix (as opposed to inconsistent
whitespace), the extra statements were not harmful.
jhb [Sun, 26 Jan 2020 14:19:08 +0000 (14:19 +0000)]
Compile hack.c with normal CFLAGS + -shared -nostdlib.
Originally, hack.c was compiled into a shard object with just -shared
-nostdlib. This assumed that ${CC} did not require any additional
flags for ABIs, cross-building, etc.
When kern.post.mk was created in r89509 by reducing duplication in
kernel Makefile.<arch> files, the -shared flag was moved into a
HACK_EXTRA_FLAGS variable so that sparc64 could override it with
-Wl,-shared. The sparc64 hack was removed in r111650, but
HACK_EXTRA_FLAGS was left in place. Over time, we have started
support toolchains that require flags to support alternate ABIs on
MIPS and PowerPC and started (ab)using HACK_EXTRA_FLAGS to set only
those flags.
I need to fix risc-v to pass -mno-relax to the hack.c build for lld in
llvm 10, and the patches to support cross-build from non-FreeBSD hosts
need to include -target for clang in CFLAGS for hack.c. Rather than
adding more hacks into HACK_EXTRA_FLAGS, just use the full set of
CFLAGS with hack.c.
mjg [Sun, 26 Jan 2020 00:41:38 +0000 (00:41 +0000)]
vfs: stop null checking routines in vop wrappers
Calls to vop_bypass pass the same argument, but type casted to something else.
Thus by replacing NULL routines with vop_bypass we avoid a runtime check.
mjg [Sun, 26 Jan 2020 00:40:27 +0000 (00:40 +0000)]
vfs: fix freevnodes count update race against preemption
vdbatch_process leaves the critical section too early, openign a time
window where another thread can get scheduled and modify vd->freevnodes.
Once it the preempted thread gets back it overrides the value with 0.
Just move critical_exit to the end of the function.
mjg [Sun, 26 Jan 2020 00:38:06 +0000 (00:38 +0000)]
ufs: add vgone calls for unconstructed vnodes in the error path
This mostly eliminates the requirement that vput never unlocks the vnode
before calling VOP_INACTIVE. Note it may still be present for other
filesystems.
See r356126 for an example bug.
Note vput stopped doing early unlock in r357070 thus this change does
not affect correctness as it is.
phil [Sat, 25 Jan 2020 21:16:45 +0000 (21:16 +0000)]
Import libxo-1.4.0:
- Two changes to encoder options:
encoder options may use plus or colon, but only one
encoder names can be specified as "@name"
This results in the syntax:
- If xo_set_program is called before xo_parse_args, honor the requested value
- add xo_errorn* function; repair newline-adding-on-xo_error bug
- test programs now use fixed name, since linux libtool prefixs "lt-"
- Fix "horse butt" comment in source code
- update test cases
projects / FreeBSD / FreeBSD.git / log