cem [Wed, 23 Oct 2019 19:03:03 +0000 (19:03 +0000)]
amd64: Add CFI directives for libc syscall stubs
No functional change (in program code). Additional DWARF metadata is
generated in the .eh_frame section. Also, it is now a compile-time
requirement that machine/asm.h ENTRY() and END() macros are paired. (This
is subject to ongoing discussion and may change.)
This DWARF metadata allows llvm-libunwind to unwind program stacks when the
program is executing the function. The goal is to collect accurate
userspace stacktraces when programs have entered syscalls.
(The motivation for "Call Frame Information," or CFI for short -- not to be
confused with Control Flow Integrity -- is to sufficiently annotate assembly
functions such that stack unwinders can unwind out of the local frame
without the requirement of a dedicated framepointer register; i.e.,
-fomit-frame-pointer. This is necessary for C++ exception handling or
collecting backtraces.)
For the curious, a more thorough description of the metadata and some
examples may be found at [1] and documentation at [2]. You can also look at
'cc -S -o - foo.c | less' and search for '.cfi_' to see the CFI directives
generated by your C compiler.
rstone [Wed, 23 Oct 2019 17:20:20 +0000 (17:20 +0000)]
Add missing M_NOWAIT flag
The LinuxKPI linux_dma code calls PCTRIE_INSERT with a
mutex held, but does not set M_NOWAIT when allocating
nodes, leading to a potential panic. All of this code
can handle an allocation failure here, so prefer an
allocation failure to sleeping on memory.
Also fix a related case where NOWAIT/WAITOK was not
specified. In this case it's not clear whether sleeping
is allowed so be conservative and assume not. There are
a lot of other paths in this code that can fail due to
a lack of memory anyway.
dim [Wed, 23 Oct 2019 17:02:45 +0000 (17:02 +0000)]
Build toolchain components as dynamically linked executables by default
Summary:
Historically, we have built toolchain components such as cc, ld, etc as
statically linked executables. One of the reasons being that you could
sometimes save yourself from botched upgrades, by e.g. recompiling a
"known good" libc and reinstalling it.
In this day and age, we have boot environments, virtual machine
snapshots, cloud backups, and other much more reliable methods to
restore systems to working order. So I think the time is ripe to flip
this default, and link the toolchain components dynamically, just like
almost all other executables on FreeBSD.
Maybe at some point they can even become PIE executables by default! :)
dim [Wed, 23 Oct 2019 16:57:11 +0000 (16:57 +0000)]
Bump clang's default target CPU for the i386 architecture (aka "x86") to
i686, as per the discussion on the freebsd-arch mailing list. Earlier
in r352030, I had already bumped it to i586, to work around missing
atomic 64 bit functions for the i386 architecture.
jhb [Wed, 23 Oct 2019 16:41:31 +0000 (16:41 +0000)]
Fix atomic_*cmpset32 on riscv64 with clang.
The lr.w instruction used to read the value from memory sign-extends
the value read from memory. GCC sign-extends the 32-bit comparison
value passed in whereas clang currently does not. As a result, if the
value being compared has the MSB set, the comparison fails for
matching 32-bit values when compiled with clang.
Use a cast to explicitly sign-extend the unsigned comparison value.
This works with both GCC and clang.
There is commentary in the RISC-V spec that suggests that GCC's
approach is more correct, but it is not clear if the commentary in the
RISC-V spec is binding.
andrew [Wed, 23 Oct 2019 13:21:15 +0000 (13:21 +0000)]
Stop enabling interrupts when reentering kdb on arm64
When we raise a data abort from the kernel we need to enable interrupts,
however we shouldn't be doing this when in the kernel debugger. In this
case interrupts can lead to a further panic as they don't expect to be
run from such a context.
manu [Wed, 23 Oct 2019 09:56:53 +0000 (09:56 +0000)]
regulator: Add a regnode_set_constraint function
This method check that boot_on or always_on is set to 1 and if it
is it will try to enable the regulator.
The binding docs aren't clear on what to do but Linux enable the regulator
if any of those properties is set so we want to do the same.
The function first check the status to see if the regulator is
already enabled it then get the voltage to check if it is in a acceptable
range and then enables it.
This will be either called from the regnode_init method (if it's needed by the platform)
or by a SYSINIT at SI_SUB_LAST
jhibbits [Wed, 23 Oct 2019 00:31:19 +0000 (00:31 +0000)]
powerpc/booke: Fix Book-E boot post-minidump
r353489 added minidump support for powerpc64, but it added a dependency on
the dump_avail array. Leaving it uninitialized caused breakage in late
boot. Initialize dump_avail, even though the 64-bit booke pmap doesn't yet
support minidumps, but will in the future.
mjg [Tue, 22 Oct 2019 22:55:46 +0000 (22:55 +0000)]
amd64 pmap: conditionalize per-superpage locks on NUMA
Instead of superpages use. The current code employs superpage-wide locking
regardless and the better locking granularity is welcome with NUMA enabled
even when superpage support is not used.
Requested by: alc
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21982
mjg [Tue, 22 Oct 2019 22:52:53 +0000 (22:52 +0000)]
pseudofs: hashed vncache
Vast majority of uses the cache are just checking if there is an entry
present on process exit (and evicting it if so). Both checking and
eviction process are very expensive and put the lock protecting it high
up on the profile during poudriere -j 104.
Convert the linked list into a hash. This allows to almost always avoid
taking the lock in the first place (and consequently almost removes it
from the profile). Note only one lock is preserved as a split did not
meaningfully impact contention.
Should the cache be used for something it will still run into contention
issues. The code needs a rewrite, but should someone want to tidy it up
further the following can be done:
1) per-chain locks (or at least an array)
2) hashing by something else than just pid
kevans [Tue, 22 Oct 2019 21:46:03 +0000 (21:46 +0000)]
arm: correct kernelstack allocation size
This appears to be a copy-pasto from previous lines that propagated to v6
over the years. Indeed, nothing references kernelstack beyond
USPACE_SVC_STACK_TOP and it would be odd if anything did.
jhb [Tue, 22 Oct 2019 20:41:54 +0000 (20:41 +0000)]
Split Chelsio send tags into a generic base tag and a ratelimit tag.
NIC KTLS will add a new TLS send tag type in cxgbe(4) that is a
distinct tag from a ratelimit tag. To support this, refactor
cxgbe_snd_tag to be a simple send tag with a type and convert the
existing ratelimit tag to a new cxgbe_rate_tag structure.
jhb [Tue, 22 Oct 2019 20:01:47 +0000 (20:01 +0000)]
Always allocate the atid table during attach.
Previously the table was allocated on first use by TOE and the
ratelimit code. The forthcoming NIC KTLS code also uses this table.
Allocate it unconditionally during attach to simplify consumers.
luporl [Tue, 22 Oct 2019 18:28:58 +0000 (18:28 +0000)]
[PPC] Avoid underflows in NUMA domains
On POWER8 systems with only one memory domain, the "ibm,associativity"
number that corresponds to it is 0, unlike POWER9 systems with two
or more domains, in which the minimum value is 1.
In POWER8 case, subtracting 1 causes an underflow on the unsigned domain
variable and a subsequent index out-of-bounds access.
glebius [Tue, 22 Oct 2019 18:05:15 +0000 (18:05 +0000)]
Allow epoch tracker to use the very last byte of the stack. Not sure
this will help to avoid panic in this function, since it will also use
some stack, but makes code more strict.
kib [Tue, 22 Oct 2019 16:17:38 +0000 (16:17 +0000)]
Fix interface between nfsclient and vnode pager.
Make the nfsclient always call vnode_pager_setsize() with the vnode
exclusively locked. This ensures that page fault always can find the
backing page if the object size check succeeded. Set VV_VMSIZEVNLOCK
flag on NFS nodes.
The main offender breaking the interface in nfsclient is
nfs_loadattrcache(), which is used whenever server responded with
updated attributes, which can happen on non-changing operations as
well. Also, iod threads only have buffers locked (and even that is
LK_KERNPROC), but they still may call nfs_loadattrcache() on RPC
response.
Instead of immediately calling vnode_pager_setsize() if server
response indicated changed file size, but the vnode is not exclusively
locked, set a new node flag NVNSETSZSKIP. When the vnode exclusively
locked, or when we can temporary upgrade the lock to exclusive, call
vnode_pager_setsize(), by providing the nfsclient VOP_LOCK() implementation.
Tested by: pho
Discussed with: rmacklem
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D21883
kib [Tue, 22 Oct 2019 16:09:25 +0000 (16:09 +0000)]
Add VV_VMSIZEVNLOCK flag.
The flag specifies that vm_fault() handler should check the vnode'
vm_object size under the vnode lock. It is converted into the object'
OBJ_SIZEVNLOCK flag in vnode_pager_alloc().
Tested by: pho
Reviewed by: markj
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D21883
glebius [Tue, 22 Oct 2019 16:06:33 +0000 (16:06 +0000)]
Execute nd6_dad_timer() in the network epoch, since nd6_dad_duplicated()
requires it.
Make nd6_dad_starttimer() require network epoch. Two calls out of three
happen from nd6_dad_timer(). Enter epoch in the remaining one.
avg [Tue, 22 Oct 2019 14:20:35 +0000 (14:20 +0000)]
nctgpio: improve performance (latency) of operation
This change consists of two parts.
First, nctgpio now supports hardware access via an I/O port window if
it's configured by firmware. For instance, PC Engines firmware
v4.10.0.2 does that. This is faster than going through the Super I/O
configuration registers.
Second, nctgpio now caches values of bits that it controls. For
example, the driver does not need to access the hardware to determine if
a pin is an output or an input, or a state of an output. Also, the
driver makes use of the fact that the hardware preserves an output state
of a pin accross a switch to the input mode and back.
With this change I am able to use the 1-Wire bus over nctgpio whereas
previously the driver introduced too much latency to be compliant with
the relatively strict protocol timings.
superio0: <Nuvoton NCT5104D/NCT6102D/NCT6106D (rev. B+)> at port 0x2e-0x2f on isa0
gpio1: <Nuvoton GPIO controller> at GPIO ldn 0x07 on superio0
pcib0: allocated type 4 (0x220-0x226) for rid 0 of gpio1
gpiobus1: <GPIO bus> on gpio1
owc0: <GPIO attached one-wire bus> at pin 4 on gpiobus1
ow0: <1 Wire Bus> on owc0
ow0: romid 28:b2:9e:45:92:10:02:34: no driver
ow_temp0: <Advanced One Wire Temperature> romid 28:b2:9e:45:92:10:02:34 on ow0
markj [Tue, 22 Oct 2019 14:20:06 +0000 (14:20 +0000)]
Avoid reloading bucket pointers in uma_vm_zone_stats().
The correctness of per-CPU cache accounting in that function is
dependent on reading per-CPU pointers exactly once. Ensure that
the compiler does not emit multiple loads of those pointers.
Reported and tested by: pho
Reviewed by: kib
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D22081
kevans [Tue, 22 Oct 2019 00:18:16 +0000 (00:18 +0000)]
tuntap(4): properly declare if_tun and if_tap modules
Simply adding MODULE_VERSION does not do the trick, because the modules
haven't been declared. This should actually fix modfind/kldstat, which
r351229 aimed and failed to do.
This should make vm-bhyve do the right thing again when using the ports
version, rather than the latest version not in ports.
These drivers have been merged into a single if_tuntap in 13.0. The
compatibility links existed only for the interim and will be MFC'd along
with the if_tuntap merge shortly.
kevans [Mon, 21 Oct 2019 20:17:31 +0000 (20:17 +0000)]
lualoader: fix setting of loader_color=NO in loader.conf(5)
Previously color.disabled would be calculated at color module load time,
then never touched again. We can detect serial boots beyond just what we're
told by loader.conf(5) so this works out in many cases, but we must
re-evaluate the situation after the config is loaded to make sure we're not
supposed to be forcing it enabled/disabled.
kevans [Mon, 21 Oct 2019 20:09:43 +0000 (20:09 +0000)]
lualoader: don't botch disabling of color
When colors are disabled, color.escape{fg,bg} would return the passed in
color rather than the proper ANSI sequence for the color.
color.escape{fg,bg} would be wrong.
Instead return '', as the associated reset* functions will also return ''.
This should get rid of the funky '2' and '4' in the kernel selector if
you're booting serial.
emaste [Mon, 21 Oct 2019 18:40:03 +0000 (18:40 +0000)]
Additional fix for -DNO_CLEAN build across r353340 and r353381
opensolaris_atomic.S is now only used on i386 with opensolaris_atomic.c
used on other platforms. After r353381 it doesn't exist on those
platforms so the stale dependency would result in a build error.
r353408 addressed this issue for cddl/lib/libzpool, but it persisted
with the opensolaris and zfs modules.