Dimitry Andric [Fri, 12 Jan 2018 18:19:14 +0000 (18:19 +0000)]
Pull in r321994 from upstream llvm trunk (by Alexey Bataev):
[SLP] Fix PR35777: Incorrect handling of aggregate values.
Summary:
Fixes the bug with incorrect handling of InsertValue|InsertElement
instrucions in SLP vectorizer. Currently, we may use incorrect
ExtractElement instructions as the operands of the original
InsertValue|InsertElement instructions.
Dimitry Andric [Tue, 9 Jan 2018 17:38:43 +0000 (17:38 +0000)]
Pull in r322041 from upstream lld trunk (by Rui Ueyama):
Do not use parallelForEach to call maybeCompress().
Currently LLVM's paralellForEach has a problem with reentracy.
That caused https://bugs.llvm.org/show_bug.cgi?id=35788 (lld somtimes
hangs while linking Ruby 2.4) because maybeCompress calls writeTo
which uses paralellForEach.
This patch is to avoid using paralellForEach to call maybeCompress to
workaround the issue.
This should fix potential hangs when linking parts of ruby24.
Dimitry Andric [Tue, 9 Jan 2018 17:37:09 +0000 (17:37 +0000)]
Pull in r321986 from upstream lld trunk (by James Henderson):
[ELF] Compress debug sections after assignAddresses and support
custom layout
Previously, in r320472, I moved the calculation of section offsets
and sizes for compressed debug sections into maybeCompress, which
happens before assignAddresses, so that the compression had the
required information. However, I failed to take account of
relocations that patch such sections. This had two effects:
1. A race condition existed when a debug section referred to a
different debug section (see PR35788).
2. References to symbols in non-debug sections would be patched
incorrectly. This is because the addresses of such symbols are not
calculated until after assignAddresses (this was a partial
regression caused by r320472, but they could still have been
broken before, in the event that a custom layout was used in a
linker script).
assignAddresses does not need to know about the output section size
of non-allocatable sections, because they do not affect the value of
Dot. This means that there is no longer a reason not to support
custom layout of compressed debug sections, as far as I'm aware.
These two points allow for delaying when maybeCompress can be called,
removing the need for the loop I previously added to calculate the
section size, and therefore the race condition. Furthermore, by
delaying, we fix the issues of relocations getting incorrect symbol
values, because they have now all been finalized.
This should fix thread race conditions when linking parts of ruby24.
Dimitry Andric [Mon, 8 Jan 2018 18:42:40 +0000 (18:42 +0000)]
Add explanatory comment for r327622: clang 6.0.0 and higher warn about
the ACPI_ROOT_OBJECT and ACPI_TO_POINTER macros from acpica's actypes.h
header, that they use arithmetic on a null pointer treated as a cast
from integer to pointer, which is a GNU extension. We turn off the
warning, because this is in contributed code.
Dimitry Andric [Sun, 7 Jan 2018 18:33:19 +0000 (18:33 +0000)]
Pull in r321963 from upstream libc++ trunk (by me):
Add pre-C++11 is_constructible wrappers for 3 arguments
Summary:
After rL319736 for D28253 (which fixes PR28929), gcc cannot compile
<memory> anymore in pre-C+11 modes, complaining:
In file included from /usr/include/c++/v1/memory:648:0,
from test.cpp:1:
/usr/include/c++/v1/memory: In static member function 'static std::__1::shared_ptr<_Tp> std::__1::shared_ptr<_Tp>::make_shared(_A0&, _A1&, _A2&)':
/usr/include/c++/v1/memory:4365:5: error: wrong number of template arguments (4, should be at least 1)
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in make_shared" );
^
In file included from /usr/include/c++/v1/memory:649:0,
from test.cpp:1:
/usr/include/c++/v1/type_traits:3198:29: note: provided for 'template<class _Tp, class _A0, class _A1> struct std::__1::is_constructible'
struct _LIBCPP_TEMPLATE_VIS is_constructible
^~~~~~~~~~~~~~~~
In file included from /usr/include/c++/v1/memory:648:0,
from test.cpp:1:
/usr/include/c++/v1/memory:4365:5: error: template argument 1 is invalid
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in make_shared" );
^
/usr/include/c++/v1/memory: In static member function 'static std::__1::shared_ptr<_Tp> std::__1::shared_ptr<_Tp>::allocate_shared(const _Alloc&, _A0&, _A1&, _A2&)':
/usr/include/c++/v1/memory:4444:5: error: wrong number of template arguments (4, should be at least 1)
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in allocate_shared" );
^
In file included from /usr/include/c++/v1/memory:649:0,
from test.cpp:1:
/usr/include/c++/v1/type_traits:3198:29: note: provided for 'template<class _Tp, class _A0, class _A1> struct std::__1::is_constructible'
struct _LIBCPP_TEMPLATE_VIS is_constructible
^~~~~~~~~~~~~~~~
In file included from /usr/include/c++/v1/memory:648:0,
from test.cpp:1:
/usr/include/c++/v1/memory:4444:5: error: template argument 1 is invalid
static_assert((is_constructible<_Tp, _A0, _A1, _A2>::value), "Can't construct object in allocate_shared" );
^
This is also reported in https://bugs.freebsd.org/224946 (FreeBSD is
apparently one of the very few projects that regularly builds
programs against libc++ with gcc).
The reason is that the static assertions are invoking
is_constructible with three arguments, while gcc does not have the
built-in is_constructible feature, and the pre-C++11 is_constructible
wrappers in <type_traits> only provide up to two arguments.
I have added additional wrappers for three arguments, modified the
is_constructible entry point to take three arguments instead, and
added a simple test to is_constructible.pass.cpp.
Dimitry Andric [Sat, 6 Jan 2018 15:52:21 +0000 (15:52 +0000)]
Revert r327338, undoing the changes to the ACPI_ADD_PTR and ACPI_SUB_PTR
macros. Instead, turn off clang 6.0.0 warnings about null pointer
arithmetic in usr.sbin/acpi/acpidb instead.
Kyle Evans [Sat, 6 Jan 2018 14:21:32 +0000 (14:21 +0000)]
Move syscon_generic to attach much later
It still needs to be before if_awg at least in order to be available for
other operations, but it should not be attaching before interrupt
controllers at the very least.
This should make errors involving syscon register space colliding with other
devices a little more innocent, but these conflicts should really be tracked
down and resolved. One such conflict is with the Raspberry Pi 3 local
interrupt controller, noticed by tuexen@
We normally want to ignore SHT_NOBITS sections when computing
offsets. The sh_offset of section itself seems to be irrelevant and
* If the section is in the middle of a PT_LOAD, it will make no
difference on the computed offset of the followup section.
* If it is in the end of a PT_LOAD, we want to avoid its alignment
changing the offset of the followup sections.
The issue is if it is at the start of the PT_LOAD. In that case we do
have to align it so that the following sections have congruent
address and offset module the page size. We were not handling this
case.
This should fix freebsd kernel link.
In particular, this fixes ctfmerge and/or objcopy throwing "Layout
constraint violation" errors when processing an lld-linked kernel.
Warner Losh [Sat, 6 Jan 2018 12:46:04 +0000 (12:46 +0000)]
The source strings are from the password database which guarantees
that the data going into it is sane. Out of an abundance of caution,
limit the string copies to prevent an overflow.
Warner Losh [Sat, 6 Jan 2018 12:45:59 +0000 (12:45 +0000)]
Sanity check sysconf return value to ensure it's positive before we
use it. Use proper cast to convert long to size_t (instead of
blksize_t) to preclude sign extension issues.
John Baldwin [Fri, 5 Jan 2018 22:26:23 +0000 (22:26 +0000)]
Remove LINK_MAX.
After recent changes to change filesystems to use filesystem-specific
limits, LINK_MAX is no longer used in the base system. Applications
should in theory be able to cope with a lack of LINK_MAX by using
pathconf().
Add cpuctl(4) ioctl CPUCTL_EVAL_CPU_FEATURES which forces re-read of
cpu_features, cpu_features2, cpu_stdext_features, and
std_stdext_features2.
The intent is to allow the kernel to see the changes in the CPU
features after micocode update. Of course, the update is not atomic
across variables and not synchronized with readers. See the man page
warning as well.
Reviewed by: imp (previous version), jilles
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D13770
Warner Losh [Fri, 5 Jan 2018 07:09:40 +0000 (07:09 +0000)]
Add a number of sanity checks to the data that we're handling from the
CIS. Coverity has tagged it as tainted. While this data is more
trusted than your average data, we still need to do some basic
validation on it. Check ioctl return value to ensure we switch memory
targets between common and attribute as well as the lseek.
Warner Losh [Fri, 5 Jan 2018 07:09:29 +0000 (07:09 +0000)]
Need to convert '/' back to '\' when creating a path. Ideally, this
would be filesystem type dependent, but that's difficult to accomplish
and it's unclear how the UEFI firmware will cope. Be conservative and
make boot loaders cope instead.
Warner Losh [Fri, 5 Jan 2018 07:09:24 +0000 (07:09 +0000)]
Set dp to NULL when we free it, and tree a NULL dp as an error
condition. This should prevent a double free. In addition, prevent a
leak by freeing dp each loop and when we're done.
Kyle Evans [Fri, 5 Jan 2018 01:46:41 +0000 (01:46 +0000)]
hexdump(1): Speed up -s flag on devices
Using the -s flag on devices is extraordinarily slow due to using fseek(3) a
little too conservatively. Address this by using fseek on character/block
devices as well, falling back to getchar(3) only if we fail to seek or we're
operating on tape drives, where fseek may succeed while not actually being
supported.
PR: 86485
Submitted by: arundel (originally; modified since then)
Reviewed by: cem
Differential Revision: https://reviews.freebsd.org/D10939
Kyle Evans [Thu, 4 Jan 2018 22:37:15 +0000 (22:37 +0000)]
if_awg: Use syscon prop if it exists
The emac bindings that are landing in Linux 4.15 specify a syscon property
on the emac node that point to /soc/syscon. Use this property if it's
specified, but maintain backwards compatibility with the old method.
The older method is still used for boards that we get .dtb from u-boot, such
as pine64, that did not yet have stable emac bindings.
Tested on: Banana Pi-M3 (a83t)
Tested on: Pine64 (a64)
Reviewed by: manu
Differential Revision: https://reviews.freebsd.org/D13296
John Baldwin [Thu, 4 Jan 2018 22:07:58 +0000 (22:07 +0000)]
Always use atomic_fetchadd() when updating per-user accounting values.
This avoids re-reading a variable after it has been updated via an
atomic op. It is just a cosmetic cleanup as the read value was only
used to control a diagnostic printf that should rarely occur (if ever).
John Baldwin [Thu, 4 Jan 2018 21:59:34 +0000 (21:59 +0000)]
Report offset relative to the backing object for kinfo_vmentry structures.
For the pathname reported in kinfo_vmentry structures (kve_path), the
sysctl handlers walk the object chain to find the bottom-most VM object.
This permits a COW mapping of a file with dirty pages to report the
pathname of the originally mapped file. Do the same for the object
offset (kve_offset) computing a cumulative offset during the same object
walk so that the reported offset is relative to the reported pathname.
Note that ptrace(PT_VM_ENTRY) already returns a cumulative offset
rather than the raw offset of the VM map entry.
Note also that this does not affect procstat -v output (even structured
output) since that output does not include the kve_offset field.
Ed Schouten [Thu, 4 Jan 2018 21:57:37 +0000 (21:57 +0000)]
Allow timed waits with relative timeouts on locks and condvars.
Even though pthreads doesn't support this, there are various alternative
APIs that use this. For example, uv_cond_timedwait() accepts a relative
timeout. So does Rust's std::sync::Condvar::wait_timeout().
Though I personally think that relative timeouts are bad (due to
imprecision for repeated operations), it does seem that people want
this. Extend the existing futex functions to keep track of whether an
absolute timeout is used in a boolean flag.
Steven Hartland [Thu, 4 Jan 2018 20:05:47 +0000 (20:05 +0000)]
Disabled the use of flowid for lagg by default
Disabled the use of RSS hash from the network card aka flowid for
lagg(4) interfaces by default as it's currently incompatible with
the lacp and loadbalance protocols.
The incompatibility is due to the fact that the flowid isn't know
for the first packet of a new outbound stream which can result in
the hash calculation method changing and hence a stream being
incorrectly split across multiple interfaces during normal
operation.
This can be re-enabled by setting the following in loader.conf:
net.link.lagg.default_use_flowid="1"
The default 80MHz clock speed returned by bhnd_pmu_si_clock() was already
correct; this just prevents the "No backplane clock specified" warning
printf from being emitted when querying backplane clock speed.
Restructure swapout tests after vm map locking was removed.
Consolidate the regions covered by the process lock.
Combine similar conditions tests into one, e.g. all process flags can
be test with one logical operation.
Add check for in-exec state, since p_vmspace is dererenced.
Remove labels and goto by explicitly tracking state.
Update comments.
Jason Unovitch [Thu, 4 Jan 2018 17:58:48 +0000 (17:58 +0000)]
Remove myself from ports-secteam
I will be moving on to other life commitments this year and will not have
the time to support contributions as a ports committer, if able, until life
settles at the end of the year.
Alan Cox [Thu, 4 Jan 2018 03:16:32 +0000 (03:16 +0000)]
Once we have decided to swap out a process, don't delay the laundering of
its per-thread kernel stack pages by making them pass through the inactive
queue first. Instead, immediately place them in the laundry so that they
might be cleaned and made available for reclamation sooner.
John Baldwin [Wed, 3 Jan 2018 17:40:51 +0000 (17:40 +0000)]
Use 'extern uint8_t' instead of 'extern void' for external symbols.
The beri boot loaders depend on symbols defined in linker scripts or
assembly files. The boot loaders do not care about the type of these
symbols but just want to extract a pointer to them. Older versions of
GCC permitted external symbols to be declared of type 'void' and then
'&foo' generated a void pointer to the memory at the symbol's address.
However, void objects are not valid C and newer versions of GCC error if
these are used. Instead, declare these symbols as being bytes (or
an array of bytes in the cheri_sdcard_vaddr case).
John Baldwin [Wed, 3 Jan 2018 17:35:38 +0000 (17:35 +0000)]
Don't clobber system LDFLAGS for beri boot loaders.
Prior to r325114, bsd.init.mk was included after assignments to CFLAGS and
LDFLAGS in these Makefiles. After r325114, bare assignments (= rather than
+=) lost system-assigned default values that are needed when compiling with
an external toolchain. CFLAGS in both Makefiles already uses +=. This
commit changes LDFLAGS to use +=. While here, depend on the LDFLAGS update
in the parent Makefile.inc to set -nostdlib.
Use the new SDM-approved way to serialize x2APIC MSR writes.
SDM editions 64 and below stated that it is enough to use MFENCe or
LFENCE to serialize x2APIC register writes. New edition 65 requires
either full serialization instruction or MFENCE;LFENCE sequence. Use
the later, FreeBSD needs serialization to ensure that writes done
before IPI request are visible to the target IPI CPU.
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Mike Karels [Wed, 3 Jan 2018 00:56:30 +0000 (00:56 +0000)]
make SW_WATCHDOG dynamic
Enable the hardclock-based watchdog previously conditional on the
SW_WATCHDOG option whenever hardware watchdogs are not found, and
watchdogd attempts to enable the watchdog. The SW_WATCHDOG option
still causes the sofware watchdog to be enabled even if there is a
hardware watchdog. This does not change the other software-based
watchdog enabled by the --softtimeout option to watchdogd.
Note that the code to reprime the watchdog during kernel core dumps is
no longer conditional on SW_WATCHDOG. I think this was previously a bug.
Mark Johnston [Tue, 2 Jan 2018 18:11:54 +0000 (18:11 +0000)]
Fix some I/O ordering issues in gmirror.
- BIO_FLUSH requests were dispatched to the disks directly from
g_mirror_start() rather than going through the mirror's I/O request
queue, so they could have been reordered with preceding writes.
Address this by processing such requests from the queue, avoiding
direct dispatch.
- Handling for collisions with synchronization requests was too
fine-grained and could cause reordering of writes. In particular,
BIO_ORDERED was not being honoured. Address this by effectively
freezing the request queue any time a collision with a synchronization
request occurs. The queue is unfrozen once the collision with the
first frozen request is over.
- The above-mentioned collision handling allowed reads to jump ahead
of writes to the same offset. Address this by freezing all request
types when a collision occurs, not just BIO_WRITEs and BIO_DELETEs.
Also add some more fail points for use in testing error handling.
Conrad Meyer [Tue, 2 Jan 2018 17:25:13 +0000 (17:25 +0000)]
rpcbind: Fix race in signal termination
If a signal was delivered while the main thread was not in poll(2) and after
check was performed, we could reenter poll and never detect termination. Fix
this with the pipefd trick. (This race was introduced very recently, in
r327482.)
There has been some fallout from the change. The change itself was not valueable
enough to spend time investigating the corner cases, let's just back it out.
Ed Maste [Tue, 2 Jan 2018 14:07:55 +0000 (14:07 +0000)]
elfcopy: copy raw (untranslated) contents to binary output
Previously elfcopy used elf_getdata to obtain data from ELF sections
being copied to binary output, but elf_getdata returns data that has
been translated - that is, data is in host byte order. When the host and
target differ in endianness (e.g., converting a big-endian MIPS ELF
object to binary on an x86 host) this resulted in byte-swapped data in
certain sections such as .dynamic.
Instead use elf_rawdata to keep data in the original, target endianness.
Reported by: Hiroki Mori <yamori83@yahoo.co.jp>, Bill Yuan
MFC after: 1 week
Sponsored by: The FreeBSD Foundation