Add support for IFCOUNTER_IPACKETS, IFCOUNTER_OPACKETS,
IFCOUNTER_OBYTES, IFCOUNTER_OMCASTS, IFCOUNTER_OERRORS, and
IFCOUNTER_OQDROPS.
This allows tools like systat to report the incoming and outgoing
bandwidth.
Reviewed by: dsl, Timo Völker
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D54893
Stub out both experimental functions introduced in ecc039be7f when
doing embedded builds.
Fixes: ecc039be7f ("nanobsd: Add a NO_ROOT build option")
MFC after: 3 days
This breaks the riscv64 build; clearly the version we have in-tree is
not new enough to work. That or our build system needs tweaking. Either
way, we're not ready for this, and enabling it without testing was
misguided.
This reverts commit c59a47dc6c.
The try_binary_file() function has inverted logic for the
FIRMWARE_GET_NOWARN flag. When the flag is set (meaning "don't warn"),
the code sets warn=true and makes noise anyway.
Invert the assignment to warn to correctly suppress warnings when
FIRMWARE_GET_NOWARN is set.
Reviewed by: kevans
Differential Revision: https://reviews.freebsd.org/D54955
maybe_resched() needs to schedule TDA_SCHED for curthread, but this
requires taking curthread lock while owning some other thread lock.
To avoid introducing the order:
- Use a scheduler-private TDP flag.
- Register an unconditional TDA_SCHED_PRIV for 4BSD.
When an AST needs to be scheduled, i.e. the current thread must do
context switch in the return to userspace path, set the flag.
Then the ast handler calls ast_scheduler(), which gives the same
effect as scheduling TDA_AST.
The overhead is a single function call on each userspace return, for
4BSD case.
Reported and tested by: pho (previous version)
Reviewed by: markj
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54945
I have hit the case multiple times that some LinuxKPI field may not be
set during driver bringup and lindebugfs would cause a panic.
The backtrace goes like:
strlen() at strlen+0x54
pfs_create_dir() at pfs_create_dir+0x41
debugfs_create_dir() at debugfs_create_dir+0xa1
...
While the problem is clearly in LinuxKPI or the driver, we likely
should at least add an assert to pfs_create_dir() if name is NULL
like we have for pfs_add_node() but for lindebugfs at least make
this a graceful error and continue without creating the dir instead
of panicing.
Sponsored by: The FreeBSD Foundation
MFC after: 3 days
Reviewed by: kevans
Differential Revision: https://reviews.freebsd.org/D54944
Globally enable both schedulers for LINT.
Enable both schedulers for GENERIC on amd64.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
It is apparently should be considered part of the ABI, and is used by
the base top(1). But do not declare the ccpu variable in headers, it is
needed only by 4bsd. So put the variable definition into sched_shim.c to
make the kernel buildable without SCHED_4BSD.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
Add sched_find_l2_neighbor(). This really should be not
scheduler-depended, in does not have anything to do with scheduler at
all. But for now keep the same code structure.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
For !SCHED_ULE, even if the fence is not needed, it is harmless.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
Move the code to decide on the timer accounting into the scheduler hook.
Since there were no inclusion of opt_sched.h, it is probably done
unconditionally anyway.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
It is nop for 4BSD.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
This is the infrastructure to allow scheduler implementation to be
selected on boot, supported by ifuncs.
The DEFINE_SHIM() macros and their usage provided by jrtc27@.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
It is not used since c72188d85a.
Noted by: jrtc27
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
There is no harm in having unused prototypes.
Reviewed by: olce
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D54831
Adopt the POSIX standard text to our implementation.
PR: 206284
Reviewed by: des, jilles, ziaee
Differential Revision: https://reviews.freebsd.org/D49895
We always use NO_ROOT for release artifact builds, so remove the
alternate code paths.
For the first step we set NO_ROOT unconditionally in cases that invoke
submakes, and turn NO_ROOT being unset into an error in lover-level
targets so that we can catch potential out-of-tree build scripts (or
missed in-tree cases) that expect to run not-NO_ROOT builds. The second
step will be to remove those entirely.
Reviewed by: cperciva
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D54179
Previously connect() or sendto() to INADDR_ANY or IN6ADDR_ANY reached
some socket bound to some host interface address. Although this was
intentional it was an artifact of a different era, and is not desirable
now.
In 417b35a97b markj added support to disallow connect() to INADDR_ANY
and IN6ADDR_ANY. Connections to INADDR_ANY were disabled by default in
cd240957d7. Follow suit with IN6ADDR_ANY.
Reviewed by: glebius, markj, zlei
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D54306
On some environement a VM can be connected to multiple ifaces but only
one will respond to dhcp and able to fetch the metadata.
Launch the early dhclient on all available ifaces and as soon as one
dhcp has successfully configured an iface, proceed with fetching the
metadata.
MFC After: 1 week
Reported by: mateusz.klejn@ovhcloud.com
Tested by: mateusz.klejn@ovhcloud.com
Sponsored by: OVHcloud
This prevents 'service -e' from emitting (/var/log/messages):
/usr/sbin/service: WARNING: $virtual_oss_enable is not set properly - see rc.conf(5)
Pull Request: https://github.com/freebsd/freebsd-src/pull/1987
Reviewed by: christos
Signed-off-by: eborisch@gmail.com
MFC after: 1 week
Clang's -Wmissing-variable-declarations flags this as an issue since the
flag is only used in `gtest.cc`. Declare the flag beforehand to ensure
that the variable scope is properly limited to `gtest.cc`.
MFC after: 1 week
MFC with: 46333229c6
Ref: https://github.com/google/googletest/pull/4898
The changes between the two versions can be found in this diff of the
two release tags:
https://github.com/google/googletest/compare/v1.15.2...v1.17.0
One notable change is that GoogleTest 1.17.0 now requires C++-17 to
build.
MFC after: 1 week
Merge commit '3a4c29b5bed4ea20266ad9371fbfdc6bca088f92'
Previously it was enabled for all architectures except but RISC-V.
Upstream made significant progress on RISC-V support over the last
year[1], and although there's still some FreeBSD-specific work to do we
can start building it now.
[1] https://jonasdevlieghere.com/post/lldb-2025/
Sponsored by: The FreeBSD Foundation
Michael Tuexen