Some of the changes in this release:
- IPv6 L4 checksum offload fixes.
- Optimization of the Tx req_id validation.
- Timer service adjustments.
- NUMA awareness for the kernel RSS mode.
Submitted by: Michal Krawczyk <mk@semihalf.com>
Obtained from: Semihalf
MFC after: 2 weeks
Sponsored by: Amazon, Inc.
Verified spelling in the README and fixed the typos.
Also updated the contact section by removing Artur and adding Dawid
Gorecki who is now the second ENA FreeBSD driver developer.
Submitted by: Michal Krawczyk <mk@semihalf.com>
Obtained from: Semihalf
MFC after: 2 weeks
Sponsored by: Amazon, Inc.
If the device becomes unresponsive, the driver will not be able to
finish the reset process correctly. Timeout during version validation
indicates that the device is currently not responding. In that case
do not perform the reset and instead reschedule timer service. Because
of that the driver will continue trying to reset the device until it
succeeds or is detached.
Submitted by: Dawid Gorecki <dgr@semihalf.com>
Obtained from: Semihalf
MFC after: 2 weeks
Sponsored by: Amazon, Inc.
The timer service was started when the interface was brought up and it
was stopped when it was brought down. Since ena_up requires the device
to be responsive, triggering the reset would become impossible if the
device became unresponsive with the interface down.
Since most of the functions in timer service already perform the check
to see if the device is running, this only requires starting the callout
in attach and stopping it when bringing the interface up or down to
avoid race between different admin queue calls.
Since callout functions for timer service are always called with the
same arguments, replace callout_{init,reset,drain} calls with
ENA_TIMER_{INIT,RESET,DRAIN} macros.
Submitted by: Dawid Gorecki <dgr@semihalf.com>
Obtained from: Semihalf
MFC after: 2 weeks
Sponsored by: Amazon, Inc.
Since `ena_com_tx_comp_req_id_get` already checks for `req_id` validity,
the logic was exiting early, never giving `validate_tx_req_id` a chance
to trigger device reset.
Rewrite the logic so that device reset is called based on return value
of `ena_com_tx_comp_req_id_get` instead.
Submitted by: Artur Rojek <ar@semihalf.com>
Obtained from: Semihalf
MFC after: 2 weeks
Sponsored by: Amazon, Inc.
ena_tx_csum function did not check if IPv6 checksum offload was
requested it only checked checksum offloading flags for IPv4 packets.
Because of that, when encountering CSUM_IP6_* flags, the function simply
returned without actually setting checksum offloading in ena_ctx.
Check CUSM_IP6_* flags to enable IPv6 checksum offload.
Additionally, only IPv4 header was being parsed regardless of EtherType
field, because of that, value of L4 protocol read when actually trying
to send IPv6 packets was wrong. Use ip6_lasthdr function to get length
of all IPv6 headers and payload protocol.
Set the DF flag to 1 in order to allow the device to offload the IPv6
checksum calculation and achieve optimal performance.
Add CSUM6_OFFLOAD and CSUM_OFFLOAD definitions into ena_datapath.h.
Submitted by: Dawid Gorecki <dgr@semihalf.com>
Obtained from: Semihalf
MFC after: 2 weeks
Sponsored by: Amazon, Inc.
Merge commit '2530eb1fa01bf28fbcfcdda58bd41e055dcb2e4a'
Adjust the driver to the upgraded ena-com part twofold:
First update is related to the driver's NUMA awareness.
Allocate I/O queue memory in NUMA domain local to the CPU bound to the
given queue, improving data access time. Since this can result in
performance hit for unaware users, this is done only when RSS
option is enabled, for other cases the driver relies on kernel to
allocate memory by itself.
Information about first CPU bound is saved in adapter structure, so
the binding persists after bringing the interface down and up again.
If there are more buckets than interface queues, the driver will try to
bind different interfaces to different CPUs using round-robin algorithm
(but it will not bind queues to CPUs which do not have any RSS buckets
associated with them). This is done to better utilize hardware
resources by spreading the load.
Add (read-only) per-queue sysctls in order to provide the following
information:
- queueN.domain: NUMA domain associated with the queue
- queueN.cpu: CPU affinity of the queue
The second change is for the CSUM_OFFLOAD constant, as ENA platform
file has removed its definition. To align to that change, it has been
added to the ena_datapath.h file.
Submitted by: Artur Rojek <ar@semihalf.com>
Submitted by: Dawid Gorecki <dgr@semihalf.com>
Obtained from: Semihalf
MFC after: 2 weeks
Sponsored by: Amazon, Inc.
Notable upstream pull request merges:
#12766 Fix error propagation from lzc_send_redacted
#12805 Updated the lz4 decompressor
#12851 FreeBSD: Provide correct file generation number
#12857 Verify dRAID empty sectors
#12874 FreeBSD: Update argument types for VOP_READDIR
#12896 Reduce number of arc_prune threads
#12934 FreeBSD: Fix zvol_*_open() locking
#12947 lz4: Cherrypick fix for CVE-2021-3520
#12961 FreeBSD: Fix leaked strings in libspl mnttab
#12964 Fix handling of errors from dmu_write_uio_dbuf() on FreeBSD
#12981 Introduce a flag to skip comparing the local mac when raw sending
#12985 Avoid memory allocations in the ARC eviction thread
Obtained from: OpenZFS
OpenZFS commit: 17b2ae0b24
Disable the gdb packet run length encoding for 3-symbol repetitions.
While it is technically possible to encode them, they have no advantage
over sending the characters verbatim (the resulting length is the same)
and they result in sending non-printable \x1f character. The protocol
has been designed with the intent of avoiding non-printable characters
and therefore the run length encoding is biased to emit \x20 (a space)
with the minimal intended run length of 4.
While at it, simplify the logic by merging the different 'if' blocks
into a single while loop, and moving 'runlen == 0' check lower.
Reviewed by: cem, emaste
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D33686
The text after .error et al is emitted verbatim.
Reviewed by: sjg
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D33904
that disables any access to ptrace(2) for all processes.
Reviewed by: emaste
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D33986
Privilege checks in both functions should allow the current process to
infer information about itself, as well as use the interfaces that are
proclaimed 'debugging', for instance, procctl(2).
Note that in p_cansee() case, explicit comparision of curproc and p
avoids a race where the process might change credentials and cause
thread to compare its cached stale credentials against updated process
creds, effectively disallowing the process to observe itself.
Reviewed by: emaste
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D33986
For non-anonymous swap objects, there is always a reference from the
owner to the object to keep it from recycling. Account for it when
deciding should we query pmap for hardware active references for the
page.
As result, we avoid unneeded calls to pmap_ts_referenced(), which for
non-mapped page means avoiding unneccessary lock and unlock of the pv list.
Reviewed by: markj
Discussed with: alc
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D33924
CLOCK_MONOTONIC_RAW is only a thing on Linux and macOS. I'm not
actually sure why the previous hardcoding of a constant didn't
error out, but when we removed it, it sure does now.
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Co-authored-by: Rich Ercolani <rincebrain@gmail.com>
Signed-off-by: Rich Ercolani <rincebrain@gmail.com>
Closes#12995
FreeBSD's implementation of zfs_uio_fault_move() returns EFAULT when a
page fault occurs while copying data in or out of user buffers. The VFS
treats such errors specially and will retry the I/O operation (which may
have made some partial progress).
When the FreeBSD and Linux implementations of zfs_write() were merged,
the handling of errors from dmu_write_uio_dbuf() changed such that
EFAULT is not handled as a partial write. For example, when appending
to a file, the z_size field of the znode is not updated after a partial
write resulting in EFAULT.
Restore the old handling of errors from dmu_write_uio_dbuf() to fix
this. This should have no impact on Linux, which has special handling
for EFAULT already.
Reviewed-by: Andriy Gapon <avg@FreeBSD.org>
Reviewed-by: Ryan Moeller <ryan@iXsystems.com>
Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes#12964
Raw receiving a snapshot back to the originating dataset is currently
impossible because of user accounting being present in the originating
dataset.
One solution would be resetting user accounting when raw receiving on
the receiving dataset. However, to recalculate it we would have to dirty
all dnodes, which may not be preferable on big datasets.
Instead, we rely on the os_phys flag
OBJSET_FLAG_USERACCOUNTING_COMPLETE to indicate that user accounting is
incomplete when raw receiving. Thus, on the next mount of the receiving
dataset the local mac protecting user accounting is zeroed out.
The flag is then cleared when user accounting of the raw received
snapshot is calculated.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: George Amanakis <gamanakis@gmail.com>
Closes#12981Closes#10523Closes#11221Closes#11294Closes#12594
Issue #11300
Add implementation for the ENA_MEM_ALLOC_NODE and
ENA_MEM_ALLOC_COHERENT_NODE_* macros. Also the signature of
ena_dma_alloc() function was updated, for which the implementation
will be updated in ENA driver's patch.
Submitted by: Artur Rojek <ar@semihalf.com>
Submitted by: Dawid Gorecki <dgr@semihalf.com>
Submitted by: Michal Krawczyk <mk@semihalf.com>
Obtained from: Semihalf
Sponsored by: Amazon, Inc.
CSUM_OFFLOAD is a constant which is used only by the core driver code.
Because of that it shouldn't be defined in the platform file, as it's
not the platform specific code which is used only by the ena_com layer.
Submitted by: Dawid Gorecki <dgr@semihalf.com>
Submitted by: Michal Krawczyk <mk@semihalf.com>
Obtained from: Semihalf
Sponsored by: Amazon, Inc.
When the eviction thread goes to shrink an ARC state, it allocates a set
of marker buffers used to hold its place in the state's sublists.
This can be problematic in low memory conditions, since
1) the allocation can be substantial, as we allocate NCPU markers;
2) on at least FreeBSD, page reclamation can block in
arc_wait_for_eviction()
In particular, in stress tests it's possible to hit a deadlock on
FreeBSD when the number of free pages is very low, wherein the system is
waiting for the page daemon to reclaim memory, the page daemon is
waiting for the ARC eviction thread to finish, and the ARC eviction
thread is blocked waiting for more memory.
Try to reduce the likelihood of such deadlocks by pre-allocating markers
for the eviction thread at ARC initialization time. When evicting
buffers from an ARC state, check to see if the current thread is the ARC
eviction thread, and use the pre-allocated markers for that purpose
rather than dynamically allocating them.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Reviewed-by: George Amanakis <gamanakis@gmail.com>
Signed-off-by: Mark Johnston <markj@FreeBSD.org>
Closes#12985
sizeof(bitfield.member) is invalid, and this shows up in some FreeBSD
build configurations: work around this by !!ing ‒
this makes the sizeof target the ! result type (_Bool), instead
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz>
Fixes: 42aaf0e ("libspl: ASSERT*: mark arguments as used")
Closes#12984Closes#12986
An AHCI controller may be specified with no connected ports. Avoid
dumping core in this case for compatibility with existing VM configs.
Reviewed by: khng, jhb
Fixes: 621b509048 Refactor configuration management in bhyve.
MFC after: 1 week
Sponsored by: iXsystems, Inc.
Differential Revision: https://reviews.freebsd.org/D33969
SystemCMOS address space is accessible for system wide.
So install address handler in \_SB space.
Reviewed by: jhb
Differential Revision: https://reviews.freebsd.org/D33892
atrtc(4) should always install a SystemCMOS address space handler unless
the RTC Not Present bit is not set in IAPC_BOOT_ARCH in the FADT.
The atrtc(4) driver already checks this bit, but _STA can return not-present
even when this bit is clear.
Reviewed by : jhb
Differential Revision: https://reviews.freebsd.org/D33891
X_ip_mrouter_done might sleep, which triggers INVARIANTS to
print additional errors on the screen.
Move it outside the lock, but provide some basic synchronization
to avoid race condition during module uninit/unload.
Obtained from: Semihalf
Sponsored by: Stormshield
As of f833ab9dd1 procctl(2) allows idtype P_PID with id = 0 as a
shortcut for the calling process ID. The shortcut also bypasses the
p_cansee / p_candebug test (since the process is able to act on itself.)
At present if the security.bsd.unprivileged_proc_debug sysctl is 0 then
procctl(P_PID, getpid(), ... for a process to act on itself will fail,
but procctl(P_PID, 0, ... will succeed. This should likely be addressed
with a kernel change.
In any case the id = 0 shortcut is a tiny optimization for a process to
act on itself and allows the self-procctl to succeed, so use it in ssh.
Reported by: Shawn Webb
Reviewed by: kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D33970
Part two: Append bhyve -K option for specified keyboard layout
with layout setting files every languages.
Since the cmd option '-k' was used in the meantime
it was changed to '-K'
PR: 246121
Submitted by: koinec@yahoo.co.jp
Reviewed by: grehan@
Differential Revision: https://reviews.freebsd.org/D29473
MFC after: 4 weeks
- Add usr/lib32/libfoo.so.N for lib/libfoo.so.N.
- Add usr/lib32/foo for usr/lib/foo.
- Treat casper libraries special since they are installed to
/usr/lib32 instead of /usr/lib32/casper and thus map
usr/lib/casper/foo to usr/lib32/foo.
Note that OLD_DIRS and MOVED_LIBS entries are not duplicated, only
OLD_FILES and OLD_LIBS.
Reviewed by: imp, emaste
Sponsored by: The University of Cambridge, Google Inc.
Differential Revision: https://reviews.freebsd.org/D33324
This is a bit more unusual in that the modules dropped their major
version suffix at the same time, so the old files being removed by
MOVED_LIBS in this case are the symlinks to the old libraries.
Reviewed by: emaste
Sponsored by: The University of Cambridge, Google Inc.
Differential Revision: https://reviews.freebsd.org/D33849
Add a MOVED_LIBS variable similar to OLD_LIBS except that MOVED_LIBS
is used for the cases that a library's name doesn't change, but it
just moves between /usr/lib and /lib. This will be used by a future
change to auto-generate lib32 old files entries for which these cases
need to be ignored (a moved library remains in /usr/lib32).
Suggested by: emaste
Reviewed by: emaste
Sponsored by: The University of Cambridge, Google Inc.
Differential Revision: https://reviews.freebsd.org/D33848
- Be more explicit in the difference between OLD_DIRS and OLD_FILES
(the former is only in delete-old-libs whereas the latter is in
delete-old).
- Document that debug symbols in /usr/lib/debug/ for files in
OLD_FILES and OLD_LIBS are removed as well.
Reviewed by: emaste
Sponsored by: The University of Cambridge, Google Inc.
Differential Revision: https://reviews.freebsd.org/D33847
Otherwise we end up copying one uninitialized byte into the socket
buffer.
Reported by: KMSAN
Reviewed by: jhb
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D33953
zdb -d <pool>/<objset ID> does not work when
other command line arguments are included i.e.
zdb -U <cachefile> -d <pool>/<objset ID>
This change fixes the command line parsing
to handle this situation. Also fix issue
where zdb -r <dataset> <file> does not handle
the root <dataset> of the pool. Introduce -N
option to force <objset ID> to be interpreted
as a numeric objsetID.
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Rich Ercolani <rincebrain@gmail.com>
Reviewed-by: Tony Nguyen <tony.nguyen@delphix.com>
Signed-off-by: Paul Zuchowski <pzuchowski@datto.com>
Closes#12845Closes#12944
These routines are used internally by GEOM to dispatch I/O requests to a
provider, typically for tasting or for updating GEOM class metadata
blocks.
These routines assumed that partial I/O did not occur without setting
BIO_ERROR, but this is possible in at least two cases:
- Some or all of the I/O range is beyond the provider's mediasize.
In this scenario g_io_check() truncates the bounds of the request
before it is handed to the target provider.
- A read from vnode-backed md(4) device returns EOF (the backing vnode
is allowed to be smaller than the device itself) or partial vnode I/O
occurs.
In these scenarios g_read_data() could return a partially uninitialized
buffer. Many consumers are not affected by the first case, since the
offsets used for provider metadata or tasting are relative to the
provider's mediasize, but in some cases metadata is read at fixed
offsets, such as when searching for a UFS superblock using the offsets
defined by SBLOCKSEARCH.
Thus, modify the routines to explicitly check for a non-zero residual
and return EIO in that case. Remove a related check from the
DIOCGDELETE ioctl handler, it is handled within g_delete_data() now.
Reviewed by: mav, imp, kib
Reported by: KMSAN
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D31293
tc_counter_mask is an unsigned int and in the TSC timecounter is equal
to UINT_MAX, so the addition tc->tc_counter_mask + 1 can overflow to 0,
resulting in a hang during boot.
Fixes: c2705ceaeb ("x86: Speed up clock calibration")
Reviewed by: cperciva
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D33956
