fget_cap() tries to do a cheaper snapshot of a file descriptor without
holding the file descriptor lock. This snapshot does not do a deep
copy of the ioctls capability array, but instead uses a different
return value to inform the caller to retry the copy with the lock
held. However, filecaps_copy() was returning 1 to indicate that a
retry was required, and fget_cap() was checking for 0 (actually
'!filecaps_copy()'). As a result, fget_cap() did not do a deep copy
of the ioctls array and just reused the original pointer. This cause
multiple file descriptor entries to think they owned the same pointer
and eventually resulted in duplicate frees.
The only code path that I'm aware of that triggers this is to create a
listen socket that has a restricted list of ioctls and then call
accept() which calls fget_cap() with a valid filecaps structure from
getsock_cap().
To fix, change the return value of filecaps_copy() to return true if
it succeeds in copying the caps and false if it fails because the lock
is required. I find this more intuitive than fixing the caller in
this case. While here, change the return type from 'int' to 'bool'.
Finally, make filecaps_copy() more robust in the failure case by not
copying any of the source filecaps structure over. This avoids the
possibility of leaking a pointer into a structure if a similar future
caller doesn't properly handle the return value from filecaps_copy()
at the expense of one more branch.
I also added a test case that panics before this change and now passes.
Reviewed by: kib
Discussed with: mjg (not a fan of the extra branch)
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D15047
This behavior is already documented by the man page, and suggested by POSIX.
Reviewed by: jhb
MFC after: 3 weeks
Differential Revision: https://reviews.freebsd.org/D15099
Do not build or install pf tests if WITHOUT_PF is set. This fixes the build
failure with WITHOUT_PF=yes.
Reported by: Vladimir Zakharov <zakharov.vv@gmail.com>
There was a memory leak in the DIOCRADDTABLES ioctl() code which could
be triggered by trying to add tables with the same name.
Try to provoke this memory leak. It was fixed in r331225.
MFC after: 1 week
Validate the DIOCRGETTABLES, DIOCRGETTSTATS, DIOCRCLRTSTATS and
DIOCRSETTFLAGS ioctls with invalid values. These may succeed (because
the kernel uses the minimally required size, not the specified size),
but should not trigger kernel panics.
MFC after: 1 week
The upstream repository is on github BLAKE2/libb2. Files landed in
sys/contrib/libb2 are the unmodified upstream files, except for one
difference: secure_zero_memory's contents have been replaced with
explicit_bzero() only because the previous implementation broke powerpc
link. Preferential use of explicit_bzero() is in progress upstream, so
it is anticipated we will be able to drop this diff in the future.
sys/crypto/blake2 contains the source files needed to port libb2 to our
build system, a wrapped (limited) variant of the algorithm to match the API
of our auth_transform softcrypto abstraction, incorporation into the Open
Crypto Framework (OCF) cryptosoft(4) driver, as well as an x86 SSE/AVX
accelerated OCF driver, blake2(4).
Optimized variants of blake2 are compiled for a number of x86 machines
(anything from SSE2 to AVX + XOP). On those machines, FPU context will need
to be explicitly saved before using blake2(4)-provided algorithms directly.
Use via cryptodev / OCF saves FPU state automatically, and use via the
auth_transform softcrypto abstraction does not use FPU.
The intent of the OCF driver is mostly to enable testing in userspace via
/dev/crypto. ATF tests are added with published KAT test vectors to
validate correctness.
Reviewed by: jhb, markj
Obtained from: github BLAKE2/libb2
Differential Revision: https://reviews.freebsd.org/D14662
It mistakenly believes the 'static' keyword must come first. Fix PPC,
Sparc64, and maybe MIPS world. Fallout from r331279.
Reported by: tinderbox (results come slowly)
The general idea here is to provide userspace programs with well-defined
sources of entropy, in a fashion that doesn't require opening a new file
descriptor (ulimits) or accessing paths (/dev/urandom may be restricted
by chroot or capsicum).
getrandom(2) is the more general API, and comes from the Linux world.
Since our urandom and random devices are identical, the GRND_RANDOM flag
is ignored.
getentropy(3) is added as a compatibility shim for the OpenBSD API.
truss(1) support is included.
Tests for both system calls are provided. Coverage is believed to be at
least as comprehensive as LTP getrandom(2) test coverage. Additionally,
instructions for running the LTP tests directly against FreeBSD are provided
in the "Test Plan" section of the Differential revision linked below. (They
pass, of course.)
PR: 194204
Reported by: David CARLIER <david.carlier AT hardenedbsd.org>
Discussed with: cperciva, delphij, jhb, markj
Relnotes: maybe
Differential Revision: https://reviews.freebsd.org/D14500
The 'physio' fast-path used by AIO requests on md(4) devices, is not
gated on the unsafe_aio knob. Prior to r327755, some AIO requests could
fail the fast-path and fall back to the slow-path (requests for devices
not supporting unmapped I/O and requests which failed with EFAULT during
the fast-path). However, those cases now return a suitable error rather
than using the slow-path.
PR: 217261
Reviewed by: asomers
Sponsored by: Chelsio Communications
Differential Revision: https://reviews.freebsd.org/D14742
The code assumed that disks (devices) used for testing are always named
like /dev/foo, but there is no reason for that restriction and we can
easily support paths like /dev/stripe/bar.
The change is to account for a different order in which the recursive
destroy may be attempted. If we first try a dataset that can be destroyed
then it will be destroyed, but if we first try a dataset that cannot be
destroyed then we will not attempt to destroy the other dataset.
It was originally written by Sun as part of the STF (Solaris test framework).
They open sourced it in OpenSolaris, then HighCloud partially ported it to
FreeBSD, and Spectra Logic finished the port. We also added many testcases,
fixed many broken ones, and converted them all to the ATF framework. We've had
help along the way from avg, araujo, smh, and brd.
By default most of the tests are disabled. Set the disks Kyua variable to
enable them.
Submitted by: asomers, will, justing, ken, brd, avg, araujo, smh
Sponsored by: Spectra Logic Corp, HighCloud
If the underlying provider's physical path is null, then the gpart device's
physical path will be, too. Otherwise, it will append the partition name,
such as "/p1" or "/s1/a". This will make gpart work better with zfsd(8).
PR: 224965
MFC after: 3 weeks
Differential Revision: https://reviews.freebsd.org/D14010
If the underlying provider's physical path is null, then the geli device's
physical path will be, too. Otherwise, it will append "/eli". This will make
geli work better with zfsd(8).
PR: 224962
MFC after: 3 weeks
Differential Revision: https://reviews.freebsd.org/D13979
Don't declare some types that FreeBSD incorrectly declares.
Fix an incorrect call to open() (missing mode).
ANSIfy prototypes.
Enable SysV message queue, semaphore, and shared memory tests.
With exception of the workaround for union semun, these fixes have been
committed to NetBSD.
Reviewed by: asomers
Approved by: CheriBSD
Sponsored by: DARPA, AFRL
Differential Revision: https://reviews.freebsd.org/D13471
* Registers TRNG source for random(4)
* Finds available queues, LSBs; allocates static objects
* Allocates a shared MSI-X for all queues. The hardware does not have
separate interrupts per queue. Working interrupt mode driver.
* Computes SHA hashes, HMAC. Passes cryptotest.py, cryptocheck tests.
* Does AES-CBC, CTR mode, and XTS. cryptotest.py and cryptocheck pass.
* Support for "authenc" (AES + HMAC). (SHA1 seems to result in
"unaligned" cleartext inputs from cryptocheck -- which the engine
cannot handle. SHA2 seems to work fine.)
* GCM passes for block-multiple AAD, input lengths
Largely based on ccr(4), part of cxgbe(4).
Rough performance averages on AMD Ryzen 1950X (4kB buffer):
aesni: SHA1: ~8300 Mb/s SHA256: ~8000 Mb/s
ccp: ~630 Mb/s SHA256: ~660 Mb/s SHA512: ~700 Mb/s
cryptosoft: ~1800 Mb/s SHA256: ~1800 Mb/s SHA512: ~2700 Mb/s
As you can see, performance is poor in comparison to aesni(4) and even
cryptosoft (due to high setup cost). At a larger buffer size (128kB),
throughput is a little better (but still worse than aesni(4)):
aesni: SHA1:~10400 Mb/s SHA256: ~9950 Mb/s
ccp: ~2200 Mb/s SHA256: ~2600 Mb/s SHA512: ~3800 Mb/s
cryptosoft: ~1750 Mb/s SHA256: ~1800 Mb/s SHA512: ~2700 Mb/s
AES performance has a similar story:
aesni: 4kB: ~11250 Mb/s 128kB: ~11250 Mb/s
ccp: ~350 Mb/s 128kB: ~4600 Mb/s
cryptosoft: ~1750 Mb/s 128kB: ~1700 Mb/s
This driver is EXPERIMENTAL. You should verify cryptographic results on
typical and corner case inputs from your application against a known- good
implementation.
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D12723
Scapy requires the Raw payload to be a string, which was not the case. This
caused the pft_ping.py script to fail, which in turn caused the test to fail.
Reduce the geli tests' runtime by about a third:
* In integrity_test:copy, use a file-backed md(4) device instead of a
malloc'd one. That way we can corrupt the underlying storage without
needing to detach and reattach the geli device.
* In integrity_test:{copy, hmac, data} and onetime_test:{onetime,
onetime_a}, move reads of /dev/random out of the loop.
MFC after: 2 weeks
The trick is not to destroy an md(4) device during a test. That can create
a "double-free" situation, because we also destroy md devices during test
cleanup.
MFC after: 2 weeks
I'm leaving readonly_test and nokey_test alone for now. In a future commit
they should be broken up into several smaller test cases and distributed
between multiple files.
Reviewed by: ngie
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D13717
The resize test used bsdlabel(8), which is not available on all
architectures. Change it to use gpart(8) instead, which should be available
everywhere.
PR: 221763
Reported by: andrew
MFC after: 2 weeks
This change mostly reverts r293436, which introduced the bug due to a belief
that geli(8) would allocate md(4) devices by itself. However, that belief is
incorrect. Instead of using linear probing to find available md(4) numbers,
it's best to use the existing attach_md function.
Reviewed by: ngie
MFC after: 2 weeks
Sponsored by: Spectra Logic Corp
Differential Revision: https://reviews.freebsd.org/D13666
Some of the ptrace tests need to wait for a child process to become a
zombie before preceding. The parent process polls the child process
via the kern.proc.pid sysctl to wait for it to become a zombie.
Previously the code polled until the sysctl failed with ESRCH. Now it
will poll until either the sysctl fails with ESRCH (for compatiblity
with older kernels) or returns a kinfo_proc structure with the ki_stat
field set to SZOMB.
Reported by: Jenkins
Tested by: markj
Discussed with: mjg
MFC after: 1 week
