Kevin Bowling [Tue, 17 May 2022 03:15:34 +0000 (20:15 -0700)]
igc: Increase rx_buffer_size local variable to 32b
Apply 6987c47569b377f4b6eba9966afdedfb1b39fca8 to igc. This is not
expected to have any benefit on current parts with current observed PBA
sizes but will avoid surprises if they are increased in future chips.
Approved by: grehan
MFC after: 3 days
Differential Revision: https://reviews.freebsd.org/D35217
Alan Somers [Mon, 16 May 2022 22:32:10 +0000 (16:32 -0600)]
makefs: fix calculation of file sizes
When a new FS image is created we need to calculate how much space each
file is going to consume.
Fix two bugs in that logic:
1) Count the space needed for indirect blocks for large files.
1) Normally the trailing data of a file is written to a block of frag
size, 4 kB by default.
However for files that use indirect blocks a full block is allocated,
32kB by default. Take that into account.
Adjust size calculations to match what is done in ffs_mkfs routine:
* Depending on the UFS version the superblock is stored at a different
offset. Take that into account.
* Add the cylinder group block size.
* All of the above has to be aligned to the block size.
Finally, Remove "ncg" variable. It's always 1 and it was used to
multiply stuff.
This amongst other things removes the deprecated BCAST_FILTERING option,
which we had disabled before due to firmware issues.
Also prepares us for the next firmware update.
The "enable_ini" module option has become an integer with a
sysctl proc equivalent handler. Disable the module parameter changing
for the moment until we'll have working LinuxKPI support or implement
it as SYSCTL_PROC in FreeBSD directly.
Sponsored by: The FreeBSD Foundation
MFC after: 3 days
Rick Macklem [Sun, 15 May 2022 18:51:56 +0000 (11:51 -0700)]
krpc: Fix NFS-over-TLS for KTLS1.3
When NFS-over-TLS uses KTLS1.3, the client can receive
post-handshake handshake records. These records can be
safely thown away, but are not handled correctly via the
rpctls_ct_handlerecord() upcall to the daemon.
Commit 373511338d95 changed soreceive_generic() so that it
will only return ENXIO for Alert records when MSG_TLSAPPDATA
is specified. As such, the post-handshake handshake
records will be returned to the krpc.
This patch modifies the krpc so that it will throw
these records away, which seems sufficient to make
NFS-over-TLS work with KTLS1.3. This change has
no effect on the use of KTLS1.2, since it does not
generate post-handshake handshake records.
Dmitry Chagin [Sun, 15 May 2022 18:10:50 +0000 (21:10 +0300)]
linux(4): Implement signal trampoline for arm64 in a FreeBSD-way
The implemenation differs from others Linuxulators.
For unwinders Linux ucontext_t is stored, however native machine context
is used to store/restore process state to avoid code duplication.
As DWARF Aarch64 does not define a register number for PC and provides no
direct way to encode the PC of the previous frame, CFI cannot describe a
signal trampoline frame. So, modified the vdso linker script to discard
unused sections.
Dmitry Chagin [Sun, 15 May 2022 18:08:12 +0000 (21:08 +0300)]
linux(4): Retire handmade DWARF annotations from signal trampolines
The Linux exports __kernel_sigreturn and __kernel_rt_sigreturn from the
vdso. Modern glibc's sigaction sets the sa_restorer field of sigaction
to the corresponding vdso __sigreturn, and sets the SA_RESTORER.
Our signal trampolines uses the FreeBSD-way to call a signal handler,
so does not use the sigaction's sa_restorer.
However, as glibc's runtime linker depends on the existment of the vdso
__sigreturn symbols, for all Linuxulators was added separate trampolines
named __sigcode with DWARF anotations and left separate __sigreturn
methods, which are exported.
Dmitry Chagin [Sun, 15 May 2022 18:05:01 +0000 (21:05 +0300)]
linux(4): Rework the definition of struct siginfo to match Linux actual one
Rework the defintion of struct siginfo so that the array padding
struct siginfo to SI_MAX_SIZE can be placed in a union along side of the
rest of the struct siginfo members. The result is that we no longer need
the __ARCH_SI_PREAMBLE_SIZE or SI_PAD_SIZE definitions.
Move struct siginfo definition under /compat/linux to reduce MD part.
To avoid headers polution include linux_siginfo.h in the MD linux.h
Dmitry Chagin [Sun, 15 May 2022 18:00:05 +0000 (21:00 +0300)]
linux(4): Cleanup signal trampolines
This is the first stage of a signal trampolines refactoring.
From trampolines retired emulation of the 'call' instruction, which is
replaced by direct call of a signal handler. The signal handler address
is in the register.
The previous trampoline implemenatation used semi-Linux-way to call
a signal handler via the 'jmp' instruction. Wherefore the trampoline
emulated a 'call' instruction to into the stack the return address for
signal handler's 'ret' instruction. Wherefore handmade DWARD annotations
was used.
While here rephrased and removed excessive comments.
Dmitry Chagin [Sun, 15 May 2022 17:58:47 +0000 (20:58 +0300)]
linux(4): Retire unneeded initialization
Both uc_flags and uc_link are zeroed above. On amd64 and i386 the
uc_link field is not used at all. The UC_FP_XSTATE bit should be set
in the uc_flags if OS xsave knob is turned on (and xsave is implemented).
Dimitry Andric [Sat, 14 May 2022 20:07:01 +0000 (22:07 +0200)]
Avoid adding -d to kernel module link command lines for lld >= 14
Since 0b3178a45cd0 we have added '-d' to the link command line for
kernel modules, so if any unexpected common symbols turn up (even though
we use -fno-common in CFLAGS), storage will be allocated in in the
module itself.
However, with lld this option did not have any effect since ~2017, and
as of lld 14 it warns: "-d, -dc, -dp, and --[no-]define-common will be
removed. See https://github.com/llvm/llvm-project/issues/53660"
Add a linker type and version check, to avoid adding the option for lld
14 and later.
Rick Macklem [Sat, 14 May 2022 19:56:50 +0000 (12:56 -0700)]
uipc_socket.c: Modify MSG_TLSAPPDATA to only do Alert Records
Without this patch, the MSG_TLSAPPDATA flag would cause
soreceive_generic() to return ENXIO for any non-application
data record in a TLS receive stream.
This works ok for TLS1.2, since Alert records appear to be
the only non-application data records received.
However, for TLS1.3, there can be post-handshake handshake
records, such as NewSessionKey sent to the client from the
server. These handshake records cannot be handled by the
upcall which does an SSL_read() with length == 0.
It appears that the client can simply throw away these
NewSessionKey records, but to do so, it needs to receive
them within the kernel.
This patch modifies the semantics of MSG_TLSAPPDATA slightly,
so that it only applies to Alert records and not Handshake
records. It is needed to allow the krpc to work with KTLS1.3.
Mitchell Horne [Sat, 14 May 2022 13:25:21 +0000 (10:25 -0300)]
dumpon: warn if the configured netdump link is down
Previously we expected the DIOCSKERNELDUMP ioctl to return ENXIO if the
interface was down, but it does not actually do this. Grab the link
status using getifaddrs(3) instead, and downgrade this case from an
error to a warning; the user might bring the link back up at a later
time.
Reviewed by: cem
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D35196
Mitchell Horne [Sat, 14 May 2022 13:23:58 +0000 (10:23 -0300)]
netdump: check the support status of the interface
If the interface does not support debugnet(4) we should bail early,
rather than having the user find this out at the time of the panic.
dumpon(8) already expects this return value and will print a helpful
error message.
Mitchell Horne [Sat, 14 May 2022 13:23:08 +0000 (10:23 -0300)]
debugnet: fix an errant assertion
We may call debugnet_free() before g_debugnet_pcb_inuse is true,
specifically in the cases where the interface is down or does not
support debugnet. pcb->dp_drv_input is used to hold the real driver
if_input callback while debugnet is in use, so we can check the status
of this field in the assertion.
This can be triggered trivially by trying to configure netdump on an
unsupported interface at the ddb prompt.
Initializing the dp_drv_input field to NULL explicitly is not necessary
but helps display the intent.
Apply libc++ fix for <ranges> not compiling at all
Merge commit 44cdca37c01a from llvm git (by Arthur O'Dwyer):
[libc++] Define `namespace views` in its own detail header.
Discovered in the comments on D118748: we would like this namespace
to exist anytime Ranges exists, regardless of whether concepts syntax
is supported. Also, we'd like to fully granularize the <ranges> header,
which means not putting any loose declarations at the top level.
Dimitry Andric [Tue, 15 Mar 2022 20:32:06 +0000 (21:32 +0100)]
Apply libc++ fix for firefox visibility push/pop issue
Merge commit 7ab1ab0db401 from llvm git (by Dimitry Andric):
[libc++] Make __dir_stream visibility declaration consistent
The class `__dir_stream` is currently declared in two places: as a
top-level forward declaration in `directory_iterator.h`, and as a friend
declaration in class `directory_entry`, in `directory_entry.h`.
The former has a `_LIBCPP_HIDDEN` attribute, but the latter does not,
causing the Firefox build to complain about the visibility not matching
the previous declaration. This is because Firefox plays games with
pushing and popping visibility.
Work around this by making both `__dir_stream` declarations consistently
use `_LIBCPP_HIDDEN`.
Dimitry Andric [Thu, 17 Feb 2022 21:49:39 +0000 (22:49 +0100)]
Apply fix for asm constraint error in www/php80-opcache
Merge commit 027c16bef4b7 from llvm git (by Nick Desaulniers):
[X86ISelLowering] permit BlockAddressSDNode "i" constraints for PIC
When building 32b x86 code as PIC, the existing handling of "i"
constraints is conservative since generally we have to go through the
GOT to find references to functions.
But generally, BlockAddresses from C code refer to the Function in the
current TU. Permit BlockAddresses to be used with the "i" constraint
for those cases.
I regressed this in
commit 4edb9983cb8c ("[SelectionDAG] treat X constrained labels as i for asm")
This updates llvm, clang, compiler-rt, libc++, libunwind, lld, lldb and
openmp to llvmorg-14-init-18294-gdb01b123d012, the last commit before
the upstream release/14.x branch was created.
Dmitry Chagin [Sat, 14 May 2022 11:07:20 +0000 (14:07 +0300)]
sysvsem: Fix a typo
Per jamie@ rpr can be NULL if the jail is created with sysvsem=disable.
But at least it doesn't appear to be fatal, since rpr is never dereferenced
but is only compared to other prison pointers.
Justin Hibbits [Fri, 13 May 2022 22:18:25 +0000 (17:18 -0500)]
libm: Add feenableexcept and fedisableexcept to library on powerpc
Summary:
These functions are missing from the library itself, and exist solely in
the header. This breaks a few ports that expect libm to have the
symbols in the library itself.
Questions on MFC-ability: Can this be MFC'd to 13.2, and how?
Eric van Gyzen [Thu, 12 May 2022 14:50:02 +0000 (09:50 -0500)]
sysv test: properly wait for children
In the msg and shm tests, if the child exited before the parent
entered sigsuspend(), the test would hang and time out. This was
also a problem in the sem test, but the misuse of atf_tc_pass()
masked it. Adding a short sleep before the sigsuspend() calls made
the hang 100% reliable. With the same sleep in the new version,
the test passes reliably.
Remove calls to atf_tc_pass(). The call in the sem test broke the test
by exiting prematurely, after only one child out of five had finished.
The other two were harmless but unhelpful.
Reduce a one-second sleep to a more reasonable duration so I can quickly
run many iterations of the test.
Where feasible, assert that wait() returns the child PID. While I'm here,
use the more succinct ATF_REQUIRE* instead of if/atf_tc_fail/else.
Flush stdout before forking to avoid double-flush.
Use errx() when errno is irrelevant.
Don't use ATF_REQUIRE* in children. Apparently, the output doesn't
get saved. The exit status works, so it fails correctly, but silently.
Mitchell Horne [Fri, 13 May 2022 13:39:29 +0000 (10:39 -0300)]
kerneldump: remove physical argument from d_dumper
The physical address argument is essentially ignored by every dumper
method. In addition, the dump routines don't actually pass a real
address; every call to dump_append() passes a value of zero for
physical.
Kyle Evans [Fri, 13 May 2022 03:03:32 +0000 (22:03 -0500)]
arm64: disable the EL2 MMU before dropping to EL1
An earlier stage may have set HCR_EL2.E2H, the clearing of which may
break address translation. We don't need the EL2 MMU at this point, so
we can avoid re-enabling it for now and just drop to EL1 as usual.
Suggested by: andrew
Reviewed by: andrew
Differential Revision: https://reviews.freebsd.org/D34644
Alan Somers [Wed, 4 May 2022 23:36:17 +0000 (17:36 -0600)]
fusefs: handle evil servers that return illegal inode numbers
* If during FUSE_CREATE, FUSE_MKDIR, etc the server returns the same
inode number for the new file as for its parent directory, reject it.
Previously this would triggers a recurse-on-non-recursive lock panic.
* If during FUSE_LINK the server returns a different inode number for
the new name as for the old one, reject it. Obviously, that can't be
a hard link.
* If during FUSE_LOOKUP the server returns the same inode number for the
new file as for its parent directory, reject it. Nothing good can
come of this.
Gleb Smirnoff [Thu, 12 May 2022 20:22:12 +0000 (13:22 -0700)]
sockets: use socket buffer mutexes in struct socket directly
Since c67f3b8b78e the sockbuf mutexes belong to the containing socket,
and socket buffers just point to it. In 74a68313b50 macros that access
this mutex directly were added. Go over the core socket code and
eliminate code that reaches the mutex by dereferencing the sockbuf
compatibility pointer.
This change requires a KPI change, as some functions were given the
sockbuf pointer only without any hint if it is a receive or send buffer.
This change doesn't cover the whole kernel, many protocols still use
compatibility pointers internally. However, it allows operation of a
protocol that doesn't use them.
Kristof Provost [Wed, 11 May 2022 16:51:21 +0000 (18:51 +0200)]
pf: fix pf_rule_to_actions()
If we already had a pipe set in the actions struct we need to take care
to clear the flag if we're overwriting it with a queue.
This can happen if we've got Ethernet rules setting a dummynet pipe.
It does this indirectly, by adding the dummynet information to a pf_mtag
associated with the mbuf.
Kristof Provost [Tue, 10 May 2022 07:23:36 +0000 (09:23 +0200)]
pf: tag dummynet'd route-to packets with their real destination
If we delay route-to/dup-to/reply-to through dummynet we are eventually
returned to pf_test(). At that point we no longer have the context for
the route-to destination. We'd just skip the pf_test() and continue
processing. This means that route-to did not work as expected.
Extend pf_mtag to carry the route-to destination so we can apply it when
we re-enter pf_test().
Kristof Provost [Mon, 9 May 2022 09:11:42 +0000 (11:11 +0200)]
pf: also apply dummynet to route-to/dup-to packets
If packets are processed by a route-to/dup-to/reply-to rule (i.e. they
pass through pf_route(6)) dummynet was not applied to them.
This is because pf_route(6) passes packets directly to ifp->if_output(),
so the dummynet functions were never called.
Factor out the dummynet code and call dummynet prior to
ifp->if_output(). This has a secondary benefit of reducing some code
duplication between the IPv4 and IPv6 paths.
Gleb Smirnoff [Thu, 12 May 2022 18:02:50 +0000 (11:02 -0700)]
sockets: delete stale comment from sofree()
First paragraph refers to old past "we used to" and is no longer
important today. Second paragraph has just a wrong statement that
socket buffer is destroyed before pru_detach.
Kevin Bowling [Thu, 12 May 2022 15:38:09 +0000 (08:38 -0700)]
e1000: Increase rx_buffer_size to 32b
Extend the size of the local rx_buffer_size variable to account for
larger buffer sizes possible on 82580, i350 chips.
From i350 datasheet, 6.2.10 Initialization Control 4 (LAN Base Address
+ Offset 0x13):
When 4 ports are enabled maximum buffer size is 36 KB. When 2 ports are
enabled maximum buffer size is 72 KB. When only a single port is
enabled maximum buffer size is 144 KB.
and 8.3:
The overall available internal buffer size in the I350 for all ports is
144 KB for receive buffers and 80 KB for transmit Buffers. Disabled
ports memory can be shared between active ports and sharing can be
asymmetric. The default buffer size for each port is loaded from the
EEPROM on initialization.
From the reporter:
But for I350 when only 2 ports are used PBA size can be set as 72KB
(see datasheet RXPbsize or e1000_rxpbs_adjust_82580 function in
e1000_82575.c). In this case calculating the rx_buffer_size overflows
as 0x0048 << 10 = 73728 or 0x12000 pushed into u16. It is then set as
0x2000 or 8192.
Ed Maste [Tue, 10 May 2022 14:08:21 +0000 (10:08 -0400)]
ssh: update sshd_config for prohibit-password option
The PermitRootLogin option "prohibit-password" was added as a synonym
for "without-password" in 2015. Then in 2017 these were swapped:
"prohibit-password" became the canonical option and "without-password"
became a deprecated synonym (in OpenSSH commit 071325f458).
The UsePAM description in sshd_config still mentioned
"without-password." Update it to match the new canonical option.
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Ed Maste [Thu, 12 May 2022 14:49:30 +0000 (10:49 -0400)]
Hide -fuse-ld= not supported message for non-build targets
In some build configurations a warning about (an absolute path for)
-fuse-ld= not being supported by GCC was emitted during cleandir or
other non-build make targets.
For these non-build targets COMPILER_TYPE is set to "none" but we
treated the .else case for COMPILER_TYPE==clang as implying gcc.
Check instead for COMPILER_TYPE==gcc.
PR: 263913
Reported by: pstef
Reviewed by: pstef
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Dmitry Chagin [Wed, 11 May 2022 07:36:01 +0000 (10:36 +0300)]
Use Linux semantics for the thread affinity syscalls.
Linux has more tolerant checks of the user supplied cpuset_t's.
Minimum cpuset_t size that the Linux kernel permits in case of
getaffinity() is the maximum CPU id, present in the system / NBBY,
the maximum size is not limited.
For setaffinity(), Linux does not limit the size of the user-provided
cpuset_t, internally using only the meaningful part of the set, where
the upper bound is the maximum CPU id, present in the system, no larger
than the size of the kernel cpuset_t.
Unlike FreeBSD, Linux ignores high bits if set in the setaffinity(),
so clear it in the sched_setaffinity() and Linuxulator itself.
Reviewed by: Pau Amma (man pages)
In collaboration with: jhb
Differential revision: https://reviews.freebsd.org/D34849
MFC after: 2 weeks
projects / FreeBSD / FreeBSD.git / log