Commit Graph

268508 Commits

Author SHA1 Message Date
Emmanuel Vadot
cc2f9bbbad arm64: rockchip: rk3328: Finish implementing clocks
This finish (almost) the clocks implementations for the RK3328 SoC.
The clocks are now correctly implemented respecting the clock hiearchy.
The missing clocks are mostly the DDR clocks, implementing those is only
useful for debugging as we will never set them in the kernel.
The ARMCLK still needs to be rewritten so it looks closer to how the
hardware is done.

Tested-on: Rock64
2021-09-15 18:43:10 +02:00
Emmanuel Vadot
deff1fb3dc arm64: rockchip: clk: Add MUXRAW macros
Some clocks in the RK3328 SoC (and possibly others) have registers not in
the CLKSEL_CON range. Add a macros for muxes which lives not in the range
of CLKSEL_CON which just takes a raw offset.
2021-09-15 18:43:07 +02:00
Emmanuel Vadot
731e418bd7 arm64: rockchip: clk_mux: Add support for mux in GRF type clock
Some clocks have their mux register in the GRF and not in the CRU.
Add support for that in the rk_clk_mux clock type.
2021-09-15 18:43:04 +02:00
Hans Petter Selasky
e3e7d95332 tcp: Avoid division by zero when KERN_TLS is enabled in tcp_account_for_send().
If the "len" variable is non-zero, we can assume that the sum of
"tp->t_snd_rxt_bytes + tp->t_sndbytes" is also non-zero.

It is also assumed that the 64-bit byte counters will never wrap around.

Differential Revision:	https://reviews.freebsd.org/D31959
Reviewed by:	gallatin, rrs and tuexen
Found by:	"I told you so", also called hselasky
MFC after:	1 week
Sponsored by:	NVIDIA Networking
2021-09-15 18:05:31 +02:00
John Baldwin
857dc1c0ec arm64: Pass the right label to END() for handle_empty_exception.
GNU as reported an error for the argument to .size not being a constant.

Reviewed by:	imp, emaste
Differential Revision:	https://reviews.freebsd.org/D31950
2021-09-15 09:03:18 -07:00
John Baldwin
8753039a8f arm64: Fix a logic bug in is_load_instr().
Logical and ('&&') was used to join two conditions instead of logical
or ('||') causing some store instructions to not be recognized.

Reported by:	GCC 9 -Wparentheses

Reviewed by:	andrew
Differential Revision:	https://reviews.freebsd.org/D31949
2021-09-15 09:03:18 -07:00
John Baldwin
d99c87c8d5 evdev: Add parentheses around '-' expression in operand of '&'.
This fixes a -Wparentheses error with GCC 9.

Reviewed by:	wulf
Differential Revision:	https://reviews.freebsd.org/D31947
2021-09-15 09:03:18 -07:00
John Baldwin
cd16a848d1 posixshmtest: Fix various warnings raised by GCC.
- Remove unused format string arguments.

- Remove a set but unused variable.

Reviewed by:	khng, kib
Differential Revision:	https://reviews.freebsd.org/D31946
2021-09-15 09:03:18 -07:00
John Baldwin
df005aa9b3 pf: Remove duplicate declaration of pf_ioctl_maxcount.
Fixes a -Wredundant-decls warning with GCC 9.

Reviewed by:	kp
Differential Revision:	https://reviews.freebsd.org/D31944
2021-09-15 09:03:18 -07:00
John Baldwin
44126818d2 infiniband: Disable -Wredundant-decl warnings.
ib_uverbs_flow_resources_free() is declard in two header files in
upstream OFED.  Disable the warning to avoid introducing diffs to fix
the build on GCC 9.

While here, fix the ibcore module to disable the same warnings
disabled in OFED_CFLAGS.

Reviewed by:	hselasky
Differential Revision:	https://reviews.freebsd.org/D31943
2021-09-15 09:03:18 -07:00
John Baldwin
9553c6af88 <linux/overflow.h>: Don't use __has_builtin().
GCC only added support for __has_builtin in GCC 10.  However, all
supported versions of GCC and clang include these builtins so just use
them unconditionally.

This fixes the build with GCC 9.

Reviewed by:	manu, hselasky, imp
Differential Revision:	https://reviews.freebsd.org/D31942
2021-09-15 09:03:17 -07:00
John Baldwin
b14cd3a833 efitable: Don't pass NULL as a format string to xo_err().
This fixes a -Wformat error reported by GCC 9.

Reviewed by:	kib
Differential Revision:	https://reviews.freebsd.org/D31939
2021-09-15 09:03:17 -07:00
John Baldwin
4c9cb057bd top: Remove a duplicate extern declaration for show_args.
This fixes a -Wnested-extern error with GCC 9.  There is an existing
extern declaration in top.h.

Reviewed by:	imp, emaste
Differential Revision:	https://reviews.freebsd.org/D31937
2021-09-15 09:03:17 -07:00
John Baldwin
5ac4ac85ca Remove an always-true check.
This fixes a -Wtype-limits error from GCC 9.

Reviewed by:	imp
Differential Revision:	https://reviews.freebsd.org/D31936
2021-09-15 09:03:17 -07:00
John Baldwin
11647befcb Only define sanitizer wrappers for unsized bus space operations once.
Previously, this was defining duplicate definitions for each size.

This fixes a redundant definition warning from GCC.

Reviewed by:	markj
Differential Revision:	https://reviews.freebsd.org/D31966
2021-09-15 09:03:17 -07:00
John Baldwin
45cb767113 Only define sanitizer wrappers for atomic fences once.
Previously, this was defining duplicate definitions for each type.

This fixes a redundat definition warning from GCC.

Reviewed by:	markj
Differential Revision:	https://reviews.freebsd.org/D31965
2021-09-15 09:03:17 -07:00
John Baldwin
d2bc7754a2 Assert that invalid bus widths can't be passed to bus_width_str().
This appeases a -Wreturn-type warning from GCC.

Reviewed by:	imp
Differential Revision:	https://reviews.freebsd.org/D31935
2021-09-15 09:03:17 -07:00
John Baldwin
828f257277 Remove nonexistent include path for arm64 crypto files.
Reviewed by:	emaste
Differential Revision:	https://reviews.freebsd.org/D31932
2021-09-15 09:03:17 -07:00
Martin Matuska
53b70c86d9 zfs: merge openzfs/zfs@4a1195ca5 (master) into main
Notable upstream pull request merges:
  #11312 Temporarily use root credentials to mount snapshots in .zfs
  #12246 arc: Drop an incorrect assert
  #12443 Fixed data integrity issue when underlying disk returns error
         to zfs
  #12522 Compressed receive with different ashift can result in incorrect
         PSIZE on disk
  #12535 Verify embedded blkptr's in arc_read()
  #12541 Allow sending corrupt snapshots even if metadata is corrupted

Obtained from:	OpenZFS
OpenZFS commit:	4a1195ca50
2021-09-15 17:30:53 +02:00
Piotr Pawel Stefaniak
6e8272f317 mount: improve error message for invalid filesystem names
For an invalid filesystem name used like this:
mount -t asdfs /dev/ada1p5 /usr/obj

emit an error message like this:
mount: /dev/ada1p5: Invalid fstype: Invalid argument

instead of:
mount: /dev/ada1p5: Operation not supported by device

Differential Revision:	https://reviews.freebsd.org/D31540
2021-09-15 16:25:31 +02:00
Pawel Anikiel
2de4c7f6d0 pci_host_generic: Add Synopsys Designware PCIe controller quirk
Due to the quirky nature of the Synopsys Designware PCIe IP,
the type 0 configuration is broadcast and whatever device
is plugged into slot, will appear at each 32 device
positions of bus0. Mitigate the issue by filtering out
duplicated devices on this bus for both DT and ACPI cases.

Reviewed by: mw
Sponsored by: Semihalf
MFC: after 3 weeks
Differential revision: https://reviews.freebsd.org/D31887
2021-09-15 15:17:40 +02:00
Pietro Cerutti
41ba691f92 tee: use queue(9) in place of a custom linked list
Approved by: cognet
2021-09-15 11:04:24 +00:00
Andrew Turner
1896a00943 Use a 64 bit read to access GICR_TYPER
The GICv3 ITS only needs to implement 32 bit access to the GICR_TYPER
when the CPU implements AArch32. As this may not always be the case
use a 64 bit read when checking if the ITS is enabled on the CPU.

PR:		258217
Reported by:	Olivier Delande <olivier.delande@provenrun.com>
Sponsored by:	The FreeBSD Foundation
2021-09-15 09:13:41 +00:00
Michael Tuexen
4542164685 sctp: cleanup, no functional change intended
MFC after:	1 week
2021-09-15 10:18:11 +02:00
Alexander Motin
6c2d440416 ipmi(4): Limit maximum watchdog pre-timeout interval.
Previous code by default setting pre-timeout interval to 120 seconds
made impossible to set timeout interval below that, resulting in error
0xcc (Invalid data field in Request) at least on Supermicro boards.

To fix that limit maximum pre-timeout interval to ~1/4 of the timeout
interval, that sounds like a reasonable default: not too short to fire
too late, but also not too long to give many false reports.

MFC after:	2 weeks
2021-09-14 21:06:39 -04:00
Allan Jude
4a1195ca50
Temporarily use root credentials to mount snapshots in .zfs
When mounting a snapshot in the .zfs/snapshots control directory,
temporarily assume roots credentials to perform the VFS_MOUNT().

This allows regular users and users inside jails to access these
snapshots.

The regular usermount code is not helpful here, since it requires
that the user performing the mount own the mountpoint, which won't
be the case for .zfs/snapshot/<snapname>

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tony Nguyen <tony.nguyen@delphix.com>
Signed-off-by: Allan Jude <allan@klarasystems.com>
Sponsored-By: Modirum MDPay
Sponsored-By: Klara Inc.
Closes #11312
2021-09-14 17:10:00 -06:00
Alexander Motin
02d8194012 mps/mpr(4): Move xpt_register_async() out of lock.
It fixes lock ordere reversal between SIM and device locks.  Also
remove registration for AC_FOUND_DEVICE, unused for a while now.

MFC after:	1 month
2021-09-14 17:40:32 -04:00
Alan Somers
5dc5f849be tools/test/upsdl: fix compiler warnings
MFC after:	2 weeks
Sponsored by:	Axcient
2021-09-14 14:50:01 -06:00
John Baldwin
9affbb0f52 cxgbe tom: Enter network epoch in t4_aiotx_task().
While here, don't restore the old vnet until after sorele().

Sponsored by:	Chelsio Communications
2021-09-14 13:46:15 -07:00
John Baldwin
5dbf8c1588 cxgbe tom: Update rcv_nxt for a FIN after handle_ddp_close().
For TCP DDP, handle_ddp_close() needs to see the pre-FIN rcv_nxt to
determine how much data was placed in the local buffer before the FIN
was received.  The changes in d59f1c49e2 broke this by updating
rcv_nxt before calling handle_ddp_close().

Fixes:		d59f1c49e2 cxgbe tom: Permit rcv_nxt mismatches on FIN for iSCSI connections on T6.
Sponsored by:	Chelsio Communications
2021-09-14 13:46:14 -07:00
John Baldwin
1ecbc1d8e9 cxgbe tom: Don't queue AIO requests on listen sockets.
This is similar to the fixes in 141fe2dcee.  One difference is that
TOE sockets do not change states (listen vs non-listen) once created,
so no lock is needed for SOLISTENING().

Sponsored by:	Chelsio Communications
2021-09-14 13:46:14 -07:00
Edward Tomasz Napierala
bdf0f24bb1 linux: implement PTRACE_GET_SYSCALL_INFO
This is one of the pieces required to make modern (ie Focal)
strace(1) work.

Reviewed By:	jhb (earlier version)
Sponsored by:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D28212
2021-09-14 20:19:55 +00:00
Edward Tomasz Napierala
6f43f86bf3 tzcode: Fix operation without WITH_DETECT_TZ_CHANGES
Reviewed By:	bdrewery, kevans, cy
Reported By:	lwhsu, bdrewery
Fixes:		ddedf2a11e
Sponsored by:	NetApp, Inc.
Sponsored by:	Klara, Inc.
Differential Revision:	https://reviews.freebsd.org/D31961
2021-09-14 20:17:50 +00:00
Ka Ho Ng
49050613ef ctl(4): Do hole-punching for UNMAP to file-backed LUNs
This adds support for SCSI UNMAP command to file-backed LUNs, if the
underlying file system has a non-zerofilling VOP_DEALLOCATE
implementation where some or all parts of the requested operation range
may be deallocated.

Sponsored by:	The FreeBSD Foundation
Reviewed by:	mav
Differential Revision:	https://reviews.freebsd.org/D31922
2021-09-15 03:51:58 +08:00
Ed Maste
8cba2003e8 Cirrus-CI: add a manually triggered arm64 task
Reviewed by:	imp
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D31953
2021-09-14 15:12:55 -04:00
Andrew Turner
90f33f34ba Remove a bogus assertion from the gic drivers
When setting a message based interrupt range we set the base and count.
In an earlier the count was implemented as an end value, however the
asserts used to check this value was correct were incorrectly left in.

Reported by:	emaste
Sponsored by:	The FreeBSD Foundation
2021-09-14 19:06:25 +00:00
John Baldwin
c782ea8bb5 Add a switch structure for send tags.
Move the type and function pointers for operations on existing send
tags (modify, query, next, free) out of 'struct ifnet' and into a new
'struct if_snd_tag_sw'.  A pointer to this structure is added to the
generic part of send tags and is initialized by m_snd_tag_init()
(which now accepts a switch structure as a new argument in place of
the type).

Previously, device driver ifnet methods switched on the type to call
type-specific functions.  Now, those type-specific functions are saved
in the switch structure and invoked directly.  In addition, this more
gracefully permits multiple implementations of the same tag within a
driver.  In particular, NIC TLS for future Chelsio adapters will use a
different implementation than the existing NIC TLS support for T6
adapters.

Reviewed by:	gallatin, hselasky, kib (older version)
Sponsored by:	Chelsio Communications
Differential Revision:	https://reviews.freebsd.org/D31572
2021-09-14 11:43:41 -07:00
Mark Johnston
cf4670fe0b kcov: Integrate with KMSAN
- kern_kcov.c needs to be compiled with -fsanitize=kernel-memory when
  KMSAN is configured since it calls into various other subsystems.
- Disable address and memory sanitizers in kcov(4)'s coverage sanitizer
  callbacks, as they do not provide useful checking.  Moreover, with
  KMSAN we may otherwise get false positives since the caller (coverage
  sanitizer runtime) is not instrumented.
- Disable KASAN and KMSAN interceptors in subr_coverage.c, as they do
  not provide any benefit but do introduce overhead when fuzzing.

Sponsored by:	The FreeBSD Foundation
2021-09-14 14:29:27 -04:00
Alexander Motin
272c4a4dc5 Allow setting NFS server scope and owner.
By default NFS server reports as scope and owner major the host UUID
value and zero for owner minor.  It works good in case of standalone
server.  But in case of CARP-based HA cluster failover the values
should remain persistent, otherwise some clients like VMware ESXi
get confused by the change and fail to reconnect automatically.

The patch makes server scope, major owner and minor owner values
configurable via sysctls.  If not set (by default) the host UUID
value is still used.

Reviewed by:	rmacklem
MFC after:	2 weeks
Differential Revision:	https://reviews.freebsd.org/D31952
2021-09-14 14:18:03 -04:00
Ed Maste
0f9bafdfc3 openssh: pass ssh context to BLACKLIST_NOTIFY
Fixes:		19261079b7 ("openssh: update to OpenSSH v8.7p1")
Sponsored by:	The FreeBSD Foundation
2021-09-14 13:44:39 -04:00
Brian Behlendorf
6954c22f35
Use fallthrough macro
As of the Linux 5.9 kernel a fallthrough macro has been added which
should be used to anotate all intentional fallthrough paths.  Once
all of the kernel code paths have been updated to use fallthrough
the -Wimplicit-fallthrough option will because the default.  To
avoid warnings in the OpenZFS code base when this happens apply
the fallthrough macro.

Additional reading: https://lwn.net/Articles/794944/

Reviewed-by: Tony Nguyen <tony.nguyen@delphix.com>
Reviewed-by: George Melikov <mail@gmelikov.ru>
Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov>
Closes #12441
2021-09-14 10:17:54 -06:00
Mark Johnston
b1746faad6 debugnet: Include some required headers
Don't depend on pollution from net/vnet.h.

PR:		258496
MFC after:	1 week
Sponsored by:	The FreeBSD Foundation
2021-09-14 11:02:45 -04:00
Mark Johnston
e6c19aa94d sctp: Allow blocking on I/O locks even with non-blocking sockets
There are two flags to request a non-blocking receive on a socket:
MSG_NBIO and MSG_DONTWAIT.  They are handled a bit differently in that
soreceive_generic() and soreceive_stream() will block on the socket I/O
lock when MSG_NBIO is set, but not if MSG_DONTWAIT is set.  In general,
MSG_NBIO seems to mean, "don't block if there is no data to receive" and
MSG_DONTWAIT means "don't go to sleep for any reason".

SCTP's soreceive implementation did not allow blocking on the I/O lock
if either flag is set, but this violates an assumption in
aio_process_sb(), which specifies MSG_NBIO but nonetheless
expects to make progress if data is available to read.  Change
sctp_sorecvmsg() to block on the I/O lock only if MSG_DONTWAIT
is not set.

Reported by:	syzbot+c7d22dbbb9aef509421d@syzkaller.appspotmail.com
Reviewed by:	tuexen
MFC after:	1 week
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D31915
2021-09-14 09:02:05 -04:00
Mark Johnston
fa0463c384 socket: De-duplicate SBLOCKWAIT() definitions
MFC after:	1 week
Sponsored by:	The FreeBSD Foundation
2021-09-14 09:01:32 -04:00
Andrew Turner
fd860ace3b Remove the DN flag from the initial arm64 fpcr value
This fixes software that expects NaN values to propagate.

Sponsored by:	The FreeBSD Foundation
2021-09-14 12:52:48 +00:00
Andrew Turner
e4d89a633e Add support for gicv2m as a child of gicv3
On some systems, e.g. Parallels set to host a Linux VM under an M1 Mac,
there is a GICv2m as a child of the GICv3. We previously assumed the
GICv2m was always a child of a GICv2. Fix this by adding the needed
support to the GICv3 driver.

PR:		258136
Reported by:	trasz
Sponsored by:	The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D31768
2021-09-14 08:24:52 +01:00
Andrew Turner
c6f3076d44 Move the GICv2m msi handling to the parent
This is in preperation for adding support for the GICv2m driver as a
child of the GICv3 driver.

PR:             258136
Reported by:    trasz
Sponsored by:	The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D31767
2021-09-14 08:24:52 +01:00
Wojciech Macek
ba4d9d9d5b Revert "if_mvneta: Build the driver as a kernel module"
This reverts commit bcf5c7a8b1.
2021-09-14 11:49:59 +02:00
Peter Holm
94c678cf19 stress2: A two second timeout is too short 2021-09-14 07:50:26 +00:00
Hubert Mazur
bcf5c7a8b1 if_mvneta: Build the driver as a kernel module
Fix device detach and attach routine. Add required Makefile
to build as a module. Remove entry from GENERIC, since now
it can be loaded automatically.

Tested on EspressoBin.

Obtained from:		Semihalf
Reviewed by:		manu
Differential revision:	https://reviews.freebsd.org/D31581
2021-09-14 08:29:53 +02:00