1519 Commits

Author SHA1 Message Date
Alan Cox
e41fde3ed7 On a failed fcmpset don't pointlessly repeat tests
In a few places, on a failed compare-and-set, both the amd64 pmap and
the arm64 pmap repeat tests on bits that won't change state while the
pmap is locked.  Eliminate some of these unnecessary tests.

Reviewed by:	andrew, kib, markj
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D31014
2021-07-05 21:07:40 -05:00
Edward Tomasz Napierala
45d99014ca linux(4): implement coredumps on arm64
Previously they only worked on amd64.

Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30975
2021-07-03 08:06:31 +01:00
Emmanuel Vadot
993e8236c3 arm64: allwinner: Add r_intc driver
The r intc interrupt controller seems to do a lot of things :
- It can handle the NMI interrupt
- It have local interrupts for some device that also can be muxed with GIC
- It can serve as an forwarder for the GIC

It's mostly used for deepsleep/wakeup if I understood correctly and we do not
support this on arm64.

For now just forward everything to the GIC so interrupts works again for device
which now have this interrupts controller set since dts v5.12

Sponsored by:	Diablotin Systems
2021-07-01 18:46:38 +02:00
Andrew Turner
244002b482 Switch the order of the ID_AA64PFR1_EL1 fields
This makes them consistent with the fields in other registers.

Sponsored by:	The FreeBSD Foundation
2021-07-01 00:48:56 +00:00
Edward Tomasz Napierala
93c3453f11 linux(4): revert arm64 part of 447636e43c0
The arm64 part of the patch was incomplete and prevented
linux64.ko from loading due to missing symbol.

Sponsored By:	EPSRC
2021-07-01 08:29:23 +00:00
Edward Tomasz Napierala
447636e43c linux(4): implement coredump support
Implement dumping core for Linux binaries on amd64, for both
32- and 64-bit executables.  Some bits are still missing.

This is based on a prototype by chuck@.

Reviewed By:	kib
Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30019
2021-06-30 22:45:06 +01:00
Mitchell Horne
8cc3815f02 hwpmc_arm64: accept raw event codes for PMC_OP_PMCALLOCATE
Make it possible to specify event codes without an offset of
PMC_EV_ARMV8_FIRST, by setting a machine-dependent flag. This is
required to make use of event definitions from pmu-events.

Reviewed by:	ray (slightly earlier version)
MFC after:	2 weeks
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D30602
2021-06-30 16:47:09 -03:00
Olivier Houchard
712c060c94 arm64: Make sure COMPAT_FREEBSD32 handles thumb entry point.
If the entry point for the binary executed is a thumb 2 entry point, make
sure we set the PSR_T bit, or the CPU will interpret it as arm32 code and
bad things will happen.

PR: 256899
MFC after: 1 week
2021-06-30 14:55:18 +02:00
Edward Tomasz Napierala
435754a59e Add infrastructure required for Linux coredump support
This adds `sv_elf_core_osabi`, `sv_elf_core_abi_vendor`,
and `sv_elf_core_prepare_notes` fields to `struct sysentvec`,
and modifies imgact_elf.c to make use of them instead
of hardcoding FreeBSD-specific values.  It also updates all
of the ABI definitions to preserve current behaviour.

This makes it possible to implement non-native ELF coredump
support without unnecessary code duplication.  It will be used
for Linux coredumps.

Reviewed By:	kib
Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30921
2021-06-29 08:49:12 +01:00
Alan Cox
26a357245f arm64: a few simplications to pmap_remove_{all,write}
Eliminate some unnecessary unlocking and relocking when we have to retry
the operation to avoid deadlock.  (All of the other pmap functions that
iterate over a PV list already implemented retries without these same
unlocking and relocking operations.)

Avoid a pointer dereference by using an existing local variable that
already holds the desired value.

Eliminate some unnecessary repetition of code on a failed fcmpset.
Specifically, there is no point in retesting the DBM bit because it
cannot change state while the pmap lock is held.

Reviewed by:	kib, markj
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D30931
2021-06-28 22:21:24 -05:00
Alan Cox
19c288b3a6 arm64: eliminate a duplicated #define 2021-06-27 01:44:58 -05:00
Mike Karels
184291b0a5 genet: pullup minimum header amount for IPv4
The genet driver (RPi4 Ethernet) had code to pull headers into the
first mbuf if there was only an Ethernet header there.  This was
originally needed for ICMPv6 replies, then for forwarded IPv6/TCP.
Now a situation has been found where it is needed for IPv4, when
using NAT with IPFW.  Generalize to do this for all protocols.
Rather than using an IPv6-related definition for the length, move
the length to a variable that can be set with sysctl
(hw.genet.tx_hdr_min).  Move an old tunable to a new RDTUN variable
with a better name.

PR:		25607
MFC after:	3 days
Reviewers:	emaste
Differential Revision: https://reviews.freebsd.org/D30831
2021-06-26 11:04:02 -05:00
Alan Cox
5dd84e315a arm64: fix a potential KVA leak in pmap_demote_l1()
In the unlikely event that the 1 GB page mapping being demoted is used
to access the L1 page table page containing the 1 GB page mapping and
the vm_page_alloc() to allocate a new L2 page table page fails, we
would leak a page of kernel virtual address space.  Fix this leak.

MFC after:	1 week
2021-06-25 23:01:32 -05:00
Alan Cox
c94249decd arm64: make it possible to define PV_STATS
Remove an #if 0 that results in a compilation error if PV_STATS is
defined.  Aside from this #if 0, there is nothing wrong with the
PV_STATS code.

MFC after:	1 week
2021-06-24 18:32:56 -05:00
Marcin Wojtas
19aa95e4b3 Introduce new driver for NXP Ethernet controller
ENETC it a gigabit Ethernet controller found on the LS1028A board.
It supports basic VLAN offloads - tag extraction, injection and hardware
filtering. Inband MDIO connectivity is used for link status
monitoring through the miibus interface. Fixed-link mode is also
supported, which allows for operation of internal cpu to switch port.
Since no admin interrupts are present in hardware, link status polling
has to be used.
Due to a hardware bug software reset of the NIC results in a external
abort. Because of that most of the hardware initialization is done
during attach. This also means that in the case of an fatal error full
board reset is required.
The enetc_hw.h header was imporoted from Linux. It is dual licensed.

Submitted by: Kornel Duleba <mindal@semihalf.com>
Obtained from: Semihalf
Sponsored by: Alstom Group
Differential Revision: https://reviews.freebsd.org/D30729
2021-06-24 13:01:13 +02:00
Colin Percival
ccadcb8f33 arm64: Instrument initarm with TSLOG
The "function entry" needs to be recorded with TSRAW() rather than
TSENTER() since curthread is not yet defined.  (The same approach
is taken in amd64's hammer_time.)
2021-06-23 22:17:55 -07:00
Colin Percival
0acc529dc3 arm64: Disable early printf if TSLOG
The warning message "ERROR loading DTB" (for systems without a device
tree blob) is printed extremely early in the boot process -- among
other things, before curthread or other pcpu data has been set up.
Unfortunately, printf is instrumented with TSLOG, which cannot run
quite this early.

Wrap the printf in #ifndef TSLOG; the situations where the printf
will be useful are not ones where TSLOG would be in use.
2021-06-23 22:10:26 -07:00
Alan Cox
0c188c06c6 arm64: replace pa_to_pvh() with page_to_pvh() in pmap_remove_l2()
Revise pmap_remove_l2() to use the constant-time function page_to_pvh()
instead of the linear-time function pa_to_pvh().

Reviewed by:	kib, markj
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D30876
2021-06-23 22:35:46 -05:00
Alan Cox
62ea198e95 arm64: remove an unneeded test from pmap_clear_modify()
The page table entry for a 4KB page mapping must be valid if a PV entry
for the mapping exists, so there is no point in testing each page table
entry's validity when iterating over a PV list.

Reviewed by:	kib, markj
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D30875
2021-06-23 14:22:46 -05:00
Alan Cox
6f6a166eaf arm64: Use page_to_pvh() when the vm_page_t is known
When support for a sparse pv_table was added, the implementation of
pa_to_pvh() changed from a simple constant-time calculation to iterating
over the array vm_phys_segs[].  To mitigate this issue, an alternative
function, page_to_pvh(), was introduced that still runs in constant time
but requires the vm_page_t to be known.  However, three cases where the
vm_page_t is known were not converted to page_to_pvh().  This change
converts those three cases.

Reviewed by:	kib, markj
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D30832
2021-06-21 17:25:06 -05:00
Dmitry Chagin
c1da89fec2 linux(4): Retire linux_kplatform.
Assuming we can't run on i486, i586 class cpu, retire linux_kplatform var
and use hardcoded 'machine' value in linux_newuname().

I have added linux_kplatform for consistency with linux_platform which is
placed in to vdso to avoid excess copyout it on stack for AT_PLATFORM at
exec time.

This is the first stage of Linuxulator's vdso revision.

Reviewed by:		trasz, imp
Differential Revision:	https://reviews.freebsd.org/D30774
MFC after:		2 weeks
2021-06-22 08:36:21 +03:00
Mike Karels
13604fb0fd genet: Fix potential crash during attach
As pointed out in the bug, the genet driver (RPi4 Ethernet) was
attaching the interrupts before the data structures were fully
initialized, causing a crash if an interrupt came in during the
attach.  Fix by reordering code blocks.

PR:		256334
Reported by:	< ghuckriede at blackberry.com >
Reviewed by:	< ghuckriede at blackberry.com > (informally)
MFC after:	3 days
2021-06-20 11:17:13 -05:00
Edward Tomasz Napierala
a397b55083 linux: implement set_cloned_tls() on arm64
This fixes random segfaults on "git commit --amend".

Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30508
2021-06-16 10:32:39 +00:00
Edward Tomasz Napierala
e505c3066b linux: implement set_upcall on aarch64
This fixes "git diff" (git-1:2.25.1-1ubuntu3).

Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30509
2021-06-16 05:13:17 +00:00
Konstantin Belousov
870e197d52 Add quirks for Linux ABI signals handling
Require queueing of the signals with default action, and disable
dequeueing SIGCHLD on wait for live process.

Reported and tested by:	dchagin
Reviewed by:	dchagin, markj
Sponsored by:	The FreeBSD Foundation
MFC after:	1 week
Differential revision:	https://reviews.freebsd.org/D30675
2021-06-16 02:01:35 +03:00
Alex Richardson
9bb8a4091c Reduce code duplication in machine/_types.h
Many of these typedefs are the same across all architectures or can
be set based on an architecture-independent compiler-provided macro
(e.g. __SIZEOF_SIZE_T__). These macros have been available since GCC 4.6
and Clang sometime before 3.0 (godbolt.org does not have any older clang
versions installed).

I originally considered using the compiler-provided `__FOO_TYPE__` directly.
However, in order to do so we have to check that those match the previous
typedef exactly (not just that they have the same size) since any change
would be an ABI break. For example, changing `long` to `long long` results
in different C++ name mangling. Additionally, Clang and GCC disagree on
the underlying type for some of (u)int*_fast_t types, so this change
only moves the definitions that are identical across all architectures
and does not touch those types.

This de-deduplication will allow us to have a smaller diff downstream in
CheriBSD: we only have to only change the (u)intptr_t definition in
sys/_types.h in CheriBSD instead of having to change machine/_types.h for
all CHERI-enabled architectures (currently RISC-V, AArch64 and MIPS).

Reviewed By: imp, kib
Differential Revision: https://reviews.freebsd.org/D29895
2021-06-14 16:30:16 +01:00
Andrew Turner
15fa52a564 Add more GIC and GICv3 registers
These aren't used by either driver, however they will be needed by
bhyve on arm64 to emulate a GICv3 interrupt controller.

Sponsored by:	Innovate UK
2021-06-14 14:53:11 +00:00
Andrew Turner
e7f369b1c1 Use the correct length when copying arm64 vfp registers
We passed the wrong length into memcpy in the arm64 get_fpcontext and
set_fpcontext. This caused us to copy two status registers we didn't
expect to copy.

These are safe as they exist in both the source and destination, although
in a different order, and we copy the correct values after the memcpy.

Obtained from:	CheriBSD
Sponsored by:	Innovate UK
2021-06-14 14:53:10 +00:00
Oleksandr Tymoshenko
cb25a94558 arm64: allwinner: Add i2s and codec support
Differential Revision:	https://reviews.freebsd.org/D27830
2021-06-11 21:06:08 +02:00
Dmitry Chagin
ee64d98204 linux(4): Regen for futex system call.
MFC after:	2 weeks
2021-06-10 14:16:40 +03:00
Dmitry Chagin
3c1de151e3 linux(4): Change Linux futex syscall definition to match Linux actual one.
MFC after:	2 weeks
2021-06-10 14:00:00 +03:00
Andriy Gapon
ffc5dc788f rk3328_cru: fix a typo in the SCLK_I2S2 gate definition
MFC after:	4 days
2021-06-09 12:41:52 +03:00
Marcin Wojtas
f0f7b0868a Remove ThunderX PCIe FDT quirks from pci_host_generic_fdt.c
ThunderX is the only board known to use them.
Move them to the ThunderX PCIe driver.

Submitted by: Kornel Duleba <mindal@semihalf.com>
Reviewed by: andrew
Obtained from: Semihalf
Sponsored by: Alstom Group
Differential Revision: https://reviews.freebsd.org/D30179
2021-06-08 17:51:11 +02:00
Mark Johnston
4e4035ef1f arm64: Fix pmap_copy()'s handling of 2MB mappings
When copying mappings from parent to child, we clear the accessed and
dirty bits.  This is done for both 4KB and 2MB PTEs.  However,
pmap_demote_l2() asserts that writable superpages must be dirty.  This
is to avoid races with the MMU setting the dirty bit during promotion
and demotion.  pmap_copy() can create clean, writable superpage
mappings, so it violates this assertion.

Modify pmap_copy() to preserve the accessed and dirty bits when copying
2MB mappings, like we do on amd64.

Fixes:		ca2cae0b4dd
Reported by:	Jenkins via mhorne
Reviewed by:	alc, kib
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D30643
2021-06-06 16:44:46 -04:00
Mark Johnston
a48f51b3d3 arm64: Use the right PTE when downgrading perms in pmap_promote_l2()
When promoting a run of small mappings to a superpage, we have to
downgrade clean, writable mappings to read-only, to handle the
possibility that the MMU will concurrently mark one of the mappings as
dirty.

The code which performed this operation for the first PTE in the run
used the wrong PTE pointer.  As a result, the comparison would always
fail, aborting the promotion.  This only occurs when promoting writable,
clean mappings.

Fixes:		ca2cae0b4dd
Reviewed by:	alc, kib
MFC after:	1 week
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D30642
2021-06-06 16:44:46 -04:00
Konstantin Belousov
598f6fb49c linuxolator: Add compat.linux.setid_allowed knob
PR:	21463
Reported by:	kris
Reviewed by:	dchagin
Tested by:	trasz
Sponsored by:	The FreeBSD Foundation
MFC after:	1 week
Differential Revision:	https://reviews.freebsd.org/D28154
2021-06-06 21:43:00 +03:00
Dmitry Chagin
e29ea22f70 Regen for ('0f8dab45404f347752470579feccc6d2739b9570') Linux
rt_sigtimedwait system call.

MFC after:	2 weeks
2021-06-07 05:39:29 +03:00
Dmitry Chagin
0f8dab4540 linux(4): Fix timeout parameter of rt_sigtimedwait syscall, which is
timespec not a timeval.

MFC after:	2 weeks
2021-06-07 05:35:35 +03:00
Bjoern A. Zeeb
002cbc89c3 arm64/rk805: remove RTC Set logging
When ntpd is synchronizing the system time, it also periodically (30m)
syncs the the RTC time.  Remove printf in rk805_settime which triggers
every 30m, as settime_task_func() will log errors under bootverbose.
We leave the RTC Read logging, which should happen only once at boot.

Commit message by:	imp
Reviewed by:		manu, imp
MFC after:		2 weeks
Differential Revision: https://reviews.freebsd.org/D30361
2021-06-05 17:07:56 +00:00
Bjoern A. Zeeb
57af163c8e arm64/rk_pcie_phy: handle assigned-clock*
Nanopi4 based SoCs (NanoPC-T4, NanoPi M4*, and NanoPi Neo4) have
assigned-clock* in the pcie_phy node.  Handle them but only fail
in case clk_set_assigned() returns an error other than
"no assigned-clock*" (as it would for all other SoCs).

Reviewed by:	manu
MFC After:	2 weeks
Differential Revision: https://reviews.freebsd.org/D30363
2021-06-05 16:16:29 +00:00
Edward Tomasz Napierala
b501b2ae52 linux: export AT_HWCAP and AT_HWCAP2 on aarch64
The flag values seem to be the same between Linux and FreeBSD.
Comparing to a Linux VM on the same hardware, we're missing
HWCAP_EVTSTRM, HWCAP_CPUID, HWCAP_DCPOP, HWCAP_USCAT, HWCAP_PACA,
and HWCAP_PACG.

Reviewed By:	mhorne, emaste
Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30540
2021-06-01 13:50:20 +01:00
Dmitry Chagin
19593f775c linux(4); Retire unnecessary __packed attribute from some struct's
definition.

Differential Revision:	https://reviews.freebsd.org/D30482
MFC after:		2 weeks
2021-05-31 21:56:34 +03:00
Edward Tomasz Napierala
83043a741d linux: deduplicate DUMMY() entries
No functional changes.

Reviewed By:	emaste
Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30524
2021-05-29 17:51:36 +00:00
Edward Tomasz Napierala
aa462cab60 linux: fix architecture returned for uname on aarch64
Previously it would return "arm64", which was breaking build
for Linux kernel.  While here, reshuffle entries in the auxv
vector to match real Linux.

Reviewed By:	emaste
Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30500
2021-05-28 23:02:02 +00:00
Edward Tomasz Napierala
023ee5392c linux: make linux64.ko loadable again on aarch64
This should probably get deduplicated; this is a quick fix
just to unbreak things.

Fixes:		6d926e850d2
Sponsored By:	EPSRC
2021-05-28 14:20:30 +00:00
Ed Maste
d097687b31 arm64: correct END() copy-pasteo for generic_bs_fault
PR:		256197
Reported by:	Nick Reilly
MFC after:	1 week
Fixes:		95a85c125d0e67 ("Add NetBSD compatible bus_space_peek_N()...")
Sponsored by:	The FreeBSD Foundation
2021-05-28 09:10:11 -04:00
Edward Tomasz Napierala
c0f171736a Regen after 6d926e850d2.
Sponsored By:	EPSRC
2021-05-28 09:04:17 +01:00
Edward Tomasz Napierala
6d926e850d linux: add new syscall numbers
Sponsored By:	EPSRC
Differential Revision:	https://reviews.freebsd.org/D30193
2021-05-28 09:02:16 +01:00
Ruslan Bukin
38e7025a60 o Fix i2c read operation for large transfers (more than 32 bytes).
o Fix slave address setting.

This allows to read the EDID from an HDMI monitor.

Reviewed by:	manu
Sponsored by:	UKRI
Differential Revision: https://reviews.freebsd.org/D27139
2021-05-26 10:55:23 +01:00
Mitchell Horne
6f4bb8ecc2 arm64, riscv: remove reference to fsu_intr_fault
This variable no longer exists.

MFC after:	3 days
2021-05-25 12:26:52 -03:00