The driver now ensures only one thread at a time is running in the API
functions (clock_gettime() and clock_settime()) by specifically requesting
ownership of the i2c bus without using IIC_RECURSIVE, then it does all IO
using IIC_RECURSIVE so that each individual IO operation doesn't try to
re-acquire the bus.
The other IO done by the driver happens at attach or intr_config_hooks time,
when there can't be multiple threads running with the same device instance.
So, the IIC_RECURSIVE flag can be safely ORed into the wait flags for all IO
done by the driver, because it's all either done in a single-threaded
environment, or protected within a block bounded by explict
iicbus_acquire_bus() and iicbus_release_bus() calls.
The driver now ensures only one thread at a time is running in the API
functions (clock_gettime() and clock_settime()) by specifically requesting
ownership of the i2c bus without using IIC_RECURSIVE, then it does all IO
using IIC_RECURSIVE so that each individual IO operation doesn't try to
re-acquire the bus.
The other IO done by the driver happens at attach or intr_config_hooks time,
when there can't be multiple threads running with the same device instance.
So, the IIC_RECURSIVE flag can be safely ORed into the wait flags for all IO
done by the driver, because it's all either done in a single-threaded
environment, or protected within a block bounded by explict
iicbus_acquire_bus() and iicbus_release_bus() calls.
The recursive ownership support added in r321584 was unconditionally in
effect all the time -- whenever a given i2c slave device instance tried to
lock the i2c bus for exclusive use when it already owned the bus, the call
returned immediately without waiting. However, many i2c slave drivers use
bus ownership to enforce that only a single thread at a time can be using
the slave device. The recursive locking changes broke this use case.
Now there is a new flag, IIC_RECURSIVE, which can be mixed in with the
other flags passed to iicbus_acquire_bus() to allow drivers to indicate
when recursive locking is desired. Using the flag implies that the driver
is managing concurrent access to the device by different threads in some way.
This immediately fixes all existing i2c slave drivers except for the two
i2c RTC drivers which use the recursive locking feature; those will be
fixed in a followup commit.
In iflib, the device-specific init() function isn't supposed to edit
the struct ifnet driver flags. If it does, it'll cause an MPASS() assert
in iflib to fail.
PR: 225312
Reported by: bhughes@
These functions deal the same type of overflows we do with mallocarray(9).
Using our mallocarray will panic, which different from the previous
behavior (returning NULL), but neither behavior is more correct.
As a sidenote, drm_calloc_large() is not currently used at all.
Reviewed by: dumbbell
Differential Revision: https://reviews.freebsd.org/D13835
When allocating memory through malloc(9), we always expect the amount of
memory requested to be unsigned as a negative value would either stand for
an error or an overflow.
Unsign some values, found when considering the use of mallocarray(9), to
avoid unnecessary casting. Also consider that indexes should be of
at least the same size/type as the upper limit they pretend to index.
MFC after: 3 weeks
Uses of mallocarray(9).
The use of mallocarray(9) has rocketed the required swap to build FreeBSD.
This is likely caused by the allocation size attributes which put extra pressure
on the compiler.
Given that most of these checks are superfluous we have to choose better
where to use mallocarray(9). We still have more uses of mallocarray(9) but
hopefully this is enough to bring swap usage to a reasonable level.
Reported by: wosch
PR: 225197
Similarly as other extres pseudo-drivers, implement phy by using kobj model.
This detaches it from provider device, so single device driver can export
multiple different phys. Additionally, this allows phy to be subclassed to
more specialized drivers, like is USB OTG phy, or PCIe phy with hot-plug
capability.
Tested by: manu (previous version, on Allwinner board)
MFC after: 1 month
During set_freq a clknode might have reparent (using a better parent that
have a higher frequency for example), before refreshing the cache, re-get
the parent frequency.
Reviewed by: mmel
possible to change string and numeric vendor and product identifiers,
as well as anything else there might be to change for a particular
device side template, eg the MAC address.
Reviewed by: hselasky@
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D13920
This fixes a panic when `EVDEV_SUPPORT` is enabled: if a trackpoint
packet was detected but there was no trackpoint, we still tried to emit an
evdev event even though the associated relative evdev device (`evdev_r`)
was not initialized.
PR: 225339
MFC after: 1 week
In psmprobe(), we set the initial `syncmask` to the vendor default value
if the `PSM_CONFIG_NOCHECKSYNC` bit is unset. However, we currently only
set it for the Elantech touchpad later in psmattach(), thus `syncmask`
is always configured.
Now, we check `PSM_CONFIG_NOCHECKSYNC` and skip sync check if it is set.
This fixes Elantech touchpad support for units which have `hascrc` set.
To clarify that, when we log the `syncmask` and `syncbits` fields, also
mention if they are actually used.
Finally, when we set `PSM_CONFIG_NOCHECKSYNC`, clear `PSM_NEED_SYNCBITS`
flag.
PR: 225338
MFC after: 1 week
any children prior to detach.
With the newbus child deletion ordering changes introduced in r307518,
parent devices are now detached (and their driver set to NULL) prior to
detaching and deleting child devices; child-related bus methods (e.g.
BUS_CHILD_DETACHED, BUS_CHILD_DELETED) are no longer be dispatched to the
parent device driver after it returns 0 (success) from DEVICE_DETACH.
Sponsored by: The FreeBSD Foundation
addressing. The host addressing constraint does not apply to device address
space, and shouldn't be passed to bhnd_get_dma_translation() as the
maximum supported device address width.
Sponsored by: The FreeBSD Foundation
- Extend the probe method to accept devclasses that inherit from the pci
devclass (e.g. cardbus).
- Some BCM4306-based CardBus adapters appear to advertise 4K SPROM, but
only the first 2K is mapped into BAR0. We can safely assume that the
SPROM data fits within the first 2K of the SPROM, rather than rejecting
the SPROM mapping as invalid.
Sponsored by: The FreeBSD Foundation
- Do not panic on siba(4) detach when the bhnd(4) bus calls
bhnd_get_pmu_info() on a PMU-less device.
- Fix bhnd_pwrctl attach/detach on fixed-clock devices:
- Treat bhnd_pwrctl_updateclk() as a no-op on fixed-clock devices.
- Use bhnd_pwrctl_updateclk() to perform the appropriate clock
transition on detach.
Sponsored by: The FreeBSD Foundation
kernel by PHYS_TO_DMAP() as previously present on amd64, arm64, riscv, and
powerpc64. This introduces a new MI macro (PMAP_HAS_DMAP) that can be
evaluated at runtime to determine if the architecture has a direct map;
if it does not (or does) unconditionally and PMAP_HAS_DMAP is either 0 or
1, the compiler can remove the conditional logic.
As part of this, implement PHYS_TO_DMAP() on sparc64 and mips64, which had
similar things but spelled differently. 32-bit MIPS has a partial direct-map
that maps poorly to this concept and is unchanged.
Reviewed by: kib
Suggestions from: marius, alc, kib
Runtime tested on: amd64, powerpc64, powerpc, mips64
r326454.
bwn(4)/bhnd(4) has been tested with most chipsets currently supported by
bwn(4), and this change should be transparent to existing bwn(4) users;
please report any regressions that you do encounter.
To revert to using siba_bwn(4) instead of bhnd(4), place the following
lines in loader.conf(5):
hw.bwn_pci.preferred="0"
Once we're satisfied that the switch to bhnd(4) has seen sufficient broader
testing, bwn(4) will be migrated to use the native bhnd(9) interface
directly, and support for siba_bwn(4) will be dropped (see D13518).
Sponsored by: The FreeBSD Foundation
pmcstat request for close will generate a close event.
This event will be in turn received by pmcstat to close the file.
Reviewed by: kib
Tested by: pho
MFC after: 1 week
Sponsored by: Stormshield
The implementation of the Kernel Page Table Isolation (KPTI) for
amd64, first version. It provides a workaround for the 'meltdown'
vulnerability. PTI is turned off by default for now, enable with the
loader tunable vm.pmap.pti=1.
The pmap page table is split into kernel-mode table and user-mode
table. Kernel-mode table is identical to the non-PTI table, while
usermode table is obtained from kernel table by leaving userspace
mappings intact, but only leaving the following parts of the kernel
mapped:
kernel text (but not modules text)
PCPU
GDT/IDT/user LDT/task structures
IST stacks for NMI and doublefault handlers.
Kernel switches to user page table before returning to usermode, and
restores full kernel page table on the entry. Initial kernel-mode
stack for PTI trampoline is allocated in PCPU, it is only 16
qwords. Kernel entry trampoline switches page tables. then the
hardware trap frame is copied to the normal kstack, and execution
continues.
IST stacks are kept mapped and no trampoline is needed for
NMI/doublefault, but of course page table switch is performed.
On return to usermode, the trampoline is used again, iret frame is
copied to the trampoline stack, page tables are switched and iretq is
executed. The case of iretq faulting due to the invalid usermode
context is tricky, since the frame for fault is appended to the
trampoline frame. Besides copying the fault frame and original
(corrupted) frame to kstack, the fault frame must be patched to make
it look as if the fault occured on the kstack, see the comment in
doret_iret detection code in trap().
Currently kernel pages which are mapped during trampoline operation
are identical for all pmaps. They are registered using
pmap_pti_add_kva(). Besides initial registrations done during boot,
LDT and non-common TSS segments are registered if user requested their
use. In principle, they can be installed into kernel page table per
pmap with some work. Similarly, PCPU can be hidden from userspace
mapping using trampoline PCPU page, but again I do not see much
benefits besides complexity.
PDPE pages for the kernel half of the user page tables are
pre-allocated during boot because we need to know pml4 entries which
are copied to the top-level paging structure page, in advance on a new
pmap creation. I enforce this to avoid iterating over the all
existing pmaps if a new PDPE page is needed for PTI kernel mappings.
The iteration is a known problematic operation on i386.
The need to flush hidden kernel translations on the switch to user
mode make global tables (PG_G) meaningless and even harming, so PG_G
use is disabled for PTI case. Our existing use of PCID is
incompatible with PTI and is automatically disabled if PTI is
enabled. PCID can be forced on only for developer's benefit.
MCE is known to be broken, it requires IST stack to operate completely
correctly even for non-PTI case, and absolutely needs dedicated IST
stack because MCE delivery while trampoline did not switched from PTI
stack is fatal. The fix is pending.
Reviewed by: markj (partially)
Tested by: pho (previous version)
Discussed with: jeff, jhb
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
On a SPROM-less device, the PCI(e) bridge core will be initialized with its
power-on-reset defaults; this can leave the SPROM-derived BHND_PCI_SRSH_PI
value pointing to the wrong backplane address. This value is used by the
PCI core when performing address translation between the static register
windows in BAR0 that map the PCI core's register block, and backplane
address space.
Previously, bhndb_pci(4) incorrectly used the potentially invalid static
BAR0 PCI register windows when attempting to correct the BHND_PCI_SRSH_PI
value in the PCI core's SPROM shadow.
Instead, we now read/update BHND_PCI_SRSH_PI by fetching the PCI core's
backplane address from the core enumeration table, and then using a dynamic
register window to explicitly map the PCI core's register block into BAR0.
Sponsored by: The FreeBSD Foundation
The implementation will follow (D12723). For now, get the changes to
commit-protected files out of the way.
Approved by: secteam (gordon)
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D13925
Focus on code where we are doing multiplications within malloc(9). None of
these is likely to overflow, however the change is still useful as some
static checkers can benefit from the allocation attributes we use for
mallocarray.
This initial sweep only covers malloc(9) calls with M_NOWAIT. No good
reason but I started doing the changes before r327796 and at that time it
was convenient to make sure the sorrounding code could handle NULL values.
Since we have no control over the name, the MAKEDEV_CHECKNAME flag must be
used to return an error on an invalid (to devfs) name instead of panicing.
r305900 that originally added this feature also introduced a few other bugs:
- Proper locking not performed
- Theoretically broke the expectation that the control event buffer would
not span more than one pages, but did not update the CTASSERT that was
in place to prevent this. However, since the struct virtio_console_control
and the bulk buffer together were quite small, this could not have happened.
Also workaround an QEMU VirtIO spec violation in that it includes the NUL
terminator in the buffer length when the spec says it is not included.
PR: 223531
MFC after: 1 week
Attaching syscon_generic earlier than BUS_PASS_DEFAULT makes it more
difficult for specific syscon drivers to attach to the syscon node and to
get ordering right. Further discussion yielded the following set of
decisions:
- Move syscon_generic to BUS_PASS_DEFAULT
- If a platform needs a syscon with different attach order or probe
behavior, it should subclass syscon_generic and match on the SoC specific
compat string
- When we come across a need for a syscon that attaches earlier but only
specifies compatible = "syscon", we should create a syscon_exclusive driver
that provides generic access but probes earlier and only matches if "syscon"
is the only compatible. Such fdt nodes do exist in the wild right now, but
we don't really use them at the moment.
Additionally:
- Any syscon provider that has needs any more complex than a spinlock solely
for syscon access and a single memory resource should subclass syscon
directly rather than attempting to subclass syscon_generic or add complexity
to it. syscon_generic's attach/detach methods may be made public should the
need arise to subclass it with additional attach/detach behavior.
We introduce aw_syscon(4) that just subclasses syscon_generic but probes
earlier to meet our requirements for if_awg and implements #2 above for this
specific situation. It currently only matches a64/a83t/h3 since these are
the only platforms that really need it at the time being.
Discussed with: ian
Reviewed by: manu, andrew, bcr (manpages, content unchanged since review)
Differential Revision: https://reviews.freebsd.org/D13793
