| Age | Commit message (Collapse) | Author |
|---|
|
directives. binutils requires it, and binutils is the authoritative
parser in this regard. This means arm64 +lld remains broken (but all
the other architectures work)
|
|
without src and comp work. That is what we've designed for, but since
we always have src and comp on our systems it is so easy to slide...
problem first noted by rpe.
|
|
expression line. Removing the extra semicolons makes both ld and
ldd accept the link script generated, although ldd produces a wrong
object file on arm64; ok kettenis@ deraadt@
|
|
we need suitable data structures. Organize the pf fragments with
two red-black trees. One is holding the address and protocol
information and the other has only the fragment id. This will allow
to drop fragemts for specific connections more aggressively. `
from markus@; OK sashan@
|
|
|
|
|
|
Hyper-V interrupts seem to be sometimes delivered before the message
becomes available on the channel ring. This is reproducible on hvs(4)
under load. This change is modelled on the workaround found in the
Linux driver.
|
|
|
|
|
|
and urtwn_r88e_dma_init() into one; from FreeBSD r291902.
- Change the way the number of pages for public queue is calculated;
Based on the vendor driver's _InitQueueReservedPage(), the formula for
calculation of the number of pages for public priority queue is as follows:
if the device has 3 endpoints:
PUBQ = total number of Tx pages - HQ - LQ - NQ
if the device has 2 endpoints (i.e. has high and normal priority queues):
PUBQ = total number of Tx pages - HQ - NQ
- Don't write MCS8..MCS15 Tx power if the WNIC is a 1-stream Tx;
- Set R88E_TX_PAGE_COUNT to a correct value;
- Minimal enhancement in urtwn_open_pipes() for upcoming rtl8192eu chipset
support;
Tested with USB-N10 Nano (rtl8188cu), DWA-131 rev B (rtl8192cu), and
TL-WN725N v2 (rtl8188eu) on amd64.
ok stsp@
|
|
M_LOOP flag. This broke IPv6 multicast. Always set M_LOOP when
going through if_input_local() and adjust the flag's comment.
report rzalamena@; OK mpi@
|
|
correct prologue if compiled with -DPROF.
ok deraadt@
|
|
hardcoding 64 is too optimistic.
|
|
previously it would figure out if there's enough items overall for
all the cpus to have full active an inactive free lists. this
included currently allocated items, which pools wont actually hold
on a free list and cannot predict when they will come back.
instead, see if there's enough items in the idle lists in the depot
that could instead go on all the free lists on the cpus. if there's
enough idle items, then we can grow.
tested by hrvoje popovski and amit kulkarni
ok visa@
|
|
architectures we must also link against a gapdummy.o to repair the ABI
of the resulting .o file. Woe is me.
Also repair install: target to update the linkkit & hash when a kernel
is installed.
ok rpe tb mlarkin and tested by others also
|
|
so that an unhibernate kernel can detect if it is running with the
kernel it booted.
ok mlarkin
|
|
not to perform redundant work. There is no direct way to flush the
cache; rather, the kernel evicts unwanted bits from the cache by
executing a block of instructions, which is expensive.
With this diff, make build time decreases about 0.4%.
Diff from miod@
|
|
ok miod@
|
|
|
|
cpu_hatch() on each cpu, allowing initialization of per-cpu features on each
cpu as it starts up.
input from reyk@ and mikeb@
ok mikeb@
|
|
|
|
|
|
valid. This is an undocumented little debugging diagnostic, much like
the '*'.
|
|
(BTW, the binutils disasm cannot handle that instruction with those
parameters, because the decoder logic is wrong..)
|
|
|
|
While here, make SCSI command completion routine more robust.
|
|
|
|
|
|
== NULL)". The problem was that setting the inp pointer in the
statekey to NULL was delayed until the statekey refcounter reached
0. So the inp could get linked to another statekey while an mbuf
in the socket buffer was keeping the refcounter at 1. Set the
statekey inp pointer to NULL in pf_state_key_detach() immediately,
then the kassert can be even stricter.
OK sashan@
|
|
|
|
similar to smashing versions written by visa
|
|
|
|
and bail out if the pointer has become invalid. This prevents
use-after-free memory accesses that corrupted the dwc2qtd pool.
This fix improves stability on the Raspberry Pi 3.
From Linux commit dc8730846948e517169f630826cd2c97615f5ee8
OK kettenis@
|
|
They're might not be fully constructed.
ok mpi@ deraadt@ bluhm@
|
|
The IFF_RUNNING flag in struct ifnet serves the same purpose.
ok mpi@ kettenis@ deraadt@
|
|
interface is attached to the net80211 layer. Prevents confusion
in cases where drivers forget to initialize the link state.
ok mpi@ kettenis@
|
|
|
|
to ipip_input_if() and always pass the ifp. Only dump the packet
to bpf if we are called with a gif(4) interface.
OK mpi@
|
|
to grab the rwlock.
Problem reported by Rivo Nurges.
ok bluhm@
|
|
pf table code iterates on a radix tree and currently relies on the
NET_LOCK() to do the serialization.
Found the hardway by jasper@
|
|
supports that feature before we try to use it.
discussed with deraadt
|
|
#TF exceptions during guest VM boot
ok brynet
|
|
if the lock around the global depot of extra cache lists is contented
a lot in between the gc task runs, consider growing the number of
entries a free list can hold.
the size of the list is bounded by the number of pool items the
current set of pages can represent to avoid having cpus starve each
other. im not sure this semantic is right (or the least worst) but
we're putting it in now to see what happens.
this also means reality matches the documentation i just committed
in pool_cache_init.9.
tested by hrvoje popovski and amit kulkarni
ok visa@
|
|
presume we booted from. If you boot from another kernel, we cannot help
you later with hibernate, sorry -- The kernel does not get a useable
filename from the bootblocks.
In the bootblocks, detect a live hibernate signature and boot from
/bsd.booted instead.
with yasuoka, lots of discussion with mlarkin, ok tom
|
|
found by regress/sys/kern/pledge/generic; OK deraadt@
|
|
bugs could easily result in use-after-free or double free. Introduce
m_freemp() which automatically resets the pointer before freeing
it. So we have less dangling pointers in the kernel.
OK krw@ mpi@ claudio@
|
|
error. Make the ip_mforward() return value consistent. Simplify
the caller logic in ipv6_input() like in IPv4.
OK mpi@
|
|
|
|
of the code.
|
|
|
