| Age | Commit message (Collapse) | Author |
|---|
|
ok deraadt@
|
|
a byte or a half-word is needed. Certain implementations of the Synopsis
Designware copy-and-paste logic blocks don't respond to transactions that
are smaller than a word.
Fixes the serial console on boards with a Rockchip RK3288.
|
|
and identified early on. For the secondary CPUs this happens late, such that
the drivers we need to spin up CPUs, such as psci(4), will be available.
This also fixes some code in simplebus(4) where the return value of
OF_getprop() was not properly checked.
Heavily based on an earlier diff from drahn@.
ok drahn@, jsg@
|
|
that it can be used in SVM and VMX.
no functional change
|
|
hardcoded number. no functional change
|
|
tested by reyk, dcoppa, and a few others.
ok kettenis@ on the fpu bits
ok deraadt@ on the vmm bits
|
|
the backout police even though it wasn't responsible for anything.
pointed out by tb@
|
|
Performance regression reported by deraadt@ and uvm_fault reported by fcambus@.
I will be poking further at this out of tree.
|
|
Since only leaf queues can have packets assigned to them,
H-FSC requires the user specified root queue to have a
parent. To simplify userland tools and the configuration
interface, the kernel can be leveraged to set it up.
ok henning
|
|
Some of these pointed out by Michael W. Bombardieri, I went through
the file and fixed the rest.
No functional change.
|
|
Since we now sync the entire Rx DMA buffer the SYNC_RESP_STRUCT() macro which
was used to sync just parts of the buffer can be replaced by assignments.
The SYNC_RESP_PTR() macro was already unused so just remove it.
The ADVANCE_RXQ() macro was used just once so expand it in place.
|
|
All callers were passing the same constant IWM_RBUF_SIZE, which is derived
from the smallest RX buffer the hardware supports (4k). Smaller sizes don't
make sense anyway and larger sizes are not on our horizon for now.
|
|
Some of these pointed out by Michael W. Bombardieri, I went through
the file and fixed the rest.
No functional change.
|
|
a frame which was retried N times was retried just once.
This punishes retry-heavy data rates even more and should reduce latency.
It does not seem to have a noticable effect on net throughput in my testing.
If this change causes throughput problems for anyone, let me know.
ok tb@ mlarkin@
|
|
ok kettenis
|
|
they have a different ID from "real" Cortex-A17 cores.
ok phessler@, patrick@
|
|
MDIO controller drivers are attached using fdt.
|
|
a device-tree. It will be needed later when more device drivers
are attached using fdt.
OK kettenis@
|
|
this was wrong and was corrected at the end of last year in rev 1.140.
Before then a 64 bit BAR was not enforced as the test was wrong.
It turns out there exist 82598 parts which have a 32 bit BAR so change
the test to only require a memory BAR and not a 64 bit memory BAR.
Problem reported by Robert Blacquiere. ok mikeb@
|
|
The magic commands to capture raw 802.11 frames on e.g. channel 1 are:
ifconfig iwm0 mediaopt monitor chan 1
tcpdump -n -i iwm0 -y IEEE802_11_RADIO -s 4096 -w /tmp/iwm.pcap
ifconfig iwm0 -mediaopt monitor -chan
|
|
We can now clear SINGLE_FRAME_MSK in the receive configuration register.
Linux has not set this bit in years, and Dragonfly stopped setting it
in commit b5eb43f0280bbcfd26af51cf5a4b8e8ff3590b67.
This is a prerequisite for monitor mode since the firmware seems to ignore
the SINGLE_FRAME_MSK bit while running in monitor mode and sends us garbage.
Tested on 7260, 7265, and 8260 devices.
|
|
of the esr_el1 register for unhandled userland exceptions (and continue to
dump the registers in that case).
Handle breakpoint traps and make sure we call refreshcreds() on userland traps.
ok jsg@
|
|
Problem reported by Colton Lewis on misc@
ok tb@
|
|
with certain rulesets and excessively noisy; move them to LOG_INFO (which was
previously unused). ok benno@
|
|
|
|
Reported and patch tested by wen heping
|
|
|
|
by the tree of given state key are filled.
ok sasha
|
|
The ternary operator was causing the result to be promoted to
int for __swap16. Fixes warning with clang. OK guenther@
|
|
the size of the corresponding id field in struct iwm_host_cmd.
Patch by Imre Vadasz.
|
|
iwm_disable_rx_dma() was unintentionally releasing the nic lock too early.
Patch by Imre Vadasz.
|
|
The iwm_{read,write}_prph() calls don't grab the nic lock themselves
so make sure to acquire it where necessary.
The iwm_{set,clear}_bits_prph() on the other hand acquire/release the
nic lock themselves. Since the nic lock doesn't do recursive locking
we need to be careful to avoid accidentally releasing the nic lock
too early that way.
Patch by Imre Vadasz.
|
|
From Hiltjo Posthuma.
|
|
This gives fast access to the address on systems that implement
the UserLocal register. TCB caching is still used when running
in the single-threaded mode in order not to penalize old systems.
The kernel counterpart of this change must be in place before
using this diff!
With guenther@
|
|
This lets programs get the address without a system call on OCTEON II
and later.
Add UserLocal load emulation for systems that do not implement
the RDHWR instruction or the UserLocal register.
OK guenther@
|
|
from Anton Lindqvist
ok semarie
|
|
without <machine/mplock.h>.
|
|
|
|
PCIE entry as it used the same product id.
Add another PLX/Avago/"Broadcom" PCIE bridge seen in the wild while here.
|
|
|
|
on amd64 and i386.
|
|
|
|
Loosely based on a diff from Christian Ludwig
|
|
|
|
Go-ahead from kettenis@, guenther@, deraadt@
|
|
on amd64 and i386.
With guenther@
|
|
declarations since clang doesn't register the %-prefixed ones.
|
|
ok deraadt@ visa@
|
|
ok deraadt@ visa@
|
|
Merge these functions together and remove icmp6_rip6_input().
OK mpi@
|
