| Age | Commit message (Collapse) | Author |
|---|
|
ok mpi@
|
|
might free it.
Prevent a use-after-free in various aynchronous cases. Found while
looking at another user-after-free pointed out by ehrhardt@.
Committing now to find if other drivers rely on this use-after-free.
tb@ confirmed ubcmtp(4) is now fixed.
|
|
noop and prevent us from fixing real use-after-free.
Since xfer->status is reset by *_device_intr_done(), checking for
USBD_IOERROR cannot be true.
Problem reported by tb@ because ubcmtp(4) triggers some I/O errors
when multiple fingers are used.
|
|
|
|
ok mpi@
|
|
This diff corrects dmesg to properly show that all others are 1T1R, since right
now it only applies to 8188C and all others incorrectly display 0T0R.
hint & OK stsp@
|
|
to determine whether the interface was reset while we were sleeping. The flag
will be set if the interface is still down when the task wakes up, but the
interface could already be up again in which case the flag will be cleared.
ok mpi@ kettenis@
|
|
|
|
locking and unlocking. Make sure we don't unlock 7000 family NICs while
a command is being processed (later NICs apparently don't need this).
Inspired by similar approaches in Linux and Dragonfly, and a patch given
to me by Imre Vadasz.
Tested by me on 7265 and 8260 devices, and on a 7260 device by Stefan Wollny.
|
|
the driver needs to enable the CCK high power feature, as already done for
other chips supported by urtwn(4).
Same change as FreeBSD r311347.
Patch submitted by Kevin Lo. Tested by me.
|
|
Same changes as FreeBSD r311948 and r287584.
Patch submitted by Kevin Lo. Tested by me.
|
|
For a HT node, ni_txrate is always zero. We should be using ni_txmcs instead.
Simplify the if-else logic to make sure of that.
The mimo delimiter in the link quality command was never set.
I don't know how important this is. But Linux sets it, so why not.
Hardcode the lowest rate at the tail of the retry table.
While debugging the old code I have encountered retry tables filled
with only 'MCS 8' which is obviously not ideal.
While here, fix a misspelled function prototype.
Tested by kettenis@, mpi@, and Stefan Wollny
ok mpi@
|
|
such that "ifconfig dwge0 lladdr 00:11:22:33:44:55" actually works.
|
|
the GMAC found on the Allwinner A20/A31.
With this diff I can receive packets on the Firefly-RK3399. Unfortunately
sending packets doesn't seem to work yet.
|
|
|
|
|
|
|
|
|
|
ok kettenis@
|
|
|
|
only the RK3399 is supported.
|
|
|
|
the eMMC controller on the Rockchip RK3399 work.
|
|
the Rockchip RK3399.
- Make it possible to override sdhc_signal_voltage().
- Make it possible to disable double-data rate modes.
ok patrick@
|
|
|
|
mpi says a transfer can't exist without a pipe. ok mpi@
|
|
|
|
|
|
the random subsystem as entropy. This value is pretty much unknown,
and anyways our entropy input ring does not saturate from knowns.
ok mikeb djm
|
|
No binary change.
OK ratchov@
|
|
instead of "device timeout". The latter is printed by the Tx watchdog already.
|
|
MBIM already uses network byte order for IP addresses, so just
use them as they are.
ok stsp@, deraadt@
|
|
When MIMO support was added several entries in the rate table index enum
got a wrong value. On most devices the firmware ended up using a Tx rate
different from the rate net80211 asked for (which is bad, but not fatal).
But on 3165 devices which do not support MIMO the firmware rightly raised
a fatal error whenever the driver mistakenly asked for a MIMO Tx rate.
Reported by Georgios Pediaditis, Steve Throckmorton, and Kai Wirt on misc@
Thanks to benno@ for providing 3165 hardware I could use to debug this.
|
|
Bombardieri" <mb@ii.net>.
|
|
|
|
for Allwinner devices. Implement support for the latter such that we can
use newer device trees.
ok patrick@
|
|
(kernel part)
This feature is for OpenBSD guests only.
ok reyk, kettenis
|
|
|
|
Using a name with the 'db_' prefix makes it invisible from the dynamic
profiler.
ok deraadt@, kettenis@, visa@
|
|
RK3399 available to other drivers through the regmap interface.
|
|
It doesn't compile und hasn't been working during the last decade.
ok kettenis@, deraadt@
|
|
spirit to the xhci_fdt.c glue. The new code sets up any attached USB PHYs
based on their "compatible" property. All the hardware supported by sxiehci.c
should be supported by this new code. In addition to that this adds support
for the EHCI controllers found on various Rockchip hardware such as the
RK3288 and RK3399 SoCs.
ok patrick@
|
|
r1.25. Fixes creation of concat volumes.
Noticed by and diff from Thordur I. Bjornsson via tech@
ok jsing@
|
|
ok mlarkin
|
|
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 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.
|
|
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@
|
|
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@
|
