Good evening Mark, tech@,
I was trying to enable brightness control on Pinebook, so it could be set via:
wsconsctl display.brightness=N%.
I added sxipwm and pwmbl drivers. First relies on "pwm" bindings,
2nd - on "pwm-backlight".
I had it working for some time now. For PWM, I tried to replicate S
The zstd package includes a zstdgrep script, which should behave like zgrep,
however it assumes a few gnu grep behaviors we don't support.
1. --label=name prints a custom label, so you can get file.txt, not
file.txt.zst in the output.
2. It uses - for stdin instead of a missing argument.
Both ar
On Sun, Oct 06, 2019 at 12:04:49PM +0200, Stefan Sperling wrote:
> The net80211 stack currently displays every received control frame
> as "discarded input control packet" in netstat(1).
>
> We do in fact process "power saving poll" and "block ack request" frames.
> Such frames should not be counte
> Date: Sun, 6 Oct 2019 23:44:28 +1000
> From: Jonathan Matthew
>
> I noticed the new dwge(4) locks up after a while running tcpbench.
> After scratching my head quite a lot I realised it wasn't restarting the
> send queue after processing tx completions, it was just clearing OACTIVE,
> Changing
get_process_info() fills the global handle structure and returns a
pointer to it, which is later passed to format_next_process() to print
each process on the screen.
Without casting back and forth between `struct handle *' and `caddr_t'
I find the code is easier to read.
No object change.
OK?
I noticed the new dwge(4) locks up after a while running tcpbench.
After scratching my head quite a lot I realised it wasn't restarting the
send queue after processing tx completions, it was just clearing OACTIVE,
Changing this to call ifq_restart() instead (as prescribed in ifq.h) things
work much
All compare_*() functions (corresponding to the various sort orders)
take void pointers as they are passed to qsort(3), so the usual casting
inside follows.
Yet, as SETORDER() eventually assigns the actual keys, it can also
directly do so from the function arguments instead of an intermediate
stru
On Sun, Oct 06, 2019 at 12:04:49PM +0200, Stefan Sperling wrote:
> The net80211 stack currently displays every received control frame
> as "discarded input control packet" in netstat(1).
>
> We do in fact process "power saving poll" and "block ack request" frames.
> Such frames should not be count
The net80211 stack currently displays every received control frame
as "discarded input control packet" in netstat(1).
We do in fact process "power saving poll" and "block ack request" frames.
Such frames should not be counted as discarded.
ok?
(diff uses 5 context lines for easier review)
diff