This patch removes explicit kernel locking from futex_wait(). I think
the original reason for the KERNEL_LOCK() was that rwsleep() with PCATCH
needed it. Now rwsleep() does the necessary locking internally.
As a subtle detail, the copyin32() call is now run outside the kernel
lock. Algorithm-wise
Scott Cheloha wrote:
> On Fri, Nov 19, 2021 at 07:42:18PM -0700, Theo de Raadt wrote:
> > +#include /* for MAXBSIZE */
> >
> > No way, that is non-POSIX namespace. We are going in precisely the opposite
> > direction, eliminating this non-portability from the tree.
> >
> > No biblical scrolls
On Fri, 19 Nov 2021 21:31:12 -0600, Scott Cheloha wrote:
> Is there a nicer way to pick a "reasonable" buffer size when we just
> want to move as many bytes as possible on a given platform without
> hogging the machine?
Not really. But I don't think you need to worry about "hogging the
machine"
On Fri, Nov 19, 2021 at 07:42:18PM -0700, Theo de Raadt wrote:
> +#include /* for MAXBSIZE */
>
> No way, that is non-POSIX namespace. We are going in precisely the opposite
> direction, eliminating this non-portability from the tree.
>
> No biblical scrolls have it written "all programs must u
+#include /* for MAXBSIZE */
No way, that is non-POSIX namespace. We are going in precisely the opposite
direction, eliminating this non-portability from the tree.
No biblical scrolls have it written "all programs must use buffer sizes
decided by a system header file".
If you want 64*1024 in t
Expanding the tee(1) read buffer trivially increases the throughput.
In most applications the writer or the disk will be the bottleneck,
but if tee(1) happens to be the bottleneck then the 8K buffer makes it
worse.
# dd(1) reading /dev/zero and writing to /dev/null
$ for i in $(jot 10); do nanotim
Updated diff. Re-lock dances were simplified in the unix(4) sockets
layer.
Reference counters added to unix(4) sockets layer too. This makes
pointer dereference of peer's control block always safe after re-lock.
The `unp_refs' list cleanup done in the unp_detach(). This removes the
case where th
This is three times the same thing since the code is copy-paste + tweak.
In genrsa there is a slight twist that involves not reaching into BIGNUM
and we can take the opportunity to get rid of some Windows 3.1 things by
calling the conversion routines instead of handrolling them.
The callbacks them
Straightforward conversion. Note that EVP_PKEY_get0_RSA() can't fail
after checking that we have an RSA key.
Index: x509.c
===
RCS file: /cvs/src/sbin/isakmpd/x509.c,v
retrieving revision 1.123
diff -u -p -r1.123 x509.c
--- x509.c
The manpage for nl(1) reads:
The nl utility treats the text it reads in terms of logical
pages. Unless specified otherwise, line numbering is reset at the
start of each logical page. A logical page consists of a header, a
body and a footer section; empty sections are valid. [...]
POSIX.
On Tue, Nov 16, 2021 at 9:56 AM Joel Knight wrote:
>
> Inspecting the REQUEST packets showed that dhclient was setting the
> ciaddr to the existing leased IP address while dhcpleased was not
> setting this field. RFC 2131 4.3.2 (with a nice summary at 4.3.6) is
> pretty strict about when ciaddr an
On Tue 2021.11.16 at 21:33 +, Lu?s Henriques wrote:
> Hi!
>
> I've tried to setup a line like:
>
> bind-key XF86MonBrightnessDown ""
>
> in my .cwmrc and the result was that no key event was sent to my windows.
>
> Looking at the code, it looks like XKeysymToKeycode() is returning 0 for
> t
I want to note all the people who testing this diff. Please be
sure your test exceeds the lifetime of the `tdb’ and some
rekeying cycles have been made.
> On 19 Nov 2021, at 20:09, Alexander Bluhm wrote:
>
> Hi,
>
> I got a new hardware in my testlab and could panic this machine
> easily with t
Hi,
I got a new hardware in my testlab and could panic this machine
easily with the tdb refcount diff.
Adding tdb refcounts for timeout fixes this for me. This allows
to get rid of the timeout reaper. I refcount all successful
timeout_add() and decrease in timeout handler. There are no
timeout
On Mon, Nov 08, 2021 at 01:07:58PM +0100, Stefan Sperling wrote:
> The last_rx timestamp which controls timeout of an Rx block ack
> session is not updated when a frame is received.
> This can result in the session timing out too early.
>
> Not a huge deal because the AP will simply request a new
> On Nov 18, 2021, at 23:59, Sebastien Marie wrote:
>
> On Thu, Nov 18, 2021 at 07:50:01PM -0600, Scott Cheloha wrote:
>>> On Thu, Nov 18, 2021 at 12:30:30PM +0100, Martin Pieuchot wrote:
>>> On 17/11/21(Wed) 09:51, Scott Cheloha wrote:
> On Nov 17, 2021, at 03:22, Martin Pieuchot wrote:
>>
On Fri, Nov 19, 2021 at 08:18:45AM +0100, Jan Stary wrote:
> On Nov 19 00:01:04, stef...@sdaoden.eu wrote:
> > Jan Stary wrote in
> > :
> > |On Nov 18 20:13:03, h...@stare.cz wrote:
> > |> On Nov 16 21:33:31, hen...@camandro.org wrote:
> > |>> I've tried to setup a line like:
> > |>> bind-key
17 matches
Mail list logo
