ok, comments below
Reyk Floeter(r...@openbsd.org) on 2016.09.02 19:03:47 +0200:
> Hi,
>
> after all the preparation, the following diff adds support for
> fork+exec in relayd; based on rzalamena@'s work for httpd.
>
> Notes:
> - proc.c is identical to the version that is already in httpd.
> - Th
> Date: Fri, 2 Sep 2016 19:45:39 +0200
> From: Christian Weisgerber
>
> I would like to sync the system time periodically back to the RTC.
>
> Currently we update the RTC
> (1) when the time is set with clock_settime() or settimeofday(), which
> never happens for a typical ntpd setup;
> (2)
On Sat, Sep 03, 2016 at 11:42:54AM +0200, Sebastian Benoit wrote:
> ok, comments below
>
The proc.c is identical (except one #include) to httpd, I'll update
the DPRINTFs in httpd's proc.c accordingly and sync it with relayd
later. In one case it is DEBUG > 1, so no DPRINTF is needed.
Reyk
> Re
On 08/30/16 02:32, Alexandr Nedvedicky wrote:
Hello,
mikeb has just pointed out the patch fell under the desk asking me to resend
it.
henning@ and mikeb@ showed some interest to change handling of once rules to
the same way as PF has it on Solaris. Just to refresh the audience on once
option o
On Fri, Sep 2, 2016 at 7:17 PM, Michal Mazurek wrote:
> I don't think the -s option is ever ignored now.
>
> OK?
ok guenther@
Hi,
I've converted ikectl(8) from getpass to readpassphrase.
===
RCS file: /cvs/src/usr.sbin/ikectl/ikeca.c,v
retrieving revision 1.40
diff -u -p -r1.40 ikeca.c
--- ikeca.c 2 Nov 2015 12:21:27 - 1.40
+++ ikeca.c 3 S
Hello,
One of the teams in Oracle Solaris uses sophisticated naming scheme for PF
rulesets. The anchor (ruleset) is identified by something like that:
root/whatever:component:name/some-virtual-instance-long-name/inbound
That particular team hit a bug in pfctl, when they were trying to load r
The next MP problem when using route entries are stale ifas.
Stale ifas exist because we do not purge corresponding route entries
when we remove an ifa. Instead we try to 'fixup' such routes when
we found them later on. This happens in two places when cloning
a route, inside rtalloc(9), or when
I have been using a 5.9 amd64 box with 2 em interfaces configured as a
bridge for some time between my ISP's modem/router and my home network.
With this setup on 5.9, things worked perfectly.
On 6.0 I have tried the same very simple configuration of creating a bridge
with em0 and em1 interfaces an
On 09/03/16 13:55, Aaron Riekenberg wrote:
I have been using a 5.9 amd64 box with 2 em interfaces configured as a
bridge for some time between my ISP's modem/router and my home network.
With this setup on 5.9, things worked perfectly.
On 6.0 I have tried the same very simple configuration of cre
The system call is only called from librthreads, and librthreads doesn't
use tickets.
There is a comment in librthreads that says:
/*
* tickets don't work yet? (or seem much slower, with lots of system time)
* until then, keep the struct around to avoid excessive changes going
* back and forth.
Michal Mazurek wrote:
> The system call is only called from librthreads, and librthreads doesn't
> use tickets.
>
> There is a comment in librthreads that says:
> /*
> * tickets don't work yet? (or seem much slower, with lots of system time)
> * until then, keep the struct around to avoid excess
On 09:53:15, 3.09.16, Ted Unangst wrote:
> please keep this line.
Index: sys/kern/kern_synch.c
===
RCS file: /cvs/src/sys/kern/kern_synch.c,v
retrieving revision 1.133
diff -u -p -r1.133 kern_synch.c
--- sys/kern/kern_synch.c 6
On Sat, Sep 3, 2016 at 2:53 PM, Ted Unangst wrote:
> Michal Mazurek wrote:
>> The system call is only called from librthreads, and librthreads doesn't
>> use tickets.
>>
>> There is a comment in librthreads that says:
>> /*
>> * tickets don't work yet? (or seem much slower, with lots of system ti
On 2016/08/17 10:11, Peter Hessler wrote:
> It sure would be nice if we could see the PID of the process that added
> routes. Heck, route(8) even tries to print them already.
I haven't checked ifdefs so some of these might not need it, but easier to bump
them all. There's also the usual junk in g
The code says that the maximum packet size is 131024, but in reality
it's 65527. Values greater than 65527 will cause ping6 to fail:
$ ping6 -s 10 ::1
PING6 ::1 (::1): 10 data bytes
ping6: sendmsg: Invalid argument
ping6: wrote ::1 18 chars, ret=-1
Doc
This splits gzip into two processes, such that the heavy lifting is done in a
process with even fewer privileges. The idea is the child decompresses the
data and feeds it to the parent over a pipe. There's also a control pipe used
for some metadata that needs to be passed up.
early version, i'm no
Hello,
> >updated version is below.
> >
> >comments? O.K.?
>
> One comment, otherwise ok.
>
> Could you assert the lock is held otherwise, this might save
> effort if/when this code is refactored:
>
> else
> rw_assert_wrlock(&pf_consistency_lock);
sure.
also mikeb came t
It's defined as 0, this diff should not introduce any functional changes.
Index: lib/librthread//rthread.c
===
RCS file: /cvs/src/lib/librthread/rthread.c,v
retrieving revision 1.92
diff -u -p -r1.92 rthread.c
--- lib/librthread//rthr
There are more valid values than just CLOCK_REALTIME and
CLOCK_MONOTONIC.
Index: lib/libc/sys/__thrsleep.2
===
RCS file: /cvs/src/lib/libc/sys/__thrsleep.2,v
retrieving revision 1.5
diff -u -p -r1.5 __thrsleep.2
--- lib/libc/sys/__thr
The -w flag to jot() allows the user to specify a format string that
will be passed to printf after it was verified that it contains only one
conversion specifier. There are some subtle differences what jot's
getformat() accepts and what printf will do. Thus, I took vfprintf.c and
carved out whate
On Fri, Sep 02, 2016 at 03:36:50PM +0200, Michal Mazurek wrote:
> Here is a working futex implementation for OpenBSD. This diff touches
> the kernel and librthread.
>
> * get rid of tickets from rthreads, they were getting in the way and are
> unused anyway
> * replace all struct _spinlock wit
Vincent Gross writes:
> On Wed, 31 Aug 2016 15:26:53 +0200
> Vincent Gross wrote:
>
>> On Thu, 11 Aug 2016 16:57:27 +0100
>> Stuart Henderson wrote:
>>
>> > On 2016/06/27 13:00, Jérémie Courrèges-Anglas wrote:
>> [...]
>> > >
>> > > I also gave my ok to vgross by IM.
>> > >
>> > > I know
Hello,
there was still one more glitch catched by mikeb:
I have to sanitize pointer when copying rule to userland.
The other thing pointed out by mike is the Expired time should
be printed for expired rules only in debug mode output (pfctl -sr -g)
Incremental patch is as follows:
8<
On Sat, Sep 3, 2016 at 4:51 PM, Michal Mazurek wrote:
> There are more valid values than just CLOCK_REALTIME and
> CLOCK_MONOTONIC.
ok guenther@
On Sat, Sep 03, 2016 at 04:36:04PM +0200, Michal Mazurek wrote:
> The code says that the maximum packet size is 131024, but in reality
> it's 65527. Values greater than 65527 will cause ping6 to fail:
>
> $ ping6 -s 10 ::1
> PING6 ::1 (::1): 10 data bytes
> ping6: sendms
I really like this. It will allow us to move forward. While I was for a
long time holding the opinion that static routes should not disapear from
the routing table because the interface address changes I came to the
conclusion that this is causing more harm and so it is better to remove
all routes
Here's the diff for UTF-8 support in column(1).
OK?
Index: column.c
===
RCS file: /cvs/src/usr.bin/column/column.c,v
retrieving revision 1.24
diff -u -p -r1.24 column.c
--- column.c31 Aug 2016 20:43:57 - 1.24
+++ column.
Hello,
mikeb pointed out I should not be copy'n'pasting buggy code.
I'm just re-sending updated patch with change below:
8<---8<---8<--8<
diff -r e2383ec80feb src/sbin/pfctl/pfctl.c
--- a/src/sbin/pfctl/pfctl.cSat Sep 03 15:39:17 2016 +
On 3 September 2016 at 18:49, Alexandr Nedvedicky
wrote:
> Hello,
>
> mikeb pointed out I should not be copy'n'pasting buggy code.
> I'm just re-sending updated patch with change below:
>
OK mikeb
On 3 September 2016 at 18:03, Alexandr Nedvedicky
wrote:
> Hello,
>
> there was still one more glitch catched by mikeb:
>
> I have to sanitize pointer when copying rule to userland.
>
> The other thing pointed out by mike is the Expired time should
> be printed for expired rules only in debug
I forgot to add the iswspace(3) support.
Updated diff below.
On 09/03/16 18:46, Martijn van Duren wrote:
> Here's the diff for UTF-8 support in column(1).
>
> OK?
>
Index: column.c
===
RCS file: /cvs/src/usr.bin/column/column.c,v
re
Hi,
Here is a diff to remove the "smtpctl stop" command.
The proper way to stop a daemon is kill(1)/pkill(1) only.
It makes no sense to have different code path for that.
Eric.
Index: control.c
===
RCS file: /cvs/src/usr.sbin/smtpd/
On Sat, Sep 03, 2016 at 08:09:25PM +0200, Eric Faurot wrote:
> Hi,
>
> Here is a diff to remove the "smtpctl stop" command.
> The proper way to stop a daemon is kill(1)/pkill(1) only.
> It makes no sense to have different code path for that.
>
much agreed, queue code was designed to be kill-safe
On Wed, Aug 10, 2016 at 03:42:34PM +0200, Martin Pieuchot wrote:
> I'd ignore it as a first step, then revisit this later.
Attached a first stab at a seperate driver. Some points of interest:
- I've decided against using a custom hid descriptor, but instead define
the needed inputs in the drive
On Sat, Sep 03, 2016 at 08:12:20PM +0200, Gilles Chehade wrote:
> On Sat, Sep 03, 2016 at 08:09:25PM +0200, Eric Faurot wrote:
> > Hi,
> >
> > Here is a diff to remove the "smtpctl stop" command.
> > The proper way to stop a daemon is kill(1)/pkill(1) only.
> > It makes no sense to have different
There is only one result mentioned: ready-to-install binary packages.
Rob
Index: faq15.html
===
RCS file: /cvs/www/faq/faq15.html,v
retrieving revision 1.141
diff -u -p -r1.141 faq15.html
--- faq15.html 1 Sep 2016 12:05:14 -
committed. thanks Rob
On Sat, Sep 03, 2016 at 02:30:17PM -0400, Rob Pierce wrote:
> There is only one result mentioned: ready-to-install binary packages.
>
> Rob
>
> Index: faq15.html
> ===
> RCS file: /cvs/www/faq/faq15.html,v
> re
This is the last step to remove ticket support from librthreads.
Replace "struct _spinlock" with "_atomic_lock_t".
Index: lib/librthread/rthread.c
===
RCS file: /cvs/src/lib/librthread/rthread.c,v
retrieving revision 1.93
diff -u -p
Hi,
the LLVM project apparently uses the .cpp extension for all their C++
files. We already handle .cpp in sys.mk, but make depend does not
handle .cpp at all. This means that on "make depend" no .depend file
will be generated if there are only .cpp source files. But as SRCS is
set it will stil
Patrick Wildt writes:
> Hi,
>
> the LLVM project apparently uses the .cpp extension for all their C++
> files. We already handle .cpp in sys.mk, but make depend does not
> handle .cpp at all. This means that on "make depend" no .depend file
> will be generated if there are only .cpp source file
On Sat, Sep 03, 2016 at 10:42:51PM +0100, Patrick Wildt wrote:
> Hi,
>
> the LLVM project apparently uses the .cpp extension for all their C++
> files. We already handle .cpp in sys.mk, but make depend does not
> handle .cpp at all. This means that on "make depend" no .depend file
> will be gene
Hi tech@,
Some cleanups and tweaks for wc(1):
- Removed unnecessary string.h include
- Changed 'format_and_print' argument type to int64_t and casting
inside the function
- Declaring 'print_counts', 'format_and_print', and 'cnt' as static
- Remove unnecessary cast for NULL, and (void) casts fro
rdaemon allows us to go in the background, saving stderr error messages
and exit status for a longer interval in the startup step.
Thoughts?
Index: rtadvd.c
===
RCS file: /cvs/src/usr.sbin/rtadvd/rtadvd.c,v
retrieving revision 1.78
On Mon, Aug 22, 2016 at 11:10:28PM +0300, Pavel Korovin wrote:
> Dear all,
>
> I have local patch which implements IP/hostname logging for all SMTP
> operations.
> It simplifies log processing for me since I don't have to keep reference
> between session ids and IPs/hostnames and check it every t
Not sure if this is the right place or the right wording, but I think it
deserves a mention somewhere in the faq.
Regards,
Rob
Index: faq6.html
===
RCS file: /cvs/www/faq/faq6.html,v
retrieving revision 1.383
diff -u -p -r1.383 faq6
46 matches
Mail list logo