On Tue, 28 Nov 2017 05:52:25 +, Theo de Raadt wrote:
> > In fact, can recallocarray be faster than plain free followed by calloc?
>
> Yes.
>
> I think you are missing some nuances. These added functions have fast paths
> and slow paths. freezero() isn't just a bzero, it also has munmap()
> se
Below is a DEBUG dump from failing OPTIONS+GET+GET, e.g. header
X-Forwarded-Port is not set.
accept_reserve: inflight incremented, now 1
relay_read_http: session 1: size 87, to read -2
relay_read_http: session 1: header 'OPTIONS: /options.php HTTP/1.1'
relay_read_http: session 1: header 'Host: te
> In fact, can recallocarray be faster than plain free followed by calloc?
Yes.
I think you are missing some nuances. These added functions have fast paths
and slow paths. freezero() isn't just a bzero, it also has munmap()
sequences. You are adding forced bzero or munmap() in circumstances
wh
On Mon, 27 Nov 2017 08:01:25 +, Otto Moerbeek wrote:
> On Mon, Nov 27, 2017 at 05:48:14AM +, kshe wrote:
> > On Mon, 27 Nov 2017 00:42:01 +, Theo de Raadt wrote:
> > > This needs worst-case performance measurements.
> >
> > The only instances where performance could be a problem are in
On Mon, Nov 27, 2017 at 09:39:03PM +0100, Stefan Sperling wrote:
>
> This makes athn(4) report similar RSSI values as iwn(4) does,
> instead of bugos positive dBm values. The driver forgot about
> adding the default noisefloor to measured RSSI values.
>
> The same is already done in the USB athn(
On Mon, Nov 27, 2017 at 09:30:59PM +0100, Stefan Sperling wrote:
>
> Fix a comment which misidentifies the field where RSSI values occur.
> Add macros to access RSSI info in ds_status4 as well.
>
> ok?
Sure, ok kevlo@
> Index: ar5008reg.h
> ==
> On Nov 27, 2017, at 9:54 AM, Jeremie Courreges-Anglas wrote:
>
> On Fri, Nov 24 2017, Scott Cheloha wrote:
>> Hi,
>>
>> [...]
>>
>> Thoughts and feedback?
>
> This seems to mix refactoring, eg changing the signature of
> timer_add_event(), with semantic changes. Could you please
> split
Hello,
adds two test cases for issues reported by Leonardo. I've created extra
pfloadanchors target in regress/sbin/pfctl/Makefile. The 'load anchor ... from
' construct still needs more love to be covered by existing targets.
consider output for command 'pfctl -o none -nvf pf113.in':
Hi,
i had forgotten about this already, but this has existed for a while.
systat does consistently show "66 tick"(around 108 for stattick), while
root@av7marsb:~ # vmstat -i
interrupt total rate
irq32/sximmc0 35222 19
irq55/sxie0
Factor out the reset / closing of a relay connection. Currently only
relay_close is using this but I want to use this in relay_http.c to close
the backend connection after a request is done.
OK?
--
:wq Claudio
Index: relay.c
===
RCS
On Mon, Nov 27, 2017 at 07:00:51PM +0100, Otto Moerbeek wrote:
> On Mon, Nov 27, 2017 at 10:34:50AM +, Jason McIntyre wrote:
>
> > On Sun, Nov 26, 2017 at 07:18:33PM +, kshe wrote:
> > > Hi,
> > >
> > > The manual page for dc(1) is very careful about signalling which
> > > commands are no
Hi,
1G is too smallto hold a cvs checkout, its 1.1G right now.
ok?
(benno_disklabel_editor)
Index: editor.c
===
RCS file: /cvs/src/sbin/disklabel/editor.c,v
retrieving revision 1.308
diff -u -p -r1.308 editor.c
--- editor.c
Hi,
This converts the pf rule structure to use the divert type.
Old semantics was:
divert.port: divert-to or divert-reply
divert.addr: divert-to
divert_packet.port: divert-packet
Now we have one divert structure with an explicit type.
ok?
bluhm
Index: sys/net/pf.c
Hello,
> Hi,
>
> The divert structure uses the port number to indicate that divert-to
> or divert-reply is active. Divert packet uses a separate structure.
> This is confusing and makes it hard to add new features. It is
> better to have a divert type that explicitly says what is configured.
>
On Mon, Nov 27, 2017 at 11:46:54PM +0100, Claudio Jeker wrote:
> On Mon, Nov 27, 2017 at 11:34:49PM +0100, Alexander Bluhm wrote:
> > On Mon, Nov 27, 2017 at 11:26:06PM +0100, Claudio Jeker wrote:
> > > Guess we should make the 3 fatalx() in that code different so that it
> > > becomes more clear o
On Mon, Nov 27, 2017 at 11:34:49PM +0100, Alexander Bluhm wrote:
> On Mon, Nov 27, 2017 at 11:26:06PM +0100, Claudio Jeker wrote:
> > Guess we should make the 3 fatalx() in that code different so that it
> > becomes more clear on which call it fails.
>
> As in the diff below?
>
> > Wonder what ki
On Mon, Nov 27, 2017 at 11:26:06PM +0100, Claudio Jeker wrote:
> Guess we should make the 3 fatalx() in that code different so that it
> becomes more clear on which call it fails.
As in the diff below?
> Wonder what kind of startup race we lose...
I can reproduce one of the errors on my laptop:
On Mon, Nov 27, 2017 at 11:11:59PM +0100, Alexander Bluhm wrote:
> On Mon, Nov 27, 2017 at 11:04:49PM +0100, Alexander Bluhm wrote:
> > And I am waiting for my loop to fail ...
>
> Now I have got a different error:
>
> execute: ssh ot2 perl -I /usr/src/regress/usr.sbin/relayd
> /usr/src/regress/
On Mon, Nov 27, 2017 at 11:04:49PM +0100, Alexander Bluhm wrote:
> On Mon, Nov 27, 2017 at 10:40:33PM +0100, Claudio Jeker wrote:
> > Does not happen here. Running
> > while make run-regress-args-https-inspect.pl ; do echo -n; done
> > for a few minutes now and no failure.
>
> It takes a while
On Mon, Nov 27, 2017 at 11:04:49PM +0100, Alexander Bluhm wrote:
> And I am waiting for my loop to fail ...
Now I have got a different error:
execute: ssh ot2 perl -I /usr/src/regress/usr.sbin/relayd
/usr/src/regress/usr.sbin/relayd/remote.pl copy 10.188.81.22 10.188.81.21 37198
/usr/src/regres
Adapt the changes Joel Sing did to httpd a while ago to move to
tls_config_set_ecdhecurves() to relayd.
This removes and changes the ecdhe config in relayd.conf but I assume not
many people are setting non default values there anyway. With this diff
multipl cureves can be selected instead of just o
On Mon, Nov 27, 2017 at 10:40:33PM +0100, Claudio Jeker wrote:
> Does not happen here. Running
> while make run-regress-args-https-inspect.pl ; do echo -n; done
> for a few minutes now and no failure.
It takes a while. I am running it on
- very old and slow i386 machine, different timing
-
On Mon, Nov 27, 2017 at 10:27:50PM +0100, Alexander Bluhm wrote:
> On Mon, Nov 27, 2017 at 08:23:29PM +0100, Claudio Jeker wrote:
> > Instead of using imsg to pass certificates, pass the fd to the cert to the
> > relay processes. This allows for large certificates and esp. ca file to
> > work. OCSP
On Mon, Nov 27, 2017 at 08:23:29PM +0100, Claudio Jeker wrote:
> Instead of using imsg to pass certificates, pass the fd to the cert to the
> relay processes. This allows for large certificates and esp. ca file to
> work. OCSP stapling will also be added through this.
relayd regression tests pass
Hi,
The divert structure uses the port number to indicate that divert-to
or divert-reply is active. Divert packet uses a separate structure.
This is confusing and makes it hard to add new features. It is
better to have a divert type that explicitly says what is configured.
This is the first par
This makes athn(4) report similar RSSI values as iwn(4) does,
instead of bugos positive dBm values. The driver forgot about
adding the default noisefloor to measured RSSI values.
The same is already done in the USB athn(4) driver.
It looks like noisefloor calibration is not yet enabled in this dr
Fix a comment which misidentifies the field where RSSI values occur.
Add macros to access RSSI info in ds_status4 as well.
ok?
Index: ar5008reg.h
===
RCS file: /cvs/src/sys/dev/ic/ar5008reg.h,v
retrieving revision 1.4
diff -u -p -r1.
On Sat, Nov 18, 2017 at 02:58:29AM +0100, Jeremie Courreges-Anglas wrote:
> On Sun, Nov 12 2017, Scott Cheloha wrote:
> > Hi,
> >
> > GNU ld has prefixed the contents of .gnu.warning.SYMBOL sections
> > with "warning: " since 2003, so the messages themselves need not
> > contain the prefix anymore
Instead of using imsg to pass certificates, pass the fd to the cert to the
relay processes. This allows for large certificates and esp. ca file to
work. OCSP stapling will also be added through this.
OK?
--
:wq Claudio
Index: ca.c
=
Here is setup which reproduces this problem. Also exists in 6.2.
Server:
Apache with mod_php serving following content:
———cut options.php——
{ 1.2.3.4 }
relay web_test {
listen on 5.6.7.8 port 80
protocol http_relay
forward to port 80 mode loadbalance check tcp
}
Cl
On Mon, Nov 27 2017, Stefan Sperling wrote:
> On Mon, Nov 27, 2017 at 02:33:59AM +0100, Stefan Sperling wrote:
>> Most people I've talked to seem to be OK with never exposing
>> these secrets to userland in the first place.
Makes sense.
> Better diff for the wireless part.
> WEP keys showed up a
On Mon, Nov 27, 2017 at 02:33:59AM +0100, Stefan Sperling wrote:
> Most people I've talked to seem to be OK with never exposing
> these secrets to userland in the first place.
Better diff for the wireless part.
WEP keys showed up as 0x0... instead of '' in
the previous diff.
Index: ieee80211_
On Mon, Nov 27, 2017 at 10:34:50AM +, Jason McIntyre wrote:
> On Sun, Nov 26, 2017 at 07:18:33PM +, kshe wrote:
> > Hi,
> >
> > The manual page for dc(1) is very careful about signalling which
> > commands are non-portable extensions, with the exception of the `e'
> > command, which is a
relayd does this evil thing of replacing the output buffer of bufferevnt
with another one. This makes it impossible to reopen connections. Instead
just write out the output buffer when connecting to the backend and move
on.
THis is needed to make HTTP keep-alive work better.
--
:wq Claudio
Index
On Sun, 26 Nov 2017 21:55:21 +0100 Jeremie Courreges-Anglas wrote:
> Yup. Before someone proposes to remove or significantly change the
> content of the version string: please think about shell scripts out
> there that might use KSH_VERSION.
Long thread from Feb 2015:
http://openbsd-archive.7691.n
On 2017/11/27 11:22, Peter Hessler wrote:
> :retrieving revision 1.319
> :diff -u -p -r1.319 ip_carp.c
> :--- netinet/ip_carp.c21 Nov 2017 09:08:55 - 1.319
> :+++ netinet/ip_carp.c27 Nov 2017 01:29:34 -
> :@@ -2158,9 +2158,8 @@ carp_ioctl(struct ifnet *ifp, u_long cmd
>
On Mon, Nov 27 2017, Helg wrote:
> fusefs_link returns the wrong error code when attempting to create a
> hard link to a directory. It returns EISDIR when it should instead
> return EPERM. Discovered while running the ffs test suite on ntfs-3g
> and confirmed by comparing to ufs.
>
> This is the
On Fri, Nov 24 2017, Scott Cheloha wrote:
> Hi,
>
> These events are meants to fire after an interval has elapsed,
> so we should use the monotonic clock to measure.
>
> The pattern throughout the daemon of loading the current time,
> adding a timeout to the structure member, and then passing a
>
On 2017 Nov 27 (Mon) at 16:27:57 +0100 (+0100), Stefan Sperling wrote:
:On Mon, Nov 27, 2017 at 11:22:46AM +0100, Peter Hessler wrote:
:> This hides the username that is used, not the password/authkey. Is the
:> username private information?
:
:Yes it seems best to avoid exposing these.
:We cannot
fusefs_link returns the wrong error code when attempting to create a
hard link to a directory. It returns EISDIR when it should instead
return EPERM. Discovered while running the ffs test suite on ntfs-3g
and confirmed by comparing to ufs.
This is the description for the test that fails:
"link
On Mon, Nov 27, 2017 at 11:22:46AM +0100, Peter Hessler wrote:
> This hides the username that is used, not the password/authkey. Is the
> username private information?
Yes it seems best to avoid exposing these.
We cannot assume that telcos follow best practices of data hygiene.
Quoting naddy@ fr
On Wed, Nov 22, 2017 at 05:26:24PM +0100, Patrick Wildt wrote:
> On 2017/06/25 21:44, Tim Stewart wrote:
> > My first patch did, in fact, break Child SAs rekeying. I have a new
> > patch at the end of this message that simply restricts DH group
> > negotiation to IKE SAs (I *think* that DH group g
On 27/11/17(Mon) 11:49, Mathieu - wrote:
> Hi everyone,
>
> I was looking / poking around the filedesc handling in kern_descrip.c
> and found the locking a bit.. weird,
Can you define "weird"? Is it a taste thing or did you find any bug?
If it's a bug how can you reproduce it?
>
On Mon, Nov 27, 2017 at 10:43:09AM +0100, Martin Pieuchot wrote:
> Here's a diff that includes that and prevent a user-after-free pointed
> out by visa@. We should not try to dereference `rt' if nd6_free() has
> been called.
>
> Hrvoje Popovski confirmed he couldn't reproduce the panic with this
On Mon, Nov 27, 2017 at 12:20:34PM +0100, Martin Pieuchot wrote:
> Questions, comments, tests?
New panic with regress. I think it was sys/kern/sosplice this time.
login: panic: kernel diagnostic assertion "_kernel_lock_held()" failed: file
"/usr/src/sys/kern/uipc_socket.c", line 1882
Stopped at
On 20/11/17(Mon) 16:22, Martin Pieuchot wrote:
> Diff below remove the KERNEL_LOCK() around all pr_input() routines.
> It's a bit rough so I'd appreciate more tests before splitting it into
> pieces.
>
> I'm using tasks to delay selwakeup/csignal calls, just like I did for
> bpf(4).
Updated versi
Hi everyone,
I was looking / poking around the filedesc handling in kern_descrip.c
and found the locking a bit.. weird, especially the fd_getfile function
is touching protected members of the filedesc w/o taking any lock.
This has been already hit previously in [1] and fixed by reordering the
mall
On Sun, Nov 26, 2017 at 07:47:01PM +, kshe wrote:
> Hi,
>
> I noticed a certain number of inaccuracies within the manual page for
> sed. The diff below corrects to most obvious ones, although further
> improvements are certainly possible.
>
> Additionally, the script given in the EXAMPLES se
On Sun, Nov 26, 2017 at 07:18:33PM +, kshe wrote:
> Hi,
>
> The manual page for dc(1) is very careful about signalling which
> commands are non-portable extensions, with the exception of the `e'
> command, which is a more recent addition.
>
> Index: dc.1
>
On 2017 Nov 27 (Mon) at 02:33:59 +0100 (+0100), Stefan Sperling wrote:
:On Mon, Nov 27, 2017 at 01:31:17AM +0100, Stefan Sperling wrote:
:> On Sun, Nov 26, 2017 at 06:17:14PM +0100, Jeremie Courreges-Anglas wrote:
:> >
:> > I don't think anything has been committed regarding this issue, right?
:>
Hello!
In my laptop, the BIOS does not configure properly the hardware. The
OpenBSD code does, instead, as regards the PCI-to-PCI bridge before my
network card.
Because no BIOS update and no other OS can do this, I would like to
know what my bridge actually needs. In sys/dev/pci/ppb.c:
void
ppb_al
On Tue, Nov 14, 2017 at 01:58:41PM +0100, Patrick Wildt wrote:
> On Thu, Nov 09, 2017 at 11:40:30AM +0100, Patrick Wildt wrote:
> > Hi,
> >
> > this diff implements MOBIKE (RFC 4555) support in iked(8), with us
> > acting as responder. In practice this support means that clients like
> > iPhones
On 23/11/17(Thu) 21:45, Helg wrote:
> On Thu, Nov 23, 2017 at 12:09:34PM +, Helg Bredow wrote:
> > - Forwarded message from Martin Pieuchot -
> >
> > Date: Sat, 18 Nov 2017 11:03:49 +0100
> > From: Martin Pieuchot
> > To: Helg Bredow
> > CC: "tech@openbsd.org"
> > Subject: Re: fuse
On 23/11/17(Thu) 15:34, Alexander Bluhm wrote:
> On Wed, Nov 22, 2017 at 04:24:22PM +0100, Martin Pieuchot wrote:
> > Diff below implements 3/ because it seems the simplest approach to
> > me and reduce differences with ARP a bit further.
>
> Yes.
>
> > void
> > -nd6_llinfo_settimer(struct llinf
On Sat, 25 Nov 2017 11:42:07 -0700 Joel Knight wrote:
> On Thu, Mar 9, 2017 at 10:02 PM, Joel Knight wrote:
> > Hi.
> >
> > snmpd(8) uses unsigned ints internally to represent the size and used
> > space of a file system. The HOST-RESOURCES-MIB defines the valid
> > values for those OIDs as 0..21
cd /usr/xenocara/lib
ftp https://mesa.freedesktop.org/archive/mesa-17.2.6.tar.gz
ftp http://jsg.id.au/mesa-update/mesa.diff.gz
tar zxf mesa-17.2.6.tar.gz
gunzip mesa.diff.gz
patch -p0 < mesa.diff
sed -i 's/mesa$/mesa-17.2.6/' Makefile
build xenocara as normal
Builds on at least amd64, i386, spar
On Mon, Nov 27, 2017 at 05:48:14AM +, kshe wrote:
> On Mon, 27 Nov 2017 00:42:01 +, Theo de Raadt wrote:
> > This needs worst-case performance measurements.
>
> The only instances where performance could be a problem are in
> vfprintf.c and vfwprintf.c, where the calls happen inside a loo
57 matches
Mail list logo
