cleanup wsconsctl 4/4

2012-07-13 Thread Alexandr Shadchin
add options noinput and noyywrap -- Alexandr Shadchin Index: map_scan.l === RCS file: /cvs/src/sbin/wsconsctl/map_scan.l,v retrieving revision 1.3 diff -u -p -r1.3 map_scan.l --- map_scan.l 26 Jun 2008 05:42:06 - 1.3 +++ m

cleanup wsconsctl 3/4

2012-07-13 Thread Alexandr Shadchin
ansify -- Alexandr Shadchin Index: map_scan.l === RCS file: /cvs/src/sbin/wsconsctl/map_scan.l,v retrieving revision 1.3 diff -u -p -r1.3 map_scan.l --- map_scan.l 26 Jun 2008 05:42:06 - 1.3 +++ map_scan.l 13 Jul 2012 21:

cleanup wsconsctl 2/4

2012-07-13 Thread Alexandr Shadchin
fix typo -- Alexandr Shadchin Index: map_parse.y === RCS file: /cvs/src/sbin/wsconsctl/map_parse.y,v retrieving revision 1.6 diff -u -p -r1.6 map_parse.y --- map_parse.y 14 May 2012 06:19:51 - 1.6 +++ map_parse.y 13 Jul 201

cleanup wsconsctl 1/4

2012-07-13 Thread Alexandr Shadchin
zap whitespace -- Alexandr Shadchin Index: keysym.c === RCS file: /cvs/src/sbin/wsconsctl/keysym.c,v retrieving revision 1.6 diff -u -p -r1.6 keysym.c --- keysym.c28 Jun 2010 20:40:39 - 1.6 +++ keysym.c13 Jul 2012 2

Re: Fix type of WSKBDIO_?ETENCODING in wsconsctl.

2012-07-13 Thread Alexandr Shadchin
more fix type -- Alexandr Shadchin Index: display.c === RCS file: /cvs/src/sbin/wsconsctl/display.c,v retrieving revision 1.16 diff -u -p -r1.16 display.c --- display.c 20 Aug 2010 00:20:55 - 1.16 +++ display.c 13 Jul 2

Re: garbage collect USE_GCC3/TARGET_USE_GCC3/4

2012-07-13 Thread Brad Smith
On Thu, Jul 05, 2012 at 09:40:00PM -0400, Brad Smith wrote: > This garbage collects the USE_GCC3/TARGET_USE_GCC3/4 variables > within bsd.own.mk and Makefile.cross. I removed the last few > instances of USE_GCC3 within the ports tree a number of releases > ago and TARGET_USE_GCC3/4 was never used f

Re: some more net80211 node leaks

2012-07-13 Thread Stefan Sperling
On Fri, Jul 13, 2012 at 08:16:56PM +0200, Stefan Sperling wrote: > The ieee80211_dup_bss() function internally calls > ieee80211_alloc_node_helper(), which means we should check > the node cache for an existing entry before calling ieee80211_dup_bss(). > > ieee80211_node.c already does this but th

some more net80211 node leaks

2012-07-13 Thread Stefan Sperling
The ieee80211_dup_bss() function internally calls ieee80211_alloc_node_helper(), which means we should check the node cache for an existing entry before calling ieee80211_dup_bss(). ieee80211_node.c already does this but there are some instances in ieee80211_input.c where we fail to check for exis

Fix type of WSKBDIO_?ETENCODING in wsconsctl.

2012-07-13 Thread Alexandr Shadchin
>From NetBSD -- Alexandr Shadchin Index: keyboard.c === RCS file: /cvs/src/sbin/wsconsctl/keyboard.c,v retrieving revision 1.10 diff -u -p -r1.10 keyboard.c --- keyboard.c 20 Aug 2010 00:20:55 - 1.10 +++ keyboard.c 13 Jul

Re: fun with libtool for masochistic guys

2012-07-13 Thread Marc Espie
Okay, current decision is to actually pass options we do not understand to cc, and to look harder at options. So far, this is actually finding problems in ports. Stuff that would be completely ignored, or miscompiled. Check ports-changes@ for details.

Re: ral(4) tx/prio queue fixes (was: rt2560 tx/prio queue fixes)

2012-07-13 Thread Mark Kettenis
> Date: Fri, 13 Jul 2012 11:37:41 +0200 > From: Stefan Sperling > > On Tue, Jul 10, 2012 at 05:56:29PM +0200, Stefan Sperling wrote: > > Updated version that includes similar fixes for the rt2661 variants. > > > > This seems to help soekris-based ral APs that get stuck with the OACTIVE > > flag

Re: rt2560 bbp/antenna diff

2012-07-13 Thread Mark Kettenis
> Date: Sat, 7 Jul 2012 16:35:46 +0200 > From: Stefan Sperling > > rt2560 is selecting antennas before initialising the baseband > processor (BBP), however initialising antennas involves tweaking > of BBP registers. > > The diff below (taken from dragonfly, written by sephe) ensures > antennas a

Re: ftp mput recursiv upload diff for testing

2012-07-13 Thread Otto Moerbeek
On Fri, Jul 13, 2012 at 06:58:48PM +0200, Alexander Bluhm wrote: > On Fri, Jul 13, 2012 at 03:23:26AM +0200, Jan Klemkow wrote: > > > + char *cmd, *tp, *xargv[] = {argv[0], NULL, NULL}; > Put spaces inside {}: { argv[0], NULL, NULL } > > > It took me a while to figure out what that code d

Re: ftp mput recursiv upload diff for testing

2012-07-13 Thread Alexander Bluhm
On Fri, Jul 13, 2012 at 03:23:26AM +0200, Jan Klemkow wrote: > + char *cmd, *tp, *xargv[] = {argv[0], NULL, NULL}; Put spaces inside {}: { argv[0], NULL, NULL } > It took me a while to figure out what that code does, so I think the > comments are usefull for everybody who tries to read it.

Simplify wsconsctl (mouse part)

2012-07-13 Thread Alexandr Shadchin
Not need call ioctl(WSMOUSEIO_GCALIBCOORDS or WSMOUSEIO_SCALIBCOORDS) twice, enough once. OK ? -- Alexandr Shadchin Index: mouse.c === RCS file: /cvs/src/sbin/wsconsctl/mouse.c,v retrieving revision 1.11 diff -u -p -r1.11 mouse.c -

Campamentos Cientificos en Tucuman

2012-07-13 Thread institucional
OBSERVATORIO ASTRONOMICO AMPIMPA TUCUMAN - ARGENTINA Declarado de Interés Educativo por el Ministerio de Educación de la Nación _ tech@openbsd.org At. Campamento Científico Internacional - SolarMax 2012 - La Ciencia en el Aula – Nuevos enfoques didácticos Organiza: Observatorio Astronómic

new pci ids

2012-07-13 Thread Kirill Bychkov
Hi. Here is the bunch of pci devices, found in some laptops and desktops I get in my hands. OK to commit? Index: pcidevs === RCS file: /cvs/src/sys/dev/pci/pcidevs,v retrieving revision 1.1651 diff -u -r1.1651 pcidevs --- pcidevs

Re: diff: missing break in switch

2012-07-13 Thread Claudio Jeker
On Fri, Jul 13, 2012 at 04:07:20PM +0200, Yasuoka Masahiko wrote: > npppd used wrong AVPs as a `calling number' because `break' in switch > case was missing. > > ok? > OK. > Index: npppd/l2tp/l2tp_call.c > === > RCS file: /cvs/src/

cosmetic netstat -W tweak

2012-07-13 Thread Stefan Sperling
"Input packets dropped" isn't really a good description of what happened when we failed to allocate a new node in the net80211 stack. This allocation doesn't happen for every packet, only for new nodes that appear in reach of the antenna. And we already mention "nodes" elsewhere in the output:

diff: missing break in switch

2012-07-13 Thread YASUOKA Masahiko
npppd used wrong AVPs as a `calling number' because `break' in switch case was missing. ok? Index: npppd/l2tp/l2tp_call.c === RCS file: /cvs/src/usr.sbin/npppd/l2tp/l2tp_call.c,v retrieving revision 1.12 diff -u -p -r1.12 l2tp_call.c

Paga $53 por Pen Drive Kingston 8gb SOLO POR HOY!!!

2012-07-13 Thread Bonus Cupon
Si no podes visualizar este mail, ingresa a: http://news1.bonuscupon.com.ar/r.html?uid=1.1y.295h.y7.3urkca5kl9

Re: Build cpu topology on amd64.

2012-07-13 Thread Christiano F. Haesbaert
On Fri, Jul 13, 2012 at 03:06:34PM +0200, Mark Kettenis wrote: > > Date: Fri, 13 Jul 2012 14:57:11 +0200 > > From: "Christiano F. Haesbaert" > > > > Ok so here is the version with #ifndef SMALL_KERNEL, the only question > > that remains is: do we keep the printf in dmesg ? or shall I take that >

Re: kbd: Use NULL instead of 0 for pointers

2012-07-13 Thread Christiano F. Haesbaert
I like these kind of diffs, some people don't, ok from me. On Fri, Jul 13, 2012 at 03:37:08PM +0600, Alexandr Shadchin wrote: > Use NULL instead of 0 for pointers > > -- > Alexandr Shadchin > > Index: kbd_wscons.c > === > RCS file:

Re: rt2560 bbp/antenna diff

2012-07-13 Thread Stefan Sperling
On Sat, Jul 07, 2012 at 08:14:34PM -0400, Brad Smith wrote: > I had posted a diff for this and a few other bug fixes here.. > > http://marc.info/?l=openbsd-tech&m=124139607313719&w=2 I've looked at the additional changes you ported from dragonfly in that diff. I don't know what impact these chang

Re: Build cpu topology on amd64.

2012-07-13 Thread Mark Kettenis
> Date: Fri, 13 Jul 2012 14:57:11 +0200 > From: "Christiano F. Haesbaert" > > Ok so here is the version with #ifndef SMALL_KERNEL, the only question > that remains is: do we keep the printf in dmesg ? or shall I take that > out ? > > I'd like to keep it so we may know if the detection is correc

Re: Build cpu topology on amd64.

2012-07-13 Thread Christiano F. Haesbaert
Ok so here is the version with #ifndef SMALL_KERNEL, the only question that remains is: do we keep the printf in dmesg ? or shall I take that out ? I'd like to keep it so we may know if the detection is correctly just by looking at sent dmesgs. Index: arch/amd64/amd64/identcpu.c

Re: ncheck_ffs DUID support

2012-07-13 Thread Jan Stary
On Jul 13 11:54:11, Alexander Hall wrote: > On 07/12/12 22:55, Alexander Hall wrote: > > On 07/11/12 23:43, Jan Stary wrote: > >> On Jul 11 19:18:21, Alexander Hall wrote: > >>> This adds DUID support to ncheck_ffs. > >>> Testers? ok? > >> > >> This indeed enables ncheck_ffs for DUIDs, > >> but bre

proto cksum madness

2012-07-13 Thread Henning Brauer
after just 15 months i got this finally working. basically moving us to the assumption that there is some cksum offload engine in each an every output path, calling the software engine very late if we figure there is none. foremost this adds this logic to the alternate output pathes like bridge, t

Re: mask support for ethernet bridge filtering

2012-07-13 Thread Henning Brauer
* sven falempin [2012-07-12 17:37]: > Inline diff even without judging on wether we'll want that at all (I'm still sceptic): > +//inline > +int > +bridge_test_ea(struct ether_addr *ea_packet, struct ether_addr *ea_rules, > +struct ether_addr *ea_mask) { > + int i; > + struct ether_addr ea_cm

I wanna share some happiness with you. Would you like some?)

2012-07-13 Thread Susanne Heineman
I am Susanne by the way. I am really excited about your answer))

Re: make acpiec _GLK aware

2012-07-13 Thread Paul Irofti
On Thu, Jul 12, 2012 at 09:56:40AM -0700, Matthew Dempsky wrote: > On Thu, Jul 12, 2012 at 9:48 AM, Paul Irofti wrote: > > + if (aml_evalname(sc->sc_acpi, sc->sc_devnode, "_GLK", 0, NULL, > > &res)) > > + sc->sc_glk = 0; > > + if (res.type != AML_OBJTYPE_INTEGER) > > +

Re: ncheck_ffs DUID support

2012-07-13 Thread Alexander Hall
On 07/12/12 22:55, Alexander Hall wrote: > On 07/11/12 23:43, Jan Stary wrote: >> On Jul 11 19:18:21, Alexander Hall wrote: >>> This adds DUID support to ncheck_ffs. >>> Testers? ok? >> >> This indeed enables ncheck_ffs for DUIDs, >> but breaks ncheck_ffs for /dev/wd0x >> >> My fstab says >> >> 5d2

Re: kbd: Use NULL instead of 0 for pointers

2012-07-13 Thread Gilles Chehade
no my area, but I like it better so ok gilles@ ;-) On Fri, Jul 13, 2012 at 03:37:08PM +0600, Alexandr Shadchin wrote: > Use NULL instead of 0 for pointers > > -- > Alexandr Shadchin > > Index: kbd_wscons.c > === > RCS file: /cvs/sr

kbd: Use NULL instead of 0 for pointers

2012-07-13 Thread Alexandr Shadchin
Use NULL instead of 0 for pointers -- Alexandr Shadchin Index: kbd_wscons.c === RCS file: /cvs/src/sbin/kbd/kbd_wscons.c,v retrieving revision 1.25 diff -u -p -r1.25 kbd_wscons.c --- kbd_wscons.c23 Jun 2008 17:41:21 -

Re: ral(4) tx/prio queue fixes (was: rt2560 tx/prio queue fixes)

2012-07-13 Thread Stefan Sperling
On Tue, Jul 10, 2012 at 05:56:29PM +0200, Stefan Sperling wrote: > Updated version that includes similar fixes for the rt2661 variants. > > This seems to help soekris-based ral APs that get stuck with the OACTIVE > flag set (see the flags line in "ifconfig ral0" output when the AP stops > respondi

Re: make net80211_input be nice to soekris

2012-07-13 Thread Stefan Sperling
On Fri, Jul 13, 2012 at 03:16:27AM -0600, Theo de Raadt wrote: > + printf(msg); > > This really should be > > printf("%s", msg); > > To avoid format string problems. > > Yes, you say you are completely in control of the string however > someone could reuse this workq handler f

Re: make net80211_input be nice to soekris

2012-07-13 Thread Theo de Raadt
+ printf(msg); This really should be printf("%s", msg); To avoid format string problems. Yes, you say you are completely in control of the string however someone could reuse this workq handler for some other purpose later.

Re: make net80211_input be nice to soekris

2012-07-13 Thread Stefan Sperling
On Fri, Jul 13, 2012 at 10:54:23AM +0200, Mark Kettenis wrote: > But doing the printf from spltty() doesn't make an awful lot of sense. > We already protect printf(9) by a mutex at IPL_HIGH. Correct, I hadn't seen that, thanks! > Also... > > +/* Work queue task that prints a received frame. A

improve hostap in noisy environments

2012-07-13 Thread Stefan Sperling
We've got lots of RF noise in the g2k12 hackroom. Edd's soekris sometimes failed to allow new nodes to associate and we found that this was due to ieee80211_node_alloc_helper() failing to add a new node to the cache. 'netstat -W ral0' showed the 'input packets dropped' counter, which corresponds to

Re: make net80211_input be nice to soekris

2012-07-13 Thread Mark Kettenis
> Date: Fri, 13 Jul 2012 10:42:02 +0200 > From: Stefan Sperling > > On Thu, Jul 12, 2012 at 05:51:32PM +0200, Stefan Sperling wrote: > > Running "ifconfig ral0 debug down up" can leave slow systems, such > > as edd@'s soekris, with an unusable wireless interface until reboot. > > > > The net80

Re: make net80211_input be nice to soekris

2012-07-13 Thread Stefan Sperling
On Thu, Jul 12, 2012 at 05:51:32PM +0200, Stefan Sperling wrote: > Running "ifconfig ral0 debug down up" can leave slow systems, such > as edd@'s soekris, with an unusable wireless interface until reboot. > > The net80211 layer will run a scan when the interface comes up. > The scan hops from ch