Discussion:
[Testers needed!] WiFi drivers changes
(too old to reply)
Gleb Smirnoff
2015-05-29 13:35:35 UTC
Permalink
Hi!

As part of the "opaque ifnet project" [1], we are doing some code shake
with all IEEE802.11 (read WiFi) drivers. The drivers, that provide a parent
interface for the wlan(4) interface.

The core idea is that parent device loses its ifnet(9) structure. The
code is already complete for the stack, but only 2 drivers are converted
to new KPI: iwn(4) and bwi(4). We got 22 more drivers left. The changes
are quite mechanical, but nevertheless testing is required before committing.

So, if you run FreeBSD head and wlan(4), please sign up here as tester:

https://wiki.freebsd.org/projects/ifnet/net80211

As soon as I see testers there, I will start converting more drivers.

For those who want to review the patch, or even help with converting,
you are welcome here:

https://reviews.freebsd.org/D2655

Waiting for your feedback :)

[1] https://wiki.freebsd.org/projects/ifnet
--
Totus tuus, Glebius.
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Kevin Lo
2015-05-29 15:14:05 UTC
Permalink
Post by Gleb Smirnoff
Hi!
Hi Gleb,
Post by Gleb Smirnoff
As part of the "opaque ifnet project" [1], we are doing some code shake
with all IEEE802.11 (read WiFi) drivers. The drivers, that provide a parent
interface for the wlan(4) interface.
Thanks for putting the effort into making this happen. :)
Post by Gleb Smirnoff
The core idea is that parent device loses its ifnet(9) structure. The
code is already complete for the stack, but only 2 drivers are converted
to new KPI: iwn(4) and bwi(4). We got 22 more drivers left. The changes
are quite mechanical, but nevertheless testing is required before committing.
https://wiki.freebsd.org/projects/ifnet/net80211
As soon as I see testers there, I will start converting more drivers.
For those who want to review the patch, or even help with converting,
https://reviews.freebsd.org/D2655
Waiting for your feedback :)
I have a bunch of usb wifi dongles, I would like to help with converting.
Post by Gleb Smirnoff
[1] https://wiki.freebsd.org/projects/ifnet
--
Totus tuus, Glebius.
Kevin
Gleb Smirnoff
2015-06-01 15:37:17 UTC
Permalink
Hi!

I've converted the ath(4), probably the most complex ieee80211 driver.

The updated diff is uploaded to https://reviews.freebsd.org/D2655.

Pretty sure it will panic or fail on first try :) Nevertheless,
asking for your help. Please try to run it and report any problems
to me.
--
Totus tuus, Glebius.
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Oliver Pinter
2015-06-01 19:47:21 UTC
Permalink
Post by Gleb Smirnoff
Hi!
I've converted the ath(4), probably the most complex ieee80211 driver.
The updated diff is uploaded to https://reviews.freebsd.org/D2655.
Pretty sure it will panic or fail on first try :) Nevertheless,
asking for your help. Please try to run it and report any problems
to me.
Hi!

Do you have compile tested the code? I got this build error:

--- if_ath.o ---
/usr/src/sys/dev/ath/if_ath.c:5732:26: error: no member named 'ic_ifp'
in 'struct ieee80211com'; did you mean 'ic_dfs'?
struct ifnet *ifp = ic->ic_ifp;
^~~~~~
ic_dfs
/usr/src/sys/net80211/ieee80211_var.h:197:29: note: 'ic_dfs' declared here
struct ieee80211_dfs_state ic_dfs; /* DFS state */
^
/usr/src/sys/dev/ath/if_ath.c:5732:16: error: initializing 'struct
ifnet *' with an expression of incompatible type 'struct
ieee80211_dfs_state'
struct ifnet *ifp = ic->ic_ifp;
^
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
_______________________________________________
http://lists.freebsd.org/mailman/listinfo/freebsd-current
Oliver Pinter
2015-06-01 19:55:32 UTC
Permalink
On Mon, Jun 1, 2015 at 9:47 PM, Oliver Pinter
Post by Oliver Pinter
Post by Gleb Smirnoff
Hi!
I've converted the ath(4), probably the most complex ieee80211 driver.
The updated diff is uploaded to https://reviews.freebsd.org/D2655.
Pretty sure it will panic or fail on first try :) Nevertheless,
asking for your help. Please try to run it and report any problems
to me.
Hi!
--- if_ath.o ---
/usr/src/sys/dev/ath/if_ath.c:5732:26: error: no member named 'ic_ifp'
in 'struct ieee80211com'; did you mean 'ic_dfs'?
struct ifnet *ifp = ic->ic_ifp;
^~~~~~
ic_dfs
/usr/src/sys/net80211/ieee80211_var.h:197:29: note: 'ic_dfs' declared here
struct ieee80211_dfs_state ic_dfs; /* DFS state */
^
/usr/src/sys/dev/ath/if_ath.c:5732:16: error: initializing 'struct
ifnet *' with an expression of incompatible type 'struct
ieee80211_dfs_state'
struct ifnet *ifp = ic->ic_ifp;
^
diff --git a/sys/dev/ath/if_ath.c b/sys/dev/ath/if_ath.c
index 53763a6..b719ed4 100644
--- a/sys/dev/ath/if_ath.c
+++ b/sys/dev/ath/if_ath.c
@@ -5729,8 +5729,7 @@ ath_scan_end(struct ieee80211com *ic)
static void
ath_update_chw(struct ieee80211com *ic)
{
- struct ifnet *ifp = ic->ic_ifp;
- struct ath_softc *sc = ifp->if_softc;
+ struct ath_softc *sc = ic->ic_softc;

DPRINTF(sc, ATH_DEBUG_STATE, "%s: called\n", __func__);
ath_set_channel(ic);
diff --git a/sys/dev/ath/if_ath_tdma.c b/sys/dev/ath/if_ath_tdma.c
index fd23db1..d4c9ccd 100644
--- a/sys/dev/ath/if_ath_tdma.c
+++ b/sys/dev/ath/if_ath_tdma.c
@@ -359,7 +359,7 @@ ath_tdma_update(struct ieee80211_node *ni,
#define TU_TO_TSF(_tu) (((u_int64_t)(_tu)) << 10)
struct ieee80211vap *vap = ni->ni_vap;
struct ieee80211com *ic = ni->ni_ic;
- struct ath_softc *sc = ic->ic_ifp->if_softc;
+ struct ath_softc *sc = ic->ic_softc;
struct ath_hal *ah = sc->sc_ah;
const HAL_RATE_TABLE *rt = sc->sc_currates;
u_int64_t tsf, rstamp, nextslot, nexttbtt, nexttbtt_full;
Post by Oliver Pinter
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
_______________________________________________
http://lists.freebsd.org/mailman/listinfo/freebsd-current
Oliver Pinter
2015-06-01 20:46:33 UTC
Permalink
On Mon, Jun 1, 2015 at 9:55 PM, Oliver Pinter
Post by Oliver Pinter
On Mon, Jun 1, 2015 at 9:47 PM, Oliver Pinter
Post by Oliver Pinter
Post by Gleb Smirnoff
Hi!
I've converted the ath(4), probably the most complex ieee80211 driver.
The updated diff is uploaded to https://reviews.freebsd.org/D2655.
Pretty sure it will panic or fail on first try :) Nevertheless,
asking for your help. Please try to run it and report any problems
to me.
Hi!
--- if_ath.o ---
/usr/src/sys/dev/ath/if_ath.c:5732:26: error: no member named 'ic_ifp'
in 'struct ieee80211com'; did you mean 'ic_dfs'?
struct ifnet *ifp = ic->ic_ifp;
^~~~~~
ic_dfs
/usr/src/sys/net80211/ieee80211_var.h:197:29: note: 'ic_dfs' declared here
struct ieee80211_dfs_state ic_dfs; /* DFS state */
^
/usr/src/sys/dev/ath/if_ath.c:5732:16: error: initializing 'struct
ifnet *' with an expression of incompatible type 'struct
ieee80211_dfs_state'
struct ifnet *ifp = ic->ic_ifp;
^
diff --git a/sys/dev/ath/if_ath.c b/sys/dev/ath/if_ath.c
index 53763a6..b719ed4 100644
--- a/sys/dev/ath/if_ath.c
+++ b/sys/dev/ath/if_ath.c
@@ -5729,8 +5729,7 @@ ath_scan_end(struct ieee80211com *ic)
static void
ath_update_chw(struct ieee80211com *ic)
{
- struct ifnet *ifp = ic->ic_ifp;
- struct ath_softc *sc = ifp->if_softc;
+ struct ath_softc *sc = ic->ic_softc;
DPRINTF(sc, ATH_DEBUG_STATE, "%s: called\n", __func__);
ath_set_channel(ic);
diff --git a/sys/dev/ath/if_ath_tdma.c b/sys/dev/ath/if_ath_tdma.c
index fd23db1..d4c9ccd 100644
--- a/sys/dev/ath/if_ath_tdma.c
+++ b/sys/dev/ath/if_ath_tdma.c
@@ -359,7 +359,7 @@ ath_tdma_update(struct ieee80211_node *ni,
#define TU_TO_TSF(_tu) (((u_int64_t)(_tu)) << 10)
struct ieee80211vap *vap = ni->ni_vap;
struct ieee80211com *ic = ni->ni_ic;
- struct ath_softc *sc = ic->ic_ifp->if_softc;
+ struct ath_softc *sc = ic->ic_softc;
struct ath_hal *ah = sc->sc_ah;
const HAL_RATE_TABLE *rt = sc->sc_currates;
u_int64_t tsf, rstamp, nextslot, nexttbtt, nexttbtt_full;
And the same test against my atheros seems like working fine, both the
secondary VAP creation and destruction. I'm able to run kismet without
panic, and that seems too working fine.

In ath case I got some LOR during boot and during kismet, see the
attached dmesgs.

One confusing thing, that the underlaying devices (ath0 and iwn0) has
gone from ifconfig, and that's a little confusing, when you have
multiple pci card and try to create multiple VAP to specific device.

***@opn ~> ifconfig wlan0 scan
SSID/MESH ID BSSID CHAN RATE S:N INT CAPS
IRA f4:ec:38:e4:48:b8 1 54M -91:-96 100 EPS RSN
Fridel 38:60:77:d4:6b:81 1 54M -83:-96 100 EP RSN HTCAP WPA WME
B4 f8:1a:67:38:9d:d8 6 54M -86:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
fagi f8:d1:11:bd:f9:f6 6 54M -92:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
Koka 10:fe:ed:b5:c7:6a 11 54M -77:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
Otti-home b0:c5:54:86:16:9a 11 54M -93:-96 100 EPS HTCAP
WPA RSN WME WPS
teszt f8:1a:67:40:1d:b0 13 54M -93:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
Linksys70385 20:aa:4b:78:31:e4 2 54M -96:-96 100 EPS RSN HTCAP WME WPS
teszt2 20:e5:2a:5d:30:2a 3 54M -93:-96 100 EP HTCAP RSN WME WPS
hellooo c0:4a:00:ea:5a:ea 5 54M -74:-96 100 EPS RSN HTCAP ATH
***@opn ~> ifconfig wlan0 list ap
SSID/MESH ID BSSID CHAN RATE S:N INT CAPS
IRA f4:ec:38:e4:48:b8 1 54M -91:-96 100 EPS RSN
Fridel 38:60:77:d4:6b:81 1 54M -83:-96 100 EP RSN HTCAP WPA WME
B4 f8:1a:67:38:9d:d8 6 54M -86:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
fagi f8:d1:11:bd:f9:f6 6 54M -92:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
Koka 10:fe:ed:b5:c7:6a 11 54M -77:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
Otti-home b0:c5:54:86:16:9a 11 54M -93:-96 100 EPS HTCAP
WPA RSN WME WPS
teszt f8:1a:67:40:1d:b0 13 54M -93:-96 100 EPS RSN
HTCAP WPA WME ATH WPS
Linksys70385 20:aa:4b:78:31:e4 2 54M -96:-96 100 EPS RSN HTCAP WME WPS
teszt2 20:e5:2a:5d:30:2a 3 54M -93:-96 100 EP HTCAP RSN WME WPS
hellooo c0:4a:00:ea:5a:ea 5 54M -74:-96 100 EPS RSN HTCAP ATH
***@opn ~> ifconfig wlan0 list sta
ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG
c0:4a:00:ea:5a:ea 4 5 135M 20.0 0 7 64 EPS AEHTRS+
RSN HTCAP ATH
Post by Oliver Pinter
Post by Oliver Pinter
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
_______________________________________________
http://lists.freebsd.org/mailman/listinfo/freebsd-current
Lev Serebryakov
2015-06-01 21:00:46 UTC
Permalink
Post by Oliver Pinter
One confusing thing, that the underlaying devices (ath0 and iwn0)
has gone from ifconfig, and that's a little confusing, when you
have multiple pci card and try to create multiple VAP to specific
device.
I think, it is a bug. Some options could be configured only on
underlying devices!

- --
// Lev Serebryakov
Gleb Smirnoff
2015-06-01 21:20:31 UTC
Permalink
On Tue, Jun 02, 2015 at 12:00:46AM +0300, Lev Serebryakov wrote:
L> > One confusing thing, that the underlaying devices (ath0 and iwn0)
L> > has gone from ifconfig, and that's a little confusing, when you
L> > have multiple pci card and try to create multiple VAP to specific
L> > device.
L> I think, it is a bug. Some options could be configured only on
L> underlying devices!

What options?
--
Totus tuus, Glebius.
Lev Serebryakov
2015-06-01 21:33:13 UTC
Permalink
On Tue, Jun 02, 2015 at 12:00:46AM +0300, Lev Serebryakov wrote: L>
Post by Oliver Pinter
One confusing thing, that the underlaying devices (ath0 and
iwn0) L> > has gone from ifconfig, and that's a little confusing,
when you L> > have multiple pci card and try to create multiple VAP
to specific L> > device. L> I think, it is a bug. Some options
could be configured only on L> underlying devices! What options?
Regulatory domain, as far as I remember. But I've checked long ago,
maybe it was changed?

And media/mode?

- --
// Lev Serebryakov
Gleb Smirnoff
2015-06-01 21:41:25 UTC
Permalink
On Tue, Jun 02, 2015 at 12:33:13AM +0300, Lev Serebryakov wrote:
L> > On Tue, Jun 02, 2015 at 12:00:46AM +0300, Lev Serebryakov wrote: L>
L> > > One confusing thing, that the underlaying devices (ath0 and
L> > iwn0) L> > has gone from ifconfig, and that's a little confusing,
L> > when you L> > have multiple pci card and try to create multiple VAP
L> > to specific L> > device. L> I think, it is a bug. Some options
L> > could be configured only on L> underlying devices! What options?
L> Regulatory domain, as far as I remember. But I've checked long ago,
L> maybe it was changed?
L>
L> And media/mode?

Looking into code, I don't see any. If I missed something, then please
point me at the problem, and I will fix it, so that the command will
work on wlanX (if it doesn't already).
--
Totus tuus, Glebius.
Gleb Smirnoff
2015-06-01 21:25:16 UTC
Permalink
On Mon, Jun 01, 2015 at 10:46:33PM +0200, Oliver Pinter wrote:
O> And the same test against my atheros seems like working fine, both the
O> secondary VAP creation and destruction. I'm able to run kismet without
O> panic, and that seems too working fine.

Thanks a lot!

O> In ath case I got some LOR during boot and during kismet, see the
O> attached dmesgs.

I will look at these.

O> One confusing thing, that the underlaying devices (ath0 and iwn0) has
O> gone from ifconfig, and that's a little confusing, when you have
O> multiple pci card and try to create multiple VAP to specific device.

This is intended. The ifnet layer was making KPI more complex and
it is absolutely superfluous.

You've got a PCI device, seen in pciconf. You've get kernel device,
seen in devinfo. But you need to create wlan0 interface.

The ath0 or iwn0 were not actually interfaces. You couldn't assign
addresses to them, send packets, receive packets on them. They were
just sitting in the list, confusing people. Now people got used to
them, and their absense on list is confusing, like you said. But
you will get used again :)
--
Totus tuus, Glebius.
Lev Serebryakov
2015-06-01 20:57:34 UTC
Permalink
also, here are bunch of things like this:

DPRINTF(sc, ATH_DEBUG_ANY, "%s: if_flags %x\n",
__func__, ifp->if_flags);


- --
// Lev Serebryakov
Gleb Smirnoff
2015-06-02 22:58:31 UTC
Permalink
Oliver,

On Mon, Jun 01, 2015 at 10:46:33PM +0200, Oliver Pinter wrote:
O> In ath case I got some LOR during boot and during kismet, see the
O> attached dmesgs.

I've fixed the first one. But I believe the second LOR happens on
unmodified head as well.

Thanks a lot for your help.
--
Totus tuus, Glebius.
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Lev Serebryakov
2015-06-01 20:58:21 UTC
Permalink
Post by Gleb Smirnoff
I've converted the ath(4), probably the most complex ieee80211
driver.
The updated diff is uploaded to https://reviews.freebsd.org/D2655.
Pretty sure it will panic or fail on first try :) Nevertheless,
asking for your help. Please try to run it and report any problems
to me.
And

#ifdef IEEE80211_SUPPORT_TDMA
if (sc->sc_tdma) {
if (sc->sc_tdmaswba == 0) {
struct ieee80211com *ic =
ifp->if_l2com;
struct ieee80211vap *vap =
TAILQ_FIRST(&ic->ic_vaps);
ath_tdma_beacon_send(sc, vap);
sc->sc_tdmaswba =
vap->iv_tdma->tdma_bintval;
} else
sc->sc_tdmaswba--;
} else
#endif


- --
// Lev Serebryakov
Thomas Mueller
2015-05-30 09:24:42 UTC
Permalink
Post by Gleb Smirnoff
As part of the "opaque ifnet project" [1], we are doing some code shake
with all IEEE802.11 (read WiFi) drivers. The drivers, that provide a parent
interface for the wlan(4) interface.
The core idea is that parent device loses its ifnet(9) structure. The
code is already complete for the stack, but only 2 drivers are converted
to new KPI: iwn(4) and bwi(4). We got 22 more drivers left. The changes
are quite mechanical, but nevertheless testing is required before committing.
https://wiki.freebsd.org/projects/ifnet/net80211
As soon as I see testers there, I will start converting more drivers.
For those who want to review the patch, or even help with converting,
https://reviews.freebsd.org/D2655
Waiting for your feedback :)
[1] https://wiki.freebsd.org/projects/ifnet
Totus tuus, Glebius.
There are two I could test: athn and rsu, but neither fits iwn or bwi, and athn is not in FreeBSD src at all.

athn is in NetBSD-current and 7.0_BETA, sometimes is able to connect, but usually I get
athn0: could not load firmware (35)

Always fails with OpenBSD on USB stick (liveusb-openbsd.sourceforge.net), also fails with kernels from OpenBSD 5.5, 5.6 and 5.7 as does Ethernet (re); no success with rsu either in OpenBSD.

I see on the website that sys/dev/re was tested.

I notice, since the most recent change (2015-04-09) in sys/dev/re, that this Ethernet now works for me on new computer where it didn't before:
MSI Z77 MPOWER motherboard.

Also, the latest fix (2015-01-05) to sys/dev/usb/wlan/if_rsu.c seems to make Hiro H50191 USB WLAN adapter work again.

This success is on current amd64 and i386; no such success on FreeBSD 10.1-STABLE.

But my installed ports date to last July 26; I installed lang/gcc-aux last August 17; started updating last May 12 but hit a snag on the png update reverse dependencies, now am starting over on a new GPT partition while preserving the old (current amd64).

Tom
Maxim V Filimonov
2015-06-01 18:01:01 UTC
Permalink
Post by Gleb Smirnoff
Hi!
As part of the "opaque ifnet project" [1], we are doing some code shake
with all IEEE802.11 (read WiFi) drivers. The drivers, that provide a parent
interface for the wlan(4) interface.
The core idea is that parent device loses its ifnet(9) structure. The
code is already complete for the stack, but only 2 drivers are converted
to new KPI: iwn(4) and bwi(4). We got 22 more drivers left. The changes
are quite mechanical, but nevertheless testing is required before committing.
https://wiki.freebsd.org/projects/ifnet/net80211
Did I get it correctly that you don't need tests for iwn(4)?
--
wbr, Maxim Filimonov <***@bein.link>
Gleb Smirnoff
2015-06-01 21:18:39 UTC
Permalink
On Mon, Jun 01, 2015 at 09:01:01PM +0300, Maxim V Filimonov wrote:
M> > As part of the "opaque ifnet project" [1], we are doing some code shake
M> > with all IEEE802.11 (read WiFi) drivers. The drivers, that provide a parent
M> > interface for the wlan(4) interface.
M> >
M> > The core idea is that parent device loses its ifnet(9) structure. The
M> > code is already complete for the stack, but only 2 drivers are converted
M> > to new KPI: iwn(4) and bwi(4). We got 22 more drivers left. The changes
M> > are quite mechanical, but nevertheless testing is required before
M> > committing.
M> >
M> > So, if you run FreeBSD head and wlan(4), please sign up here as tester:
M> >
M> > https://wiki.freebsd.org/projects/ifnet/net80211
M> >
M>
M> Did I get it correctly that you don't need tests for iwn(4)?

I've got one in my laptop and run it with the patch. However, extra
testing never hurts :)
--
Totus tuus, Glebius.
Gleb Smirnoff
2015-06-04 11:18:08 UTC
Permalink
Hi Kevin and Olivier,

you signed up as testers of ral(4). I have converted ral(4)
and uploaded new patch at:

https://reviews.freebsd.org/D2655

Please try, report and update the project page.

https://wiki.freebsd.org/projects/ifnet/net80211

Thanks a lot for your help.
--
Totus tuus, Glebius.
Gleb Smirnoff
2015-06-04 12:52:56 UTC
Permalink
Hi Kevin and Glen,

you signed up as testers of urtwn(4). I have converted urtwn(4)
and uploaded new patch at:

https://reviews.freebsd.org/D2655

Please try, report and update the project page.

https://wiki.freebsd.org/projects/ifnet/net80211

Thanks a lot for your help.
--
Totus tuus, Glebius.
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Kevin Lo
2015-06-05 03:18:42 UTC
Permalink
Post by Gleb Smirnoff
Hi Kevin and Glen,
Hi Gleb,
Post by Gleb Smirnoff
you signed up as testers of urtwn(4). I have converted urtwn(4)
https://reviews.freebsd.org/D2655
Please try, report and update the project page.
https://wiki.freebsd.org/projects/ifnet/net80211
Thanks a lot for your help.
Seems to be working fine. Tested on RTL8188CUS, RTL8192CU and RTL8188EUS.
Thanks.
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
Kevin
Glen Barber
2015-06-05 03:32:51 UTC
Permalink
Post by Kevin Lo
Post by Gleb Smirnoff
Hi Kevin and Glen,
Hi Gleb,
Post by Gleb Smirnoff
you signed up as testers of urtwn(4). I have converted urtwn(4)
https://reviews.freebsd.org/D2655
Please try, report and update the project page.
https://wiki.freebsd.org/projects/ifnet/net80211
Thanks a lot for your help.
Seems to be working fine. Tested on RTL8188CUS, RTL8192CU and RTL8188EUS.
Thanks.
Kevin,

Thank you for testing. I won't be able to test with mine until at least
tomorrow morning, but the timing scares me a bit, as I will be
traveling. I might pull Gleb aside and test while at BSDCan, so he can
see in person the success (and hopefully no failures!).

Glen
Kevin Lo
2015-06-05 03:16:06 UTC
Permalink
Post by Gleb Smirnoff
Hi Kevin and Olivier,
Hi Gleb,
Post by Gleb Smirnoff
you signed up as testers of ral(4). I have converted ral(4)
https://reviews.freebsd.org/D2655
Please try, report and update the project page.
https://wiki.freebsd.org/projects/ifnet/net80211
Thanks a lot for your help.
It works fine for me. Tested on RT5390, thanks.
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
Kevin
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Kevin Lo
2015-06-08 02:27:47 UTC
Permalink
Post by Kevin Lo
Post by Gleb Smirnoff
Hi Kevin and Olivier,
Hi Gleb,
Post by Gleb Smirnoff
you signed up as testers of ral(4). I have converted ral(4)
https://reviews.freebsd.org/D2655
Please try, report and update the project page.
https://wiki.freebsd.org/projects/ifnet/net80211
Thanks a lot for your help.
It works fine for me. Tested on RT5390, thanks.
As I mentioned in a private message, the diff below changes ifp->if_drv_flags
to sc->sc_flags, thanks.

--- D2655.diff.orig 2015-06-08 10:20:12.220844488 +0800
+++ D2655.diff 2015-06-08 10:20:38.020180596 +0800
@@ -6279,7 +6279,7 @@
RAL_LOCK_ASSERT(sc);

- KASSERT(ifp->if_drv_flags & IFF_DRV_RUNNING, ("not running"));
-+ KASSERT(ifp->if_drv_flags & RAL_RUNNING, ("not running"));
++ KASSERT(sc->sc_flags & RAL_RUNNING, ("not running"));

if (sc->sc_invalid) /* card ejected */
return;
@@ -7109,7 +7109,7 @@
RAL_LOCK_ASSERT(sc);

- KASSERT(ifp->if_drv_flags & IFF_DRV_RUNNING, ("not running"));
-+ KASSERT(ifp->if_drv_flags & RT2860_RUNNNING, ("not running"));
++ KASSERT(sc->sc_flags & RT2860_RUNNNING, ("not running"));

if (sc->sc_invalid) /* card ejected */
return;
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Gleb Smirnoff
2015-06-05 15:46:36 UTC
Permalink
Hi Kevin,
you signed up as testers of ural(4). I have converted ural(4)
and uploaded new patch at:

https://reviews.freebsd.org/D2655

Please try, report and update the project page.

https://wiki.freebsd.org/projects/ifnet/net80211

Thanks a lot for your help.
--
Totus tuus, Glebius.
Kevin Lo
2015-06-08 02:22:49 UTC
Permalink
Post by Gleb Smirnoff
Hi Kevin,
Hi Gleb,
Post by Gleb Smirnoff
you signed up as testers of ural(4). I have converted ural(4)
https://reviews.freebsd.org/D2655
Please try, report and update the project page.
https://wiki.freebsd.org/projects/ifnet/net80211
Thanks a lot for your help.
Works fine for me. Tested on Asus WL-167g.

# dmesg |grep ural
ural0: <ASUS 802.11g WLAN + Pen Drive, class 0/0, rev 2.00/0.01, addr 2> on usbus0
ural0: MAC/BBP RT2570 (rev 0x03), RF RT2526

# netstat -hI wlan0 1
input wlan0 output
packets errs idrops bytes packets errs bytes colls
743 0 0 1.1M 370 0 38K 0
436 0 0 642K 218 3 22K 0
650 0 0 958K 325 0 33K 0
618 0 0 900K 305 2 31K 0
378 0 0 553K 188 1 19K 0
361 0 0 526K 179 1 18K 0
405 0 0 591K 200 0 20K 0
406 0 0 588K 200 0 20K 0
412 0 0 597K 202 0 21K 0
389 0 0 562K 191 0 20K 0
389 0 0 566K 192 2 20K 0
405 0 0 589K 200 0 20K 0
288 0 0 415K 141 0 14K 0
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
Kevin
Gleb Smirnoff
2015-06-13 15:00:20 UTC
Permalink
Hi Kevin,

you signed up as tester of run(4). I have converted run(4)
and uploaded new patch at:

https://reviews.freebsd.org/D2655

Please try, report and update the project page.

https://wiki.freebsd.org/projects/ifnet/net80211

Thanks a lot for your help.
--
Totus tuus, Glebius.
Kevin Lo
2015-06-15 01:52:29 UTC
Permalink
Post by Gleb Smirnoff
Hi Kevin,
Hi Gleb,
Post by Gleb Smirnoff
you signed up as tester of run(4). I have converted run(4)
https://reviews.freebsd.org/D2655
Please try, report and update the project page.
https://wiki.freebsd.org/projects/ifnet/net80211
Thanks a lot for your help.
run(4) loses statistics after conversion.

# netstat -hI wlan0 1
input wlan0 output
packets errs idrops bytes packets errs bytes colls
79 0 0 91K 0 2 0 0
85 0 0 121K 0 1 0 0
92 0 0 133K 0 3 0 0
60 0 0 83K 0 1 0 0
32 0 0 46K 0 0 0 0
32 0 0 37K 0 1 0 0
29 0 0 32K 0 0 0 0
21 0 0 31K 0 0 0 0
22 0 0 25K 0 0 0 0
31 0 0 36K 0 0 0 0
83 0 0 107K 0 0 0 0
114 0 0 164K 0 0 0 0
62 0 0 89K 0 0 0 0
22 0 0 33K 0 1 0 0
16 0 0 24K 0 0 0 0
19 0 0 25K 0 0 0 0
22 0 0 28K 0 0 0 0
26 0 0 34K 0 0 0 0
31 0 0 35K 0 0 0 0
91 0 0 123K 0 0 0 0
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
Kevin
Gleb Smirnoff
2015-06-15 06:26:10 UTC
Permalink
On Mon, Jun 15, 2015 at 09:52:29AM +0800, Kevin Lo wrote:
K> > you signed up as tester of run(4). I have converted run(4)
K> > and uploaded new patch at:
K> >
K> > https://reviews.freebsd.org/D2655
K> >
K> > Please try, report and update the project page.
K> >
K> > https://wiki.freebsd.org/projects/ifnet/net80211
K> >
K> > Thanks a lot for your help.
K>
K> run(4) loses statistics after conversion.

I have uplodaded corrected patch. Please, test it.
--
Totus tuus, Glebius.
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Kevin Lo
2015-06-16 01:44:53 UTC
Permalink
Post by Gleb Smirnoff
K> > you signed up as tester of run(4). I have converted run(4)
K> >
K> > https://reviews.freebsd.org/D2655
K> >
K> > Please try, report and update the project page.
K> >
K> > https://wiki.freebsd.org/projects/ifnet/net80211
K> >
K> > Thanks a lot for your help.
K>
K> run(4) loses statistics after conversion.
I have uplodaded corrected patch. Please, test it.
I tested it on various chipsets and it now works fine for me, thanks.
Post by Gleb Smirnoff
--
Totus tuus, Glebius.
Kevin
_______________________________________________
freebsd-***@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-***@freebsd.org"

--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-***@muc.de
Loading...