Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 23 Nov 2004 12:04:55 +0800 (CST)
From:      Tai-hwa Liang <avatar@mmlab.cse.yzu.edu.tw>
To:        Sam Leffler <sam@errno.com>
Cc:        Freddie Cash <fcash@ocis.net>
Subject:   Re: wifi on 5.3, sort of OT
Message-ID:  <04112312005318.8132@www.mmlab.cse.yzu.edu.tw>
In-Reply-To: <41780961.3040808@errno.com>
References:  <415C59C7.7090408@pythonemproject.com> <200410210901.09392.fcash@ocis.net><41780961.3040808@errno.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 21 Oct 2004, Sam Leffler wrote:
> doug@polands.org wrote:
>> On Thu, Oct 21, 2004 at 09:01:09AM -0700, Freddie Cash wrote:
>>> On September 30, 2004 12:08 pm, Rob wrote:
>>>> Problem seems to be right now that my Netgear WG511T cardbus card
>>>> doesn't seem to be reconginized.
>>> 
>>> I have the same card at home.  In order to get it to work with BETA7 I had 
>>> to install a new Atheros HAL.  This can be downloaded from the 
>>> people.freebsd.org/internal/ website, although I forget the name of the 
>>> person who did the patchset.  You download the patchset, patch the source 
>>> tree, and rebuild the Atheros modules.  After that, everything worked just 
>>> tickety-boo.
>>> 
>> Have these patches made it into RC1?  If not, are they planned for
>> 5.3-RELEASE?
>> 
> I cannot commit any of the changes because other drivers need to be updated 
> and noone has stepped up to help do that.  Until then the code in the tree 
> will remain ~1 year behind my current work.
>
> 	Sam

Hi Sam,

   I'm curious about what those "other drivers" are and what kind of
update they may require to cooperate with the latest 802.11 stack?

   A short grep of 80211 on sys/dev shows following hits:

 	sys/dev/an/if_aironet_ieee.h
 	sys/dev/an/if_an.c
 	sys/dev/an/if_anreg.h
 	sys/dev/arl/if_arl.c
 	sys/dev/awi/am79c930reg.h
 	sys/dev/awi/awi.c
 	sys/dev/awi/awireg.h
 	sys/dev/awi/awivar.h
 	sys/dev/awi/if_awi_pccard.c
 	sys/dev/if_ndis/if_ndis_pccard.c
 	sys/dev/if_ndis/if_ndis.c
 	sys/dev/if_ndis/if_ndis_pci.c
 	sys/dev/if_ndis/if_ndisvar.h
 	sys/dev/owi/if_ieee80211.h
 	sys/dev/owi/if_owi.c
 	sys/dev/owi/if_owi_pccard.c
 	sys/dev/ray/if_ray.c
 	sys/dev/ray/if_raymib.h
 	sys/dev/wi/if_wavelan_ieee.h
 	sys/dev/wi/if_wi.c
 	sys/dev/wi/if_wi_pccard.c
 	sys/dev/wi/if_wi_pci.c
 	sys/dev/wi/if_wireg.h
 	sys/dev/wi/if_wivar.h

   It also appears that net80211+ath-20041015.tgz had already dealt with
the following files:

 	sys/dev/awi/awi.c
 	sys/dev/awi/awivar.h
 	sys/dev/if_ndis/if_ndis.c
 	sys/dev/if_ndis/if_ndisvar.h
 	sys/dev/ray/if_ray.c
 	sys/dev/wi/if_wi.c
 	sys/dev/wi/if_wi_pci.c
 	sys/dev/wi/if_wivar.h

   Though I don't have the real hardware to test; however, after a
quick glance at arl, owi and ray, I guess that they should just work
with the latest 802.11 stack since they didn't seem too much
depend on net80211 to complete their tasks.

   I've applied your patch and tested them with Linksys WPC55AG
pccard(if_ath), IBM High Rate Wireless LAN Mini PCI adapter(if_wi) and
Cisco Aironet Mini PCI adapter(if_an). Here are my testing result on
a Thinkpad R40:

 			Infrastructure		Adhoc(with another if_wi)

 	if_an		able to associate(1)	able to associate(1)

 	if_ath		somewhat works(2)	freeze if "too many" frames
 						were sent/received(3)

 	if_wi		works			works

(1) Despite the "an0: device timeout" starts to appear after 4 ~ 5 ping
     packets were sent, the card can associate with AP or other adhoc
     station. I'm still trying to figure out what's really going on with
     this timeout since it also happens *before* the patch was applied. The
     timeout would occupy an0 a lot of time to do the device re-init and
     thus render the card almost unusable -- taking "ping" as an example,
     there would be one "device timeout" every 4 ping packets, which raises
     the RTT to 1xxx ms per ping packet.

(2) Interactive ssh session to/from another host works okay; however, if
     there're packets burst, the connection would hang. For example, ssh
     from if_ath to another host and issue a dmesg(about 8 Kbytes of data),
     the connection will hang. Meanwhile, this box is still repsonse to
     remote ping, and is still able to ssh to another host.

(3) If I issue a ping from each other at the same time(10.0.0.2 -> 10.0.0.5
     and 10.0.0.5 -> 10.0.0.2), the system froze; that is, it didn't response
     to ping and keyboard input anymore. However, if I manually eject the
     pccard, the laptop would response again. I'm not sure whether this
     is correlated to the "an0: device timeout" aforementioned in (1) since
     according to systat -vm, there're only about 20 interrupts per second
     on cbb0 during the packet transmission.
     In addition to that, sometime if the device didn't freeze but
     timeout'ed(ath0: device timeout), the ssid will be cleared.

   BTW, Following is the patch to make device awi compile after applying
net80211+ath-20041015.patch. I would be very appreciate it if anyone who
happens to get this card to test this.

--- /sys/dev/awi/awi.c.bak	Mon Nov 22 17:54:16 2004
+++ /sys/dev/awi/awi.c	Mon Nov 22 17:54:54 2004
@@ -663,7 +663,7 @@
  		ni->ni_intval = ic->ic_lintval;
  		ni->ni_rssi = 0;
  		ni->ni_rstamp = 0;
-		memset(ni->ni_tstamp, 0, sizeof(ni->ni_tstamp));
+		ni->ni_tstamp.tsf = 0;
  		ni->ni_rates =
  		    ic->ic_sup_rates[ieee80211_chan2mode(ic, ni->ni_chan)];
  		IEEE80211_ADDR_COPY(ni->ni_macaddr, ic->ic_myaddr);
@@ -2048,14 +2048,14 @@
  				awi_write_2(sc, AWI_CA_SYNC_DWELL, 0);
  			}
  			if (ic->ic_flags & IEEE80211_F_SIBSS) {
-				memset(ni->ni_tstamp, 0, sizeof(ni->ni_tstamp));
+				ni->ni_tstamp.tsf = 0;
  				ni->ni_rstamp = 0;
  				awi_write_1(sc, AWI_CA_SYNC_STARTBSS, 1);
  			} else
  				awi_write_1(sc, AWI_CA_SYNC_STARTBSS, 0);
  			awi_write_2(sc, AWI_CA_SYNC_MBZ, 0);
  			awi_write_bytes(sc, AWI_CA_SYNC_TIMESTAMP,
-			    ni->ni_tstamp, 8);
+			    ni->ni_tstamp.data, 8);
  			awi_write_4(sc, AWI_CA_SYNC_REFTIME, ni->ni_rstamp);
  			sc->sc_cur_chan = ieee80211_chan2ieee(ic, ni->ni_chan);
  			if ((error = awi_cmd(sc, AWI_CMD_SYNC, AWI_NOWAIT))


---------- Related system information after applying the patch ----------

ath_rate: <Atsushi Onoe's rate control algorithm>
ath_hal: <Atheros Hardware Access Layer>version 0.9.12.13
found->	vendor=0x168c, dev=0x0013, revid=0x01
 	bus=3, slot=0, func=0
 	class=02-00-00, hdrtype=0x00, mfdev=0
 	cmdreg=0x0000, statreg=0x0290, cachelnsz=8 (dwords)
 	lattimer=0xa8 (5040 ns), mingnt=0x0a (2500 ns), maxlat=0x1c (7000 ns)
 	intpin=a, irq=11
 	powerspec 2  supports D0 D3  current D0
pcib2: device cardbus0 requested decoded memory range 0xd0200000-0xdfffffff
pcib2: device cardbus0 requested decoded memory range 0xd0200000-0xdfffffff
ath0: <Atheros 5212> mem 0xd0210000-0xd021ffff irq 11 at device 0.0 on cardbus0
pcib2: device ath0 requested decoded memory range 0xd0210000-0xd021ffff
ath0: [MPSAFE]
ath0: bpf attached
ath0: Ethernet address: 00:0c:41:fc:6c:54
ath0: bpf attached
ath0: bpf attached
ath0: 11a rates: 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
ath0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps
ath0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
ath0: turbo rates: 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
ath0: mac 5.6 phy 4.1 5ghz radio 3.6
ath0: Use hw queue 1 for WME_AC_BE traffic
ath0: Use hw queue 0 for WME_AC_BK traffic
ath0: Use hw queue 2 for WME_AC_VI traffic
ath0: Use hw queue 3 for WME_AC_VO traffic
pci0: driver added
found->	vendor=0x8086, dev=0x24c3, revid=0x03
 	bus=0, slot=31, func=3
 	class=0c-05-00, hdrtype=0x00, mfdev=0
 	cmdreg=0x0001, statreg=0x0280, cachelnsz=0 (dwords)
 	lattimer=0x00 (0 ns), mingnt=0x00 (0 ns), maxlat=0x00 (0 ns)
 	intpin=b, irq=11
pci0:31:3: reprobing on driver added
pci0:31:3: Transition from D0 to D3
found->	vendor=0x8086, dev=0x24c5, revid=0x03
 	bus=0, slot=31, func=5
 	class=04-01-00, hdrtype=0x00, mfdev=0
 	cmdreg=0x0007, statreg=0x0290, cachelnsz=0 (dwords)
 	lattimer=0x00 (0 ns), mingnt=0x00 (0 ns), maxlat=0x00 (0 ns)
 	intpin=b, irq=11
 	powerspec 2  supports D0 D3  current D3
pci0:31:5: reprobing on driver added
pci0:31:5: Transition from D3 to D0
pci0:31:5: Transition from D0 to D3
found->	vendor=0x8086, dev=0x24c6, revid=0x03
 	bus=0, slot=31, func=6
 	class=07-03-00, hdrtype=0x00, mfdev=0
 	cmdreg=0x0005, statreg=0x0290, cachelnsz=0 (dwords)
 	lattimer=0x00 (0 ns), mingnt=0x00 (0 ns), maxlat=0x00 (0 ns)
 	intpin=b, irq=11
 	powerspec 2  supports D0 D3  current D3
pci0:31:6: reprobing on driver added
pci0:31:6: Transition from D3 to D0
pci0:31:6: Transition from D0 to D3
pci1: driver added
pci2: driver added

interrupt                          total       rate
irq0: clk                         828213        999
irq1: atkbd0                        3126          3
irq8: rtc                         105994        127
irq9: acpi0                           97          0
irq11: cbb0 fwohci+                30616         36
irq12: psm0                           51          0
irq13: npx0                            1          0
irq14: ata0                         2637          3
irq15: ata1                           53          0
Total                             970788       1171



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?04112312005318.8132>