acx100-devel
You can subscribe to this list here.
| 2003 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(9) |
Aug
(16) |
Sep
(11) |
Oct
(113) |
Nov
(136) |
Dec
(123) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2004 |
Jan
(61) |
Feb
(42) |
Mar
(114) |
Apr
(34) |
May
(5) |
Jun
(26) |
Jul
(4) |
Aug
(29) |
Sep
(62) |
Oct
(24) |
Nov
(3) |
Dec
(9) |
| 2005 |
Jan
(23) |
Feb
(8) |
Mar
(36) |
Apr
(75) |
May
(10) |
Jun
(109) |
Jul
(54) |
Aug
(135) |
Sep
(163) |
Oct
(83) |
Nov
(32) |
Dec
(44) |
| 2006 |
Jan
(134) |
Feb
(78) |
Mar
(27) |
Apr
(10) |
May
(16) |
Jun
(5) |
Jul
(6) |
Aug
(3) |
Sep
(6) |
Oct
(2) |
Nov
|
Dec
(9) |
| 2007 |
Jan
(8) |
Feb
(4) |
Mar
(2) |
Apr
|
May
(7) |
Jun
(14) |
Jul
(16) |
Aug
(4) |
Sep
(2) |
Oct
(1) |
Nov
(2) |
Dec
(1) |
| 2008 |
Jan
|
Feb
|
Mar
(3) |
Apr
|
May
(27) |
Jun
(2) |
Jul
(29) |
Aug
(5) |
Sep
|
Oct
(2) |
Nov
(3) |
Dec
(5) |
| 2009 |
Jan
(1) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(3) |
Nov
|
Dec
|
| 2010 |
Jan
(9) |
Feb
|
Mar
(7) |
Apr
(5) |
May
(2) |
Jun
|
Jul
|
Aug
(1) |
Sep
(3) |
Oct
|
Nov
|
Dec
|
| 2012 |
Jan
|
Feb
|
Mar
|
Apr
(152) |
May
(248) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2013 |
Jan
(8) |
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(3) |
Oct
|
Nov
|
Dec
|
| 2014 |
Jan
(2) |
Feb
(3) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2016 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(4) |
Nov
|
Dec
|
| S | M | T | W | T | F | S |
|---|---|---|---|---|---|---|
|
|
|
|
|
|
|
1
|
|
2
|
3
(1) |
4
|
5
|
6
(1) |
7
(1) |
8
(1) |
|
9
(1) |
10
|
11
|
12
|
13
|
14
|
15
|
|
16
|
17
|
18
|
19
|
20
|
21
(1) |
22
(7) |
|
23
(3) |
24
(3) |
25
(2) |
26
|
27
|
28
(1) |
29
|
|
30
|
31
(1) |
|
|
|
|
|
|
From: Luis Padilla Visdomine <padilla@ga...> - 2005-01-31 21:37:12
|
Hi there, I'm new and I have a question/comment/request:
I have a D-Link DWL-G650+ cardbus card which have an ACX111 chipset and
I've been using its monitor mode extensively. It works fine but at first, when
I was using debug=0x01, I noticed many IRQ interrupts for IV_ICV_FAILURE but I
didn't pay attention, I thought it was somehow normal. Then I realized that
I was receiving very few WEP packets, I saw lots of acknowledge packets but
almost no wep packets, in a proportion of 1 to 1000 or even worse. I though
the reason was a poor signal level, which might be enough for short low speed
packets (like beacons and ack), but not for long full speed packets, which are
subject to a higher error probability.
After many trials and a deep insight of the driver I came to the conclusion
that (I know this will already obvious for you, experts ;-) the ACX111 chipset
is able to encrypt/decrypt WEP packets by itself, which is in principle a very
nice feature, because it simplifies wep management in the driver and frees
computer resources. The problem is that sometimes we do want to see the wep
packets unmodified. Reading the source code of the driver I saw that the ACX100
has a bit of the WEPOption byte of the scruct acx100_ie_wep_options to override
wep decryption by the chipset (monitor_settings), fantastic... if it would work
for the ACX111, but it doesn't.
Desperately I modified the driver to set all bits in the WEPOption byte just
in case they moved the bit to another position in the acx111 chipset. No success.
Another desperate try was to force the read out of a packet after an IV_ICV_FAILURE
IRQ, but no success, always buffer empty. I even modified the rx_config_1 and
2 shorts to set all bits, even those not documented in acx100.h just in case any
of them implemented the passthrough of wep decryption failed packets. Again no
success. If a don't initialize the key I get an IRQ of type KEY_NOT_FOUND for each
received wep packet and if I initialize it then I get IRQs of type IV_ICV_FAILURE,
but no way to receive the original wep packets unmodified, by the way, there is a
lot of curiosity out there about both kinds of messages as seen in the forum, most
of the questions are unanswered. The few wep packets that I'm able to see I think
are due to random transmission errors, which modifies the FC field, unsetting the
wep bit (among who knows what other things), so the acx111 let it pass through.
None of them has the wep bit enabled but they have a clear wep structure, with its
IV + key id and IV increasing progressively.
Now my question is: does anybody know how to override the wep decrypt in the
acx111 chipset? Is it possible at all? If not, I'm afraid I'll have to look for
another wireless card... Have anybody got the programmer's manual of the acx111
chipset? Alternatively, has anyone disassembled the firmware? because surely one
might at least disable wep decryption changing the firmware, not clean but if it
works it would be enough for me. I don't even know what processor is the firmware
written for. I suspect someone has the source code of the firmware because there
are some comments in the driver like this:
/*--- Controller Commands ----------------------------------------------------*/
/* can be found in table cmdTable in firmware "Rev. 1.5.0" (FW150) */
On the other hand I like very much this card when used with this driver
because it supports very nice features. For example this card is able to send
packets while in monitor mode, in fact I modified the driver to allow it to send
raw packets while in monitor mode. I also developed a little utility to send
the packets through the driver. However I will not release the modifications
until they are well tested, but for that I need to disable the f****** automatic
wep decryption. By the way I'm working on the snapshot version 43 because I noticed
that version 44 seems to hang after a few hours in monitor mode (you have to reload
the driver). Maybe it was just a coincidence, it happened a couple of times and I
switched back to version 43. Now I see version 45 is out, so I will give it a try
provided I can fix the wep decryption problem.
As you can see I like this card and this driver, and I'm willing to help
developing this driver, particularly the wep support that seems to be long awaited
by many people (including me ;) but I need technical information about the acx111
chipset.
Sorry for this long email, I hope it is of interest to you and we can
help each other.
Regards.
Luis.
|
|
From: Markus Kolb <acx100-mkolb@to...> - 2005-01-28 22:21:36
|
Markus Kolb wrote on Mon, Jan 24, 2005 at 17:42:36 +0100: > Denis Vlasenko wrote on Sun, Jan 23, 2005 at 23:30:52 +0200: > > On Sunday 23 January 2005 21:56, Markus Kolb wrote: > [...] > > > > > > SysRq : Show Regs > > > > > > Pid: 1129, comm: modprobe.moduti > > > EIP: 0010:[<c010a4a0>] CPU: 0 EFLAGS: 00000246 Tainted: P > > ^^^^^^^^^^ > > What proprietary module do you use? > > It is quota v2. > > [...] > > Either enable kallsyms (if you run 2.6 kernels) > > or run this through ksymoops after reboot. > > Pid: 3323, comm: modprobe.moduti > EIP: 0010:[<c010a4a0>] CPU: 0 EFLAGS: 00000246 Tainted: P > Using defaults from ksymoops -t elf32-i386 -a i386 > EAX: 04000000 EBX: c4b2a380 ECX: 0000000b EDX: 00000018 > ESI: 00000001 EDI: c1643e30 EBP: 0000000b DS: 0018 ES: 0018 > Warning (Oops_set_regs): garbage 'DS: 0018 ES: 0018' at end of register > line ign > ored > CR0: 8005003b CR2: 40018000 CR3: 04a33000 CR4: 000006d0 > Call Trace: [<c010a664>] [<c010caa8>] [<c010a4a0>] [<c010a664>] > [<c010caa8>] > [<c01205fe>] [<c010a69a>] [<c010caa8>] [<c89c32e0>] [<c89c0018>] > [<c89c18eb>] > [<c89b020d>] [<c011e02a>] [<c011d38f>] [<c0108c0b>] > Warning (Oops_read): Code line not seen, dumping what data is available > > > >>EIP; c010a4a0 <handle_IRQ_event+30/70> <===== > > >>EBX; c4b2a380 <_end+482d880/8517560> > >>EDI; c1643e30 <_end+1347330/8517560> > > Trace; c010a664 <do_IRQ+64/a0> > Trace; c010caa8 <call_do_IRQ+5/d> > Trace; c010a4a0 <handle_IRQ_event+30/70> > Trace; c010a664 <do_IRQ+64/a0> > Trace; c010caa8 <call_do_IRQ+5/d> > Trace; c01205fe <do_softirq+3e/a0> > Trace; c010a69a <do_IRQ+9a/a0> > Trace; c010caa8 <call_do_IRQ+5/d> > Trace; c89c32e0 <[acx_pci]IO_ACX111+0/38> > Trace; c89c0018 <[acx_pci]acx100_create_rx_host_desc_queue+d4/272> > Trace; c89c18eb <[acx_pci]acx_issue_cmd+23f/44f> > Trace; c89b020d <[acx_pci]acx_cleanup_module+49/100> > Trace; c011e02a <free_module+aa/c0> > Trace; c011d38f <sys_delete_module+9f/1d0> > Trace; c0108c0b <system_call+33/38> Do you get this mail? Do you have an idea where the problem can be? Markus |
|
From: Andreas Mohr <andim2@us...> - 2005-01-25 19:31:20
|
Hi, On Tue, Jan 25, 2005 at 05:31:24PM +0100, Minetti Valerio wrote: > ok, i'm looking at the code.. for now i've figured out that > > First step: set get parameter. > I have to write 2 functions from the scratch and put them here: > (taken from acx100_ioctl.c) > > (iw_handler) NULL /* acx_ioctl_set_frag FIXME */, /* SIOCSIWFRAG */ > (iw_handler) NULL /* acx_ioctl_get_frag FIXME */, /* SIOCGIWFRAG */ > > they should read and write the frag threshold attribute (described in wlandevice.h ? ). Yup, easy, many other functions do the (almost) same thing... > Second step: use of this parameter. > a) (TX) when the data is transmitted i have to split it in chunks of thresohold size modifying the field that holds the fragment sequence number and looping the transmission while all fragments are received succesfully (ap should reply an ACK) Yup, something like that. > b) (RX) recognize that a packet is part of a fragment burst, receive, reorder and assembly of the package. (are there chanches that packets are received out of order?) Yup, same here. Note that this may have been a large source of problems, since our driver doesn't support Rx defrag. > is it correct? AFAICS yes (but I'm far from knowledgeable on this area). Andreas Mohr |
|
From: Minetti Valerio <valerio.minetti@ml...> - 2005-01-25 16:32:27
|
ok, i'm looking at the code.. for now i've figured out that First step: set get parameter. I have to write 2 functions from the scratch and put them here:=20 (taken from acx100_ioctl.c) (iw_handler) NULL /* acx_ioctl_set_frag FIXME */, /* SIOCSIWFRAG */ (iw_handler) NULL /* acx_ioctl_get_frag FIXME */, /* SIOCGIWFRAG */ they should read and write the frag threshold attribute (described in = wlandevice.h ? ). Second step: use of this parameter.=20 a) (TX) when the data is transmitted i have to split it in chunks of = thresohold size modifying the field that holds the fragment sequence = number and looping the transmission while all fragments are received = succesfully (ap should reply an ACK) b) (RX) recognize that a packet is part of a fragment burst, receive, = reorder and assembly of the package. (are there chanches that packets = are received out of order?) is it correct? |
|
From: Markus Kolb <acx100-mkolb@to...> - 2005-01-24 16:43:17
|
Denis Vlasenko wrote on Sun, Jan 23, 2005 at 23:30:52 +0200: > On Sunday 23 January 2005 21:56, Markus Kolb wrote: [...] > > > > SysRq : Show Regs > > > > Pid: 1129, comm: modprobe.moduti > > EIP: 0010:[<c010a4a0>] CPU: 0 EFLAGS: 00000246 Tainted: P > ^^^^^^^^^^ > What proprietary module do you use? It is quota v2. [...] > Either enable kallsyms (if you run 2.6 kernels) > or run this through ksymoops after reboot. Pid: 3323, comm: modprobe.moduti EIP: 0010:[<c010a4a0>] CPU: 0 EFLAGS: 00000246 Tainted: P Using defaults from ksymoops -t elf32-i386 -a i386 EAX: 04000000 EBX: c4b2a380 ECX: 0000000b EDX: 00000018 ESI: 00000001 EDI: c1643e30 EBP: 0000000b DS: 0018 ES: 0018 Warning (Oops_set_regs): garbage 'DS: 0018 ES: 0018' at end of register line ign ored CR0: 8005003b CR2: 40018000 CR3: 04a33000 CR4: 000006d0 Call Trace: [<c010a664>] [<c010caa8>] [<c010a4a0>] [<c010a664>] [<c010caa8>] [<c01205fe>] [<c010a69a>] [<c010caa8>] [<c89c32e0>] [<c89c0018>] [<c89c18eb>] [<c89b020d>] [<c011e02a>] [<c011d38f>] [<c0108c0b>] Warning (Oops_read): Code line not seen, dumping what data is available >>EIP; c010a4a0 <handle_IRQ_event+30/70> <===== >>EBX; c4b2a380 <_end+482d880/8517560> >>EDI; c1643e30 <_end+1347330/8517560> Trace; c010a664 <do_IRQ+64/a0> Trace; c010caa8 <call_do_IRQ+5/d> Trace; c010a4a0 <handle_IRQ_event+30/70> Trace; c010a664 <do_IRQ+64/a0> Trace; c010caa8 <call_do_IRQ+5/d> Trace; c01205fe <do_softirq+3e/a0> Trace; c010a69a <do_IRQ+9a/a0> Trace; c010caa8 <call_do_IRQ+5/d> Trace; c89c32e0 <[acx_pci]IO_ACX111+0/38> Trace; c89c0018 <[acx_pci]acx100_create_rx_host_desc_queue+d4/272> Trace; c89c18eb <[acx_pci]acx_issue_cmd+23f/44f> Trace; c89b020d <[acx_pci]acx_cleanup_module+49/100> Trace; c011e02a <free_module+aa/c0> Trace; c011d38f <sys_delete_module+9f/1d0> Trace; c0108c0b <system_call+33/38> |
|
From: Andreas Mohr <andim2@us...> - 2005-01-24 12:49:56
|
Hi, On Mon, Jan 24, 2005 at 01:17:26PM +0100, Minetti Valerio wrote: > I'm using a sitecom wl-120 and 2.6.9 kernel. when i plug the card in my laptop i get this messager logged on dmesg: > > > get_mask 0x00004d82, set_mask 0x00000000 > Don't know how to get sensitivity for radio type 0x16, please try to add that! > Got sensitivity value 0 > Got antenna value 0x4A > acx_issue_cmd FAILED: Unknown Command [4999800 uSec] Cmd: 1h, Result: 2h > Got Energy Detect (ED) threshold 0 > acx_issue_cmd FAILED: Unknown Command [4998800 uSec] Cmd: 1h, Result: 2h > Got Channel Clear Assessment (CCA) value 0 Ah, that's the message that has been reported few days ago already. I'm certain it's due to the change in acx_set_defaults() in pf44. It shouldn't matter much (in fact it's simply done in order to "initialize" a potentially uninitialized firmware value by retrieving it). I will probably change it again, since it doesn't seem to make much sense anyway (as you can see by the failure here ;). > The card anyway seems to work fine except for the famed tx eccessive retries (and consequend recalibration) problem. Which I also just found out one of the reasons for... The ACX111 recalibration I configure on Tx timeout was a "manual" one, not one for repeated "automatic" recalibration by firmware, which is why this might have caused a deadlock due to the firmware not recalibrating and the driver possibly being asleep at the wheel in case of card Tx issues... The best way probably is to do one manual recalib on Tx timeout and then immediately configure the card to do automatic recalibration with a 30s(?) interval (the firmware does NOT seem to do an initial recalibration immediately upon configuring an automatic recalibration interval, as I observed from one particular lockup behaviour yesterday, that's why I need to do a manual recalib first to make sure I get a recalib as soon as I had that Tx timeout, and not only 30 seconds later!). God, this radio drift stuff is absolutely horrible. It has been taking many, many months to figure out good ways to provide suitable service which takes into account both card overheating due to prolonged operation after recalib *and* making sure to have tolerable service due to those necessary radio recalib pauses. And that's ACX100 only. ACX111 is still largely unsolved since I wasn't using it that often (until I was FORCED to use it last weekend, that is, since my ACX100 doesn't write its firmware any more for some reason - hopefully until the next reboot in several days only!). And if a reboot fixes that, then I'll have to look into why my driver isn't able to fix the card setup for proper firmware upload itself... (possibly some PCI configuration breakage due to excessive suspend/resume...) > BTW: i have noticed that support for fragmentation threshold is missing and i'd need it really bad, is it scheduled to be implemented soon? if not i could try to work on it maybe looking at rts threshold code.. RTS threshold is an *entirely* different thing, since it's just a matter of telling the firmware from which packet size threshold to start sending the RTS/CTS sequence, whereas with frag threshold the driver has to do the packet splitting logic itself. I.e. you have to check whether to do fragmentation (which is being configured in acx100_ioctl.c) and in this case send packet fragments only and modify the DESC_CTL_FIRSTFRAG flag in the Tx descr accordingly. One would have to do similar things on Rx, I guess (maybe this is a MAJOR problem of our driver in case of fragmentation on the peer side, since we don't implement that yet?????). Feel free to ask me about that. BTW, I could also make use of that, since my signal is buried in the basement and thus is very weak, so I actually glanced over it last weekend ;) > Ty for the nice work You're welcome. Andreas Mohr |
|
From: Minetti Valerio <valerio.minetti@ml...> - 2005-01-24 12:18:27
|
I'm using a sitecom wl-120 and 2.6.9 kernel. when i plug the card in my = laptop i get this messager logged on dmesg: get_mask 0x00004d82, set_mask 0x00000000 Don't know how to get sensitivity for radio type 0x16, please try to add = that! Got sensitivity value 0 Got antenna value 0x4A acx_issue_cmd FAILED: Unknown Command [4999800 uSec] Cmd: 1h, Result: 2h Got Energy Detect (ED) threshold 0 acx_issue_cmd FAILED: Unknown Command [4998800 uSec] Cmd: 1h, Result: 2h Got Channel Clear Assessment (CCA) value 0 Got regulatory domain 0x30 get_mask 0x00000000, set_mask 0x00000000 - after update new ratevector: 82 84 8b 0c 12 96 18 24 2c 30 48 60 6c SSID =3D STA00E3C8, len =3D 9 Beacon length:71 SSID =3D STA00E3C8, len =3D 9 hw_unavailable-- acx100: form factor 0x01 ((mini-)PCI / CardBus), radio type 0x16 = (Radia), EEPROM version 0x05. Uploaded firmware 'Rev 0.1.0.11' = (0x03010101). The card anyway seems to work fine except for the famed tx eccessive = retries (and consequend recalibration) problem. BTW: i have noticed that support for fragmentation threshold is missing = and i'd need it really bad, is it scheduled to be implemented soon? if = not i could try to work on it maybe looking at rts threshold code.. Ty for the nice work |
|
From: Denis Vlasenko <vda@po...> - 2005-01-23 21:31:06
|
On Sunday 23 January 2005 21:56, Markus Kolb wrote:
> Denis Vlasenko wrote on Sun, Jan 23, 2005 at 16:15:02 +0200:
> > On Friday 21 January 2005 22:31, Markus Kolb wrote:
> > > Hello,
> > >
> [...]
> > > I can not remove the acx_pci module without a kernel crash.
> >
> > Pleace more details on that crash, it shouldn't do that...
>
> I don't get a kernel oops.
> The last message I get is the remove of /proc/driver/acx_wlan0.
>
> Then I can only do the magic sysreq things.
>
> SysRq : Show Regs
>
> Pid: 1129, comm: modprobe.moduti
> EIP: 0010:[<c010a4a0>] CPU: 0 EFLAGS: 00000246 Tainted: P
^^^^^^^^^^
What proprietary module do you use?
Is it reproducible without it?
> EAX: 04000000 EBX: c337e320 ECX: 0000000b EDX: 00000018
> ESI: 00000001 EDI: c0e3ddbc EBP: 0000000b DS: 0018 ES: 0018
> CR0: 8005003b CR2: 40018000 CR3: 00e7f000 CR4: 000006d0
> Call Trace: [<c010a664>] [<c010caa8>] [<c010a4a0>] [<c010a664>]
> [<c010caa8>] [<c010a4a0>] [<c010a664>] [<c010caa8>] [<c01205fe>]
> [<c010a69a>] [<c010caa8>] [<c89c32e0>] [<c89c0018>] [<c89c18eb>]
> [<c89b020d>] [<c011e02a>] [<c011d38f>] [<c0108c0b>]
Either enable kallsyms (if you run 2.6 kernels)
or run this through ksymoops after reboot.
You will need to save /proc/kallsyms before crash
(see ksymoops --help for more info)
--
vda
|
|
From: Markus Kolb <acx100-mkolb@to...> - 2005-01-23 19:59:32
|
Denis Vlasenko wrote on Sun, Jan 23, 2005 at 16:15:02 +0200: > On Friday 21 January 2005 22:31, Markus Kolb wrote: > > Hello, > > [...] > > I can not remove the acx_pci module without a kernel crash. > > Pleace more details on that crash, it shouldn't do that... I don't get a kernel oops. The last message I get is the remove of /proc/driver/acx_wlan0. Then I can only do the magic sysreq things. SysRq : Show Regs Pid: 1129, comm: modprobe.moduti EIP: 0010:[<c010a4a0>] CPU: 0 EFLAGS: 00000246 Tainted: P EAX: 04000000 EBX: c337e320 ECX: 0000000b EDX: 00000018 ESI: 00000001 EDI: c0e3ddbc EBP: 0000000b DS: 0018 ES: 0018 CR0: 8005003b CR2: 40018000 CR3: 00e7f000 CR4: 000006d0 Call Trace: [<c010a664>] [<c010caa8>] [<c010a4a0>] [<c010a664>] [<c010caa8>] [<c010a4a0>] [<c010a664>] [<c010caa8>] [<c01205fe>] [<c010a69a>] [<c010caa8>] [<c89c32e0>] [<c89c0018>] [<c89c18eb>] [<c89b020d>] [<c011e02a>] [<c011d38f>] [<c0108c0b>] cu Markus |
|
From: Denis Vlasenko <vda@po...> - 2005-01-23 14:15:27
|
On Friday 21 January 2005 22:31, Markus Kolb wrote: > Hello, > > after 24h I have thousands of tx errors 0x10 in my logs. > I have a Netgear WG311 v2 PCI card and after reboot it works quite well. > The transfer rates are good. > But after a while of usage the transfer rates breaks down and these tx > errors start. > After a reboot everything is fine again for a while. > > I can not remove the acx_pci module without a kernel crash. Pleace more details on that crash, it shouldn't do that... -- vda |
|
From: Markus Kolb <acx100-mkolb@to...> - 2005-01-22 20:15:25
|
Ivor Hewitt wrote on Sat, Jan 22, 2005 at 17:04:47 +0000: > > OK and that's the most recent version of the code, and you can see it is > frequently updated as fixes are made athough there haven't been any huge > changes for some time, mainly because the code is pretty mature at the > moment. > > > So my question again ... where is development? > > > What exactly are you looking for? Are you asking how you can contribute? This > is the -devel list after all if you have problems with the code or suggested > patches, this is the right place. I thought there is some place to get the really latest code where ongoing development is done. I would test it first to see if my problems still exist and then I would try to fix my problems if I can. But ok, if there are no real changes I can use the released code, too. Markus |
|
From: Sascha Sommer <saschasommer@fr...> - 2005-01-22 18:42:21
|
On Saturday 22 January 2005 19:23, Ivor Hewitt wrote: > On Saturday 22 Jan 2005 18:06, Olivier Fauchon wrote: > > > So my question again ... where is development? > > > > Marcus 's question is right: > > > > Why don't we have anymore an up-to-date CVS ? > > Is there a special reason CVS was leaved? > > sourceforge cvs was a pain and problematic a while ago, so Andreas moved > the cvs to a private server and started releasing tar package snapshots > instead. > Is there a cvslog mailinglist available for this cvs server or can the messages be forwarded here? Sascha |
|
From: Ivor Hewitt <ivor@iv...> - 2005-01-22 18:23:49
|
On Saturday 22 Jan 2005 18:06, Olivier Fauchon wrote: > > So my question again ... where is development? > > Marcus 's question is right: > > Why don't we have anymore an up-to-date CVS ? > Is there a special reason CVS was leaved? > sourceforge cvs was a pain and problematic a while ago, so Andreas moved the cvs to a private server and started releasing tar package snapshots instead. and... if I recall rightly... back then people kept moaning that the source-code was only available from cvs and that it should be released as frequent tarballs instead. :) > mmm, maybe i missed something. > Well no-ones said what they think they've lost or what they're missing with the current arrangement. The code is available and up-to date and any problems can be discussed here. -- Ivor Hewitt. http://www.ivor.it - tech | http://www.ivor.org - hedge |
|
From: Olivier Fauchon <olivier@ai...> - 2005-01-22 18:06:05
|
> So my question again ... where is development? > Marcus 's question is right: Why don't we have anymore an up-to-date CVS ? Is there a special reason CVS was leaved? mmm, maybe i missed something. Thanks you for your work, guys! Olivier Fauchon http://www.aixmarseille.com |
|
From: Ivor Hewitt <ivor@iv...> - 2005-01-22 17:05:32
|
On Saturday 22 Jan 2005 16:52, Markus Kolb wrote: > Carlos Martin wrote on Sat, Jan 22, 2005 at 14:07:03 +0100: > > On Fri, 21 Jan 2005 21:31:06 +0100, Markus Kolb > > > > <acx100-mkolb@...> wrote: > > > Then I am suprised if there is any ongoing development. I've exported > > > the CVS source and there is no change since the latest prelease nearly > > > one year ago. > > > > > > What's up with the development? > > > > > > Please give some comments. > > > > The CVS hasn't been used for ages. As it says in the sf.net page, go > > to http://lisas.de/~andi/acx100 to get the latest. > > I am using 0.2.0pre8_plus_fixes_44. > That's a prerelease and no development tree. OK and that's the most recent version of the code, and you can see it is frequently updated as fixes are made athough there haven't been any huge changes for some time, mainly because the code is pretty mature at the moment. > So my question again ... where is development? > What exactly are you looking for? Are you asking how you can contribute? This is the -devel list after all if you have problems with the code or suggested patches, this is the right place. -- Ivor Hewitt. http://www.ivor.it - tech | http://www.ivor.org - hedge |
|
From: Markus Kolb <acx100-mkolb@to...> - 2005-01-22 16:44:46
|
Carlos Martin wrote on Sat, Jan 22, 2005 at 14:07:03 +0100: > On Fri, 21 Jan 2005 21:31:06 +0100, Markus Kolb > <acx100-mkolb@...> wrote: > > Then I am suprised if there is any ongoing development. I've exported > > the CVS source and there is no change since the latest prelease nearly > > one year ago. > > > > What's up with the development? > > > > Please give some comments. > > The CVS hasn't been used for ages. As it says in the sf.net page, go > to http://lisas.de/~andi/acx100 to get the latest. I am using 0.2.0pre8_plus_fixes_44. That's a prerelease and no development tree. So my question again ... where is development? Markus |
|
From: Carlos Martin <carlosmn@gm...> - 2005-01-22 13:07:08
|
On Fri, 21 Jan 2005 21:31:06 +0100, Markus Kolb <acx100-mkolb@...> wrote: > Then I am suprised if there is any ongoing development. I've exported > the CVS source and there is no change since the latest prelease nearly > one year ago. >=20 > What's up with the development? >=20 > Please give some comments. The CVS hasn't been used for ages. As it says in the sf.net page, go to http://lisas.de/~andi/acx100 to get the latest. --=20 Carlos Mart=EDn "I'll wager it's the most extraordinary thing to hapen round here since Queen Elizabeth's handmaid got hit by lightning and sprouted a beard" -- T. C. Boyle, "Water Music" |
|
From: Markus Kolb <acx100-mkolb@to...> - 2005-01-21 20:31:50
|
Hello, after 24h I have thousands of tx errors 0x10 in my logs. I have a Netgear WG311 v2 PCI card and after reboot it works quite well. The transfer rates are good. But after a while of usage the transfer rates breaks down and these tx errors start. After a reboot everything is fine again for a while. I can not remove the acx_pci module without a kernel crash. Shouldn't there be an automatic cell usage in ad-hoc mode without a manual iwlist scan? Then I am suprised if there is any ongoing development. I've exported the CVS source and there is no change since the latest prelease nearly one year ago. What's up with the development? Please give some comments. Thanks. Markus |
|
From: Andreas Mohr <andim2@us...> - 2005-01-09 09:09:29
|
Hi, On Thu, Jan 06, 2005 at 10:59:00PM +0100, Moritz Angermann wrote: > Hi Andreas, hi everyone. > Well you wouldn't believe it, would you?. > Well at least I did not. But I just installed > Ubuntu Linux on my Laptop (yea gentoo is just wrong for 850MHz). > And apart from the fact that it installed withought a problem. > A 'updatedb'. And later 'locate acx' returned the acx100-pci.ko. Same with SuSE. A user also reported an ACX111 "factory-installed" using acx_pci on 9.2. > PS: if you WLAN card in Windows dies (device can not be started) try > linux. It will work withought a problem. (OurACXDriver++; > WindowsACXDriver-- ) Oops. Not bad :) Andreas Mohr |
|
From: Andreas Mohr <andim2@us...> - 2005-01-08 13:44:28
|
On Sat, Jan 08, 2005 at 02:23:42AM +1100, sean dwyer wrote: > I gave up on ad-hoc mode working and got a D-Link DI-514. It's working > perfectly, only the occasional Tx retransmit panic. My only question now is, > does the driver play nice with bridging? No idea/experience, our README says: Wired-Wireless bridging? Driver would perhaps need to support WDS (whatever that is). For now, proxy ARP should do: http://www.hazard.maks.net/parprouted/ Good luck! (reports welcome) Andreas Mohr |
|
From: sean dwyer <ewe2@aa...> - 2005-01-07 15:23:59
|
I gave up on ad-hoc mode working and got a D-Link DI-514. It's working perfectly, only the occasional Tx retransmit panic. My only question now is, does the driver play nice with bridging? -- Oh My God! They Killed /etc/init! You Bastards! |
|
From: Moritz Angermann <moritz.angermann@gm...> - 2005-01-06 22:00:10
|
Hi Andreas, hi everyone. Well you wouldn't believe it, would you?. Well at least I did not. But I just installed Ubuntu Linux on my Laptop (yea gentoo is just wrong for 850MHz). And apart from the fact that it installed withought a problem. A 'updatedb'. And later 'locate acx' returned the acx100-pci.ko. So ubuntu comes bundelded with our driver! They even have the RADIO firmware files on the disk (at least they are there after installation) I'm going to try ubuntu out on my workstation ASAP. I'll put the source on a USB MMC. Just in case it doesn't work out of the box with the DLink cards. But I'd say we should credit ubuntu on our frontpage for this. What would YOU say? kindest regards, Moritz Angermann PS: if you WLAN card in Windows dies (device can not be started) try linux. It will work withought a problem. (OurACXDriver++; WindowsACXDriver-- ) |
|
From: sean dwyer <ewe2@aa...> - 2005-01-03 14:46:31
|
I have an odd situation with two DWL-520+ cards (firmware version 1.9.98) that I am testing. I'm running kernel 2.6.10, acx100-0.2.0pre8_plus_fixes_43, with WEP keys in Ad-hoc mode on both. One is timing out within less than 5 minutes, the other is stable. Running both on identical Debian packages. I've done this on two different computers and even swapped the cards over, and used external antenna. Is this a weird power/room issue? I'm going to try a completely different chipset and card to test whether it's indeed environmental issues, or that the cards are just too power-sensitive. I'll post more when I've done so. -- Oh My God! They Killed /etc/init! You Bastards! |