Re: [Ndiswrapper-general] ACPI and battery powered

[Will Chow <williamc@CSUA.Berkeley.EDU>]

This is an acpi issue, not an ndiswrapper issue.
I have no idea what distro you are using, but the acpi=3Dforced directive i=
s
used only when you have an old BIOS, so I'm assumming your bios is out of
date. I would suggest that you upgrade your BIOS so that you don't require
the forced directive as it breaks certain things.
You may also want to try to compile a custom kernel that doesn't load up
the battery modules of ACPI and see if that works for you.

Will




On Fri, 30 Apr 2004, Bent Terp wrote:

> Hi!
> I've installed ndiswrapper on my lappy (Dell Latitude D600) and found
> that I need to boot with acpi=3Dforced in order to have acpi support as
> required by ndiswrapper.
>
> Fine, but... if I boot with acpi=3Dforced and no AC power supply, the boo=
t
> hangs itself immediately after the "starting loopback netadapter [OK]"
> message.
>
> If I've understood this ACPI business right, I need it especially when
> running batterypowered, right?
>
> What am I doing wrong?
>
> --
> with kind regards,
>         Bent Nagstrup Terp
>         Systemadministrator
>         Center for Genomics and Bioinformatics
>         Karolinska Instituttet
>         Berzelius V=E4g 35
>         171 77 Stockholm, Sweden
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: Oracle 10g
> Get certified on the hottest thing ever to hit the market... Oracle 10g.
> Take an Oracle 10g class now, and we'll give you the exam FREE.
> http://ads.osdn.com/?ad_id149&alloc_id=8166&op=CCk
> _______________________________________________
> Ndiswrapper-general mailing list
> Ndiswrapper-general@...
> https://lists.sourceforge.net/lists/listinfo/ndiswrapper-general
>

Re: [Ndiswrapper-general] 5211 atheros (ibm t40 wireless) partly working (mostly not)

[Jan Kiszka <jan.kiszka@we...>]

Brad Langhorst schrieb:

> with the release version of ndiswrapper i'm able to insert ndiswrapper
> and set the essid of the interface .
> 
> if i also specify the channel iwconfig reports the correct mac address
> for the AP
> 
> i've also tried with cvs and debugging enabled... (see the attached log)
> 
> what can i do to help resolv this?
> i've been lurking a while but haven't seen anything about this
> serialized driver issue. nothing explanitory in the archive or the faq
> either.  What is the problem with "serialized" drivers - and what does
> "serialized" mean in this context?
> 
> 

I'm also trying to get the internal Atheros-based a/b/g card of my R30 
work. A few weeks ago I developed a patch to fix the problem with the 
timers you described. It works, but then the driver chrashes as soon as 
you try to send some data. And you can still not switch it to ad-hoc 
mode (that's my personal requirement).

My patch is currently under examination by the maintainers. One of them 
have access to similar hardware now, and I'm also trying to look at this 
issue again. So we may see a solution soon - hopefully :)

Jan

Re: [Ndiswrapper-general] modprobe ndiswrapper 0.7 fails on ASUS m2400n kernel 2.6.4

[David Boersma <davidb@xs...>]

Replying to my own message: found the solution <blush>in the
FAQ...</blush>

On Fri, 30 Apr 2004, David Boersma wrote:

> Hello,
>
> I am trying to get the wireless stuff running on my ASUS m2400n laptop.
> Debian Woody with lots of backported stuff, kernel 2.6.4-plus,
> configuration from http://www.tuxmobil.org/asus_m2400n.html (scroll down
> to "Patched 2.6.4").

[...snip...]

>  [<c016ab98>] sys_ioctl+0xb2/0x22a
>  [<c010a477>] syscall_call+0x7/0xb
>
> failed to allocate shared mem
> ndiswrapper: error log: C000138D, length: 2 (00000000)
> ndiswrapper: error log: C000138D, length: 2 (00000000)
> ndiswrapper: Driver init returned error
> w70n51.sys: probe of 0000:01:05.0 failed with error -22

Item 10 in the FAQ specs that I should two lines to sysctl.conf. Did that,
gave the sysctl command but saw no result, did a reboot and then after the
modprobe ndiswrapper I saw the wlan0 finding message in the dmesg output.
Yippie.

The iwconfig command (without args) still did not find it by itself, it
just said that eth0 is not wireless and it did not mention wlan0. I just
proceeded with the other iwconfig stated in the INSTALL document. I
skipped the setting of the network parameters, since we have DHCP where I
work. It works now and I am a happy ndiswrapper user.

Cheers,
David

[Ndiswrapper-general] Trendnet TEW-421pc : No wireless extensions

[Bellegueulle Damien <d.bellegueulle@as...>]

Hello ,


No wireless extensions for wlan0

ndiswrapper: disagrees about version of symbol per_cpu__softnet_data
ndiswrapper: Unknown symbol per_cpu__softnet_data
ndiswrapper: disagrees about version of symbol netif_rx
ndiswrapper: Unknown symbol netif_rx
ndiswrapper version 0.7 loaded
ndiswrapper adding tnet1130.sys


Why ?

[Ndiswrapper-general] Trendnet TEW-421pc : probleme

[Bellegueulle Damien <d.bellegueulle@as...>]

Hello ,

After install ndiswrapper with windows Xp drivers , the system is very very 
slow and it's reboot after 2 minutes ( black screnn , hot cpu .... ) !! Why ?


Infos :

Mdk 10
Trendnet TEW-421pc 802.11g ( tiacx111 ) cardbus

Dmesg :

ndiswrapper version 0.7 loaded
ndiswrapper adding tnet1130.sys
wlan0: ndiswrapper ethernet device 00:40:f4:9e:34:68 using driver tnet1130.sys

lspci -n :

06:00.0 Class 0280: 104c:9066


Help me :)

[Ndiswrapper-general] Trendnet TEW-421pc : probleme

[Bellegueulle Damien <d.bellegueulle@as...>]

Hello ,

After install ndiswrapper with windows Xp drivers , the system is very very 
slow and it's reboot after 2 minutes ( black screnn , hot cpu .... ) !! Why ?


Infos :

Mdk 10
Trendnet TEW-421pc 802.11g ( tiacx111 ) cardbus

Dmesg :

ndiswrapper version 0.7 loaded
ndiswrapper adding tnet1130.sys
wlan0: ndiswrapper ethernet device 00:40:f4:9e:34:68 using driver tnet1130.sys

lspci -n :

06:00.0 Class 0280: 104c:9066


Help me :)

[Ndiswrapper-general] modprobe ndiswrapper 0.7 fails on ASUS m2400n kernel 2.6.4

[David Boersma <davidb@xs...>]

Hello,

I am trying to get the wireless stuff running on my ASUS m2400n laptop.
Debian Woody with lots of backported stuff, kernel 2.6.4-plus,
configuration from http://www.tuxmobil.org/asus_m2400n.html (scroll down
to "Patched 2.6.4").

After doing a make install of ndiswrapper 0.7 I installed the windows
driver. From your "Supported Chipsets" page I took driver2, installed it
with ndiswrapper -i /path/w70n51.inf. That seems to go without problem.
Then I try:

mercurius:/# ndiswrapper -l
pcilib: Cannot open /sys/bus/pci/devices
Installed ndis drivers:
w70n51  present

Maybe I should not worry about the absence of /sys. When I do modprobe
ndiswrapper I get no error messages, but with dmesg I see:

ndiswrapper version 0.7 loaded
ndiswrapper adding w70n51.sys
loadndisdriver: page allocation failure. order:0, mode:0x20
Call Trace:
 [<c014123d>] __alloc_pages+0x308/0x35e
 [<c01412b2>] __get_free_pages+0x1f/0x41
 [<c0110227>] dma_alloc_coherent+0x4b/0x81
 [<d08fb116>] NdisMAllocateSharedMemory+0x43/0x78 [ndiswrapper]
 [<d08f6415>] call_init+0xa2/0xb1 [ndiswrapper]
 [<d08f74f4>] ndis_init_one+0x2e2/0x39f [ndiswrapper]
 [<c0189c97>] sysfs_create+0x71/0x9d
 [<c01e94cf>] pci_device_probe_static+0x52/0x61
 [<c01e9519>] __pci_device_probe+0x3b/0x4e
 [<c01e9558>] pci_device_probe+0x2c/0x4a
 [<c022b16d>] bus_match+0x3f/0x6a
 [<c022b27f>] driver_attach+0x56/0x80
 [<c022b4fd>] bus_add_driver+0x8b/0x9d
 [<c022b8e1>] driver_register+0x2f/0x33
 [<d08f64b3>] call_entry+0x36/0x3d [ndiswrapper]
 [<c01e9714>] pci_register_driver+0x5c/0x84
 [<d08f7837>] start_driver+0x137/0x151 [ndiswrapper]
 [<d08f806e>] misc_ioctl+0x15d/0x291 [ndiswrapper]
 [<c0159487>] __fput+0x9f/0xf5
 [<c0156ddb>] filp_close+0x59/0x86
 [<c016ab98>] sys_ioctl+0xb2/0x22a
 [<c010a477>] syscall_call+0x7/0xb

failed to allocate shared mem
ndiswrapper: error log: C000138D, length: 2 (00000000)
ndiswrapper: error log: C000138D, length: 2 (00000000)
ndiswrapper: Driver init returned error
w70n51.sys: probe of 0000:01:05.0 failed with error -22


With the driver1 I get something similar.

Hints? Ideas? Solutions?

Thanks in advance,
David Boersma

[Ndiswrapper-general] ACPI and battery powered

[Bent Terp <Bent.Terp@cg...>]

Hi!
I've installed ndiswrapper on my lappy (Dell Latitude D600) and found
that I need to boot with acpi=3Dforced in order to have acpi support as
required by ndiswrapper.

Fine, but... if I boot with acpi=3Dforced and no AC power supply, the boo=
t
hangs itself immediately after the "starting loopback netadapter [OK]"
message.

If I've understood this ACPI business right, I need it especially when
running batterypowered, right?

What am I doing wrong?

--=20
with kind regards,
        Bent Nagstrup Terp
        Systemadministrator
        Center for Genomics and Bioinformatics
        Karolinska Instituttet
        Berzelius V=E4g 35
        171 77 Stockholm, Sweden

Re: [Ndiswrapper-general] REF Emachines M5310 Fedora core 2 test 3 Kernel 2.6.5-1.327

[Pontus Fuchs <pontus.fuchs@ta...>]

Redhat/Fedora might have the 4K stack patch in their kernel?

Pontus

On Fri, 2004-04-30 at 06:11, John Hawk wrote:
> The Kernel options changed:
> Auditing off
> 4GB Kernel Mem off
> AMD non fatal error catcher on
> Module force unload on (for test)
> REGPARAM OFF I think this is the culprit of Wep Killer
> 
> I will attempt to narow it down but at a minimum the WEP is back on
> line!
> 
> John
> 
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by: Oracle 10g
> Get certified on the hottest thing ever to hit the market... Oracle 10g. 
> Take an Oracle 10g class now, and we'll give you the exam FREE.
> http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
> _______________________________________________
> Ndiswrapper-general mailing list
> Ndiswrapper-general@...
> https://lists.sourceforge.net/lists/listinfo/ndiswrapper-general
> 
> 
> 
> 
> !DSPAM:4091d30836271768415162!
> 
> 

Re: [Ndiswrapper-general] REF Emachines M5310 Fedora core 2 test 3 Kernel 2.6.5-1.327

[Giridhar Pemmasani <giri@lm...>]

On Thu, 29 Apr 2004 23:11:25 -0500, John Hawk <john.hawk@...> said:

  John> REGPARAM OFF I think this is the culprit of Wep Killer

If that were the case, it would've broken for others too, so I doubt
that is the problem (at least, not all by itself).

-- 
Giri

[Ndiswrapper-general] REF Emachines M5310 Fedora core 2 test 3 Kernel 2.6.5-1.327

[John Hawk <john.hawk@ne...>]

The Kernel options changed:
Auditing off
4GB Kernel Mem off
AMD non fatal error catcher on
Module force unload on (for test)
REGPARAM OFF I think this is the culprit of Wep Killer

I will attempt to narow it down but at a minimum the WEP is back on
line!

John

[Ndiswrapper-general] ndiswrapper and compaq nx9110

[Ard van Breemen <ard@kw...>]

Hi,
The compaq nx9110 contains a bcm94306.
I've tried to use the debian package ndiswrapper-bcm94306, but
that did not work. It installed the dell drivers.
I've copied the bcmwl5.sys driver of the original installation
over the dell driver, and it works like that.

I will look into how to install it from cd.
Just to let people know: ndiswrapper works with the bcm94306 on
the compaq nx9110, but only with the supplied drivers.

BTW: are there anywhere up to date mailing lists archives
available? I have a lot of problems (local_bh_enable f.i.), and I
want to read before I start to nag :-).

Re: [Ndiswrapper-general] BROADCOM 4309 Emachines M5310 WEP Failure

[RJ <hellcat@hi...>]

Giridhar Pemmasani wrote:

> If you know how to do it, can you put printk's in ndis_set_wep and see
> what causes the problem?
> 

Ok tried vanilla kernel from kernel.org 2.6.5.
After some initial kernel config option mistakes, ndiswrapper works fine now.
With and Without WEP.

So it looks like the Fedora Core 2 Test 2(3) costum Kernels are just not compatible with
ndiswrapper.
First i thought it is SELinux that is enabled on Fedora Core, but i compiled
the vanilla kernel with selinux and enabled it so it isn't a problem of selinux or the policies.

But i can take a look at the sourcecode and try to find the problem this weekend.

---
RJ

Re: [Ndiswrapper-general] BROADCOM 4309 Emachines M5310 WEP Failure

[Giridhar Pemmasani <giri@lm...>]

If you know how to do it, can you put printk's in ndis_set_wep and see
what causes the problem?

-- 
Giri

Re: [Ndiswrapper-general] BROADCOM 4309 Emachines M5310 WEP Failure

[RJ <hellcat@hi...>]

Giridhar Pemmasani wrote:
> I remember someone else having problems with WEP + Fedora kernel
> (don't remember kernel version). Can you use a vanilla 2.6 kernel and
> see how that works?
> 

That would be me.
Works fine without any encryption and as soon as i set wep key i get a kernel Ooops

Will try it with a kernel from kernel.org as soon as i have some time.

Re: [Ndiswrapper-general] BROADCOM 4309 Emachines M5310 WEP Failure

[Giridhar Pemmasani <giri@lm...>]

I remember someone else having problems with WEP + Fedora kernel
(don't remember kernel version). Can you use a vanilla 2.6 kernel and
see how that works?

-- 
Giri

[Ndiswrapper-general] BROADCOM 4309 Emachines M5310 WEP Failure

[John Hawk <john.hawk@ne...>]

Has anyone been able to get a handle on the  WEP intermittent kernel
lock up (No Blue Screen) same result. I have had ndiswrapper woking fine
with wep "0.7" then I installed Fedora Core 2 Test 3 rebuilt the kernel
for the laptop and I'm back to no WEP :-(.

I think we have a kernel options problem here that goes beyond the need
for wireless support in the Kernel 2.6.5-1.327. 

I have a license for Linuxant but it does not play well with sleep at
all.

I ahve tried all the the krazy options of switching things around (i.e
iwconfig wlan0 mode managed key xxx / iwconfig wlan0 key xxx mode auto
essid mypoint etc)

Re: [Ndiswrapper-general] How to eliminate "failed to allocate DMA coherent memory" errors?

[<q3qlhh002@sn...>]

Actually I ended up buying a different wireless card (a Linksys WPC11 v.3) 
which has a different chipset (uses orinoco_cs), and it was no problem at all 
to get working. So I can't really help you with this anymore. Sorry!

Paul

On Thursday 29 April 2004 03:17, Pontus Fuchs pontus.fuchs-at-tactel.se |
ndiswrapper| wrote:
> On Wed, 2004-04-28 at 19:39, Giridhar Pemmasani wrote:
> > You can try increasing the shared memory even more. It seems you have
> > some other processes that are taking up shared memory.
>
> Giri, you got it all wrong. The name AllocateSharedMemory is actually
> pretty unfortunate because it does make one think of Linux kernel shared
> memory which is a totally different thing. AllocateSharedMemory
> allocated DMA-coherent memory which has nothing to do with shared memory
> in Linux.
>
> The problem is probably that the driver is allocate more DMA-coherent
> memory than the kernel can provide. It would be nice to know just how
> much memory the driver tried to allocate. Can you look into that Paul?
>
> Pontus
>
>
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: Oracle 10g
> Get certified on the hottest thing ever to hit the market... Oracle 10g.
> Take an Oracle 10g class now, and we'll give you the exam FREE.
> http://ads.osdn.com/?ad_id=3149&alloc_id=8166&op=click
> _______________________________________________
> Ndiswrapper-general mailing list
> Ndiswrapper-general@...
> https://lists.sourceforge.net/lists/listinfo/ndiswrapper-general

[Ndiswrapper-general] DWL-610 again!

[Julien V. <julien@sk...>]

Hi,

I have not given up, and I will not give up ;-)
I am trying to make my D-Link DWL-610 PCMCIA card work on my system
(Mandrake 10 Community with 2.6.3 kernel).

After compiling and installing yesterday's CVS version on a freshly
installed system, I do not have any problem with my keyboard (which used to
act abnormally with previous versions). I also installed the latest D-Link
driver from their website.
That is a very good thing :)

BUT (yes, there is a but), I cannot manage to make it work in managed mode.
Either, I cannot change my Ess-ID other than being in ad-hoc mode (auto does
not change anything). 
To make the tests easier, I disabled WEP encryption, and configured the
wlan0 interface manually (thus no DHCP).

Here is a kind of sum up of what I have tried, with the results:

# modprobe ndiswrapper
# tail /var/log/messages
Apr 29 17:34:28 athena kernel: ndiswrapper: no version for "struct_module"
found: kernel tainted.
Apr 29 17:34:28 athena kernel: ndiswrapper version 0.7 loaded
Apr 29 17:34:28 athena kernel: ndiswrapper adding netdlwl.sys
Apr 29 17:34:28 athena ifplugd(wlan0)[2056]: ifplugd 0.21b initializing.
Apr 29 17:34:28 athena ifplugd(wlan0)[2056]: Using interface
wlan0/00:0D:88:56:90:47
Apr 29 17:34:28 athena ifplugd(wlan0)[2056]: Using detection mode:
SIOCETHTOOL
Apr 29 17:34:28 athena ifplugd(wlan0)[2056]: Initialization complete, link
beat not detected.
Apr 29 17:34:28 athena kernel: wlan0: ndiswrapper ethernet device
00:0d:88:56:90:47 using driver netdlwl.sys

# iwconfig wlan0

wlan0     IEEE 802.11b  ESSID:""
          Mode:Auto  Frequency:2.437GHz  Access Point: FF:FF:FF:FF:FF:FF
          Bit Rate:11Mb/s   Tx-Power:20 dBm   Sensitivity=0/3
          RTS thr:2432 B   Fragment thr:2432 B
          Encryption key:off
          Power Management:off
          Link Quality:100/100  Signal level:-95 dBm  Noise level:-256 dBm
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:0   Missed beacon:0

# iwconfig wlan0 mode ad-hoc
# iwconfig wlan0 essid My-Essid

wlan0     IEEE 802.11b  ESSID:"My-Essid"
          Mode:Ad-Hoc  Frequency:2.437GHz  Cell: 02:E0:A8:9D:65:24
          Bit Rate:11Mb/s   Tx-Power:20 dBm   Sensitivity=0/3
          RTS thr:2432 B   Fragment thr:2432 B
          Encryption key:off
          Power Management:off
          Link Quality:100/100  Signal level:-95 dBm  Noise level:-256 dBm
          Rx invalid nwid:0  Rx invalid crypt:0  Rx invalid frag:0
          Tx excessive retries:0  Invalid misc:0   Missed beacon:0

I really do not know what represents this MAC address! My network is very
simple (an usb DSL modem, a firewall, an AP and a lap-top with a wi-fi
card). Of course, my AP does not list my client.

Moreover, the logs always repeat this line:
Apr 29 17:35:19 athena kernel: wlan0: wireless mode set earlier (1) is not
same as wireless mode now (2)

I have read yesterday on one French forum that this card works with
ndiswrapper, what encourages me a lot!
Here is the link: http://linuxfr.org/comments/394080.html (the commentar)
For non-French speaker, I try to sum up:
"I own a DWL-610 card and I have tried to follow the guide on
http://www.canit.se/~oli/wl/ with no success. Nothing to do, it just make
the machine hanging up.
Though, with ndiswrapper, it works! But you have to rename the config files
generated by ndiswrapper -i... "

I do not understand what this guy means! Of course I tried to contact him,
but no email address! Someone left a message yesterday. I hope he will reply
;-)

Well, I think that is enough... I really hope someone can try to help me!
Do not hesitate to ask for further details (but as I am quite a newbie in
this matter, please try and be clear :-P )

I thank you in advance for all your help and suggestions!

Cheers,
Julien

[Ndiswrapper-general] Problems with Intel PRO/Wireless LAN 2200 BG Mini PCI

[Lasse Bang Mikkelsen <lbm@fa...>]

Hi

I've just upgraded my old Intel PRO/Wireless LAN 2100 3B Mini PCI
wireless adapter with a 802.11g Intel PRO/Wireless LAN 2200 BG Mini PCI,
but I have problems!

As stated on the website, I installed the newest ndiswrapper from CVS.
The card is apparently found and in some way initialized: "wlan0:
ndiswrapper ethernet device 00:0e:35:0f:**:** using driver w22n51.sys".

But it isn't possible to set any essid. iwconfig tells me that: "wlan0:
getting configuration failed (S000...1)".

My 2100 3B didn't output much (with debug enabled), but my 2200 BG
certainly do!

The following is output again and again (with a different timer id
everytime):

hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
hangcheck_bh: Hangcheck timer
NdisMCancelTimer
wrapper_cancel_timer
NdisMCancelTimer
wrapper_cancel_timer
Canceling timer d9ae1580

Any ideas?

-- 
Regards

Lasse Bang Mikkelsen
lbm@...

[Ndiswrapper-general] Machine slow-down

[Kevin Penrose <kevin.penrose@mo...>]

Recently someone posted a message to this mailing list that concerned a
machine slow-down when 
the ndiswrapper loaded the drivers for a broadcom based card.  My reason for
interest in this is that an
earlier message that I had posted about the same symptoms was mentioned as a
reference to the fact that
other people were experiencing the same symptoms.

Does anyone know if there was any resolution to this problem?  I haven't
tried the new version (0.7) yet, as I'm
waiting to see if someone that's running an emachines notebook with the
broadcom chip set has any success, 
because it seems to be the one set-up that has foiled the ndiswrapper
solution so far.

-- Kevin

Re: [Ndiswrapper-general] How to eliminate "failed to allocate DMA coherent memory" errors?

[Giridhar Pemmasani <giri@lm...>]

Sorry, didn't look at the function implementation.

I see that in 2.6 kernels, we can use dma_alloc_coherent with
GFP_KERNEL flag. If that is okay, it may solve this problem (for 2.6
at least)? 

-- 
Giri

[Ndiswrapper-general] DWL-G520+ success

[<sxbarral@ya...>]

Hi there !

I have just installed my new shiny DWL-G520+ in a
somewhat old IBM Pentium II box (400Mhz, 256MbRAM),
with Mandrake Linux 10.0 (kernel 2.6.3). 

I have downloaded ndiswrapper 0.7 and compiled it. 
Then used WindowsXP gplus driver (the one that comes
alongside with the card) following HOWTO guidelines. 

And it works pretty well !

Thank you ndiswrapper team, you're doing a wonderful
work !

Serxio Barral.


		
______________________________________________________________________
Correo Yahoo! - 6MB, más protección contra el spam ¡Gratis!
http://correo.yahoo.es

Re: [Ndiswrapper-general] How to eliminate "failed to allocate DMA coherent memory" errors?

[Pontus Fuchs <pontus.fuchs@ta...>]

On Wed, 2004-04-28 at 19:39, Giridhar Pemmasani wrote:
> You can try increasing the shared memory even more. It seems you have
> some other processes that are taking up shared memory. 

Giri, you got it all wrong. The name AllocateSharedMemory is actually
pretty unfortunate because it does make one think of Linux kernel shared
memory which is a totally different thing. AllocateSharedMemory
allocated DMA-coherent memory which has nothing to do with shared memory
in Linux.

The problem is probably that the driver is allocate more DMA-coherent
memory than the kernel can provide. It would be nice to know just how
much memory the driver tried to allocate. Can you look into that Paul?

Pontus

[Ndiswrapper-general] Realtek 8180L Hangcheck

[Andy O'Neill <oneill@ph...>]

Hi,

I've already expressed my gratitude for the whole project in the forums,
but I'll say it again - thanks a lot!

I'm running a Rt8180 with the latest XP driver on Fedora Core 1, with
the latest Fedora kernel, lastest CVS(today) of ndiswrapper.

Under moderately heavy loads (uploading or downloading, it seems) the
card freezes, although the system stays up just fine.  I applied the
patch that was given on the mailing list last week.  Prior to applying
that patch, I would need to restart the kernel pcmcia services to get
the thing back up.  Now I can just 
ifdown wlan0
ifup wlan0

and the connection comes back up.

The patch:
--- wrapper.c   2 Apr 2004 08:10:21 -0000       1.169
+++ wrapper.c   23 Apr 2004 09:17:32 -0000
@@ -264,6 +264,7 @@ static void hangcheck_bh(void *data)
        struct ndis_handle *handle = (struct ndis_handle *)data;
        DBGTRACE("%s: Hangcheck timer\n", __FUNCTION__);
+       return;

    if(handle->driver->miniport_char.hangcheck(handle->adapter_ctx))
    {
                int res;




>From /var/log/messages:


Apr 29 01:57:25 dixon kernel: send_packets returning 00000103
Apr 29 01:57:25 dixon kernel: NdisMSendComplete 00000000
Apr 29 01:57:26 dixon kernel: hangcheck_bh: Hangcheck timer
Apr 29 01:57:29 dixon kernel: hangcheck_bh: Hangcheck timer
Apr 29 01:57:30 dixon kernel: NdisQueryBuffer: entry
Apr 29 01:57:30 dixon last message repeated 2 times
Apr 29 01:57:30 dixon kernel: send_packets returning 00000103
Apr 29 01:57:30 dixon kernel: NdisMSendComplete 00000000
Apr 29 01:57:32 dixon kernel: hangcheck_bh: Hangcheck timer
Apr 29 01:57:35 dixon kernel: NdisQueryBuffer: entry
Apr 29 01:57:35 dixon last message repeated 2 times
Apr 29 01:57:35 dixon kernel: send_packets returning 00000103
Apr 29 01:57:35 dixon kernel: NdisMSendComplete 00000000
Apr 29 01:57:35 dixon kernel: hangcheck_bh: Hangcheck timer
Apr 29 01:57:37 dixon kernel: NdisQueryBuffer: entry
Apr 29 01:57:37 dixon last message repeated 2 times
Apr 29 01:57:37 dixon kernel: send_packets returning 00000103
Apr 29 01:57:37 dixon kernel: NdisMSendComplete 00000000
Apr 29 01:57:38 dixon kernel: hangcheck_bh: Hangcheck timer
Apr 29 01:57:40 dixon kernel: NdisQueryBuffer: entry
Apr 29 01:57:40 dixon last message repeated 2 times
Apr 29 01:57:40 dixon kernel: send_packets returning 00000103
Apr 29 01:57:40 dixon kernel: NdisMSendComplete 00000000
Apr 29 01:57:41 dixon kernel: hangcheck_bh: Hangcheck timer
Apr 29 01:57:50 dixon last message repeated 3 times
Apr 29 01:57:53 dixon kernel: hangcheck_bh: Hangcheck timer

Anything else I can do?  I'm going to try reverting the patch see if it
still hangs completely doing the hangcheck reset.

Andy

-- 
|========================================================|
| Andrew C. O'Neill                |       o_            |
| http://phys.columbia.edu/~oneill |      _ >/           |
|                                  |     (_) \(_)        |
|  "Evil will always triumph,      |  How's My Driving?  |
|   because good is dumb"          |   1-212-854-6156    |
|========================================================|