Segmentation fault (hostapd)

[Nazeer Khan]

Hi,

> On Mon, Mar 05, 2007 at 09:57:17PM +1100, Nazeer Khan wrote:
>
>> I am using hostapd-0.5.7 with madwifi-ng. I am getting segmentation
>> fault at
>>
>> HOSTAPD_DEBUG(HOSTAPD_DEBUG_VERBOSE,
>> 		"%s: dev_up=%d\n", __func__, dev_up);
>
> There is no such line in hostapd 0.5.7 driver_madwifi.c, but there is
> one in driver_bsd.c. I've only seen madwifi-ng being used to refer to
> the Linux driver, so I think I would need bit more explanation on what
> exactly you are testing.
>

I am sorry, i was using hostapd-0.4.9. And

HOSTAPD_DEBUG(HOSTAPD_DEBUG_VERBOSE,
 		"%s: dev_up=%d\n", __func__, dev_up);

is in madwifi_set_iface_flags. I am worried tht where ever it is called, i
will get a segmentation fault. :(

>> Actually there are two deamons, both running on seperate machines and i
>> am
>> trying to control one with the other.
>
> This sounds like something that would required modifications to the
> source code. Is the modified source code available somewhere?
>
>> Can someone tell why am i getting segmentation fault here. I am getting
>> this on both machines but in different function calls.
>
> Where are you seeing the segmentation fault in the other case?
>

I modified the code myself. I am sending the config data from one PC1 to
another PC2 where it fills the structure "hapd->conf". We have OPS for
madwifi driver. The functions are called on PC1, but the ultimate action
is taken at PC2. I mean let suppose we generate the keys and in the end
want to set the keys for the driver. The keys are generated on PC1, and
the data is sent to PC2 where it calls "set80211_priv" to set the keys etc
etc. To be very precise it is like running hostapd on PC1 but the point
where hostapd needs to do IOCTL calls the required data is transfered to
PC2 in form of buffer and then given as argument to "set80211_priv" or
"set80211_param" or made an IOCTL calls independently as in case of
"madwifi_set_iface_flags" and "madwifi_get_we_version" etc etc

Anyways i got some help from google :) I did'nt understand it. Can someone
be of little help in this case ???
-----------
This release is from the development branch (0.5.x). Please note that
0.4.x branch continues to be the current source of stable releases.

hostapd:
* fixed stdarg use in hostapd_logger(): if both stdout and syslog
  logging was enabled, hostapd could trigger a segmentation fault in
  vsyslog on some CPU -- C library combinations
-------------

I tried to disable syslog by setting its value to "0" but was getting
segmentation fault again.

Thank you

Nazeer


--------------------------------------------------------------------------
This email and any attachments may be confidential. They may contain legally
privileged information or copyright material. You should not read, copy,
use or disclose them without authorisation. If you are not an intended
recipient, please contact us at once by return email and then delete both
messages. We do not accept liability in connection with computer virus,
data corruption, delay, interruption, unauthorised access or unauthorised
amendment. This notice should not be removed.