[maemo-users] Memory corruption during WLAN use: detailled analysis and workaround

From: Siarhei Siamashka siarhei.siamashka at gmail.com
Date: Wed Sep 12 17:31:49 EEST 2007
On 12 September 2007, Frantisek Dufka wrote:
> Well done, thank you. This was already observed but solution was not
> found. See details here https://bugs.maemo.org/show_bug.cgi?id=677
> (comment #7). Please can you add your findings to the bug report?

Maybe it is better to submit a new bugreport with the descriptive summary and
all the details explained to keep bugzilla in a good shape? The following
bugreports (hijacked by me later) are too vague and it is not clear if their
initial reporters had the same problem or a different one, probably it is even
better to close them as they are reported against older firmware release:

On the other hand, the wifi related memory corruption problem is reproducible
and confirmed by a lot of people, so it deserves its own bugzilla entry and
can be handled individually.

No matter if Nokia is going to do anything about it or not, it is better to
have this bug recorded somewhere with all the relevant links (to cx3110x
driver sources for example), ideas and other information, so that one of the
community kernel hackers could try to fix it without the need to search and
collect vital information from various places. Too bad I did not know about
cx3110x garage project at that time and nobody cared to notify me, otherwise 
I would try doing something about this issue when I was investigating it.

Unfortunately Nokia used to open parts of previously closed code without
notifying anyone properly and then probably thinking: "look, this idiot is
complaining again about something that is already open" :)

Here is one spectacular example:

Well, what do you think a community developer would do after receiving an
answer like: 'this stuff is currently closed but we probably will consider
what can be done about it'? Would he subscribe to kernel mailing lists
and waste time tracking them every day in the hope that something gets
suddenly committed there knowing that it may just as well never happen
actually? Nokia has no responsibility to notify anyone about anything, but
it's a communication problem that hurts everyone in the end.

By the way, if the same wifi memory corruption gets confirmed on N800 too,
probably the following issues can be also related to it:

More information about the maemo-users mailing list