Good evening,
Is there any reason why I need to fragment the outgoing packet myself? I
guess it has something to do with the memory allocation but I can't put my
finger on it.
If I set TCP_BLOCKS_SIZE in the code below to KB(14) I receive the data I
requested on the client side, however if I set
On 01/17/2012 08:34 PM, Kieran Mansley wrote:
It's not obvious what the problem is. Which version of lwIP?
Thanks for your time! I forgot some essential info yeah, sorry about that.
I'm using LwIP v1.4.0 with a Tripple Speed Ethernet MAC driver modified
for multi-tasking. The system is a Nios
Oke, I found the problem! The problem was that MEM_SIZE was set at 16000
bytes and the TCP_SND_BUF was set to 32KB this leads to an wrong error
handling in tcp_write(). Because of this configuration tcp_write() thinks
there is 32KB of buffer free, so when it's trying to fragment the packet
19KB tot
Note to self, I'll fill a bug when http://savannah.nongnu.org is up again...
___
lwip-users mailing list
lwip-users@nongnu.org
https://lists.nongnu.org/mailman/listinfo/lwip-users
This will only fail if there is a space after de backslash. Or am I overlooking
something?
l...@piwos.de wrote:
>
>
>Hi Simon,
>
>i just noticed a problem with the GCC compiler for ARM,
>in api.h there is the line:
>
>#define netconn_write(conn, dataptr, size,
>apiflags)
> netconn_write_partl
Hi Fabi,
I verified this in LwIP 1.4.1:
Here the source included to have a better view at the problem. The one
you named in IGMP however I don't see. Because line 801 here contains
'IGMP_STATS_INC(igmp.memerr);'
Please fill in a bug report at
https://savannah.nongnu.org/bugs/?group=lwip
This is not the same variable, this macro expands to:
'++lwip_stats.igmp.memerr;'
Cheers,
Nick
On 2013-02-07 13:53, Fabian Cenedese wrote:
So igmp could still be NULL if p allocation fails and is used to
get igmp.memerr.
bye Fabi
___
lwip-users ma
Hi Thomas,
Can you create a pastebin with the code of lpc_low_level_input and what
line it crashed?
What variable is incorrect? etc. etc.
Only a posting a callstack isn't use full in any way...
Nick
On 2013-02-06 14:57, ThomasJ wrote:
Hi again
I also can add, that it seems not to crash if
On 2013-03-05 15:07, Fabian Cenedese wrote:
That's what I already tried. However if I netif_set_up the second
interface then the communication over the first interface doesn't
work anymore. I found out that the interfaces must have IP
addresses differing inside the network mask. That may be
obvio
On 2013-03-05 16:17, Fabian Cenedese wrote:
Thanks for the link, I'll have a look at it. There are several
changes,
I don't know if they're all related to 2 interfaces. On the other hand
I'd like to stay close to the original sources so an update is easier.
I'll play a bit with it.
Thanks
by
10 matches
Mail list logo
