Marcus Bäckman wrote:

> I have encounted a collision problem in PBUF_RAM. udp.c uses
pbuf_alloc(PBUF_IP, 

> UDP_HLEN, PBUF_RAM) and this buffer will be queued for transmission in the


> ethernet driver. The reference count of this buffer is increased in the
ethernet layer, 

> and the buffer is freed at a later point when it has been transmitted.

> I have noticed that (sometimes) the pbuf allocated in udp.c points to the
same payload 

> which is used by a pbuf (also previously allocated in udp.c) waiting to be
transmitted. 

> The pbuf's differs but they point to the same payload. 

> The pbuf waiting to be transmitted has not been freed at the point udp.c
allocates 

> a new pbuf (the reference count is 2).

 

Just wanted to give you an update of my findings. It turned out to be
explained by "bug #29361: ip_frag has problems with zero-copy DMA MACs". I
am using a zero-copy approach.

 

Regards,

Marcus 

_______________________________________________
lwip-users mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/lwip-users

Reply via email to