RE: [lwip-users] DHCP: malloc within the rx isr

[Tom C. Barker]

Thanks Stephen,

That makes sense. It is UDP, but the thread dispatches
via ip_input. That will make it work. I have the mailbox
running and blocking tcpip_thread so all that's left is
making the netif input as a post call.

Thanks again,
Tom

-----Original Message-----
From: stephen [mailto:address@hidden
Sent: Monday, February 23, 2004 4:42 PM
To: Mailing list for lwIP users
Subject: Re: [lwip-users] DHCP: malloc within the rx isr


Mr.  Baker:

In my case, eth_input() should call tcpip_input() to send a message to wake
up tcpip thread.
Then tcpip thread will call ip_input() to handle the incoming dhcp packet.
In RxISR, we just send a message and exit.
Who actually waiting for resource is tcpip thread.
Is that correct?

Best Regards,
Stephen Chen


----- Original Message -----
From: "Tom C. Barker" <address@hidden>
To: "'Mailing list for lwIP users'" <address@hidden>
Sent: Tuesday, February 24, 2004 8:15 AM
Subject: [lwip-users] DHCP: malloc within the rx isr


> Hello,
>
> I'm finding that in a dhcp response, I get following
> procedure stack:
>
> RxIsr
> eth_input
> ip_input( pbuf*, netif* )
> udp_input( pbuf*, netif* )
> dhcp_recv
> dhcp_unfold_reply
> mem_malloc( unsigned long )
> sys_sem_wait()
> sysarch_timeouts()
>
> This essentially leads to the possibility that the isr could
> need to wait for resources in the malloc function.
>
> Is there an implementation of dhcp using the api/mailboxes?
> Or could someone suggest a way around this situation?
>
> Tom
>
>
> _______________________________________________
> lwip-users mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/lwip-users



_______________________________________________
lwip-users mailing list
address@hidden
http://mail.gnu.org/mailman/listinfo/lwip-users