Re: [lwip-users] tcp_write() errors on snd_queuelen

[Andrew Foster]

Tim,

For the sake of our discussion I'm going to assume 
MAP_IntMasterDisable() does what it as advertised to do.

Moving forward, how does your project use lwIP?

More specifically:
Are you running an RTOS? If so which?

How does your application call tcp_write? I.e. from a thread or other 
context?

Can you include your lwipopts.h in this post?

Andrew

On 03/16/2011 11:56 AM, Tim Lambrix wrote:
> Andrew,
> Thanks for your help too. I will take all I can get right now.  It looks like 
> SYS_ARCH_PROTECT is defined as follows and it does hit this line of code (only 
> from the interrupt itself however):
>         sys_prot_t
>         sys_arch_protect(void)
>         {
>                 return((sys_prot_t)MAP_IntMasterDisable());
>         }
> I think this is correct?  However, the stellarisif_transmit function you 
> mention is called in three places and only one of them has the SYS_ARCH_PROTECT 
> called before it.  I don't see the calls in the interrupt itself 
> stellarisif_interrupt.
>
> Maybe I am missing something here but the problem I seem to have is the Ethernet 
> interrupt goes off while I am in the tcp_write->tcp_enqueue function and changes 
> the values of pcb->snd_queuelen. I have modified the local variable in that 
> function queuelen to not use the value read at line 195 from:
>
>         queuelen = pcb->snd_queuelen;    to
>         queuelen = 0;
>
> and line 411 from:
>
>         pcb->snd_queuelen = queuelen;    to
>         pcb->snd_queuelen += queuelen;
>
> This fixes the issue but I welcome your insight on why it should not happen in the 
> first place.  I don't see any calls close to SYS_ARCH_PROTECT in the tcp_write 
> function to prevent the interrupt from going off.  I also tried disabling the 
> Ethernet ISR for both ETH_INT_RX and ETH_INT_TX (the only two enabled) before 
> calling tcp_write and enabling both after the call.  That also works to prevent 
> the corruption of pcb->  snd_queuelen.
>
>
>  From what I have traced in code, the protection prevents the interrupt from 
> going off again while in the interrupt.  I have been trying to find where lwIP 
> gets processed outside of the interrupts if the interrupt is not suppose to 
> mess with the pbufs.  It looks to me like they can.
>
> My original post of findings is here:
> http://lists.nongnu.org/archive/html/lwip-users/2011-03/msg00069.html
>
> Tim
>
>
>
> -----Original Message-----
> From: address@hidden [mailto:address@hidden On Behalf Of Andrew Foster
> Sent: Wednesday, March 16, 2011 11:17 AM
> To: Mailing list for lwIP users
> Subject: Re: [lwip-users] tcp_write() errors on snd_queuelen
>
> Tim,
>
> Sorry for jumping in on this late in the thread. I've used the TI
> Port(1.3.2) a couple of times on different projects. Typically, I've
> used it in conjunction with FreeRTOS. One thing to be mindful of is
> making sure you have SYS_ARCH_PROTECT defined correctly.
>
> stellarisif_output and stellarisif_transmit must be ran in critical
> sections with both the PBUF structure and the ETH TX FIFO protected.
> Otherwise, the stellarisif_interrupt can come in and modify the PBUF chain.
>
> Again, I apologize if I'm stating the obvious here or things you've
> already ruled out. I wasn't able to find the original post.
>
> Andrew
> _______________________________________________
> lwip-users mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/lwip-users
>
> _______________________________________________
> lwip-users mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/lwip-users