Like he said, nothing really, since it's immediately rejected, however
if there was an interface that could accept multiple packets at once
this would create a problem.
Ali
On Aug 25, 2008, at 5:28 PM, nathan binkert wrote:
> Nice catch. Ali, do you know what the effect of this bug is? What
> happens when you try to sendTiming on the same packet twice?
>
> Nate
>
> On Mon, Aug 25, 2008 at 2:24 PM, Ali Saidi <[EMAIL PROTECTED]> wrote:
>> You're right, it's a bug. I'll commit it. Is Clint Smullen <[EMAIL PROTECTED]
>>> ok for the change log?
>>
>>
>> Thanks,
>> Ali
>>
>> On Aug 25, 2008, at 5:18 PM, Clint Smullen wrote:
>>
>>> While looking at traces, I discovered some odd behavior with the DMA
>>> ports. After successfully sending a packet, it keeps trying to send
>>> it. I determined that the problem is that even though the DmaPort
>>> pops the packet off the queue, it does not get the new front. This
>>> does not typically cause a problem, because the attempt to send the
>>> packet a second time will always fail, but it doesn't change the
>>> fact that it is a bug. It is present from 2.0b5 to the latest
>>> unstable. I have included a simple patch below.
>>>
>>> - Clint Smullen
>>>
>>> --- io_device.cc 2008-08-25 16:05:51.321310000 -0400
>>> +++ io_device.cc 2008-08-25 17:12:14.139920000 -0400
>>> @@ -190,9 +190,9 @@
>>> DmaPort::recvRetry()
>>> {
>>> assert(transmitList.size());
>>> - PacketPtr pkt = transmitList.front();
>>> bool result = true;
>>> do {
>>> + PacketPtr pkt = transmitList.front();
>>> DPRINTF(DMA, "Retry on %s addr %#x\n",
>>> pkt->cmdString(), pkt->getAddr());
>>> result = sendTiming(pkt);
>>> _______________________________________________
>>> m5-dev mailing list
>>> [email protected]
>>> http://m5sim.org/mailman/listinfo/m5-dev
>>
>> _______________________________________________
>> m5-dev mailing list
>> [email protected]
>> http://m5sim.org/mailman/listinfo/m5-dev
>>
>>
> _______________________________________________
> m5-dev mailing list
> [email protected]
> http://m5sim.org/mailman/listinfo/m5-dev
>
_______________________________________________
m5-dev mailing list
[email protected]
http://m5sim.org/mailman/listinfo/m5-dev
