On 10/26/05, Bruce Barnett <[EMAIL PROTECTED]> wrote: > > > Same problem as before. The kernel message looks the same except the > > CCID messages were missing. There were many other errors sent to the > > console. Many daemons were reporting Out of Memory. I'll have to get a > > serial cable to capture these. > > Got the log. On the sender, I typed > ./ttcp-dccp -c -n5000000 -l1 -t 192.168.0.4 > after I typed the line below (using two 1.7GHz laptops)
> ttcp-dccp: page allocation failure. order:0, mode:0x20 > [<c013edbb>] __alloc_pages+0x35b/0x403 > [<c0141689>] kmem_getpages+0x2a/0x77 > [<c0142311>] cache_grow+0xab/0x155 > [<c014256b>] cache_alloc_refill+0x1b0/0x1e5 > [<c0142830>] __kmalloc+0x78/0x7a > [<c029bac3>] __alloc_skb+0x45/0x114 Thanks Bruce, just shows that the packet history not being accounted to the sock is hurting us, yesterday I started to look at how to fix this, my goal is to try and avoid two allocations per packet like we have now, dccp_sendmsg would allocate anything needed by the CCID being used, etc. - Arnaldo - To unsubscribe from this list: send the line "unsubscribe dccp" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
