From mboxrd@z Thu Jan 1 00:00:00 1970 From: achandran@mvista.com (Arun Chandran) Date: Thu, 25 Jun 2015 17:50:36 +0530 Subject: BUG: Bad page state in process swapper pfn:XXXXX In-Reply-To: <558BDF49.7020908@arm.com> References: <558BDF49.7020908@arm.com> Message-ID: To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Thu, Jun 25, 2015 at 4:30 PM, Robin Murphy wrote: > Hi Arun, > > On 25/06/15 11:08, Arun Chandran wrote: > [...] > >> A u-boot memtest for those areas resulted in errors[1 in 5 boots] >> >> SOCFPGA_ARRIA5 # mtest 0x1ff3a000 0x1ff4ffff >> Pattern 00000002 Writing... Reading... >> Mem error @ 0x1FF4A714: found FFFFFFFF, expected 000041C7 >> >> Mem error @ 0x1FF4A718: found 1C00FFFF, expected 000041C8 >> >> Mem error @ 0x1FF4A71C: found 143DCB25, expected 000041C9 >> >> Mem error @ 0x1FF4A720: found 01000608, expected 000041CA >> >> Mem error @ 0x1FF4A724: found 04060008, expected 000041CB >> >> Mem error @ 0x1FF4A728: found 1C000100, expected 000041CC >> >> Mem error @ 0x1FF4A72C: found 143DCB25, expected 000041CD >> >> Mem error @ 0x1FF4A730: found 1967A20A, expected 000041CE >> >> Mem error @ 0x1FF4A734: found 00000000, expected 000041CF >> >> Mem error @ 0x1FF4A738: found A20A0000, expected 000041D0 >> >> Mem error @ 0x1FF4A73C: found 00001667, expected 000041D1 >> >> Mem error @ 0x1FF4A740: found 00000000, expected 000041D2 >> >> Mem error @ 0x1FF4A744: found 00000000, expected 000041D3 >> >> Mem error @ 0x1FF4A748: found 00000000, expected 000041D4 >> >> Mem error @ 0x1FF4A74C: found 00000000, expected 000041D5 >> >> Mem error @ 0x1FF4A750: found E3D2A784, expected 000041D6 >> Pattern 0000000A Writing... Reading... >> Mem error @ 0x1FF3AE14: found 005A0320, expected 0000038F >> Pattern FFFFFFF4 Writing... Reading... >> Mem error @ 0x1FF3AE14: found 00400320, expected FFFFFC6F >> Pattern FFFFFFEF Writing... Reading... >> >> >> That means there is problem with my RAM [or board reset]. > > > Or that your bootloader is leaving the network controller DMA'ing over > kernel memory. That 'corruption' has a suspiciously familiar shape to it, > and sure enough, turns out to be a valid ethernet packet - Wireshark had > this to say about it: > > 1 0.000000000 HonHaiPr_cb:3d:14 Broadcast ARP 64 > Who has 10.162.103.22? Tell 10.162.103.25 > > Robin. Hi Robin, Nice catch. Thanks for pointing that out. --Arun