One of my customers is reporting frequent occurrences of Bugcheck 124. Crash
dump analysis show this to be caused by GCACHEL2_ERR_ERR. At the time of the
failure, one of my PCIExpress devices is performing large, high speed DMA
transfers to system memory at the same time that the CPU is processing
previously transferred data.
Am I correct in saying that this is an underlying motherboard problem?
Is there a way of further decoding the status bits in the error descriptor?
Error : GCACHEL2_ERR_ERR (Proc 8 Bank 19)
Status : 0xbe200000000b110a
Address : 0x0000000ed5135340
Misc. : 0xacfe221088000086
Best regards
Chris Read
.
We had a similar issue with our PCIe board in combination with a particular motherboard. It would give this error about 1/1000 during shutdown. At that time, the board was performing high speed DMA to system memory. The processor core number would vary, but the memory bank (19 in your case) would always be the same.
Have a look in the Intel programmer’s manual p 2491:
http://www.intel.com/content/dam/www/public/us/en/documents/manuals/64-ia-32-architectures-software-developer-manual-325462.pdf
And at some errata:
http://www.intel.com/content/www/us/en/processors/atom/atom-z6xx-specification-update.html
p13 at the bottom
We were never able to find out the root cause of our (spurious) problem and have switched to another motherboard in the mean time.
Hope this gives you some direction on where to look.
Or you can apply for an OSR Crash dump analysis course.
Definitely worthwhile !