On Thu, 2010-07-29 at 17:19 +0200, Axel Walsleben wrote: > my genTable selftest runs only on 1 GPU. > > This is the Output from Dmesg: > > [ 6.441438] fglrx: module license 'Proprietary. (C) 2002 - ATI > Technologies, Starnberg, GERMANY' taints kernel. > [ 6.441443] Disabling lock debugging due to kernel taint > [ 6.501099] [fglrx] Maximum main memory to use for locked dma > buffers: 3801 MBytes. > [ 6.501153] [fglrx] vendor: 1002 device: 6898 count: 1 > [ 6.501156] [fglrx] vendor: 1002 device: 6898 count: 2 > [ 6.501159] [fglrx] vendor: 1002 device: 6898 count: 3 > [ 6.501521] [fglrx] ioport: bar 4, base 0x9000, size: 0x100 > > It Seems, that all 3 HD5870 were detected. > The Sample Program FindNumDevices tolds me that i only have 1 Device. > Does anybody have an Hint for me ? > > thxs
You need to edit GenTable.py - it is clearly marked where you should edit the number of GPUs to use. (I know I could use all by default) While you are at it, take some time to study the auto increment feature. You will have to edit the increment (10) and the cutoff point (690). When these settings are correctly adjusted, you can leave your machine running for weeks, and it will produce 1 table after another without any intervention from your side, just make sure you have enough free disk space. As the tables are completed, you run sort2 on the output data, and you can delete the original files, freeing up space as you go along. (You can also use TableConvert and compress the tables to free up even more space.) cheers, Frank _______________________________________________ A51 mailing list A51@lists.reflextor.com http://lists.lists.reflextor.com/cgi-bin/mailman/listinfo/a51
