On 5 June 2014 21:36, <[email protected]> wrote: > Hi! >> When testing on ppc64 arch, `./hugemmap03 -s 64` failed message as: >> hugemmap03 1 TFAIL : mmap into high region failed unexpectedly >> - expect errno=ENOMEM, got: errno=EINVAL(22): Invalid >> >> But it is not the system issue, MMAP(2) Programmer's Manual define >> the mmap() error return value like: >> EINVAL We don't like addr, length, or offset (e.g., they are too >> large, or not aligned on a page boundary). >> ENOMEM No memory is available, or the process's maximum number of >> mappings would have been exceeded. >> So, we should consider the EINVAL return situation on ppc64. > > Which kernel version does this happen on? > > Can you capture kernel function trace of the test? > > We looked at the test (I've got help from Mel Gorman) and we concluded > that the test has nothing to do with hugepages and that it simply checks > that addresses outside of the process space are rejected by the kernel. > Which should return ENOMEM on ppc64 too (because the high address > constant in the test should be high enough for ppc64 too). > > -- > Cyril Hrubis > [email protected]
Hi Cyril, Thanks for your response. I hit the issue on kernel-2.6.32. But now I found it had already been fixed by Jerome's patch( https://lkml.org/lkml/2014/1/15/301 ). Obviously you are right! We should not change the LTP related code again. Thank you very much. :) -- Regards, Li Wang Email: [email protected] ------------------------------------------------------------------------------ HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions Find What Matters Most in Your Big Data with HPCC Systems Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. Leverages Graph Analysis for Fast Processing & Easy Data Exploration http://p.sf.net/sfu/hpccsystems _______________________________________________ Ltp-list mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ltp-list
