test_map and tracex3 uses percpu array. Currently it might fail due to number of CPU mismatch between kernel and userspace. We will provide patch later.
Regards, William On Thu, Sep 8, 2016 at 3:36 AM, Louie Lu via iovisor-dev <[email protected]> wrote: > Hi everyone, > > I'm running kernel 4.8.0-rc5, and after running test_maps under > samples/bpf, I found that `test_maps` just failed in > test_percpu_hashmap_sanity() at line 146 > > key = 2; > /* check that key=2 is not found */ > assert(bpf_lookup_elem(map_fd, &key, value) == -1 && errno == ENOENT); > > After checking William Tu's post in Jun 21, setting ulimit to > unlimited, and using patch from > here:https://patchwork.ozlabs.org/patch/640485/, it just still don't > work. > > And, the sample which used per cpu map "tracex3" got a segmentation fault. > > ➜ bpf git:(master) ✗ ./tracex3 > heatmap of IO latency > - many events with this latency > - few events > |1us |10us |100us |1ms |10ms |100ms |1s |10s > # 0 > [1] 15738 segmentation fault (core dumped) ./tracex3 > > after some ad hoc testing, found that bpf_lookup_elem() was the > problem to get segmentation fault. > > Are there any suggestion for debugging this problem and how to fix this? > > > Thanks. > > Louie Lu > _______________________________________________ > iovisor-dev mailing list > [email protected] > https://lists.iovisor.org/mailman/listinfo/iovisor-dev _______________________________________________ iovisor-dev mailing list [email protected] https://lists.iovisor.org/mailman/listinfo/iovisor-dev
