遇到这个问题是在测试项目的性能时发现的，每个对象分配一页大小的空间然后mprotect() 保护起来，当系统分配3W多个页的时候会出现这个问题。
google到在一份邮件列表中也曾提到该问题.https://sourceware.org/ml/libc-help/2010-04/msg00026.html
摘抄部分如下：
----------------------------------------------------------------------------------------------------------------------
On Wed, 2010-04-21 at 01:05 +0200, Pawe? Sikora wrote:
> hi,
>
> i‘m trying to debug an ugly application with ElectricFence.
> in fact, on x86-64 box with 8GB ram and 16GB swap i‘m getting following error:
>
> "ElectricFence Exiting: mprotect() failed: Cannot allocate memory"
>
> the program has been compiled with gcc-4.5, glibc-2.11.1, kernel-2.6.32.
> did you ever come across such (kernel/glibc) limitations?
>
> here‘s a simple testcase which triggs -ENOMEM in mprotect().
You probably depleted the max map count, see:
/proc/sys/vm/max_map_count
We have a limit on the number of maps you can have, those mprotect()
calls split you maps like crazy, see also /proc/$pid/maps.
eg. change your second test program to include something like:
char buf[128];
snprintf(buf, sizeof(buf), "cat /proc/%d/maps", (int)getpid());
system(buf);
at the end after lowering your NN count to fit, and observe the result
of those mprotect() calls.
yes, that is the clue :)
the limit in /proc/sys/vm/max_map_count was set to 65530.
with `echo 128000 > /proc/sys/vm/max_map_count` the testcase passes.
----------------------------------------------------------------------------------------------------------------------
所以 解决方案是设置/proc/sys/vm/max_map_count为更大的值。
查看 max_map_count值：
more /proc/sys/vm/max_map_count
vm.max_map_count=128000直接写到/etc/sysctl.conf中,
然后执行
sysctl -p
或者直接执行
echo 128000 > /proc/sys/vm/max_map_count
OK 到这里问题就解决了。总结一下原因：
proc文件系统给用户提供了很多内核信息帮助，使得用户可以通过修改内核参数达到提高系统性能的目的.
-----------------------------------------------------------------------------------------------------------------------------------------------------
[email protected]:/proc/sys/vm$ ls
block_dump legacy_va_layout oom_dump_tasks
compact_memory lowmem_reserve_ratio oom_kill_allocating_task
dirty_background_bytes max_map_count overcommit_memory
dirty_background_ratio memory_failure_early_kill overcommit_ratio
dirty_bytes memory_failure_recovery page-cluster
dirty_expire_centisecs min_free_kbytes panic_on_oom
dirty_ratio min_slab_ratio percpu_pagelist_fraction
dirty_writeback_centisecs min_unmapped_ratio scan_unevictable_pages
drop_caches mmap_min_addr stat_interval
extfrag_threshold nr_hugepages swappiness
extra_free_kbytes nr_hugepages_mempolicy vfs_cache_pressure
hugepages_treat_as_movable nr_overcommit_hugepages would_have_oomkilled
hugetlb_shm_group nr_pdflush_threads zone_reclaim_mode
laptop_mode numa_zonelist_order
---------------------------------------------------------------------------------
/proc/sys/vm 下每个文件的信息详见http://www.linuxinsight.com/proc_sys_vm_hierarchy.html
max_map_count : This file contains the maximum number of memory map areas a process may have. Memory map areas are used as a side-effect of calling
malloc, directly by mmap and mprotect, and also when loading shared libraries. While most applications need less than a thousand maps, certain programs, particularly malloc debuggers, may consume lots of them, e.g., up to one or two maps per allocation. The
default value is 65536.