Hi -
I wrote:
> [...] I've run valgrind on pmmgr in the past against a larger
> network-scanny installation, and again no complaints. Am rerunning
> now.
A brief scanny run now produces actual valgrind leak reports ...
all within libpcp, not pmmgr:
==17174== HEAP SUMMARY:
==17174== in use at exit: 203,575 bytes in 114 blocks
==17174== total heap usage: 240,631 allocs, 240,517 frees, 60,114,048 bytes
allocated
==17174==
==17174== 324 (72 direct, 252 indirect) bytes in 9 blocks are definitely lost
in loss record 14 of 21
==17174== at 0x4C2B946: calloc (in
/usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==17174== by 0x4E80824: __pmHashAdd (hash.c:51)
==17174== by 0x4E6D346: parseAttributeSpec (spec.c:842)
==17174== by 0x4E6D346: __pmParseHostAttrsSpec (spec.c:937)
==17174== by 0x4E49545: pmNewContext (context.c:468)
==17174== by 0x116B8A: pmmgr_job_spec::compute_hostid(std::string const&)
(pmmgr.cxx:236)
==17174== by 0x1178A6: pmmgr_job_spec::poll() (pmmgr.cxx:430)
==17174== by 0x10DE48: main (pmmgr.cxx:1313)
==17174==
==17174== 1,509 bytes in 50 blocks are definitely lost in loss record 16 of 21
==17174== at 0x4C29BCF: malloc (in
/usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==17174== by 0x5B87129: strdup (strdup.c:42)
==17174== by 0x4E70B94: posix_formatter (config.c:159)
==17174== by 0x4E70EA3: __pmConfig (config.c:218)
==17174== by 0x4E71069: pmGetConfig (config.c:242)
==17174== by 0x10DA53: main (pmmgr.cxx:1201)
==17174==
==17174== LEAK SUMMARY:
==17174== definitely lost: 1,581 bytes in 59 blocks
==17174== indirectly lost: 252 bytes in 18 blocks
==17174== possibly lost: 0 bytes in 0 blocks
==17174== still reachable: 201,742 bytes in 37 blocks
==17174== suppressed: 0 bytes in 0 blocks
==17174== Reachable blocks (those to which a pointer was found) are not shown.
==17174== To see them, rerun with: --leak-check=full --show-leak-kinds=all
- FChE
|