Wireshark-bugs: [Wireshark-bugs] [Bug 12656] Buildbot crash output: fuzz-2016-07-24-1421.pcap

Date: Sat, 13 Aug 2016 22:42:11 +0000

Comment # 33 on bug 12656 from
So if I allocate the hash tables in question with wmem_epan_scope(), and don't
bother freeing them (leaving it up to the libwireshark code to free that
scope), I get

==35054== LEAK SUMMARY:
==35054==    definitely lost: 12,818 bytes in 204 blocks
==35054==    indirectly lost: 10,698 bytes in 316 blocks
==35054==      possibly lost: 1,172,867 bytes in 9,249 blocks
==35054==    still reachable: 269,464 bytes in 1,045 blocks
==35054==         suppressed: 645,431 bytes in 357 blocks

Running tshark on an empty capture shows no CPU time differences between
"allocate with global scope" and "allocate with epan scope"; is there any
compelling reason *not* to use epan scope here?


You are receiving this mail because:
  • You are watching all bug changes.