| commit | 0eb394e2b9fa06ff225084969300fe65b8f17dfc | |
| author | eroman <eroman@chromium.org> | |
| Thu, 30 Apr 2015 23:25:46 +0000 (30 16:25 -0700) | ||
| committer | Commit bot <commit-bot@chromium.org> | |
| Thu, 30 Apr 2015 23:26:30 +0000 (30 23:26 +0000) | ||
| tree | 38a4cfa1dd8bcdc88ef077f311bf576475f07fe8 | treesnapshot (tar.gz zip) |
| parent | d522cf5a1615d2b1fdf5e1e9b3f385f75d687c1a | commitdiff |
Narrow the instrumentation to track jank in HostResolverImpl.
I am fairly confident this is coming from IsGloballyReachable(), which seems to have a dependency on UDP sockets, so focused the instrumentation there.
One weakness of this instrumentation is it doesn't separate out some of the non-trivial work being done by local variable's destructors (most notably deletion of the UDP sockets, which will likely be hitting issue 165382).
BUG=455942
TBR=pkasting@chromium.org
Review URL: https://codereview.chromium.org/1120733002
Cr-Commit-Position: refs/heads/master@{#327831}
I am fairly confident this is coming from IsGloballyReachable(), which seems to have a dependency on UDP sockets, so focused the instrumentation there.
One weakness of this instrumentation is it doesn't separate out some of the non-trivial work being done by local variable's destructors (most notably deletion of the UDP sockets, which will likely be hitting issue 165382).
BUG=455942
TBR=pkasting@chromium.org
Review URL: https://codereview.chromium.org/1120733002
Cr-Commit-Position: refs/heads/master@{#327831}
| net/dns/host_resolver_impl.cc | diffblobblamehistory |