Compare those two crash reports: bp-0d82972a-ab33-4f14-b68e-1bed92160129 (no e10s) bp-8bb19104-fedd-41fd-9b1f-8ab772160129 (e10s enabled) The first one has plugin name and version annotations, the second doesn't. At least in 45.0b1, this happens with all e10s crashes, see https://crash-stats.mozilla.com/search/?product=Firefox&version=45.0b1&process_type=plugin&dom_ipc_enabled=!__null__&_facets=plugin_name&_facets=plugin_version#facet-plugin_version (e10s on) compared to https://crash-stats.mozilla.com/search/?product=Firefox&version=45.0b1&process_type=plugin&dom_ipc_enabled=__null__&_facets=plugin_name&_facets=plugin_version#facet-plugin_version (e10s off) We should fix this before shipping e10s.
Agree this is pretty important. These annotations are written here: http://hg.mozilla.org/mozilla-central/annotate/5f9ba76eb3b1/dom/plugins/ipc/PluginModuleParent.cpp#l836 I'd expect PluginModuleChromeParent::WriteExtraDataForMinidump to be the codepath for both chrome and plugin crashes. Bill am I wrong about that class being used (in the chrome process) in the e10s case?
something wee need for crash analysis.
Assignee: nobody → wmccloskey
tracking-e10s: ? → m9+
I just tested this in a local mc build and the annotations are getting set properly. https://crash-stats.mozilla.com/report/index/a009eb6d-01e9-4743-b82a-41fe42160317
Very very strange, I'm not sure what's going on here. https://crash-stats.mozilla.com/search/?product=Firefox&version=48.0a1&process_type=plugin&dom_ipc_enabled=!__null__&_facets=plugin_name&_facets=plugin_version&_facets=plugin_filename&_facets=cpu_arch&_columns=signature&_columns=platform&_columns=plugin_name&_columns=cpu_arch#crash-reports
I knew this code looked familiar! Fixed by bug 1256943. http://hg.mozilla.org/mozilla-central/rev/3c8ea88e6916
Status: NEW → RESOLVED
Last Resolved: 3 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.