Created attachment 8609552 [details] Screenshot 2015-05-22 11.30.37.png See, for example, this crash: https://crash-stats.mozilla.com/report/index/02a8cec2-88fc-4b65-b6da-ccf272150522 It has, under "App Notes":: AdapterVendorID: 0x10de, AdapterDeviceID: 0x fe9 This space means we can't ever find it a graphics adapter name for this. E.g. See "Graphics Adapter Report" tab on https://crash-stats.mozilla.com/report/list?product=Firefox&signature=mozilla%3A%3Aipc%3A%3ASerializeInputStream%28nsIInputStream*%2C+mozilla%3A%3Aipc%3A%3AInputStreamParams%26%2C+nsTArray%3Cmozilla%3A%3Aipc%3A%3AFileDescriptor%3E%26%29
Apparently we have an API to spoof the adapter device ID: http://mxr.mozilla.org/mozilla-central/source/widget/nsIGfxInfoDebug.idl#14 It's possible some add-on or user has set this to a bad value. I'm not sure if it's worth trying to deal with this. WONTFIX?
This crash was from peterbe's computer, so feel free to ask him if he has something weird configured!
Kartikaya, What's your question? I have add-ons installed. Several of them. None are called Screw-Up-The-Hex-Values :)
Shoulda guessed it wouldn't be that obvious :) If you start Firefox with a clean profile and trigger a crash, does the crash report still have the spaces in the AdapterHexIDs? Alternatively, if you have a build you can debug, stick a breakpoint at http://mxr.mozilla.org/mozilla-central/source/widget/cocoa/GfxInfo.mm?rev=f25e20a0f238#396 and see if it gets hit. If there's no spoofing happening, then I'd be interested in knowing if the value obtained at http://mxr.mozilla.org/mozilla-central/source/widget/cocoa/GfxInfo.mm?rev=f25e20a0f238#96 has the space, or if the space is somehow inserted afterwards. As far as I can tell there's no way this should be happening unless there's some spoofing going on.
Here's another crash  I just made with 41.0a1 (2015-06-01). It has no extensions installed at all.  https://crash-stats.mozilla.com/report/index/ea292326-fda9-4804-8178-9f5132150601
Created attachment 8613610 [details] [diff] [review] Patch Doh.
Comment on attachment 8613610 [details] [diff] [review] Patch Review of attachment 8613610 [details] [diff] [review]: ----------------------------------------------------------------- nice