Adding UI>Browser>Navigation component, since this is navigation-related, ish.

I'm not able to repro on Linux 53.0.2774.3.  Can you repro on other platforms?  Without symbols in the crash dump, there's not much we can do here.

Re #2: I'm able to repro on both my Chromebox and another ChromeOS device.

I found that my personal account, in a different profile on the same device, does not have the problem, so I suspect it's tickled by some extension or other.

Does not repro on 53.0.2773.3, nor 2782.6 for me, only on 2773.0.  

This issue has started reproing again; seems it may be related to up-time or installed extensions injecting stuff into pages, perhaps, hence only repros after a while?

See report Ids 7e53a23600000000, 35b2623600000000, 3d7a623600000000 and 7b4327f100000000.

FYI,  issue 610902  is tracking the symbolization problem.

Sorry, I'm not able to help here.  I don't have a ChromeOS debugging setup, and there's no indication of what code this is happening in.  Since it doesn't repro on other platforms, it's unlikely to be navigation code.  I'd suggest pinging someone on ChromeOS?

Have we checked what extensions were installed at the time? Is this reproing on the build under test today? 8530.6.0 and 53.0.2785.4.

dchan@ can you please confirm on comment #9 above?

Making this a RBB instead since we don't yet have confirmation on whether this exists on today's build.

Adding dgarrett@ and mmandlis@ because of the symbolization issue w/ the crash reports.

The symbols that were missing when these reports were processed, were uploaded since.
Let me try to resubmit these reports, and see if the stacktraces look better in the new version. I'll update the bug with new report IDs shortly.

Looking at the first report I uploaded, I see two chrome modules listed as
missing symbols, one with a null identifier, so I'd guess we're hitting the
issue w/ multiple executable segments in the binary?

I resubmitted and resulting report IDs are: f88cddb600000000, ac21850900000000
Unfortunately, this hasn't helped much, because the processor is still trying to use the invalid debug ID 000000000000000000000000000000000.
I'll check if there is anything else that I could do.

cannot reproduce on veyron_speedy 54.0.2794.0, is this M53 only? Btw, I am also interested in comment 9, any updates?

Since there is a symbolization issue with this crash report, I really want a repro.

Unfortunately I don't have a repro; it just happened (reliably) when I filed the report, and then after a reboot (not sure if there was an update or not) it no longer happened.  I then observed it once again and re-opened the bug. Right now it isn't repro'ing.

However, I did also file  issue 627137  (browser crashes if you click Details for any Extension under chrome://extension), which still reliably repros, though I'm not sure it's the same issue.

Are they (this bug and 627137) happened on the same DUT, same profile? Then it might be related. But from the comments in 627137, I cannot see a relation with this. I tried several times today and was not able to repro. Any luck on your side? wez@

Re #22: No, I have not repro'd this issue recently. I did just file  issue 629247 , another browser crasher, but I suspect that's another case of  issue 627137 .

All I can suggest is to grab one of the minidumps I uploaded, hack the modules-stream to fix the memory maps to show the chrome binary correctly (see issue 628040) and then use dump_syms to get correctly symbolized stacks. :P

Since we don't have a reliable repro currently wez@ can we make this non-RBB? 

Since it doesn't seem to be reproing on the current dev, that seems reasonable, yes

Marking this as non-RBB.

Moving this nonessential bug to the next milestone.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

mark as wontfix, please reopen it when this becomes an issue again.