Usmar A Padow Reporter
	Description • 8 years ago

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.109 Safari/537.36

Steps to reproduce:

at first I thought "how can a server side program crash a client side program?"
on my server, I have a php file with the following code:
<?php for (;;) { echo "<b>MMmm</b><br />\n"; } ?>



Actual results:

firefox froze and crashed and then finally it would start again with the same page, had to quickly close it to avoid the same problem again


Expected results:

well, chrome does not crash, it displays a long list of the 
MMmm
MMmm
MMmm
MMmm
MMmm
MMmm
MMmm
MMmm
MMmm
MMmm

	Usmar A Padow Reporter
	Comment 1 • 8 years ago

also, sorry I don't give my link, but I don't want to receive many users to my server

	Ryan VanderMeulen [:RyanVM]
	Comment 2 • 8 years ago

OOM from the looks of it. Page janks heavily until eventually exhausting available memory and crashing.
https://crash-stats.mozilla.com/report/index/ffb0cb70-69d7-45b9-8310-00b722160216

Chrome also spikes the CPU and memory use appears to grow indefinitely, albeit at a much slower rate. I strongly suspect that it would also crash if I were patient enough to let it run for long enough.

If someone needs a live example of this testcase, I've uploaded to my personal hosting account. Ping me for the URL.

	Andrew McCreight [:mccr8]
	Comment 3 • 8 years ago

If our memory usage is growing much faster than Chrome, this might be interesting to look into as an extreme test case.

But this looks like a safe OOM crash, so it doesn't need to be a security bug.

	Henri Sivonen (:hsivonen)
	Comment 4 • 8 years ago

Weird. I thought I made this fail more gracefully in bug 1029671.

	Sylvestre Ledru [:Sylvestre]
	Comment 5 • 6 years ago

Closing because no crash reported since 12 weeks.