1590777 - Crash in [@ DOMSecurityManager::ParseCSPAndEnforceFrameAncestorCheck]

Status: RESOLVED FIXED

(Core :: DOM: Security, defect, P1)

Description

[Calixte Denizet (:calixte)]

This bug is for crash report bp-13bef52a-dcea-4184-a4ad-1efd20191023.

Top 10 frames of crashing thread:

0 xul.dll nsresult DOMSecurityManager::ParseCSPAndEnforceFrameAncestorCheck dom/security/DOMSecurityManager.cpp:170
1 xul.dll nsresult DOMSecurityManager::Observe dom/security/DOMSecurityManager.cpp:111
2 xul.dll nsObserverService::NotifyObservers xpcom/ds/nsObserverService.cpp:291
3 xul.dll mozilla::net::nsHttpHandler::NotifyObservers netwerk/protocol/http/nsHttpHandler.cpp:806
4 xul.dll nsresult mozilla::net::nsHttpChannel::ProcessResponse netwerk/protocol/http/nsHttpChannel.cpp:2474
5 xul.dll nsresult mozilla::net::nsHttpChannel::OnStartRequest netwerk/protocol/http/nsHttpChannel.cpp:7705
6 xul.dll unsigned int nsInputStreamPump::OnStateStart netwerk/base/nsInputStreamPump.cpp:487
7 xul.dll nsresult nsInputStreamPump::OnInputStreamReady netwerk/base/nsInputStreamPump.cpp:396
8 xul.dll nsresult nsInputStreamReadyEvent::Run xpcom/io/nsStreamUtils.cpp:91
9 xul.dll nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1225

There are 32 crashes (from 15 installations) in nightly 72 starting with buildid 20191022214314. In analyzing the backtrace, the regression may have been introduced by patch [1] to fix bug 1584993.

[1] https://hg.mozilla.org/mozilla-central/rev?node=e21ad27bfd0a

Comment 1

[Christoph Kerschbaumer [:ckerschb}]

Attachment: Bug 1590777: Add Null check for referrerinfo within ParseCSPAndEnforceFrameAncestorCheck. r=tnguyen

Comment 2

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/6dffdf63221a
Add Null check for referrerinfo within ParseCSPAndEnforceFrameAncestorCheck. r=tnguyen

Comment 3

[Arthur Iakab [arthur_iakab]]

https://hg.mozilla.org/mozilla-central/rev/6dffdf63221a

Comment 5

[Catalin Sasca, Desktop Test Engineering [:csasca]]

Hi Cristoph, is there something manually verificable? And if yes, could you provide some steps? Thanks!

Comment 6

[Christoph Kerschbaumer [:ckerschb}]

(In reply to Catalin Sasca, QA [:csasca] from comment #5)

Hi Cristoph, is there something manually verificable? And if yes, could you provide some steps? Thanks!

I don't think there is actually, we just looked at the crash signature and added a null check before derefing the pointer, but there was not testcase and I also can't think of a testcase which would trigger that codeline.