Bug 12467 – REGRESSION: Loading webarchive with empty frame source crashes WebKit ToT

• Component: WebKit Misc.
• Status: RESOLVED FIXED

[David Kilzer (:ddkilzer)]

Summary:

Loading a webarchive file containing a frame with no source crashes WebKit ToT.

Steps to reproduce.

1. Open Safari/WebKit.
2. Open test webarchive.

Expected results:

Safari/WebKit should not crash.

Actual results:

Safari/WebKit crashes.

Regression:

This does not occur on shipping Safari 2.0.4 (419.3) on Mac OS X 10.4.8 (8N1037).

Confirmed crash with locally-built debug build of WebKit r19232 with Safari 2.0.4 (419.3) on Mac OS X 10.4.8 (8N1037).

Notes:

This bug was separate from Bug 12463 after I misunderstood the original issue and happened upon a different bug.

Jim Correia filed this Radar bug for this issue (which is probably missing a nil check in the frame loader code per Bug 12463 Comment #6):

<rdar://problem/4961915>

[David Kilzer (:ddkilzer)]

Created attachment 12762 [details]
Test webarchive (will crash ToT!)

This is a webarchive of Attachment 12750 [details] from Bug 12463.

[David Kilzer (:ddkilzer)]

Created attachment 12763 [details]
Stack trace

[David Kilzer (:ddkilzer)]

Jim Correia wrote:

(In reply to comment #0)
> Jim Correia filed this Radar bug for this issue (which is probably missing a
> nil check in the frame loader code per Bug 12463 Comment #6):
>
> <rdar://problem/4961915>

Sorry if my brevity introduced confusion. That radar refers to duplicate page elements resultant from Safari archiving the current DOM instead of the data source. It was meant as an explanation for why I am archiving the data source and not the DOM (therefore why this 12463 affects my application and not Safari.)

This bug - 12467 - is a wholly separate issue.

[David Kilzer (:ddkilzer)]

(In reply to comment #3)
> (In reply to comment #0)
> > Jim Correia filed this Radar bug for this issue (which is probably missing a
> > nil check in the frame loader code per Bug 12463 Comment #6):
> >
> > <rdar://problem/4961915>
> 
> Sorry if my brevity introduced confusion. That radar refers to duplicate page
> elements resultant from Safari archiving the current DOM instead of the data
> source. It was meant as an explanation for why I am archiving the data source
> and not the DOM (therefore why this 12463 affects my application and not
> Safari.)

That sounds like Bug 7266.  I have a fix for that, but I need to fix Bug 11882 first.

[Jim Correia]

rdar://problem/4424647 also fits in (archives don't work in "offline" mode)

I filed that in radar only on 1/27/2006 and it was closed as a duplicate. (I couldn't find an equivalent bug in WebKit Bugzilla, but may have overlooked it.0

[David Kilzer (:ddkilzer)]

(In reply to comment #5)
> rdar://problem/4424647 also fits in (archives don't work in "offline" mode)
> 
> I filed that in radar only on 1/27/2006 and it was closed as a duplicate. (I
> couldn't find an equivalent bug in WebKit Bugzilla, but may have overlooked
> it.

That's probably because the current implementation misses some resources when it saves the webarchive, most notably anything referred to from CSS.  Some relevant bugs:

Bug 11839
Bug 11850
Bug 12025

[David Kilzer (:ddkilzer)]

<rdar://problem/4963871>

[David Kilzer (:ddkilzer)]

Committed by mjs in r19281.