Closed
Bug 53935
Opened 25 years ago
Closed 25 years ago
Crash viewing long page
Categories
(Core :: DOM: HTML Parser, defect, P1)
Tracking
()
VERIFIED
FIXED
M18
People
(Reporter: bugzilla, Assigned: rickg)
References
()
Details
(Keywords: crash, Whiteboard: [rtm++][Fix in hand])
Attachments
(4 files)
|
2.11 KB,
patch
|
Details | Diff | Splinter Review | |
|
2.21 KB,
patch
|
Details | Diff | Splinter Review | |
|
2.77 KB,
text/plain
|
Details | |
|
6.75 KB,
text/plain
|
Details |
Build ID: just pulled
I crash in GKHTML.DLL viewing the URL (don't ask why I was there :P) will try
in a debug build asap.
|
||
Comment 1•25 years ago
|
||
I get a crash in kernel32.dll on
http://www.palosverdes.com/rpv/munidatabase/rpv/index.cfm after
hitting "search" (2000 092212), but I'm not getting it on 2000 092408.
http://www.ex-mozilla.org/date.html does crash for me, in gkhtml.dll, on 2000
092408.
It's asserting in HandleAttributeToken(), which will only happen in a debug
build. I'll dig into the page later today to find out why.
Status: NEW → ASSIGNED
Min test case:
<table>
<body>
<a href="foo">foo</a>
</body>
<FONT SIZE="+2"> Crash me!</FONT>
Requesting nsbeta3+ status because it's a very simple testcase, and an even
easier fix.
Keywords: nsbeta3
OS: Windows 98 → All
Priority: P3 → P1
Whiteboard: [nsbeta3+]
Target Milestone: --- → M18
|
||
Comment 5•25 years ago
|
||
PDT marking nsbeta3-, nominate for rtm if appropriate.
Whiteboard: [nsbeta3+] → [nsbeta3-]
We simply *cant* ship with this. Please approve for RTM.
|
|
Reporter | |
Updated•25 years ago
|
Whiteboard: [rtm]
|
|
Reporter | |
Comment 8•25 years ago
|
||
[RTM] in the status whiteboard doesn't mean anything. It should either be
[RTM+] (if you're allowed to + bugs), or the `rtm' keyword (in the keyword
field) to nominate it to be fixed for RTM.
marking ++ for review.
Whiteboard: [RTM+] -- fix in hand. → [RTM++] -- fix in hand.
|
|
Assignee | |
Comment 10•25 years ago
|
||
This is caused by a logic error in HandleOmittedTags() and/or
HandleSavedTokens(). The <table> is getting closed in HandleSavedTokens() when
it's autoclosing the stack to the marked position in order to deal with the
misplaced stuff. That would be fine, so long as it gets pushed back later (it
doesnt). We also fail to capture an attribute token before jumping into
HandleSavedTokens(), which later gets handled normally, which is why we assert.
Overall -- this code is *scary* -- please review.
Assignee: rickg → harishd
Status: ASSIGNED → NEW
Whiteboard: [RTM++] -- fix in hand. → [RTM+]
|
|
||
Comment 11•25 years ago
|
||
Darn, it definitely looks scary!!! I'll take a look into this. Thanx Rickg for
the reduced case.
Status: NEW → ASSIGNED
|
|
||
Comment 12•25 years ago
|
||
|
|
||
Comment 13•25 years ago
|
||
|
|
||
Comment 14•25 years ago
|
||
PDT marking [rtm need info] until code reviews are available. If the patch is
scary, can we validate it on the trunk before applying to the RTM branch?
Whiteboard: [RTM+][Fix in hand] → [RTM need info][Fix in hand]
|
|
||
Comment 15•25 years ago
|
||
PDT: The code was scary before the patch not after the patch.
|
|
||
Comment 16•25 years ago
|
||
a=buster. recommend for rtm++. I talked at length with harish about this, and
it seems like a significant stability win. We did a detailed code review, and
the changes look good.
|
|
||
Comment 17•25 years ago
|
||
Marking rtm+ to get back onto the PDT radar.
Whiteboard: [RTM need info][Fix in hand] → [rtm+][Fix in hand]
|
|
Assignee | |
Comment 18•25 years ago
|
||
The patch looks fine to me.
|
|
||
Comment 19•25 years ago
|
||
PDT marking [rtm++]
Whiteboard: [rtm+][Fix in hand] → [rtm++][Fix in hand]
|
|
||
Comment 20•25 years ago
|
||
Fix is in the branch & trunk. Marking fixed.
Status: ASSIGNED → RESOLVED
Closed: 25 years ago
Resolution: --- → FIXED
|
|
||
Comment 21•25 years ago
|
||
Works (very slowly) on the trunk build but not on the branch build.
tested on:
trunk build -
2000-10-12-09-Mtrunk : win32
branch builds-
2000-10-12-13-MN6 : win32
2000-10-12-09-MN6 : linux
2000-10-11-10-MN6 : mac
attaching talkback reports from windows crash and from linux crash
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
|
|
||
Comment 22•25 years ago
|
||
|
|
||
Comment 23•25 years ago
|
||
|
|
||
Comment 24•25 years ago
|
||
This is a different stack trace altogether. That is, the bug is not the same
anymore.
BlockFrame = buster :-)
Reassigning to buster and ccing waterson.
Assignee: harishd → buster
Status: REOPENED → NEW
|
|
||
Comment 25•25 years ago
|
||
the reported crash has nothing to do with this bug. Resetting this bug to
rickg and fixed, filing another bug for the newly discovered problem.
|
|
||
Comment 26•25 years ago
|
||
this bug really is fixed
Status: NEW → RESOLVED
Closed: 25 years ago → 25 years ago
Resolution: --- → FIXED
You need to log in
before you can comment on or make changes to this bug.
Description
•