Closed
Bug 4018
Opened 26 years ago
Closed 26 years ago
NECKO: [PP] Crash with object tag on Linux
Categories
(Core :: Networking, defect, P2)
Tracking
()
VERIFIED
FIXED
M9
People
(Reporter: glynn, Assigned: gagan)
References
()
Details
This appears to be a slightly different crash from the others reported already,
so I am submitting to make sure bases are covered.
March 18 M4 optimized apprunner or viewer builds, cross platform Mac and Linux
1. Launch apprunner and go to: about:glynn, which routes you to my Netscape
personal page: http://people.netscape.com/glynn/, a simple frameset created eons
ago which contains an embedded qtime movie and a link. Frameset code:
<HTML>
<HEAD><TITLE>Greg Lynn Netscape Home Page</TITLE></HEAD>
<FRAMESET ROWS="40%,34%,*" FRAMEBORDER=NO BORDER=0 FRAMECOLOR="#FFFFFF">
<FRAME SCROLLING="NO" SRC="contents.html" NAME="Contents">
<FRAME SCROLLING="NO" SRC="display.html" NAME="Display">
<FRAME SCROLLING="NO" SRC="display2.html" NAME="Display2">
<NOFRAME>
<A HREF="contents.html">Greg Lynn's Home Page</A>
</NOFRAME>
</FRAMESET>
</HTML>
• Crash on Mac into Macsbug PPC access exception. Linux simply quits itself
unexpectedly, Windows seems to work, though page layout is not great.
Marked plugins initially, may be issue with frames layout. Will try to narrow
it down a little later.
Updated•26 years ago
|
Status: NEW → ASSIGNED
Comment 1•26 years ago
|
||
cc'ing Patrick as he has been working on the Mac plugin code recently
Updated•26 years ago
|
Status: ASSIGNED → RESOLVED
Closed: 26 years ago
Resolution: --- → FIXED
Comment 2•26 years ago
|
||
This no longer crashes for me on either Mac or Windows.
March 23 builds, Mac and Windows don't crash, but RedHat Linux still quits
itself. Reopening.
Updated•26 years ago
|
Status: REOPENED → ASSIGNED
OS: Mac System 8.5 → Linux
Hardware: All → Other
Comment 4•26 years ago
|
||
Changing OS to Linux
Comment 5•26 years ago
|
||
Changing QA contact to sujay
Updated•26 years ago
|
Target Milestone: M5
Summary: App quits/crashes access page with embedded plugin content → [PP]App quits/crashes access page with embedded plugin content
Updated•26 years ago
|
Priority: P3 → P2
Comment 6•26 years ago
|
||
Andrei - can you test this out on Linux?
This home page is no longer there anymore..
Greg, can you re-direct us to a similar page on your
personal home page or another site?
thanks!
Yesterday build on Linux behaves as described with 'Segmentation fault'.
Comment 10•26 years ago
|
||
Andrei - what URL did you use?
Comment 11•26 years ago
|
||
I just typed about:glyn. The crash is not consistent but when it happens I can
see that it tries to load home.netscape.com, and that's what I got in the
terminal window:
Area(td)(0)@0x8251dc8: WARNING: max-element-size:3000,860 desired:7320,640
maxSize:1073741824,1073741824
Warning - table cell content max element height 900 greater than desired height
680
Segmentation fault (core dumped)
Looks like it has nothing to do with the glynn's page, besides I saw the same
with about:[somebodyelse].
Comment 12•26 years ago
|
||
Looks like a crasher with embeded content on Linux ONLY.
Go to the above URL and you will crash on Linux.
Comment 13•26 years ago
|
||
Looks like a crasher with embeded content on Linux ONLY.
Go to the above URL and you will crash on Linux.
Comment 14•26 years ago
|
||
To looks like it has nothing to do with object tag, but rather with redirection.
I type in about:glynn which redirects me to home.netscape.com.
So, what is happening on redirection.
Here is the call stack
#0 0x408e95bc in PR_Close (fd=0x74656e2e) at priometh.c:109
#1 0x4066071e in net_ProcessHTTP (ce=0x8179738) at
../../../../network/protocol/http/mkhttp.c:3625
#2 0x406d046f in NET_ProcessNet (ready_fd=0x81b8f08, fd_type=2) at
../../../network/main/mkgeturl.c:3355
#3 0x406d83fd in NET_PollSockets () at ../../../network/main/mkselect.c:320
#4 0x406a8e9a in nsNetlibService::NetPollSocketsCallback (aTimer=0x81768e0,
aClosure=0x816a258)
at ../../../network/module/nsNetService.cpp:1256
#5 0x40510de9 in TimerImpl::FireTimeout (this=0x81768e0) at
../../../../base/src/gtk/nsTimer.cpp:73
#6 0x405112d2 in nsTimerExpired (aCallData=0x81768e0) at
../../../../base/src/gtk/nsTimer.cpp:189
#7 0x40a88990 in g_timeout_dispatch (source_data=0x81b8d98,
current_time=0xbffff758, user_data=0x81768e0)
at gmain.c:1147
#8 0x40a87c83 in g_main_dispatch (current_time=0xbffff758) at gmain.c:647
#9 0x40a8820f in g_main_iterate (block=1, dispatch=1) at gmain.c:854
#10 0x40a88391 in g_main_run (loop=0x8183810) at gmain.c:912
#11 0x409b544b in gtk_main () at gtkmain.c:475
#12 0x4004d158 in nsAppShell::Run (this=0x8076500) at
../../../../widget/src/gtk/nsAppShell.cpp:208
#13 0x8053bd2 in ?? ()
#14 0x8053de6 in ?? ()
The code vicinity where it happens for frame #1 is:
3619
TIMING_STOPCLOCK_NAME("http:request",ce->URL_s->address,ce->window_id,"error");
3620 TIMING_STOPCLOCK_NAME("http:complete",
ce->URL_s->address,ce->window_id,"error");
3621 if(cd->connection->sock != NULL) {
3622 NET_ClearDNSSelect(ce->window_id, cd->connection->sock);
3623 NET_ClearReadSelect(ce->window_id, cd->connection->sock);
3624 NET_ClearConnectSelect(ce->window_id, cd->connection->sock);
3625 PR_Close(cd->connection->sock);
3626 NET_TotalNumberOfOpenConnections--;
3627 cd->connection->sock = NULL;
3628 }
cd->connection looks like this:
(gdb) p *cd->connection
$3 = {hostname = 0x656d6f68 <Address 0x656d6f68 out of bounds>, sock =
0x74656e2e, busy = 115 's',
prev_cache = 99 'c', secure = 97 'a', doNotSendConHdr = 112 'p'}
where sock is corrupted:
(gdb) p *cd->connection->sock
Cannot access memory at address 0x74656e2e.
To me it looks like netlib bug, so reassigning.
Assignee | ||
Comment 15•26 years ago
|
||
Marking till Necko lands...
Assignee | ||
Comment 16•26 years ago
|
||
Per DP's suggestion marking these till M8. Though Necko lands with M7, we will
be able to verify it for M8.
Comment 17•26 years ago
|
||
I'm moving this to target M9, Necko will be enabled somewhere during late M8 or
early M9. We will need to get on this and it cannot be postponed past the M9
milestone.
Comment 18•26 years ago
|
||
Changing all Networking Library/Browser bugs to Networking-Core component for
Browser.
Occasionally, Bugzilla will burp and cause Verified bugs to reopen when I do
this in a bulk change. If this happens, I will fix. ;-)
Summary: [PP] Crash with object tag on Linux → NECKO: [PP] Crash with object tag on Linux
Assignee | ||
Comment 19•26 years ago
|
||
Please check if its still valid with Necko.
Updated•26 years ago
|
Status: ASSIGNED → RESOLVED
Closed: 26 years ago → 26 years ago
Resolution: --- → FIXED
Updated•26 years ago
|
Status: RESOLVED → VERIFIED
Comment 20•26 years ago
|
||
well, the redirect is fine now. Greg's page is gone now, so we can't test the
original problem. Oh well.
Comment 21•25 years ago
|
||
Bulk move of all Networking-Core (to be deleted component) bugs to new
Networking component.
You need to log in
before you can comment on or make changes to this bug.
Description
•