32.1.0 beta with Google WebComponents

Talk about code development, features, specific bugs, enhancements, patches, and similar things.
Forum rules
Please keep everything here strictly on-topic.
This board is meant for Pale Moon source code development related subjects only like code snippets, patches, specific bugs, git, the repositories, etc.

This is not for tech support! Please do not post tech support questions in the "Development" board!
Please make sure not to use this board for support questions. Please post issues with specific websites, extensions, etc. in the relevant boards for those topics.

Please keep things on-topic as this forum will be used for reference for Pale Moon development. Expect topics that aren't relevant as such to be moved or deleted.
User avatar
THX-1139
Lunatic
Lunatic
Posts: 453
Joined: 2019-06-13, 13:42
Location: In a place with no YT

Re: 32.1.0 beta with Google WebComponents

Unread post by THX-1139 » 2023-03-05, 22:18

Okay, Thanks...I think I get it now: Install Palemoon-Portable-32.1.0b3.win64.exe to the T drive
Then extract palemoon-32.1.0b3.win64.7z to "wherever" and then replace the bin/palemoon contents with those from the zip. :thumbup:
Pale Moon 32.5.0 (64) Win7 pro (64) Intel Core i5-3570 3.4GHz-16 GB DDR3

"The biggest joke on mankind is that computers have started asking humans to prove they're not a robot."

"Man needs difficulties; they are necessary for health." ~Carl Jung

dbsoft
Project Contributor
Project Contributor
Posts: 407
Joined: 2020-02-21, 17:35

Re: 32.1.0 beta with Google WebComponents

Unread post by dbsoft » 2023-03-06, 16:08

I have experienced a few crashes while testing on both Mac and Windows. I am using master branch from last night... but I got some of the crashes with the beta 3 code base.... however I didn't have debugging setup. This call stack is from the code base from about 12 hours ago... not sure if anything new has been committed... not sure this crash is in anything WebComponents related, it seems to be in the JS engine.... I was starting to watch a video on Odysee, one of the sites people noted as heavily using WebComponents. I also had a few other WebComponents sites open in other tabs but not actively using them (Rotten Tomatoes, YouTube, GitHub, NBC News). Here is the crash log and call stack from my debugger on Mac:

https://pastebin.com/cSYAMu5U

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35477
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by Moonchild » 2023-03-06, 16:26

The callstack points to the generational garbage collector. Just triggered normally from the refresh driver so not really pointing to anything in particular. It may have something to do with shadow DOM if it's WC related but there's no evidence pointing at that specifically here, so there isn't much here that points at an explicit bug, unfortunately.
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

dbsoft
Project Contributor
Project Contributor
Posts: 407
Joined: 2020-02-21, 17:35

Re: 32.1.0 beta with Google WebComponents

Unread post by dbsoft » 2023-03-06, 16:46

Moonchild wrote:
2023-03-06, 16:26
The callstack points to the generational garbage collector. Just triggered normally from the refresh driver so not really pointing to anything in particular. It may have something to do with shadow DOM if it's WC related but there's no evidence pointing at that specifically here, so there isn't much here that points at an explicit bug, unfortunately.
Yeah, that was my assessment as well, but figured I'd post it here just in case I missed anything... I'll do an ASan build on Linux and see if I can replicate it there.

Got another crash this time on Windows... wasn't able to get an ASan crash on Linux yet, but also did a Mac ASan build... running them both currently.

https://pastebin.com/8rHqSnED

dbsoft
Project Contributor
Project Contributor
Posts: 407
Joined: 2020-02-21, 17:35

Re: 32.1.0 beta with Google WebComponents

Unread post by dbsoft » 2023-03-07, 18:43

Another crash running Mac ASan build, but this one is even more confusing.... it actually crashed in the OpenGL driver and while I was AFK... I came back after leaving it running for a few hours and it was crashed like this:

https://pastebin.com/TVmUwVhL

I am not sure if these crashes are related to WebComponents or not, but even if they are and there is something causing instability... with the exception of the one I had while I was AFK, I have really been having to work to get it to crash. I'd probably even if we don't identify what the issue is I am having, I would vote in favor of enabling WebComponents either way.

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35477
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by Moonchild » 2023-03-07, 20:42

dbsoft wrote:
2023-03-06, 16:46
Got another crash this time on Windows...
hmm. maybe that one is fallout from the changes that were made to regexpshared to make it a GC thing; that has nothing to do with WC though, but maybe should look into, if you have steps to reproduce that would help? If it's hard to make crash and is intermittent then that will make it harder (but also less likely to his people in the wild)

The OpenGL crash is pretty clearly a crash in the GeForce driver so we can't do anything about that.
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

dbsoft
Project Contributor
Project Contributor
Posts: 407
Joined: 2020-02-21, 17:35

Re: 32.1.0 beta with Google WebComponents

Unread post by dbsoft » 2023-03-07, 21:25

Just been browsing videos on Odysee, Facebook, Youtube, GitHub, NBC News, Rotten Tomatoes primarily. It has not crashed doing the same thing any time. Was doing something different each time it crashed, wish I could provide a way to reproduce. Here is a new one, null pointer dereference this time in some code recently touched:

https://pastebin.com/LSAUy42V

This latest one I was on Facebook's tab loading conversation with all that other stuff in background tabs.

GetUncomposedDoc() at nsINode.cpp line 1634 returned NULL which caused the null pointer dereference, many of the changes were switching GetUncomposedDoc() for GetComposedDoc()... is this something necessary here?

Looks like it was GetUncomposedDoc() in Mozilla's code base until it was removed entirely in: https://bugzilla.mozilla.org/show_bug.cgi?id=1469521

User avatar
FranklinDM
Add-ons Team
Add-ons Team
Posts: 575
Joined: 2017-01-14, 02:40
Location: Philippines
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by FranklinDM » 2023-03-07, 23:42

Hmm... I suspect it's this commit for the last crash. I was initially hesitant if this change was necessary since I can't reproduce the issue in the provided test case in the linked m-c bug. Can you check if reverting that commit fixes the issue?

dbsoft
Project Contributor
Project Contributor
Posts: 407
Joined: 2020-02-21, 17:35

Re: 32.1.0 beta with Google WebComponents

Unread post by dbsoft » 2023-03-08, 00:08

FranklinDM wrote:
2023-03-07, 23:42
Hmm... I suspect it's this commit for the last crash. I was initially hesitant if this change was necessary since I can't reproduce the issue in the provided test case in the linked m-c bug. Can you check if reverting that commit fixes the issue?
That change definitely would cause it, since it removes the NULL check... the Mozilla code did not have that later use of aDocument, so they didn't need the NULL check. If we still need the aDocument->BlockOnload(); code then we need the NULL check in that if() statement.

Or is it possible that GetOurCurrentDoc() would return something different from aDocument? In that case we should store that in a "doc" variable like we do in UnbindFromTree().

FranklinDM: Your PR is exactly what I was testing with... and I haven't crashed since I made that change... although not going to assume that is the only issue at this point due to the various call stacks I've crashed with so far. Thanks for the PR!!
Last edited by dbsoft on 2023-03-08, 04:40, edited 1 time in total.

User avatar
FranklinDM
Add-ons Team
Add-ons Team
Posts: 575
Joined: 2017-01-14, 02:40
Location: Philippines
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by FranklinDM » 2023-03-08, 01:31

dbsoft wrote:
2023-03-08, 00:08
Or is it possible that GetOurCurrentDoc() would return something different from aDocument? In that case we should store that in a "doc" variable like we do in UnbindFromTree().
Yeah, we should do that instead. As for aDocument->BlockOnload();, we still need it since we're not affected by the issue described in bug 1404422 (which is fallout from the async changes in bug 1359833).

EDIT: PR Issue #2147 (UXP)

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35477
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by Moonchild » 2023-03-08, 08:29

dbsoft wrote:
2023-03-08, 00:08
Your PR is exactly what I was testing with... and I haven't crashed since I made that change...
FranklinDM wrote:
2023-03-08, 01:31
PR Issue #2147 (UXP)
:thumbup: :thumbup:

Merged. One less crash to worry about :)
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

dbsoft
Project Contributor
Project Contributor
Posts: 407
Joined: 2020-02-21, 17:35

Re: 32.1.0 beta with Google WebComponents

Unread post by dbsoft » 2023-03-08, 08:54

Address Sanitizer on Mac reported that same crash I had in Windows with some more information:

https://pastebin.com/hS1Rmzdh

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35477
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by Moonchild » 2023-03-08, 09:21

Hmm so it looks like that is a double free operation when GC/CC is run over SharedRegexp.

Filed Issue #2148 (UXP) for this.
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

User avatar
Martok
Contributing developer
Contributing developer
Posts: 5
Joined: 2019-08-29, 00:46

Re: 32.1.0 beta with Google WebComponents

Unread post by Martok » 2023-03-08, 19:04

Just to quickly chime in here, "a few" days late...
back2themoon wrote:
2023-03-01, 17:24
Thank you. I am guessing the Palefill extension needs to be disabled, or even uninstalled before testing this version?
Palefill checks for natively available features before installing fixes, as do the CustomElements polyfill and DarkTree. So in theory, it should "just work" and recognize it needs to do nothing. If something in PF breaks in this beta, that's a bug and needs to be fixed for the next release anyway. (Btw: I can already tell you that version will likely be delayed since I'm away the week 32.1 is currently scheduled)
Mæstro wrote:
2023-03-01, 21:32
Will the release version of 32·1 enable WebComponents by default? If so, will I be able to disable it? My experiences with Fanbox, Pixiv proper and the Archive have made me prefer Palefill.
As per published policy, once a Pale Moon version with native ShadowDOM and ShadowCSS is released, related polyfills will be superseded and no longer maintained. I won't outright remove them so disabling builtin support should get picked up per the mechanism above, but if there are weird interactions these won't be fixed.
"For more information please reread."

User avatar
UCyborg
Fanatic
Fanatic
Posts: 172
Joined: 2019-01-10, 09:37

Re: 32.1.0 beta with Google WebComponents

Unread post by UCyborg » 2023-03-08, 20:45

The crashes you guys talked about, does one of those occur in mozglue.dll at offset 0x10c74, exception code 0xc0000005 ?

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35477
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by Moonchild » 2023-03-08, 21:22

UCyborg wrote:
2023-03-08, 20:45
The crashes you guys talked about, does one of those occur in mozglue.dll at offset 0x10c74, exception code 0xc0000005 ?
Yeah the first one would crash in mozglue because of the double free, as that is where the memory allocator lives. We're working on getting all crashes we know of taken care of before final release.
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

User avatar
Mæstro
Lunatic
Lunatic
Posts: 463
Joined: 2019-08-13, 00:30
Location: Casumia

Re: 32.1.0 beta with Google WebComponents

Unread post by Mæstro » 2023-03-08, 21:53

Martok wrote:
2023-03-08, 19:04
As per published policy, once a Pale Moon version with native ShadowDOM and ShadowCSS is released, related polyfills will be superseded and no longer maintained. I won't outright remove them so disabling builtin support should get picked up per the mechanism above, but if there are weird interactions these won't be fixed.
Firstly, welcome to our board. :thumbup:
Should I therefore disable updating the Palefill extension if I want to keep my current, WebComponents-free behaviour, or would Palefill (1·26) already detect that enabling WebComponents could do such-and-such and refuse to work?

※In principle, I am willing to try the WebComponents versions on the example sites listed, but I only feel safe in so doing knowing I can fall back to the present, comfortable order. I would have tested these sites already in the beta, but I have been too busy with my academic life to back up my profile and look into how to use it in the tarball version.
Browser: Pale Moon (Pusser’s repository for Debian)
Operating System: Linux Mint Debian Edition 4 (amd64)
※Receiving Debian 10 LTS security upgrades
Hardware: HP Pavilion DV6-7010 (1400 MHz, 6 GB)
Formerly user TheRealMaestro: æsc is the best letter.

User avatar
jobbautista9
Keeps coming back
Keeps coming back
Posts: 782
Joined: 2020-11-03, 06:47
Location: Philippines
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by jobbautista9 » 2023-03-10, 10:00

I'm not sure where I should post this, but the latest tree in the master branch currently crashes in this page, both on Windows and GNU/Linux: https://www.cbsnews.com/news/asteroid-2 ... =204708560

Here's the SEGV and backtrace from gdb:

Code: Select all

Thread 1 "palemoon" received signal SIGSEGV, Segmentation fault.
0x00007fb465c5dd63 in mozilla::css::Loader::LoadStyleLink (this=0x7fb441903d40, aElement=aElement@entry=0x0, aURL=0x7fb43df55b60, aTitle=..., aMedia=..., aHasAlternateRel=aHasAlternateRel@entry=false, aCORSMode=mozilla::CORS_NONE, aReferrerPolicy=mozilla::net::RP_No_Referrer_When_Downgrade, aIntegrity=..., aObserver=0x7fb43f1a1400, aIsAlternate=0x7fff32910ee6, aIsExplicitlyEnabled=0x7fff32910ee7) at /home/job/Software/uxp-work/platform/dom/base/nsWrapperCache.h:209
209         return !!(mFlags & aFlag);
(gdb) bt
#0  0x00007fb465c5dd63 in mozilla::css::Loader::LoadStyleLink(nsIContent*, nsIURI*, nsAString_internal const&, nsAString_internal const&, bool, mozilla::CORSMode, mozilla::net::ReferrerPolicy, nsAString_internal const&, nsICSSLoaderObserver*, bool*, bool*)
    (this=0x7fb441903d40, aElement=aElement@entry=0x0, aURL=0x7fb43df55b60, aTitle=..., aMedia=..., aHasAlternateRel=aHasAlternateRel@entry=false, aCORSMode=mozilla::CORS_NONE, aReferrerPolicy=mozilla::net::RP_No_Referrer_When_Downgrade, aIntegrity=..., aObserver=0x7fb43f1a1400, aIsAlternate=0x7fff32910ee6, aIsExplicitlyEnabled=0x7fff32910ee7)
    at /home/job/Software/uxp-work/platform/dom/base/nsWrapperCache.h:209
#1  0x00007fb464c43f70 in nsContentSink::ProcessStyleLink(nsIContent*, nsAString_internal const&, bool, nsAString_internal const&, nsAString_internal const&, nsAString_internal const&)
     (this=0x7fb43f1a1400, aElement=0x0, aHref=<optimized out>, aAlternate=<optimized out>, aTitle=..., aType=..., aMedia=...) at /home/job/Software/uxp-work/platform/dom/base/nsContentSink.cpp:792
#2  0x00007fb464c49998 in nsContentSink::ProcessLinkHeader(nsAString_internal const&) (this=this@entry=0x7fb43f1a1400, aLinkData=...)
    at /home/job/Software/uxp-work/platform/dom/base/nsContentSink.cpp:665
#3  0x00007fb464c49dbf in nsContentSink::DoProcessLinkHeader() (this=this@entry=0x7fb43f1a1400) at /home/job/Software/uxp-work/platform/dom/base/nsContentSink.cpp:367
#4  0x00007fb464c4acea in nsContentSink::WillBuildModelImpl() (this=this@entry=0x7fb43f1a1400) at /home/job/Software/uxp-work/platform/dom/base/nsContentSink.cpp:1607
#5  0x00007fb4645e856d in nsHtml5TreeOpExecutor::WillBuildModel(nsDTDMode) (this=0x7fb43f1a1400, aDTDMode=aDTDMode@entry=eDTDMode_unknown)
    at /home/job/Software/uxp-work/platform/parser/html/nsHtml5TreeOpExecutor.cpp:102
#6  0x00007fb4645ff8be in nsHtml5StreamParser::OnStartRequest(nsIRequest*, nsISupports*) (this=0x7fb43f1a1800, aRequest=<optimized out>, aContext=<optimized out>)
    at /home/job/Software/uxp-work/platform/parser/html/nsHtml5StreamParser.cpp:920
#7  0x00007fb464582cab in nsDocumentOpenInfo::OnStartRequest(nsIRequest*, nsISupports*) (this=0x7fb43f34d6f0, request=0x7fb43e784848, aCtxt=0x0)
    at /home/job/Software/uxp-work/platform/uriloader/base/nsURILoader.cpp:287
#8  0x00007fb46421320f in mozilla::net::nsHttpChannel::CallOnStartRequest() (this=this@entry=0x7fb43e784800) at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:1295
#9  0x00007fb46421377f in mozilla::net::nsHttpChannel::ContinueProcessNormal(nsresult) (rv=<optimized out>, this=0x7fb43e784800)
    at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:2229
#10 mozilla::net::nsHttpChannel::ContinueProcessNormal(nsresult) (this=0x7fb43e784800, rv=<optimized out>) at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:2168
#11 0x00007fb4642138fb in mozilla::net::nsHttpChannel::ProcessNormal() (this=this@entry=0x7fb43e784800) at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:2164
#12 0x00007fb4642145e6 in mozilla::net::nsHttpChannel::ContinueProcessResponse2(nsresult) (this=this@entry=0x7fb43e784800, rv=nsresult::NS_OK)
    at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:2086
#13 0x00007fb4642149f7 in mozilla::net::nsHttpChannel::ContinueProcessResponse2(nsresult) (rv=nsresult::NS_BINDING_FAILED, this=0x7fb43e784800)
    at /home/job/Software/uxp-work/palemoon/obj-x86_64-pc-linux-gnu/dist/include/nsError.h:152
#14 mozilla::net::nsHttpChannel::ContinueProcessResponse1() (this=this@entry=0x7fb43e784800) at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:1943
#15 0x00007fb464214bf8 in mozilla::net::nsHttpChannel::ContinueProcessResponse1() (this=0x7fb43e784800) at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:1864
#16 mozilla::net::nsHttpChannel::ProcessResponse() (this=this@entry=0x7fb43e784800) at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:1863
#17 0x00007fb464214e3a in mozilla::net::nsHttpChannel::OnStartRequest(nsIRequest*, nsISupports*) (this=0x7fb43e784800, request=<optimized out>, ctxt=<optimized out>)
    at /home/job/Software/uxp-work/platform/netwerk/protocol/http/nsHttpChannel.cpp:6307
#18 0x00007fb463fc31e7 in nsInputStreamPump::OnStateStart() (this=this@entry=0x7fb4418c42c0) at /home/job/Software/uxp-work/platform/netwerk/base/nsInputStreamPump.cpp:527
#19 0x00007fb463fc46b8 in nsInputStreamPump::OnInputStreamReady(nsIAsyncInputStream*) (this=0x7fb4418c42c0, stream=<optimized out>)
    at /home/job/Software/uxp-work/platform/netwerk/base/nsInputStreamPump.cpp:429
#20 0x00007fb463f35efc in nsInputStreamReadyEvent::Run() (this=0x7fb43b4fc080) at /home/job/Software/uxp-work/platform/xpcom/io/nsStreamUtils.cpp:94
#21 0x00007fb463f5000c in nsThread::ProcessNextEvent(bool, bool*) (aResult=0x7fff32911f2f, aMayWait=<optimized out>, this=0x7fb46a3417b0)
    at /home/job/Software/uxp-work/platform/xpcom/threads/nsThread.cpp:1140
#22 nsThread::ProcessNextEvent(bool, bool*) (this=0x7fb46a3417b0, aMayWait=<optimized out>, aResult=0x7fff32911f2f) at /home/job/Software/uxp-work/platform/xpcom/threads/nsThread.cpp:1076
#23 0x00007fb463f7177a in NS_ProcessNextEvent(nsIThread*, bool) (aThread=<optimized out>, aThread@entry=0x7fb46a3417b0, aMayWait=aMayWait@entry=true)
    at /home/job/Software/uxp-work/platform/xpcom/glue/nsThreadUtils.cpp:355
#24 0x00007fb464270658 in mozilla::ipc::MessagePump::Run(base::MessagePump::Delegate*) (this=0x7fb463533c40, aDelegate=0x7fb46a3e24e0)
    at /home/job/Software/uxp-work/platform/ipc/glue/MessagePump.cpp:123
#25 0x00007fb464251e55 in MessageLoop::RunInternal() (this=<optimized out>) at /home/job/Software/uxp-work/platform/ipc/chromium/src/base/message_loop.cc:228
#26 MessageLoop::RunHandler() (this=<optimized out>) at /home/job/Software/uxp-work/platform/ipc/chromium/src/base/message_loop.cc:221
#27 MessageLoop::Run() (this=<optimized out>) at /home/job/Software/uxp-work/platform/ipc/chromium/src/base/message_loop.cc:201
#28 0x00007fb465b42578 in nsBaseAppShell::Run() (this=0x7fb45888ee60) at /home/job/Software/uxp-work/platform/widget/nsBaseAppShell.cpp:153
#29 0x00007fb46620a36e in nsAppStartup::Run() (this=0x7fb4588743d0) at /home/job/Software/uxp-work/platform/toolkit/components/startup/nsAppStartup.cpp:282
#30 0x00007fb46622a154 in XREMain::XRE_mainRun() (this=this@entry=0x7fff32912190) at /home/job/Software/uxp-work/platform/toolkit/xre/nsAppRunner.cpp:3811
#31 0x00007fb46622a9fb in XREMain::XRE_main(int, char**, nsXREAppData const*)
    (this=this@entry=0x7fff32912190, argc=argc@entry=4, argv=argv@entry=0x7fff32913548, aAppData=aAppData@entry=0x7fff329123c0)
    at /home/job/Software/uxp-work/platform/toolkit/xre/nsAppRunner.cpp:3921
#32 0x00007fb46622acca in XRE_main(int, char**, nsXREAppData const*, uint32_t) (argc=4, argv=0x7fff32913548, aAppData=0x7fff329123c0, aFlags=<optimized out>)
    at /home/job/Software/uxp-work/platform/toolkit/xre/nsAppRunner.cpp:4003
#33 0x00005622dd5066ca in do_main(int, char**, char**, nsIFile*) (argc=argc@entry=4, argv=argv@entry=0x7fff32913548, envp=envp@entry=0x7fff32913570, xreDirectory=0x7fb46a3a2600)
    at /home/job/Software/uxp-work/palemoon/palemoon/app/nsBrowserApp.cpp:253
#34 0x00005622dd505a29 in main(int, char**, char**) (argc=4, argv=0x7fff32913548, envp=0x7fff32913570) at /home/job/Software/uxp-work/palemoon/palemoon/app/nsBrowserApp.cpp:377
Image

merry mimas

XUL add-ons developer. You can find a list of add-ons I manage at http://rw.rs/~job/software.html.

Mima avatar by 絵虎. Pixiv post: https://www.pixiv.net/en/artworks/15431817

Image

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35477
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by Moonchild » 2023-03-10, 10:26

jobbautista9 wrote:
2023-03-10, 10:00
I'm not sure where I should post this
Here is fine actually. we've been using this thread to report any issues prior to release.

verified on master. it's shadow DOM related.

Code: Select all

The thread 0x6680 has exited with code 0 (0x0).
Exception thrown: read access violation.
**aElement** was nullptr.
Loader::LoadStyleLink in Loader.cpp gets passed in aElement. Apparently that can be null, but isn't checked for further down.
This results in a null dereferencing crash. Should be easy enough to fix or at least hotpatch.

EDIT: straightforward fix in https://repo.palemoon.org/MoonchildProd ... 4f32e5b558 fixes that crash.
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

User avatar
jobbautista9
Keeps coming back
Keeps coming back
Posts: 782
Joined: 2020-11-03, 06:47
Location: Philippines
Contact:

Re: 32.1.0 beta with Google WebComponents

Unread post by jobbautista9 » 2023-03-10, 12:31

Thanks, can confirm the fix resolved the crash!
Image

merry mimas

XUL add-ons developer. You can find a list of add-ons I manage at http://rw.rs/~job/software.html.

Mima avatar by 絵虎. Pixiv post: https://www.pixiv.net/en/artworks/15431817

Image

Locked