Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|

Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

ISHIKAWA,chiaki
I filed a bug regarding an assertion failure that appears suddenly after
a week or so blank period of testing locally.

(It is triggered in many places it seems, and so I had to abort the
whole |make mozmill| test.)


https://bugzilla.mozilla.org/show_bug.cgi?id=1573939

Bug 1573939
Assertion failure: sheet->ParsingMode() == aParsingMode, at
/NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

Does this ring a bell?

TIA

Chiaki

_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

Jörg Knobloch
On 14 Aug 2019 20:35, ISHIKAWA,chiaki wrote:
> Assertion failure: sheet->ParsingMode() == aParsingMode, at
> /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989
>
> Does this ring a bell?

You filed a duplicate ;-)

_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

ISHIKAWA,chiaki
On 2019/08/15 5:59, Jörg Knobloch wrote:
> On 14 Aug 2019 20:35, ISHIKAWA,chiaki wrote:
>> Assertion failure: sheet->ParsingMode() == aParsingMode, at
>> /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989
>>
>> Does this ring a bell?
>
> You filed a duplicate ;-)
>

Did I? bugzilla became unresponsive on the first try and when I checked
my dashboard to see if the attempt to post succeeded, I did not see the
post and so I retried. Hmm... Let me check.


> _______________________________________________
> dev-apps-thunderbird mailing list
> [hidden email]
> https://lists.mozilla.org/listinfo/dev-apps-thunderbird
>
>

_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

ISHIKAWA,chiaki
Hi,

Now I see what you mean.

https://bugzilla.mozilla.org/show_bug.cgi?id=1573639

Almost total MozMill failure in debug builds on 2019-08-13


I initially thought I posted the same/similar content twice, thus
posting a duplicate myself :-)
I have no idea what was wrong with bugzilla.

Makes me wonder, though, whether the current work flow for incorporating
patches checks that the
new patch does not break testing (Hmm. Maybe checking only FF tree builds.)

One other thing:

Jorg, would it not be better to have the descriptive string from a
compilation or test failure,

|sheet->ParsingMode() == aParsingMode in this case in the title of the
bugzilla entry so that searching for it is easy? I usually use the
simple search function and with it, I cannot find the bug 1573639 when I
type aParsingMode as search string. (Strange, I can't even find my own
post. Hmm...) Using advanced search, I can find MY OWN duplicate
bugzilla entry, https://bugzilla.mozilla.org/show_bug.cgi?id=1573939 ||when I type aParsingMode. I still cannot find your |
https://bugzilla.mozilla.org/show_bug.cgi?id=1573639

Or is there a clever setting of search to make it possible to find bug 1573639 by typing, say,  aParsingMode?

Anyway, thank you for the pointer to the bugzilla entry.
I wish I could find it earlier. Thus my comment and question above.

Chiaki

||


On 2019/08/15 8:23, ISHIKAWA,chiaki wrote:

> On 2019/08/15 5:59, Jörg Knobloch wrote:
>> On 14 Aug 2019 20:35, ISHIKAWA,chiaki wrote:
>>> Assertion failure: sheet->ParsingMode() == aParsingMode, at
>>> /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989
>>>
>>> Does this ring a bell?
>>
>> You filed a duplicate ;-)
>>
>
> Did I? bugzilla became unresponsive on the first try and when I
> checked my dashboard to see if the attempt to post succeeded, I did
> not see the
> post and so I retried. Hmm... Let me check.
>
_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

Jörg Knobloch
On 16 Aug 2019 00:35, ISHIKAWA,chiaki wrote:
> Makes me wonder, though, whether the current work flow for
> incorporating patches checks that the
> new patch does not break testing (Hmm. Maybe checking only FF tree
> builds.)

FF land stuff all the time and bust us, that's nothing new, see
https://treeherder.mozilla.org/#/jobs?repo=comm-central right now, more
bustages than we had for a while.

As for the bug summary: You're right, it should have been more
descriptive. But usually no one cares and my workflow is that I start
with the "massive test failure" and forget the details. I'll try to do
better next time.

Jörg.

_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

ISHIKAWA,chiaki
Hi Jörg.

On 2019/08/16 7:41, Jörg Knobloch wrote:

> On 16 Aug 2019 00:35, ISHIKAWA,chiaki wrote:
>> Makes me wonder, though, whether the current work flow for
>> incorporating patches checks that the
>> new patch does not break testing (Hmm. Maybe checking only FF tree
>> builds.)
>
> FF land stuff all the time and bust us, that's nothing new, see
> https://treeherder.mozilla.org/#/jobs?repo=comm-central right now,
> more bustages than we had for a while.
>
>

Looks bad.: those failures really makes it hard for me to figure out if
my local changes break TB or not.

Now for the last two days, but I have a tough time simply compiling TB.

Is it only me who sees the following COMPILE TIME FAILURE after updating
the source tree in the last 48 hours or so?
I bet that there are issues with CPP_THROW_NEW macro although I cannot
find the reference to related issues of CPP_THROW_NEW despite there are
old CPP_THROW_NEW entries in bugzilla.
E.g.
https://bugzilla.mozilla.org/show_bug.cgi?id=304982
CPP_THROW_NEW should be defined to throw() even if exceptions are enabled

https://bugzilla.mozilla.org/show_bug.cgi?id=1295192
micro-optimize layout code by removing null checks associated with
placement operator new
(This bug refers to the definition of CPP_THROW_NEW.)
I wish bugzilla search can be made more powerful. I found the above two
entries by google search :-(

Anyway, the compilation bug:
~~~~~~~~~
  3:54.17 comm/mailnews/news/src
  3:59.02 comm/db/mork/src
  3:59.75 In file included from
/NREF-COMM-CENTRAL/mozilla/comm/db/mork/src/morkArray.cpp:17:
  3:59.75 /NREF-COMM-CENTRAL/mozilla/comm/db/mork/src/morkNode.h:157:33:
error: expected ‘;’ at end of member declaration
  3:59.75                       morkEnv* ev) CPP_THROW_NEW {
  3:59.75                                  ^
  3:59.75                                   ;
  3:59.75 /NREF-COMM-CENTRAL/mozilla/comm/db/mork/src/morkNode.h:157:35:
error: ‘CPP_THROW_NEW’ does not name a type
  3:59.75                       morkEnv* ev) CPP_THROW_NEW {
  3:59.75                                    ^~~~~~~~~~~~~
  3:59.75 In file included from
/NREF-COMM-CENTRAL/mozilla/comm/db/mork/src/morkPool.h:14,
  3:59.75                  from
/NREF-COMM-CENTRAL/mozilla/comm/db/mork/src/morkEnv.h:18,
  3:59.76                  from
/NREF-COMM-CENTRAL/mozilla/comm/db/mork/src/morkArray.cpp:21:
  3:59.76 /NREF-COMM-CENTRAL/mozilla/comm/db/mork/src/morkDeque.h:59:33:
error: expected ‘;’ at end of member declaration
  3:59.76                       morkEnv* ev) CPP_THROW_NEW {
  3:59.76                                  ^

I have no clue. Maybe M-C stopped defining CPP_THROW_NEW ?

 > As for the bug summary: You're right, it should have been more
descriptive. But usually no one cares and my workflow is that I start
with the "massive test failure" and forget the details. I'll try to do
better next time.

Thanks. I noticed you changed the title.

"usually no one cares": well, there may be a handful: aceman, ...
hmm...,  and me. I am not sure whether many developers compile and test
locally. At least I do.


>
> Jörg.
>
> _______________________________________________
> dev-apps-thunderbird mailing list
> [hidden email]
> https://lists.mozilla.org/listinfo/dev-apps-thunderbird


_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

Jörg Knobloch
On 16 Aug 2019 16:54, ISHIKAWA,chiaki wrote:
>
> Looks bad.: those failures really makes it hard for me to figure out
> if my local changes break TB or not.
>
> Now for the last two days, but I have a tough time simply compiling TB.

Yes, that assert is annoying. It's also affecting my development work,
so I had to remove it locally:

-      MOZ_ASSERT(sheet->ParsingMode() == aParsingMode);
+      // MOZ_ASSERT(sheet->ParsingMode() == aParsingMode);

The compilation error related to CPP_THROW_NEW was fixed yesterday, see:
https://hg.mozilla.org/comm-central/rev/82fde34332727c5a9e1128299cba660089b72289

Keep in mind that we are constantly fixing bustage, so you need to
rebase frequently.

Jörg.

_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

ISHIKAWA,chiaki
Dear Jorg,

I figured out the CPP_THROW_NEW bug quickly after my own post, but thank
you for posting the patch already.

On 2019/08/17 0:11, Jörg Knobloch wrote:

> On 16 Aug 2019 16:54, ISHIKAWA,chiaki wrote:
>>
>> Looks bad.: those failures really makes it hard for me to figure out
>> if my local changes break TB or not.
>>
>> Now for the last two days, but I have a tough time simply compiling TB.
>
> Yes, that assert is annoying. It's also affecting my development work,
> so I had to remove it locally:
>
> -      MOZ_ASSERT(sheet->ParsingMode() == aParsingMode);
> +      // MOZ_ASSERT(sheet->ParsingMode() == aParsingMode);
>
I tried to see if the patch for this landed by updating M-C portion
only, but it is not yet officially,

and instead bitten by the cleanup work related to
CPP_THROW_NEW.

> The compilation error related to CPP_THROW_NEW was fixed yesterday, see:
> https://hg.mozilla.org/comm-central/rev/82fde34332727c5a9e1128299cba660089b72289 
>
>
> Keep in mind that we are constantly fixing bustage, so you need to
> rebase frequently.
>
>
Thank you again for the quick fix.
This is a bit annoying, but nobody has promised us a rose garden, I
suppose. Hm...

Chiaki

> Jörg.
>
> _______________________________________________
> dev-apps-thunderbird mailing list
> [hidden email]
> https://lists.mozilla.org/listinfo/dev-apps-thunderbird


_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird
Reply | Threaded
Open this post in threaded view
|

Re: Bug 1573939, Assertion failure: sheet->ParsingMode() == aParsingMode, at /NREF-COMM-CENTRAL/mozilla/layout/style/Loader.cpp:989

Jörg Knobloch
On 16 Aug 2019 19:47, ISHIKAWA,chiaki wrote:
Yes, that assert is annoying. It's also affecting my development work, so I had to remove it locally:

-      MOZ_ASSERT(sheet->ParsingMode() == aParsingMode);
+      // MOZ_ASSERT(sheet->ParsingMode() == aParsingMode);

I tried to see if the patch for this landed by updating M-C portion only, but it is not yet officially,

Emilio told me that the issue should be fixed now, see https://bugzilla.mozilla.org/show_bug.cgi?id=1573639#c4.

We'll see how the last C-C push turns out.


_______________________________________________
dev-apps-thunderbird mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-apps-thunderbird