edbrowse-dev - development list for edbrowse
 help / color / mirror / Atom feed
* [Edbrowse-dev] js shell
@ 2014-02-08 14:56 Karl Dahlke
  2014-02-08 16:16 ` Chris Brannon
  0 siblings, 1 reply; 2+ messages in thread
From: Karl Dahlke @ 2014-02-08 14:56 UTC (permalink / raw)
  To: Edbrowse-dev

Again, wondering why edbrowse works with our compiled js 24 library
but not with their distributed library.
Has anyone looked at their shell application?

/usr/local/mozjs-24.2.0/js/src/shell/js.cpp

If you take this stand alone program and compile it
against the debian library does it run?
If not I would think that worthy of a bug report.
If yes then there is information there.

I did a grep for NewRuntime
and there are only a couple calls between that and NewContext,
which is the point where we segfault.
I don't think the intervening calls are important,
although SetNativeStackQuota might be,
don't know what that does.

Well it seemed like something to look at,
an application that they provide, that uses their library,
that is suppose to work.

Karl Dahlke

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [Edbrowse-dev] js shell
  2014-02-08 14:56 [Edbrowse-dev] js shell Karl Dahlke
@ 2014-02-08 16:16 ` Chris Brannon
  0 siblings, 0 replies; 2+ messages in thread
From: Chris Brannon @ 2014-02-08 16:16 UTC (permalink / raw)
  To: Edbrowse-dev

Karl Dahlke <eklhad@comcast.net> writes:

> If you take this stand alone program and compile it
> against the debian library does it run?

It isn't quite standalone, as it uses headers private to mozjs, some of
which are autogenerated.  I can't seem to get it to build without
building the whole libmozjs.
On Debian, the libmozjs source is intertwined with the source for
iceweasel, which is basically their rebranded firefox.  So in order to
build the js shell, I have to build a whole web browser!
Or at the very least, I need to get through its ./configure phase.
Unfortunately, that's going to require installing a bunch of development
packages for GUI libraries, and my VM disk image just doesn't have enough
space for that.

They do package the JS shell, and it works fine,
but it is statically linked against Spidermonkey.  IMHO, that isn't a
good test of whether the dynamic libmozjs library is working properly.

A better question is this.  Do they have any package that does link
against the libmozjs dynamic library?
Yes they do.

Script started on Sat Feb  8 08:12:07 2014
chris@somehost:~$ apt-cache rdepends libmozjs24d
libmozjs24d
Reverse Depends:
  xulrunner-24.0
  libmozjs24d-dbg
  libmozjs-dev
chris@somehost:~$ apt-cache rdepends xulrunner-24.0
xulrunner-24.0
Reverse Depends:
  zotero-standalone
 |tuxguitar
  xulrunner-dev
  xulrunner-24.0-dbg
  iceweasel
 |conkeror
chris@somehost:~$ exit
exit

Script done on Sat Feb  8 08:12:48 2014

So presumably all of that stuff that depends on xulrunner -- and
indirectly on libmozjs -- works just fine.  Can we conclude that their
package doesn't have a bug and we're doing something wrong in JS
initialization?

-- Chris

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2014-02-08 16:17 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-02-08 14:56 [Edbrowse-dev] js shell Karl Dahlke
2014-02-08 16:16 ` Chris Brannon

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).