After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 319873 - [gnomevfssrc] lockup if ctrl-c is pressed at just the right time
[gnomevfssrc] lockup if ctrl-c is pressed at just the right time
Status: RESOLVED DUPLICATE of bug 338827
Product: GStreamer
Classification: Platform
Component: gst-plugins-base
git master
Other Linux
: Normal minor
: git master
Assigned To: Andy Wingo
GStreamer Maintainers
Depends on: 338827
Blocks:
 
 
Reported: 2005-10-26 17:00 UTC by Andy Wingo
Modified: 2006-10-22 12:14 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Andy Wingo 2005-10-26 17:00:58 UTC
here's a backtrace of gst-launch gnomevfssrc ! decodebin ! xvimagesink, only one
thread:

  • #0 __select_nocancel
    from /lib/libc.so.6
  • #1 gnome_vfs_inet_connection_get_address
    from /usr/lib/libgnomevfs-2.so.0
  • #2 gnome_vfs_socket_buffer_new
    from /usr/lib/libgnomevfs-2.so.0
  • #3 gnome_vfs_socket_buffer_read_until
    from /usr/lib/libgnomevfs-2.so.0
  • #4 ne_sock_readline
    from /usr/lib/gnome-vfs-2.0/modules/libhttp.so
  • #5 ne_request_dispatch
    from /usr/lib/gnome-vfs-2.0/modules/libhttp.so
  • #6 ne_begin_request
    from /usr/lib/gnome-vfs-2.0/modules/libhttp.so
  • #7 ??
    from /usr/lib/gnome-vfs-2.0/modules/libhttp.so
  • #8 ??
    from /usr/lib/gnome-vfs-2.0/modules/libhttp.so
  • #9 gnome_vfs_open_uri_cancellable
    from /usr/lib/libgnomevfs-2.so.0
  • #10 gst_gnomevfssrc_start
    at gstgnomevfssrc.c line 1090
  • #11 gst_base_src_start
    at gstbasesrc.c line 936
  • #12 gst_base_src_activate_pull
    at gstbasesrc.c line 1088
  • #13 gst_pad_activate_pull
    at gstpad.c line 691
  • #14 gst_pad_activate_pull
    at gstpad.c line 681
  • #15 gst_ghost_pad_internal_do_activate_pull
    at gstghostpad.c line 534
  • #16 gst_pad_activate_pull
    at gstpad.c line 691
  • #17 gst_pad_activate_pull
    at gstpad.c line 681
  • #18 gst_type_find_element_activate
    at gsttypefindelement.c line 794
  • #19 gst_pad_set_active
    at gstpad.c line 613
  • #20 activate_pads
    at gstelement.c line 2203
  • #21 gst_iterator_fold
    at gstiterator.c line 468
  • #22 iterator_fold_with_resync
    at gstelement.c line 2221
  • #23 gst_element_pads_activate
    at gstelement.c line 2264
  • #24 gst_element_change_state_func
    at gstelement.c line 2301

not really sure what's going on here tho. just noting the bug.
Comment 1 Andy Wingo 2005-10-26 17:04:27 UTC
actually i'm seeing this without ctrl-c, renaming the bug. with this pipeline:

gst-launch gnomevfssrc location=http://localhost:8800/ ! decodebin ! xvimagesink

something is broadcasting on 8800, but it could be that no data is coming out...
Comment 2 Wim Taymans 2006-03-09 15:51:05 UTC
This can only be made to work if we use async gnomevfs handling and gnomevfssrc returns ASYNC from the state change to PAUSED. Is this worth it?
Comment 3 Alex Lancaster 2006-10-09 02:09:16 UTC
(In reply to comment #2)
> Is this worth it?

Yes, definitely, it causes problems in apps like rhythmbox, see bug #313049 comment #76. 
Comment 4 Wim Taymans 2006-10-09 08:54:26 UTC
A simple alternative would be to do nothing in the _start method, but to open the connection right before producing data in the streaming thread.
the pipeline will appear to not complete the state change to PAUSED (as is now) but the application will be able to cancel the operation.
Comment 5 René Stadler 2006-10-11 18:49:17 UTC
This is closely related or even identical to bug #338827 (posted patch there), maybe either should be closed to keep the discussion central.
Comment 6 Alex Lancaster 2006-10-18 17:05:22 UTC
(In reply to comment #5)
> This is closely related or even identical to bug #338827 (posted patch there),
> maybe either should be closed to keep the discussion central.

Yep, this bug should be closed as a duplicate of bug #338827.  If there are no objections, I'll do it.

Comment 7 Wim Taymans 2006-10-22 12:14:30 UTC

*** This bug has been marked as a duplicate of 338827 ***