Previously the default and full modes were the same. Now the default
mode is like before: it accumulates all buffers in a buffer list until
the threshold is reached and then writes them all out, potentially in
multiple writes.
The new full mode works by always copying memory to a single memory area
and writing everything out with a single write once the threshold is
reached.
This seems to happen when another client is accessing the file at the
same time, and retrying after a short amount of time solves it.
Sometimes partial data is written at that point already but we have no
idea how much it is, or if what was written is correct (it sometimes
isn't) so we always first seek back to the current position and repeat
the whole failed write.
It happens at least on Linux and macOS on SMB/CIFS and NFS file systems.
Between write attempts that failed with EACCES we wait 10ms, and after
enough consecutive tries that failed with EACCES we simply time out.
In theory a valid EACCES for files to which we simply have no access
should've happened already during the call to open(), except for NFS
(see open(2)).
This can be enabled with the new max-transient-error-timeout property, and
a new o-sync boolean property was added to open the file in O_SYNC mode
as without that it's not guaranteed that we get EACCES for the actual
writev() call that failed but might only get it at a later time.
Fixes https://gitlab.freedesktop.org/gstreamer/gstreamer/issues/305
fflush() has no effect because we use writev() directly, so fsync()
should be used instead which is actually flushing the kernel-side
buffers.
As a next step, a non-line-buffered buffering mode is to be added.
https://bugzilla.gnome.org/show_bug.cgi?id=794173
Original commit message from CVS:
Based on Patch by: Laurent Glayal <spglegle at yahoo dot fr>
* configure.ac:
Check for stdio_ext.h for the filesink changes.
* plugins/elements/gstfilesink.c: (buffer_mode_get_type),
(gst_file_sink_class_init), (gst_file_sink_init),
(gst_file_sink_dispose), (gst_file_sink_set_property),
(gst_file_sink_get_property), (gst_file_sink_open_file),
(gst_file_sink_close_file):
* plugins/elements/gstfilesink.h:
Add two properties to control the buffering mode and size.
API: GstFileSink::buffer-mode
API: GstFileSink::buffer-size
Original commit message from CVS:
Patch by: Mark Nauwelaerts <manauw at skynet be>
* plugins/elements/gstfilesink.c: (gst_file_sink_open_file),
(gst_file_sink_query), (gst_file_sink_do_seek),
(gst_file_sink_get_current_offset), (gst_file_sink_render):
* plugins/elements/gstfilesink.h:
Fix position reporting; rename data_written member to current_pos to
reflect its real meaning (fixes#412648).
Original commit message from CVS:
* plugins/elements/gstfilesink.c: (gst_file_sink_open_file),
(gst_file_sink_close_file), (gst_file_sink_do_seek),
(gst_file_sink_event), (gst_file_sink_render):
* plugins/elements/gstfilesink.h:
Check if we can seek in the file instead of assuming
we always can. Post an error when we are asked to seek in a
non-seekable file (like a fifo). Fixes#343312.
Some cleanups.
Original commit message from CVS:
implement URI schemes
Elements can now register as a source or sink for a protocol and applications can use gst_element_make_from_uri () to get an element that handles a given URI.
This patch provides:
- removal of old broken URI handling scheme.
- new URI handling using interfaces.
- updates for registry to save handled URIs.
- interface for URI handlers.
- implementation of that in filesrc and filesink for the file:// URI
- extension to pipeline parsing to allow specifying only a URI instead of element
Does not include:
- tests
- inclusion in docs build
Original commit message from CVS:
GST_DEBUG reorganization
This is a big diff (ca 450k), containing loads of stuff:
- gstinfo.[ch] complete rewrite
- changing of all GST_DEBUG messages to reflect that change
- reorganization of subsystem disabling
- addition of gstconfig.h.in so we can track the disablings
- <gst/gst.h> does not include <unistd.h> and <config.h> anymore
- documentation updated for gstinfo stuff (build the docs yourself to know what changed)
- bugfixes for making of the docs (files from CVS are not deleted anymore
- testsuite for debugging changes in testsuite/debug
expect breakage
Original commit message from CVS:
I don't know how I ever got to introducing this madness, but I was wrong. *REALLY* wrong. It hasn't even worked, it still doesn't work and it will never work because the whole concept of doing a multifile output in filesink is completely braindead. Signal EOS, close file, set new filename and reset to PLAYING is so much easier. This patch removes all the multifile output cruft that I introduced and makes filesink simply work.
