search for: exportname

...the "protocol: " line. Thus it's not correct to test for the output file being non-empty. Remove these lines which caused the test to fail. However I replaced them with "cat" for diagnostic purposes. Fixes: commit 7623b2cc45078cca88fdd2d96c70c7f82a0db49d --- tests/test-exportname.sh | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/test-exportname.sh b/tests/test-exportname.sh index 2ec45023..54e04f86 100755 --- a/tests/test-exportname.sh +++ b/tests/test-exportname.sh @@ -172,12 +172,12 @@ test "$(jq -c "$query" exportname.out)&qu...

....open is called only with an export name that the plugin was willing to advertise, but for that, we'll need a way to call the plugin's .list_exports from within the context of .open (since we can't guarantee the guest will call NBD_OPT_LIST). So for now, strict mode requires redundant exportname= command line parameters. It may also be worth adding a way to pass exportname-file=/path/to/file, which then gets parsed the same was as the sh plugin (right now with NAMES, INTERLEAVED, or NAMES+DESCRIPTIONS), to avoid having to do so many export names directly on the command line. Signed-off-b...

...y with an export name that the plugin was willing to > advertise, but for that, we'll need a way to call the plugin's > .list_exports from within the context of .open (since we can't > guarantee the guest will call NBD_OPT_LIST). So for now, strict mode > requires redundant exportname= command line parameters. It may also > be worth adding a way to pass exportname-file=/path/to/file, which > then gets parsed the same was as the sh plugin (right now with NAMES, > INTERLEAVED, or NAMES+DESCRIPTIONS), to avoid having to do so many > export names directly on the command...

...gh the layers this commit makes several changes: The filter .open callback now takes an extra parameter, the export name. This is always non-NULL (for oldstyle it is ""). This string has a short lifetime and filters that need to hang on to it must take a copy. The filter must pass the exportname parameter down to the next layer (or potentially modify it). The plugin .open callback is unchanged, but the binding to it in server/plugins.c takes a copy of the exportname in the handle and this is what nbdkit_export_name returns to plugins. Eventually we will deprecate that function and the V3...

When using ext2file=exportname to pick the file to server from the client's export name, we do not want to leak the underlying plugin's export list. While touching this, take advantage of string lifetimes via nbdkit_strdup_intern and similar for less cleanup bookkeeping. Note that we don't actually implement a full...

Here's changes to the file plugin (which I'm happy with) and a new exportname filter (which is still at RFC stage; I need to finish implementing strict mode in .open, and add tests). I also discovered that we really want .list_exports and .open to know when they are used on plaintext vs. tls clients for --tls=on, and we may want to split out a new .default_export callback r...

...nbdkit-filter.pod index 12343dbf..b6ed5504 100644 --- a/docs/nbdkit-filter.pod +++ b/docs/nbdkit-filter.pod @@ -403,7 +403,7 @@ Returns a copy of the C<i>'th export. =head2 C<.open> void * (*open) (nbdkit_next_open *next, void *nxdata, - int readonly, const char *exportname); + int readonly, const char *exportname, int is_tls); This is called when a new client connection is opened and can be used to allocate any per-connection data structures needed by the filter. @@ -420,8 +420,9 @@ error message and return C<NULL>. This callback is optiona...

...makes several changes: > > The filter .open callback now takes an extra parameter, the export > name. This is always non-NULL (for oldstyle it is ""). This string > has a short lifetime and filters that need to hang on to it must take > a copy. The filter must pass the exportname parameter down to the > next layer (or potentially modify it). > > The plugin .open callback is unchanged, but the binding to it in > server/plugins.c takes a copy of the exportname in the handle and this > is what nbdkit_export_name returns to plugins. Eventually we will > depr...

...nothing at all. There are some visible but very minor changes resulting from this commit: (1) NBD_OPT_LIST advertises a single export called "" (instead of the -e parameter). We intend to replace this implementation soon with a correct one. (2) The --run option no longer sets $exportname (to -e) nor puts the export name into the $uri. However this was always the wrong thing to do since export names are per connection, not per server. Existing --run scripts will see $exportname expand to "" which is most likely what they saw before and overwhelmingly more...

Applies on top of my pending series for the exportname filter, addressing one of the todo's in that cover letter. Eric Blake (2): filters: Add .export_description wrappers ext2: Supply .list_exports and .default_export filters/ext2/nbdkit-ext2-filter.pod | 3 +- tests/Makefile.am | 16 +++- filters/ext2/ext2.c...

This is a revision of my .default_export work, plus new work on .export_descriptions and a new exportname filter. I think it is now ready to check in. Things I'd still like in 1.22: - the file plugin should implement .list_exports (patch already posted, but it needs rebasing on this series) - the ext2 filter should override .list_exports when in exportname mode - the nbd plugin should be able to...

...use in source and binary forms, with or without * modification, are permitted provided that the following conditions are @@ -113,45 +113,6 @@ struct retry_handle { bool open; }; -static void * -retry_open (nbdkit_next_open *next, nbdkit_context *nxdata, - int readonly, const char *exportname, int is_tls) -{ - struct retry_handle *h; - - if (next (nxdata, readonly, exportname) == -1) - return NULL; - - h = malloc (sizeof *h); - if (h == NULL) { - nbdkit_error ("malloc: %m"); - return NULL; - } - - h->readonly = readonly; - h->exportname = strdup (exportna...

...We intend to replace this implementation soon with a > correct one. And in the meantime, we still allow clients to connect with any name at all, and that still makes no difference for plugins that don't read the client's request. > > (2) The --run option no longer sets $exportname (to -e) nor puts the > export name into the $uri. However this was always the wrong > thing to do since export names are per connection, not per server. > Existing --run scripts will see $exportname expand to "" which is > most likely what they saw before...

...hard W.M. Jones wrote: > >+++ b/server/plugins.c > >@@ -278,12 +278,30 @@ plugin_preconnect (struct backend *b, int readonly) > > } > > static void * > >-plugin_open (struct backend *b, int readonly) > >+plugin_open (struct backend *b, int readonly, const char *exportname) > > { > >+ GET_CONN; > > struct backend_plugin *p = container_of (b, struct backend_plugin, backend); > > assert (p->plugin.open != NULL); > >+ /* Save the exportname since the lifetime of the string passed in > >+ * here is likely to be brief. In...

...dkit_use_default_export' 003/14:[0059] [FC] 'server: Respond to NBD_INFO_NAME request' 004/14:[0021] [FC] 'sh, eval: Implement .default_export' 005/14:[0036] [FC] 'api: Alter .list_exports' 006/14:[0032] [FC] 'api: Add .export_description' 007/14:[0015] [FC] 'exportname: New filter' 008/14:[0049] [FC] 'filters: Add .export_description wrappers' 009/14:[0002] [FC] 'ext2: Supply .list_exports and .default_export' 010/14:[----] [--] 'nbd: Implement .default_export, .export_description' 011/14:[0010] [FC] 'nbd: Add dynamic-export=true o...

...ut improve the documentation to mention that -e is now useful _only_ with --run, and only if the client uses it. This partially reverts commit e71178e9b71403, although for -Wshadow reasons, we now have to name the global variable export_name due to other code in the meantime using parameters named exportname. Signed-off-by: Eric Blake <eblake@redhat.com> --- docs/nbdkit-captive.pod | 12 +++++++++++- docs/nbdkit-protocol.pod | 2 +- docs/nbdkit.pod | 18 ++++++++++++++++++ docs/synopsis.txt | 3 ++- server/internal.h | 1 + server/captive.c | 20 ++++++++++++++...

On Tuesday 11 March 2014 23:13:46 Richard W.M. Jones wrote: > diff --git a/src/drives.c b/src/drives.c > index 2c85b52..68e37f7 100644 > --- a/src/drives.c > +++ b/src/drives.c > @@ -115,7 +115,8 @@ static struct drive * > create_drive_file (guestfs_h *g, const char *path, > bool readonly, const char *format, > const char *iface,

...lter registers a callback but in that callback it doesn't call the C<next> function then the corresponding method in the @@ -450,8 +451,10 @@ requires write access to the underlying device in case a journal needs to be replayed for consistency as part of the mounting process. The C<exportname> string is only guaranteed to be available during the -call. If the filter needs to use it (other than immediately passing -it down to the next layer) it must take a copy. The C<exportname> and +call (different than the lifetime for the return of C<nbdkit_export_name> +used by plug...

...lter registers a callback but in that callback it doesn't call the C<next> function then the corresponding method in the @@ -450,8 +451,10 @@ requires write access to the underlying device in case a journal needs to be replayed for consistency as part of the mounting process. The C<exportname> string is only guaranteed to be available during the -call. If the filter needs to use it (other than immediately passing -it down to the next layer) it must take a copy. The C<exportname> and +call (different than the lifetime for the return of C<nbdkit_export_name> +used by plug...

Note this patch requires a non-upstream qemu patch that I've been experimenting with. See qemu-devel list. Rich.