search for: can_write

...od exercise, because I really like how it forced me to spend a good hour on the commit message of patch 1 explaining things, and how patch 2 then fell out in 10 minutes of hacking to fix the problem identified in passing in the patch 1 commit message. Eric Blake (2): server: Cache per-connection can_write server: Remember .open(readonly) status server/internal.h | 5 ++- server/backend.c | 43 ++++++++++++++++++++- server/connections.c | 5 ++- server/filters.c | 6 +-- server/plugins.c | 4 +- server/protocol-handshake.c | 74 +++++++++++++++...

...t. An alternate way I can think to fix this would be for the core server to maintain a readonly flag for each layer (instead of just per- server). You could also argue that our readonly test in server/connections.c:compute_eflags is wrong and/or that the implementations of server/plugins.c:plugin_can_write and server/filters.c:filter_can_write have the wrong default - they should consult the readonly flag which was passed to that layer's open call. However the fix only affects one filter, and we're not maintaining an API for filters, so maybe we can defer the problem. Rich.

...((0 + $(od --endian=big -An -N2 -d < eflags.out))) + printf "eflags 0x%04x" $eflags + for f in $all_flags; do + [ $(( eflags & ${!f} )) -ne 0 ] && echo -n " $f" + done + echo +} + +fail () +{ + echo "$@" + exit 1 +} + +# no -r +# can_write=false + +do_nbdkit <<'EOF' +case "$1" in + get_size) echo 1M ;; + *) exit 2 ;; +esac +EOF + +[ $eflags -eq $(( HAS_FLAGS|READ_ONLY )) ] || + fail "expected HAS_FLAGS|READ_ONLY" + +# -r +# can_write=false + +do_nbdkit -r <<'EOF' +case "...

...this error: nbdkit: file[1]: debug: xz: pwrite count=4096 offset=1048576 flags=0x0 nbdkit: file[1]: debug: pwrite count=4096 offset=1048576 fua=0 nbdkit: file[1]: error: pwrite: Bad file descriptor nbdkit: file[1]: debug: sending error reply: Bad file descriptor Since the filter did not intercept can_write, it was passed through to the layer below (the file plugin). The filter opens the file plugin passing readonly=1, but in the case where there are no filters, nbdkit core would never call can_write() (assuming that because we opened the server readonly, it would return false). However with the fil...

The previous patch argued that globally affecting .can_write based on '-r' (the global 'readonly') prevents the ability for a filter to call next_open(nxdata, true) to purposefully write to the plugin, even while advertising .can_write=0 to the client. But it also demonstrated that when a filter passes false to next_open, we were still makin...

...plugin.pod b/plugins/python/nbdkit-python-plugin.pod index 51e0f57..0fd4dcb 100644 --- a/plugins/python/nbdkit-python-plugin.pod +++ b/plugins/python/nbdkit-python-plugin.pod @@ -184,25 +184,25 @@ contents will be garbage collected. def get_size(h): # return the size of the disk -=item C<can_write> +=item C<is_rotational> (Optional) - def can_write(h): + def is_rotational(h): # return a boolean -=item C<can_flush> +=item C<can_write> (Optional) - def can_flush(h): + def can_write(h): # return a boolean -=item C<is_rotational> +=item C<can_f...

...lugin.pod > index 51e0f57..0fd4dcb 100644 > --- a/plugins/python/nbdkit-python-plugin.pod > +++ b/plugins/python/nbdkit-python-plugin.pod > @@ -184,25 +184,25 @@ contents will be garbage collected. > def get_size(h): > # return the size of the disk > > -=item C<can_write> > +=item C<is_rotational> > > (Optional) > > - def can_write(h): > + def is_rotational(h): > # return a boolean > > -=item C<can_flush> > +=item C<can_write> > > (Optional) > > - def can_flush(h): > + def can...

Hi - the attached patch provides implementations and tests for sys::fs::can_read() and sys::fs::can_write() for Unix. These were adapted from the previous implementations in lib/Support/Unix/Path.inc. can_execute() could be implemented quite similarly, but I wasn't sure where the best place to add a test for it would be, so left it out for now. Thanks, Liam -------------- next part --------------...

v1 was here: https://www.redhat.com/archives/libguestfs/2018-December/thread.html#00123 v2: - Document "-" instead of "script=-" and use it in the test; and verify this also works on FreeBSD; and verify that it doesn't depend on the particular behaviour of our wrapper script and should work with installed nbdkit too. - Fix handling of zero flags parameter. -

...const char *export_description (struct nbdkit_next_ops *next_ops, + void *nxdata, void *handle); + +This intercepts the plugin C<.export_description> method and can be +used to read or modify the export description that the NBD client +will see. + =head2 C<.can_write> =head2 C<.can_flush> diff --git a/docs/nbdkit-plugin.pod b/docs/nbdkit-plugin.pod index f1476032..62feae47 100644 --- a/docs/nbdkit-plugin.pod +++ b/docs/nbdkit-plugin.pod @@ -825,6 +825,26 @@ to get the size (in bytes) of the block device being exported. The returned size must be E&lt...

Some feature additions to the shell script plugin allow us to test the export flags field reasonably easily. Rich.

In the just-added nbdkit-sh-plugin.pod, you documented that unlike other plugins (where introspection or struct member population) can be used to determine which functionalities are supported, the shell plugin has to implement can_write and friends to return a special status 2 to document not supported, and 3 to indicate false, in part because you can't detect if a pwrite function is present without running it. But since pwrite normally takes additional parameters, could we instead document that nbdkit calling '/path/t...

I think this is my last round of patches for issues I identified with the retry filter. With this in place, it should be safe to interject another filter in between retry and the plugin. Eric Blake (5): retry: Don't call into closed plugin tests: Refactor test-retry-reopen-fail.sh tests: Enhance retry test to cover failed reopen server: Move prepare/finalize/close recursion to

...ssed, so .close is needed */ + HANDLE_CONNECTED = 2, /* Set if .prepare passed, so .finalize is needed */ + HANDLE_FAILED = 4, /* Set if .finalize failed */ +}; + struct b_conn_handle { void *handle; + unsigned char state; /* Bitmask of HANDLE_* values */ + uint64_t exportsize; int can_write; int can_flush; @@ -173,6 +181,7 @@ static inline void reset_b_conn_handle (struct b_conn_handle *h) { h->handle = NULL; + h->state = 0; h->exportsize = -1; h->can_write = -1; h->can_flush = -1; diff --git a/server/backend.c b/server/backend.c index 702c9b96..bdc5bb...

...ckend_reopen (struct backend *b, struct connection *conn, int readonly) > { > + struct b_conn_handle *h = &conn->handles[b->i]; > + > debug ("%s: reopen", b->name); > > b->close (b, conn); > + > + /* This forces .open to recalculate h->can_write, which might have > + * changed since we may have a new readonly value. > + */ > + h->can_write = -1; > + > return backend_open (b, conn, readonly); > } Obviously this hunk should be in the first patch ... Rich. -- Richard Jones, Virtualization Group, Red Hat http:...

...to the plugin more than once per connection on any of the flag determination callbacks. The following script demonstrates the speedup, where we avoid repeated calls into a slow can_fua. Pre-patch: $ cat script case "$1" in get_size) echo 1m;; can_fua) sleep 1; echo native;; can_write | can_zero | pwrite | zero) ;; *) exit 2 ;; esac $ /bin/time -f %e ./nbdkit --filter=blocksize sh script maxlen=4k \ --run 'qemu-io -f raw -c "w -z -f 0 16k" $nbd' wrote 16384/16384 bytes at offset 0 16 KiB, 1 ops; 0:00:05.07 (3.157 KiB/sec and 0.1973 ops/sec) 6.14 there ar...

On Mon, Mar 16, 2020 at 10:36:14PM -0500, Eric Blake wrote: > The documentation for .can_write and friends mentioned merely that the > callback must return a boolean value, or -1 on failure. Historically, > we have treated any non-zero value other than -1 as true, and some > plugins have sloppily relied on this behavior, including the > standalone nbd plugin prior to the use of...

On Fri, Jan 19, 2018 at 10:45:51AM -0600, Eric Blake wrote: > I'm wondering if we're missing documentation here (and/or in the plugins > document) that if .can_write returns true, the plugin must supply a > .pwrite; likewise for .can_trim implying a .trim, and .can_flush > implying a .flush. It's a strange one. For example if a plugin returns .can_pwrite == 1 but doesn't implement .pwrite then the NBD connection will appear to be writable, but a...

And fill out most of the missing bits of the API. Rich.

...&buf, &len); + + if (flags & NBDKIT_FLAG_FAST_ZERO) + flag_append("fast", &comma, &buf, &len); } static void @@ -536,7 +539,7 @@ sh_pwrite (void *handle, const void *buf, uint32_t count, uint64_t offset, /* Common code for handling all boolean methods like can_write etc. */ static int -boolean_method (void *handle, const char *method_name) +boolean_method (void *handle, const char *method_name, int def) { char *h = handle; const char *args[] = { script, method_name, h, NULL }; @@ -546,8 +549,8 @@ boolean_method (void *handle, const char *method_name)...