similar to: [PATCH nbdkit 0/2] python: Implement pread passing buffer for v2 API.

v3 was here: https://www.redhat.com/archives/libguestfs/2019-November/msg00209.html In v4: - Rebase on top of current master. Includes various fixes and updates required because of Nir's patches that went into master. - Fix api_version() -> API_VERSION in patch 2 noted previously on the mailing list. Rich.

v2 was here: https://www.redhat.com/archives/libguestfs/2019-November/msg00163.html I pushed patch 1 (with spelling fix), patch 4 and patch 5 since those were previously ACKed on the list. Differences in v3: - Add error checking to PyModule_AddIntConstant. - Use API_VERSION constant instead of function. - Add max API version supported to --dump-plugin output. - Print API_VERSION selected by

pread() can return now any object supporting the buffer protocol[1] such as bytearray, bytes, or memoryview. [1] https://docs.python.org/3/c-api/buffer.html --- When the new python tests will be in master we can add some automated tests. For now I tested this manually using the example plugin as is and with these changes:  def pread(h, count, offset):      global disk     return

It's more efficient if we pass the C buffer directly to Python code. In some cases the Python code will be able to write directly into the C buffer using functions like file.readinto and socket.recv_into. This avoids an extra copy. Thanks: Nir Soffer --- plugins/python/example.py | 8 ++++--- plugins/python/nbdkit-python-plugin.pod | 16 +++++-------- plugins/python/python.c

On Sat, Nov 23, 2019 at 3:10 PM Richard W.M. Jones <rjones@redhat.com> wrote: > > On Sat, Nov 23, 2019 at 01:42:15AM +0200, Nir Soffer wrote: > > On Fri, Nov 22, 2019 at 9:55 PM Richard W.M. Jones <rjones@redhat.com> wrote: > > > +def pread(h, count, offset, flags): > > > + assert flags == 0 > > > + return

Passing a memoryview we avoid unneeded copy of the original buffer. On the python side memoryview object can be used for slicing, writing to file, or sending to socket. This may break plugins assuming that the they get a bytearray, but good python code should not care about the type of the buffer, only about the behaviour. Testing with a plugin writing to /dev/null shows 2.7x speedup. Real

On Fri, Nov 22, 2019 at 9:55 PM Richard W.M. Jones <rjones@redhat.com> wrote: > > This tests the Python plugin thoroughly by issuing client commands > through libnbd and checking we get the expected results. > --- > tests/Makefile.am | 13 +-- > tests/test-python-plugin.py | 134 ++++++++++++++++++++++++++++ > tests/test-python.py | 172

On Sun, Nov 24, 2019, 12:42 Richard W.M. Jones <rjones@redhat.com> wrote: > On Sat, Nov 23, 2019 at 06:11:47PM +0200, Nir Soffer wrote: > > On Sat, Nov 23, 2019 at 3:10 PM Richard W.M. Jones <rjones@redhat.com> > wrote: > > > > > > On Sat, Nov 23, 2019 at 01:42:15AM +0200, Nir Soffer wrote: > > > > On Fri, Nov 22, 2019 at 9:55 PM Richard W.M.

v1: https://www.redhat.com/archives/libguestfs/2019-November/msg00153.html v2: - Fix implementation of can_cache. - Add implementation of can_fua. - Add a very thorough test suite which tests every command + flag combination.

Replace python 2 only "buffer" with "memoryview". Falling back to emulated zero would fail with: NameError: name 'buffer' is not defined I did not test the changed code but it was not tested before so it is unlikely to be worse. Detected by pylint. --- v2v/rhv-upload-plugin.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git

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

--- plugins/python/nbdkit-python-plugin.pod | 19 ++++++- plugins/python/python.c | 75 +++++++++++++++++++++++++ tests/test-python-plugin.py | 8 ++- tests/test_python.py | 73 +++++++++++++++++++++++- 4 files changed, 169 insertions(+), 6 deletions(-) diff --git a/plugins/python/nbdkit-python-plugin.pod b/plugins/python/nbdkit-python-plugin.pod

As mentioned in my reviews, I wonder if we should make our python callbacks look a bit more Pythonic by having kwargs added for each new flag that we want to expose. The idea was first floated here: https://www.redhat.com/archives/libguestfs/2018-April/msg00108.html Note that with my proposal, there is no need for a python script to expose a global API_VERSION variable; new flags are added

Hi, the nbdkit python plugin example has suboptimal memory management: - it creates the disk image as a string on init - it casts the string to bytearray on every read - it copies the string before and the string after the written region, then reassembles those pieces together with the written region to a new disk image string This is not a problem as long as the image is small, but in my tests

First out of our four language bindings to add FUA support (for reference, I added 'zero' support for python, perl, and ruby back in 1.1.13, then Rich had to add it for ocaml in 1.1.20). I tested this heavily under python 2, but for now only compile tested under python 3; I plan to do further testing there and make any tweaks if necessary. I wrote patch 5 early on, but then realized I

On 11/22/19 1:53 PM, Richard W.M. Jones wrote: > To avoid breaking existing plugins, Python plugins wishing to use > version 2 of the API must opt in by declaring: > > def api_version(): > return 2 > > (Plugins which do not do this are assumed to want API version 1). Could we also permit the python code to declare a global variable instead of a function? But a

On Sat, Nov 23, 2019 at 06:11:47PM +0200, Nir Soffer wrote: > On Sat, Nov 23, 2019 at 3:10 PM Richard W.M. Jones <rjones@redhat.com> wrote: > > > > On Sat, Nov 23, 2019 at 01:42:15AM +0200, Nir Soffer wrote: > > > On Fri, Nov 22, 2019 at 9:55 PM Richard W.M. Jones <rjones@redhat.com> wrote: > > > > +def pread(h, count, offset, flags): > > >

I'm planning on tweaking the language callbacks to support fua; first up is the python bindings. I want to move from: def zero(h, count, offset, may_trim): to a nicer: def zero(h, count, offset, may_trim=False, fua=False): where the C code passes keywords for the flags (we can add new flags as needed), perhaps by using introspection to learn whether the plugin has a mandatory may_trim

This doesn't yet work. However it does make me more convinced than ever that we really need to sort out persistent buffer lifetimes in the library (similar to what we did for closures). Rich.

Hi, I'm new to nbdkit and am trying to run a simple "single file block device using mmap" setup, using this python script: import mmap def config(key, value): print ("ignored parameter %s=%s" % (key, value)) def open(readonly): print ("open: readonly=%d" % readonly) f = open("disk","r+") m = mmap.mmap(f.fileno(), 0,