search for: get_handshake_flag

...defaults.py index 47c1d02..fb961cf 100644 --- a/python/t/110-defaults.py +++ b/python/t/110-defaults.py @@ -22,6 +22,5 @@ assert h.get_export_name() == "" assert h.get_full_info() is False assert h.get_tls() == nbd.TLS_DISABLE assert h.get_request_structured_replies() is True -assert h.get_handshake_flags() == (nbd.HANDSHAKE_FLAG_FIXED_NEWSTYLE | - nbd.HANDSHAKE_FLAG_NO_ZEROES) +assert h.get_handshake_flags() == nbd.HANDSHAKE_FLAG_MASK assert h.get_opt_mode() is False diff --git a/python/t/120-set-non-defaults.py b/python/t/120-set-non-defaults.py index 3963021..3...

Well, the golang changes (patch 1 and 2/3 of v1) were already committed, all that was left was the OCaml changes. I'm a lot happier with how things turned out with an UNKNOWN constructor in the OCaml variants. Eric Blake (3): tests: Enhance coverage of enum/flag range checking ocaml: Support unknown values for Enum/Flags ocaml: Typesafe returns for REnum/RFlags generator/OCaml.ml

...) = let info = NBD.get_full_info nbd in assert (info = false); let tls = NBD.get_tls nbd in - assert (tls = 0); (* XXX Add REnum, to get NBD.TLS.DISABLE? *) + assert (tls = NBD.TLS.DISABLE); let sr = NBD.get_request_structured_replies nbd in assert (sr = true); let flags = NBD.get_handshake_flags nbd in - assert (flags = 3); (* XXX Add RFlags, to get NBD.HANDSHAKE_FLAG list? *) + assert (flags = [ NBD.HANDSHAKE_FLAG.FIXED_NEWSTYLE; + NBD.HANDSHAKE_FLAG.NO_ZEROES ]); let opt = NBD.get_opt_mode nbd in assert (opt = false) diff --git a/ocaml/tests/test_120_set_non_...

...> >This is good because Enum is passed as int. > > > >>[...] > > > >You used unsigned for RFlags, but shouldn't it be this instead? > > > >+ | RFlags (_) -> "uint32_t" > > Hmm. For Flags as input, we take uint32_t, but for > get_handshake_flags which used to return RUint, we returned > unsigned; as written, my patch preserved generated code in its > entirety. Changing it to return uint32_t makes sense, but is a > minor API change. Oh I see, good point. I think although technically it might not be an ABI change, best to leave...

On Sat, Sep 05, 2020 at 08:40:58PM -0500, Eric Blake wrote: > diff --git a/generator/API.mli b/generator/API.mli > index 712e837..e45b5c0 100644 > --- a/generator/API.mli > +++ b/generator/API.mli > @@ -78,6 +78,8 @@ and ret = > | RString (** return a newly allocated string, > caller frees, NULL for error *) > | RUInt

...= 2 + +If the client is forced to use C<NBD_OPT_EXPORT_NAME> instead of +the preferred C<NBD_OPT_GO>, this flag allows the server to send +fewer all-zero padding bytes over the connection. + +=back + +Future NBD extensions may add further flags. +"; + see_also = ["L<nbd_get_handshake_flags(3)>"; + "L<nbd_set_request_structured_replies(3)>"]; + }; + + "get_handshake_flags", { + default_call with + args = []; ret = RUInt; + may_set_error = false; + shortdesc = "see which handshake flags are supported"; + longd...

In v2: - now based on my proposal to add LIBNBD_SHUTDOWN_IMMEDIATE - four flags instead of two: STRICT_FLAGS is new (patch 4), and STRICT_BOUNDS is separate from STRICT_ZERO_SIZE (patch 5) - various refactorings for more shared code and less duplication Eric Blake (5): api: Add xxx_MASK constant for each Flags type generator: Refactor filtering of accepted OFlags api: Add

...) = let info = NBD.get_full_info nbd in assert (info = false); let tls = NBD.get_tls nbd in - assert (tls = 0); (* XXX Add REnum, to get NBD.TLS.DISABLE? *) + assert (tls = NBD.TLS.DISABLE); let sr = NBD.get_request_structured_replies nbd in assert (sr = true); let flags = NBD.get_handshake_flags nbd in - assert (flags = 3); (* XXX Add RFlags, to get NBD.HANDSHAKE_FLAG list? *) + assert (flags = [ NBD.HANDSHAKE_FLAG.NO_ZEROES; + NBD.HANDSHAKE_FLAG.FIXED_NEWSTYLE ]); let opt = NBD.get_opt_mode nbd in assert (opt = false) diff --git a/ocaml/tests/test_120_set_non_...

...Int | REnum (_) -> "int" > > This is good because Enum is passed as int. > >> [...] > > You used unsigned for RFlags, but shouldn't it be this instead? > > + | RFlags (_) -> "uint32_t" Hmm. For Flags as input, we take uint32_t, but for get_handshake_flags which used to return RUint, we returned unsigned; as written, my patch preserved generated code in its entirety. Changing it to return uint32_t makes sense, but is a minor API change. Fortunately, it would not be an ABI change on any of the platforms we compile on. So I can go ahead and re...

On 9/17/19 5:02 AM, Richard W.M. Jones wrote: > This commit adds a new API which can be used from the connected to > state to read back which NBD protocol (eg. oldstyle, newstyle-fixed) > we are using. Somewhat of an overlap with my get_handshake_flags (as newstyle-fixed corresponds to whether handshake_flags includes the FIXED_NEWSTYLE flag), but I don't see the redundancy as an issue. > > It was helpful to add a new state in newstyle negotiation > (%NEWSTYLE.FINISHED) so we can route all successful option > negotiations throu...

...n (see below). I'm also not sure if there is a more efficient way to avoid outputting Val_FOO() converters for types not referenced in REnum/RFlags than my hack of __attribute__((unused)). Hence I'll wait for a review on these. Now, for the future compatibility lesson, promised above. nbd_get_handshake_flags is documented as potentially returning additional bits from a newer libnbd than what the current client code was compiled against. How? Let's say a future NBD protocol addition adds LIBNBD_HANDSHAKE_FLAG_64BIT (as that one might actually be a way that we implement 64-bit requests...). Presuma...

..."get_tls", { default_call with - args = []; ret = RInt; + args = []; ret = REnum (tls_enum); may_set_error = false; shortdesc = "get the TLS request setting"; longdesc = "\ @@ -678,7 +680,7 @@ Future NBD extensions may add further flags. "get_handshake_flags", { default_call with - args = []; ret = RUInt; + args = []; ret = RFlags (handshake_flags); may_set_error = false; shortdesc = "see which handshake flags are supported"; longdesc = "\ @@ -2708,11 +2710,13 @@ let () = failwithf "%s: if may...

For commands with long names and/or short descriptors, you can end up going over 72 characters in the first line of the man page (causing podwrapper to complain). Wrap these lines. --- generator/generator | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/generator/generator b/generator/generator index 7d3f656..ad1cb6b 100755 --- a/generator/generator +++ b/generator/generator

...n L<nbd_connect_socket(3)>. + You can check if the connection is still connecting by calling L<nbd_aio_is_connecting(3)>, or if it has connected to the server and completed the NBD handshake by calling L<nbd_aio_is_ready(3)>, @@ -2758,6 +2791,8 @@ let first_version = [ "get_handshake_flags", (1, 2); "connect_systemd_socket_activation", (1, 2); "aio_connect_systemd_socket_activation", (1, 2); + "connect_socket", (1, 2); + "aio_connect_socket", (1, 2); (* These calls are proposed for a future version of libnbd, but * have n...

...necting by calling L<nbd_aio_is_connecting(3)>, or if it has connected to the server and completed the NBD handshake by calling L<nbd_aio_is_ready(3)>, @@ -2678,6 +2719,8 @@ let first_version = [ "get_protocol", (1, 2); "set_handshake_flags", (1, 2); "get_handshake_flags", (1, 2); + "connect_socket_activation", (1, 2); + "aio_connect_socket_activation", (1, 2); (* These calls are proposed for a future version of libnbd, but * have not been added to any released version so far. @@ -3035,7 +3078,8 @@ end = struct let all_external...

...sts. Also, since both initial state bits and block +status are orthogonal extensions within the NBD protocol, there is no +requirement that a server's advertisement match what could be learned +by checking block status. +" +^ non_blocking_test_call_description; + see_also = [Link "get_handshake_flags"; Link "block_status"]; + }; + "get_size", { default_call with args = []; ret = RInt64; @@ -2292,7 +2317,7 @@ return only one extent per metadata context where that extent does not exceed C<count> bytes; however, libnbd does not validate that the serv...

When LIBNBD_TLS_ALLOW is used we don't have a way to find out if TLS was really negotiated. This adds a flag and a way to read it back. Unfortunately there is no test yet, because LIBNBD_TLS_ALLOW is not tested -- it really should be but requires quite a complicated set of tests because ideally we'd like to find out whether it falls back correctly for all supported servers. --- TODO

Necessary when writing a client that wants to avoid unnecessary EINVAL errors from sending unaligned requests. At some point, we may want to add synchronous convenience API wrappers that do request splitting or read-modify-write to obey server constraints while still appearing to the library client as accepting any possible request. But such a wrapper should only be synchronous and not copied to

...necting by calling L<nbd_aio_is_connecting(3)>, or if it has connected to the server and completed the NBD handshake by calling L<nbd_aio_is_ready(3)>, @@ -2678,6 +2719,8 @@ let first_version = [ "get_protocol", (1, 2); "set_handshake_flags", (1, 2); "get_handshake_flags", (1, 2); + "connect_socket_activation", (1, 2); + "aio_connect_socket_activation", (1, 2); (* These calls are proposed for a future version of libnbd, but * have not been added to any released version so far. @@ -3036,7 +3079,8 @@ end = struct let all_external...

...Link "set_handshake_flags"; Link "set_strict_mode"; Link "get_structured_replies_negotiated"; Link "can_meta_context"; Link "can_df"]; }; @@ -692,7 +705,7 @@ blindly setting a constant value."; "get_handshake_flags", { default_call with - args = []; ret = RFlags (handshake_flags); + args = []; ret = RFlags handshake_flags; may_set_error = false; shortdesc = "see which handshake flags are supported"; longdesc = "\ @@ -706,10 +719,62 @@ protocol defines new handsha...