search for: rsizet

--- generator/API.ml | 1 + generator/API.mli | 1 + generator/C.ml | 14 ++++++++++---- generator/GoLang.ml | 5 +++++ generator/OCaml.ml | 5 +++++ generator/Python.ml | 11 +++++++++-- 6 files changed, 31 insertions(+), 6 deletions(-) diff --git a/generator/API.ml b/generator/API.ml index 0a876c4..b93580d 100644 --- a/generator/API.ml +++ b/generator/API.ml @@ -47,6 +47,7 @@

...rator/API.mli | 1 + > generator/C.ml | 14 ++++++++++---- > generator/GoLang.ml | 5 +++++ > generator/OCaml.ml | 5 +++++ > generator/Python.ml | 11 +++++++++-- > 6 files changed, 31 insertions(+), 6 deletions(-) Looks good, and I'll push a counterpart patch adding RSizeT for returning ssize_t values. Question: Should nbd_opt_list() return RSizeT instead of RInt, as it is returning a count of times the callback was called? The same question would apply to my upcoming nbd_opt_list_meta_context(). Changing a return type from 'int' to 'ssize_t' can re...

On Mon, Sep 28, 2020 at 05:05:16PM -0500, Eric Blake wrote: > + "get_meta_context", { > + default_call with > + args = [ Int "i" ]; ret = RString; Previously we have only used the Int parameter type for things which are plausibly integers (timeouts, signal numbers). This one really ought to be an unsigned or better still a size_t. Would it help if I provided

...signal numbers). This one really > ought to be an unsigned or better still a size_t. Would it help if I > provided a SizeT type for the generator? If we take SizeT as parameter for this one, we probably want to return an ssize_t for get_nr_meta_contexts, rather than RInt. Does the name RSizeT that returns ssize_t make sense? > > But in general yes this patch is obvious so ACK. > > Rich. > -- Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org